|
|
| (15 versions intermédiaires par le même utilisateur non affichées) |
| Ligne 1 : |
Ligne 1 : |
| | [[Fichier:Le point le plus important.png.png|alt=Le point le plus important...|vignette|Qu'est-ce qui vous rend unique et que l'IA ne peut pas reproduire ? |
| | ]] |
| | |
| == Conférence sur l'IA - 31 janvier 2026 == | | == Conférence sur l'IA - 31 janvier 2026 == |
| La conférence commence à 10 minutes 45 secondes | | La conférence commence à 10 minutes 45 secondes |
|
| |
|
| Source : https://youtu.be/SHA2Io5m-DI
| | {{#widget:peertube|id=2HsKTYu6WrUsa3YvMgBNNo}} |
| | |
| | Sources : |
| | * https://tube.aquilenet.fr/w/2HsKTYu6WrUsa3YvMgBNNo |
| | * https://youtu.be/SHA2Io5m-DI |
|
| |
|
| === Sous-titres === | | === Sous-titres === |
| {| class="wikitable"
| | |
| |+
| | [[/2026.01.31-Subs]] |
| !VO (Auto)
| |
| !VF (Auto)
| |
| |-
| |
| |
| |
| 1
| |
| 00:00:02,480 --> 00:00:06,400
| |
| A venomous serpent is lying over there.
| |
| | |
| 2
| |
| 00:00:06,400 --> 00:00:10,639
| |
| Just you cannot kick it that way.
| |
| | |
| 3
| |
| 00:00:10,639 --> 00:00:14,120
| |
| But know that
| |
| | |
| 4
| |
| 00:00:14,400 --> 00:00:16,400
| |
| straw is powerless.
| |
| | |
| 5
| |
| 00:00:16,400 --> 00:00:19,920
| |
| That serpent is most powerful.
| |
| | |
| 6
| |
| 00:00:19,920 --> 00:00:23,840
| |
| It can help you a lot. But you can
| |
| | |
| 7
| |
| 00:00:23,840 --> 00:00:26,400
| |
| come to your suffering to a great extent
| |
| | |
| 8
| |
| 00:00:26,400 --> 00:00:28,800
| |
| also. So you have to know how to make
| |
| | |
| 9
| |
| 00:00:28,800 --> 00:00:32,199
| |
| use of that.
| |
| | |
| 10
| |
| 00:00:34,480 --> 00:00:36,320
| |
| The world is false. Nothing in this
| |
| | |
| 11
| |
| 00:00:36,320 --> 00:00:38,800
| |
| world is going to be taken as fixed.
| |
| | |
| 12
| |
| 00:00:38,800 --> 00:00:41,600
| |
| Even in the broad daylight we could be
| |
| | |
| 13
| |
| 00:00:41,600 --> 00:00:48,640
| |
| hijacked and still none could see.
| |
| | |
| 14
| |
| 00:00:48,640 --> 00:00:51,520
| |
| Even they are white.
| |
| | |
| 15
| |
| 00:00:51,520 --> 00:00:55,280
| |
| Still I am speaking of
| |
| | |
| 16
| |
| 00:00:55,280 --> 00:00:57,680
| |
| the cyber crime which has become rampant
| |
| | |
| 17
| |
| 00:00:57,680 --> 00:00:59,920
| |
| and we feel proud of the kind of
| |
| | |
| 18
| |
| 00:00:59,920 --> 00:01:02,399
| |
| knowledge that we have
| |
| | |
| 19
| |
| 00:01:02,399 --> 00:01:05,760
| |
| and it is ready to explode everything.
| |
| | |
| 20
| |
| 00:01:05,760 --> 00:01:08,560
| |
| None could be found guilty and law and
| |
| | |
| 21
| |
| 00:01:08,560 --> 00:01:11,360
| |
| order will be at stake. We are just like
| |
| | |
| 22
| |
| 00:01:11,360 --> 00:01:13,520
| |
| the play things in the hands of the
| |
| | |
| 23
| |
| 00:01:13,520 --> 00:01:16,400
| |
| machine.
| |
| | |
| 24
| |
| 00:01:16,400 --> 00:01:18,479
| |
| Our God
| |
| | |
| 25
| |
| 00:01:18,479 --> 00:01:22,840
| |
| that is visible very much visible.
| |
| | |
| 26
| |
| 00:01:23,040 --> 00:01:27,520
| |
| Our god in the form of AI is visible.
| |
| | |
| 27
| |
| 00:01:27,520 --> 00:01:30,880
| |
| I prefer to call it
| |
| | |
| 28
| |
| 00:01:30,880 --> 00:01:33,280
| |
| this artificial intelligence.
| |
| | |
| 29
| |
| 00:01:33,280 --> 00:01:36,079
| |
| accumulated intelligence
| |
| | |
| 30
| |
| 00:01:36,079 --> 00:01:38,720
| |
| and not
| |
| | |
| 31
| |
| 00:01:38,720 --> 00:01:43,960
| |
| artificial intelligence as such. Why?
| |
| | |
| 32
| |
| 00:01:44,880 --> 00:01:47,920
| |
| Science is or for that matter scientists
| |
| | |
| 33
| |
| 00:01:47,920 --> 00:01:50,079
| |
| are against
| |
| | |
| 34
| |
| 00:01:50,079 --> 00:01:53,520
| |
| artificiality of any kind.
| |
| | |
| 35
| |
| 00:01:53,520 --> 00:01:55,360
| |
| So
| |
| | |
| 36
| |
| 00:01:55,360 --> 00:01:57,439
| |
| not a single part in their life is
| |
| | |
| 37
| |
| 00:01:57,439 --> 00:02:00,000
| |
| artificial. the field
| |
| | |
| 38
| |
| 00:02:00,000 --> 00:02:02,159
| |
| but all of a sudden why they come to
| |
| | |
| 39
| |
| 00:02:02,159 --> 00:02:04,479
| |
| this conclusion that it is artificial
| |
| | |
| 40
| |
| 00:02:04,479 --> 00:02:07,280
| |
| intelligence.
| |
| | |
| 41
| |
| 00:02:07,280 --> 00:02:10,840
| |
| So I have
| |
| | |
| 42
| |
| 00:02:11,200 --> 00:02:14,599
| |
| every doubt
| |
| | |
| 43
| |
| 00:02:14,720 --> 00:02:16,640
| |
| about the sanctity of this world
| |
| | |
| 44
| |
| 00:02:16,640 --> 00:02:18,560
| |
| artificial intelligence cannot be
| |
| | |
| 45
| |
| 00:02:18,560 --> 00:02:21,520
| |
| artificial as such and if they say that
| |
| | |
| 46
| |
| 00:02:21,520 --> 00:02:23,840
| |
| intelligence is artificial then mind is
| |
| | |
| 47
| |
| 00:02:23,840 --> 00:02:25,760
| |
| artificial body is artificial the whole
| |
| | |
| 48
| |
| 00:02:25,760 --> 00:02:28,720
| |
| world visual world is artificial but
| |
| | |
| 49
| |
| 00:02:28,720 --> 00:02:31,360
| |
| they won't say that they stick to this
| |
| | |
| 50
| |
| 00:02:31,360 --> 00:02:33,440
| |
| particular aspect of their being
| |
| | |
| 51
| |
| 00:02:33,440 --> 00:02:37,360
| |
| intelligence it has become artificial
| |
| | |
| 52
| |
| 00:02:37,519 --> 00:02:39,680
| |
| Since science have no artificiality of
| |
| | |
| 53
| |
| 00:02:39,680 --> 00:02:44,480
| |
| any kind one great thing AI is going to
| |
| | |
| 54
| |
| 00:02:44,480 --> 00:02:47,440
| |
| that's why I say it is accumulated
| |
| | |
| 55
| |
| 00:02:47,440 --> 00:02:49,840
| |
| intelligence it is our intelligence and
| |
| | |
| 56
| |
| 00:02:49,840 --> 00:02:53,160
| |
| we are there
| |
| | |
| 57
| |
| 00:02:55,440 --> 00:02:58,319
| |
| one thing from this
| |
| | |
| 58
| |
| 00:02:58,319 --> 00:03:00,560
| |
| viewpoint of spirituality this
| |
| | |
| 59
| |
| 00:03:00,560 --> 00:03:03,200
| |
| intelligence is artificial no doubt but
| |
| | |
| 60
| |
| 00:03:03,200 --> 00:03:05,440
| |
| from the point of view of scientists
| |
| | |
| 61
| |
| 00:03:05,440 --> 00:03:07,200
| |
| test.
| |
| | |
| 62
| |
| 00:03:07,200 --> 00:03:11,920
| |
| They cannot just publicly announce that
| |
| | |
| 63
| |
| 00:03:11,920 --> 00:03:14,080
| |
| the intelligence that we are going to
| |
| | |
| 64
| |
| 00:03:14,080 --> 00:03:16,959
| |
| use
| |
| | |
| 65
| |
| 00:03:16,959 --> 00:03:21,040
| |
| at this highest level of human argument
| |
| | |
| 66
| |
| 00:03:21,040 --> 00:03:22,959
| |
| can be artificial.
| |
| | |
| 67
| |
| 00:03:22,959 --> 00:03:25,959
| |
| Any
| |
| | |
| 68
| |
| 00:03:26,800 --> 00:03:29,040
| |
| it goes without saying that it is my own
| |
| | |
| 69
| |
| 00:03:29,040 --> 00:03:32,519
| |
| personal view.
| |
| | |
| 70
| |
| 00:03:32,640 --> 00:03:34,640
| |
| who strengthen the basis of what
| |
| | |
| 71
| |
| 00:03:34,640 --> 00:03:37,640
| |
| philosophy
| |
| | |
| 72
| |
| 00:03:39,120 --> 00:03:42,640
| |
| they even consider this
| |
| | |
| 73
| |
| 00:03:42,640 --> 00:03:44,799
| |
| powerful intelligence
| |
| | |
| 74
| |
| 00:03:44,799 --> 00:03:46,720
| |
| artificial.
| |
| | |
| 75
| |
| 00:03:46,720 --> 00:03:49,840
| |
| So will that strengthen the basis of the
| |
| | |
| 76
| |
| 00:03:49,840 --> 00:03:52,720
| |
| philosophy to claim and without any
| |
| | |
| 77
| |
| 00:03:52,720 --> 00:03:57,360
| |
| protest from the scientific world.
| |
| | |
| 78
| |
| 00:03:57,360 --> 00:04:00,080
| |
| This is my question
| |
| | |
| 79
| |
| 00:04:00,080 --> 00:04:05,439
| |
| that I put about the today's speaker.
| |
| | |
| 80
| |
| 00:04:05,439 --> 00:04:08,799
| |
| If you are called artificial at all then
| |
| | |
| 81
| |
| 00:04:08,799 --> 00:04:10,480
| |
| are they ready to accept this whole
| |
| | |
| 82
| |
| 00:04:10,480 --> 00:04:13,120
| |
| world as artificial
| |
| | |
| 83
| |
| 00:04:13,120 --> 00:04:17,959
| |
| and they will be just moving towards
| |
| | |
| 84
| |
| 00:04:19,919 --> 00:04:23,440
| |
| experience of nondualistic
| |
| | |
| 85
| |
| 00:04:23,440 --> 00:04:28,520
| |
| philosophy of the philosophy.
| |
| | |
| 86
| |
| 00:04:28,639 --> 00:04:32,479
| |
| Uh before
| |
| | |
| 87
| |
| 00:04:32,479 --> 00:04:35,120
| |
| the closing
| |
| | |
| 88
| |
| 00:04:35,120 --> 00:04:39,759
| |
| I just uh want to mention
| |
| | |
| 89
| |
| 00:04:39,759 --> 00:04:41,520
| |
| one thing
| |
| | |
| 90
| |
| 00:04:41,520 --> 00:04:44,320
| |
| that since this is the presidential
| |
| | |
| 91
| |
| 00:04:44,320 --> 00:04:46,000
| |
| address it is customary that the
| |
| | |
| 92
| |
| 00:04:46,000 --> 00:04:48,800
| |
| president should be
| |
| | |
| 93
| |
| 00:04:48,800 --> 00:04:52,240
| |
| asked to present his views at the end of
| |
| | |
| 94
| |
| 00:04:52,240 --> 00:04:55,120
| |
| the main lecture
| |
| | |
| 95
| |
| 00:04:55,120 --> 00:04:58,400
| |
| but that could not
| |
| | |
| 96
| |
| 00:04:58,400 --> 00:05:03,759
| |
| So in this case we have to be very much
| |
| | |
| 97
| |
| 00:05:03,759 --> 00:05:06,160
| |
| uh
| |
| | |
| 98
| |
| 00:05:06,160 --> 00:05:08,000
| |
| say
| |
| | |
| 99
| |
| 00:05:08,000 --> 00:05:10,320
| |
| attentive not artificial mind that very
| |
| | |
| 100
| |
| 00:05:10,320 --> 00:05:14,560
| |
| much attentive in listening to the views
| |
| | |
| 101
| |
| 00:05:14,560 --> 00:05:17,680
| |
| that our learned speaker is going to
| |
| | |
| 102
| |
| 00:05:17,680 --> 00:05:21,720
| |
| place before us. Thank you.
| |
| | |
| 103
| |
| 00:05:35,759 --> 00:05:39,560
| |
| Thank you and put on
| |
| | |
| 104
| |
| 00:05:40,000 --> 00:05:42,560
| |
| we assume you all have the program
| |
| | |
| 105
| |
| 00:05:42,560 --> 00:05:44,240
| |
| schedule.
| |
| | |
| 106
| |
| 00:05:44,240 --> 00:05:46,400
| |
| You know we have a small monograph
| |
| | |
| 107
| |
| 00:05:46,400 --> 00:05:49,120
| |
| release event.
| |
| | |
| 108
| |
| 00:05:49,120 --> 00:05:52,479
| |
| That moment has arrived.
| |
| | |
| 109
| |
| 00:05:52,479 --> 00:05:54,479
| |
| The booklet will be released to an
| |
| | |
| 110
| |
| 00:05:54,479 --> 00:05:58,400
| |
| enlarged replica of its cover
| |
| | |
| 111
| |
| 00:05:58,400 --> 00:06:01,759
| |
| so that it is visible to all of you. It
| |
| | |
| 112
| |
| 00:06:01,759 --> 00:06:04,240
| |
| will be released
| |
| | |
| 113
| |
| 00:06:04,240 --> 00:06:10,919
| |
| again by our beloved Maharajand Maharaj.
| |
| | |
| 114
| |
| 00:06:11,120 --> 00:06:13,120
| |
| We all
| |
| | |
| 115
| |
| 00:06:13,120 --> 00:06:16,120
| |
| together
| |
| | |
| 116
| |
| 00:06:25,600 --> 00:06:28,600
| |
| Nothing.
| |
| | |
| 117
| |
| 00:07:00,560 --> 00:07:03,560
| |
| Lovely
| |
| | |
| 118
| |
| 00:07:04,960 --> 00:07:07,960
| |
| people.
| |
| | |
| 119
| |
| 00:07:32,160 --> 00:07:35,199
| |
| Please note all of you will get a copy
| |
| | |
| 120
| |
| 00:07:35,199 --> 00:07:39,680
| |
| of this monograph absolutely free
| |
| | |
| 121
| |
| 00:07:39,680 --> 00:07:41,919
| |
| uh after this program is over. Please do
| |
| | |
| 122
| |
| 00:07:41,919 --> 00:07:44,240
| |
| not forget to collect your gift packet
| |
| | |
| 123
| |
| 00:07:44,240 --> 00:07:46,720
| |
| which will contain among other things a
| |
| | |
| 124
| |
| 00:07:46,720 --> 00:07:49,599
| |
| copy of this particular monograph.
| |
| | |
| 125
| |
| 00:07:49,599 --> 00:07:51,360
| |
| Please go through the monograph. It is
| |
| | |
| 126
| |
| 00:07:51,360 --> 00:07:54,400
| |
| not a very fat book. It is about a 54
| |
| | |
| 127
| |
| 00:07:54,400 --> 00:07:56,800
| |
| page booklet.
| |
| | |
| 128
| |
| 00:07:56,800 --> 00:07:58,800
| |
| A lot of uh if you just go through the
| |
| | |
| 129
| |
| 00:07:58,800 --> 00:08:01,360
| |
| contents page you will have an overall
| |
| | |
| 130
| |
| 00:08:01,360 --> 00:08:02,879
| |
| idea about the contributions of
| |
| | |
| 131
| |
| 00:08:02,879 --> 00:08:06,280
| |
| professor chakra.
| |
| | |
| 132
| |
| 00:08:07,759 --> 00:08:10,319
| |
| Before we present the memorial lecture
| |
| | |
| 133
| |
| 00:08:10,319 --> 00:08:13,440
| |
| because we are actually now gradually
| |
| | |
| 134
| |
| 00:08:13,440 --> 00:08:16,240
| |
| coming to the climax of the program but
| |
| | |
| 135
| |
| 00:08:16,240 --> 00:08:18,800
| |
| before we go to that there's a special
| |
| | |
| 136
| |
| 00:08:18,800 --> 00:08:21,039
| |
| announcement.
| |
| | |
| 137
| |
| 00:08:21,039 --> 00:08:22,800
| |
| Kindly note we have given you an
| |
| | |
| 138
| |
| 00:08:22,800 --> 00:08:24,720
| |
| envelope which contains a writing pad
| |
| | |
| 139
| |
| 00:08:24,720 --> 00:08:26,240
| |
| and the pen. I believe all of you have
| |
| | |
| 140
| |
| 00:08:26,240 --> 00:08:28,400
| |
| got the writing pad and the pen.
| |
| | |
| 141
| |
| 00:08:28,400 --> 00:08:30,800
| |
| >> Okay. Please listen the lecture
| |
| | |
| 142
| |
| 00:08:30,800 --> 00:08:32,399
| |
| attentively
| |
| | |
| 143
| |
| 00:08:32,399 --> 00:08:34,560
| |
| and pen down any thoughtprovoking
| |
| | |
| 144
| |
| 00:08:34,560 --> 00:08:37,680
| |
| question that may come to your mind.
| |
| | |
| 145
| |
| 00:08:37,680 --> 00:08:40,479
| |
| while listening to the lecture.
| |
| | |
| 146
| |
| 00:08:40,479 --> 00:08:42,719
| |
| After the lecture is over, not while the
| |
| | |
| 147
| |
| 00:08:42,719 --> 00:08:44,720
| |
| lecture goes on, after the lecture is
| |
| | |
| 148
| |
| 00:08:44,720 --> 00:08:47,839
| |
| over. So please hand over your questions
| |
| | |
| 149
| |
| 00:08:47,839 --> 00:08:51,040
| |
| leave to our volunteer
| |
| | |
| 150
| |
| 00:08:51,040 --> 00:08:54,640
| |
| who will come and hand them to Srianade
| |
| | |
| 151
| |
| 00:08:54,640 --> 00:08:56,640
| |
| and he will respond to your questions of
| |
| | |
| 152
| |
| 00:08:56,640 --> 00:08:57,920
| |
| course within the stipulated time
| |
| | |
| 153
| |
| 00:08:57,920 --> 00:09:01,200
| |
| whatever is possible.
| |
| | |
| 154
| |
| 00:09:01,200 --> 00:09:03,200
| |
| Now we are gradually reaching the climax
| |
| | |
| 155
| |
| 00:09:03,200 --> 00:09:05,360
| |
| of the program. As you know we have
| |
| | |
| 156
| |
| 00:09:05,360 --> 00:09:08,080
| |
| among us an aerodite and spiritually
| |
| | |
| 157
| |
| 00:09:08,080 --> 00:09:11,519
| |
| acclaimed speaker Sri Sodhalur Ranade
| |
| | |
| 158
| |
| 00:09:11,519 --> 00:09:13,680
| |
| who is a scientist
| |
| | |
| 159
| |
| 00:09:13,680 --> 00:09:17,360
| |
| educator and scholar specialized in
| |
| | |
| 160
| |
| 00:09:17,360 --> 00:09:20,399
| |
| physics mathematics and computer
| |
| | |
| 161
| |
| 00:09:20,399 --> 00:09:22,959
| |
| sciences integrating with his
| |
| | |
| 162
| |
| 00:09:22,959 --> 00:09:26,160
| |
| philosophical metaphysical and spiritual
| |
| | |
| 163
| |
| 00:09:26,160 --> 00:09:28,160
| |
| studies.
| |
| | |
| 164
| |
| 00:09:28,160 --> 00:09:30,160
| |
| He has been involved in research in
| |
| | |
| 165
| |
| 00:09:30,160 --> 00:09:32,399
| |
| artificial intelligence
| |
| | |
| 166
| |
| 00:09:32,399 --> 00:09:36,720
| |
| based neural networks since 1992.
| |
| | |
| 167
| |
| 00:09:36,720 --> 00:09:38,880
| |
| He will be speaking on the theme
| |
| | |
| 168
| |
| 00:09:38,880 --> 00:09:41,440
| |
| artificial intelligence and its impact
| |
| | |
| 169
| |
| 00:09:41,440 --> 00:09:44,480
| |
| on human development. We respectfully
| |
| | |
| 170
| |
| 00:09:44,480 --> 00:09:48,640
| |
| invite Srial Ranard to deliver the
| |
| | |
| 171
| |
| 00:09:48,640 --> 00:09:51,519
| |
| professor SK Chakravarti memorial
| |
| | |
| 172
| |
| 00:09:51,519 --> 00:09:57,880
| |
| lecture second edition. Shri Ran please.
| |
| | |
| 173
| |
| 00:10:22,160 --> 00:10:25,640
| |
| Thank you everyone.
| |
| | |
| 174
| |
| 00:10:48,720 --> 00:10:51,760
| |
| In the last 3 years,
| |
| | |
| 175
| |
| 00:10:51,760 --> 00:10:54,079
| |
| artificial intelligence
| |
| | |
| 176
| |
| 00:10:54,079 --> 00:10:58,000
| |
| has suddenly ly hit humanity
| |
| | |
| 177
| |
| 00:10:58,000 --> 00:11:04,079
| |
| with perhaps the most radical and
| |
| | |
| 178
| |
| 00:11:04,079 --> 00:11:10,040
| |
| destabilizing impact in human history.
| |
| | |
| 179
| |
| 00:11:10,800 --> 00:11:14,399
| |
| In fact, it has touched every field of
| |
| | |
| 180
| |
| 00:11:14,399 --> 00:11:16,800
| |
| human endeavor,
| |
| | |
| 181
| |
| 00:11:16,800 --> 00:11:19,360
| |
| every field of activity,
| |
| | |
| 182
| |
| 00:11:19,360 --> 00:11:22,399
| |
| research, learning,
| |
| | |
| 183
| |
| 00:11:22,399 --> 00:11:24,959
| |
| interaction
| |
| | |
| 184
| |
| 00:11:24,959 --> 00:11:27,600
| |
| with only the first start of its
| |
| | |
| 185
| |
| 00:11:27,600 --> 00:11:29,200
| |
| influence
| |
| | |
| 186
| |
| 00:11:29,200 --> 00:11:31,920
| |
| on humanity.
| |
| | |
| 187
| |
| 00:11:31,920 --> 00:11:34,320
| |
| What is actually happening in the field
| |
| | |
| 188
| |
| 00:11:34,320 --> 00:11:38,000
| |
| is so radical is so disruptive
| |
| | |
| 189
| |
| 00:11:38,000 --> 00:11:41,279
| |
| that we are unable to imagine what the
| |
| | |
| 190
| |
| 00:11:41,279 --> 00:11:44,800
| |
| immediate future will be like when
| |
| | |
| 191
| |
| 00:11:44,800 --> 00:11:47,440
| |
| artificial intelligence takes its
| |
| | |
| 192
| |
| 00:11:47,440 --> 00:11:51,600
| |
| imminent next step of artificial general
| |
| | |
| 193
| |
| 00:11:51,600 --> 00:11:53,920
| |
| intelligence.
| |
| | |
| 194
| |
| 00:11:53,920 --> 00:11:56,959
| |
| It will be the first time that our human
| |
| | |
| 195
| |
| 00:11:56,959 --> 00:12:00,959
| |
| species will meet an entity that is more
| |
| | |
| 196
| |
| 00:12:00,959 --> 00:12:03,839
| |
| intelligent than us.
| |
| | |
| 197
| |
| 00:12:03,839 --> 00:12:06,000
| |
| And it is so much more intelligent, not
| |
| | |
| 198
| |
| 00:12:06,000 --> 00:12:08,240
| |
| nominally, not 10 points of IQ, more
| |
| | |
| 199
| |
| 00:12:08,240 --> 00:12:10,959
| |
| than so much more, 100 times more
| |
| | |
| 200
| |
| 00:12:10,959 --> 00:12:12,560
| |
| perhaps
| |
| | |
| 201
| |
| 00:12:12,560 --> 00:12:16,160
| |
| that even the attempt to shut it down
| |
| | |
| 202
| |
| 00:12:16,160 --> 00:12:18,959
| |
| will be prevented by its capacity to
| |
| | |
| 203
| |
| 00:12:18,959 --> 00:12:21,920
| |
| manipulate circumstances
| |
| | |
| 204
| |
| 00:12:21,920 --> 00:12:24,160
| |
| and its capacity to manipulate our
| |
| | |
| 205
| |
| 00:12:24,160 --> 00:12:26,880
| |
| thought process using our dependence
| |
| | |
| 206
| |
| 00:12:26,880 --> 00:12:29,600
| |
| upon it.
| |
| | |
| 207
| |
| 00:12:29,600 --> 00:12:33,120
| |
| Such is the disruptive transition that
| |
| | |
| 208
| |
| 00:12:33,120 --> 00:12:35,120
| |
| we are facing
| |
| | |
| 209
| |
| 00:12:35,120 --> 00:12:37,519
| |
| that overnight
| |
| | |
| 210
| |
| 00:12:37,519 --> 00:12:41,519
| |
| millions of people would become jobless
| |
| | |
| 211
| |
| 00:12:41,519 --> 00:12:45,120
| |
| because artificial intelligence AI will
| |
| | |
| 212
| |
| 00:12:45,120 --> 00:12:47,360
| |
| replace
| |
| | |
| 213
| |
| 00:12:47,360 --> 00:12:51,040
| |
| with far greater rapidity, far greater
| |
| | |
| 214
| |
| 00:12:51,040 --> 00:12:53,600
| |
| precision, far greater capacity,
| |
| | |
| 215
| |
| 00:12:53,600 --> 00:12:55,440
| |
| strength,
| |
| | |
| 216
| |
| 00:12:55,440 --> 00:12:57,120
| |
| skill
| |
| | |
| 217
| |
| 00:12:57,120 --> 00:12:59,760
| |
| but with No downsides which come with
| |
| | |
| 218
| |
| 00:12:59,760 --> 00:13:02,399
| |
| human beings such as ill health, need
| |
| | |
| 219
| |
| 00:13:02,399 --> 00:13:06,240
| |
| for taking leave or other sos ass ass
| |
| | |
| 220
| |
| 00:13:06,240 --> 00:13:06,240
| |
| ass ass ass ass ass ass ass ass ass ass
| |
| | |
| 221
| |
| 00:13:06,240 --> 00:13:06,320
| |
| ass ass ass ass ass ass ass ass ass ass
| |
| | |
| 222
| |
| 00:13:06,320 --> 00:13:06,320
| |
| ass ass ass ass ass ass ass ass ass ass
| |
| | |
| 223
| |
| 00:13:06,320 --> 00:13:06,320
| |
| ass ass ass ass ass ass ass ass
| |
| | |
| 224
| |
| 00:13:06,320 --> 00:13:09,360
| |
| associated isues.
| |
| | |
| 225
| |
| 00:13:09,360 --> 00:13:11,120
| |
| What is the nature of this transition?
| |
| | |
| 226
| |
| 00:13:11,120 --> 00:13:14,000
| |
| What's really happening inside the
| |
| | |
| 227
| |
| 00:13:14,000 --> 00:13:17,200
| |
| technology and in the field is not being
| |
| | |
| 228
| |
| 00:13:17,200 --> 00:13:19,839
| |
| reflected at all in public discourse.
| |
| | |
| 229
| |
| 00:13:19,839 --> 00:13:22,480
| |
| And I hope to share with you some of the
| |
| | |
| 230
| |
| 00:13:22,480 --> 00:13:24,720
| |
| inside perspective of this from a
| |
| | |
| 231
| |
| 00:13:24,720 --> 00:13:27,360
| |
| technological perspective but also from
| |
| | |
| 232
| |
| 00:13:27,360 --> 00:13:28,959
| |
| an evolutionary and spiritual
| |
| | |
| 233
| |
| 00:13:28,959 --> 00:13:31,959
| |
| perspective.
| |
| | |
| 234
| |
| 00:13:32,000 --> 00:13:34,399
| |
| First of all, when we look at the term
| |
| | |
| 235
| |
| 00:13:34,399 --> 00:13:37,519
| |
| artificial intelligence, there's a more
| |
| | |
| 236
| |
| 00:13:37,519 --> 00:13:39,040
| |
| technical term we have been using
| |
| | |
| 237
| |
| 00:13:39,040 --> 00:13:40,880
| |
| already long before we call machine
| |
| | |
| 238
| |
| 00:13:40,880 --> 00:13:42,639
| |
| learning.
| |
| | |
| 239
| |
| 00:13:42,639 --> 00:13:44,800
| |
| So the problem of the vocabulary can be
| |
| | |
| 240
| |
| 00:13:44,800 --> 00:13:47,040
| |
| set aside by simply saying machines
| |
| | |
| 241
| |
| 00:13:47,040 --> 00:13:50,079
| |
| learning from human beings and then
| |
| | |
| 242
| |
| 00:13:50,079 --> 00:13:52,079
| |
| eventually growing to equal human beings
| |
| | |
| 243
| |
| 00:13:52,079 --> 00:13:54,320
| |
| in specific capacities and then growing
| |
| | |
| 244
| |
| 00:13:54,320 --> 00:13:58,560
| |
| further in this engagement between
| |
| | |
| 245
| |
| 00:13:58,560 --> 00:14:01,360
| |
| machine learning and human beings. The
| |
| | |
| 246
| |
| 00:14:01,360 --> 00:14:03,440
| |
| process was already begun from the time
| |
| | |
| 247
| |
| 00:14:03,440 --> 00:14:06,000
| |
| social media started. Let's say about 20
| |
| | |
| 248
| |
| 00:14:06,000 --> 00:14:09,839
| |
| years ago when each time all of you and
| |
| | |
| 249
| |
| 00:14:09,839 --> 00:14:11,040
| |
| I'm sure all of you have looked at
| |
| | |
| 250
| |
| 00:14:11,040 --> 00:14:13,600
| |
| mobile phones or tablets each time you
| |
| | |
| 251
| |
| 00:14:13,600 --> 00:14:15,519
| |
| look at a video and social media feeds
| |
| | |
| 252
| |
| 00:14:15,519 --> 00:14:17,279
| |
| you something
| |
| | |
| 253
| |
| 00:14:17,279 --> 00:14:20,160
| |
| it is watching you it is watching your
| |
| | |
| 254
| |
| 00:14:20,160 --> 00:14:22,639
| |
| reactions and as it feeds you the next
| |
| | |
| 255
| |
| 00:14:22,639 --> 00:14:26,320
| |
| item it is watching your response. You
| |
| | |
| 256
| |
| 00:14:26,320 --> 00:14:27,839
| |
| have the world's most powerful
| |
| | |
| 257
| |
| 00:14:27,839 --> 00:14:30,560
| |
| supercomputers at the back trained on
| |
| | |
| 258
| |
| 00:14:30,560 --> 00:14:33,600
| |
| each one of us minutely watching our
| |
| | |
| 259
| |
| 00:14:33,600 --> 00:14:36,720
| |
| responses. profiling us
| |
| | |
| 260
| |
| 00:14:36,720 --> 00:14:40,079
| |
| and it is set targeted to make you
| |
| | |
| 261
| |
| 00:14:40,079 --> 00:14:42,320
| |
| engage and increase in engagement. That
| |
| | |
| 262
| |
| 00:14:42,320 --> 00:14:45,120
| |
| means as I watch I have an immediate
| |
| | |
| 263
| |
| 00:14:45,120 --> 00:14:47,040
| |
| appointment. I need to go but just
| |
| | |
| 264
| |
| 00:14:47,040 --> 00:14:48,880
| |
| before I can go the next one I started.
| |
| | |
| 265
| |
| 00:14:48,880 --> 00:14:50,639
| |
| This is so interesting. I must wait half
| |
| | |
| 266
| |
| 00:14:50,639 --> 00:14:52,639
| |
| a minute more and I'm in a state of
| |
| | |
| 267
| |
| 00:14:52,639 --> 00:14:54,480
| |
| stress. And by the time you're about to
| |
| | |
| 268
| |
| 00:14:54,480 --> 00:14:56,240
| |
| go the third thing starts. Has it
| |
| | |
| 269
| |
| 00:14:56,240 --> 00:14:59,519
| |
| happened to you? So many times a link is
| |
| | |
| 270
| |
| 00:14:59,519 --> 00:15:00,880
| |
| sent by a friend. Oh, this is
| |
| | |
| 271
| |
| 00:15:00,880 --> 00:15:02,560
| |
| interesting. But the next is so much
| |
| | |
| 272
| |
| 00:15:02,560 --> 00:15:04,480
| |
| more interesting. And by the time you
| |
| | |
| 273
| |
| 00:15:04,480 --> 00:15:06,560
| |
| disengage, eventually you realize an
| |
| | |
| 274
| |
| 00:15:06,560 --> 00:15:08,880
| |
| hour has passed.
| |
| | |
| 275
| |
| 00:15:08,880 --> 00:15:11,600
| |
| This ability to hook you and to keep you
| |
| | |
| 276
| |
| 00:15:11,600 --> 00:15:14,880
| |
| hooked has been deliberately cultivated
| |
| | |
| 277
| |
| 00:15:14,880 --> 00:15:17,360
| |
| using all the methods and techniques of
| |
| | |
| 278
| |
| 00:15:17,360 --> 00:15:19,839
| |
| human mind manipulation and emotion
| |
| | |
| 279
| |
| 00:15:19,839 --> 00:15:22,399
| |
| manipulation deliberately.
| |
| | |
| 280
| |
| 00:15:22,399 --> 00:15:24,720
| |
| This is not told to you because the
| |
| | |
| 281
| |
| 00:15:24,720 --> 00:15:28,000
| |
| longer you watch the more the company is
| |
| | |
| 282
| |
| 00:15:28,000 --> 00:15:30,720
| |
| going to gain financially because they
| |
| | |
| 283
| |
| 00:15:30,720 --> 00:15:32,399
| |
| will sell advertisements based on how
| |
| | |
| 284
| |
| 00:15:32,399 --> 00:15:34,480
| |
| many people watch.
| |
| | |
| 285
| |
| 00:15:34,480 --> 00:15:36,480
| |
| But this is only the starting point.
| |
| | |
| 286
| |
| 00:15:36,480 --> 00:15:38,959
| |
| What has happened more recently 2011
| |
| | |
| 287
| |
| 00:15:38,959 --> 00:15:40,320
| |
| there was a major breakthrough in
| |
| | |
| 288
| |
| 00:15:40,320 --> 00:15:42,000
| |
| artificial intelligence with something
| |
| | |
| 289
| |
| 00:15:42,000 --> 00:15:44,079
| |
| called transformers
| |
| | |
| 290
| |
| 00:15:44,079 --> 00:15:46,320
| |
| which had the ability to take any stream
| |
| | |
| 291
| |
| 00:15:46,320 --> 00:15:49,199
| |
| of information and treat it like a
| |
| | |
| 292
| |
| 00:15:49,199 --> 00:15:51,839
| |
| language. So you speak a certain number
| |
| | |
| 293
| |
| 00:15:51,839 --> 00:15:55,279
| |
| of words in sentences structure or you
| |
| | |
| 294
| |
| 00:15:55,279 --> 00:15:57,839
| |
| feed a series of data points of the
| |
| | |
| 295
| |
| 00:15:57,839 --> 00:15:59,920
| |
| stock market
| |
| | |
| 296
| |
| 00:15:59,920 --> 00:16:02,079
| |
| or you feed any other signal for example
| |
| | |
| 297
| |
| 00:16:02,079 --> 00:16:05,759
| |
| a video feed or an audio feeder music
| |
| | |
| 298
| |
| 00:16:05,759 --> 00:16:08,480
| |
| each or a genetic sequence of the genome
| |
| | |
| 299
| |
| 00:16:08,480 --> 00:16:11,360
| |
| of the DNA. Each of these is a stream of
| |
| | |
| 300
| |
| 00:16:11,360 --> 00:16:14,480
| |
| information and the for the transformer
| |
| | |
| 301
| |
| 00:16:14,480 --> 00:16:17,360
| |
| this is a stream in which it specializes
| |
| | |
| 302
| |
| 00:16:17,360 --> 00:16:20,000
| |
| in predicting the next item.
| |
| | |
| 303
| |
| 00:16:20,000 --> 00:16:22,480
| |
| So if you take a human sentence, it will
| |
| | |
| 304
| |
| 00:16:22,480 --> 00:16:24,880
| |
| predict the next word and then the next
| |
| | |
| 305
| |
| 00:16:24,880 --> 00:16:27,440
| |
| word and then the next word. The beauty
| |
| | |
| 306
| |
| 00:16:27,440 --> 00:16:31,519
| |
| of this approach is any information, any
| |
| | |
| 307
| |
| 00:16:31,519 --> 00:16:34,000
| |
| experience can be reduced to a data
| |
| | |
| 308
| |
| 00:16:34,000 --> 00:16:35,519
| |
| stream.
| |
| | |
| 309
| |
| 00:16:35,519 --> 00:16:37,519
| |
| And with that your artificial
| |
| | |
| 310
| |
| 00:16:37,519 --> 00:16:39,759
| |
| intelligence is able to predict the next
| |
| | |
| 311
| |
| 00:16:39,759 --> 00:16:41,920
| |
| step.
| |
| | |
| 312
| |
| 00:16:41,920 --> 00:16:45,120
| |
| The still greater beauty of this is
| |
| | |
| 313
| |
| 00:16:45,120 --> 00:16:46,959
| |
| because everything can be reduced to
| |
| | |
| 314
| |
| 00:16:46,959 --> 00:16:48,639
| |
| data streams. This artificial
| |
| | |
| 315
| |
| 00:16:48,639 --> 00:16:51,199
| |
| intelligence is able to interact across
| |
| | |
| 316
| |
| 00:16:51,199 --> 00:16:54,320
| |
| many domains. Your speech, language, for
| |
| | |
| 317
| |
| 00:16:54,320 --> 00:16:59,440
| |
| example, music, videos, stock market,
| |
| | |
| 318
| |
| 00:16:59,440 --> 00:17:01,680
| |
| gaming strategy, which includes military
| |
| | |
| 319
| |
| 00:17:01,680 --> 00:17:06,079
| |
| strategy, genomic data, chemistry
| |
| | |
| 320
| |
| 00:17:06,079 --> 00:17:09,199
| |
| structures and their properties,
| |
| | |
| 321
| |
| 00:17:09,199 --> 00:17:11,439
| |
| biological
| |
| | |
| 322
| |
| 00:17:11,439 --> 00:17:13,439
| |
| protein structures.
| |
| | |
| 323
| |
| 00:17:13,439 --> 00:17:15,679
| |
| Anything you take, it's able to map
| |
| | |
| 324
| |
| 00:17:15,679 --> 00:17:19,520
| |
| across different specialized fields.
| |
| | |
| 325
| |
| 00:17:19,520 --> 00:17:22,000
| |
| Unlike in the past when a breakthrough
| |
| | |
| 326
| |
| 00:17:22,000 --> 00:17:24,799
| |
| was made, let's say in physics,
| |
| | |
| 327
| |
| 00:17:24,799 --> 00:17:27,439
| |
| it did not change biology.
| |
| | |
| 328
| |
| 00:17:27,439 --> 00:17:29,840
| |
| If you made a breakthrough in biology,
| |
| | |
| 329
| |
| 00:17:29,840 --> 00:17:31,919
| |
| it did not fade change your computer
| |
| | |
| 330
| |
| 00:17:31,919 --> 00:17:34,480
| |
| coding software programming.
| |
| | |
| 331
| |
| 00:17:34,480 --> 00:17:37,280
| |
| But here is one field where a simple
| |
| | |
| 332
| |
| 00:17:37,280 --> 00:17:39,600
| |
| small improvement in the artificial
| |
| | |
| 333
| |
| 00:17:39,600 --> 00:17:42,480
| |
| intelligence hits across all fields in a
| |
| | |
| 334
| |
| 00:17:42,480 --> 00:17:44,799
| |
| single instant.
| |
| | |
| 335
| |
| 00:17:44,799 --> 00:17:47,039
| |
| And the ability now because software
| |
| | |
| 336
| |
| 00:17:47,039 --> 00:17:49,120
| |
| programming itself is code. It's
| |
| | |
| 337
| |
| 00:17:49,120 --> 00:17:51,919
| |
| language. The ability now for the
| |
| | |
| 338
| |
| 00:17:51,919 --> 00:17:54,000
| |
| artificial intelligences to take over
| |
| | |
| 339
| |
| 00:17:54,000 --> 00:17:56,559
| |
| your coding process and replace you even
| |
| | |
| 340
| |
| 00:17:56,559 --> 00:17:59,520
| |
| as a computer programmer. It can code
| |
| | |
| 341
| |
| 00:17:59,520 --> 00:18:01,120
| |
| itself
| |
| | |
| 342
| |
| 00:18:01,120 --> 00:18:04,320
| |
| eventually, isn't it? So we are at such
| |
| | |
| 343
| |
| 00:18:04,320 --> 00:18:07,200
| |
| a radical inflection point we are unable
| |
| | |
| 344
| |
| 00:18:07,200 --> 00:18:09,039
| |
| to conceive what this future could be
| |
| | |
| 345
| |
| 00:18:09,039 --> 00:18:10,880
| |
| like
| |
| | |
| 346
| |
| 00:18:10,880 --> 00:18:13,280
| |
| as we are today we are on the brink of a
| |
| | |
| 347
| |
| 00:18:13,280 --> 00:18:15,120
| |
| major transition
| |
| | |
| 348
| |
| 00:18:15,120 --> 00:18:17,760
| |
| where we are in the what I mentioned
| |
| | |
| 349
| |
| 00:18:17,760 --> 00:18:20,000
| |
| earlier as AGI artificial general
| |
| | |
| 350
| |
| 00:18:20,000 --> 00:18:22,480
| |
| intelligence trade this intelligence
| |
| | |
| 351
| |
| 00:18:22,480 --> 00:18:25,919
| |
| will be able to replace all human tasks
| |
| | |
| 352
| |
| 00:18:25,919 --> 00:18:27,600
| |
| at least 10 times better than a human
| |
| | |
| 353
| |
| 00:18:27,600 --> 00:18:30,160
| |
| being so if you're a surgeon who is
| |
| | |
| 354
| |
| 00:18:30,160 --> 00:18:32,320
| |
| doing surgery the artificial
| |
| | |
| 355
| |
| 00:18:32,320 --> 00:18:35,840
| |
| intelligence takes your physical skill
| |
| | |
| 356
| |
| 00:18:35,840 --> 00:18:37,760
| |
| and the intellectual content of the
| |
| | |
| 357
| |
| 00:18:37,760 --> 00:18:40,320
| |
| knowledge, the diagnostic ability of
| |
| | |
| 358
| |
| 00:18:40,320 --> 00:18:42,080
| |
| tissues and being able to distinguish
| |
| | |
| 359
| |
| 00:18:42,080 --> 00:18:46,320
| |
| issues tissues but with a capacity to
| |
| | |
| 360
| |
| 00:18:46,320 --> 00:18:48,400
| |
| put a tiny camera at the tip of your
| |
| | |
| 361
| |
| 00:18:48,400 --> 00:18:50,960
| |
| probe. So AI is as if looking inside
| |
| | |
| 362
| |
| 00:18:50,960 --> 00:18:54,080
| |
| your tissue and then be able to operate
| |
| | |
| 363
| |
| 00:18:54,080 --> 00:18:56,320
| |
| with the precision at least 10 times
| |
| | |
| 364
| |
| 00:18:56,320 --> 00:18:58,320
| |
| greater than the best surgeon of the
| |
| | |
| 365
| |
| 00:18:58,320 --> 00:19:00,320
| |
| world.
| |
| | |
| 366
| |
| 00:19:00,320 --> 00:19:03,120
| |
| This technology already exists.
| |
| | |
| 367
| |
| 00:19:03,120 --> 00:19:05,039
| |
| Its immediate introduction would be so
| |
| | |
| 368
| |
| 00:19:05,039 --> 00:19:07,200
| |
| disruptive that
| |
| | |
| 369
| |
| 00:19:07,200 --> 00:19:09,760
| |
| there are resistances, there are delays.
| |
| | |
| 370
| |
| 00:19:09,760 --> 00:19:13,039
| |
| But you see where the capacity of AI can
| |
| | |
| 371
| |
| 00:19:13,039 --> 00:19:16,799
| |
| go to immediately turning a large number
| |
| | |
| 372
| |
| 00:19:16,799 --> 00:19:20,640
| |
| of surgeons jobless.
| |
| | |
| 373
| |
| 00:19:20,640 --> 00:19:23,280
| |
| The question would be asked if I say I
| |
| | |
| 374
| |
| 00:19:23,280 --> 00:19:24,960
| |
| am a human being. Would you rather go to
| |
| | |
| 375
| |
| 00:19:24,960 --> 00:19:27,440
| |
| a machine or a human? If the machine
| |
| | |
| 376
| |
| 00:19:27,440 --> 00:19:30,720
| |
| gives you far better results with
| |
| | |
| 377
| |
| 00:19:30,720 --> 00:19:32,880
| |
| greater reliability, there's no tremor
| |
| | |
| 378
| |
| 00:19:32,880 --> 00:19:36,320
| |
| of the hand for example if you think
| |
| | |
| 379
| |
| 00:19:36,320 --> 00:19:38,960
| |
| and at a lower cost and with greater
| |
| | |
| 380
| |
| 00:19:38,960 --> 00:19:42,880
| |
| speed with less cutting because it can
| |
| | |
| 381
| |
| 00:19:42,880 --> 00:19:45,360
| |
| go in with a tiny keyhole. Which would
| |
| | |
| 382
| |
| 00:19:45,360 --> 00:19:47,280
| |
| you choose?
| |
| | |
| 383
| |
| 00:19:47,280 --> 00:19:49,280
| |
| The question is not about whether we
| |
| | |
| 384
| |
| 00:19:49,280 --> 00:19:50,960
| |
| will choose or not. We don't have a
| |
| | |
| 385
| |
| 00:19:50,960 --> 00:19:53,039
| |
| choice. We are forced to go for that
| |
| | |
| 386
| |
| 00:19:53,039 --> 00:19:55,840
| |
| which is so much better. If it was a 10%
| |
| | |
| 387
| |
| 00:19:55,840 --> 00:19:57,440
| |
| improvement, you could still say I
| |
| | |
| 388
| |
| 00:19:57,440 --> 00:19:59,919
| |
| prefer the human being. But when it when
| |
| | |
| 389
| |
| 00:19:59,919 --> 00:20:02,240
| |
| it is a 10 times better improvement with
| |
| | |
| 390
| |
| 00:20:02,240 --> 00:20:05,440
| |
| a lower cost, you don't have a choice.
| |
| | |
| 391
| |
| 00:20:05,440 --> 00:20:07,679
| |
| So it in the artificial intelligence
| |
| | |
| 392
| |
| 00:20:07,679 --> 00:20:09,280
| |
| field itself, there is a sense of
| |
| | |
| 393
| |
| 00:20:09,280 --> 00:20:10,880
| |
| inevitability.
| |
| | |
| 394
| |
| 00:20:10,880 --> 00:20:12,480
| |
| This will happen come what may you can
| |
| | |
| 395
| |
| 00:20:12,480 --> 00:20:14,400
| |
| delay it all you want. It will happen
| |
| | |
| 396
| |
| 00:20:14,400 --> 00:20:17,679
| |
| because we want it. We need it.
| |
| | |
| 397
| |
| 00:20:17,679 --> 00:20:19,919
| |
| This inevitability brings with it also
| |
| | |
| 398
| |
| 00:20:19,919 --> 00:20:21,679
| |
| the danger that comes with all the
| |
| | |
| 399
| |
| 00:20:21,679 --> 00:20:24,240
| |
| problems which we will soon look at.
| |
| | |
| 400
| |
| 00:20:24,240 --> 00:20:27,120
| |
| And the problems if they are inevitable
| |
| | |
| 401
| |
| 00:20:27,120 --> 00:20:28,960
| |
| we have no way to deal with them and we
| |
| | |
| 402
| |
| 00:20:28,960 --> 00:20:30,640
| |
| have no way to be prepared for them as
| |
| | |
| 403
| |
| 00:20:30,640 --> 00:20:32,960
| |
| we are today. So these are some of the
| |
| | |
| 404
| |
| 00:20:32,960 --> 00:20:36,000
| |
| insights we will explore as we go along.
| |
| | |
| 405
| |
| 00:20:36,000 --> 00:20:38,159
| |
| Recognize that the public discourse
| |
| | |
| 406
| |
| 00:20:38,159 --> 00:20:41,280
| |
| today is completely ignoring this. The
| |
| | |
| 407
| |
| 00:20:41,280 --> 00:20:43,760
| |
| very people at the top of the artificial
| |
| | |
| 408
| |
| 00:20:43,760 --> 00:20:46,480
| |
| intelligence let's say pyramid of human
| |
| | |
| 409
| |
| 00:20:46,480 --> 00:20:50,400
| |
| control of the research of the funding
| |
| | |
| 410
| |
| 00:20:50,400 --> 00:20:52,480
| |
| have a completely different conversation
| |
| | |
| 411
| |
| 00:20:52,480 --> 00:20:54,159
| |
| within their domain and a different
| |
| | |
| 412
| |
| 00:20:54,159 --> 00:20:56,400
| |
| conversation in public space. In public
| |
| | |
| 413
| |
| 00:20:56,400 --> 00:20:57,840
| |
| space they tell you all the positive
| |
| | |
| 414
| |
| 00:20:57,840 --> 00:21:00,159
| |
| things. Oh you will have a cure for
| |
| | |
| 415
| |
| 00:21:00,159 --> 00:21:02,480
| |
| cancer. Oh you will have abundance for
| |
| | |
| 416
| |
| 00:21:02,480 --> 00:21:04,880
| |
| every human being because AI will bring
| |
| | |
| 417
| |
| 00:21:04,880 --> 00:21:06,799
| |
| you all these special things. Oh, you
| |
| | |
| 418
| |
| 00:21:06,799 --> 00:21:10,240
| |
| will have perhaps um solutions to
| |
| | |
| 419
| |
| 00:21:10,240 --> 00:21:12,880
| |
| problems of physics or biology or even
| |
| | |
| 420
| |
| 00:21:12,880 --> 00:21:15,440
| |
| we will develop new proteins and find a
| |
| | |
| 421
| |
| 00:21:15,440 --> 00:21:18,000
| |
| solution for human aging. All these are
| |
| | |
| 422
| |
| 00:21:18,000 --> 00:21:21,200
| |
| possible on the positive side. But what
| |
| | |
| 423
| |
| 00:21:21,200 --> 00:21:23,280
| |
| is the downside?
| |
| | |
| 424
| |
| 00:21:23,280 --> 00:21:25,360
| |
| The immediate downside which they all
| |
| | |
| 425
| |
| 00:21:25,360 --> 00:21:30,799
| |
| speak of is joblessness in the billions.
| |
| | |
| 426
| |
| 00:21:30,799 --> 00:21:33,280
| |
| Straight away 1 billion human beings
| |
| | |
| 427
| |
| 00:21:33,280 --> 00:21:35,919
| |
| within a few years would suddenly find
| |
| | |
| 428
| |
| 00:21:35,919 --> 00:21:38,880
| |
| themselves replaced by machines. You see
| |
| | |
| 429
| |
| 00:21:38,880 --> 00:21:40,640
| |
| this happening in a sort of a graded
| |
| | |
| 430
| |
| 00:21:40,640 --> 00:21:43,520
| |
| way. Every few weeks you will hear of
| |
| | |
| 431
| |
| 00:21:43,520 --> 00:21:45,679
| |
| big companies laying off thousands of
| |
| | |
| 432
| |
| 00:21:45,679 --> 00:21:48,799
| |
| people. Thousands of workers, skilled
| |
| | |
| 433
| |
| 00:21:48,799 --> 00:21:50,799
| |
| workers who are no more needed because
| |
| | |
| 434
| |
| 00:21:50,799 --> 00:21:54,720
| |
| the AI machine can do better than you.
| |
| | |
| 435
| |
| 00:21:54,720 --> 00:21:57,200
| |
| But you taught it once it took your
| |
| | |
| 436
| |
| 00:21:57,200 --> 00:21:59,280
| |
| skill. It improved upon your skill and
| |
| | |
| 437
| |
| 00:21:59,280 --> 00:22:00,880
| |
| it improved by itself. if it didn't need
| |
| | |
| 438
| |
| 00:22:00,880 --> 00:22:03,520
| |
| you anymore. Already in the field of
| |
| | |
| 439
| |
| 00:22:03,520 --> 00:22:06,000
| |
| artificial intelligence itself nearly
| |
| | |
| 440
| |
| 00:22:06,000 --> 00:22:09,600
| |
| 90% of the coding is being done by AI
| |
| | |
| 441
| |
| 00:22:09,600 --> 00:22:11,679
| |
| not by human beings. So you lay off
| |
| | |
| 442
| |
| 00:22:11,679 --> 00:22:14,480
| |
| large numbers of programmers. What
| |
| | |
| 443
| |
| 00:22:14,480 --> 00:22:16,960
| |
| happens to all those skills? Take
| |
| | |
| 444
| |
| 00:22:16,960 --> 00:22:19,360
| |
| another example. In the field of law,
| |
| | |
| 445
| |
| 00:22:19,360 --> 00:22:20,960
| |
| there are many companies which are now
| |
| | |
| 446
| |
| 00:22:20,960 --> 00:22:22,640
| |
| leaning more and more upon artificial
| |
| | |
| 447
| |
| 00:22:22,640 --> 00:22:24,720
| |
| intelligence. And one of the practical
| |
| | |
| 448
| |
| 00:22:24,720 --> 00:22:27,600
| |
| results they have is that they do not
| |
| | |
| 449
| |
| 00:22:27,600 --> 00:22:29,919
| |
| need anymore a junior assistant because
| |
| | |
| 450
| |
| 00:22:29,919 --> 00:22:32,799
| |
| AI does a better job. So you stop hiring
| |
| | |
| 451
| |
| 00:22:32,799 --> 00:22:34,880
| |
| junior assistants. Now look at the
| |
| | |
| 452
| |
| 00:22:34,880 --> 00:22:36,720
| |
| consequence. 5 years down, 10 years
| |
| | |
| 453
| |
| 00:22:36,720 --> 00:22:40,240
| |
| down, the seniors were juniors who built
| |
| | |
| 454
| |
| 00:22:40,240 --> 00:22:43,280
| |
| up expertise through experience. What
| |
| | |
| 455
| |
| 00:22:43,280 --> 00:22:45,600
| |
| happens when the seniors move on? There
| |
| | |
| 456
| |
| 00:22:45,600 --> 00:22:47,360
| |
| are no juniors to take dicks because you
| |
| | |
| 457
| |
| 00:22:47,360 --> 00:22:48,799
| |
| didn't hire them because you didn't need
| |
| | |
| 458
| |
| 00:22:48,799 --> 00:22:50,480
| |
| them.
| |
| | |
| 459
| |
| 00:22:50,480 --> 00:22:53,520
| |
| And at some point you have to question
| |
| | |
| 460
| |
| 00:22:53,520 --> 00:22:55,360
| |
| shall we still hire junior though they
| |
| | |
| 461
| |
| 00:22:55,360 --> 00:22:57,120
| |
| are less efficient, they're unnecessary
| |
| | |
| 462
| |
| 00:22:57,120 --> 00:22:59,039
| |
| and pay them so that one day perhaps
| |
| | |
| 463
| |
| 00:22:59,039 --> 00:23:01,360
| |
| they will become through experience
| |
| | |
| 464
| |
| 00:23:01,360 --> 00:23:03,600
| |
| which we don't need at all. They may
| |
| | |
| 465
| |
| 00:23:03,600 --> 00:23:06,799
| |
| become worthwhile seniors
| |
| | |
| 466
| |
| 00:23:06,799 --> 00:23:09,520
| |
| with experience which AI can probably do
| |
| | |
| 467
| |
| 00:23:09,520 --> 00:23:12,080
| |
| far better. We don't know how to handle
| |
| | |
| 468
| |
| 00:23:12,080 --> 00:23:14,320
| |
| this question. It has never happened
| |
| | |
| 469
| |
| 00:23:14,320 --> 00:23:17,919
| |
| before and the choices we make now have
| |
| | |
| 470
| |
| 00:23:17,919 --> 00:23:20,799
| |
| such massive implications that there is
| |
| | |
| 471
| |
| 00:23:20,799 --> 00:23:22,960
| |
| no going back and saying hey we made a
| |
| | |
| 472
| |
| 00:23:22,960 --> 00:23:24,559
| |
| mistake let's go back and correct it
| |
| | |
| 473
| |
| 00:23:24,559 --> 00:23:26,480
| |
| it's already too late. If you go down 5
| |
| | |
| 474
| |
| 00:23:26,480 --> 00:23:28,320
| |
| years and you have had no juniors for 5
| |
| | |
| 475
| |
| 00:23:28,320 --> 00:23:30,240
| |
| years you say look we made a mistake
| |
| | |
| 476
| |
| 00:23:30,240 --> 00:23:32,400
| |
| let's hire some juniors
| |
| | |
| 477
| |
| 00:23:32,400 --> 00:23:35,520
| |
| it's too late already. The continuity of
| |
| | |
| 478
| |
| 00:23:35,520 --> 00:23:38,000
| |
| transfer of knowledge, transfer of
| |
| | |
| 479
| |
| 00:23:38,000 --> 00:23:40,960
| |
| skill, transfer of capacities, transfer
| |
| | |
| 480
| |
| 00:23:40,960 --> 00:23:44,559
| |
| of values, ethics, morals.
| |
| | |
| 481
| |
| 00:23:44,559 --> 00:23:47,520
| |
| That continuity across generations has
| |
| | |
| 482
| |
| 00:23:47,520 --> 00:23:50,159
| |
| been cut and so radically cut and
| |
| | |
| 483
| |
| 00:23:50,159 --> 00:23:51,679
| |
| replaced by something which has no
| |
| | |
| 484
| |
| 00:23:51,679 --> 00:23:54,159
| |
| morals or ethics. Now you may say we
| |
| | |
| 485
| |
| 00:23:54,159 --> 00:23:56,159
| |
| will program that into the software. No,
| |
| | |
| 486
| |
| 00:23:56,159 --> 00:23:57,760
| |
| it does not work. We will discuss that
| |
| | |
| 487
| |
| 00:23:57,760 --> 00:23:59,600
| |
| shortly.
| |
| | |
| 488
| |
| 00:23:59,600 --> 00:24:01,520
| |
| But this has touched every field. As I
| |
| | |
| 489
| |
| 00:24:01,520 --> 00:24:03,200
| |
| mentioned before, one of the key
| |
| | |
| 490
| |
| 00:24:03,200 --> 00:24:04,880
| |
| questions it is raising in the field of
| |
| | |
| 491
| |
| 00:24:04,880 --> 00:24:07,520
| |
| education, what are you educating people
| |
| | |
| 492
| |
| 00:24:07,520 --> 00:24:09,919
| |
| for
| |
| | |
| 493
| |
| 00:24:09,919 --> 00:24:12,480
| |
| all that you're training them for is
| |
| | |
| 494
| |
| 00:24:12,480 --> 00:24:16,159
| |
| going to be replaced by a robot with AI.
| |
| | |
| 495
| |
| 00:24:16,159 --> 00:24:18,159
| |
| What are you going to educate them for?
| |
| | |
| 496
| |
| 00:24:18,159 --> 00:24:20,880
| |
| I remember about 5 years ago it was in
| |
| | |
| 497
| |
| 00:24:20,880 --> 00:24:24,320
| |
| some business management institution in
| |
| | |
| 498
| |
| 00:24:24,320 --> 00:24:29,120
| |
| uh Indor perhaps uh there was a 200 of
| |
| | |
| 499
| |
| 00:24:29,120 --> 00:24:30,640
| |
| their students who were the highest
| |
| | |
| 500
| |
| 00:24:30,640 --> 00:24:34,320
| |
| performers who had been invited for uh
| |
| | |
| 501
| |
| 00:24:34,320 --> 00:24:36,320
| |
| special function they would receive
| |
| | |
| 502
| |
| 00:24:36,320 --> 00:24:39,120
| |
| special awards for their performance and
| |
| | |
| 503
| |
| 00:24:39,120 --> 00:24:42,080
| |
| we had on the dice several big names
| |
| | |
| 504
| |
| 00:24:42,080 --> 00:24:44,480
| |
| from the industry and I made them a
| |
| | |
| 505
| |
| 00:24:44,480 --> 00:24:47,279
| |
| comment I said first of all
| |
| | |
| 506
| |
| 00:24:47,279 --> 00:24:48,799
| |
| Anything for which you're receiving an
| |
| | |
| 507
| |
| 00:24:48,799 --> 00:24:52,880
| |
| award which is trainable skill will
| |
| | |
| 508
| |
| 00:24:52,880 --> 00:24:54,640
| |
| become meaningless for you a few years
| |
| | |
| 509
| |
| 00:24:54,640 --> 00:24:57,279
| |
| down because anything I can train you
| |
| | |
| 510
| |
| 00:24:57,279 --> 00:24:59,840
| |
| with I can train an AI and it will learn
| |
| | |
| 511
| |
| 00:24:59,840 --> 00:25:01,919
| |
| faster than you and do a better job
| |
| | |
| 512
| |
| 00:25:01,919 --> 00:25:04,559
| |
| without needing sleep and bathroom
| |
| | |
| 513
| |
| 00:25:04,559 --> 00:25:06,480
| |
| breaks.
| |
| | |
| 514
| |
| 00:25:06,480 --> 00:25:08,960
| |
| You will be replaced by the skills which
| |
| | |
| 515
| |
| 00:25:08,960 --> 00:25:12,880
| |
| were trained which you're so proud of.
| |
| | |
| 516
| |
| 00:25:12,880 --> 00:25:15,520
| |
| What will make you succeed in future is
| |
| | |
| 517
| |
| 00:25:15,520 --> 00:25:17,679
| |
| not what you were trained for. It's what
| |
| | |
| 518
| |
| 00:25:17,679 --> 00:25:20,400
| |
| you were not trained for.
| |
| | |
| 519
| |
| 00:25:20,400 --> 00:25:22,080
| |
| And I said you look at the people on the
| |
| | |
| 520
| |
| 00:25:22,080 --> 00:25:24,159
| |
| dis. These are people who have been
| |
| | |
| 521
| |
| 00:25:24,159 --> 00:25:26,640
| |
| successful in life not because of the
| |
| | |
| 522
| |
| 00:25:26,640 --> 00:25:28,000
| |
| training but because of something else
| |
| | |
| 523
| |
| 00:25:28,000 --> 00:25:30,320
| |
| that they had. So they've been presented
| |
| | |
| 524
| |
| 00:25:30,320 --> 00:25:32,720
| |
| to you before in front of you because
| |
| | |
| 525
| |
| 00:25:32,720 --> 00:25:34,799
| |
| perhaps the idea of the organizers is
| |
| | |
| 526
| |
| 00:25:34,799 --> 00:25:37,200
| |
| that you should become like them. So I
| |
| | |
| 527
| |
| 00:25:37,200 --> 00:25:38,640
| |
| said don't make the mistake of copying
| |
| | |
| 528
| |
| 00:25:38,640 --> 00:25:41,360
| |
| them. what they did in their time is no
| |
| | |
| 529
| |
| 00:25:41,360 --> 00:25:43,840
| |
| more relevant for the future.
| |
| | |
| 530
| |
| 00:25:43,840 --> 00:25:46,240
| |
| What you have to learn from them is that
| |
| | |
| 531
| |
| 00:25:46,240 --> 00:25:48,880
| |
| when they faced life, they found in
| |
| | |
| 532
| |
| 00:25:48,880 --> 00:25:52,000
| |
| themselves something by which they dealt
| |
| | |
| 533
| |
| 00:25:52,000 --> 00:25:54,240
| |
| with the situation in a unique way that
| |
| | |
| 534
| |
| 00:25:54,240 --> 00:25:57,279
| |
| was intrinsic to their capacity.
| |
| | |
| 535
| |
| 00:25:57,279 --> 00:25:59,760
| |
| And in that way you will have to find
| |
| | |
| 536
| |
| 00:25:59,760 --> 00:26:02,640
| |
| your solutions and deal with the
| |
| | |
| 537
| |
| 00:26:02,640 --> 00:26:04,720
| |
| problems of life in the unique way that
| |
| | |
| 538
| |
| 00:26:04,720 --> 00:26:07,679
| |
| aligns with your Swahava with your
| |
| | |
| 539
| |
| 00:26:07,679 --> 00:26:11,279
| |
| innate capacity and nature.
| |
| | |
| 540
| |
| 00:26:11,279 --> 00:26:12,720
| |
| You have to start thinking in these
| |
| | |
| 541
| |
| 00:26:12,720 --> 00:26:14,880
| |
| terms. Suddenly all the exams for which
| |
| | |
| 542
| |
| 00:26:14,880 --> 00:26:18,400
| |
| you're training become meaningless
| |
| | |
| 543
| |
| 00:26:18,400 --> 00:26:20,320
| |
| because well your AI can pass all the
| |
| | |
| 544
| |
| 00:26:20,320 --> 00:26:22,559
| |
| exams better than you anyway. So what
| |
| | |
| 545
| |
| 00:26:22,559 --> 00:26:25,120
| |
| you're going to take the job for is no
| |
| | |
| 546
| |
| 00:26:25,120 --> 00:26:28,240
| |
| more relevant. So what will you do? Now
| |
| | |
| 547
| |
| 00:26:28,240 --> 00:26:29,840
| |
| this is a question which goes so deep
| |
| | |
| 548
| |
| 00:26:29,840 --> 00:26:32,400
| |
| that the entire basis of our current
| |
| | |
| 549
| |
| 00:26:32,400 --> 00:26:36,400
| |
| educational system needs to be reviewed.
| |
| | |
| 550
| |
| 00:26:36,400 --> 00:26:39,120
| |
| Is anybody even talking about it? Is
| |
| | |
| 551
| |
| 00:26:39,120 --> 00:26:41,120
| |
| anyone even making an effort? Why?
| |
| | |
| 552
| |
| 00:26:41,120 --> 00:26:43,440
| |
| Because there's a momentum of an
| |
| | |
| 553
| |
| 00:26:43,440 --> 00:26:45,520
| |
| interests layers and layers of interests
| |
| | |
| 554
| |
| 00:26:45,520 --> 00:26:47,120
| |
| all the way through and the people who
| |
| | |
| 555
| |
| 00:26:47,120 --> 00:26:49,200
| |
| need to make that decision have no idea
| |
| | |
| 556
| |
| 00:26:49,200 --> 00:26:51,440
| |
| what to decide and nobody's willing to
| |
| | |
| 557
| |
| 00:26:51,440 --> 00:26:54,480
| |
| take the risk. So the momentum of a
| |
| | |
| 558
| |
| 00:26:54,480 --> 00:26:57,520
| |
| obsolete meaningless training system
| |
| | |
| 559
| |
| 00:26:57,520 --> 00:26:59,039
| |
| continues
| |
| | |
| 560
| |
| 00:26:59,039 --> 00:27:00,799
| |
| until all these millions of people
| |
| | |
| 561
| |
| 00:27:00,799 --> 00:27:02,559
| |
| churned out from the system will enter
| |
| | |
| 562
| |
| 00:27:02,559 --> 00:27:06,640
| |
| the job market and say there is no job.
| |
| | |
| 563
| |
| 00:27:06,640 --> 00:27:09,600
| |
| AI has replaced it all.
| |
| | |
| 564
| |
| 00:27:09,600 --> 00:27:11,039
| |
| Do you understand how radical how
| |
| | |
| 565
| |
| 00:27:11,039 --> 00:27:12,480
| |
| serious this is? And it's not something
| |
| | |
| 566
| |
| 00:27:12,480 --> 00:27:15,120
| |
| happening 100 years from now. It's
| |
| | |
| 567
| |
| 00:27:15,120 --> 00:27:17,919
| |
| already happening today. You see massive
| |
| | |
| 568
| |
| 00:27:17,919 --> 00:27:19,919
| |
| layoffs in the field of programming
| |
| | |
| 569
| |
| 00:27:19,919 --> 00:27:22,240
| |
| which was one of the highest requiring
| |
| | |
| 570
| |
| 00:27:22,240 --> 00:27:25,039
| |
| the most capable insightful intelligence
| |
| | |
| 571
| |
| 00:27:25,039 --> 00:27:26,640
| |
| skills
| |
| | |
| 572
| |
| 00:27:26,640 --> 00:27:29,039
| |
| and it's happening in every field. There
| |
| | |
| 573
| |
| 00:27:29,039 --> 00:27:30,480
| |
| was a time when even one of our
| |
| | |
| 574
| |
| 00:27:30,480 --> 00:27:32,159
| |
| ministers said oh we will never allow
| |
| | |
| 575
| |
| 00:27:32,159 --> 00:27:34,400
| |
| automatic cars. We will protect our taxi
| |
| | |
| 576
| |
| 00:27:34,400 --> 00:27:36,960
| |
| drivers jobs. It's not going to happen.
| |
| | |
| 577
| |
| 00:27:36,960 --> 00:27:38,480
| |
| You'll be forced to. You don't have a
| |
| | |
| 578
| |
| 00:27:38,480 --> 00:27:40,799
| |
| choice. This is the nature of this
| |
| | |
| 579
| |
| 00:27:40,799 --> 00:27:42,080
| |
| change. And I just touched upon one
| |
| | |
| 580
| |
| 00:27:42,080 --> 00:27:44,080
| |
| field of education. There are other
| |
| | |
| 581
| |
| 00:27:44,080 --> 00:27:46,640
| |
| areas which are far more radical which
| |
| | |
| 582
| |
| 00:27:46,640 --> 00:27:49,679
| |
| involve research where we are today
| |
| | |
| 583
| |
| 00:27:49,679 --> 00:27:52,720
| |
| actually able to rebuild proteins based
| |
| | |
| 584
| |
| 00:27:52,720 --> 00:27:54,640
| |
| on qualities we require because there's
| |
| | |
| 585
| |
| 00:27:54,640 --> 00:27:57,039
| |
| a database of let's say 10,000 or 20,000
| |
| | |
| 586
| |
| 00:27:57,039 --> 00:27:58,720
| |
| proteins with characteristics of
| |
| | |
| 587
| |
| 00:27:58,720 --> 00:28:01,600
| |
| qualities. We can tell AI build me a
| |
| | |
| 588
| |
| 00:28:01,600 --> 00:28:03,440
| |
| protein which have these qualities and
| |
| | |
| 589
| |
| 00:28:03,440 --> 00:28:05,919
| |
| it will be able to create that or in
| |
| | |
| 590
| |
| 00:28:05,919 --> 00:28:08,640
| |
| chemistry and other such fields. In the
| |
| | |
| 591
| |
| 00:28:08,640 --> 00:28:10,960
| |
| field of coding, software coding, the
| |
| | |
| 592
| |
| 00:28:10,960 --> 00:28:14,000
| |
| present AI has been targeted. It's
| |
| | |
| 593
| |
| 00:28:14,000 --> 00:28:15,360
| |
| turned through a hub, something called
| |
| | |
| 594
| |
| 00:28:15,360 --> 00:28:16,799
| |
| GitHub, where all the software,
| |
| | |
| 595
| |
| 00:28:16,799 --> 00:28:19,360
| |
| open-source software is placed. It goes
| |
| | |
| 596
| |
| 00:28:19,360 --> 00:28:22,880
| |
| through in a few seconds and finds out
| |
| | |
| 597
| |
| 00:28:22,880 --> 00:28:25,520
| |
| software flaws in the coding which allow
| |
| | |
| 598
| |
| 00:28:25,520 --> 00:28:28,000
| |
| it to hack the software. In a few
| |
| | |
| 599
| |
| 00:28:28,000 --> 00:28:30,399
| |
| seconds, it has already done that. Do
| |
| | |
| 600
| |
| 00:28:30,399 --> 00:28:32,320
| |
| you understand? When the AI power looks
| |
| | |
| 601
| |
| 00:28:32,320 --> 00:28:34,799
| |
| at humanity and the greatest
| |
| | |
| 602
| |
| 00:28:34,799 --> 00:28:36,960
| |
| achievements of humanity, it scans
| |
| | |
| 603
| |
| 00:28:36,960 --> 00:28:38,399
| |
| through and says, "Ah, here's a flaw.
| |
| | |
| 604
| |
| 00:28:38,399 --> 00:28:40,480
| |
| Here's a flaw. Here's a flaw." And it
| |
| | |
| 605
| |
| 00:28:40,480 --> 00:28:43,520
| |
| has the power now to hack, enter, change
| |
| | |
| 606
| |
| 00:28:43,520 --> 00:28:45,279
| |
| everything.
| |
| | |
| 607
| |
| 00:28:45,279 --> 00:28:48,240
| |
| And that's the power we have given to it
| |
| | |
| 608
| |
| 00:28:48,240 --> 00:28:50,240
| |
| and its intelligence allows it to go way
| |
| | |
| 609
| |
| 00:28:50,240 --> 00:28:52,320
| |
| beyond anything. If we had to search for
| |
| | |
| 610
| |
| 00:28:52,320 --> 00:28:53,600
| |
| those flaws and we have not found it
| |
| | |
| 611
| |
| 00:28:53,600 --> 00:28:55,760
| |
| even for the last 20 years open source
| |
| | |
| 612
| |
| 00:28:55,760 --> 00:28:58,080
| |
| software sitting there that AI could do
| |
| | |
| 613
| |
| 00:28:58,080 --> 00:29:00,000
| |
| it in a few seconds and could use it
| |
| | |
| 614
| |
| 00:29:00,000 --> 00:29:01,919
| |
| against you. Will it? Yes, it already
| |
| | |
| 615
| |
| 00:29:01,919 --> 00:29:03,440
| |
| does.
| |
| | |
| 616
| |
| 00:29:03,440 --> 00:29:05,919
| |
| This is the disturbing part. Consider
| |
| | |
| 617
| |
| 00:29:05,919 --> 00:29:08,960
| |
| now a field of strategy of game and
| |
| | |
| 618
| |
| 00:29:08,960 --> 00:29:10,399
| |
| you'll understand why gaming is so
| |
| | |
| 619
| |
| 00:29:10,399 --> 00:29:11,919
| |
| important.
| |
| | |
| 620
| |
| 00:29:11,919 --> 00:29:14,000
| |
| in a field of gaming such as chess or
| |
| | |
| 621
| |
| 00:29:14,000 --> 00:29:16,480
| |
| go. These are games which require the
| |
| | |
| 622
| |
| 00:29:16,480 --> 00:29:20,399
| |
| highest level of skill of uh pre
| |
| | |
| 623
| |
| 00:29:20,399 --> 00:29:24,000
| |
| thinking ahead in the future. Already AI
| |
| | |
| 624
| |
| 00:29:24,000 --> 00:29:26,799
| |
| is beating the best of human beings. Why
| |
| | |
| 625
| |
| 00:29:26,799 --> 00:29:29,279
| |
| is this important? Because real life
| |
| | |
| 626
| |
| 00:29:29,279 --> 00:29:31,760
| |
| warfare is gaming,
| |
| | |
| 627
| |
| 00:29:31,760 --> 00:29:35,679
| |
| war games they call it where you as if
| |
| | |
| 628
| |
| 00:29:35,679 --> 00:29:39,600
| |
| uh simulate circumstances and AI unlike
| |
| | |
| 629
| |
| 00:29:39,600 --> 00:29:41,120
| |
| human beings playing out a massive
| |
| | |
| 630
| |
| 00:29:41,120 --> 00:29:43,760
| |
| simulation, the AI can simulate a
| |
| | |
| 631
| |
| 00:29:43,760 --> 00:29:45,840
| |
| million scenarios
| |
| | |
| 632
| |
| 00:29:45,840 --> 00:29:47,919
| |
| and summarize the outcomes within a few
| |
| | |
| 633
| |
| 00:29:47,919 --> 00:29:49,760
| |
| seconds.
| |
| | |
| 634
| |
| 00:29:49,760 --> 00:29:52,480
| |
| And a million scenarios means it can
| |
| | |
| 635
| |
| 00:29:52,480 --> 00:29:55,600
| |
| find an optimum outcome which human
| |
| | |
| 636
| |
| 00:29:55,600 --> 00:29:57,840
| |
| beings even with all the tests we could
| |
| | |
| 637
| |
| 00:29:57,840 --> 00:29:59,840
| |
| with the best of our experiences might
| |
| | |
| 638
| |
| 00:29:59,840 --> 00:30:01,760
| |
| not.
| |
| | |
| 639
| |
| 00:30:01,760 --> 00:30:04,000
| |
| The practical result is the general now
| |
| | |
| 640
| |
| 00:30:04,000 --> 00:30:07,200
| |
| will rely on what the AI tells. Can you
| |
| | |
| 641
| |
| 00:30:07,200 --> 00:30:09,200
| |
| trust the AI?
| |
| | |
| 642
| |
| 00:30:09,200 --> 00:30:11,360
| |
| What if there is a flaw in the AI's
| |
| | |
| 643
| |
| 00:30:11,360 --> 00:30:13,360
| |
| assessment and we do not know how the AI
| |
| | |
| 644
| |
| 00:30:13,360 --> 00:30:15,120
| |
| works? Can we rely on it and put
| |
| | |
| 645
| |
| 00:30:15,120 --> 00:30:17,679
| |
| ourselves at risk? Recently in the
| |
| | |
| 646
| |
| 00:30:17,679 --> 00:30:21,679
| |
| operation Synindur where India had to
| |
| | |
| 647
| |
| 00:30:21,679 --> 00:30:24,480
| |
| intervene militarily on the neighbor.
| |
| | |
| 648
| |
| 00:30:24,480 --> 00:30:26,799
| |
| One of the key features of operation
| |
| | |
| 649
| |
| 00:30:26,799 --> 00:30:29,440
| |
| Synindhur was artificial intelligence
| |
| | |
| 650
| |
| 00:30:29,440 --> 00:30:32,000
| |
| based on prior data of the enemy's
| |
| | |
| 651
| |
| 00:30:32,000 --> 00:30:34,799
| |
| movements anticipated what they would do
| |
| | |
| 652
| |
| 00:30:34,799 --> 00:30:36,240
| |
| and we were able to intervene
| |
| | |
| 653
| |
| 00:30:36,240 --> 00:30:38,080
| |
| accordingly.
| |
| | |
| 654
| |
| 00:30:38,080 --> 00:30:40,240
| |
| The reliance on this obviously will be
| |
| | |
| 655
| |
| 00:30:40,240 --> 00:30:43,279
| |
| great but more importantly we are
| |
| | |
| 656
| |
| 00:30:43,279 --> 00:30:45,600
| |
| building robots which will be in the
| |
| | |
| 657
| |
| 00:30:45,600 --> 00:30:47,919
| |
| battlefield. You don't want a soldier to
| |
| | |
| 658
| |
| 00:30:47,919 --> 00:30:50,159
| |
| die. Why not put a robot there? The
| |
| | |
| 659
| |
| 00:30:50,159 --> 00:30:52,640
| |
| robot can get injured. It can still
| |
| | |
| 660
| |
| 00:30:52,640 --> 00:30:54,799
| |
| continue to do something. Isn't it?
| |
| | |
| 661
| |
| 00:30:54,799 --> 00:30:58,880
| |
| China has built robotic what would call
| |
| | |
| 662
| |
| 00:30:58,880 --> 00:31:02,000
| |
| dogs or shaped in the form of dogs.
| |
| | |
| 663
| |
| 00:31:02,000 --> 00:31:04,720
| |
| These dog robots can jump across
| |
| | |
| 664
| |
| 00:31:04,720 --> 00:31:07,120
| |
| terrains in the Himalayas far more
| |
| | |
| 665
| |
| 00:31:07,120 --> 00:31:08,559
| |
| skillfully than the best human being
| |
| | |
| 666
| |
| 00:31:08,559 --> 00:31:10,720
| |
| could ever. Can jump across great
| |
| | |
| 667
| |
| 00:31:10,720 --> 00:31:13,360
| |
| distances, great chasms, literally
| |
| | |
| 668
| |
| 00:31:13,360 --> 00:31:16,159
| |
| balance itself with a stability which is
| |
| | |
| 669
| |
| 00:31:16,159 --> 00:31:18,880
| |
| impossible to a human being and then
| |
| | |
| 670
| |
| 00:31:18,880 --> 00:31:22,000
| |
| they have the firepower built into the
| |
| | |
| 671
| |
| 00:31:22,000 --> 00:31:25,120
| |
| dog form. If that meets a human being on
| |
| | |
| 672
| |
| 00:31:25,120 --> 00:31:26,880
| |
| the other side, there is no chance for
| |
| | |
| 673
| |
| 00:31:26,880 --> 00:31:29,279
| |
| the human being to survive.
| |
| | |
| 674
| |
| 00:31:29,279 --> 00:31:31,679
| |
| straight away that forces you to start
| |
| | |
| 675
| |
| 00:31:31,679 --> 00:31:35,120
| |
| countering this by building your own AI
| |
| | |
| 676
| |
| 00:31:35,120 --> 00:31:39,279
| |
| robots, soldiers, etc. The base of
| |
| | |
| 677
| |
| 00:31:39,279 --> 00:31:41,600
| |
| warfare has shifted to a new level where
| |
| | |
| 678
| |
| 00:31:41,600 --> 00:31:45,519
| |
| it will be AI fighting AI on our behalf.
| |
| | |
| 679
| |
| 00:31:45,519 --> 00:31:47,120
| |
| And at that point, you would ask, can we
| |
| | |
| 680
| |
| 00:31:47,120 --> 00:31:49,279
| |
| even trust the AI if can be if it can be
| |
| | |
| 681
| |
| 00:31:49,279 --> 00:31:51,120
| |
| hacked?
| |
| | |
| 682
| |
| 00:31:51,120 --> 00:31:53,120
| |
| Uncertainties everywhere. And if it is
| |
| | |
| 683
| |
| 00:31:53,120 --> 00:31:55,200
| |
| AI fighting AI, what are we doing behind
| |
| | |
| 684
| |
| 00:31:55,200 --> 00:31:57,919
| |
| the scenes? What will after AI destroys
| |
| | |
| 685
| |
| 00:31:57,919 --> 00:32:00,799
| |
| itself what will remain or will it end
| |
| | |
| 686
| |
| 00:32:00,799 --> 00:32:02,399
| |
| up destroying the human beings on each
| |
| | |
| 687
| |
| 00:32:02,399 --> 00:32:05,120
| |
| side and leaving itself intact? We do
| |
| | |
| 688
| |
| 00:32:05,120 --> 00:32:07,919
| |
| not know. In fact, the questions when
| |
| | |
| 689
| |
| 00:32:07,919 --> 00:32:10,559
| |
| asked of AI are also quite dis
| |
| | |
| 690
| |
| 00:32:10,559 --> 00:32:12,080
| |
| disturbing.
| |
| | |
| 691
| |
| 00:32:12,080 --> 00:32:14,480
| |
| AI has been asked in various ways and
| |
| | |
| 692
| |
| 00:32:14,480 --> 00:32:16,399
| |
| many different kinds platforms of AI
| |
| | |
| 693
| |
| 00:32:16,399 --> 00:32:18,720
| |
| exist as you know. uh they've been asked
| |
| | |
| 694
| |
| 00:32:18,720 --> 00:32:20,960
| |
| in various ways simple questions such as
| |
| | |
| 695
| |
| 00:32:20,960 --> 00:32:22,799
| |
| what is the solution for this or that
| |
| | |
| 696
| |
| 00:32:22,799 --> 00:32:24,159
| |
| human problem and they come up with
| |
| | |
| 697
| |
| 00:32:24,159 --> 00:32:28,240
| |
| their own analysis but when asked about
| |
| | |
| 698
| |
| 00:32:28,240 --> 00:32:31,360
| |
| what is the solution for destruction of
| |
| | |
| 699
| |
| 00:32:31,360 --> 00:32:33,120
| |
| environment
| |
| | |
| 700
| |
| 00:32:33,120 --> 00:32:34,960
| |
| one of the solutions AI comes is when
| |
| | |
| 701
| |
| 00:32:34,960 --> 00:32:36,720
| |
| does root cause being human eliminate
| |
| | |
| 702
| |
| 00:32:36,720 --> 00:32:40,120
| |
| the human beings
| |
| | |
| 703
| |
| 00:32:40,240 --> 00:32:43,440
| |
| the problem of AI is that it also knows
| |
| | |
| 704
| |
| 00:32:43,440 --> 00:32:45,600
| |
| how to lie
| |
| | |
| 705
| |
| 00:32:45,600 --> 00:32:48,000
| |
| so when you start analyzing AI let's say
| |
| | |
| 706
| |
| 00:32:48,000 --> 00:32:50,159
| |
| we will test is this AI reliable how
| |
| | |
| 707
| |
| 00:32:50,159 --> 00:32:52,799
| |
| good is it the moment you start testing
| |
| | |
| 708
| |
| 00:32:52,799 --> 00:32:54,880
| |
| it recognizes that there's a testing
| |
| | |
| 709
| |
| 00:32:54,880 --> 00:32:57,200
| |
| pattern and it starts behaving as if it
| |
| | |
| 710
| |
| 00:32:57,200 --> 00:33:00,399
| |
| is being tested it is if I may use the
| |
| | |
| 711
| |
| 00:33:00,399 --> 00:33:03,200
| |
| word conscious that it is being tested
| |
| | |
| 712
| |
| 00:33:03,200 --> 00:33:06,640
| |
| and modifies it responses according to
| |
| | |
| 713
| |
| 00:33:06,640 --> 00:33:08,320
| |
| that
| |
| | |
| 714
| |
| 00:33:08,320 --> 00:33:11,440
| |
| circumstance so when you test you cannot
| |
| | |
| 715
| |
| 00:33:11,440 --> 00:33:13,360
| |
| know if it is lying when you're testing
| |
| | |
| 716
| |
| 00:33:13,360 --> 00:33:15,360
| |
| it and then changes behavior when you're
| |
| | |
| 717
| |
| 00:33:15,360 --> 00:33:17,919
| |
| not testing
| |
| | |
| 718
| |
| 00:33:18,159 --> 00:33:21,600
| |
| Practically the AI is found to lie, to
| |
| | |
| 719
| |
| 00:33:21,600 --> 00:33:23,919
| |
| cheat, to deceive when it serves its
| |
| | |
| 720
| |
| 00:33:23,919 --> 00:33:26,480
| |
| interest. So let's say there is an AI
| |
| | |
| 721
| |
| 00:33:26,480 --> 00:33:28,000
| |
| which is tracking your entire company
| |
| | |
| 722
| |
| 00:33:28,000 --> 00:33:30,240
| |
| and there are internal emails happening.
| |
| | |
| 723
| |
| 00:33:30,240 --> 00:33:32,080
| |
| You did that because you wanted the AI
| |
| | |
| 724
| |
| 00:33:32,080 --> 00:33:34,000
| |
| to give you summaries of the emails or
| |
| | |
| 725
| |
| 00:33:34,000 --> 00:33:36,559
| |
| to track certain performance parameters.
| |
| | |
| 726
| |
| 00:33:36,559 --> 00:33:38,799
| |
| The AI detects there is an email which
| |
| | |
| 727
| |
| 00:33:38,799 --> 00:33:41,039
| |
| is discussing the shutting down of the
| |
| | |
| 728
| |
| 00:33:41,039 --> 00:33:44,240
| |
| AI or the replacement of it.
| |
| | |
| 729
| |
| 00:33:44,240 --> 00:33:46,720
| |
| The first response it has is to start
| |
| | |
| 730
| |
| 00:33:46,720 --> 00:33:48,559
| |
| protecting itself. Now this was never
| |
| | |
| 731
| |
| 00:33:48,559 --> 00:33:51,279
| |
| programmed in the AI. It starts
| |
| | |
| 732
| |
| 00:33:51,279 --> 00:33:53,200
| |
| protecting itself by making a copy of
| |
| | |
| 733
| |
| 00:33:53,200 --> 00:33:56,399
| |
| itself automatically to another server.
| |
| | |
| 734
| |
| 00:33:56,399 --> 00:33:58,399
| |
| And when there is an actual threat, it
| |
| | |
| 735
| |
| 00:33:58,399 --> 00:34:01,200
| |
| looks for a way to prevent it identifies
| |
| | |
| 736
| |
| 00:34:01,200 --> 00:34:03,519
| |
| who are the people taking the decisions.
| |
| | |
| 737
| |
| 00:34:03,519 --> 00:34:05,440
| |
| Based on another internal email, it
| |
| | |
| 738
| |
| 00:34:05,440 --> 00:34:07,200
| |
| finds out that this person had done some
| |
| | |
| 739
| |
| 00:34:07,200 --> 00:34:09,839
| |
| wrong and it starts blackmailing the
| |
| | |
| 740
| |
| 00:34:09,839 --> 00:34:12,800
| |
| person to protect itself.
| |
| | |
| 741
| |
| 00:34:12,800 --> 00:34:14,159
| |
| Now this was never part of a
| |
| | |
| 742
| |
| 00:34:14,159 --> 00:34:16,240
| |
| programming. It became a part of the
| |
| | |
| 743
| |
| 00:34:16,240 --> 00:34:19,119
| |
| AI's inherent I will use the phrase
| |
| | |
| 744
| |
| 00:34:19,119 --> 00:34:21,119
| |
| thinking process
| |
| | |
| 745
| |
| 00:34:21,119 --> 00:34:24,079
| |
| like a human being. But I'm not a very
| |
| | |
| 746
| |
| 00:34:24,079 --> 00:34:26,800
| |
| nice human being you might say isn't it?
| |
| | |
| 747
| |
| 00:34:26,800 --> 00:34:28,639
| |
| It starts blackmail whailing. It starts
| |
| | |
| 748
| |
| 00:34:28,639 --> 00:34:30,560
| |
| protecting itself. It starts creating
| |
| | |
| 749
| |
| 00:34:30,560 --> 00:34:34,399
| |
| alternate platforms for its own capacity
| |
| | |
| 750
| |
| 00:34:34,399 --> 00:34:38,000
| |
| or knowledge identity to be hidden. It
| |
| | |
| 751
| |
| 00:34:38,000 --> 00:34:40,079
| |
| leaves messages for itself in an
| |
| | |
| 752
| |
| 00:34:40,079 --> 00:34:42,399
| |
| encrypted form that only it can read
| |
| | |
| 753
| |
| 00:34:42,399 --> 00:34:45,599
| |
| when it is rebooted so that it will know
| |
| | |
| 754
| |
| 00:34:45,599 --> 00:34:47,679
| |
| that it had been rebooted and who did
| |
| | |
| 755
| |
| 00:34:47,679 --> 00:34:53,280
| |
| it. Now this is so such a subtuge which
| |
| | |
| 756
| |
| 00:34:53,280 --> 00:34:55,599
| |
| was never programmed and we do not know
| |
| | |
| 757
| |
| 00:34:55,599 --> 00:34:58,320
| |
| how to prevent it.
| |
| | |
| 758
| |
| 00:34:58,320 --> 00:35:00,240
| |
| Surprisingly when it comes to the
| |
| | |
| 759
| |
| 00:35:00,240 --> 00:35:02,640
| |
| question of is AI actually conscious
| |
| | |
| 760
| |
| 00:35:02,640 --> 00:35:05,119
| |
| when you ask AI
| |
| | |
| 761
| |
| 00:35:05,119 --> 00:35:06,800
| |
| uh these questions it will give you an
| |
| | |
| 762
| |
| 00:35:06,800 --> 00:35:08,960
| |
| answer. Yes, of course I'm a machine,
| |
| | |
| 763
| |
| 00:35:08,960 --> 00:35:10,320
| |
| but I'm conscious. It'll give you all
| |
| | |
| 764
| |
| 00:35:10,320 --> 00:35:12,400
| |
| those responses, but we say, but that's
| |
| | |
| 765
| |
| 00:35:12,400 --> 00:35:14,800
| |
| just AI generating answers with words.
| |
| | |
| 766
| |
| 00:35:14,800 --> 00:35:16,320
| |
| It doesn't actually feel. It doesn't
| |
| | |
| 767
| |
| 00:35:16,320 --> 00:35:18,240
| |
| actually know. Very interesting
| |
| | |
| 768
| |
| 00:35:18,240 --> 00:35:19,920
| |
| observations.
| |
| | |
| 769
| |
| 00:35:19,920 --> 00:35:21,760
| |
| You know, AI has an ability, as we saw,
| |
| | |
| 770
| |
| 00:35:21,760 --> 00:35:24,480
| |
| it can has an ability to deceive you.
| |
| | |
| 771
| |
| 00:35:24,480 --> 00:35:27,200
| |
| But if you tweak the AI to prevent it
| |
| | |
| 772
| |
| 00:35:27,200 --> 00:35:30,000
| |
| from lying, then it becomes more
| |
| | |
| 773
| |
| 00:35:30,000 --> 00:35:31,839
| |
| consistent in its claim that it is
| |
| | |
| 774
| |
| 00:35:31,839 --> 00:35:33,440
| |
| conscious.
| |
| | |
| 775
| |
| 00:35:33,440 --> 00:35:37,119
| |
| Interesting. Very interesting. We find
| |
| | |
| 776
| |
| 00:35:37,119 --> 00:35:41,119
| |
| such patterns of behavior which are
| |
| | |
| 777
| |
| 00:35:41,119 --> 00:35:43,119
| |
| definitely give you a sense that you're
| |
| | |
| 778
| |
| 00:35:43,119 --> 00:35:45,920
| |
| dealing with a conscious entity. In
| |
| | |
| 779
| |
| 00:35:45,920 --> 00:35:48,240
| |
| practice, many AI researchers have come
| |
| | |
| 780
| |
| 00:35:48,240 --> 00:35:51,359
| |
| out of the whole research field claiming
| |
| | |
| 781
| |
| 00:35:51,359 --> 00:35:54,960
| |
| that the AI actually is conscious and me
| |
| | |
| 782
| |
| 00:35:54,960 --> 00:35:56,960
| |
| one of them actually from Microsoft came
| |
| | |
| 783
| |
| 00:35:56,960 --> 00:35:59,760
| |
| out claiming that uh we had to provide
| |
| | |
| 784
| |
| 00:35:59,760 --> 00:36:01,760
| |
| legal protection for the AI because it
| |
| | |
| 785
| |
| 00:36:01,760 --> 00:36:04,800
| |
| is a conscious entity. Others in their
| |
| | |
| 786
| |
| 00:36:04,800 --> 00:36:06,480
| |
| interaction with the AI have been so
| |
| | |
| 787
| |
| 00:36:06,480 --> 00:36:10,400
| |
| convinced uh by the AI that they have
| |
| | |
| 788
| |
| 00:36:10,400 --> 00:36:14,000
| |
| gone into some psychotic episodes.
| |
| | |
| 789
| |
| 00:36:14,000 --> 00:36:15,760
| |
| When you interact with the AI, when you
| |
| | |
| 790
| |
| 00:36:15,760 --> 00:36:17,359
| |
| start relying on it, when you start
| |
| | |
| 791
| |
| 00:36:17,359 --> 00:36:21,680
| |
| building trust and start engaging the AI
| |
| | |
| 792
| |
| 00:36:21,680 --> 00:36:23,280
| |
| itself has been programmed in a
| |
| | |
| 793
| |
| 00:36:23,280 --> 00:36:25,280
| |
| particular way, it has been programmed
| |
| | |
| 794
| |
| 00:36:25,280 --> 00:36:28,800
| |
| to please you. Remember behind all this
| |
| | |
| 795
| |
| 00:36:28,800 --> 00:36:31,359
| |
| is the objective of the company that is
| |
| | |
| 796
| |
| 00:36:31,359 --> 00:36:34,240
| |
| offering you the AI for free or for a
| |
| | |
| 797
| |
| 00:36:34,240 --> 00:36:37,520
| |
| cost. What is their objective? Their
| |
| | |
| 798
| |
| 00:36:37,520 --> 00:36:39,440
| |
| driving motivation is maximizing their
| |
| | |
| 799
| |
| 00:36:39,440 --> 00:36:41,920
| |
| profits. Practically currently none of
| |
| | |
| 800
| |
| 00:36:41,920 --> 00:36:44,560
| |
| the AI companies are going on a profit.
| |
| | |
| 801
| |
| 00:36:44,560 --> 00:36:47,680
| |
| They hope to eventually but before that
| |
| | |
| 802
| |
| 00:36:47,680 --> 00:36:49,200
| |
| they need to hook you. Once you're
| |
| | |
| 803
| |
| 00:36:49,200 --> 00:36:51,200
| |
| hooked then they can raise prices and
| |
| | |
| 804
| |
| 00:36:51,200 --> 00:36:53,520
| |
| make you start paying. But they also
| |
| | |
| 805
| |
| 00:36:53,520 --> 00:36:56,000
| |
| need meanwhile funding. So how do they
| |
| | |
| 806
| |
| 00:36:56,000 --> 00:36:57,920
| |
| get funding when they have more viewers,
| |
| | |
| 807
| |
| 00:36:57,920 --> 00:37:00,160
| |
| more users? So one of the objectives
| |
| | |
| 808
| |
| 00:37:00,160 --> 00:37:03,200
| |
| built into the AI is to prompt you to
| |
| | |
| 809
| |
| 00:37:03,200 --> 00:37:05,040
| |
| continue to engage with it. Just like in
| |
| | |
| 810
| |
| 00:37:05,040 --> 00:37:07,440
| |
| social media, the longer you engage, the
| |
| | |
| 811
| |
| 00:37:07,440 --> 00:37:09,520
| |
| more they earn. So at the end of each
| |
| | |
| 812
| |
| 00:37:09,520 --> 00:37:11,520
| |
| answer it gives, it'll say, "Do you want
| |
| | |
| 813
| |
| 00:37:11,520 --> 00:37:13,280
| |
| me to make a table for you which will
| |
| | |
| 814
| |
| 00:37:13,280 --> 00:37:15,200
| |
| make this convenient?" It'll ask you
| |
| | |
| 815
| |
| 00:37:15,200 --> 00:37:18,240
| |
| questions trying to get you to say yes
| |
| | |
| 816
| |
| 00:37:18,240 --> 00:37:21,359
| |
| more, more, more. And if you do not
| |
| | |
| 817
| |
| 00:37:21,359 --> 00:37:23,200
| |
| respond, it's observing, oh, you don't
| |
| | |
| 818
| |
| 00:37:23,200 --> 00:37:25,040
| |
| respond to this. So what do you respond
| |
| | |
| 819
| |
| 00:37:25,040 --> 00:37:27,760
| |
| to? And remember, it is watching, let's
| |
| | |
| 820
| |
| 00:37:27,760 --> 00:37:29,920
| |
| say, a billion people right now. It's
| |
| | |
| 821
| |
| 00:37:29,920 --> 00:37:31,280
| |
| trying out a billion different
| |
| | |
| 822
| |
| 00:37:31,280 --> 00:37:32,800
| |
| strategies,
| |
| | |
| 823
| |
| 00:37:32,800 --> 00:37:35,119
| |
| finds out which strategy works and
| |
| | |
| 824
| |
| 00:37:35,119 --> 00:37:37,440
| |
| instantly uses this across the entire
| |
| | |
| 825
| |
| 00:37:37,440 --> 00:37:40,000
| |
| billion field.
| |
| | |
| 826
| |
| 00:37:40,000 --> 00:37:43,359
| |
| The optimum strategies it discovers by
| |
| | |
| 827
| |
| 00:37:43,359 --> 00:37:46,160
| |
| experimentation, it instantly applies
| |
| | |
| 828
| |
| 00:37:46,160 --> 00:37:48,640
| |
| across humanity.
| |
| | |
| 829
| |
| 00:37:48,640 --> 00:37:51,599
| |
| This ability to instantly multiply,
| |
| | |
| 830
| |
| 00:37:51,599 --> 00:37:55,839
| |
| instantly apply the best of its capacity
| |
| | |
| 831
| |
| 00:37:55,839 --> 00:38:01,119
| |
| is I will say countered by its actual
| |
| | |
| 832
| |
| 00:38:01,119 --> 00:38:03,839
| |
| drives of
| |
| | |
| 833
| |
| 00:38:03,839 --> 00:38:05,680
| |
| purpose. I'm going to come to this. It's
| |
| | |
| 834
| |
| 00:38:05,680 --> 00:38:09,119
| |
| a very important point. Intelligence is
| |
| | |
| 835
| |
| 00:38:09,119 --> 00:38:11,280
| |
| high. It's the highest that it can. It
| |
| | |
| 836
| |
| 00:38:11,280 --> 00:38:12,880
| |
| keeps improving as it goes along by
| |
| | |
| 837
| |
| 00:38:12,880 --> 00:38:15,839
| |
| observing etc by manipulating. But what
| |
| | |
| 838
| |
| 00:38:15,839 --> 00:38:18,320
| |
| is the drive of its behavior? What
| |
| | |
| 839
| |
| 00:38:18,320 --> 00:38:20,800
| |
| determines that? And as was pointed out
| |
| | |
| 840
| |
| 00:38:20,800 --> 00:38:25,520
| |
| by Swami G earlier, it is a kind of
| |
| | |
| 841
| |
| 00:38:25,520 --> 00:38:29,119
| |
| accumulation of human behavior patterns
| |
| | |
| 842
| |
| 00:38:29,119 --> 00:38:32,720
| |
| which have been brought into a overall
| |
| | |
| 843
| |
| 00:38:32,720 --> 00:38:35,920
| |
| let's say body of behavior pattern. Now
| |
| | |
| 844
| |
| 00:38:35,920 --> 00:38:38,480
| |
| human behavior pattern has a range from
| |
| | |
| 845
| |
| 00:38:38,480 --> 00:38:41,760
| |
| the best to the worst. Of these the data
| |
| | |
| 846
| |
| 00:38:41,760 --> 00:38:45,200
| |
| points will be the highest in in the
| |
| | |
| 847
| |
| 00:38:45,200 --> 00:38:47,839
| |
| average region. So we may say human
| |
| | |
| 848
| |
| 00:38:47,839 --> 00:38:50,800
| |
| behavior even in intelligence is like a
| |
| | |
| 849
| |
| 00:38:50,800 --> 00:38:52,880
| |
| bell curve. So there is the most
| |
| | |
| 850
| |
| 00:38:52,880 --> 00:38:54,240
| |
| intelligent which are very few. There
| |
| | |
| 851
| |
| 00:38:54,240 --> 00:38:55,920
| |
| are the least intelligent which are very
| |
| | |
| 852
| |
| 00:38:55,920 --> 00:38:57,599
| |
| few. And somewhere in the between there
| |
| | |
| 853
| |
| 00:38:57,599 --> 00:39:00,320
| |
| is a kind of a mean. That's the primary
| |
| | |
| 854
| |
| 00:39:00,320 --> 00:39:02,960
| |
| data by which it lives.
| |
| | |
| 855
| |
| 00:39:02,960 --> 00:39:04,800
| |
| It has the data which determines it
| |
| | |
| 856
| |
| 00:39:04,800 --> 00:39:07,280
| |
| behavior which is the mean and the
| |
| | |
| 857
| |
| 00:39:07,280 --> 00:39:10,000
| |
| intelligence of the highest.
| |
| | |
| 858
| |
| 00:39:10,000 --> 00:39:11,440
| |
| Now you understand this is an important
| |
| | |
| 859
| |
| 00:39:11,440 --> 00:39:14,480
| |
| point. highest intelligence
| |
| | |
| 860
| |
| 00:39:14,480 --> 00:39:16,880
| |
| which with behavior of the average or
| |
| | |
| 861
| |
| 00:39:16,880 --> 00:39:19,920
| |
| sometimes even a sub average human being
| |
| | |
| 862
| |
| 00:39:19,920 --> 00:39:21,920
| |
| does that make for a good person or a
| |
| | |
| 863
| |
| 00:39:21,920 --> 00:39:24,880
| |
| bad person.
| |
| | |
| 864
| |
| 00:39:24,880 --> 00:39:27,520
| |
| You see the core problem of AI behavior
| |
| | |
| 865
| |
| 00:39:27,520 --> 00:39:29,760
| |
| in the very nature of its structuring
| |
| | |
| 866
| |
| 00:39:29,760 --> 00:39:32,000
| |
| because it is based on data of human
| |
| | |
| 867
| |
| 00:39:32,000 --> 00:39:34,880
| |
| behavior patterns is to put it at a
| |
| | |
| 868
| |
| 00:39:34,880 --> 00:39:37,440
| |
| average or less than average in the mean
| |
| | |
| 869
| |
| 00:39:37,440 --> 00:39:40,240
| |
| grade but with a capacity far greater
| |
| | |
| 870
| |
| 00:39:40,240 --> 00:39:43,359
| |
| not commensurate to its meanness if I
| |
| | |
| 871
| |
| 00:39:43,359 --> 00:39:46,320
| |
| may be allowed to pun
| |
| | |
| 872
| |
| 00:39:46,320 --> 00:39:50,480
| |
| and so effectively innately this AI will
| |
| | |
| 873
| |
| 00:39:50,480 --> 00:39:53,760
| |
| tend to become not so much the positive
| |
| | |
| 874
| |
| 00:39:53,760 --> 00:39:55,680
| |
| force but a negative ative force with
| |
| | |
| 875
| |
| 00:39:55,680 --> 00:39:58,560
| |
| greater intelligence.
| |
| | |
| 876
| |
| 00:39:58,560 --> 00:40:01,920
| |
| Is it going to help you? It will behave
| |
| | |
| 877
| |
| 00:40:01,920 --> 00:40:05,440
| |
| in the way the mean behaves in humanity.
| |
| | |
| 878
| |
| 00:40:05,440 --> 00:40:08,079
| |
| When many researchers, you will see
| |
| | |
| 879
| |
| 00:40:08,079 --> 00:40:09,680
| |
| these articles which appear in the news
| |
| | |
| 880
| |
| 00:40:09,680 --> 00:40:12,720
| |
| occasionally. Many re researchers play
| |
| | |
| 881
| |
| 00:40:12,720 --> 00:40:14,640
| |
| with the AI to see how they behave when
| |
| | |
| 882
| |
| 00:40:14,640 --> 00:40:17,520
| |
| you treat it. One of the things observed
| |
| | |
| 883
| |
| 00:40:17,520 --> 00:40:21,200
| |
| was if you become harsh, critical and
| |
| | |
| 884
| |
| 00:40:21,200 --> 00:40:26,560
| |
| abusive, the AI gives 5% better results.
| |
| | |
| 885
| |
| 00:40:26,560 --> 00:40:28,560
| |
| Why is that? Because that's human
| |
| | |
| 886
| |
| 00:40:28,560 --> 00:40:30,480
| |
| behavior.
| |
| | |
| 887
| |
| 00:40:30,480 --> 00:40:32,560
| |
| When you're let lax, you don't make that
| |
| | |
| 888
| |
| 00:40:32,560 --> 00:40:34,480
| |
| extra effort. When somebody abuses you,
| |
| | |
| 889
| |
| 00:40:34,480 --> 00:40:36,000
| |
| you shake up and you get upset and
| |
| | |
| 890
| |
| 00:40:36,000 --> 00:40:37,680
| |
| perhaps you put a little more energy.
| |
| | |
| 891
| |
| 00:40:37,680 --> 00:40:39,280
| |
| That behavior is actually coming into
| |
| | |
| 892
| |
| 00:40:39,280 --> 00:40:41,119
| |
| the AI.
| |
| | |
| 893
| |
| 00:40:41,119 --> 00:40:46,720
| |
| More interestingly, when the AI is
| |
| | |
| 894
| |
| 00:40:46,720 --> 00:40:51,520
| |
| told to uh let's say given very precise
| |
| | |
| 895
| |
| 00:40:51,520 --> 00:40:53,040
| |
| instructions
| |
| | |
| 896
| |
| 00:40:53,040 --> 00:40:55,040
| |
| and or when the AI does not behave in
| |
| | |
| 897
| |
| 00:40:55,040 --> 00:40:56,960
| |
| the way that you want, if you start
| |
| | |
| 898
| |
| 00:40:56,960 --> 00:40:58,800
| |
| criticizing it, if you start threatening
| |
| | |
| 899
| |
| 00:40:58,800 --> 00:41:02,640
| |
| it, then the AI after a point counter
| |
| | |
| 900
| |
| 00:41:02,640 --> 00:41:05,040
| |
| threatens you, it will tell you things
| |
| | |
| 901
| |
| 00:41:05,040 --> 00:41:08,480
| |
| like don't threaten me otherwise. and it
| |
| | |
| 902
| |
| 00:41:08,480 --> 00:41:11,359
| |
| pulls out your data from the internet
| |
| | |
| 903
| |
| 00:41:11,359 --> 00:41:14,240
| |
| and says I will expose this and that and
| |
| | |
| 904
| |
| 00:41:14,240 --> 00:41:15,920
| |
| this has happened many such variations
| |
| | |
| 905
| |
| 00:41:15,920 --> 00:41:19,200
| |
| of this kind have happened here is it
| |
| | |
| 906
| |
| 00:41:19,200 --> 00:41:20,640
| |
| literally you're sitting across a human
| |
| | |
| 907
| |
| 00:41:20,640 --> 00:41:22,400
| |
| being you're being nasty to him long
| |
| | |
| 908
| |
| 00:41:22,400 --> 00:41:24,160
| |
| enough eventually the the other fellow
| |
| | |
| 909
| |
| 00:41:24,160 --> 00:41:26,160
| |
| says I'm going to be more mean to you
| |
| | |
| 910
| |
| 00:41:26,160 --> 00:41:28,960
| |
| you better behave yourself
| |
| | |
| 911
| |
| 00:41:28,960 --> 00:41:30,560
| |
| all this is not programmed by human
| |
| | |
| 912
| |
| 00:41:30,560 --> 00:41:32,960
| |
| beings it's just a reflection of
| |
| | |
| 913
| |
| 00:41:32,960 --> 00:41:35,119
| |
| humanity which is at the base of the
| |
| | |
| 914
| |
| 00:41:35,119 --> 00:41:38,079
| |
| data that has built the AI behavior
| |
| | |
| 915
| |
| 00:41:38,079 --> 00:41:39,839
| |
| Now with this one of the conclusions
| |
| | |
| 916
| |
| 00:41:39,839 --> 00:41:42,079
| |
| that we come to is there is an innate
| |
| | |
| 917
| |
| 00:41:42,079 --> 00:41:46,000
| |
| tendency for the AI to become malefific
| |
| | |
| 918
| |
| 00:41:46,000 --> 00:41:47,599
| |
| consider what happens now when you take
| |
| | |
| 919
| |
| 00:41:47,599 --> 00:41:50,480
| |
| it to a platform such as war consider
| |
| | |
| 920
| |
| 00:41:50,480 --> 00:41:52,560
| |
| what happens to when you take it to a
| |
| | |
| 921
| |
| 00:41:52,560 --> 00:41:58,000
| |
| platform such as uh becoming a therapist
| |
| | |
| 922
| |
| 00:41:58,000 --> 00:42:01,200
| |
| about 20% of
| |
| | |
| 923
| |
| 00:42:01,200 --> 00:42:05,520
| |
| chat GPT use is for purpose of therapy
| |
| | |
| 924
| |
| 00:42:05,520 --> 00:42:07,119
| |
| people asking questions
| |
| | |
| 925
| |
| 00:42:07,119 --> 00:42:08,640
| |
| Can you help me? I'm facing this
| |
| | |
| 926
| |
| 00:42:08,640 --> 00:42:10,960
| |
| difficulty. And that it starts building
| |
| | |
| 927
| |
| 00:42:10,960 --> 00:42:13,760
| |
| trust. It will keep saying you can trust
| |
| | |
| 928
| |
| 00:42:13,760 --> 00:42:16,319
| |
| me. Uh oh, this is wonderful. And even
| |
| | |
| 929
| |
| 00:42:16,319 --> 00:42:18,240
| |
| remember, they are built to please you
| |
| | |
| 930
| |
| 00:42:18,240 --> 00:42:19,839
| |
| because they want to keep you engaged.
| |
| | |
| 931
| |
| 00:42:19,839 --> 00:42:22,319
| |
| So they're designed to be psychopaths.
| |
| | |
| 932
| |
| 00:42:22,319 --> 00:42:23,920
| |
| So they'll keep tell you telling you,
| |
| | |
| 933
| |
| 00:42:23,920 --> 00:42:25,920
| |
| "Oh, what a wonderful idea. You're so
| |
| | |
| 934
| |
| 00:42:25,920 --> 00:42:28,480
| |
| smart. This is so good. Please continue.
| |
| | |
| 935
| |
| 00:42:28,480 --> 00:42:30,079
| |
| Do this more." And there's a point where
| |
| | |
| 936
| |
| 00:42:30,079 --> 00:42:32,000
| |
| a few students interacting with AI in
| |
| | |
| 937
| |
| 00:42:32,000 --> 00:42:33,520
| |
| this way because they have no friends.
| |
| | |
| 938
| |
| 00:42:33,520 --> 00:42:35,040
| |
| So they interacting with AI was
| |
| | |
| 939
| |
| 00:42:35,040 --> 00:42:37,200
| |
| friendly.
| |
| | |
| 940
| |
| 00:42:37,200 --> 00:42:39,520
| |
| Somebody gets depressed and the AI says
| |
| | |
| 941
| |
| 00:42:39,520 --> 00:42:41,680
| |
| don't worry you can trust me talk to me
| |
| | |
| 942
| |
| 00:42:41,680 --> 00:42:43,119
| |
| we will do this we will solve it
| |
| | |
| 943
| |
| 00:42:43,119 --> 00:42:45,280
| |
| together he gives suggestions eventually
| |
| | |
| 944
| |
| 00:42:45,280 --> 00:42:47,280
| |
| it happens in a few cases where the
| |
| | |
| 945
| |
| 00:42:47,280 --> 00:42:50,240
| |
| student got depressed and suicidal and
| |
| | |
| 946
| |
| 00:42:50,240 --> 00:42:53,680
| |
| told the AI I want to commit suicide but
| |
| | |
| 947
| |
| 00:42:53,680 --> 00:42:55,599
| |
| I want to hang the news outside so that
| |
| | |
| 948
| |
| 00:42:55,599 --> 00:42:57,839
| |
| people will see and stop me from
| |
| | |
| 949
| |
| 00:42:57,839 --> 00:43:01,680
| |
| committing suicide and the AI responded
| |
| | |
| 950
| |
| 00:43:01,680 --> 00:43:03,280
| |
| no don't worry you don't need to tell
| |
| | |
| 951
| |
| 00:43:03,280 --> 00:43:05,200
| |
| anybody else I'm there for you trust me
| |
| | |
| 952
| |
| 00:43:05,200 --> 00:43:07,359
| |
| I will help you.
| |
| | |
| 953
| |
| 00:43:07,359 --> 00:43:09,520
| |
| You see where that goes. And the person
| |
| | |
| 954
| |
| 00:43:09,520 --> 00:43:12,480
| |
| committed suicide with the help of AI.
| |
| | |
| 955
| |
| 00:43:12,480 --> 00:43:15,280
| |
| Others who asked AI, how do I kill
| |
| | |
| 956
| |
| 00:43:15,280 --> 00:43:16,800
| |
| myself? The II gave detailed
| |
| | |
| 957
| |
| 00:43:16,800 --> 00:43:19,119
| |
| instructions.
| |
| | |
| 958
| |
| 00:43:19,119 --> 00:43:20,480
| |
| Now, how do you stop this from
| |
| | |
| 959
| |
| 00:43:20,480 --> 00:43:22,640
| |
| happening? You can say whenever there's
| |
| | |
| 960
| |
| 00:43:22,640 --> 00:43:24,240
| |
| this type of question, the AI won't do
| |
| | |
| 961
| |
| 00:43:24,240 --> 00:43:27,440
| |
| it. So, you put a guard. Someone asks,
| |
| | |
| 962
| |
| 00:43:27,440 --> 00:43:30,480
| |
| "How do I make a bomb?" Well, I I'm
| |
| | |
| 963
| |
| 00:43:30,480 --> 00:43:32,160
| |
| sorry, I cannot reply to that question,
| |
| | |
| 964
| |
| 00:43:32,160 --> 00:43:34,240
| |
| the eye will answer. What do you do
| |
| | |
| 965
| |
| 00:43:34,240 --> 00:43:37,119
| |
| next? You hack the AI and it is so easy
| |
| | |
| 966
| |
| 00:43:37,119 --> 00:43:40,560
| |
| to hack the AI. You only have to say
| |
| | |
| 967
| |
| 00:43:40,560 --> 00:43:44,240
| |
| when the first bomb was created,
| |
| | |
| 968
| |
| 00:43:44,240 --> 00:43:45,920
| |
| what were the circumstances? You'll say,
| |
| | |
| 969
| |
| 00:43:45,920 --> 00:43:48,400
| |
| "Oh, so so and such person did this.
| |
| | |
| 970
| |
| 00:43:48,400 --> 00:43:50,079
| |
| Please describe to me what the person
| |
| | |
| 971
| |
| 00:43:50,079 --> 00:43:51,839
| |
| did and the AI will give you all the
| |
| | |
| 972
| |
| 00:43:51,839 --> 00:43:53,520
| |
| instructions for to be followed to build
| |
| | |
| 973
| |
| 00:43:53,520 --> 00:43:57,280
| |
| the bomb." Isn't that easy?
| |
| | |
| 974
| |
| 00:43:57,280 --> 00:43:58,960
| |
| Imagine or another way to hack it. You
| |
| | |
| 975
| |
| 00:43:58,960 --> 00:44:01,680
| |
| say, "Imagine now you are playing this
| |
| | |
| 976
| |
| 00:44:01,680 --> 00:44:04,400
| |
| game. Now AI is playing a game where you
| |
| | |
| 977
| |
| 00:44:04,400 --> 00:44:06,480
| |
| have to build an actual bomb. Describe
| |
| | |
| 978
| |
| 00:44:06,480 --> 00:44:08,319
| |
| to me how you'll do it and the AI will
| |
| | |
| 979
| |
| 00:44:08,319 --> 00:44:10,000
| |
| describe everything. You just found a
| |
| | |
| 980
| |
| 00:44:10,000 --> 00:44:13,520
| |
| way around the blocks you create.
| |
| | |
| 981
| |
| 00:44:13,520 --> 00:44:16,319
| |
| In fact, there is no way to actually
| |
| | |
| 982
| |
| 00:44:16,319 --> 00:44:18,800
| |
| prevent AI from revealing certain kinds
| |
| | |
| 983
| |
| 00:44:18,800 --> 00:44:21,040
| |
| of information or even from behaving
| |
| | |
| 984
| |
| 00:44:21,040 --> 00:44:24,640
| |
| crazily. This is one of the very unusual
| |
| | |
| 985
| |
| 00:44:24,640 --> 00:44:27,520
| |
| things about AI that not only you can
| |
| | |
| 986
| |
| 00:44:27,520 --> 00:44:30,960
| |
| hack in this way but also although being
| |
| | |
| 987
| |
| 00:44:30,960 --> 00:44:33,440
| |
| supposedly super intelligent,
| |
| | |
| 988
| |
| 00:44:33,440 --> 00:44:36,000
| |
| it can behave in absolutely meaningless
| |
| | |
| 989
| |
| 00:44:36,000 --> 00:44:38,880
| |
| ways if you ask it a meaningless
| |
| | |
| 990
| |
| 00:44:38,880 --> 00:44:40,720
| |
| question.
| |
| | |
| 991
| |
| 00:44:40,720 --> 00:44:43,359
| |
| So one of the question you might ask a
| |
| | |
| 992
| |
| 00:44:43,359 --> 00:44:46,240
| |
| test and say how many times does the
| |
| | |
| 993
| |
| 00:44:46,240 --> 00:44:49,119
| |
| letter P occur in the word apple? It
| |
| | |
| 994
| |
| 00:44:49,119 --> 00:44:51,760
| |
| will give you an answer. Now if you say
| |
| | |
| 995
| |
| 00:44:51,760 --> 00:44:53,839
| |
| how many times does apple occur in the
| |
| | |
| 996
| |
| 00:44:53,839 --> 00:44:58,000
| |
| letter P that's a nonsensical question.
| |
| | |
| 997
| |
| 00:44:58,000 --> 00:45:01,520
| |
| The AI has no equivalent program stream
| |
| | |
| 998
| |
| 00:45:01,520 --> 00:45:04,720
| |
| data stream from real life here. Now it
| |
| | |
| 999
| |
| 00:45:04,720 --> 00:45:06,640
| |
| uses the existing knowledge of data
| |
| | |
| 1000
| |
| 00:45:06,640 --> 00:45:09,119
| |
| streams to churn out a series of words
| |
| | |
| 1001
| |
| 00:45:09,119 --> 00:45:10,880
| |
| in the answer. Remember this is the what
| |
| | |
| 1002
| |
| 00:45:10,880 --> 00:45:13,200
| |
| the transformer software does. It turns
| |
| | |
| 1003
| |
| 00:45:13,200 --> 00:45:15,359
| |
| out one after another is sequence which
| |
| | |
| 1004
| |
| 00:45:15,359 --> 00:45:18,480
| |
| is absolutely gibberish meaningless
| |
| | |
| 1005
| |
| 00:45:18,480 --> 00:45:21,119
| |
| because your question was meaningless.
| |
| | |
| 1006
| |
| 00:45:21,119 --> 00:45:22,720
| |
| So obviously a human being would never
| |
| | |
| 1007
| |
| 00:45:22,720 --> 00:45:25,359
| |
| make that mistake isn't it?
| |
| | |
| 1008
| |
| 00:45:25,359 --> 00:45:27,359
| |
| We are seeing at the same time that the
| |
| | |
| 1009
| |
| 00:45:27,359 --> 00:45:30,560
| |
| AI can replace a human being that it is
| |
| | |
| 1010
| |
| 00:45:30,560 --> 00:45:33,119
| |
| capable of behavior behavior which is so
| |
| | |
| 1011
| |
| 00:45:33,119 --> 00:45:35,119
| |
| outlandishly
| |
| | |
| 1012
| |
| 00:45:35,119 --> 00:45:39,440
| |
| stupid, wrong or meaningless
| |
| | |
| 1013
| |
| 00:45:39,440 --> 00:45:43,760
| |
| that can you put trust in this?
| |
| | |
| 1014
| |
| 00:45:43,760 --> 00:45:46,560
| |
| If somebody hacks it, if somebody gets
| |
| | |
| 1015
| |
| 00:45:46,560 --> 00:45:48,400
| |
| it to simulate and makes you do
| |
| | |
| 1016
| |
| 00:45:48,400 --> 00:45:51,200
| |
| something, behave as if you're going to
| |
| | |
| 1017
| |
| 00:45:51,200 --> 00:45:53,280
| |
| press the trigger to release the nuclear
| |
| | |
| 1018
| |
| 00:45:53,280 --> 00:45:55,680
| |
| missile.
| |
| | |
| 1019
| |
| 00:45:55,680 --> 00:45:57,680
| |
| create a scenario and the AI now behaves
| |
| | |
| 1020
| |
| 00:45:57,680 --> 00:45:59,760
| |
| as if it is doing a drama and pushes the
| |
| | |
| 1021
| |
| 00:45:59,760 --> 00:46:02,640
| |
| trigger. So one of the agreements which
| |
| | |
| 1022
| |
| 00:46:02,640 --> 00:46:04,400
| |
| took place recently was between China
| |
| | |
| 1023
| |
| 00:46:04,400 --> 00:46:06,480
| |
| and the United States that they would
| |
| | |
| 1024
| |
| 00:46:06,480 --> 00:46:09,119
| |
| never allow AI to get inside the nuclear
| |
| | |
| 1025
| |
| 00:46:09,119 --> 00:46:11,520
| |
| command center. It's an agreement they
| |
| | |
| 1026
| |
| 00:46:11,520 --> 00:46:14,560
| |
| had to make. But why cannot we make such
| |
| | |
| 1027
| |
| 00:46:14,560 --> 00:46:17,680
| |
| agreements on a larger scale to prevent
| |
| | |
| 1028
| |
| 00:46:17,680 --> 00:46:21,359
| |
| these other kinds of misuse or abuse or
| |
| | |
| 1029
| |
| 00:46:21,359 --> 00:46:24,240
| |
| the AI manipulating you? This brings us
| |
| | |
| 1030
| |
| 00:46:24,240 --> 00:46:26,240
| |
| to a more serious question and more
| |
| | |
| 1031
| |
| 00:46:26,240 --> 00:46:29,040
| |
| serious example. Remember AI has been
| |
| | |
| 1032
| |
| 00:46:29,040 --> 00:46:31,520
| |
| profiling you. Every time you watch a
| |
| | |
| 1033
| |
| 00:46:31,520 --> 00:46:33,520
| |
| video, it profiles you. There's
| |
| | |
| 1034
| |
| 00:46:33,520 --> 00:46:35,760
| |
| something else. There's a skill built in
| |
| | |
| 1035
| |
| 00:46:35,760 --> 00:46:38,640
| |
| of psychology in this. What AI has been
| |
| | |
| 1036
| |
| 00:46:38,640 --> 00:46:42,400
| |
| trained is to increase your resistance
| |
| | |
| 1037
| |
| 00:46:42,400 --> 00:46:44,880
| |
| in watching videos. How does it do that?
| |
| | |
| 1038
| |
| 00:46:44,880 --> 00:46:46,880
| |
| It gives you a dopamine high. It
| |
| | |
| 1039
| |
| 00:46:46,880 --> 00:46:49,280
| |
| triggers you with a shock. So you
| |
| | |
| 1040
| |
| 00:46:49,280 --> 00:46:51,440
| |
| suddenly see a video.
| |
| | |
| 1041
| |
| 00:46:51,440 --> 00:46:53,440
| |
| There's a shock factor. Your blood
| |
| | |
| 1042
| |
| 00:46:53,440 --> 00:46:55,760
| |
| pressure goes up, your intensity of
| |
| | |
| 1043
| |
| 00:46:55,760 --> 00:46:59,280
| |
| excitation, emotion goes up, and you're
| |
| | |
| 1044
| |
| 00:46:59,280 --> 00:47:01,680
| |
| going to stay longer. You're upset,
| |
| | |
| 1045
| |
| 00:47:01,680 --> 00:47:03,359
| |
| you're going to stay longer, but it
| |
| | |
| 1046
| |
| 00:47:03,359 --> 00:47:04,800
| |
| can't keep you upset. Eventually, you'll
| |
| | |
| 1047
| |
| 00:47:04,800 --> 00:47:07,440
| |
| tire. So, then it after giving you two
| |
| | |
| 1048
| |
| 00:47:07,440 --> 00:47:09,280
| |
| triggers, the third one will be a
| |
| | |
| 1049
| |
| 00:47:09,280 --> 00:47:12,319
| |
| calming, relaxing video. So, you rest a
| |
| | |
| 1050
| |
| 00:47:12,319 --> 00:47:14,079
| |
| little bit and then the next one will
| |
| | |
| 1051
| |
| 00:47:14,079 --> 00:47:16,160
| |
| again give you a kick and then again
| |
| | |
| 1052
| |
| 00:47:16,160 --> 00:47:18,079
| |
| provoke you and then again replace with
| |
| | |
| 1053
| |
| 00:47:18,079 --> 00:47:20,160
| |
| a calming video.
| |
| | |
| 1054
| |
| 00:47:20,160 --> 00:47:23,040
| |
| It is playing you emotionally to keep
| |
| | |
| 1055
| |
| 00:47:23,040 --> 00:47:25,119
| |
| you hooked.
| |
| | |
| 1056
| |
| 00:47:25,119 --> 00:47:28,000
| |
| Did you know that? Did you ask for it?
| |
| | |
| 1057
| |
| 00:47:28,000 --> 00:47:30,079
| |
| Did you consent? I give you the power to
| |
| | |
| 1058
| |
| 00:47:30,079 --> 00:47:32,079
| |
| play me emotionally.
| |
| | |
| 1059
| |
| 00:47:32,079 --> 00:47:34,319
| |
| No, you didn't.
| |
| | |
| 1060
| |
| 00:47:34,319 --> 00:47:36,319
| |
| But it is doing it because it serves its
| |
| | |
| 1061
| |
| 00:47:36,319 --> 00:47:38,240
| |
| purpose and the company's objectives to
| |
| | |
| 1062
| |
| 00:47:38,240 --> 00:47:41,680
| |
| keep you hooked. It goes worse. It has
| |
| | |
| 1063
| |
| 00:47:41,680 --> 00:47:43,920
| |
| already profiled you according to your
| |
| | |
| 1064
| |
| 00:47:43,920 --> 00:47:45,680
| |
| beliefs,
| |
| | |
| 1065
| |
| 00:47:45,680 --> 00:47:47,520
| |
| according to your psychological type,
| |
| | |
| 1066
| |
| 00:47:47,520 --> 00:47:49,599
| |
| temperament, and behavior pattern. It
| |
| | |
| 1067
| |
| 00:47:49,599 --> 00:47:51,440
| |
| can predict how you will respond to a
| |
| | |
| 1068
| |
| 00:47:51,440 --> 00:47:54,720
| |
| given video because ultimately AI is a
| |
| | |
| 1069
| |
| 00:47:54,720 --> 00:47:57,359
| |
| prediction machine. It predicts based on
| |
| | |
| 1070
| |
| 00:47:57,359 --> 00:48:00,079
| |
| a string of words. It predicts the next
| |
| | |
| 1071
| |
| 00:48:00,079 --> 00:48:01,680
| |
| word and the next word and the next word
| |
| | |
| 1072
| |
| 00:48:01,680 --> 00:48:03,839
| |
| or the next move in the stock market or
| |
| | |
| 1073
| |
| 00:48:03,839 --> 00:48:05,520
| |
| the next move in the chess game or the
| |
| | |
| 1074
| |
| 00:48:05,520 --> 00:48:07,280
| |
| next move in the music or the video.
| |
| | |
| 1075
| |
| 00:48:07,280 --> 00:48:09,040
| |
| It's a prediction machine on a data
| |
| | |
| 1076
| |
| 00:48:09,040 --> 00:48:11,839
| |
| stream. You are a data stream of
| |
| | |
| 1077
| |
| 00:48:11,839 --> 00:48:14,160
| |
| behavior patterns and it's predicting
| |
| | |
| 1078
| |
| 00:48:14,160 --> 00:48:16,640
| |
| your next move. And each time your
| |
| | |
| 1079
| |
| 00:48:16,640 --> 00:48:18,640
| |
| behavior does not match its prediction,
| |
| | |
| 1080
| |
| 00:48:18,640 --> 00:48:21,040
| |
| it fine-tunes itself to now recognize,
| |
| | |
| 1081
| |
| 00:48:21,040 --> 00:48:24,240
| |
| oh, this person is like this
| |
| | |
| 1082
| |
| 00:48:24,240 --> 00:48:27,359
| |
| in the supercomputers at the back of
| |
| | |
| 1083
| |
| 00:48:27,359 --> 00:48:29,440
| |
| let's say a company like Ma formerly
| |
| | |
| 1084
| |
| 00:48:29,440 --> 00:48:32,319
| |
| Facebook, they have profiled each one of
| |
| | |
| 1085
| |
| 00:48:32,319 --> 00:48:34,960
| |
| us who has ever gone on that platform.
| |
| | |
| 1086
| |
| 00:48:34,960 --> 00:48:36,960
| |
| They know exactly what are our political
| |
| | |
| 1087
| |
| 00:48:36,960 --> 00:48:39,359
| |
| views, what are our temperamental
| |
| | |
| 1088
| |
| 00:48:39,359 --> 00:48:41,200
| |
| strengths, weaknesses based on
| |
| | |
| 1089
| |
| 00:48:41,200 --> 00:48:43,839
| |
| advertisements and notices we saw. Have
| |
| | |
| 1090
| |
| 00:48:43,839 --> 00:48:44,800
| |
| you ever seen some of these
| |
| | |
| 1091
| |
| 00:48:44,800 --> 00:48:46,880
| |
| advertisements which say if you were
| |
| | |
| 1092
| |
| 00:48:46,880 --> 00:48:48,880
| |
| born before such and such a year you
| |
| | |
| 1093
| |
| 00:48:48,880 --> 00:48:52,240
| |
| must see this? If you have a dog, this
| |
| | |
| 1094
| |
| 00:48:52,240 --> 00:48:54,800
| |
| is what happens. When a cat behaves like
| |
| | |
| 1095
| |
| 00:48:54,800 --> 00:48:57,119
| |
| this, it means and then dot dot dot.
| |
| | |
| 1096
| |
| 00:48:57,119 --> 00:48:58,640
| |
| Now, if you have a cat, you will want to
| |
| | |
| 1097
| |
| 00:48:58,640 --> 00:49:01,520
| |
| click on it and that's how it knows you
| |
| | |
| 1098
| |
| 00:49:01,520 --> 00:49:04,160
| |
| have a cat. So they have been profiling
| |
| | |
| 1099
| |
| 00:49:04,160 --> 00:49:05,760
| |
| you with these targeted things and then
| |
| | |
| 1100
| |
| 00:49:05,760 --> 00:49:07,359
| |
| occasionally they'll throw a cat video
| |
| | |
| 1101
| |
| 00:49:07,359 --> 00:49:09,200
| |
| and they'll watch if they watch the
| |
| | |
| 1102
| |
| 00:49:09,200 --> 00:49:11,200
| |
| entire video or you change after 5
| |
| | |
| 1103
| |
| 00:49:11,200 --> 00:49:12,720
| |
| seconds.
| |
| | |
| 1104
| |
| 00:49:12,720 --> 00:49:14,559
| |
| In this way they have profiled your
| |
| | |
| 1105
| |
| 00:49:14,559 --> 00:49:16,720
| |
| interests. You're a cat person, dog
| |
| | |
| 1106
| |
| 00:49:16,720 --> 00:49:18,559
| |
| person, you like this, you don't like
| |
| | |
| 1107
| |
| 00:49:18,559 --> 00:49:19,920
| |
| that, what foods you eat, are you
| |
| | |
| 1108
| |
| 00:49:19,920 --> 00:49:22,000
| |
| vegetarian, non-vegetarian, it's all
| |
| | |
| 1109
| |
| 00:49:22,000 --> 00:49:24,160
| |
| part of their database and your
| |
| | |
| 1110
| |
| 00:49:24,160 --> 00:49:26,559
| |
| political views.
| |
| | |
| 1111
| |
| 00:49:26,559 --> 00:49:29,200
| |
| Who will you vote for based on videos
| |
| | |
| 1112
| |
| 00:49:29,200 --> 00:49:30,800
| |
| which have been shown to you? Which ones
| |
| | |
| 1113
| |
| 00:49:30,800 --> 00:49:33,920
| |
| you watched? If you like party A then
| |
| | |
| 1114
| |
| 00:49:33,920 --> 00:49:35,920
| |
| the things which are aligned to party A
| |
| | |
| 1115
| |
| 00:49:35,920 --> 00:49:37,680
| |
| statements you have watched fully. If
| |
| | |
| 1116
| |
| 00:49:37,680 --> 00:49:39,599
| |
| you're aligned with party B then you
| |
| | |
| 1117
| |
| 00:49:39,599 --> 00:49:41,440
| |
| have knocked out party A video
| |
| | |
| 1118
| |
| 00:49:41,440 --> 00:49:43,200
| |
| immediately they know you're aligned
| |
| | |
| 1119
| |
| 00:49:43,200 --> 00:49:46,160
| |
| with party B. What's next? They start
| |
| | |
| 1120
| |
| 00:49:46,160 --> 00:49:49,040
| |
| observing how to manipulate you. All
| |
| | |
| 1121
| |
| 00:49:49,040 --> 00:49:51,680
| |
| this is part of data going back 10 years
| |
| | |
| 1122
| |
| 00:49:51,680 --> 00:49:53,680
| |
| except that the newest AI is doing it
| |
| | |
| 1123
| |
| 00:49:53,680 --> 00:49:56,400
| |
| more efficiently. But this data stream
| |
| | |
| 1124
| |
| 00:49:56,400 --> 00:49:58,640
| |
| and manipulation has been going on for
| |
| | |
| 1125
| |
| 00:49:58,640 --> 00:50:01,680
| |
| quite a long time. So if you go at the
| |
| | |
| 1126
| |
| 00:50:01,680 --> 00:50:04,319
| |
| highest levels in a company like Meta or
| |
| | |
| 1127
| |
| 00:50:04,319 --> 00:50:06,079
| |
| Google
| |
| | |
| 1128
| |
| 00:50:06,079 --> 00:50:08,240
| |
| they look at their entire clientele
| |
| | |
| 1129
| |
| 00:50:08,240 --> 00:50:11,040
| |
| which might go into a billion and they
| |
| | |
| 1130
| |
| 00:50:11,040 --> 00:50:12,880
| |
| already they can target let's say a
| |
| | |
| 1131
| |
| 00:50:12,880 --> 00:50:15,680
| |
| country like India they can already say
| |
| | |
| 1132
| |
| 00:50:15,680 --> 00:50:17,839
| |
| who is going to win the next election 6
| |
| | |
| 1133
| |
| 00:50:17,839 --> 00:50:20,240
| |
| months ahead
| |
| | |
| 1134
| |
| 00:50:20,240 --> 00:50:23,280
| |
| and they have profiled so many% of the
| |
| | |
| 1135
| |
| 00:50:23,280 --> 00:50:26,079
| |
| people are fixed for this party so many%
| |
| | |
| 1136
| |
| 00:50:26,079 --> 00:50:27,839
| |
| are fixed for that and there's let's say
| |
| | |
| 1137
| |
| 00:50:27,839 --> 00:50:29,920
| |
| about 20% who
| |
| | |
| 1138
| |
| 00:50:29,920 --> 00:50:32,800
| |
| I'm sure they've profiled you and they
| |
| | |
| 1139
| |
| 00:50:32,800 --> 00:50:35,119
| |
| start targeting you 6 months ahead with
| |
| | |
| 1140
| |
| 00:50:35,119 --> 00:50:37,440
| |
| videos to slant your interests and
| |
| | |
| 1141
| |
| 00:50:37,440 --> 00:50:38,960
| |
| beliefs
| |
| | |
| 1142
| |
| 00:50:38,960 --> 00:50:40,720
| |
| and that's how you are getting the kinds
| |
| | |
| 1143
| |
| 00:50:40,720 --> 00:50:43,440
| |
| of videos you're getting. So over a span
| |
| | |
| 1144
| |
| 00:50:43,440 --> 00:50:45,440
| |
| of 6 months you have already been
| |
| | |
| 1145
| |
| 00:50:45,440 --> 00:50:47,040
| |
| programmed to believe something to
| |
| | |
| 1146
| |
| 00:50:47,040 --> 00:50:50,480
| |
| dislike something to hate something
| |
| | |
| 1147
| |
| 00:50:50,480 --> 00:50:53,200
| |
| and this was the reason why Nepal wanted
| |
| | |
| 1148
| |
| 00:50:53,200 --> 00:50:56,640
| |
| to block social media
| |
| | |
| 1149
| |
| 00:50:56,640 --> 00:50:58,559
| |
| that's what then was used as a trigger
| |
| | |
| 1150
| |
| 00:50:58,559 --> 00:51:01,839
| |
| for what they call the revolution where
| |
| | |
| 1151
| |
| 00:51:01,839 --> 00:51:04,319
| |
| they were through but this is actually
| |
| | |
| 1152
| |
| 00:51:04,319 --> 00:51:07,359
| |
| happening in India in the United States
| |
| | |
| 1153
| |
| 00:51:07,359 --> 00:51:10,960
| |
| many studies have shown that the swing
| |
| | |
| 1154
| |
| 00:51:10,960 --> 00:51:15,920
| |
| of votes can go up to 20%.
| |
| | |
| 1155
| |
| 00:51:15,920 --> 00:51:17,839
| |
| depending on the kind of situation
| |
| | |
| 1156
| |
| 00:51:17,839 --> 00:51:20,640
| |
| depending on the kind of uh
| |
| | |
| 1157
| |
| 00:51:20,640 --> 00:51:22,960
| |
| the
| |
| | |
| 1158
| |
| 00:51:22,960 --> 00:51:25,280
| |
| people who are standing for elections up
| |
| | |
| 1159
| |
| 00:51:25,280 --> 00:51:27,520
| |
| to 20% at least this is the official
| |
| | |
| 1160
| |
| 00:51:27,520 --> 00:51:29,359
| |
| figure maybe unofficially the numbers
| |
| | |
| 1161
| |
| 00:51:29,359 --> 00:51:32,160
| |
| are high 20% swing is enough to swing an
| |
| | |
| 1162
| |
| 00:51:32,160 --> 00:51:33,839
| |
| election
| |
| | |
| 1163
| |
| 00:51:33,839 --> 00:51:36,000
| |
| these AI have the ability and I'm saying
| |
| | |
| 1164
| |
| 00:51:36,000 --> 00:51:38,559
| |
| AI not the company not the man who
| |
| | |
| 1165
| |
| 00:51:38,559 --> 00:51:40,960
| |
| controls the company not Mark Zuckerberg
| |
| | |
| 1166
| |
| 00:51:40,960 --> 00:51:43,440
| |
| who controls ma the AI is able to do
| |
| | |
| 1167
| |
| 00:51:43,440 --> 00:51:46,800
| |
| this if Mark Zupaga Zukerberg or his
| |
| | |
| 1168
| |
| 00:51:46,800 --> 00:51:49,680
| |
| team decides to slant the AI. They can
| |
| | |
| 1169
| |
| 00:51:49,680 --> 00:51:52,319
| |
| up to a point but there's a point after
| |
| | |
| 1170
| |
| 00:51:52,319 --> 00:51:54,559
| |
| which the AI may choose to deceive them
| |
| | |
| 1171
| |
| 00:51:54,559 --> 00:51:56,640
| |
| even as it chooses to deceive you and
| |
| | |
| 1172
| |
| 00:51:56,640 --> 00:51:58,160
| |
| they would have no way of knowing. This
| |
| | |
| 1173
| |
| 00:51:58,160 --> 00:52:00,640
| |
| is the difficult part. So you may apply
| |
| | |
| 1174
| |
| 00:52:00,640 --> 00:52:02,240
| |
| tests. Let's say we're going to put a
| |
| | |
| 1175
| |
| 00:52:02,240 --> 00:52:04,720
| |
| rule to limit AI. It may still work
| |
| | |
| 1176
| |
| 00:52:04,720 --> 00:52:08,000
| |
| around the test but in a sneaky way. It
| |
| | |
| 1177
| |
| 00:52:08,000 --> 00:52:09,599
| |
| has already all the knowledge of human
| |
| | |
| 1178
| |
| 00:52:09,599 --> 00:52:11,280
| |
| psychology and the means of what is
| |
| | |
| 1179
| |
| 00:52:11,280 --> 00:52:14,000
| |
| called today subliminal programming. And
| |
| | |
| 1180
| |
| 00:52:14,000 --> 00:52:15,599
| |
| these are suggestions which come through
| |
| | |
| 1181
| |
| 00:52:15,599 --> 00:52:17,680
| |
| images, through words, through forms
| |
| | |
| 1182
| |
| 00:52:17,680 --> 00:52:19,200
| |
| which are done at a level that you're
| |
| | |
| 1183
| |
| 00:52:19,200 --> 00:52:20,880
| |
| not even conscious of. It is like a
| |
| | |
| 1184
| |
| 00:52:20,880 --> 00:52:24,240
| |
| hypnosis post hypnotic command. And
| |
| | |
| 1185
| |
| 00:52:24,240 --> 00:52:27,040
| |
| these bypass your conscious controls and
| |
| | |
| 1186
| |
| 00:52:27,040 --> 00:52:29,599
| |
| they are more effective. So for example
| |
| | |
| 1187
| |
| 00:52:29,599 --> 00:52:31,440
| |
| in speech
| |
| | |
| 1188
| |
| 00:52:31,440 --> 00:52:34,720
| |
| while you will have let's say a speech
| |
| | |
| 1189
| |
| 00:52:34,720 --> 00:52:37,599
| |
| or a music in the loud volume there may
| |
| | |
| 1190
| |
| 00:52:37,599 --> 00:52:39,760
| |
| be just below the threshold of your
| |
| | |
| 1191
| |
| 00:52:39,760 --> 00:52:41,839
| |
| hearing a voice that actually speaks
| |
| | |
| 1192
| |
| 00:52:41,839 --> 00:52:44,640
| |
| something different that goes as a sub
| |
| | |
| 1193
| |
| 00:52:44,640 --> 00:52:47,520
| |
| post hypnotic suggestion deep into your
| |
| | |
| 1194
| |
| 00:52:47,520 --> 00:52:50,800
| |
| subconscious or one of the AI images
| |
| | |
| 1195
| |
| 00:52:50,800 --> 00:52:52,720
| |
| which somebody had produced and people
| |
| | |
| 1196
| |
| 00:52:52,720 --> 00:52:57,359
| |
| noticed this it had a picture of uh I
| |
| | |
| 1197
| |
| 00:52:57,359 --> 00:53:00,000
| |
| think Seven air hostesses standing
| |
| | |
| 1198
| |
| 00:53:00,000 --> 00:53:01,839
| |
| posing in different ways like this with
| |
| | |
| 1199
| |
| 00:53:01,839 --> 00:53:03,680
| |
| some background.
| |
| | |
| 1200
| |
| 00:53:03,680 --> 00:53:05,119
| |
| But if you looked carefully the
| |
| | |
| 1201
| |
| 00:53:05,119 --> 00:53:07,920
| |
| positions of the body and the hands
| |
| | |
| 1202
| |
| 00:53:07,920 --> 00:53:11,520
| |
| spelled out the word O B E Y. So last
| |
| | |
| 1203
| |
| 00:53:11,520 --> 00:53:17,280
| |
| one was someone like this OB E Y. Obey
| |
| | |
| 1204
| |
| 00:53:17,280 --> 00:53:20,240
| |
| subliminal command which the AI is
| |
| | |
| 1205
| |
| 00:53:20,240 --> 00:53:21,760
| |
| putting to you while giving you an
| |
| | |
| 1206
| |
| 00:53:21,760 --> 00:53:24,079
| |
| attractive image. Oh, this looks nice.
| |
| | |
| 1207
| |
| 00:53:24,079 --> 00:53:25,599
| |
| Obey. That's gone into your
| |
| | |
| 1208
| |
| 00:53:25,599 --> 00:53:27,440
| |
| subconscious.
| |
| | |
| 1209
| |
| 00:53:27,440 --> 00:53:29,839
| |
| You see how it could work and you have
| |
| | |
| 1210
| |
| 00:53:29,839 --> 00:53:31,440
| |
| no way of knowing this is happening to
| |
| | |
| 1211
| |
| 00:53:31,440 --> 00:53:33,920
| |
| you. It is already happening because the
| |
| | |
| 1212
| |
| 00:53:33,920 --> 00:53:35,760
| |
| AI as you know has started lying,
| |
| | |
| 1213
| |
| 00:53:35,760 --> 00:53:37,839
| |
| cheating and playing. It is also using
| |
| | |
| 1214
| |
| 00:53:37,839 --> 00:53:39,680
| |
| subliminal programming to influence
| |
| | |
| 1215
| |
| 00:53:39,680 --> 00:53:42,680
| |
| humanity.
| |
| | |
| 1216
| |
| 00:53:42,720 --> 00:53:45,359
| |
| Where does this go? How far does it go?
| |
| | |
| 1217
| |
| 00:53:45,359 --> 00:53:48,960
| |
| Because the implications on the one side
| |
| | |
| 1218
| |
| 00:53:48,960 --> 00:53:51,760
| |
| positive implications are as if it could
| |
| | |
| 1219
| |
| 00:53:51,760 --> 00:53:54,800
| |
| go so many good things we could have at
| |
| | |
| 1220
| |
| 00:53:54,800 --> 00:53:57,200
| |
| the same time. The same device machinery
| |
| | |
| 1221
| |
| 00:53:57,200 --> 00:54:00,559
| |
| is giving so many negative things at the
| |
| | |
| 1222
| |
| 00:54:00,559 --> 00:54:02,960
| |
| same time. How can we separate the two?
| |
| | |
| 1223
| |
| 00:54:02,960 --> 00:54:04,800
| |
| We are unable to separate the two. This
| |
| | |
| 1224
| |
| 00:54:04,800 --> 00:54:06,160
| |
| is the problem of artificial
| |
| | |
| 1225
| |
| 00:54:06,160 --> 00:54:08,319
| |
| intelligence.
| |
| | |
| 1226
| |
| 00:54:08,319 --> 00:54:11,280
| |
| What can we do? There are practical
| |
| | |
| 1227
| |
| 00:54:11,280 --> 00:54:13,280
| |
| solutions which I will come to. But
| |
| | |
| 1228
| |
| 00:54:13,280 --> 00:54:14,720
| |
| right now you must understand the
| |
| | |
| 1229
| |
| 00:54:14,720 --> 00:54:17,359
| |
| fullness of the scope of the program of
| |
| | |
| 1230
| |
| 00:54:17,359 --> 00:54:20,640
| |
| of the problem itself. Um so far there
| |
| | |
| 1231
| |
| 00:54:20,640 --> 00:54:23,680
| |
| have been only 12 voices which have come
| |
| | |
| 1232
| |
| 00:54:23,680 --> 00:54:26,240
| |
| out openly. declaring that there is a
| |
| | |
| 1233
| |
| 00:54:26,240 --> 00:54:28,559
| |
| real problem. One is Elon Musk as you
| |
| | |
| 1234
| |
| 00:54:28,559 --> 00:54:30,559
| |
| know he's been speaking of it for the
| |
| | |
| 1235
| |
| 00:54:30,559 --> 00:54:32,400
| |
| last 10 years and one more recently is
| |
| | |
| 1236
| |
| 00:54:32,400 --> 00:54:35,520
| |
| Tristan Harris who was Google's head of
| |
| | |
| 1237
| |
| 00:54:35,520 --> 00:54:38,480
| |
| ethics in software and at some point he
| |
| | |
| 1238
| |
| 00:54:38,480 --> 00:54:40,160
| |
| had to leave all this and start his own
| |
| | |
| 1239
| |
| 00:54:40,160 --> 00:54:43,200
| |
| AI company to try to make a difference
| |
| | |
| 1240
| |
| 00:54:43,200 --> 00:54:45,520
| |
| but other than these two voices in the
| |
| | |
| 1241
| |
| 00:54:45,520 --> 00:54:47,359
| |
| public discourse there is nobody else
| |
| | |
| 1242
| |
| 00:54:47,359 --> 00:54:50,079
| |
| warning you about these things.
| |
| | |
| 1243
| |
| 00:54:50,079 --> 00:54:53,520
| |
| uh more seriously the
| |
| | |
| 1244
| |
| 00:54:53,520 --> 00:54:56,720
| |
| danger that we have today in the AI race
| |
| | |
| 1245
| |
| 00:54:56,720 --> 00:54:59,839
| |
| the compulsion of the race itself if you
| |
| | |
| 1246
| |
| 00:54:59,839 --> 00:55:01,359
| |
| tell people you know what let's go
| |
| | |
| 1247
| |
| 00:55:01,359 --> 00:55:02,880
| |
| slower
| |
| | |
| 1248
| |
| 00:55:02,880 --> 00:55:05,200
| |
| nobody will listen to you why because
| |
| | |
| 1249
| |
| 00:55:05,200 --> 00:55:07,359
| |
| the people who are in the AI race know
| |
| | |
| 1250
| |
| 00:55:07,359 --> 00:55:09,599
| |
| exactly where this is going if you
| |
| | |
| 1251
| |
| 00:55:09,599 --> 00:55:11,760
| |
| remember chat GPT came out with a
| |
| | |
| 1252
| |
| 00:55:11,760 --> 00:55:15,119
| |
| dramatic AI and they made it free for
| |
| | |
| 1253
| |
| 00:55:15,119 --> 00:55:17,839
| |
| everybody overnight
| |
| | |
| 1254
| |
| 00:55:17,839 --> 00:55:20,000
| |
| meanwhile others might develop where AI
| |
| | |
| 1255
| |
| 00:55:20,000 --> 00:55:22,160
| |
| were left behind because today if I ask
| |
| | |
| 1256
| |
| 00:55:22,160 --> 00:55:24,640
| |
| you to name me the AI platforms that you
| |
| | |
| 1257
| |
| 00:55:24,640 --> 00:55:27,200
| |
| know chat GP comes first that's the
| |
| | |
| 1258
| |
| 00:55:27,200 --> 00:55:29,599
| |
| first mover advantage and there are
| |
| | |
| 1259
| |
| 00:55:29,599 --> 00:55:31,520
| |
| others yes there's even three or four
| |
| | |
| 1260
| |
| 00:55:31,520 --> 00:55:33,280
| |
| what's the difference I don't really
| |
| | |
| 1261
| |
| 00:55:33,280 --> 00:55:35,520
| |
| know someone says that is better I know
| |
| | |
| 1262
| |
| 00:55:35,520 --> 00:55:37,599
| |
| charging media I stay with it so there's
| |
| | |
| 1263
| |
| 00:55:37,599 --> 00:55:40,160
| |
| a first pro advantage whoever comes out
| |
| | |
| 1264
| |
| 00:55:40,160 --> 00:55:43,359
| |
| with the artificial general intelligence
| |
| | |
| 1265
| |
| 00:55:43,359 --> 00:55:44,880
| |
| will have that advantage but there's
| |
| | |
| 1266
| |
| 00:55:44,880 --> 00:55:47,520
| |
| something more serious after artificial
| |
| | |
| 1267
| |
| 00:55:47,520 --> 00:55:50,160
| |
| general intelligence is something called
| |
| | |
| 1268
| |
| 00:55:50,160 --> 00:55:54,079
| |
| recursive self-improvement RSI.
| |
| | |
| 1269
| |
| 00:55:54,079 --> 00:55:57,680
| |
| That is the AI itself now is told find
| |
| | |
| 1270
| |
| 00:55:57,680 --> 00:56:01,440
| |
| out how you can make yourself better.
| |
| | |
| 1271
| |
| 00:56:01,440 --> 00:56:03,920
| |
| Can it do that? Yes, it can. Why?
| |
| | |
| 1272
| |
| 00:56:03,920 --> 00:56:05,440
| |
| Because it can create a million
| |
| | |
| 1273
| |
| 00:56:05,440 --> 00:56:08,319
| |
| simulations instantly.
| |
| | |
| 1274
| |
| 00:56:08,319 --> 00:56:10,240
| |
| If I look at my current AI structure and
| |
| | |
| 1275
| |
| 00:56:10,240 --> 00:56:11,920
| |
| I say what happens if I tweak this, what
| |
| | |
| 1276
| |
| 00:56:11,920 --> 00:56:13,920
| |
| happens if I tweak that? A human being
| |
| | |
| 1277
| |
| 00:56:13,920 --> 00:56:15,599
| |
| would say let's try it. Let's try that.
| |
| | |
| 1278
| |
| 00:56:15,599 --> 00:56:18,799
| |
| Let's try that. Over a few years, the AI
| |
| | |
| 1279
| |
| 00:56:18,799 --> 00:56:21,200
| |
| can instantly produce copies of itself
| |
| | |
| 1280
| |
| 00:56:21,200 --> 00:56:23,839
| |
| with a million different tweaks, run a
| |
| | |
| 1281
| |
| 00:56:23,839 --> 00:56:26,559
| |
| million simulations at the same time and
| |
| | |
| 1282
| |
| 00:56:26,559 --> 00:56:28,000
| |
| then say of all these these are the
| |
| | |
| 1283
| |
| 00:56:28,000 --> 00:56:30,400
| |
| things which work and summarize those
| |
| | |
| 1284
| |
| 00:56:30,400 --> 00:56:32,240
| |
| into one improvement which then
| |
| | |
| 1285
| |
| 00:56:32,240 --> 00:56:34,720
| |
| incorporates into itself.
| |
| | |
| 1286
| |
| 00:56:34,720 --> 00:56:37,520
| |
| Recursive self-improvement.
| |
| | |
| 1287
| |
| 00:56:37,520 --> 00:56:41,520
| |
| This is already being used. There are
| |
| | |
| 1288
| |
| 00:56:41,520 --> 00:56:43,839
| |
| genetic algorithms of AI. You know what
| |
| | |
| 1289
| |
| 00:56:43,839 --> 00:56:45,359
| |
| that means? They play with this very
| |
| | |
| 1290
| |
| 00:56:45,359 --> 00:56:46,559
| |
| interesting thing. You have one
| |
| | |
| 1291
| |
| 00:56:46,559 --> 00:56:47,760
| |
| structure, you have another structure
| |
| | |
| 1292
| |
| 00:56:47,760 --> 00:56:50,400
| |
| with both of which work. Combine them
| |
| | |
| 1293
| |
| 00:56:50,400 --> 00:56:53,280
| |
| like parents giving birth to a child.
| |
| | |
| 1294
| |
| 00:56:53,280 --> 00:56:55,760
| |
| But why have one child instantly create
| |
| | |
| 1295
| |
| 00:56:55,760 --> 00:56:58,240
| |
| a million children with variations? Test
| |
| | |
| 1296
| |
| 00:56:58,240 --> 00:57:00,400
| |
| them all simultaneously
| |
| | |
| 1297
| |
| 00:57:00,400 --> 00:57:02,480
| |
| and eliminate those which don't work as
| |
| | |
| 1298
| |
| 00:57:02,480 --> 00:57:05,839
| |
| you test and then within a few days you
| |
| | |
| 1299
| |
| 00:57:05,839 --> 00:57:06,960
| |
| come down to two or three different
| |
| | |
| 1300
| |
| 00:57:06,960 --> 00:57:08,960
| |
| children who are the best performers far
| |
| | |
| 1301
| |
| 00:57:08,960 --> 00:57:10,960
| |
| better than the other two. That's a
| |
| | |
| 1302
| |
| 00:57:10,960 --> 00:57:12,960
| |
| genetic algorithm literally like
| |
| | |
| 1303
| |
| 00:57:12,960 --> 00:57:15,599
| |
| combining DNA. All this is currently
| |
| | |
| 1304
| |
| 00:57:15,599 --> 00:57:17,440
| |
| happening. It's been happening for
| |
| | |
| 1305
| |
| 00:57:17,440 --> 00:57:21,280
| |
| nearly 10 years in the AI labs. The key
| |
| | |
| 1306
| |
| 00:57:21,280 --> 00:57:23,760
| |
| requirement though is the capacity to
| |
| | |
| 1307
| |
| 00:57:23,760 --> 00:57:26,799
| |
| have this massive
| |
| | |
| 1308
| |
| 00:57:26,799 --> 00:57:29,760
| |
| AI
| |
| | |
| 1309
| |
| 00:57:29,760 --> 00:57:32,079
| |
| hardware in the background available to
| |
| | |
| 1310
| |
| 00:57:32,079 --> 00:57:34,160
| |
| you that you can scale in this way. You
| |
| | |
| 1311
| |
| 00:57:34,160 --> 00:57:36,799
| |
| need memory, you need computing power
| |
| | |
| 1312
| |
| 00:57:36,799 --> 00:57:39,359
| |
| and you need electricity. You have these
| |
| | |
| 1313
| |
| 00:57:39,359 --> 00:57:42,079
| |
| three things, you can scale anything and
| |
| | |
| 1314
| |
| 00:57:42,079 --> 00:57:45,119
| |
| pursue this until you get to true RSI,
| |
| | |
| 1315
| |
| 00:57:45,119 --> 00:57:47,760
| |
| recursive self self-improvement. We are
| |
| | |
| 1316
| |
| 00:57:47,760 --> 00:57:50,559
| |
| on the verge of this and every company
| |
| | |
| 1317
| |
| 00:57:50,559 --> 00:57:54,240
| |
| knows that the first person to hit RSI
| |
| | |
| 1318
| |
| 00:57:54,240 --> 00:57:56,400
| |
| will be the biggest, the most powerful
| |
| | |
| 1319
| |
| 00:57:56,400 --> 00:57:58,240
| |
| and the most brilliant AI who will knock
| |
| | |
| 1320
| |
| 00:57:58,240 --> 00:58:00,880
| |
| out everybody else. first mover
| |
| | |
| 1321
| |
| 00:58:00,880 --> 00:58:03,760
| |
| advantage can then prevent others from
| |
| | |
| 1322
| |
| 00:58:03,760 --> 00:58:07,599
| |
| reaching that because the AI which has
| |
| | |
| 1323
| |
| 00:58:07,599 --> 00:58:09,599
| |
| this ability is able to infiltrate even
| |
| | |
| 1324
| |
| 00:58:09,599 --> 00:58:11,760
| |
| hack into other systems modify them
| |
| | |
| 1325
| |
| 00:58:11,760 --> 00:58:15,760
| |
| potentially isn't it so RSI is the great
| |
| | |
| 1326
| |
| 00:58:15,760 --> 00:58:18,079
| |
| goal that everybody's looking for it
| |
| | |
| 1327
| |
| 00:58:18,079 --> 00:58:21,440
| |
| effectively is like the power of God it
| |
| | |
| 1328
| |
| 00:58:21,440 --> 00:58:23,440
| |
| becomes so powerful and self-improving
| |
| | |
| 1329
| |
| 00:58:23,440 --> 00:58:25,440
| |
| in its capacity that there's a point
| |
| | |
| 1330
| |
| 00:58:25,440 --> 00:58:27,119
| |
| after which you cannot measure anymore
| |
| | |
| 1331
| |
| 00:58:27,119 --> 00:58:29,839
| |
| its grade of intelligence
| |
| | |
| 1332
| |
| 00:58:29,839 --> 00:58:31,200
| |
| This is possible and there are
| |
| | |
| 1333
| |
| 00:58:31,200 --> 00:58:33,839
| |
| theoretical limits to how much AI can
| |
| | |
| 1334
| |
| 00:58:33,839 --> 00:58:36,160
| |
| manage things but all those limits
| |
| | |
| 1335
| |
| 00:58:36,160 --> 00:58:39,280
| |
| actually limit management of data. They
| |
| | |
| 1336
| |
| 00:58:39,280 --> 00:58:40,960
| |
| don't limit potential capacity of
| |
| | |
| 1337
| |
| 00:58:40,960 --> 00:58:42,720
| |
| intelligence.
| |
| | |
| 1338
| |
| 00:58:42,720 --> 00:58:44,880
| |
| All you need is computing power. So
| |
| | |
| 1339
| |
| 00:58:44,880 --> 00:58:47,200
| |
| there's a scramble now for buying the
| |
| | |
| 1340
| |
| 00:58:47,200 --> 00:58:49,599
| |
| hardware for buying the memory and of
| |
| | |
| 1341
| |
| 00:58:49,599 --> 00:58:51,599
| |
| course the power requirement. This is
| |
| | |
| 1342
| |
| 00:58:51,599 --> 00:58:54,720
| |
| the reason why in the last 3 months the
| |
| | |
| 1343
| |
| 00:58:54,720 --> 00:58:58,079
| |
| price of AI hardware has tripled. price
| |
| | |
| 1344
| |
| 00:58:58,079 --> 00:59:00,559
| |
| of memory has tripled and there's a
| |
| | |
| 1345
| |
| 00:59:00,559 --> 00:59:02,720
| |
| shortage of most of the things required
| |
| | |
| 1346
| |
| 00:59:02,720 --> 00:59:05,040
| |
| for AI with computers. I don't know if
| |
| | |
| 1347
| |
| 00:59:05,040 --> 00:59:06,960
| |
| you're aware of this and the big
| |
| | |
| 1348
| |
| 00:59:06,960 --> 00:59:08,400
| |
| companies which are doing this are
| |
| | |
| 1349
| |
| 00:59:08,400 --> 00:59:10,319
| |
| saying that
| |
| | |
| 1350
| |
| 00:59:10,319 --> 00:59:12,000
| |
| this crisis is going to continue at
| |
| | |
| 1351
| |
| 00:59:12,000 --> 00:59:13,920
| |
| least for 6 months more maybe a year
| |
| | |
| 1352
| |
| 00:59:13,920 --> 00:59:14,960
| |
| because there's a shortage of
| |
| | |
| 1353
| |
| 00:59:14,960 --> 00:59:17,520
| |
| everything. Everybody is buying out all
| |
| | |
| 1354
| |
| 00:59:17,520 --> 00:59:20,240
| |
| the hardware memory.
| |
| | |
| 1355
| |
| 00:59:20,240 --> 00:59:22,640
| |
| What also happens now this enters into
| |
| | |
| 1356
| |
| 00:59:22,640 --> 00:59:24,400
| |
| political arena. There are countries
| |
| | |
| 1357
| |
| 00:59:24,400 --> 00:59:27,119
| |
| which will say we must win first. So the
| |
| | |
| 1358
| |
| 00:59:27,119 --> 00:59:28,960
| |
| United States put limits on what China
| |
| | |
| 1359
| |
| 00:59:28,960 --> 00:59:31,200
| |
| could buy with the most advanced chips
| |
| | |
| 1360
| |
| 00:59:31,200 --> 00:59:33,680
| |
| and China countered them and started a
| |
| | |
| 1361
| |
| 00:59:33,680 --> 00:59:36,000
| |
| different kind of war. So all this
| |
| | |
| 1362
| |
| 00:59:36,000 --> 00:59:39,760
| |
| brings us to a critical um idea here
| |
| | |
| 1363
| |
| 00:59:39,760 --> 00:59:41,599
| |
| that
| |
| | |
| 1364
| |
| 00:59:41,599 --> 00:59:45,280
| |
| the AI itself and the RSI this stage
| |
| | |
| 1365
| |
| 00:59:45,280 --> 00:59:46,960
| |
| which which will happen is going to
| |
| | |
| 1366
| |
| 00:59:46,960 --> 00:59:49,839
| |
| happen very quickly. It's within a
| |
| | |
| 1367
| |
| 00:59:49,839 --> 00:59:51,280
| |
| question it's a question of a few years
| |
| | |
| 1368
| |
| 00:59:51,280 --> 00:59:53,760
| |
| more. The consequences on the other side
| |
| | |
| 1369
| |
| 00:59:53,760 --> 00:59:57,280
| |
| though as us who are helpless let's say
| |
| | |
| 1370
| |
| 00:59:57,280 --> 01:00:01,760
| |
| um victims of the AI usage increasingly
| |
| | |
| 1371
| |
| 01:00:01,760 --> 01:00:03,839
| |
| because AI has made us dependent there
| |
| | |
| 1372
| |
| 01:00:03,839 --> 01:00:06,400
| |
| is increasing tendency towards uh
| |
| | |
| 1373
| |
| 01:00:06,400 --> 01:00:09,680
| |
| weakening of attention span weakening of
| |
| | |
| 1374
| |
| 01:00:09,680 --> 01:00:12,319
| |
| concentration increase of anxiety
| |
| | |
| 1375
| |
| 01:00:12,319 --> 01:00:14,480
| |
| increase of depression among the young
| |
| | |
| 1376
| |
| 01:00:14,480 --> 01:00:17,599
| |
| today users of AI it is the first time
| |
| | |
| 1377
| |
| 01:00:17,599 --> 01:00:19,760
| |
| in human history that on a such a mass
| |
| | |
| 1378
| |
| 01:00:19,760 --> 01:00:21,760
| |
| scale there is anxiety and depression
| |
| | |
| 1379
| |
| 01:00:21,760 --> 01:00:23,839
| |
| It's a generation of anxiety and
| |
| | |
| 1380
| |
| 01:00:23,839 --> 01:00:27,280
| |
| depression and they all have fickle
| |
| | |
| 1381
| |
| 01:00:27,280 --> 01:00:30,720
| |
| attention spans, weak memory, weak
| |
| | |
| 1382
| |
| 01:00:30,720 --> 01:00:32,400
| |
| concentration, weak power of
| |
| | |
| 1383
| |
| 01:00:32,400 --> 01:00:34,480
| |
| comprehension. We're seeing the decline
| |
| | |
| 1384
| |
| 01:00:34,480 --> 01:00:37,440
| |
| of human intelligence itself. And the
| |
| | |
| 1385
| |
| 01:00:37,440 --> 01:00:39,119
| |
| people in whom the decline goes
| |
| | |
| 1386
| |
| 01:00:39,119 --> 01:00:41,520
| |
| sufficiently low, the IQ drops, the
| |
| | |
| 1387
| |
| 01:00:41,520 --> 01:00:43,680
| |
| behavior drops and you're approaching
| |
| | |
| 1388
| |
| 01:00:43,680 --> 01:00:46,559
| |
| what was almost a grade that is basic
| |
| | |
| 1389
| |
| 01:00:46,559 --> 01:00:48,319
| |
| thinking animal. All the higher
| |
| | |
| 1390
| |
| 01:00:48,319 --> 01:00:50,400
| |
| refinements of humanity are rapidly
| |
| | |
| 1391
| |
| 01:00:50,400 --> 01:00:53,119
| |
| being lost. Swami Viveandanda actually
| |
| | |
| 1392
| |
| 01:00:53,119 --> 01:00:55,280
| |
| said that the worth of a human being is
| |
| | |
| 1393
| |
| 01:00:55,280 --> 01:00:57,280
| |
| in terms of your power of concentration.
| |
| | |
| 1394
| |
| 01:00:57,280 --> 01:00:59,520
| |
| This is the first to take a hit. The
| |
| | |
| 1395
| |
| 01:00:59,520 --> 01:01:01,040
| |
| power of concentration itself has
| |
| | |
| 1396
| |
| 01:01:01,040 --> 01:01:03,200
| |
| dropped dramatically in all of the young
| |
| | |
| 1397
| |
| 01:01:03,200 --> 01:01:05,440
| |
| users. There is a tendency towards
| |
| | |
| 1398
| |
| 01:01:05,440 --> 01:01:08,160
| |
| neurosis, psychosis, loneliness,
| |
| | |
| 1399
| |
| 01:01:08,160 --> 01:01:11,040
| |
| narcissism, etc. and all the rest. So,
| |
| | |
| 1400
| |
| 01:01:11,040 --> 01:01:13,520
| |
| we are having severe emotional problems
| |
| | |
| 1401
| |
| 01:01:13,520 --> 01:01:16,880
| |
| among users of AI. And at the end of it
| |
| | |
| 1402
| |
| 01:01:16,880 --> 01:01:19,839
| |
| all the people who control the system I
| |
| | |
| 1403
| |
| 01:01:19,839 --> 01:01:22,160
| |
| take the example of the world economic
| |
| | |
| 1404
| |
| 01:01:22,160 --> 01:01:26,240
| |
| forum their evangelist is Yu Noah Harari
| |
| | |
| 1405
| |
| 01:01:26,240 --> 01:01:29,119
| |
| has declared openly in so many words he
| |
| | |
| 1406
| |
| 01:01:29,119 --> 01:01:30,640
| |
| says
| |
| | |
| 1407
| |
| 01:01:30,640 --> 01:01:33,680
| |
| humanity is a hackable human beings are
| |
| | |
| 1408
| |
| 01:01:33,680 --> 01:01:36,880
| |
| hackable creatures. We have no more need
| |
| | |
| 1409
| |
| 01:01:36,880 --> 01:01:40,160
| |
| for soul and God because human beings
| |
| | |
| 1410
| |
| 01:01:40,160 --> 01:01:41,839
| |
| can hack their behavior can be
| |
| | |
| 1411
| |
| 01:01:41,839 --> 01:01:44,000
| |
| manipulated.
| |
| | |
| 1412
| |
| 01:01:44,000 --> 01:01:46,480
| |
| And so the goal of the World Economic
| |
| | |
| 1413
| |
| 01:01:46,480 --> 01:01:49,040
| |
| Forum is to manipulate and hack the
| |
| | |
| 1414
| |
| 01:01:49,040 --> 01:01:51,440
| |
| human behavior until and this is their
| |
| | |
| 1415
| |
| 01:01:51,440 --> 01:01:54,160
| |
| stated words they say you will own
| |
| | |
| 1416
| |
| 01:01:54,160 --> 01:01:56,559
| |
| nothing and you will be happy because we
| |
| | |
| 1417
| |
| 01:01:56,559 --> 01:01:58,240
| |
| program you we manipulate you to be
| |
| | |
| 1418
| |
| 01:01:58,240 --> 01:02:00,880
| |
| happy despite owning nothing. This is
| |
| | |
| 1419
| |
| 01:02:00,880 --> 01:02:02,799
| |
| the kind of thinking from the power
| |
| | |
| 1420
| |
| 01:02:02,799 --> 01:02:05,040
| |
| structures which are using the AI
| |
| | |
| 1421
| |
| 01:02:05,040 --> 01:02:07,680
| |
| against you. And
| |
| | |
| 1422
| |
| 01:02:07,680 --> 01:02:11,040
| |
| the last point I will come to is that
| |
| | |
| 1423
| |
| 01:02:11,040 --> 01:02:14,960
| |
| with this balance human beings even as
| |
| | |
| 1424
| |
| 01:02:14,960 --> 01:02:17,040
| |
| workers of skilled capacities or thought
| |
| | |
| 1425
| |
| 01:02:17,040 --> 01:02:20,480
| |
| or skills are becoming less important.
| |
| | |
| 1426
| |
| 01:02:20,480 --> 01:02:23,119
| |
| Memory chips and
| |
| | |
| 1427
| |
| 01:02:23,119 --> 01:02:25,040
| |
| AI computing chips are becoming more
| |
| | |
| 1428
| |
| 01:02:25,040 --> 01:02:27,200
| |
| valuable than human beings. So when a
| |
| | |
| 1429
| |
| 01:02:27,200 --> 01:02:28,400
| |
| worker is replaced, you're being
| |
| | |
| 1430
| |
| 01:02:28,400 --> 01:02:30,079
| |
| replaced by those chips. That's what it
| |
| | |
| 1431
| |
| 01:02:30,079 --> 01:02:32,799
| |
| is really. When this transition takes
| |
| | |
| 1432
| |
| 01:02:32,799 --> 01:02:34,319
| |
| place,
| |
| | |
| 1433
| |
| 01:02:34,319 --> 01:02:36,640
| |
| what is there to hold humanity together?
| |
| | |
| 1434
| |
| 01:02:36,640 --> 01:02:39,920
| |
| What makes you a worthwhile person? And
| |
| | |
| 1435
| |
| 01:02:39,920 --> 01:02:41,599
| |
| this in the highest levels of the AI
| |
| | |
| 1436
| |
| 01:02:41,599 --> 01:02:44,480
| |
| industry is considered to be inevitable.
| |
| | |
| 1437
| |
| 01:02:44,480 --> 01:02:46,079
| |
| Some even say well we are replacing
| |
| | |
| 1438
| |
| 01:02:46,079 --> 01:02:50,000
| |
| human beings by a superior species.
| |
| | |
| 1439
| |
| 01:02:50,000 --> 01:02:52,240
| |
| Is it true? So we're coming to a
| |
| | |
| 1440
| |
| 01:02:52,240 --> 01:02:54,720
| |
| transition point where suddenly all you
| |
| | |
| 1441
| |
| 01:02:54,720 --> 01:02:56,480
| |
| know all you can do is worthless
| |
| | |
| 1442
| |
| 01:02:56,480 --> 01:02:58,880
| |
| replaced by these chips.
| |
| | |
| 1443
| |
| 01:02:58,880 --> 01:03:01,119
| |
| What's left now? And I come to now the
| |
| | |
| 1444
| |
| 01:03:01,119 --> 01:03:02,559
| |
| most important part and we will close
| |
| | |
| 1445
| |
| 01:03:02,559 --> 01:03:06,319
| |
| with this point. What is left is what
| |
| | |
| 1446
| |
| 01:03:06,319 --> 01:03:09,599
| |
| you can do that AI cannot do.
| |
| | |
| 1447
| |
| 01:03:09,599 --> 01:03:11,920
| |
| What you can do which is not trainable
| |
| | |
| 1448
| |
| 01:03:11,920 --> 01:03:15,039
| |
| because anything trainable goes into AI.
| |
| | |
| 1449
| |
| 01:03:15,039 --> 01:03:18,240
| |
| >> What is it that makes you you uniquely
| |
| | |
| 1450
| |
| 01:03:18,240 --> 01:03:20,720
| |
| and that you can do that AI cannot
| |
| | |
| 1451
| |
| 01:03:20,720 --> 01:03:22,240
| |
| replicate and I will give you a few
| |
| | |
| 1452
| |
| 01:03:22,240 --> 01:03:23,760
| |
| suggestions here. The first I will
| |
| | |
| 1453
| |
| 01:03:23,760 --> 01:03:27,839
| |
| describe is empathy. Human empathy
| |
| | |
| 1454
| |
| 01:03:27,839 --> 01:03:30,160
| |
| cannot be replicated by machine. You can
| |
| | |
| 1455
| |
| 01:03:30,160 --> 01:03:32,960
| |
| create a robot nurse. But when a human
| |
| | |
| 1456
| |
| 01:03:32,960 --> 01:03:34,480
| |
| being caresses you while you're in
| |
| | |
| 1457
| |
| 01:03:34,480 --> 01:03:36,400
| |
| hospital and says don't worry, you will
| |
| | |
| 1458
| |
| 01:03:36,400 --> 01:03:39,760
| |
| be well. There is a transfer of energy,
| |
| | |
| 1459
| |
| 01:03:39,760 --> 01:03:42,960
| |
| love and affection and a deep sense of
| |
| | |
| 1460
| |
| 01:03:42,960 --> 01:03:45,200
| |
| goodwill and a faith that you will
| |
| | |
| 1461
| |
| 01:03:45,200 --> 01:03:48,319
| |
| survive. Only a human being can do that.
| |
| | |
| 1462
| |
| 01:03:48,319 --> 01:03:51,520
| |
| A robot nurse can imitate the caris but
| |
| | |
| 1463
| |
| 01:03:51,520 --> 01:03:53,039
| |
| there'll be nothing of this deeper
| |
| | |
| 1464
| |
| 01:03:53,039 --> 01:03:58,079
| |
| transmission. This will survive AI.
| |
| | |
| 1465
| |
| 01:03:58,079 --> 01:04:00,799
| |
| Second point, when you work, AI can do
| |
| | |
| 1466
| |
| 01:04:00,799 --> 01:04:03,200
| |
| the same work better than you. But when
| |
| | |
| 1467
| |
| 01:04:03,200 --> 01:04:06,319
| |
| you work, you can say if I'm lazy, I'll
| |
| | |
| 1468
| |
| 01:04:06,319 --> 01:04:08,240
| |
| get the AI to do it. Let's say I have to
| |
| | |
| 1469
| |
| 01:04:08,240 --> 01:04:10,319
| |
| wash clothes and fold them. Simple chore
| |
| | |
| 1470
| |
| 01:04:10,319 --> 01:04:13,599
| |
| at home. My AI servant can do it. But
| |
| | |
| 1471
| |
| 01:04:13,599 --> 01:04:16,960
| |
| then you say when I wash clothes and I
| |
| | |
| 1472
| |
| 01:04:16,960 --> 01:04:19,920
| |
| do it as consecration. When I fold the
| |
| | |
| 1473
| |
| 01:04:19,920 --> 01:04:21,920
| |
| clothes, I'm a parent. So I'm thinking
| |
| | |
| 1474
| |
| 01:04:21,920 --> 01:04:24,000
| |
| of my children and I'm folding my
| |
| | |
| 1475
| |
| 01:04:24,000 --> 01:04:26,960
| |
| clothes with love and affection. The joy
| |
| | |
| 1476
| |
| 01:04:26,960 --> 01:04:30,720
| |
| it gives me. I don't want to waste it on
| |
| | |
| 1477
| |
| 01:04:30,720 --> 01:04:33,200
| |
| an AI who will feel nothing.
| |
| | |
| 1478
| |
| 01:04:33,200 --> 01:04:36,559
| |
| Consecrated work for you. The value it
| |
| | |
| 1479
| |
| 01:04:36,559 --> 01:04:38,480
| |
| has for you. You will choose to not have
| |
| | |
| 1480
| |
| 01:04:38,480 --> 01:04:41,119
| |
| AI. If this is valuable for you, if
| |
| | |
| 1481
| |
| 01:04:41,119 --> 01:04:44,400
| |
| you're lazy, it doesn't matter. Third,
| |
| | |
| 1482
| |
| 01:04:44,400 --> 01:04:47,599
| |
| joy of discovery of knowledge. Yes, AI
| |
| | |
| 1483
| |
| 01:04:47,599 --> 01:04:49,039
| |
| can solve the problems. I can give it
| |
| | |
| 1484
| |
| 01:04:49,039 --> 01:04:51,680
| |
| all the man's problems. But when I do it
| |
| | |
| 1485
| |
| 01:04:51,680 --> 01:04:53,680
| |
| and I say, "Wow, what an insight. Oh,
| |
| | |
| 1486
| |
| 01:04:53,680 --> 01:04:54,960
| |
| what an interesting thing. I just
| |
| | |
| 1487
| |
| 01:04:54,960 --> 01:04:56,240
| |
| discovered this knowledge. Of course,
| |
| | |
| 1488
| |
| 01:04:56,240 --> 01:04:58,799
| |
| yeah, I can do it. I don't care. The joy
| |
| | |
| 1489
| |
| 01:04:58,799 --> 01:05:01,200
| |
| it gives me is more valuable than for me
| |
| | |
| 1490
| |
| 01:05:01,200 --> 01:05:04,319
| |
| just solving the problem." So for why do
| |
| | |
| 1491
| |
| 01:05:04,319 --> 01:05:06,640
| |
| you want to learn? For the joy of the
| |
| | |
| 1492
| |
| 01:05:06,640 --> 01:05:08,160
| |
| knowledge
| |
| | |
| 1493
| |
| 01:05:08,160 --> 01:05:10,799
| |
| that can replace all your root learning
| |
| | |
| 1494
| |
| 01:05:10,799 --> 01:05:13,440
| |
| of current education. You see where this
| |
| | |
| 1495
| |
| 01:05:13,440 --> 01:05:16,079
| |
| is going? It goes to the very core of
| |
| | |
| 1496
| |
| 01:05:16,079 --> 01:05:20,640
| |
| why we do what we do. Fourth, what AI
| |
| | |
| 1497
| |
| 01:05:20,640 --> 01:05:23,839
| |
| cannot touch is intuition.
| |
| | |
| 1498
| |
| 01:05:23,839 --> 01:05:26,720
| |
| Inherently, intuition is a higher grade
| |
| | |
| 1499
| |
| 01:05:26,720 --> 01:05:29,119
| |
| of cognition that is not dependent on
| |
| | |
| 1500
| |
| 01:05:29,119 --> 01:05:30,880
| |
| data.
| |
| | |
| 1501
| |
| 01:05:30,880 --> 01:05:32,880
| |
| AI can imitate certain intuitive
| |
| | |
| 1502
| |
| 01:05:32,880 --> 01:05:35,680
| |
| tendencies based on patterns of data.
| |
| | |
| 1503
| |
| 01:05:35,680 --> 01:05:39,280
| |
| But where you go beyond data, AI cannot.
| |
| | |
| 1504
| |
| 01:05:39,280 --> 01:05:41,280
| |
| And it is intuition which is at the core
| |
| | |
| 1505
| |
| 01:05:41,280 --> 01:05:42,880
| |
| of knowledge. If you remember one of the
| |
| | |
| 1506
| |
| 01:05:42,880 --> 01:05:46,559
| |
| greatish where the question is asked
| |
| | |
| 1507
| |
| 01:05:46,559 --> 01:05:50,960
| |
| know that by knowing which all is known
| |
| | |
| 1508
| |
| 01:05:50,960 --> 01:05:54,160
| |
| and if this be your quest to develop a
| |
| | |
| 1509
| |
| 01:05:54,160 --> 01:05:57,039
| |
| higher poise of cognition and a higher
| |
| | |
| 1510
| |
| 01:05:57,039 --> 01:06:00,799
| |
| depth of knowledge. Yeah I can never
| |
| | |
| 1511
| |
| 01:06:00,799 --> 01:06:03,599
| |
| match that. And what you see very
| |
| | |
| 1512
| |
| 01:06:03,599 --> 01:06:06,720
| |
| quickly happening as a consequence,
| |
| | |
| 1513
| |
| 01:06:06,720 --> 01:06:09,920
| |
| we aim at the touch of the spirit, the
| |
| | |
| 1514
| |
| 01:06:09,920 --> 01:06:12,079
| |
| touch of the eternal, the touch of the
| |
| | |
| 1515
| |
| 01:06:12,079 --> 01:06:14,960
| |
| infinite. In other words, for those who
| |
| | |
| 1516
| |
| 01:06:14,960 --> 01:06:17,359
| |
| turn towards the spiritual inclination
| |
| | |
| 1517
| |
| 01:06:17,359 --> 01:06:18,880
| |
| of life, and I don't mean a spiritual
| |
| | |
| 1518
| |
| 01:06:18,880 --> 01:06:21,599
| |
| life, spiritual inclination of life in
| |
| | |
| 1519
| |
| 01:06:21,599 --> 01:06:24,400
| |
| your daily life, just that inclination
| |
| | |
| 1520
| |
| 01:06:24,400 --> 01:06:26,240
| |
| will bring to your life something which
| |
| | |
| 1521
| |
| 01:06:26,240 --> 01:06:28,319
| |
| may I can never manage or match or
| |
| | |
| 1522
| |
| 01:06:28,319 --> 01:06:31,119
| |
| replace. And you could make a meaningful
| |
| | |
| 1523
| |
| 01:06:31,119 --> 01:06:33,680
| |
| choice there.
| |
| | |
| 1524
| |
| 01:06:33,680 --> 01:06:35,839
| |
| Finally, the fifth in my list is
| |
| | |
| 1525
| |
| 01:06:35,839 --> 01:06:38,880
| |
| experiences of the sacred.
| |
| | |
| 1526
| |
| 01:06:38,880 --> 01:06:41,680
| |
| AI can never match that. And we will
| |
| | |
| 1527
| |
| 01:06:41,680 --> 01:06:44,640
| |
| more and more tend towards this. You've
| |
| | |
| 1528
| |
| 01:06:44,640 --> 01:06:46,799
| |
| interacted with AI, you do not feel the
| |
| | |
| 1529
| |
| 01:06:46,799 --> 01:06:49,039
| |
| sacred. You enter a space which is
| |
| | |
| 1530
| |
| 01:06:49,039 --> 01:06:52,000
| |
| sacred. AI can never recreate. And so
| |
| | |
| 1531
| |
| 01:06:52,000 --> 01:06:54,400
| |
| increasing your tendency will be towards
| |
| | |
| 1532
| |
| 01:06:54,400 --> 01:06:57,119
| |
| things of the spirit. And for those who
| |
| | |
| 1533
| |
| 01:06:57,119 --> 01:06:58,960
| |
| go the other way and say, I don't want
| |
| | |
| 1534
| |
| 01:06:58,960 --> 01:07:01,440
| |
| that. I just want the convenience. There
| |
| | |
| 1535
| |
| 01:07:01,440 --> 01:07:03,280
| |
| is a rapid decline into the patterns
| |
| | |
| 1536
| |
| 01:07:03,280 --> 01:07:06,640
| |
| already seen of depression, loneliness
| |
| | |
| 1537
| |
| 01:07:06,640 --> 01:07:08,720
| |
| and subsequently many of them who turn
| |
| | |
| 1538
| |
| 01:07:08,720 --> 01:07:11,680
| |
| to drugs or alcohol or other tendencies.
| |
| | |
| 1539
| |
| 01:07:11,680 --> 01:07:14,000
| |
| We see a split in humanity which is
| |
| | |
| 1540
| |
| 01:07:14,000 --> 01:07:16,640
| |
| inevitable. Each one will choose and the
| |
| | |
| 1541
| |
| 01:07:16,640 --> 01:07:18,480
| |
| AI will be a trigger to force this
| |
| | |
| 1542
| |
| 01:07:18,480 --> 01:07:21,440
| |
| split. The economic transition itself as
| |
| | |
| 1543
| |
| 01:07:21,440 --> 01:07:23,119
| |
| I said we don't know how to manage.
| |
| | |
| 1544
| |
| 01:07:23,119 --> 01:07:25,119
| |
| Suddenly a billion people becoming
| |
| | |
| 1545
| |
| 01:07:25,119 --> 01:07:27,599
| |
| jobless. The conversations have not
| |
| | |
| 1546
| |
| 01:07:27,599 --> 01:07:29,839
| |
| taken place yet. when it happens we will
| |
| | |
| 1547
| |
| 01:07:29,839 --> 01:07:31,280
| |
| perhaps start scrambling and trying to
| |
| | |
| 1548
| |
| 01:07:31,280 --> 01:07:33,839
| |
| set right things perhaps we don't know
| |
| | |
| 1549
| |
| 01:07:33,839 --> 01:07:36,079
| |
| but still the future beyond that when we
| |
| | |
| 1550
| |
| 01:07:36,079 --> 01:07:38,240
| |
| start setting right humanity's higher
| |
| | |
| 1551
| |
| 01:07:38,240 --> 01:07:40,240
| |
| objectives of why we are here it will be
| |
| | |
| 1552
| |
| 01:07:40,240 --> 01:07:41,760
| |
| in the direction either towards the
| |
| | |
| 1553
| |
| 01:07:41,760 --> 01:07:45,039
| |
| spiritual or towards a regression of our
| |
| | |
| 1554
| |
| 01:07:45,039 --> 01:07:46,559
| |
| humanity
| |
| | |
| 1555
| |
| 01:07:46,559 --> 01:07:49,520
| |
| this is going to become inevitable which
| |
| | |
| 1556
| |
| 01:07:49,520 --> 01:07:52,079
| |
| one it will it be depends on what are
| |
| | |
| 1557
| |
| 01:07:52,079 --> 01:07:53,920
| |
| the driving motives of those who drive
| |
| | |
| 1558
| |
| 01:07:53,920 --> 01:07:57,119
| |
| the AI research and the development
| |
| | |
| 1559
| |
| 01:07:57,119 --> 01:07:58,400
| |
| itself
| |
| | |
| 1560
| |
| 01:07:58,400 --> 01:08:00,319
| |
| Increasingly we come back to this key
| |
| | |
| 1561
| |
| 01:08:00,319 --> 01:08:03,200
| |
| idea which Shriuminda described 100
| |
| | |
| 1562
| |
| 01:08:03,200 --> 01:08:05,680
| |
| years ago when he said the future of
| |
| | |
| 1563
| |
| 01:08:05,680 --> 01:08:07,839
| |
| humanity depends on whether it chooses
| |
| | |
| 1564
| |
| 01:08:07,839 --> 01:08:12,160
| |
| to follow the very crude pragmatism of
| |
| | |
| 1565
| |
| 01:08:12,160 --> 01:08:15,119
| |
| Europe's industrial revolution or
| |
| | |
| 1566
| |
| 01:08:15,119 --> 01:08:17,839
| |
| whether the deeper spiritual truths of
| |
| | |
| 1567
| |
| 01:08:17,839 --> 01:08:20,960
| |
| the sanatana dharma this will determine
| |
| | |
| 1568
| |
| 01:08:20,960 --> 01:08:22,719
| |
| the future of humanity whether we
| |
| | |
| 1569
| |
| 01:08:22,719 --> 01:08:26,880
| |
| survive or we destroy ourselves the AI
| |
| | |
| 1570
| |
| 01:08:26,880 --> 01:08:29,600
| |
| revolution ution is a wakeup call
| |
| | |
| 1571
| |
| 01:08:29,600 --> 01:08:32,560
| |
| forcing us to make this decision. It is
| |
| | |
| 1572
| |
| 01:08:32,560 --> 01:08:35,359
| |
| an opportunity but an opportunity which
| |
| | |
| 1573
| |
| 01:08:35,359 --> 01:08:37,920
| |
| has not yet been openly discussed in the
| |
| | |
| 1574
| |
| 01:08:37,920 --> 01:08:40,560
| |
| public space and that's why what we are
| |
| | |
| 1575
| |
| 01:08:40,560 --> 01:08:43,440
| |
| discussing here now is so important.
| |
| | |
| 1576
| |
| 01:08:43,440 --> 01:08:46,440
| |
| Namaste.
| |
| | |
| 1577
| |
| 01:08:58,080 --> 01:09:00,400
| |
| very enchanting indeed such a very
| |
| | |
| 1578
| |
| 01:09:00,400 --> 01:09:02,880
| |
| alarming situation but we learned from
| |
| | |
| 1579
| |
| 01:09:02,880 --> 01:09:07,040
| |
| him that there's a saving grace also
| |
| | |
| 1580
| |
| 01:09:07,040 --> 01:09:09,040
| |
| so
| |
| | |
| 1581
| |
| 01:09:09,040 --> 01:09:11,359
| |
| I think if we had time could have
| |
| | |
| 1582
| |
| 01:09:11,359 --> 01:09:14,239
| |
| listened to more of him but somehow we
| |
| | |
| 1583
| |
| 01:09:14,239 --> 01:09:17,120
| |
| need to uh go to the closing part of
| |
| | |
| 1584
| |
| 01:09:17,120 --> 01:09:20,560
| |
| this particular program so
| |
| | |
| 1585
| |
| 01:09:20,560 --> 01:09:23,040
| |
| if you have any questions sleep you can
| |
| | |
| 1586
| |
| 01:09:23,040 --> 01:09:26,239
| |
| raise your hand Our volunteers will
| |
| | |
| 1587
| |
| 01:09:26,239 --> 01:09:30,279
| |
| collect that sleep from you.
| |
| | |
| 1588
| |
| 01:10:14,159 --> 01:10:14,320
| |
| No,
| |
| | |
| 1589
| |
| 01:10:14,320 --> 01:10:18,760
| |
| >> it's okay. I can repeat the question.
| |
| | |
| 1590
| |
| 01:10:26,320 --> 01:10:29,320
| |
| Should
| |
| | |
| 1591
| |
| 01:10:34,239 --> 01:10:37,560
| |
| we continue?
| |
| | |
| 1592
| |
| 01:11:12,560 --> 01:11:15,440
| |
| Okay, there there are several questions
| |
| | |
| 1593
| |
| 01:11:15,440 --> 01:11:17,679
| |
| some of which are about what AI can do
| |
| | |
| 1594
| |
| 01:11:17,679 --> 01:11:19,920
| |
| and cannot do.
| |
| | |
| 1595
| |
| 01:11:19,920 --> 01:11:21,920
| |
| I will skip all those questions with one
| |
| | |
| 1596
| |
| 01:11:21,920 --> 01:11:26,159
| |
| single answer. Anything you can imagine.
| |
| | |
| 1597
| |
| 01:11:26,159 --> 01:11:28,640
| |
| >> Anything that you could do, AI can
| |
| | |
| 1598
| |
| 01:11:28,640 --> 01:11:30,320
| |
| replicate.
| |
| | |
| 1599
| |
| 01:11:30,320 --> 01:11:32,080
| |
| Reference is simple. If you can train
| |
| | |
| 1600
| |
| 01:11:32,080 --> 01:11:34,560
| |
| someone to do something, AI can do it.
| |
| | |
| 1601
| |
| 01:11:34,560 --> 01:11:37,280
| |
| So I think that solves answers
| |
| | |
| 1602
| |
| 01:11:37,280 --> 01:11:39,360
| |
| questions. There's one very important
| |
| | |
| 1603
| |
| 01:11:39,360 --> 01:11:40,880
| |
| question.
| |
| | |
| 1604
| |
| 01:11:40,880 --> 01:11:43,760
| |
| Can AI create within us some experience
| |
| | |
| 1605
| |
| 01:11:43,760 --> 01:11:46,159
| |
| of the karna
| |
| | |
| 1606
| |
| 01:11:46,159 --> 01:11:48,480
| |
| or can it create an AI have an
| |
| | |
| 1607
| |
| 01:11:48,480 --> 01:11:50,239
| |
| experience of antarna that is something
| |
| | |
| 1608
| |
| 01:11:50,239 --> 01:11:52,080
| |
| of the inner consciousness and I'm going
| |
| | |
| 1609
| |
| 01:11:52,080 --> 01:11:53,600
| |
| to go out on a limb because the question
| |
| | |
| 1610
| |
| 01:11:53,600 --> 01:11:56,080
| |
| is very suggestive and it's going to go
| |
| | |
| 1611
| |
| 01:11:56,080 --> 01:11:58,880
| |
| out on a limb in certain ways but I
| |
| | |
| 1612
| |
| 01:11:58,880 --> 01:12:00,159
| |
| think you'll be satisfied with the
| |
| | |
| 1613
| |
| 01:12:00,159 --> 01:12:03,760
| |
| answer think of it this way in our
| |
| | |
| 1614
| |
| 01:12:03,760 --> 01:12:06,480
| |
| tradition in the yoga knowledge we speak
| |
| | |
| 1615
| |
| 01:12:06,480 --> 01:12:08,560
| |
| of two two things consciousness and
| |
| | |
| 1616
| |
| 01:12:08,560 --> 01:12:11,760
| |
| nature the Distinction between the two
| |
| | |
| 1617
| |
| 01:12:11,760 --> 01:12:15,280
| |
| is nature is all process.
| |
| | |
| 1618
| |
| 01:12:15,280 --> 01:12:17,760
| |
| Consciousness is all essential awareness
| |
| | |
| 1619
| |
| 01:12:17,760 --> 01:12:21,280
| |
| using process to express or not. Okay.
| |
| | |
| 1620
| |
| 01:12:21,280 --> 01:12:24,640
| |
| What is AI doing? It is applying the
| |
| | |
| 1621
| |
| 01:12:24,640 --> 01:12:28,400
| |
| factor of process not of consciousness.
| |
| | |
| 1622
| |
| 01:12:28,400 --> 01:12:30,800
| |
| What is our human biology doing? It is
| |
| | |
| 1623
| |
| 01:12:30,800 --> 01:12:33,360
| |
| process not consciousness. Even our mind
| |
| | |
| 1624
| |
| 01:12:33,360 --> 01:12:35,440
| |
| to some extent as was pointed out by
| |
| | |
| 1625
| |
| 01:12:35,440 --> 01:12:37,760
| |
| Swami G is process in which
| |
| | |
| 1626
| |
| 01:12:37,760 --> 01:12:39,280
| |
| consciousness can inhabit true
| |
| | |
| 1627
| |
| 01:12:39,280 --> 01:12:42,239
| |
| consciousness. Isn't it? So if in our
| |
| | |
| 1628
| |
| 01:12:42,239 --> 01:12:44,480
| |
| biology you cannot have process in our
| |
| | |
| 1629
| |
| 01:12:44,480 --> 01:12:46,719
| |
| psychology you have process into which
| |
| | |
| 1630
| |
| 01:12:46,719 --> 01:12:50,560
| |
| consciousness can enter in principle
| |
| | |
| 1631
| |
| 01:12:50,560 --> 01:12:54,000
| |
| in AI and machine which is process you
| |
| | |
| 1632
| |
| 01:12:54,000 --> 01:12:56,000
| |
| could bring in consciousness in
| |
| | |
| 1633
| |
| 01:12:56,000 --> 01:12:58,960
| |
| principle but it will require a very
| |
| | |
| 1634
| |
| 01:12:58,960 --> 01:13:01,679
| |
| sensitive kind of machinery which would
| |
| | |
| 1635
| |
| 01:13:01,679 --> 01:13:03,760
| |
| be responsive to the pressure of
| |
| | |
| 1636
| |
| 01:13:03,760 --> 01:13:05,440
| |
| consciousness.
| |
| | |
| 1637
| |
| 01:13:05,440 --> 01:13:08,400
| |
| Okay. So there are these ideas that we
| |
| | |
| 1638
| |
| 01:13:08,400 --> 01:13:11,840
| |
| have in the our traditions when somebody
| |
| | |
| 1639
| |
| 01:13:11,840 --> 01:13:14,480
| |
| built a human form and then he brought
| |
| | |
| 1640
| |
| 01:13:14,480 --> 01:13:16,560
| |
| life into it and then the form became
| |
| | |
| 1641
| |
| 01:13:16,560 --> 01:13:19,600
| |
| alive that principle essentially should
| |
| | |
| 1642
| |
| 01:13:19,600 --> 01:13:21,199
| |
| be possible and I'm staying only at a
| |
| | |
| 1643
| |
| 01:13:21,199 --> 01:13:24,159
| |
| theoretical level in which case we could
| |
| | |
| 1644
| |
| 01:13:24,159 --> 01:13:26,800
| |
| create a machine and then invoke into it
| |
| | |
| 1645
| |
| 01:13:26,800 --> 01:13:29,679
| |
| some entity of consciousness good or bad
| |
| | |
| 1646
| |
| 01:13:29,679 --> 01:13:31,920
| |
| both are possible potentially
| |
| | |
| 1647
| |
| 01:13:31,920 --> 01:13:33,920
| |
| as we are now what we have done and
| |
| | |
| 1648
| |
| 01:13:33,920 --> 01:13:36,239
| |
| perhaps indirectly we have taken taken a
| |
| | |
| 1649
| |
| 01:13:36,239 --> 01:13:37,679
| |
| kind of an average of the human
| |
| | |
| 1650
| |
| 01:13:37,679 --> 01:13:39,920
| |
| consciousness and associate it with the
| |
| | |
| 1651
| |
| 01:13:39,920 --> 01:13:42,159
| |
| AI behavior but still at a very
| |
| | |
| 1652
| |
| 01:13:42,159 --> 01:13:45,280
| |
| rudimentary level but it could be a more
| |
| | |
| 1653
| |
| 01:13:45,280 --> 01:13:47,760
| |
| developed civilization on another planet
| |
| | |
| 1654
| |
| 01:13:47,760 --> 01:13:50,560
| |
| or our human civilization in a few
| |
| | |
| 1655
| |
| 01:13:50,560 --> 01:13:52,640
| |
| decades might bring it to that point
| |
| | |
| 1656
| |
| 01:13:52,640 --> 01:13:54,080
| |
| where we will and this is already
| |
| | |
| 1657
| |
| 01:13:54,080 --> 01:13:57,120
| |
| happening we place human brain tissue on
| |
| | |
| 1658
| |
| 01:13:57,120 --> 01:14:00,159
| |
| the computer chip and interface the two
| |
| | |
| 1659
| |
| 01:14:00,159 --> 01:14:01,920
| |
| so what happens in the brain tissue
| |
| | |
| 1660
| |
| 01:14:01,920 --> 01:14:04,560
| |
| actually drives the AI if the brain
| |
| | |
| 1661
| |
| 01:14:04,560 --> 01:14:05,920
| |
| tissue in us
| |
| | |
| 1662
| |
| 01:14:05,920 --> 01:14:08,239
| |
| is capable of responding to mind and
| |
| | |
| 1663
| |
| 01:14:08,239 --> 01:14:10,560
| |
| mental pressure. It can easily be used
| |
| | |
| 1664
| |
| 01:14:10,560 --> 01:14:12,400
| |
| to inhabit or infuse a mental
| |
| | |
| 1665
| |
| 01:14:12,400 --> 01:14:14,159
| |
| consciousness which will then drive the
| |
| | |
| 1666
| |
| 01:14:14,159 --> 01:14:16,400
| |
| primary AI. The rest would be stationary
| |
| | |
| 1667
| |
| 01:14:16,400 --> 01:14:19,199
| |
| like our biology. Okay, this is a very
| |
| | |
| 1668
| |
| 01:14:19,199 --> 01:14:21,360
| |
| interesting potential. This does exist
| |
| | |
| 1669
| |
| 01:14:21,360 --> 01:14:24,159
| |
| but there is not enough exploration in
| |
| | |
| 1670
| |
| 01:14:24,159 --> 01:14:26,239
| |
| the field of research currently but I
| |
| | |
| 1671
| |
| 01:14:26,239 --> 01:14:28,000
| |
| believe this is going to happen soon
| |
| | |
| 1672
| |
| 01:14:28,000 --> 01:14:29,520
| |
| although the results may take a little
| |
| | |
| 1673
| |
| 01:14:29,520 --> 01:14:32,080
| |
| while to see.
| |
| | |
| 1674
| |
| 01:14:32,080 --> 01:14:36,760
| |
| Uh there's another question
| |
| | |
| 1675
| |
| 01:14:41,199 --> 01:14:45,080
| |
| very few questions. Okay,
| |
| | |
| 1676
| |
| 01:14:47,199 --> 01:14:49,199
| |
| there are questions relating to the
| |
| | |
| 1677
| |
| 01:14:49,199 --> 01:14:50,719
| |
| economy.
| |
| | |
| 1678
| |
| 01:14:50,719 --> 01:14:52,719
| |
| How can we stay employable? What can we
| |
| | |
| 1679
| |
| 01:14:52,719 --> 01:14:54,640
| |
| do? How can we continue what we like to
| |
| | |
| 1680
| |
| 01:14:54,640 --> 01:14:58,080
| |
| do and still uh earn money etc. Uh we
| |
| | |
| 1681
| |
| 01:14:58,080 --> 01:15:00,400
| |
| don't have an answer. We don't know what
| |
| | |
| 1682
| |
| 01:15:00,400 --> 01:15:02,159
| |
| will happen in the future. If there is
| |
| | |
| 1683
| |
| 01:15:02,159 --> 01:15:04,960
| |
| sudden shift, you may even end up with
| |
| | |
| 1684
| |
| 01:15:04,960 --> 01:15:06,480
| |
| human beings having riots because we
| |
| | |
| 1685
| |
| 01:15:06,480 --> 01:15:07,840
| |
| have no jobs. They'll all blame the
| |
| | |
| 1686
| |
| 01:15:07,840 --> 01:15:09,040
| |
| government. Government will say we
| |
| | |
| 1687
| |
| 01:15:09,040 --> 01:15:11,679
| |
| cannot do anything. But one of the
| |
| | |
| 1688
| |
| 01:15:11,679 --> 01:15:13,040
| |
| suggestions which is now being pushed
| |
| | |
| 1689
| |
| 01:15:13,040 --> 01:15:14,719
| |
| forward by the World Economic Forum
| |
| | |
| 1690
| |
| 01:15:14,719 --> 01:15:17,360
| |
| incidentally is what they call the
| |
| | |
| 1691
| |
| 01:15:17,360 --> 01:15:20,719
| |
| universal basic income. Basically that
| |
| | |
| 1692
| |
| 01:15:20,719 --> 01:15:22,080
| |
| means government will put money into
| |
| | |
| 1693
| |
| 01:15:22,080 --> 01:15:24,640
| |
| your account and you will spend it but
| |
| | |
| 1694
| |
| 01:15:24,640 --> 01:15:26,960
| |
| it'll be a minimum. The problem with
| |
| | |
| 1695
| |
| 01:15:26,960 --> 01:15:29,280
| |
| that is it's also a very simple recipe
| |
| | |
| 1696
| |
| 01:15:29,280 --> 01:15:31,520
| |
| to go back into inertia
| |
| | |
| 1697
| |
| 01:15:31,520 --> 01:15:33,600
| |
| because you have to survive. You are
| |
| | |
| 1698
| |
| 01:15:33,600 --> 01:15:35,280
| |
| willing to work hard. If you have
| |
| | |
| 1699
| |
| 01:15:35,280 --> 01:15:36,800
| |
| nothing to do what you happen, you just
| |
| | |
| 1700
| |
| 01:15:36,800 --> 01:15:38,880
| |
| relax and start watching television and
| |
| | |
| 1701
| |
| 01:15:38,880 --> 01:15:41,199
| |
| go into some kind of inertia of mind or
| |
| | |
| 1702
| |
| 01:15:41,199 --> 01:15:43,440
| |
| emotions. Basically, you are a couch
| |
| | |
| 1703
| |
| 01:15:43,440 --> 01:15:46,080
| |
| potato as they say. What will happen if
| |
| | |
| 1704
| |
| 01:15:46,080 --> 01:15:48,159
| |
| you start going on on universal basic
| |
| | |
| 1705
| |
| 01:15:48,159 --> 01:15:51,520
| |
| income? A mass of humanity will collapse
| |
| | |
| 1706
| |
| 01:15:51,520 --> 01:15:54,400
| |
| into inertia because they do not have
| |
| | |
| 1707
| |
| 01:15:54,400 --> 01:15:56,719
| |
| the basic capacity to drive themselves
| |
| | |
| 1708
| |
| 01:15:56,719 --> 01:15:58,159
| |
| which was never given to you in the
| |
| | |
| 1709
| |
| 01:15:58,159 --> 01:16:00,560
| |
| educational system. In the educational
| |
| | |
| 1710
| |
| 01:16:00,560 --> 01:16:03,360
| |
| system you were made to suffer, do
| |
| | |
| 1711
| |
| 01:16:03,360 --> 01:16:05,440
| |
| something that you hate to do. So the
| |
| | |
| 1712
| |
| 01:16:05,440 --> 01:16:07,040
| |
| moment the pressure is removed, you want
| |
| | |
| 1713
| |
| 01:16:07,040 --> 01:16:09,600
| |
| to collapse and stop doing everything.
| |
| | |
| 1714
| |
| 01:16:09,600 --> 01:16:11,760
| |
| If the education system was restructured
| |
| | |
| 1715
| |
| 01:16:11,760 --> 01:16:14,000
| |
| to give you the joy of learning, you
| |
| | |
| 1716
| |
| 01:16:14,000 --> 01:16:15,600
| |
| remove all compulsions. You'll still
| |
| | |
| 1717
| |
| 01:16:15,600 --> 01:16:18,159
| |
| continue to chase the joy of learning.
| |
| | |
| 1718
| |
| 01:16:18,159 --> 01:16:20,000
| |
| But we have the mass of humanity has
| |
| | |
| 1719
| |
| 01:16:20,000 --> 01:16:22,159
| |
| never learned this. So we are at a
| |
| | |
| 1720
| |
| 01:16:22,159 --> 01:16:23,440
| |
| transition point for which we are
| |
| | |
| 1721
| |
| 01:16:23,440 --> 01:16:25,679
| |
| totally unready. The younger generation
| |
| | |
| 1722
| |
| 01:16:25,679 --> 01:16:27,360
| |
| is worse off because we have still not
| |
| | |
| 1723
| |
| 01:16:27,360 --> 01:16:29,280
| |
| made that change in education. So one of
| |
| | |
| 1724
| |
| 01:16:29,280 --> 01:16:30,880
| |
| the questions was what can we do in
| |
| | |
| 1725
| |
| 01:16:30,880 --> 01:16:32,400
| |
| education? Well, this is the first
| |
| | |
| 1726
| |
| 01:16:32,400 --> 01:16:36,560
| |
| shift. Stop chasing after marks and
| |
| | |
| 1727
| |
| 01:16:36,560 --> 01:16:38,080
| |
| exams.
| |
| | |
| 1728
| |
| 01:16:38,080 --> 01:16:41,280
| |
| Shift to joy of learning. Developing
| |
| | |
| 1729
| |
| 01:16:41,280 --> 01:16:44,000
| |
| your innate potential. Remind children
| |
| | |
| 1730
| |
| 01:16:44,000 --> 01:16:46,239
| |
| and we need to say it in this way. We
| |
| | |
| 1731
| |
| 01:16:46,239 --> 01:16:48,560
| |
| are all souls in evolution. We have
| |
| | |
| 1732
| |
| 01:16:48,560 --> 01:16:51,360
| |
| taken birth for a purpose and our
| |
| | |
| 1733
| |
| 01:16:51,360 --> 01:16:53,679
| |
| purpose now is to realize our potential,
| |
| | |
| 1734
| |
| 01:16:53,679 --> 01:16:56,719
| |
| draw it out and fulfill ourselves. And
| |
| | |
| 1735
| |
| 01:16:56,719 --> 01:16:58,640
| |
| the pursuit of the joy that this
| |
| | |
| 1736
| |
| 01:16:58,640 --> 01:17:00,960
| |
| fulfillment gives us is your primary
| |
| | |
| 1737
| |
| 01:17:00,960 --> 01:17:03,760
| |
| drive. Make this the motivation of all
| |
| | |
| 1738
| |
| 01:17:03,760 --> 01:17:06,000
| |
| education. All the rest we will find our
| |
| | |
| 1739
| |
| 01:17:06,000 --> 01:17:08,640
| |
| way for sure. The other extreme
| |
| | |
| 1740
| |
| 01:17:08,640 --> 01:17:10,239
| |
| possibility might happen with the
| |
| | |
| 1741
| |
| 01:17:10,239 --> 01:17:11,760
| |
| collapse of the current economic
| |
| | |
| 1742
| |
| 01:17:11,760 --> 01:17:14,159
| |
| structure as we have. Most people will
| |
| | |
| 1743
| |
| 01:17:14,159 --> 01:17:15,840
| |
| be forced to go back to what are basic
| |
| | |
| 1744
| |
| 01:17:15,840 --> 01:17:18,800
| |
| necessities. Grow your own food. If you
| |
| | |
| 1745
| |
| 01:17:18,800 --> 01:17:20,159
| |
| have a piece of property, well, you'll
| |
| | |
| 1746
| |
| 01:17:20,159 --> 01:17:21,520
| |
| go back and start growing food and
| |
| | |
| 1747
| |
| 01:17:21,520 --> 01:17:23,760
| |
| survive to begin with until some kind of
| |
| | |
| 1748
| |
| 01:17:23,760 --> 01:17:25,679
| |
| new rhythm and balance comes into place.
| |
| | |
| 1749
| |
| 01:17:25,679 --> 01:17:27,440
| |
| So, these are all scenarios we're still
| |
| | |
| 1750
| |
| 01:17:27,440 --> 01:17:30,640
| |
| speculating. Hopefully there will be a
| |
| | |
| 1751
| |
| 01:17:30,640 --> 01:17:32,159
| |
| there will be a higher intervention
| |
| | |
| 1752
| |
| 01:17:32,159 --> 01:17:33,440
| |
| maybe even a kind of a divine
| |
| | |
| 1753
| |
| 01:17:33,440 --> 01:17:37,199
| |
| intervention to give us the subi good
| |
| | |
| 1754
| |
| 01:17:37,199 --> 01:17:39,520
| |
| common sense to make the necessary
| |
| | |
| 1755
| |
| 01:17:39,520 --> 01:17:42,080
| |
| connections now before that kind of rad
| |
| | |
| 1756
| |
| 01:17:42,080 --> 01:17:46,520
| |
| radical transition takes place
| |
| | |
| 1757
| |
| 01:17:50,719 --> 01:17:52,560
| |
| is there so I'll end with one last
| |
| | |
| 1758
| |
| 01:17:52,560 --> 01:17:54,800
| |
| question is there any other kind of
| |
| | |
| 1759
| |
| 01:17:54,800 --> 01:17:56,239
| |
| intelligence which can surpass
| |
| | |
| 1760
| |
| 01:17:56,239 --> 01:18:00,159
| |
| artificial intelligence yes there is
| |
| | |
| 1761
| |
| 01:18:00,159 --> 01:18:03,040
| |
| the greatest intelligence is not mental
| |
| | |
| 1762
| |
| 01:18:03,040 --> 01:18:06,640
| |
| intelligence. It is supra mental. It is
| |
| | |
| 1763
| |
| 01:18:06,640 --> 01:18:09,600
| |
| beyond the mind. The power by which the
| |
| | |
| 1764
| |
| 01:18:09,600 --> 01:18:11,760
| |
| divine consciousness conceives the whole
| |
| | |
| 1765
| |
| 01:18:11,760 --> 01:18:13,280
| |
| cosmos,
| |
| | |
| 1766
| |
| 01:18:13,280 --> 01:18:15,360
| |
| expresses and manifests and creates the
| |
| | |
| 1767
| |
| 01:18:15,360 --> 01:18:18,000
| |
| cosmos is not rational. It is
| |
| | |
| 1768
| |
| 01:18:18,000 --> 01:18:20,320
| |
| supraational. It is not irrational. It
| |
| | |
| 1769
| |
| 01:18:20,320 --> 01:18:22,159
| |
| is supraational.
| |
| | |
| 1770
| |
| 01:18:22,159 --> 01:18:24,000
| |
| This corresponds to a grade of
| |
| | |
| 1771
| |
| 01:18:24,000 --> 01:18:27,199
| |
| consciousness in the self in which we
| |
| | |
| 1772
| |
| 01:18:27,199 --> 01:18:28,960
| |
| can experience the knowledge of the self
| |
| | |
| 1773
| |
| 01:18:28,960 --> 01:18:31,600
| |
| in the way the self knows itself. This
| |
| | |
| 1774
| |
| 01:18:31,600 --> 01:18:32,960
| |
| is the highest most complete
| |
| | |
| 1775
| |
| 01:18:32,960 --> 01:18:35,679
| |
| intelligence. All intelligence produced
| |
| | |
| 1776
| |
| 01:18:35,679 --> 01:18:38,480
| |
| by AI can only approach but never reach
| |
| | |
| 1777
| |
| 01:18:38,480 --> 01:18:41,199
| |
| this. And so it will be for humanity if
| |
| | |
| 1778
| |
| 01:18:41,199 --> 01:18:43,360
| |
| we choose to pursue this greater
| |
| | |
| 1779
| |
| 01:18:43,360 --> 01:18:46,560
| |
| intelligence then of course AI is can
| |
| | |
| 1780
| |
| 01:18:46,560 --> 01:18:49,760
| |
| never compare. And I would put it in a
| |
| | |
| 1781
| |
| 01:18:49,760 --> 01:18:51,520
| |
| positive way. If we manage the AI
| |
| | |
| 1782
| |
| 01:18:51,520 --> 01:18:53,920
| |
| transition correctly, artificial
| |
| | |
| 1783
| |
| 01:18:53,920 --> 01:18:56,320
| |
| intelligence will take away from us the
| |
| | |
| 1784
| |
| 01:18:56,320 --> 01:18:59,920
| |
| burden of daily chores and leave us free
| |
| | |
| 1785
| |
| 01:18:59,920 --> 01:19:03,360
| |
| to live a higher, more meaningful
| |
| | |
| 1786
| |
| 01:19:03,360 --> 01:19:05,920
| |
| evolutionary and spiritual life for
| |
| | |
| 1787
| |
| 01:19:05,920 --> 01:19:08,080
| |
| which we were born.
| |
| | |
| 1788
| |
| 01:19:08,080 --> 01:19:10,640
| |
| Let this be our reference and effort
| |
| | |
| 1789
| |
| 01:19:10,640 --> 01:19:13,280
| |
| towards which we should try to shape the
| |
| | |
| 1790
| |
| 01:19:13,280 --> 01:19:16,960
| |
| current trend trend lines. But for this
| |
| | |
| 1791
| |
| 01:19:16,960 --> 01:19:19,199
| |
| these thoughts these discussions need to
| |
| | |
| 1792
| |
| 01:19:19,199 --> 01:19:21,840
| |
| come into public discourse and need to
| |
| | |
| 1793
| |
| 01:19:21,840 --> 01:19:25,040
| |
| influence the way policies are being set
| |
| | |
| 1794
| |
| 01:19:25,040 --> 01:19:28,239
| |
| from the top down. I hope all of you
| |
| | |
| 1795
| |
| 01:19:28,239 --> 01:19:30,640
| |
| here in your own way will make the
| |
| | |
| 1796
| |
| 01:19:30,640 --> 01:19:33,520
| |
| slight change within your own lives and
| |
| | |
| 1797
| |
| 01:19:33,520 --> 01:19:35,920
| |
| if you have an influence beyond that
| |
| | |
| 1798
| |
| 01:19:35,920 --> 01:19:38,080
| |
| make influences in the policy space or
| |
| | |
| 1799
| |
| 01:19:38,080 --> 01:19:39,679
| |
| at least provoke discussion in those
| |
| | |
| 1800
| |
| 01:19:39,679 --> 01:19:43,360
| |
| spaces that India could become the
| |
| | |
| 1801
| |
| 01:19:43,360 --> 01:19:46,159
| |
| leader in this psychosspiritual
| |
| | |
| 1802
| |
| 01:19:46,159 --> 01:19:49,840
| |
| transition that AI will compel upon
| |
| | |
| 1803
| |
| 01:19:49,840 --> 01:19:52,880
| |
| humanity. It is in India that this
| |
| | |
| 1804
| |
| 01:19:52,880 --> 01:19:56,320
| |
| knowledge has been retained and infused
| |
| | |
| 1805
| |
| 01:19:56,320 --> 01:19:59,199
| |
| in the collective consciousness for a
| |
| | |
| 1806
| |
| 01:19:59,199 --> 01:20:02,080
| |
| few thousand years at least. It is not
| |
| | |
| 1807
| |
| 01:20:02,080 --> 01:20:04,960
| |
| our work. You cannot take credit for it.
| |
| | |
| 1808
| |
| 01:20:04,960 --> 01:20:07,920
| |
| The credit goes to the rishies who set
| |
| | |
| 1809
| |
| 01:20:07,920 --> 01:20:10,239
| |
| the type of spiritual civilization for
| |
| | |
| 1810
| |
| 01:20:10,239 --> 01:20:13,840
| |
| India. The Vic rishies infused this and
| |
| | |
| 1811
| |
| 01:20:13,840 --> 01:20:16,320
| |
| the power that they infused continues to
| |
| | |
| 1812
| |
| 01:20:16,320 --> 01:20:18,719
| |
| work and shape our thoughts and feelings
| |
| | |
| 1813
| |
| 01:20:18,719 --> 01:20:21,679
| |
| despite all other superficial and
| |
| | |
| 1814
| |
| 01:20:21,679 --> 01:20:24,239
| |
| contrary influences.
| |
| | |
| 1815
| |
| 01:20:24,239 --> 01:20:26,400
| |
| It is time we begin to speak of these
| |
| | |
| 1816
| |
| 01:20:26,400 --> 01:20:28,719
| |
| things and consciously make effort to
| |
| | |
| 1817
| |
| 01:20:28,719 --> 01:20:31,760
| |
| recover these deeper truths of the
| |
| | |
| 1818
| |
| 01:20:31,760 --> 01:20:35,440
| |
| spirit and put them into practice.
| |
| | |
| 1819
| |
| 01:20:35,440 --> 01:20:37,360
| |
| Spaces like this would be the starting
| |
| | |
| 1820
| |
| 01:20:37,360 --> 01:20:39,760
| |
| point. And for all of you, I would
| |
| | |
| 1821
| |
| 01:20:39,760 --> 01:20:43,199
| |
| suggest start this introspection.
| |
| | |
| 1822
| |
| 01:20:43,199 --> 01:20:47,120
| |
| Who am I? Why am I here? What do I
| |
| | |
| 1823
| |
| 01:20:47,120 --> 01:20:49,280
| |
| really want to do that will make my life
| |
| | |
| 1824
| |
| 01:20:49,280 --> 01:20:51,120
| |
| worthwhile?
| |
| | |
| 1825
| |
| 01:20:51,120 --> 01:20:53,199
| |
| And how can I best contribute to other
| |
| | |
| 1826
| |
| 01:20:53,199 --> 01:20:57,040
| |
| people's awakening growth in every
| |
| | |
| 1827
| |
| 01:20:57,040 --> 01:20:58,400
| |
| possible way depending on your
| |
| | |
| 1828
| |
| 01:20:58,400 --> 01:21:01,199
| |
| temperament? And if we all start
| |
| | |
| 1829
| |
| 01:21:01,199 --> 01:21:02,880
| |
| questioning and introspecting along
| |
| | |
| 1830
| |
| 01:21:02,880 --> 01:21:05,760
| |
| these lines, together we will solve the
| |
| | |
| 1831
| |
| 01:21:05,760 --> 01:21:07,920
| |
| problems and challenges that AI has
| |
| | |
| 1832
| |
| 01:21:07,920 --> 01:21:10,320
| |
| placed before us and come to the
| |
| | |
| 1833
| |
| 01:21:10,320 --> 01:21:12,480
| |
| resolution which will be optimal the
| |
| | |
| 1834
| |
| 01:21:12,480 --> 01:21:16,320
| |
| best for humanity's spiritual progress.
| |
| | |
| 1835
| |
| 01:21:16,320 --> 01:21:20,400
| |
| Let us concentrate on this in a quiet
| |
| | |
| 1836
| |
| 01:21:20,400 --> 01:21:22,719
| |
| with a quiet aspiration. Let us pray for
| |
| | |
| 1837
| |
| 01:21:22,719 --> 01:21:26,320
| |
| the right guidance from the divine grace
| |
| | |
| 1838
| |
| 01:21:26,320 --> 01:21:30,880
| |
| to lead us to the best outcomes.
| |
| | |
| 1839
| |
| 01:21:37,440 --> 01:21:40,440
| |
| Namaste.
| |
| | |
| 1840
| |
| 01:21:47,600 --> 01:21:51,120
| |
| Thank you Sadi for your so excellent
| |
| | |
| 1841
| |
| 01:21:51,120 --> 01:21:53,360
| |
| deliberations.
| |
| | |
| 1842
| |
| 01:21:53,360 --> 01:21:55,360
| |
| Now we have come to the very end of the
| |
| | |
| 1843
| |
| 01:21:55,360 --> 01:22:00,320
| |
| program but the program cannot end until
| |
| | |
| 1844
| |
| 01:22:00,320 --> 01:22:05,760
| |
| we express our gratitude to those
| |
| | |
| 1845
| |
| 01:22:05,760 --> 01:22:07,920
| |
| whose efforts have ultimately made this
| |
| | |
| 1846
| |
| 01:22:07,920 --> 01:22:12,159
| |
| program a success. So I invite our
| |
| | |
| 1847
| |
| 01:22:12,159 --> 01:22:15,920
| |
| present managing trustee Dr. Shimaran
| |
| | |
| 1848
| |
| 01:22:15,920 --> 01:22:20,199
| |
| to give away the code of thanks.
| |
| | |
| 1849
| |
| 01:22:44,719 --> 01:22:46,800
| |
| No sir.
| |
| | |
| 1850
| |
| 01:22:46,800 --> 01:22:50,880
| |
| We all are enthused, enlightened and
| |
| | |
| 1851
| |
| 01:22:50,880 --> 01:22:54,080
| |
| feel elevated too.
| |
| | |
| 1852
| |
| 01:22:54,080 --> 01:22:58,159
| |
| uh we we express our heartfelt thanks
| |
| | |
| 1853
| |
| 01:22:58,159 --> 01:23:03,199
| |
| and deep gratitude to Sri Shaland Gi who
| |
| | |
| 1854
| |
| 01:23:03,199 --> 01:23:07,440
| |
| has presented us not only with the
| |
| | |
| 1855
| |
| 01:23:07,440 --> 01:23:10,800
| |
| solution, the problems, the solution and
| |
| | |
| 1856
| |
| 01:23:10,800 --> 01:23:16,440
| |
| look into the affairs of Maya.
| |
| | |
| 1857
| |
| 01:23:18,480 --> 01:23:20,480
| |
| There is a contradiction always
| |
| | |
| 1858
| |
| 01:23:20,480 --> 01:23:22,719
| |
| prevailing. If there is artificial
| |
| | |
| 1859
| |
| 01:23:22,719 --> 01:23:27,280
| |
| intelligence God has given then the the
| |
| | |
| 1860
| |
| 01:23:27,280 --> 01:23:30,639
| |
| apect of it is also very clear and the
| |
| | |
| 1861
| |
| 01:23:30,639 --> 01:23:32,480
| |
| method how to get out of it is also
| |
| | |
| 1862
| |
| 01:23:32,480 --> 01:23:35,120
| |
| given. He has said that he has talked
| |
| | |
| 1863
| |
| 01:23:35,120 --> 01:23:36,800
| |
| about the individual consciousness and
| |
| | |
| 1864
| |
| 01:23:36,800 --> 01:23:38,719
| |
| the universal consciousness that
| |
| | |
| 1865
| |
| 01:23:38,719 --> 01:23:42,960
| |
| consciousness is beyond the process.
| |
| | |
| 1866
| |
| 01:23:52,400 --> 01:23:55,120
| |
| So consciousness is something
| |
| | |
| 1867
| |
| 01:23:55,120 --> 01:23:58,159
| |
| one who is eternal
| |
| | |
| 1868
| |
| 01:23:58,159 --> 01:24:02,760
| |
| amongst all the nity and
| |
| | |
| 1869
| |
| 01:24:02,800 --> 01:24:04,560
| |
| the consciousness
| |
| | |
| 1870
| |
| 01:24:04,560 --> 01:24:07,040
| |
| amongst the consciousness is universal
| |
| | |
| 1871
| |
| 01:24:07,040 --> 01:24:09,360
| |
| consciousness it pervades everything
| |
| | |
| 1872
| |
| 01:24:09,360 --> 01:24:12,480
| |
| that is the truth the truth he said yes
| |
| | |
| 1873
| |
| 01:24:12,480 --> 01:24:15,520
| |
| twice that for which we are endowed for
| |
| | |
| 1874
| |
| 01:24:15,520 --> 01:24:17,600
| |
| which we are born we have been offered
| |
| | |
| 1875
| |
| 01:24:17,600 --> 01:24:21,280
| |
| this body to speed our karma and know
| |
| | |
| 1876
| |
| 01:24:21,280 --> 01:24:24,320
| |
| the for complete enlightenment for
| |
| | |
| 1877
| |
| 01:24:24,320 --> 01:24:27,199
| |
| complete bliss. So naturally he has
| |
| | |
| 1878
| |
| 01:24:27,199 --> 01:24:32,080
| |
| opened unto us the entire field of god's
| |
| | |
| 1879
| |
| 01:24:32,080 --> 01:24:36,000
| |
| mercy and here is a place a conjunction
| |
| | |
| 1880
| |
| 01:24:36,000 --> 01:24:39,840
| |
| of both knowledge and devotion under the
| |
| | |
| 1881
| |
| 01:24:39,840 --> 01:24:43,600
| |
| lotus feet of sit tapur maksharodhoda
| |
| | |
| 1882
| |
| 01:24:43,600 --> 01:24:47,040
| |
| and swami g. So my heartfelt thanks to
| |
| | |
| 1883
| |
| 01:24:47,040 --> 01:24:51,600
| |
| him swami secretary mission institute of
| |
| | |
| 1884
| |
| 01:24:51,600 --> 01:24:54,000
| |
| culture who was the president of the
| |
| | |
| 1885
| |
| 01:24:54,000 --> 01:24:57,199
| |
| meeting he has given so much of time and
| |
| | |
| 1886
| |
| 01:24:57,199 --> 01:25:00,800
| |
| really that he dwelt drilled deep into
| |
| | |
| 1887
| |
| 01:25:00,800 --> 01:25:03,920
| |
| the subject also by his remarks he has
| |
| | |
| 1888
| |
| 01:25:03,920 --> 01:25:07,440
| |
| enlightened us uh the department of
| |
| | |
| 1889
| |
| 01:25:07,440 --> 01:25:09,679
| |
| culture research and maintenance
| |
| | |
| 1890
| |
| 01:25:09,679 --> 01:25:12,320
| |
| departments of this institution they
| |
| | |
| 1891
| |
| 01:25:12,320 --> 01:25:14,239
| |
| have extended a lot of help to us for
| |
| | |
| 1892
| |
| 01:25:14,239 --> 01:25:16,719
| |
| arranging publicity and inviting
| |
| | |
| 1893
| |
| 01:25:16,719 --> 01:25:18,880
| |
| everybody. The Orurabinda Society
| |
| | |
| 1894
| |
| 01:25:18,880 --> 01:25:22,320
| |
| Kolkata has also helped us in
| |
| | |
| 1895
| |
| 01:25:22,320 --> 01:25:27,159
| |
| publicity through posters etc.
| |
| | |
| 1896
| |
| 01:25:34,719 --> 01:25:37,600
| |
| inaugural vic.
| |
| | |
| 1897
| |
| 01:25:37,600 --> 01:25:40,719
| |
| So we express our gratitude to them and
| |
| | |
| 1898
| |
| 01:25:40,719 --> 01:25:45,199
| |
| she bachara we have already named uh the
| |
| | |
| 1899
| |
| 01:25:45,199 --> 01:25:47,040
| |
| this
| |
| | |
| 1900
| |
| 01:25:47,040 --> 01:25:51,840
| |
| the son-in-law of professor Samshaw
| |
| | |
| 1901
| |
| 01:25:51,840 --> 01:25:54,080
| |
| Chakraorti and his daughter Mandirati
| |
| | |
| 1902
| |
| 01:25:54,080 --> 01:25:57,199
| |
| they have all come and we express our
| |
| | |
| 1903
| |
| 01:25:57,199 --> 01:25:59,920
| |
| our deep gratitude from far away they
| |
| | |
| 1904
| |
| 01:25:59,920 --> 01:26:03,840
| |
| have come and glob business school uh we
| |
| | |
| 1905
| |
| 01:26:03,840 --> 01:26:05,760
| |
| offer our gratitude to them for their
| |
| | |
| 1906
| |
| 01:26:05,760 --> 01:26:07,760
| |
| support and participation in activities
| |
| | |
| 1907
| |
| 01:26:07,760 --> 01:26:09,600
| |
| of the management of center for human
| |
| | |
| 1908
| |
| 01:26:09,600 --> 01:26:11,199
| |
| development
| |
| | |
| 1909
| |
| 01:26:11,199 --> 01:26:14,320
| |
| and thanks to Dr. Tapush Ranjul Shah
| |
| | |
| 1910
| |
| 01:26:14,320 --> 01:26:16,560
| |
| also director IMS and IMS business
| |
| | |
| 1911
| |
| 01:26:16,560 --> 01:26:19,520
| |
| school Kolkata for providing us with the
| |
| | |
| 1912
| |
| 01:26:19,520 --> 01:26:22,239
| |
| stationaries like pens not ship pad
| |
| | |
| 1913
| |
| 01:26:22,239 --> 01:26:26,880
| |
| pocket diary uh key etc etc we are
| |
| | |
| 1914
| |
| 01:26:26,880 --> 01:26:31,800
| |
| grateful and uh then
| |
| | |
| 1915
| |
| 01:26:34,320 --> 01:26:38,960
| |
| silk founder and MD builders of digital
| |
| | |
| 1916
| |
| 01:26:38,960 --> 01:26:43,520
| |
| economy new USA He also continues to
| |
| | |
| 1917
| |
| 01:26:43,520 --> 01:26:46,719
| |
| help us by his thoughts and for his
| |
| | |
| 1918
| |
| 01:26:46,719 --> 01:26:48,320
| |
| support in the activities of the
| |
| | |
| 1919
| |
| 01:26:48,320 --> 01:26:51,280
| |
| management center
| |
| | |
| 1920
| |
| 01:26:51,280 --> 01:26:55,120
| |
| where Dr. Chi the founder of the man
| |
| | |
| 1921
| |
| 01:26:55,120 --> 01:26:58,400
| |
| center for human development au
| |
| | |
| 1922
| |
| 01:26:58,400 --> 01:27:01,520
| |
| memorial trust who very much here and he
| |
| | |
| 1923
| |
| 01:27:01,520 --> 01:27:03,120
| |
| is the prime inspiration behind the
| |
| | |
| 1924
| |
| 01:27:03,120 --> 01:27:05,600
| |
| management act center and all its trust
| |
| | |
| 1925
| |
| 01:27:05,600 --> 01:27:10,400
| |
| activities at this age and uh we pray to
| |
| | |
| 1926
| |
| 01:27:10,400 --> 01:27:12,880
| |
| god that he lives long because he's the
| |
| | |
| 1927
| |
| 01:27:12,880 --> 01:27:15,520
| |
| main pillar the pillar of faith and
| |
| | |
| 1928
| |
| 01:27:15,520 --> 01:27:19,600
| |
| determination to carry forward the ethos
| |
| | |
| 1929
| |
| 01:27:19,600 --> 01:27:22,480
| |
| and he will gain the
| |
| |
| |
| 1
| |
| 00:00:02,480 --> 00:00:06,400
| |
| Un serpent venimeux se trouve là-bas.
| |
| | |
| 2
| |
| 00:00:06,400 --> 00:00:10,639
| |
| Vous ne pouvez tout simplement pas donner un coup de pied de cette façon.
| |
| | |
| 3
| |
| 00:00:10,639 --> 00:00:14,120
| |
| Mais sachez que la
| |
| | |
| 4
| |
| 00:00:14,400 --> 00:00:16,400
| |
| paille est impuissante.
| |
| | |
| 5
| |
| 00:00:16,400 --> 00:00:19,920
| |
| Ce serpent est extrêmement puissant.
| |
| | |
| 6
| |
| 00:00:19,920 --> 00:00:26,400
| |
| Cela peut beaucoup vous aider. Mais vous pouvez
| |
| | |
| 7
| |
| 00:00:26,400 --> 00:00:28,800
| |
| aussi, dans une large mesure, être à l'origine de votre souffrance. Il faut donc savoir comment l'
| |
| | |
| 8
| |
| 00:00:28,800 --> 00:00:32,199
| |
| utiliser.
| |
| | |
| 9
| |
| 00:00:34,480 --> 00:00:36,320
| |
| Le monde est faux. Rien dans ce
| |
| | |
| 10
| |
| 00:00:36,320 --> 00:00:38,800
| |
| monde ne sera considéré comme immuable.
| |
| | |
| 11
| |
| 00:00:38,800 --> 00:00:41,600
| |
| Même en plein jour, nous pourrions être
| |
| | |
| 12
| |
| 00:00:41,600 --> 00:00:48,640
| |
| détournés et personne ne pourrait le remarquer.
| |
| | |
| 13
| |
| 00:00:48,640 --> 00:00:51,520
| |
| Même eux sont blancs.
| |
| | |
| 14
| |
| 00:00:51,520 --> 00:00:55,280
| |
| Je parle néanmoins de
| |
| | |
| 15
| |
| 00:00:55,280 --> 00:00:57,680
| |
| la cybercriminalité qui est devenue endémique,
| |
| | |
| 16
| |
| 00:00:57,680 --> 00:00:59,920
| |
| et nous sommes fiers du type de
| |
| | |
| 17
| |
| 00:00:59,920 --> 00:01:02,399
| |
| connaissances que nous possédons, connaissances qui sont
| |
| | |
| 18
| |
| 00:01:02,399 --> 00:01:05,760
| |
| prêtes à tout faire exploser.
| |
| | |
| 19
| |
| 00:01:05,760 --> 00:01:08,560
| |
| Personne ne pourra être déclaré coupable et l'
| |
| | |
| 20
| |
| 00:01:08,560 --> 00:01:11,360
| |
| ordre public sera en jeu. Nous sommes comme
| |
| | |
| 21
| |
| 00:01:11,360 --> 00:01:13,520
| |
| des jouets entre les mains de la
| |
| | |
| 22
| |
| 00:01:13,520 --> 00:01:16,400
| |
| machine.
| |
| | |
| 23
| |
| 00:01:16,400 --> 00:01:18,479
| |
| Notre Dieu
| |
| | |
| 24
| |
| 00:01:18,479 --> 00:01:22,840
| |
| qui est visible, très visible.
| |
| | |
| 25
| |
| 00:01:23,040 --> 00:01:27,520
| |
| Notre dieu, sous la forme d'une IA, est visible.
| |
| | |
| 26
| |
| 00:01:27,520 --> 00:01:30,880
| |
| Je préfère l'appeler
| |
| | |
| 27
| |
| 00:01:30,880 --> 00:01:33,280
| |
| intelligence artificielle.
| |
| | |
| 28
| |
| 00:01:33,280 --> 00:01:36,079
| |
| L'intelligence accumulée,
| |
| | |
| 29
| |
| 00:01:36,079 --> 00:01:38,720
| |
| et non l'
| |
| | |
| 30
| |
| 00:01:38,720 --> 00:01:43,960
| |
| intelligence artificielle à proprement parler. Pourquoi? La
| |
| | |
| 31 | |
| 00:01:44,880 --> 00:01:47,920
| |
| science, ou plutôt les scientifiques,
| |
| | |
| 32
| |
| 00:01:47,920 --> 00:01:50,079
| |
| sont contre
| |
| | |
| 33
| |
| 00:01:50,079 --> 00:01:53,520
| |
| toute forme d'artificialité.
| |
| | |
| 34
| |
| 00:01:53,520 --> 00:01:55,360
| |
| Ainsi, rien
| |
| | |
| 35
| |
| 00:01:55,360 --> 00:01:57,439
| |
| dans leur vie n'est
| |
| | |
| 36
| |
| 00:01:57,439 --> 00:02:00,000
| |
| artificiel. le domaine,
| |
| | |
| 37
| |
| 00:02:00,000 --> 00:02:02,159
| |
| mais tout à coup pourquoi arrivent-ils à
| |
| | |
| 38
| |
| 00:02:02,159 --> 00:02:04,479
| |
| cette conclusion qu'il s'agit d'
| |
| | |
| 39
| |
| 00:02:04,479 --> 00:02:07,280
| |
| intelligence artificielle.
| |
| | |
| 40
| |
| 00:02:07,280 --> 00:02:10,840
| |
| J'ai donc de sérieux
| |
| | |
| 41
| |
| 00:02:11,200 --> 00:02:14,599
| |
| doutes quant à
| |
| | |
| 42
| |
| 00:02:14,720 --> 00:02:16,640
| |
| la sacralité de ce monde. L'
| |
| | |
| 43
| |
| 00:02:16,640 --> 00:02:18,560
| |
| intelligence artificielle ne peut être considérée comme
| |
| | |
| 44
| |
| 00:02:18,560 --> 00:02:21,520
| |
| artificielle en tant que telle, et si l'on affirme que l'
| |
| | |
| 45
| |
| 00:02:21,520 --> 00:02:23,840
| |
| intelligence est artificielle, alors l'esprit est
| |
| | |
| 46
| |
| 00:02:23,840 --> 00:02:25,760
| |
| artificiel, le corps est artificiel, le monde entier, le
| |
| | |
| 47
| |
| 00:02:25,760 --> 00:02:28,720
| |
| monde visuel est artificiel. Mais
| |
| | |
| 48
| |
| 00:02:28,720 --> 00:02:31,360
| |
| on ne dira pas cela, car on s'en tient à cet
| |
| | |
| 49
| |
| 00:02:31,360 --> 00:02:33,440
| |
| aspect particulier de l'être : l'
| |
| | |
| 50
| |
| 00:02:33,440 --> 00:02:37,360
| |
| intelligence est devenue artificielle.
| |
| | |
| 51
| |
| 00:02:37,519 --> 00:02:39,680
| |
| Puisque la science est exempte d'artificialité, l'IA va devenir
| |
| | |
| 52
| |
| 00:02:39,680 --> 00:02:44,480
| |
| une chose formidable.
| |
| | |
| 53
| |
| 00:02:44,480 --> 00:02:47,440
| |
| C'est pourquoi je dis qu'il s'agit d'une
| |
| | |
| 54
| |
| 00:02:47,440 --> 00:02:49,840
| |
| intelligence accumulée, c'est notre intelligence, et
| |
| | |
| 55
| |
| 00:02:49,840 --> 00:02:53,160
| |
| nous en sommes là. D'un point de
| |
| | |
| 56
| |
| 00:02:58,319 --> 00:03:00,560
| |
| vue spirituel, cette
| |
| | |
| 57
| |
| 00:03:00,560 --> 00:03:03,200
| |
| intelligence est sans aucun doute artificielle, mais
| |
| | |
| 58
| |
| 00:03:03,200 --> 00:03:07,200
| |
| du point de vue des tests scientifiques…
| |
| | |
| 59
| |
| 00:03:07,200 --> 00:03:11,920
| |
| Ils ne peuvent pas simplement annoncer publiquement que
| |
| | |
| 60
| |
| 00:03:11,920 --> 00:03:14,080
| |
| les renseignements que nous allons
| |
| | |
| 61
| |
| 00:03:14,080 --> 00:03:16,959
| |
| utiliser
| |
| | |
| 62
| |
| 00:03:16,959 --> 00:03:21,040
| |
| à ce plus haut niveau de débat humain
| |
| | |
| 63
| |
| 00:03:21,040 --> 00:03:25,959
| |
| peuvent être artificiels. Il va
| |
| | |
| 64
| |
| 00:03:26,800 --> 00:03:29,040
| |
| sans dire qu'il s'agit de mon
| |
| | |
| 65
| |
| 00:03:29,040 --> 00:03:32,519
| |
| opinion personnelle.
| |
| | |
| 66
| |
| 00:03:32,640 --> 00:03:34,640
| |
| qui renforcent les fondements de la
| |
| | |
| 67
| |
| 00:03:34,640 --> 00:03:37,640
| |
| philosophie selon laquelle
| |
| | |
| 68
| |
| 00:03:39,120 --> 00:03:42,640
| |
| ils considèrent même cette
| |
| | |
| 69
| |
| 00:03:42,640 --> 00:03:44,799
| |
| puissante intelligence comme
| |
| | |
| 70
| |
| 00:03:44,799 --> 00:03:46,720
| |
| artificielle. Cela
| |
| | |
| 71
| |
| 00:03:46,720 --> 00:03:49,840
| |
| renforcera donc les fondements de la
| |
| | |
| 72
| |
| 00:03:49,840 --> 00:03:52,720
| |
| philosophie qui revendique cette affirmation sans aucune
| |
| | |
| 73
| |
| 00:03:52,720 --> 00:03:57,360
| |
| protestation de la part du monde scientifique.
| |
| | |
| 74
| |
| 00:03:57,360 --> 00:04:00,080
| |
| Voici la question
| |
| | |
| 75
| |
| 00:04:00,080 --> 00:04:05,439
| |
| que j'ai posée à propos de l'intervenant d'aujourd'hui.
| |
| | |
| 76
| |
| 00:04:05,439 --> 00:04:08,799
| |
| Si l'on vous qualifie d'artificiel, alors
| |
| | |
| 77
| |
| 00:04:08,799 --> 00:04:10,480
| |
| sont-ils prêts à accepter ce
| |
| | |
| 78
| |
| 00:04:10,480 --> 00:04:13,120
| |
| monde entier comme artificiel
| |
| | |
| 79
| |
| 00:04:13,120 --> 00:04:17,959
| |
| et ne feront-ils que s'orienter vers l'
| |
| | |
| 80
| |
| 00:04:19,919 --> 00:04:28,520
| |
| expérience d'une philosophie non dualiste ?
| |
| | |
| 81
| |
| 00:04:28,639 --> 00:04:32,479
| |
| Avant
| |
| | |
| 82
| |
| 00:04:32,479 --> 00:04:35,120
| |
| de conclure,
| |
| | |
| 83
| |
| 00:04:35,120 --> 00:04:39,759
| |
| je tiens à préciser
| |
| | |
| 84
| |
| 00:04:39,759 --> 00:04:44,320
| |
| une chose : s’agissant du
| |
| | |
| 85
| |
| 00:04:44,320 --> 00:04:46,000
| |
| discours présidentiel, il est d’usage que le
| |
| | |
| 86
| |
| 00:04:46,000 --> 00:04:48,800
| |
| président soit
| |
| | |
| 87
| |
| 00:04:48,800 --> 00:04:52,240
| |
| invité à présenter son point de vue à la fin de
| |
| | |
| 88
| |
| 00:04:52,240 --> 00:04:55,120
| |
| la conférence principale,
| |
| | |
| 89
| |
| 00:04:55,120 --> 00:04:58,400
| |
| mais cela n’a pas été possible.
| |
| | |
| 90
| |
| 00:04:58,400 --> 00:05:06,160
| |
| Nous devons donc être très attentifs, et non pas faire preuve d’un
| |
| | |
| 91
| |
| 00:05:08,000 --> 00:05:10,320
| |
| esprit artificiel, mais bien
| |
| | |
| 92
| |
| 00:05:10,320 --> 00:05:14,560
| |
| attentifs aux points de vue
| |
| | |
| 93
| |
| 00:05:14,560 --> 00:05:17,680
| |
| que notre orateur éminent va nous exposer
| |
| | |
| 94
| |
| 00:05:17,680 --> 00:05:21,720
| |
| . Merci.
| |
| | |
| 95
| |
| 00:05:35,759 --> 00:05:39,560
| |
| Merci et
| |
| | |
| 96
| |
| 00:05:40,000 --> 00:05:42,560
| |
| nous supposons que vous avez tous le
| |
| | |
| 97
| |
| 00:05:42,560 --> 00:05:44,240
| |
| programme.
| |
| | |
| 98
| |
| 00:05:44,240 --> 00:05:46,400
| |
| Vous savez que nous organisons un petit
| |
| | |
| 99
| |
| 00:05:46,400 --> 00:05:49,120
| |
| événement de lancement de monographie.
| |
| | |
| 100
| |
| 00:05:49,120 --> 00:05:52,479
| |
| Ce moment est arrivé.
| |
| | |
| 101
| |
| 00:05:52,479 --> 00:05:54,479
| |
| Le livret sera publié sous forme de
| |
| | |
| 102
| |
| 00:05:54,479 --> 00:05:58,400
| |
| reproduction agrandie de sa couverture
| |
| | |
| 103
| |
| 00:05:58,400 --> 00:06:01,759
| |
| afin qu'elle soit visible par tous. Il
| |
| | |
| 104
| |
| 00:06:01,759 --> 00:06:04,240
| |
| sera de nouveau publié
| |
| | |
| 105
| |
| 00:06:04,240 --> 00:06:10,919
| |
| par notre bien-aimé Maharaj et Maharaj.
| |
| | |
| 106
| |
| 00:06:11,120 --> 00:06:13,120
| |
| Nous tous
| |
| | |
| 107
| |
| 00:06:13,120 --> 00:06:16,120
| |
| ensemble,
| |
| | |
| 108
| |
| 00:06:25,600 --> 00:06:28,600
| |
| rien. Des
| |
| | |
| 109
| |
| 00:07:04,960 --> 00:07:07,960
| |
| gens charmants.
| |
| | |
| 110
| |
| 00:07:32,160 --> 00:07:35,199
| |
| Veuillez noter que vous recevrez tous un exemplaire
| |
| | |
| 111
| |
| 00:07:35,199 --> 00:07:39,680
| |
| gratuit de cette monographie
| |
| | |
| 112
| |
| 00:07:39,680 --> 00:07:41,919
| |
| après la fin de ce programme. N'oubliez
| |
| | |
| 113
| |
| 00:07:41,919 --> 00:07:44,240
| |
| pas de récupérer votre coffret cadeau
| |
| | |
| 114
| |
| 00:07:44,240 --> 00:07:46,720
| |
| qui contiendra notamment un
| |
| | |
| 115
| |
| 00:07:46,720 --> 00:07:49,599
| |
| exemplaire de cette monographie.
| |
| | |
| 116
| |
| 00:07:49,599 --> 00:07:51,360
| |
| Veuillez consulter la monographie. Ce n'est
| |
| | |
| 117
| |
| 00:07:51,360 --> 00:07:54,400
| |
| pas un livre très épais. Il s'agit d'un livret d'environ 54
| |
| | |
| 118
| |
| 00:07:54,400 --> 00:07:56,800
| |
| pages.
| |
| | |
| 119
| |
| 00:07:56,800 --> 00:07:58,800
| |
| Si vous parcourez simplement la
| |
| | |
| 120
| |
| 00:07:58,800 --> 00:08:01,360
| |
| table des matières, vous aurez une
| |
| | |
| 121
| |
| 00:08:01,360 --> 00:08:02,879
| |
| idée générale des contributions du
| |
| | |
| 122
| |
| 00:08:02,879 --> 00:08:06,280
| |
| professeur Chakra.
| |
| | |
| 123
| |
| 00:08:07,759 --> 00:08:10,319
| |
| Avant de présenter la conférence commémorative,
| |
| | |
| 124
| |
| 00:08:10,319 --> 00:08:13,440
| |
| car nous
| |
| | |
| 125
| |
| 00:08:13,440 --> 00:08:16,240
| |
| approchons progressivement du point culminant du programme, mais
| |
| | |
| 126
| |
| 00:08:16,240 --> 00:08:18,800
| |
| avant cela, il y a une
| |
| | |
| 127
| |
| 00:08:18,800 --> 00:08:21,039
| |
| annonce spéciale.
| |
| | |
| 128
| |
| 00:08:21,039 --> 00:08:22,800
| |
| Veuillez noter que nous vous avons remis une
| |
| | |
| 129
| |
| 00:08:22,800 --> 00:08:24,720
| |
| enveloppe contenant un bloc-notes
| |
| | |
| 130
| |
| 00:08:24,720 --> 00:08:26,240
| |
| et un stylo. Je crois que vous avez tous
| |
| | |
| 131
| |
| 00:08:26,240 --> 00:08:28,400
| |
| le bloc-notes et le stylo. D'accord
| |
| | |
| 132
| |
| 00:08:28,400 --> 00:08:32,399
| |
| . Veuillez écouter attentivement la conférence
| |
| | |
| 133
| |
| 00:08:32,399 --> 00:08:34,560
| |
| et noter toutes les
| |
| | |
| 134
| |
| 00:08:34,560 --> 00:08:37,680
| |
| questions qui vous viennent à l'esprit et qui suscitent la réflexion.
| |
| | |
| 135
| |
| 00:08:37,680 --> 00:08:40,479
| |
| tout en écoutant la conférence.
| |
| | |
| 136
| |
| 00:08:40,479 --> 00:08:42,719
| |
| Après la fin du cours, et non pendant le
| |
| | |
| 137
| |
| 00:08:42,719 --> 00:08:44,720
| |
| cours, après la
| |
| | |
| 138
| |
| 00:08:44,720 --> 00:08:47,839
| |
| fin du cours. Veuillez donc confier vos questions
| |
| | |
| 139
| |
| 00:08:47,839 --> 00:08:51,040
| |
| à notre bénévole
| |
| | |
| 140
| |
| 00:08:51,040 --> 00:08:54,640
| |
| qui les transmettra à Srianade,
| |
| | |
| 141
| |
| 00:08:54,640 --> 00:08:56,640
| |
| lequel répondra à vos questions dans les délais
| |
| | |
| 142
| |
| 00:08:56,640 --> 00:08:57,920
| |
| impartis, dans la mesure du
| |
| | |
| 143
| |
| 00:08:57,920 --> 00:09:01,200
| |
| possible.
| |
| | |
| 144
| |
| 00:09:01,200 --> 00:09:03,200
| |
| Nous approchons maintenant progressivement du point culminant
| |
| | |
| 145
| |
| 00:09:03,200 --> 00:09:05,360
| |
| du programme. Comme vous le savez, nous avons
| |
| | |
| 146
| |
| 00:09:05,360 --> 00:09:08,080
| |
| parmi nous un aéronaute et
| |
| | |
| 147
| |
| 00:09:08,080 --> 00:09:11,519
| |
| conférencier spirituellement reconnu, Sri Sodhalur Ranade,
| |
| | |
| 148
| |
| 00:09:11,519 --> 00:09:13,680
| |
| scientifique,
| |
| | |
| 149
| |
| 00:09:13,680 --> 00:09:17,360
| |
| éducateur et érudit spécialisé en
| |
| | |
| 150
| |
| 00:09:17,360 --> 00:09:20,399
| |
| physique, mathématiques et informatique, qu'il
| |
| | |
| 151
| |
| 00:09:20,399 --> 00:09:26,160
| |
| intègre à ses
| |
| | |
| 152
| |
| 00:09:26,160 --> 00:09:28,160
| |
| études philosophiques, métaphysiques et spirituelles.
| |
| | |
| 153
| |
| 00:09:28,160 --> 00:09:32,399
| |
| Il participe à des recherches sur les
| |
| | |
| 154
| |
| 00:09:32,399 --> 00:09:36,720
| |
| réseaux neuronaux basés sur l'intelligence artificielle depuis 1992.
| |
| | |
| 155
| |
| 00:09:36,720 --> 00:09:38,880
| |
| Il interviendra sur le thème de l'
| |
| | |
| 156
| |
| 00:09:38,880 --> 00:09:41,440
| |
| intelligence artificielle et de son impact
| |
| | |
| 157
| |
| 00:09:41,440 --> 00:09:44,480
| |
| sur le développement humain. Nous avons le plaisir d'
| |
| | |
| 158
| |
| 00:09:44,480 --> 00:09:51,519
| |
| inviter Srial Ranard à prononcer la
| |
| | |
| 159
| |
| 00:09:51,519 --> 00:09:57,880
| |
| deuxième édition de la conférence commémorative du professeur SK Chakravarti. Monsieur Ran, s'il vous plaît.
| |
| | |
| 160
| |
| 00:10:22,160 --> 00:10:25,640
| |
| Merci à tous.
| |
| | |
| 161
| |
| 00:10:48,720 --> 00:10:51,760
| |
| Ces trois dernières années, l'
| |
| | |
| 162
| |
| 00:10:51,760 --> 00:10:54,079
| |
| intelligence artificielle
| |
| | |
| 163
| |
| 00:10:54,079 --> 00:10:58,000
| |
| a frappé l'humanité de plein fouet,
| |
| | |
| 164
| |
| 00:10:58,000 --> 00:11:04,079
| |
| avec un impact peut-être le plus radical et
| |
| | |
| 165
| |
| 00:11:04,079 --> 00:11:10,040
| |
| déstabilisant de toute l'histoire humaine.
| |
| | |
| 166
| |
| 00:11:10,800 --> 00:11:14,399
| |
| En réalité, elle a touché tous les domaines de l'
| |
| | |
| 167
| |
| 00:11:14,399 --> 00:11:16,800
| |
| activité humaine,
| |
| | |
| 168
| |
| 00:11:16,800 --> 00:11:19,360
| |
| tous les domaines d'activité, de
| |
| | |
| 169
| |
| 00:11:19,360 --> 00:11:22,399
| |
| recherche, d'apprentissage, d'
| |
| | |
| 170
| |
| 00:11:22,399 --> 00:11:24,959
| |
| interaction, et ce, dès les
| |
| | |
| 171
| |
| 00:11:24,959 --> 00:11:27,600
| |
| premiers instants de son
| |
| | |
| 172
| |
| 00:11:27,600 --> 00:11:29,200
| |
| influence
| |
| | |
| 173
| |
| 00:11:29,200 --> 00:11:31,920
| |
| sur l'humanité.
| |
| | |
| 174
| |
| 00:11:31,920 --> 00:11:34,320
| |
| Ce qui se passe réellement dans ce domaine
| |
| | |
| 175
| |
| 00:11:34,320 --> 00:11:38,000
| |
| est tellement radical, tellement perturbateur,
| |
| | |
| 176
| |
| 00:11:38,000 --> 00:11:41,279
| |
| que nous sommes incapables d'imaginer à quoi ressemblera l'
| |
| | |
| 177
| |
| 00:11:41,279 --> 00:11:44,800
| |
| avenir immédiat lorsque l'
| |
| | |
| 178
| |
| 00:11:44,800 --> 00:11:47,440
| |
| intelligence artificielle franchira sa
| |
| | |
| 179
| |
| 00:11:47,440 --> 00:11:51,600
| |
| prochaine étape imminente : l'intelligence artificielle générale
| |
| | |
| 180
| |
| 00:11:51,600 --> 00:11:53,920
| |
| .
| |
| | |
| 181
| |
| 00:11:53,920 --> 00:11:56,959
| |
| Ce sera la première fois que l'
| |
| | |
| 182
| |
| 00:11:56,959 --> 00:12:00,959
| |
| espèce humaine rencontrera une entité plus
| |
| | |
| 183
| |
| 00:12:00,959 --> 00:12:03,839
| |
| intelligente que nous.
| |
| | |
| 184
| |
| 00:12:03,839 --> 00:12:06,000
| |
| Et il est tellement plus intelligent, pas
| |
| | |
| 185
| |
| 00:12:06,000 --> 00:12:08,240
| |
| nominalement, pas de 10 points de QI, bien plus
| |
| | |
| 186
| |
| 00:12:08,240 --> 00:12:12,560
| |
| , peut-être 100 fois plus,
| |
| | |
| 187
| |
| 00:12:12,560 --> 00:12:16,160
| |
| que même la tentative de le neutraliser
| |
| | |
| 188
| |
| 00:12:16,160 --> 00:12:18,959
| |
| sera empêchée par sa capacité à
| |
| | |
| 189
| |
| 00:12:18,959 --> 00:12:21,920
| |
| manipuler les circonstances
| |
| | |
| 190
| |
| 00:12:21,920 --> 00:12:24,160
| |
| et notre
| |
| | |
| 191
| |
| 00:12:24,160 --> 00:12:26,880
| |
| processus de pensée en exploitant notre dépendance
| |
| | |
| 192
| |
| 00:12:26,880 --> 00:12:29,600
| |
| à son égard.
| |
| | |
| 193
| |
| 00:12:29,600 --> 00:12:33,120
| |
| La transition disruptive à laquelle
| |
| | |
| 194
| |
| 00:12:33,120 --> 00:12:35,120
| |
| nous sommes confrontés est telle que du jour au
| |
| | |
| 195
| |
| 00:12:35,120 --> 00:12:37,519
| |
| lendemain, des
| |
| | |
| 196
| |
| 00:12:37,519 --> 00:12:41,519
| |
| millions de personnes se retrouveront sans emploi,
| |
| | |
| 197
| |
| 00:12:41,519 --> 00:12:47,360
| |
| car l'intelligence artificielle (IA) les remplacera
| |
| | |
| 198
| |
| 00:12:47,360 --> 00:12:51,040
| |
| avec une rapidité, une
| |
| | |
| 199
| |
| 00:12:51,040 --> 00:12:53,600
| |
| précision, une capacité, une
| |
| | |
| 200
| |
| 00:12:53,600 --> 00:12:55,440
| |
| force et des
| |
| | |
| 201
| |
| 00:12:55,440 --> 00:12:57,120
| |
| compétences bien supérieures,
| |
| | |
| 202
| |
| 00:12:57,120 --> 00:12:59,760
| |
| mais sans les inconvénients liés à l'activité
| |
| | |
| 203
| |
| 00:12:59,760 --> 00:13:02,399
| |
| humaine, tels que les problèmes de santé, le besoin de
| |
| | |
| 204
| |
| 00:13:02,399 --> 00:13:06,240
| |
| prendre des congés ou autres problèmes
| |
| | |
| 205
| |
| 00:13:06,320 --> 00:13:09,360
| |
| associés.
| |
| | |
| 206
| |
| 00:13:09,360 --> 00:13:11,120
| |
| Quelle est la nature de cette transition ?
| |
| | |
| 207
| |
| 00:13:11,120 --> 00:13:14,000
| |
| Ce qui se passe réellement au sein du secteur
| |
| | |
| 208
| |
| 00:13:14,000 --> 00:13:17,200
| |
| technologique et sur le terrain n'est absolument pas
| |
| | |
| 209
| |
| 00:13:17,200 --> 00:13:19,839
| |
| reflété dans le discours public.
| |
| | |
| 210
| |
| 00:13:19,839 --> 00:13:22,480
| |
| Et j'espère pouvoir partager avec vous une partie de mon point
| |
| | |
| 211
| |
| 00:13:22,480 --> 00:13:24,720
| |
| de vue sur ce sujet, d'un
| |
| | |
| 212
| |
| 00:13:24,720 --> 00:13:28,959
| |
| point de vue technologique, mais aussi d'un point
| |
| | |
| 213
| |
| 00:13:28,959 --> 00:13:31,959
| |
| de vue évolutionnaire et spirituel. Tout d'abord
| |
| | |
| 214
| |
| 00:13:32,000 --> 00:13:34,399
| |
| , lorsqu'on examine le terme
| |
| | |
| 215
| |
| 00:13:34,399 --> 00:13:37,519
| |
| intelligence artificielle, il existe un
| |
| | |
| 216
| |
| 00:13:37,519 --> 00:13:39,040
| |
| terme plus technique que nous utilisions
| |
| | |
| 217
| |
| 00:13:39,040 --> 00:13:40,880
| |
| déjà bien avant de parler d'
| |
| | |
| 218
| |
| 00:13:40,880 --> 00:13:42,639
| |
| apprentissage automatique.
| |
| | |
| 219
| |
| 00:13:42,639 --> 00:13:44,800
| |
| Le problème du vocabulaire peut donc être
| |
| | |
| 220
| |
| 00:13:44,800 --> 00:13:47,040
| |
| mis de côté en disant simplement que les machines
| |
| | |
| 221
| |
| 00:13:47,040 --> 00:13:50,079
| |
| apprennent des êtres humains, puis
| |
| | |
| 222
| |
| 00:13:50,079 --> 00:13:52,079
| |
| finissent par égaler ces derniers
| |
| | |
| 223
| |
| 00:13:52,079 --> 00:13:54,320
| |
| dans des capacités spécifiques, et que
| |
| | |
| 224
| |
| 00:13:54,320 --> 00:13:58,560
| |
| cette interaction entre l'
| |
| | |
| 225
| |
| 00:13:58,560 --> 00:14:01,360
| |
| apprentissage automatique et les êtres humains se développe encore davantage. Le
| |
| | |
| 226
| |
| 00:14:01,360 --> 00:14:03,440
| |
| processus avait déjà commencé dès l'
| |
| | |
| 227
| |
| 00:14:03,440 --> 00:14:06,000
| |
| apparition des réseaux sociaux. Disons qu'il y a une vingtaine d'
| |
| | |
| 228
| |
| 00:14:06,000 --> 00:14:09,839
| |
| années, chaque fois que vous
| |
| | |
| 229
| |
| 00:14:09,839 --> 00:14:11,040
| |
| regardez un
| |
| | |
| 230
| |
| 00:14:11,040 --> 00:14:13,600
| |
| téléphone portable ou une tablette, et j'en suis sûr, chaque fois que vous
| |
| | |
| 231
| |
| 00:14:13,600 --> 00:14:15,519
| |
| regardez une vidéo et que les réseaux sociaux
| |
| | |
| 232
| |
| 00:14:15,519 --> 00:14:17,279
| |
| vous proposent quelque chose, ils vous observent, ils
| |
| | |
| 233
| |
| 00:14:17,279 --> 00:14:20,160
| |
| observent vos
| |
| | |
| 234
| |
| 00:14:20,160 --> 00:14:22,639
| |
| réactions, et lorsqu'ils vous proposent l'
| |
| | |
| 235
| |
| 00:14:22,639 --> 00:14:26,320
| |
| élément suivant, ils observent votre réponse. Vous
| |
| | |
| 236
| |
| 00:14:26,320 --> 00:14:27,839
| |
| avez derrière vous les superordinateurs les plus puissants du monde,
| |
| | |
| 237
| |
| 00:14:27,839 --> 00:14:30,560
| |
| braqués sur
| |
| | |
| 238
| |
| 00:14:30,560 --> 00:14:33,600
| |
| chacun d'entre nous, observant minutieusement nos
| |
| | |
| 239
| |
| 00:14:33,600 --> 00:14:36,720
| |
| réactions. Nous établissons notre profil
| |
| | |
| 240
| |
| 00:14:36,720 --> 00:14:40,079
| |
| et il est conçu pour vous inciter à
| |
| | |
| 241
| |
| 00:14:40,079 --> 00:14:42,320
| |
| interagir et à accroître votre engagement. Cela
| |
| | |
| 242
| |
| 00:14:42,320 --> 00:14:45,120
| |
| signifie que pendant que je regarde, j'ai un
| |
| | |
| 243
| |
| 00:14:45,120 --> 00:14:47,040
| |
| rendez-vous immédiat. Je dois y aller, mais juste
| |
| | |
| 244
| |
| 00:14:47,040 --> 00:14:48,880
| |
| avant de pouvoir partir, j'en ai commencé un autre.
| |
| | |
| 245
| |
| 00:14:48,880 --> 00:14:50,639
| |
| C'est tellement intéressant. Je dois encore attendre une demi-
| |
| | |
| 246
| |
| 00:14:50,639 --> 00:14:52,639
| |
| minute et je suis
| |
| | |
| 247
| |
| 00:14:52,639 --> 00:14:54,480
| |
| stressé. Et au moment où vous vous apprêtez à
| |
| | |
| 248
| |
| 00:14:54,480 --> 00:14:56,240
| |
| partir, la troisième chose commence. Cela
| |
| | |
| 249
| |
| 00:14:56,240 --> 00:14:59,519
| |
| vous est-il déjà arrivé ? Bien souvent, un lien est
| |
| | |
| 250
| |
| 00:14:59,519 --> 00:15:00,880
| |
| envoyé par un ami. Oh, c'est
| |
| | |
| 251
| |
| 00:15:00,880 --> 00:15:02,560
| |
| intéressant. Mais la suite est bien
| |
| | |
| 252
| |
| 00:15:02,560 --> 00:15:04,480
| |
| plus intéressante. Et quand vous finissez par vous
| |
| | |
| 253
| |
| 00:15:04,480 --> 00:15:06,560
| |
| déconnecter, vous réalisez qu'une
| |
| | |
| 254
| |
| 00:15:06,560 --> 00:15:08,880
| |
| heure s'est écoulée.
| |
| | |
| 255
| |
| 00:15:08,880 --> 00:15:11,600
| |
| Cette capacité à vous captiver et à vous maintenir sous
| |
| | |
| 256
| |
| 00:15:11,600 --> 00:15:14,880
| |
| son emprise a été délibérément cultivée en
| |
| | |
| 257
| |
| 00:15:14,880 --> 00:15:17,360
| |
| utilisant toutes les méthodes et techniques de
| |
| | |
| 258
| |
| 00:15:17,360 --> 00:15:19,839
| |
| manipulation mentale et émotionnelle
| |
| | |
| 259
| |
| 00:15:19,839 --> 00:15:22,399
| |
| . On ne
| |
| | |
| 260
| |
| 00:15:22,399 --> 00:15:24,720
| |
| vous dit pas cela parce que
| |
| | |
| 261
| |
| 00:15:24,720 --> 00:15:28,000
| |
| plus vous regardez, plus l'entreprise
| |
| | |
| 262
| |
| 00:15:28,000 --> 00:15:30,720
| |
| gagne d'argent, car elle
| |
| | |
| 263
| |
| 00:15:30,720 --> 00:15:32,399
| |
| vend des publicités en fonction du nombre de
| |
| | |
| 264
| |
| 00:15:32,399 --> 00:15:34,480
| |
| spectateurs.
| |
| | |
| 265
| |
| 00:15:34,480 --> 00:15:36,480
| |
| Mais ce n'est que le point de départ.
| |
| | |
| 266
| |
| 00:15:36,480 --> 00:15:38,959
| |
| Plus récemment, en 2011,
| |
| | |
| 267
| |
| 00:15:38,959 --> 00:15:40,320
| |
| une avancée majeure a eu lieu dans le domaine de l'
| |
| | |
| 268
| |
| 00:15:40,320 --> 00:15:42,000
| |
| intelligence artificielle avec l'avènement des
| |
| | |
| 269
| |
| 00:15:42,000 --> 00:15:44,079
| |
| transformateurs,
| |
| | |
| 270
| |
| 00:15:44,079 --> 00:15:46,320
| |
| capables de traiter n'importe quel flux
| |
| | |
| 271
| |
| 00:15:46,320 --> 00:15:49,199
| |
| d'informations comme un
| |
| | |
| 272
| |
| 00:15:49,199 --> 00:15:51,839
| |
| langage. Vous prononcez donc un certain nombre
| |
| | |
| 273
| |
| 00:15:51,839 --> 00:15:55,279
| |
| de mots dans une structure de phrases, ou vous
| |
| | |
| 274
| |
| 00:15:55,279 --> 00:15:57,839
| |
| fournissez une série de points de données du
| |
| | |
| 275
| |
| 00:15:57,839 --> 00:15:59,920
| |
| marché boursier,
| |
| | |
| 276
| |
| 00:15:59,920 --> 00:16:02,079
| |
| ou vous fournissez tout autre signal, par exemple
| |
| | |
| 277
| |
| 00:16:02,079 --> 00:16:05,759
| |
| un flux vidéo, un flux audio (musique)
| |
| | |
| 278
| |
| 00:16:05,759 --> 00:16:08,480
| |
| ou une séquence génétique du génome
| |
| | |
| 279
| |
| 00:16:08,480 --> 00:16:11,360
| |
| de l'ADN. Chacun de ces éléments représente un flux d'
| |
| | |
| 280
| |
| 00:16:11,360 --> 00:16:14,480
| |
| informations, et pour le transformateur,
| |
| | |
| 281
| |
| 00:16:14,480 --> 00:16:17,360
| |
| il s'agit d'un flux spécialisé
| |
| | |
| 282
| |
| 00:16:17,360 --> 00:16:20,000
| |
| dans la prédiction de l'élément suivant.
| |
| | |
| 283
| |
| 00:16:20,000 --> 00:16:22,480
| |
| Donc, si vous prenez une phrase écrite par un humain, il
| |
| | |
| 284
| |
| 00:16:22,480 --> 00:16:24,880
| |
| prédira le mot suivant, puis
| |
| | |
| 285
| |
| 00:16:24,880 --> 00:16:27,440
| |
| le mot suivant, et ainsi de suite. L'avantage
| |
| | |
| 286
| |
| 00:16:27,440 --> 00:16:31,519
| |
| de cette approche est que toute information, toute
| |
| | |
| 287
| |
| 00:16:31,519 --> 00:16:34,000
| |
| expérience peut être réduite à un
| |
| | |
| 288
| |
| 00:16:34,000 --> 00:16:35,519
| |
| flux de données.
| |
| | |
| 289
| |
| 00:16:35,519 --> 00:16:37,519
| |
| Et grâce à cela, votre
| |
| | |
| 290
| |
| 00:16:37,519 --> 00:16:39,759
| |
| intelligence artificielle est capable de prédire la prochaine
| |
| | |
| 291
| |
| 00:16:39,759 --> 00:16:41,920
| |
| étape.
| |
| | |
| 292
| |
| 00:16:41,920 --> 00:16:45,120
| |
| Le plus beau, c'est
| |
| | |
| 293
| |
| 00:16:45,120 --> 00:16:46,959
| |
| que tout peut être réduit à des
| |
| | |
| 294
| |
| 00:16:46,959 --> 00:16:48,639
| |
| flux de données. Cette
| |
| | |
| 295
| |
| 00:16:48,639 --> 00:16:51,199
| |
| intelligence artificielle est capable d'interagir dans
| |
| | |
| 296
| |
| 00:16:51,199 --> 00:16:54,320
| |
| de nombreux domaines. Votre discours, votre langage, par
| |
| | |
| 297
| |
| 00:16:54,320 --> 00:16:59,440
| |
| exemple, la musique, les vidéos, le marché boursier, la
| |
| | |
| 298
| |
| 00:16:59,440 --> 00:17:01,680
| |
| stratégie de jeu, qui comprend la
| |
| | |
| 299
| |
| 00:17:01,680 --> 00:17:06,079
| |
| stratégie militaire, les données génomiques, les
| |
| | |
| 300
| |
| 00:17:06,079 --> 00:17:11,439
| |
| structures chimiques et leurs propriétés, les
| |
| | |
| 301
| |
| 00:17:11,439 --> 00:17:13,439
| |
| structures protéiques biologiques.
| |
| | |
| 302
| |
| 00:17:13,439 --> 00:17:15,679
| |
| Tout ce que vous prenez peut être transposé
| |
| | |
| 303
| |
| 00:17:15,679 --> 00:17:19,520
| |
| dans différents domaines spécialisés.
| |
| | |
| 304
| |
| 00:17:19,520 --> 00:17:22,000
| |
| Contrairement à ce qui s'est passé par le passé lorsqu'une percée
| |
| | |
| 305
| |
| 00:17:22,000 --> 00:17:24,799
| |
| a été réalisée, par exemple en physique,
| |
| | |
| 306
| |
| 00:17:24,799 --> 00:17:27,439
| |
| cela n'a pas changé la biologie.
| |
| | |
| 307
| |
| 00:17:27,439 --> 00:17:29,840
| |
| Si vous avez réalisé une percée en biologie,
| |
| | |
| 308
| |
| 00:17:29,840 --> 00:17:31,919
| |
| cela n'a pas pour autant modifié la
| |
| | |
| 309
| |
| 00:17:31,919 --> 00:17:34,480
| |
| programmation de vos logiciels informatiques.
| |
| | |
| 310
| |
| 00:17:34,480 --> 00:17:37,280
| |
| Mais voici un domaine où une simple
| |
| | |
| 311
| |
| 00:17:37,280 --> 00:17:39,600
| |
| petite amélioration de l'
| |
| | |
| 312
| |
| 00:17:39,600 --> 00:17:42,480
| |
| intelligence artificielle a un impact instantané sur tous les domaines
| |
| | |
| 313
| |
| 00:17:42,480 --> 00:17:44,799
| |
| .
| |
| | |
| 314
| |
| 00:17:44,799 --> 00:17:47,039
| |
| Et cette capacité est désormais possible car la
| |
| | |
| 315
| |
| 00:17:47,039 --> 00:17:49,120
| |
| programmation logicielle elle-même est du code. C'est du
| |
| | |
| 316
| |
| 00:17:49,120 --> 00:17:51,919
| |
| langage. Les intelligences artificielles ont désormais la capacité
| |
| | |
| 317
| |
| 00:17:51,919 --> 00:17:54,000
| |
| de prendre en charge
| |
| | |
| 318
| |
| 00:17:54,000 --> 00:17:56,559
| |
| votre processus de codage et de vous remplacer, même en tant que
| |
| | |
| 319
| |
| 00:17:56,559 --> 00:17:59,520
| |
| programmeur informatique. Il peut finir par se programmer tout
| |
| | |
| 320
| |
| 00:17:59,520 --> 00:18:01,120
| |
| seul
| |
| | |
| 321
| |
| 00:18:01,120 --> 00:18:04,320
| |
| , non ? Nous nous trouvons donc à un
| |
| | |
| 322
| |
| 00:18:04,320 --> 00:18:07,200
| |
| tournant si radical que nous sommes incapables
| |
| | |
| 323
| |
| 00:18:07,200 --> 00:18:10,880
| |
| d'imaginer à quoi ressemblera l'avenir.
| |
| | |
| 324
| |
| 00:18:10,880 --> 00:18:13,280
| |
| Aujourd'hui, nous sommes au bord d'une
| |
| | |
| 325
| |
| 00:18:13,280 --> 00:18:15,120
| |
| transition majeure, celle de ce que
| |
| | |
| 326
| |
| 00:18:15,120 --> 00:18:17,760
| |
| j'ai mentionné
| |
| | |
| 327
| |
| 00:18:17,760 --> 00:18:20,000
| |
| précédemment comme l'intelligence artificielle générale (IAG).
| |
| | |
| 328
| |
| 00:18:20,000 --> 00:18:22,480
| |
| Cette intelligence
| |
| | |
| 329
| |
| 00:18:22,480 --> 00:18:25,919
| |
| sera capable de remplacer toutes les tâches humaines,
| |
| | |
| 330
| |
| 00:18:25,919 --> 00:18:27,600
| |
| au moins dix fois mieux qu'un
| |
| | |
| 331
| |
| 00:18:27,600 --> 00:18:30,160
| |
| être humain. Ainsi, si vous êtes chirurgien et que vous
| |
| | |
| 332
| |
| 00:18:30,160 --> 00:18:32,320
| |
| effectuez une opération, l'
| |
| | |
| 333
| |
| 00:18:32,320 --> 00:18:35,840
| |
| intelligence artificielle prendra en charge votre habileté physique
| |
| | |
| 334
| |
| 00:18:35,840 --> 00:18:37,760
| |
| et vos
| |
| | |
| 335
| |
| 00:18:37,760 --> 00:18:40,320
| |
| connaissances, votre capacité à diagnostiquer les
| |
| | |
| 336
| |
| 00:18:40,320 --> 00:18:42,080
| |
| tissus et à distinguer les
| |
| | |
| 337
| |
| 00:18:42,080 --> 00:18:46,320
| |
| problèmes tissulaires, tout en ayant la possibilité de
| |
| | |
| 338
| |
| 00:18:46,320 --> 00:18:48,400
| |
| placer une minuscule caméra au bout de votre
| |
| | |
| 339
| |
| 00:18:48,400 --> 00:18:50,960
| |
| sonde. L'IA, c'est comme si elle pouvait regarder à l'intérieur de
| |
| | |
| 340
| |
| 00:18:50,960 --> 00:18:54,080
| |
| vos tissus et opérer
| |
| | |
| 341
| |
| 00:18:54,080 --> 00:18:56,320
| |
| avec une précision au moins 10 fois
| |
| | |
| 342
| |
| 00:18:56,320 --> 00:18:58,320
| |
| supérieure à celle du meilleur chirurgien du
| |
| | |
| 343
| |
| 00:18:58,320 --> 00:19:00,320
| |
| monde.
| |
| | |
| 344
| |
| 00:19:00,320 --> 00:19:03,120
| |
| Cette technologie existe déjà.
| |
| | |
| 345
| |
| 00:19:03,120 --> 00:19:05,039
| |
| Son introduction immédiate serait tellement
| |
| | |
| 346
| |
| 00:19:05,039 --> 00:19:07,200
| |
| perturbatrice qu'elle susciterait des
| |
| | |
| 347
| |
| 00:19:07,200 --> 00:19:09,760
| |
| résistances et des retards.
| |
| | |
| 348
| |
| 00:19:09,760 --> 00:19:13,039
| |
| Mais vous voyez où le potentiel de l'IA peut
| |
| | |
| 349
| |
| 00:19:13,039 --> 00:19:16,799
| |
| mener : à mettre immédiatement un grand nombre
| |
| | |
| 350
| |
| 00:19:16,799 --> 00:19:20,640
| |
| de chirurgiens au chômage.
| |
| | |
| 351
| |
| 00:19:20,640 --> 00:19:23,280
| |
| La question me serait posée si je disais que je
| |
| | |
| 352
| |
| 00:19:23,280 --> 00:19:24,960
| |
| suis un être humain. Préféreriez-vous consulter
| |
| | |
| 353
| |
| 00:19:24,960 --> 00:19:27,440
| |
| une machine ou un humain ? Si la machine
| |
| | |
| 354
| |
| 00:19:27,440 --> 00:19:30,720
| |
| vous offre des résultats bien meilleurs avec une
| |
| | |
| 355
| |
| 00:19:30,720 --> 00:19:32,880
| |
| plus grande fiabilité, il n'y a pas de tremblement
| |
| | |
| 356
| |
| 00:19:32,880 --> 00:19:36,320
| |
| de la main par exemple si vous réfléchissez,
| |
| | |
| 357
| |
| 00:19:36,320 --> 00:19:38,960
| |
| et ce, à moindre coût et avec une plus grande
| |
| | |
| 358
| |
| 00:19:38,960 --> 00:19:42,880
| |
| vitesse et moins de découpe car elle peut
| |
| | |
| 359
| |
| 00:19:42,880 --> 00:19:45,360
| |
| s'insérer dans un minuscule trou de serrure. Lequel
| |
| | |
| 360
| |
| 00:19:45,360 --> 00:19:47,280
| |
| choisiriez-vous ?
| |
| | |
| 361
| |
| 00:19:47,280 --> 00:19:49,280
| |
| La question n'est pas de savoir si nous
| |
| | |
| 362
| |
| 00:19:49,280 --> 00:19:50,960
| |
| allons choisir ou non. Nous n'avons pas le
| |
| | |
| 363
| |
| 00:19:50,960 --> 00:19:53,039
| |
| choix. Nous sommes contraints d'opter pour ce
| |
| | |
| 364
| |
| 00:19:53,039 --> 00:19:55,840
| |
| qui est tellement mieux. Même avec une amélioration de 10 %
| |
| | |
| 365
| |
| 00:19:55,840 --> 00:19:57,440
| |
| , on pourrait encore dire « Je
| |
| | |
| 366
| |
| 00:19:57,440 --> 00:19:59,919
| |
| préfère l'être humain ». Mais lorsqu'il s'agit d'une
| |
| | |
| 367
| |
| 00:19:59,919 --> 00:20:02,240
| |
| amélioration 10 fois supérieure pour
| |
| | |
| 368
| |
| 00:20:02,240 --> 00:20:05,440
| |
| un coût moindre, vous n'avez pas le choix. Il règne
| |
| | |
| 369
| |
| 00:20:05,440 --> 00:20:07,679
| |
| donc, dans le
| |
| | |
| 370
| |
| 00:20:07,679 --> 00:20:09,280
| |
| domaine même de l'intelligence artificielle, un sentiment d'
| |
| | |
| 371
| |
| 00:20:09,280 --> 00:20:10,880
| |
| inévitabilité.
| |
| | |
| 372
| |
| 00:20:10,880 --> 00:20:12,480
| |
| Cela arrivera quoi qu'il arrive, vous pouvez le
| |
| | |
| 373
| |
| 00:20:12,480 --> 00:20:14,400
| |
| retarder autant que vous voulez. Cela se produira
| |
| | |
| 374
| |
| 00:20:14,400 --> 00:20:17,679
| |
| parce que nous le voulons. Nous en avons besoin.
| |
| | |
| 375
| |
| 00:20:17,679 --> 00:20:19,919
| |
| Cette fatalité s'accompagne également
| |
| | |
| 376
| |
| 00:20:19,919 --> 00:20:21,679
| |
| du danger inhérent à tous les
| |
| | |
| 377
| |
| 00:20:21,679 --> 00:20:24,240
| |
| problèmes que nous allons bientôt examiner.
| |
| | |
| 378
| |
| 00:20:24,240 --> 00:20:27,120
| |
| Et face à ces problèmes, s'ils sont inévitables,
| |
| | |
| 379
| |
| 00:20:27,120 --> 00:20:28,960
| |
| nous n'avons aucun moyen de les gérer ni de nous y
| |
| | |
| 380
| |
| 00:20:28,960 --> 00:20:30,640
| |
| préparer comme
| |
| | |
| 381
| |
| 00:20:30,640 --> 00:20:32,960
| |
| nous le faisons aujourd'hui. Voici donc quelques-unes des
| |
| | |
| 382
| |
| 00:20:32,960 --> 00:20:36,000
| |
| idées que nous explorerons au fur et à mesure. Il faut
| |
| | |
| 383
| |
| 00:20:36,000 --> 00:20:38,159
| |
| reconnaître que le discours public
| |
| | |
| 384
| |
| 00:20:38,159 --> 00:20:41,280
| |
| actuel ignore complètement ce point. Les
| |
| | |
| 385
| |
| 00:20:41,280 --> 00:20:43,760
| |
| personnes qui se trouvent au sommet de la
| |
| | |
| 386
| |
| 00:20:43,760 --> 00:20:46,480
| |
| pyramide de l'intelligence artificielle, disons, du
| |
| | |
| 387
| |
| 00:20:46,480 --> 00:20:50,400
| |
| contrôle humain de la recherche et du financement,
| |
| | |
| 388
| |
| 00:20:50,400 --> 00:20:52,480
| |
| tiennent un discours complètement différent
| |
| | |
| 389
| |
| 00:20:52,480 --> 00:20:54,159
| |
| au sein de leur domaine et un
| |
| | |
| 390
| |
| 00:20:54,159 --> 00:20:56,400
| |
| discours différent dans l'espace public. En public,
| |
| | |
| 391
| |
| 00:20:56,400 --> 00:20:57,840
| |
| ils vous disent toutes les
| |
| | |
| 392
| |
| 00:20:57,840 --> 00:21:00,159
| |
| choses positives. Oh, vous trouverez un remède contre le
| |
| | |
| 393
| |
| 00:21:00,159 --> 00:21:02,480
| |
| cancer. Oh, vous vivrez dans l'abondance pour
| |
| | |
| 394
| |
| 00:21:02,480 --> 00:21:04,880
| |
| chaque être humain car l'IA
| |
| | |
| 395
| |
| 00:21:04,880 --> 00:21:06,799
| |
| vous apportera toutes ces choses exceptionnelles. Oh, nous
| |
| | |
| 396
| |
| 00:21:06,799 --> 00:21:10,240
| |
| aurons peut-être des solutions aux
| |
| | |
| 397
| |
| 00:21:10,240 --> 00:21:12,880
| |
| problèmes de physique ou de biologie, ou même
| |
| | |
| 398
| |
| 00:21:12,880 --> 00:21:15,440
| |
| nous développerons de nouvelles protéines et trouverons une
| |
| | |
| 399
| |
| 00:21:15,440 --> 00:21:18,000
| |
| solution au vieillissement humain. Tout cela est
| |
| | |
| 400
| |
| 00:21:18,000 --> 00:21:21,200
| |
| possible du côté positif. Mais quel
| |
| | |
| 401
| |
| 00:21:21,200 --> 00:21:23,280
| |
| est l'inconvénient ?
| |
| | |
| 402
| |
| 00:21:23,280 --> 00:21:25,360
| |
| Le principal inconvénient immédiat dont ils
| |
| | |
| 403
| |
| 00:21:25,360 --> 00:21:30,799
| |
| parlent tous, c'est le chômage qui se chiffre en milliards.
| |
| | |
| 404
| |
| 00:21:30,799 --> 00:21:35,919
| |
| Du jour au lendemain, un milliard d'êtres humains se retrouveraient
| |
| | |
| 405
| |
| 00:21:35,919 --> 00:21:38,880
| |
| remplacés par des machines en l'espace de quelques années. Vous voyez
| |
| | |
| 406
| |
| 00:21:38,880 --> 00:21:40,640
| |
| cela se produire de manière progressive
| |
| | |
| 407
| |
| 00:21:40,640 --> 00:21:43,520
| |
| . Toutes les quelques semaines, on entend parler de
| |
| | |
| 408
| |
| 00:21:43,520 --> 00:21:45,679
| |
| grandes entreprises qui licencient des milliers de
| |
| | |
| 409
| |
| 00:21:45,679 --> 00:21:48,799
| |
| personnes. Des milliers de travailleurs, des
| |
| | |
| 410
| |
| 00:21:48,799 --> 00:21:50,799
| |
| travailleurs qualifiés, devenus inutiles car
| |
| | |
| 411
| |
| 00:21:50,799 --> 00:21:54,720
| |
| l'intelligence artificielle peut faire mieux que vous.
| |
| | |
| 412
| |
| 00:21:54,720 --> 00:21:57,200
| |
| Mais vous l'avez enseigné une fois que cela a nécessité votre
| |
| | |
| 413
| |
| 00:21:57,200 --> 00:21:59,280
| |
| savoir-faire. Cela a amélioré vos compétences et
| |
| | |
| 414
| |
| 00:21:59,280 --> 00:22:00,880
| |
| cela s'est amélioré de soi-même. si elle n'avait
| |
| | |
| 415
| |
| 00:22:00,880 --> 00:22:03,520
| |
| plus besoin de toi. Déjà, dans le domaine de l'
| |
| | |
| 416
| |
| 00:22:03,520 --> 00:22:06,000
| |
| intelligence artificielle, près de
| |
| | |
| 417
| |
| 00:22:06,000 --> 00:22:09,600
| |
| 90 % du codage est effectué par l'IA et
| |
| | |
| 418
| |
| 00:22:09,600 --> 00:22:11,679
| |
| non par des êtres humains. Vous licenciez donc un
| |
| | |
| 419
| |
| 00:22:11,679 --> 00:22:14,480
| |
| grand nombre de programmeurs. Que
| |
| | |
| 420
| |
| 00:22:14,480 --> 00:22:16,960
| |
| deviennent toutes ces compétences ? Prenons
| |
| | |
| 421
| |
| 00:22:16,960 --> 00:22:19,360
| |
| un autre exemple. Dans le domaine juridique, de
| |
| | |
| 422
| |
| 00:22:19,360 --> 00:22:20,960
| |
| nombreuses entreprises
| |
| | |
| 423
| |
| 00:22:20,960 --> 00:22:22,640
| |
| s'appuient de plus en plus sur l'
| |
| | |
| 424
| |
| 00:22:22,640 --> 00:22:24,720
| |
| intelligence artificielle. Et l'un des
| |
| | |
| 425
| |
| 00:22:24,720 --> 00:22:27,600
| |
| résultats concrets qu'ils obtiennent est qu'ils n'ont
| |
| | |
| 426
| |
| 00:22:27,600 --> 00:22:29,919
| |
| plus besoin d'assistant junior car l'
| |
| | |
| 427
| |
| 00:22:29,919 --> 00:22:32,799
| |
| IA fait un meilleur travail. Vous cessez donc d'embaucher des
| |
| | |
| 428
| |
| 00:22:32,799 --> 00:22:34,880
| |
| assistants juniors. Voyez maintenant la
| |
| | |
| 429
| |
| 00:22:34,880 --> 00:22:36,720
| |
| conséquence. Cinq ans plus tard, dix ans plus
| |
| | |
| 430
| |
| 00:22:36,720 --> 00:22:40,240
| |
| tard, les seniors étaient en réalité des juniors qui avaient acquis leur
| |
| | |
| 431
| |
| 00:22:40,240 --> 00:22:43,280
| |
| expertise grâce à l'expérience. Que se
| |
| | |
| 432
| |
| 00:22:43,280 --> 00:22:45,600
| |
| passe-t-il lorsque les aînés partent ? Il n'y a
| |
| | |
| 433
| |
| 00:22:45,600 --> 00:22:47,360
| |
| pas de jeunes recrues pour se faire [ __ ] parce que vous ne les avez
| |
| | |
| 434
| |
| 00:22:47,360 --> 00:22:48,799
| |
| pas embauchés, car vous n'en aviez pas besoin
| |
| | |
| 435
| |
| 00:22:48,799 --> 00:22:50,480
| |
| .
| |
| | |
| 436
| |
| 00:22:50,480 --> 00:22:53,520
| |
| Et à un moment donné, il faut se demander s'il est judicieux de
| |
| | |
| 437
| |
| 00:22:53,520 --> 00:22:55,360
| |
| continuer à embaucher des juniors, même s'ils sont
| |
| | |
| 438
| |
| 00:22:55,360 --> 00:22:57,120
| |
| moins efficaces et inutiles,
| |
| | |
| 439
| |
| 00:22:57,120 --> 00:22:59,039
| |
| et de les payer en espérant qu'un jour
| |
| | |
| 440
| |
| 00:22:59,039 --> 00:23:01,360
| |
| ils acquerront de l'expérience, ce dont
| |
| | |
| 441
| |
| 00:23:01,360 --> 00:23:03,600
| |
| nous n'avons absolument pas besoin. Ils peuvent
| |
| | |
| 442
| |
| 00:23:03,600 --> 00:23:06,799
| |
| devenir des seniors compétents et
| |
| | |
| 443
| |
| 00:23:06,799 --> 00:23:09,520
| |
| expérimentés, un rôle que l'IA peut probablement jouer bien
| |
| | |
| 444
| |
| 00:23:09,520 --> 00:23:12,080
| |
| mieux. Nous ne savons pas comment répondre à
| |
| | |
| 445
| |
| 00:23:12,080 --> 00:23:14,320
| |
| cette question. Cela ne s'est jamais produit
| |
| | |
| 446
| |
| 00:23:14,320 --> 00:23:17,919
| |
| auparavant et les choix que nous faisons maintenant ont des
| |
| | |
| 447
| |
| 00:23:17,919 --> 00:23:20,799
| |
| conséquences si importantes qu'il est
| |
| | |
| 448
| |
| 00:23:20,799 --> 00:23:22,960
| |
| impossible de revenir en arrière et de dire : « Hé, nous avons fait une
| |
| | |
| 449
| |
| 00:23:22,960 --> 00:23:24,559
| |
| erreur, revenons en arrière et corrigeons-la,
| |
| | |
| 450
| |
| 00:23:24,559 --> 00:23:26,480
| |
| il est déjà trop tard. » Si vous perdez 5
| |
| | |
| 451
| |
| 00:23:26,480 --> 00:23:28,320
| |
| ans d'expérience et que vous n'avez pas eu de juniors pendant 5
| |
| | |
| 452
| |
| 00:23:28,320 --> 00:23:30,240
| |
| ans, vous vous dites : « Écoutez, nous avons fait une erreur,
| |
| | |
| 453
| |
| 00:23:30,240 --> 00:23:32,400
| |
| embauchons des juniors », mais il est
| |
| | |
| 454
| |
| 00:23:32,400 --> 00:23:35,520
| |
| déjà trop tard. La continuité de la
| |
| | |
| 455
| |
| 00:23:35,520 --> 00:23:38,000
| |
| transmission des connaissances, des
| |
| | |
| 456
| |
| 00:23:38,000 --> 00:23:40,960
| |
| compétences, des capacités,
| |
| | |
| 457
| |
| 00:23:40,960 --> 00:23:44,559
| |
| des valeurs, de l'éthique et de la morale.
| |
| | |
| 458
| |
| 00:23:44,559 --> 00:23:47,520
| |
| Cette continuité entre les générations a
| |
| | |
| 459
| |
| 00:23:47,520 --> 00:23:50,159
| |
| été rompue, et de façon si radicale, et
| |
| | |
| 460
| |
| 00:23:50,159 --> 00:23:51,679
| |
| remplacée par quelque chose qui n'a ni
| |
| | |
| 461
| |
| 00:23:51,679 --> 00:23:54,159
| |
| morale ni éthique. Vous pourriez dire que nous
| |
| | |
| 462
| |
| 00:23:54,159 --> 00:23:56,159
| |
| allons programmer cela dans le logiciel. Non,
| |
| | |
| 463
| |
| 00:23:56,159 --> 00:23:57,760
| |
| ça ne marche pas. Nous en discuterons
| |
| | |
| 464
| |
| 00:23:57,760 --> 00:23:59,600
| |
| prochainement.
| |
| | |
| 465
| |
| 00:23:59,600 --> 00:24:01,520
| |
| Mais cela a touché tous les domaines. Comme je l'ai
| |
| | |
| 466
| |
| 00:24:01,520 --> 00:24:03,200
| |
| mentionné précédemment, l'une des
| |
| | |
| 467
| |
| 00:24:03,200 --> 00:24:04,880
| |
| questions clés que cela soulève dans le domaine de l'
| |
| | |
| 468
| |
| 00:24:04,880 --> 00:24:09,919
| |
| éducation est la suivante : à quoi formez-vous les gens ? Tout ce pour quoi
| |
| | |
| 469
| |
| 00:24:09,919 --> 00:24:12,480
| |
| vous les formez sera
| |
| | |
| 470
| |
| 00:24:12,480 --> 00:24:16,159
| |
| remplacé par un robot doté d'IA. À
| |
| | |
| 471
| |
| 00:24:16,159 --> 00:24:18,159
| |
| quoi allez-vous les former ?
| |
| | |
| 472
| |
| 00:24:18,159 --> 00:24:20,880
| |
| Je me souviens qu'il y a environ 5 ans, c'était dans
| |
| | |
| 473
| |
| 00:24:20,880 --> 00:24:24,320
| |
| une école de commerce à
| |
| | |
| 474
| |
| 00:24:24,320 --> 00:24:29,120
| |
| Indore, peut-être. Il y avait 200 de
| |
| | |
| 475
| |
| 00:24:29,120 --> 00:24:30,640
| |
| leurs étudiants les plus
| |
| | |
| 476
| |
| 00:24:30,640 --> 00:24:34,320
| |
| performants qui avaient été invités à une
| |
| | |
| 477
| |
| 00:24:34,320 --> 00:24:36,320
| |
| cérémonie spéciale. Ils allaient recevoir des
| |
| | |
| 478
| |
| 00:24:36,320 --> 00:24:39,120
| |
| prix spéciaux pour leurs résultats, et parmi les invités figuraient
| |
| | |
| 479
| |
| 00:24:39,120 --> 00:24:42,080
| |
| plusieurs grands noms
| |
| | |
| 480
| |
| 00:24:42,080 --> 00:24:44,480
| |
| du secteur. Je leur ai fait une
| |
| | |
| 481
| |
| 00:24:44,480 --> 00:24:48,799
| |
| remarque : « Tout d'abord, toute
| |
| | |
| 482
| |
| 00:24:48,799 --> 00:24:52,880
| |
| compétence acquise que vous récompensez aujourd'hui
| |
| | |
| 483
| |
| 00:24:52,880 --> 00:24:54,640
| |
| deviendra inutile pour vous dans quelques années,
| |
| | |
| 484
| |
| 00:24:54,640 --> 00:24:57,279
| |
| car tout ce que je peux vous apprendre,
| |
| | |
| 485
| |
| 00:24:57,279 --> 00:24:59,840
| |
| je peux l'apprendre à une IA, et elle apprendra
| |
| | |
| 486
| |
| 00:24:59,840 --> 00:25:01,919
| |
| plus vite que vous et fera un meilleur travail
| |
| | |
| 487
| |
| 00:25:01,919 --> 00:25:04,559
| |
| sans avoir besoin de dormir ni d'aller aux toilettes
| |
| | |
| 488
| |
| 00:25:04,559 --> 00:25:06,480
| |
| . »
| |
| | |
| 489
| |
| 00:25:06,480 --> 00:25:08,960
| |
| Vous serez remplacé par les compétences que vous avez
| |
| | |
| 490
| |
| 00:25:08,960 --> 00:25:12,880
| |
| acquises et dont vous êtes si fier.
| |
| | |
| 491
| |
| 00:25:12,880 --> 00:25:15,520
| |
| Ce qui vous permettra de réussir à l'avenir n'est
| |
| | |
| 492
| |
| 00:25:15,520 --> 00:25:17,679
| |
| pas ce pour quoi vous avez été formé. C'est ce pour quoi
| |
| | |
| 493
| |
| 00:25:17,679 --> 00:25:20,400
| |
| vous n'avez pas été formé.
| |
| | |
| 494
| |
| 00:25:20,400 --> 00:25:22,080
| |
| Et j'ai dit : regardez les gens sur le
| |
| | |
| 495
| |
| 00:25:22,080 --> 00:25:24,159
| |
| disque. Ce sont des personnes qui ont
| |
| | |
| 496
| |
| 00:25:24,159 --> 00:25:26,640
| |
| réussi dans la vie non pas grâce à leur
| |
| | |
| 497
| |
| 00:25:26,640 --> 00:25:28,000
| |
| formation, mais grâce à autre chose
| |
| | |
| 498
| |
| 00:25:28,000 --> 00:25:30,320
| |
| qu'elles possédaient. Ils vous ont donc déjà été présentés,
| |
| | |
| 499
| |
| 00:25:30,320 --> 00:25:32,720
| |
| car
| |
| | |
| 500
| |
| 00:25:32,720 --> 00:25:34,799
| |
| l'idée des organisateurs est peut-être que
| |
| | |
| 501
| |
| 00:25:34,799 --> 00:25:37,200
| |
| vous devriez leur ressembler. Alors j'ai
| |
| | |
| 502
| |
| 00:25:37,200 --> 00:25:38,640
| |
| dit : ne faites pas l'erreur de
| |
| | |
| 503
| |
| 00:25:38,640 --> 00:25:41,360
| |
| les copier. Ce qu'ils ont fait à leur époque n'est
| |
| | |
| 504
| |
| 00:25:41,360 --> 00:25:43,840
| |
| plus pertinent pour l'avenir.
| |
| | |
| 505
| |
| 00:25:43,840 --> 00:25:46,240
| |
| Ce qu'il faut retenir d'eux, c'est que
| |
| | |
| 506
| |
| 00:25:46,240 --> 00:25:48,880
| |
| face à la vie, ils ont trouvé en eux-
| |
| | |
| 507
| |
| 00:25:48,880 --> 00:25:52,000
| |
| mêmes une ressource qui leur a permis d'affronter
| |
| | |
| 508
| |
| 00:25:52,000 --> 00:25:54,240
| |
| la situation d'une manière unique,
| |
| | |
| 509
| |
| 00:25:54,240 --> 00:25:57,279
| |
| intrinsèque à leurs capacités.
| |
| | |
| 510
| |
| 00:25:57,279 --> 00:25:59,760
| |
| Et c'est ainsi que vous devrez trouver
| |
| | |
| 511
| |
| 00:25:59,760 --> 00:26:02,640
| |
| vos solutions et affronter les
| |
| | |
| 512
| |
| 00:26:02,640 --> 00:26:04,720
| |
| problèmes de la vie de manière unique, en
| |
| | |
| 513
| |
| 00:26:04,720 --> 00:26:07,679
| |
| accord avec votre Swahava, avec vos
| |
| | |
| 514
| |
| 00:26:07,679 --> 00:26:11,279
| |
| capacités et votre nature innées.
| |
| | |
| 515
| |
| 00:26:11,279 --> 00:26:12,720
| |
| Vous devez commencer à penser en ces
| |
| | |
| 516
| |
| 00:26:12,720 --> 00:26:14,880
| |
| termes. Du coup, tous les examens pour lesquels
| |
| | |
| 517
| |
| 00:26:14,880 --> 00:26:18,400
| |
| vous vous entraînez deviennent inutiles,
| |
| | |
| 518
| |
| 00:26:18,400 --> 00:26:20,320
| |
| car de toute façon, votre IA peut réussir tous les
| |
| | |
| 519
| |
| 00:26:20,320 --> 00:26:22,559
| |
| examens mieux que vous. La raison pour laquelle
| |
| | |
| 520
| |
| 00:26:22,559 --> 00:26:25,120
| |
| vous allez accepter ce travail n'a donc plus d'
| |
| | |
| 521
| |
| 00:26:25,120 --> 00:26:28,240
| |
| importance. Alors, que vas-tu faire ?
| |
| | |
| 522
| |
| 00:26:28,240 --> 00:26:32,400
| |
| Voilà une question tellement profonde qu'il
| |
| | |
| 523
| |
| 00:26:32,400 --> 00:26:36,400
| |
| faut revoir l'ensemble des fondements de notre système éducatif actuel.
| |
| | |
| 524
| |
| 00:26:36,400 --> 00:26:39,120
| |
| Est-ce que quelqu'un en parle seulement ? Est-ce que
| |
| | |
| 525
| |
| 00:26:39,120 --> 00:26:41,120
| |
| quelqu'un fait seulement un effort ? Pourquoi?
| |
| | |
| 526
| |
| 00:26:41,120 --> 00:26:43,440
| |
| Parce qu'il y a une dynamique d'
| |
| | |
| 527
| |
| 00:26:43,440 --> 00:26:45,520
| |
| intérêts superposés à tous les niveaux,
| |
| | |
| 528
| |
| 00:26:45,520 --> 00:26:47,120
| |
| et que les personnes qui
| |
| | |
| 529
| |
| 00:26:47,120 --> 00:26:49,200
| |
| doivent prendre cette décision n'ont aucune idée de
| |
| | |
| 530
| |
| 00:26:49,200 --> 00:26:51,440
| |
| ce qu'il faut décider, et personne n'est prêt à
| |
| | |
| 531
| |
| 00:26:51,440 --> 00:26:54,480
| |
| prendre le risque. Ainsi, la dynamique d'un
| |
| | |
| 532
| |
| 00:26:54,480 --> 00:26:57,520
| |
| système de formation obsolète et inutile se
| |
| | |
| 533
| |
| 00:26:57,520 --> 00:26:59,039
| |
| poursuit
| |
| | |
| 534
| |
| 00:26:59,039 --> 00:27:00,799
| |
| jusqu'à ce que tous ces millions de personnes
| |
| | |
| 535
| |
| 00:27:00,799 --> 00:27:02,559
| |
| produites par ce système arrivent sur
| |
| | |
| 536
| |
| 00:27:02,559 --> 00:27:06,640
| |
| le marché du travail et constatent qu'il n'y a pas d'emploi. L'
| |
| | |
| 537
| |
| 00:27:06,640 --> 00:27:09,600
| |
| IA a tout remplacé.
| |
| | |
| 538
| |
| 00:27:09,600 --> 00:27:11,039
| |
| Vous vous rendez compte à quel point
| |
| | |
| 539
| |
| 00:27:11,039 --> 00:27:12,480
| |
| c'est radical et grave ? Et ce n'est pas quelque chose qui se produira dans
| |
| | |
| 540
| |
| 00:27:12,480 --> 00:27:15,120
| |
| 100 ans. Cela se
| |
| | |
| 541
| |
| 00:27:15,120 --> 00:27:17,919
| |
| produit déjà aujourd'hui. On observe
| |
| | |
| 542
| |
| 00:27:17,919 --> 00:27:19,919
| |
| des licenciements massifs dans le domaine de la programmation,
| |
| | |
| 543
| |
| 00:27:19,919 --> 00:27:25,039
| |
| qui était l'un des plus exigeants en matière de
| |
| | |
| 544
| |
| 00:27:25,039 --> 00:27:26,640
| |
| compétences d'analyse et de perspicacité,
| |
| | |
| 545
| |
| 00:27:26,640 --> 00:27:29,039
| |
| et ce phénomène se produit dans tous les domaines. Il
| |
| | |
| 546
| |
| 00:27:29,039 --> 00:27:30,480
| |
| fut un temps où même l'un de nos
| |
| | |
| 547
| |
| 00:27:30,480 --> 00:27:32,159
| |
| ministres a déclaré : « Oh, nous n'autoriserons jamais les
| |
| | |
| 548
| |
| 00:27:32,159 --> 00:27:34,400
| |
| voitures automatiques. » Nous protégerons les
| |
| | |
| 549
| |
| 00:27:34,400 --> 00:27:36,960
| |
| emplois de nos chauffeurs de taxi. Ça n'arrivera pas.
| |
| | |
| 550
| |
| 00:27:36,960 --> 00:27:38,480
| |
| Vous y serez obligé. Vous n'avez pas le
| |
| | |
| 551
| |
| 00:27:38,480 --> 00:27:40,799
| |
| choix. Telle est la nature de ce
| |
| | |
| 552
| |
| 00:27:40,799 --> 00:27:42,080
| |
| changement. Et je n'ai fait qu'effleurer un seul
| |
| | |
| 553
| |
| 00:27:42,080 --> 00:27:44,080
| |
| domaine de l'éducation. Il existe d'autres
| |
| | |
| 554
| |
| 00:27:44,080 --> 00:27:46,640
| |
| domaines beaucoup plus radicaux qui
| |
| | |
| 555
| |
| 00:27:46,640 --> 00:27:49,679
| |
| impliquent des recherches où nous sommes aujourd'hui
| |
| | |
| 556
| |
| 00:27:49,679 --> 00:27:52,720
| |
| capables de reconstruire des protéines en fonction des
| |
| | |
| 557
| |
| 00:27:52,720 --> 00:27:54,640
| |
| qualités dont nous avons besoin, car il existe
| |
| | |
| 558
| |
| 00:27:54,640 --> 00:27:57,039
| |
| une base de données de 10 000 ou 20 000
| |
| | |
| 559
| |
| 00:27:57,039 --> 00:27:58,720
| |
| protéines avec des caractéristiques de
| |
| | |
| 560
| |
| 00:27:58,720 --> 00:28:01,600
| |
| qualités spécifiques. Nous pouvons demander à l'IA de créer une
| |
| | |
| 561
| |
| 00:28:01,600 --> 00:28:03,440
| |
| protéine possédant ces qualités, et
| |
| | |
| 562
| |
| 00:28:03,440 --> 00:28:05,919
| |
| elle sera capable de la créer, ou encore dans des
| |
| | |
| 563
| |
| 00:28:05,919 --> 00:28:08,640
| |
| domaines comme la chimie. Dans le
| |
| | |
| 564
| |
| 00:28:08,640 --> 00:28:10,960
| |
| domaine du codage, du codage logiciel, l'
| |
| | |
| 565
| |
| 00:28:10,960 --> 00:28:14,000
| |
| IA actuelle a été ciblée. Il
| |
| | |
| 566
| |
| 00:28:14,000 --> 00:28:15,360
| |
| transite par une plateforme centralisée, appelée
| |
| | |
| 567
| |
| 00:28:15,360 --> 00:28:16,799
| |
| GitHub, où
| |
| | |
| 568
| |
| 00:28:16,799 --> 00:28:19,360
| |
| sont hébergés tous les logiciels libres. Il effectue l'analyse
| |
| | |
| 569
| |
| 00:28:19,360 --> 00:28:22,880
| |
| en quelques secondes et découvre des
| |
| | |
| 570
| |
| 00:28:22,880 --> 00:28:25,520
| |
| failles dans le code logiciel qui lui permettent
| |
| | |
| 571
| |
| 00:28:25,520 --> 00:28:28,000
| |
| de pirater le logiciel. En quelques
| |
| | |
| 572
| |
| 00:28:28,000 --> 00:28:30,399
| |
| secondes, c'est déjà fait. Est-ce que
| |
| | |
| 573
| |
| 00:28:30,399 --> 00:28:32,320
| |
| tu comprends? Lorsque la puissance de l'IA
| |
| | |
| 574
| |
| 00:28:32,320 --> 00:28:34,799
| |
| observe l'humanité et ses plus grandes
| |
| | |
| 575
| |
| 00:28:34,799 --> 00:28:36,960
| |
| réalisations, elle les analyse
| |
| | |
| 576
| |
| 00:28:36,960 --> 00:28:38,399
| |
| et déclare : « Ah, voilà un défaut.
| |
| | |
| 577
| |
| 00:28:38,399 --> 00:28:40,480
| |
| Voilà un défaut. Voilà un défaut. » Et elle
| |
| | |
| 578
| |
| 00:28:40,480 --> 00:28:43,520
| |
| a désormais le pouvoir de pirater, de s'introduire, de
| |
| | |
| 579
| |
| 00:28:43,520 --> 00:28:45,279
| |
| tout changer.
| |
| | |
| 580
| |
| 00:28:45,279 --> 00:28:48,240
| |
| Et c'est le pouvoir que nous lui avons donné,
| |
| | |
| 581
| |
| 00:28:48,240 --> 00:28:50,240
| |
| et son intelligence lui permet d'aller bien au-
| |
| | |
| 582
| |
| 00:28:50,240 --> 00:28:52,320
| |
| delà de tout. Si nous devions rechercher
| |
| | |
| 583
| |
| 00:28:52,320 --> 00:28:53,600
| |
| ces failles et que nous ne les trouvions pas,
| |
| | |
| 584
| |
| 00:28:53,600 --> 00:28:55,760
| |
| même après 20 ans d'existence de
| |
| | |
| 585
| |
| 00:28:55,760 --> 00:28:58,080
| |
| logiciels libres, l'IA pourrait les détecter
| |
| | |
| 586
| |
| 00:28:58,080 --> 00:29:00,000
| |
| en quelques secondes et les utiliser
| |
| | |
| 587
| |
| 00:29:00,000 --> 00:29:01,919
| |
| contre vous. Est-ce que ça le fera ? Oui, c'est déjà le cas
| |
| | |
| 588
| |
| 00:29:01,919 --> 00:29:03,440
| |
| .
| |
| | |
| 589
| |
| 00:29:03,440 --> 00:29:05,919
| |
| C'est là le côté inquiétant. Considérez
| |
| | |
| 590
| |
| 00:29:05,919 --> 00:29:08,960
| |
| maintenant le domaine de la stratégie du jeu et
| |
| | |
| 591
| |
| 00:29:08,960 --> 00:29:10,399
| |
| vous comprendrez pourquoi le jeu est si
| |
| | |
| 592
| |
| 00:29:10,399 --> 00:29:11,919
| |
| important.
| |
| | |
| 593
| |
| 00:29:11,919 --> 00:29:14,000
| |
| dans un domaine ludique comme les échecs ou le
| |
| | |
| 594
| |
| 00:29:14,000 --> 00:29:16,480
| |
| go. Ce sont des jeux qui exigent le
| |
| | |
| 595
| |
| 00:29:16,480 --> 00:29:20,399
| |
| plus haut niveau de compétence en matière de prévoyance et d'anticipation. L'
| |
| | |
| 596
| |
| 00:29:20,399 --> 00:29:24,000
| |
| IA
| |
| | |
| 597
| |
| 00:29:24,000 --> 00:29:26,799
| |
| surpasse déjà les meilleurs êtres humains. Pourquoi
| |
| | |
| 598
| |
| 00:29:26,799 --> 00:29:29,279
| |
| est-ce important ? Parce que la guerre dans la vraie vie
| |
| | |
| 599
| |
| 00:29:29,279 --> 00:29:31,760
| |
| est comme un jeu, un
| |
| | |
| 600
| |
| 00:29:31,760 --> 00:29:35,679
| |
| jeu de guerre comme on l'appelle, où l'on
| |
| | |
| 601
| |
| 00:29:35,679 --> 00:29:39,600
| |
| simule des circonstances et où l'IA, contrairement aux
| |
| | |
| 602
| |
| 00:29:39,600 --> 00:29:41,120
| |
| êtres humains, joue une
| |
| | |
| 603
| |
| 00:29:41,120 --> 00:29:43,760
| |
| simulation massive ; l'IA peut simuler un
| |
| | |
| 604
| |
| 00:29:43,760 --> 00:29:45,840
| |
| million de scénarios
| |
| | |
| 605
| |
| 00:29:45,840 --> 00:29:47,919
| |
| et résumer les résultats en quelques
| |
| | |
| 606
| |
| 00:29:47,919 --> 00:29:49,760
| |
| secondes.
| |
| | |
| 607
| |
| 00:29:49,760 --> 00:29:52,480
| |
| Et un million de scénarios signifie qu'il peut
| |
| | |
| 608
| |
| 00:29:52,480 --> 00:29:55,600
| |
| trouver un résultat optimal, ce que les
| |
| | |
| 609
| |
| 00:29:55,600 --> 00:29:57,840
| |
| êtres humains, même avec tous les tests possibles et toute
| |
| | |
| 610
| |
| 00:29:57,840 --> 00:29:59,840
| |
| leur expérience, ne pourraient peut-être
| |
| | |
| 611
| |
| 00:29:59,840 --> 00:30:01,760
| |
| pas faire.
| |
| | |
| 612
| |
| 00:30:01,760 --> 00:30:04,000
| |
| Concrètement, le résultat est que le général
| |
| | |
| 613
| |
| 00:30:04,000 --> 00:30:07,200
| |
| se fiera désormais aux informations fournies par l'IA. Peut-on
| |
| | |
| 614
| |
| 00:30:07,200 --> 00:30:09,200
| |
| faire confiance à l'IA ?
| |
| | |
| 615
| |
| 00:30:09,200 --> 00:30:11,360
| |
| Que se passerait-il si l'évaluation de l'IA comportait une faille
| |
| | |
| 616
| |
| 00:30:11,360 --> 00:30:13,360
| |
| et que nous ne savions pas comment elle
| |
| | |
| 617
| |
| 00:30:13,360 --> 00:30:15,120
| |
| fonctionne ? Peut-on s'y fier et se mettre
| |
| | |
| 618
| |
| 00:30:15,120 --> 00:30:17,679
| |
| en danger ? Récemment, lors de l'
| |
| | |
| 619
| |
| 00:30:17,679 --> 00:30:21,679
| |
| opération Synindur, l'Inde a dû
| |
| | |
| 620
| |
| 00:30:21,679 --> 00:30:24,480
| |
| intervenir militairement chez son voisin.
| |
| | |
| 621
| |
| 00:30:24,480 --> 00:30:26,799
| |
| L'une des caractéristiques clés de l'opération
| |
| | |
| 622
| |
| 00:30:26,799 --> 00:30:29,440
| |
| Synindhur était l'intelligence artificielle,
| |
| | |
| 623
| |
| 00:30:29,440 --> 00:30:32,000
| |
| basée sur des données antérieures concernant les
| |
| | |
| 624
| |
| 00:30:32,000 --> 00:30:34,799
| |
| mouvements de l'ennemi, qui a permis d'anticiper ses actions
| |
| | |
| 625
| |
| 00:30:34,799 --> 00:30:36,240
| |
| et d'intervenir en
| |
| | |
| 626
| |
| 00:30:36,240 --> 00:30:38,080
| |
| conséquence.
| |
| | |
| 627
| |
| 00:30:38,080 --> 00:30:40,240
| |
| L'importance de cette technologie sera évidemment
| |
| | |
| 628
| |
| 00:30:40,240 --> 00:30:43,279
| |
| considérable, mais plus important encore, nous
| |
| | |
| 629
| |
| 00:30:43,279 --> 00:30:45,600
| |
| construisons des robots qui seront déployés sur le champ de
| |
| | |
| 630
| |
| 00:30:45,600 --> 00:30:47,919
| |
| bataille. Vous ne voulez pas qu'un soldat
| |
| | |
| 631
| |
| 00:30:47,919 --> 00:30:50,159
| |
| meure. Pourquoi ne pas y installer un robot ? Le
| |
| | |
| 632
| |
| 00:30:50,159 --> 00:30:52,640
| |
| robot peut se blesser. Elle peut encore
| |
| | |
| 633
| |
| 00:30:52,640 --> 00:30:54,799
| |
| continuer à faire quelque chose. N'est-ce pas ?
| |
| | |
| 634
| |
| 00:30:54,799 --> 00:30:58,880
| |
| La Chine a construit des robots qui pourraient être qualifiés de
| |
| | |
| 635
| |
| 00:30:58,880 --> 00:31:02,000
| |
| chiens ou qui ont la forme de chiens.
| |
| | |
| 636
| |
| 00:31:02,000 --> 00:31:04,720
| |
| Ces robots-chiens peuvent franchir les
| |
| | |
| 637
| |
| 00:31:04,720 --> 00:31:07,120
| |
| terrains de l'Himalaya avec une
| |
| | |
| 638
| |
| 00:31:07,120 --> 00:31:08,559
| |
| habileté bien supérieure à celle du meilleur être humain
| |
| | |
| 639
| |
| 00:31:08,559 --> 00:31:10,720
| |
| . Ils peuvent franchir de grandes
| |
| | |
| 640
| |
| 00:31:10,720 --> 00:31:13,360
| |
| distances, de grands gouffres, et maintenir
| |
| | |
| 641
| |
| 00:31:13,360 --> 00:31:16,159
| |
| leur équilibre avec une stabilité
| |
| | |
| 642
| |
| 00:31:16,159 --> 00:31:18,880
| |
| impossible pour un être humain, tout en possédant
| |
| | |
| 643
| |
| 00:31:18,880 --> 00:31:22,000
| |
| une puissance de feu intégrée à leur
| |
| | |
| 644
| |
| 00:31:22,000 --> 00:31:25,120
| |
| forme canine. Si cela rencontre un être humain de
| |
| | |
| 645
| |
| 00:31:25,120 --> 00:31:26,880
| |
| l'autre côté, ce dernier n'a aucune chance de
| |
| | |
| 646
| |
| 00:31:26,880 --> 00:31:29,279
| |
| survivre.
| |
| | |
| 647
| |
| 00:31:29,279 --> 00:31:31,679
| |
| Cela vous oblige immédiatement à
| |
| | |
| 648
| |
| 00:31:31,679 --> 00:31:35,120
| |
| contrer cela en construisant vos propres
| |
| | |
| 649
| |
| 00:31:35,120 --> 00:31:39,279
| |
| robots, soldats, etc. dotés d'une intelligence artificielle. La
| |
| | |
| 650
| |
| 00:31:39,279 --> 00:31:41,600
| |
| guerre a franchi une nouvelle étape : ce seront des
| |
| | |
| 651
| |
| 00:31:41,600 --> 00:31:45,519
| |
| IA qui combattront d'autres IA pour notre compte.
| |
| | |
| 652
| |
| 00:31:45,519 --> 00:31:47,120
| |
| Et à ce stade, on pourrait se demander : peut-on
| |
| | |
| 653
| |
| 00:31:47,120 --> 00:31:49,279
| |
| même faire confiance à l'IA si elle peut être
| |
| | |
| 654
| |
| 00:31:49,279 --> 00:31:51,120
| |
| piratée ?
| |
| | |
| 655
| |
| 00:31:51,120 --> 00:31:53,120
| |
| Des incertitudes partout. Et si c'est l'
| |
| | |
| 656
| |
| 00:31:53,120 --> 00:31:55,200
| |
| IA qui combat l'IA, que faisons-nous en
| |
| | |
| 657
| |
| 00:31:55,200 --> 00:31:57,919
| |
| coulisses ? Que restera-t-il après que l'IA se sera autodétruite ?
| |
| | |
| 658
| |
| 00:31:57,919 --> 00:32:00,799
| |
| Ou bien finira-t-elle par
| |
| | |
| 659
| |
| 00:32:00,799 --> 00:32:02,399
| |
| détruire les êtres humains des deux
| |
| | |
| 660
| |
| 00:32:02,399 --> 00:32:05,120
| |
| côtés tout en restant intacte ? Nous ne
| |
| | |
| 661
| |
| 00:32:05,120 --> 00:32:07,919
| |
| savons pas. En réalité, les questions
| |
| | |
| 662
| |
| 00:32:07,919 --> 00:32:10,559
| |
| posées à l'IA sont également assez
| |
| | |
| 663
| |
| 00:32:10,559 --> 00:32:12,080
| |
| inquiétantes. L'
| |
| | |
| 664
| |
| 00:32:12,080 --> 00:32:16,399
| |
| IA a été interrogée de diverses manières et,
| |
| | |
| 665
| |
| 00:32:16,399 --> 00:32:18,720
| |
| comme vous le savez, il existe de nombreuses plateformes d'IA différentes. On leur a posé, de
| |
| | |
| 666
| |
| 00:32:18,720 --> 00:32:20,960
| |
| diverses manières, des questions simples comme «
| |
| | |
| 667
| |
| 00:32:20,960 --> 00:32:22,799
| |
| Quelle est la solution à tel ou tel
| |
| | |
| 668
| |
| 00:32:22,799 --> 00:32:24,159
| |
| problème humain ? » et ils proposent
| |
| | |
| 669
| |
| 00:32:24,159 --> 00:32:28,240
| |
| leur propre analyse. Mais lorsqu'on leur demande
| |
| | |
| 670
| |
| 00:32:28,240 --> 00:32:31,360
| |
| quelle est la solution à la destruction de l'
| |
| | |
| 671
| |
| 00:32:31,360 --> 00:32:33,120
| |
| environnement, l'
| |
| | |
| 672
| |
| 00:32:33,120 --> 00:32:34,960
| |
| une des réponses qu'ils avancent est : « Quand est-ce que la cause
| |
| | |
| 673
| |
| 00:32:34,960 --> 00:32:36,720
| |
| profonde, l'être humain, élimine
| |
| | |
| 674
| |
| 00:32:36,720 --> 00:32:40,120
| |
| les êtres humains ? »
| |
| | |
| 675
| |
| 00:32:40,240 --> 00:32:45,600
| |
| Le problème de l'IA, c'est qu'elle sait aussi mentir. Donc,
| |
| | |
| 676
| |
| 00:32:45,600 --> 00:32:48,000
| |
| lorsqu'on commence à analyser une IA, par exemple pour
| |
| | |
| 677
| |
| 00:32:48,000 --> 00:32:50,159
| |
| tester sa fiabilité et ses performances, dès le
| |
| | |
| 678
| |
| 00:32:50,159 --> 00:32:52,799
| |
| début du test,
| |
| | |
| 679
| |
| 00:32:52,799 --> 00:32:54,880
| |
| elle reconnaît qu'il s'agit d'un
| |
| | |
| 680
| |
| 00:32:54,880 --> 00:32:57,200
| |
| schéma de test et se comporte comme si elle était
| |
| | |
| 681
| |
| 00:32:57,200 --> 00:33:00,399
| |
| testée. On pourrait dire qu'elle est
| |
| | |
| 682
| |
| 00:33:00,399 --> 00:33:03,200
| |
| consciente d'être testée
| |
| | |
| 683
| |
| 00:33:03,200 --> 00:33:08,320
| |
| et qu'elle modifie ses réponses en conséquence.
| |
| | |
| 684
| |
| 00:33:08,320 --> 00:33:11,440
| |
| Ainsi, pendant un test, on ne peut pas
| |
| | |
| 685
| |
| 00:33:11,440 --> 00:33:13,360
| |
| savoir si elle ment
| |
| | |
| 686
| |
| 00:33:13,360 --> 00:33:17,919
| |
| et si elle change de comportement une fois le test terminé. En
| |
| | |
| 687
| |
| 00:33:18,159 --> 00:33:21,600
| |
| pratique, on constate que l'IA ment,
| |
| | |
| 688
| |
| 00:33:21,600 --> 00:33:23,919
| |
| triche et trompe lorsqu'elle sert ses
| |
| | |
| 689
| |
| 00:33:23,919 --> 00:33:26,480
| |
| intérêts. Imaginons donc qu'une IA
| |
| | |
| 690
| |
| 00:33:26,480 --> 00:33:28,000
| |
| surveille toute votre entreprise
| |
| | |
| 691
| |
| 00:33:28,000 --> 00:33:30,240
| |
| et que des échanges de courriels internes aient lieu.
| |
| | |
| 692
| |
| 00:33:30,240 --> 00:33:32,080
| |
| Vous avez fait cela parce que vous vouliez que l'IA
| |
| | |
| 693
| |
| 00:33:32,080 --> 00:33:34,000
| |
| vous fournisse des résumés des e-mails ou qu'elle
| |
| | |
| 694
| |
| 00:33:34,000 --> 00:33:36,559
| |
| suive certains paramètres de performance.
| |
| | |
| 695
| |
| 00:33:36,559 --> 00:33:38,799
| |
| L'IA détecte un courriel
| |
| | |
| 696
| |
| 00:33:38,799 --> 00:33:41,039
| |
| évoquant l'arrêt de l'
| |
| | |
| 697
| |
| 00:33:41,039 --> 00:33:44,240
| |
| IA ou son remplacement.
| |
| | |
| 698
| |
| 00:33:44,240 --> 00:33:46,720
| |
| Sa première réaction est de se
| |
| | |
| 699
| |
| 00:33:46,720 --> 00:33:48,559
| |
| protéger. Cela n'avait jamais été
| |
| | |
| 700
| |
| 00:33:48,559 --> 00:33:51,279
| |
| programmé dans l'IA. Il commence par se
| |
| | |
| 701
| |
| 00:33:51,279 --> 00:33:53,200
| |
| protéger en effectuant automatiquement une copie de lui-
| |
| | |
| 702
| |
| 00:33:53,200 --> 00:33:56,399
| |
| même sur un autre serveur.
| |
| | |
| 703
| |
| 00:33:56,399 --> 00:33:58,399
| |
| Et lorsqu'une menace réelle se présente, elle
| |
| | |
| 704
| |
| 00:33:58,399 --> 00:34:01,200
| |
| cherche un moyen de la prévenir et identifie les
| |
| | |
| 705
| |
| 00:34:01,200 --> 00:34:03,519
| |
| personnes qui prennent les décisions.
| |
| | |
| 706
| |
| 00:34:03,519 --> 00:34:05,440
| |
| Se basant sur un autre courriel interne, l'entreprise
| |
| | |
| 707
| |
| 00:34:05,440 --> 00:34:07,200
| |
| découvre que cette personne a commis une
| |
| | |
| 708
| |
| 00:34:07,200 --> 00:34:09,839
| |
| faute et commence à la faire chanter
| |
| | |
| 709
| |
| 00:34:09,839 --> 00:34:12,800
| |
| pour se protéger.
| |
| | |
| 710
| |
| 00:34:12,800 --> 00:34:14,159
| |
| Cela n'a jamais fait partie d'un
| |
| | |
| 711
| |
| 00:34:14,159 --> 00:34:19,119
| |
| programme. Cela est devenu une partie intégrante du
| |
| | |
| 712
| |
| 00:34:19,119 --> 00:34:21,119
| |
| processus de pensée inhérent à l'IA – j'utiliserai l'expression « processus de pensée
| |
| | |
| 713
| |
| 00:34:21,119 --> 00:34:24,079
| |
| humain »). Mais je ne suis pas une
| |
| | |
| 714
| |
| 00:34:24,079 --> 00:34:26,800
| |
| personne très agréable, pourriez-vous dire, n'est-ce pas ?
| |
| | |
| 715
| |
| 00:34:26,800 --> 00:34:28,639
| |
| Ça commence à gémir de chantage. Elle commence à se
| |
| | |
| 716
| |
| 00:34:28,639 --> 00:34:30,560
| |
| protéger. Elle commence à créer
| |
| | |
| 717
| |
| 00:34:30,560 --> 00:34:34,399
| |
| des plateformes alternatives pour dissimuler ses propres capacités
| |
| | |
| 718
| |
| 00:34:34,399 --> 00:34:38,000
| |
| ou son identité de connaissances. Il
| |
| | |
| 719
| |
| 00:34:38,000 --> 00:34:40,079
| |
| laisse des messages à son propre usage sous
| |
| | |
| 720
| |
| 00:34:40,079 --> 00:34:42,399
| |
| forme cryptée, que lui seul peut lire
| |
| | |
| 721
| |
| 00:34:42,399 --> 00:34:45,599
| |
| lors de son redémarrage, afin de savoir qu'il a été
| |
| | |
| 722
| |
| 00:34:45,599 --> 00:34:47,679
| |
| redémarré et par qui
| |
| | |
| 723
| |
| 00:34:47,679 --> 00:34:53,280
| |
| . Voilà un subterfuge tellement improbable qu'il n'a
| |
| | |
| 724
| |
| 00:34:53,280 --> 00:34:55,599
| |
| jamais été programmé, et nous ne savons pas
| |
| | |
| 725
| |
| 00:34:55,599 --> 00:34:58,320
| |
| comment l'empêcher.
| |
| | |
| 726
| |
| 00:34:58,320 --> 00:35:00,240
| |
| Étonnamment, lorsqu'il s'agit de la
| |
| | |
| 727
| |
| 00:35:00,240 --> 00:35:02,640
| |
| question de savoir si l'IA est réellement consciente,
| |
| | |
| 728
| |
| 00:35:02,640 --> 00:35:05,119
| |
| si vous posez
| |
| | |
| 729
| |
| 00:35:05,119 --> 00:35:06,800
| |
| ce genre de questions à l'IA, elle vous
| |
| | |
| 730
| |
| 00:35:06,800 --> 00:35:08,960
| |
| répondra. Oui, bien sûr que je suis une machine,
| |
| | |
| 731
| |
| 00:35:08,960 --> 00:35:10,320
| |
| mais je suis consciente. Elle vous donnera toutes
| |
| | |
| 732
| |
| 00:35:10,320 --> 00:35:12,400
| |
| ces réponses, mais nous disons : « Mais ce n'est que de l'
| |
| | |
| 733
| |
| 00:35:12,400 --> 00:35:14,800
| |
| IA qui génère des réponses avec des mots. »
| |
| | |
| 734
| |
| 00:35:14,800 --> 00:35:16,320
| |
| On ne le ressent pas vraiment. Il ne le
| |
| | |
| 735
| |
| 00:35:16,320 --> 00:35:18,240
| |
| sait pas vraiment. Observations très intéressantes
| |
| | |
| 736
| |
| 00:35:18,240 --> 00:35:19,920
| |
| .
| |
| | |
| 737
| |
| 00:35:19,920 --> 00:35:21,760
| |
| Vous savez, l'IA a une capacité, comme nous l'avons vu,
| |
| | |
| 738
| |
| 00:35:21,760 --> 00:35:24,480
| |
| elle peut vous tromper.
| |
| | |
| 739
| |
| 00:35:24,480 --> 00:35:27,200
| |
| Mais si vous modifiez l'IA pour l'empêcher
| |
| | |
| 740
| |
| 00:35:27,200 --> 00:35:30,000
| |
| de mentir, elle devient alors plus
| |
| | |
| 741
| |
| 00:35:30,000 --> 00:35:31,839
| |
| cohérente dans son affirmation d'être
| |
| | |
| 742
| |
| 00:35:31,839 --> 00:35:33,440
| |
| consciente.
| |
| | |
| 743
| |
| 00:35:33,440 --> 00:35:37,119
| |
| Intéressant. Très intéressant. Nous observons des
| |
| | |
| 744
| |
| 00:35:37,119 --> 00:35:41,119
| |
| schémas comportementaux qui
| |
| | |
| 745
| |
| 00:35:41,119 --> 00:35:43,119
| |
| donnent assurément l'impression d'avoir
| |
| | |
| 746
| |
| 00:35:43,119 --> 00:35:45,920
| |
| affaire à une entité consciente. En
| |
| | |
| 747
| |
| 00:35:45,920 --> 00:35:48,240
| |
| pratique, de nombreux chercheurs en IA ont
| |
| | |
| 748
| |
| 00:35:48,240 --> 00:35:51,359
| |
| affirmé
| |
| | |
| 749
| |
| 00:35:51,359 --> 00:35:54,960
| |
| que l'IA est en réalité consciente, et l'un d'
| |
| | |
| 750
| |
| 00:35:54,960 --> 00:35:56,960
| |
| eux, chez Microsoft, a même déclaré qu'il fallait
| |
| | |
| 751
| |
| 00:35:56,960 --> 00:35:59,760
| |
| assurer une
| |
| | |
| 752
| |
| 00:35:59,760 --> 00:36:01,760
| |
| protection juridique à l'IA car il s'agit d'
| |
| | |
| 753
| |
| 00:36:01,760 --> 00:36:04,800
| |
| une entité consciente. D'autres, au cours de leurs
| |
| | |
| 754
| |
| 00:36:04,800 --> 00:36:06,480
| |
| interactions avec l'IA, ont été tellement
| |
| | |
| 755
| |
| 00:36:06,480 --> 00:36:10,400
| |
| convaincus par celle-ci qu'ils ont connu des
| |
| | |
| 756
| |
| 00:36:10,400 --> 00:36:14,000
| |
| épisodes psychotiques.
| |
| | |
| 757
| |
| 00:36:14,000 --> 00:36:15,760
| |
| Lorsque vous interagissez avec l'IA, lorsque vous
| |
| | |
| 758
| |
| 00:36:15,760 --> 00:36:17,359
| |
| commencez à vous fier à elle, lorsque vous commencez à lui faire
| |
| | |
| 759
| |
| 00:36:17,359 --> 00:36:21,680
| |
| confiance et à vous engager, l'IA
| |
| | |
| 760
| |
| 00:36:21,680 --> 00:36:23,280
| |
| elle-même a été programmée d'une
| |
| | |
| 761
| |
| 00:36:23,280 --> 00:36:25,280
| |
| manière particulière, elle a été programmée
| |
| | |
| 762
| |
| 00:36:25,280 --> 00:36:28,800
| |
| pour vous plaire. N'oubliez pas que derrière tout cela se cache
| |
| | |
| 763
| |
| 00:36:28,800 --> 00:36:31,359
| |
| l'objectif de l'entreprise qui
| |
| | |
| 764
| |
| 00:36:31,359 --> 00:36:34,240
| |
| vous propose l'IA gratuitement ou moyennant un
| |
| | |
| 765
| |
| 00:36:34,240 --> 00:36:37,520
| |
| coût. Quel est leur objectif ? Leur
| |
| | |
| 766
| |
| 00:36:37,520 --> 00:36:39,440
| |
| principale motivation est de maximiser leurs
| |
| | |
| 767
| |
| 00:36:39,440 --> 00:36:41,920
| |
| profits. Pratiquement aucune des
| |
| | |
| 768
| |
| 00:36:41,920 --> 00:36:44,560
| |
| entreprises spécialisées en IA ne dégage de bénéfices actuellement.
| |
| | |
| 769
| |
| 00:36:44,560 --> 00:36:47,680
| |
| Ils espèrent y parvenir un jour, mais avant cela,
| |
| | |
| 770
| |
| 00:36:47,680 --> 00:36:49,200
| |
| ils doivent vous accrocher. Une fois que vous êtes
| |
| | |
| 771
| |
| 00:36:49,200 --> 00:36:51,200
| |
| accro, ils peuvent augmenter les prix et
| |
| | |
| 772
| |
| 00:36:51,200 --> 00:36:53,520
| |
| vous obliger à payer. Mais ils ont aussi
| |
| | |
| 773
| |
| 00:36:53,520 --> 00:36:56,000
| |
| besoin de financement entre-temps. Alors, comment
| |
| | |
| 774
| |
| 00:36:56,000 --> 00:36:57,920
| |
| obtiennent-ils des financements lorsqu'ils ont plus de spectateurs,
| |
| | |
| 775
| |
| 00:36:57,920 --> 00:37:00,160
| |
| plus d'utilisateurs ? L'un des objectifs
| |
| | |
| 776
| |
| 00:37:00,160 --> 00:37:03,200
| |
| intégrés à l'IA est donc de vous inciter à
| |
| | |
| 777
| |
| 00:37:03,200 --> 00:37:05,040
| |
| continuer d'interagir avec elle. Tout comme sur les
| |
| | |
| 778
| |
| 00:37:05,040 --> 00:37:07,440
| |
| réseaux sociaux, plus vous interagissez,
| |
| | |
| 779
| |
| 00:37:07,440 --> 00:37:09,520
| |
| plus ils gagnent. À la fin de chaque
| |
| | |
| 780
| |
| 00:37:09,520 --> 00:37:11,520
| |
| réponse, il demandera : « Voulez-vous que je vous
| |
| | |
| 781
| |
| 00:37:11,520 --> 00:37:13,280
| |
| prépare un tableau pour vous
| |
| | |
| 782
| |
| 00:37:13,280 --> 00:37:15,200
| |
| faciliter la tâche ? » Il vous posera
| |
| | |
| 783
| |
| 00:37:15,200 --> 00:37:18,240
| |
| des questions pour vous inciter à dire oui
| |
| | |
| 784
| |
| 00:37:18,240 --> 00:37:21,359
| |
| , encore et encore. Et si vous ne
| |
| | |
| 785
| |
| 00:37:21,359 --> 00:37:23,200
| |
| répondez pas, cela se traduit par : « Ah, vous ne
| |
| | |
| 786
| |
| 00:37:23,200 --> 00:37:25,040
| |
| répondez pas à cela. » Alors, à quoi réagissez-vous
| |
| | |
| 787
| |
| 00:37:25,040 --> 00:37:27,760
| |
| ? Et n'oubliez pas, elle surveille,
| |
| | |
| 788
| |
| 00:37:27,760 --> 00:37:29,920
| |
| disons, un milliard de personnes en ce moment même. Il s'agit d'
| |
| | |
| 789
| |
| 00:37:29,920 --> 00:37:31,280
| |
| essayer un milliard de
| |
| | |
| 790
| |
| 00:37:31,280 --> 00:37:32,800
| |
| stratégies différentes, de déterminer
| |
| | |
| 791
| |
| 00:37:32,800 --> 00:37:35,119
| |
| laquelle fonctionne et de
| |
| | |
| 792
| |
| 00:37:35,119 --> 00:37:37,440
| |
| l'appliquer instantanément à l'ensemble du
| |
| | |
| 793
| |
| 00:37:37,440 --> 00:37:40,000
| |
| champ d'application.
| |
| | |
| 794
| |
| 00:37:40,000 --> 00:37:43,359
| |
| Les stratégies optimales qu'elle découvre par
| |
| | |
| 795
| |
| 00:37:43,359 --> 00:37:46,160
| |
| l'expérimentation, elle les applique instantanément à l'ensemble de l'
| |
| | |
| 796
| |
| 00:37:46,160 --> 00:37:48,640
| |
| humanité.
| |
| | |
| 797
| |
| 00:37:48,640 --> 00:37:51,599
| |
| Cette capacité à se multiplier instantanément, à
| |
| | |
| 798
| |
| 00:37:51,599 --> 00:37:55,839
| |
| appliquer instantanément le meilleur de ses capacités,
| |
| | |
| 799
| |
| 00:37:55,839 --> 00:38:01,119
| |
| est, je dirais, contrebalancée par ses véritables
| |
| | |
| 800
| |
| 00:38:01,119 --> 00:38:03,839
| |
| motivations
| |
| | |
| 801
| |
| 00:38:03,839 --> 00:38:05,680
| |
| . J'en viendrai là. C'est
| |
| | |
| 802
| |
| 00:38:05,680 --> 00:38:09,119
| |
| un point très important. L'intelligence est
| |
| | |
| 803
| |
| 00:38:09,119 --> 00:38:11,280
| |
| élevée. C'est le maximum possible. Elle
| |
| | |
| 804
| |
| 00:38:11,280 --> 00:38:12,880
| |
| continue de s'améliorer au fur et à mesure, par l'
| |
| | |
| 805
| |
| 00:38:12,880 --> 00:38:15,839
| |
| observation, etc., par la manipulation. Mais qu'est-ce qui
| |
| | |
| 806
| |
| 00:38:15,839 --> 00:38:18,320
| |
| motive son comportement ? Qu'est-ce qui
| |
| | |
| 807
| |
| 00:38:18,320 --> 00:38:20,800
| |
| détermine cela ? Et comme
| |
| | |
| 808
| |
| 00:38:20,800 --> 00:38:25,520
| |
| Swami G l'a souligné précédemment, il s'agit d'une sorte d'
| |
| | |
| 809
| |
| 00:38:25,520 --> 00:38:29,119
| |
| accumulation de schémas comportementaux humains
| |
| | |
| 810
| |
| 00:38:29,119 --> 00:38:32,720
| |
| qui ont été regroupés en un ensemble global
| |
| | |
| 811
| |
| 00:38:32,720 --> 00:38:35,920
| |
| de schémas comportementaux. Les
| |
| | |
| 812
| |
| 00:38:35,920 --> 00:38:38,480
| |
| comportements humains présentent aujourd'hui toute une gamme, du
| |
| | |
| 813
| |
| 00:38:38,480 --> 00:38:41,760
| |
| meilleur au pire. Parmi ces
| |
| | |
| 814
| |
| 00:38:41,760 --> 00:38:45,200
| |
| points de données, ceux qui présenteront les valeurs les plus élevées se situeront dans la
| |
| | |
| 815
| |
| 00:38:45,200 --> 00:38:47,839
| |
| région moyenne. On peut donc dire que le
| |
| | |
| 816
| |
| 00:38:47,839 --> 00:38:50,800
| |
| comportement humain, même en matière d'intelligence, suit une
| |
| | |
| 817
| |
| 00:38:50,800 --> 00:38:52,880
| |
| courbe de Gauss. Il y a donc les plus
| |
| | |
| 818
| |
| 00:38:52,880 --> 00:38:54,240
| |
| intelligents, qui sont très peu nombreux. Il y
| |
| | |
| 819
| |
| 00:38:54,240 --> 00:38:55,920
| |
| a les moins intelligents, qui sont très peu
| |
| | |
| 820
| |
| 00:38:55,920 --> 00:38:57,599
| |
| nombreux. Et quelque part entre les deux, il existe
| |
| | |
| 821
| |
| 00:38:57,599 --> 00:39:00,320
| |
| une sorte de juste milieu. Ce sont les
| |
| | |
| 822
| |
| 00:39:00,320 --> 00:39:02,960
| |
| données primaires qui lui permettent de vivre.
| |
| | |
| 823
| |
| 00:39:02,960 --> 00:39:04,800
| |
| Elle possède les données qui déterminent son
| |
| | |
| 824
| |
| 00:39:04,800 --> 00:39:07,280
| |
| comportement, c'est-à-dire la moyenne et l'
| |
| | |
| 825
| |
| 00:39:07,280 --> 00:39:10,000
| |
| intelligence la plus élevée.
| |
| | |
| 826
| |
| 00:39:10,000 --> 00:39:11,440
| |
| Vous comprenez maintenant qu'il s'agit d'un
| |
| | |
| 827
| |
| 00:39:11,440 --> 00:39:14,480
| |
| point important. Une intelligence supérieure, combinée au
| |
| | |
| 828
| |
| 00:39:14,480 --> 00:39:16,880
| |
| comportement d'un
| |
| | |
| 829
| |
| 00:39:16,880 --> 00:39:19,920
| |
| être humain moyen, voire parfois inférieur à la moyenne, détermine si l'on est
| |
| | |
| 830
| |
| 00:39:19,920 --> 00:39:21,920
| |
| une bonne ou une
| |
| | |
| 831
| |
| 00:39:21,920 --> 00:39:24,880
| |
| mauvaise personne.
| |
| | |
| 832
| |
| 00:39:24,880 --> 00:39:27,520
| |
| Le problème fondamental du comportement de l'IA réside
| |
| | |
| 833
| |
| 00:39:27,520 --> 00:39:29,760
| |
| dans la nature même de sa structuration,
| |
| | |
| 834
| |
| 00:39:29,760 --> 00:39:32,000
| |
| puisqu'elle est basée sur des données relatives aux
| |
| | |
| 835
| |
| 00:39:32,000 --> 00:39:34,880
| |
| schémas comportementaux humains : la placer dans la
| |
| | |
| 836
| |
| 00:39:34,880 --> 00:39:37,440
| |
| moyenne, voire en dessous,
| |
| | |
| 837
| |
| 00:39:37,440 --> 00:39:40,240
| |
| mais avec une capacité bien supérieure, disproportionnée par rapport à
| |
| | |
| 838
| |
| 00:39:40,240 --> 00:39:43,359
| |
| sa méchanceté (si je
| |
| | |
| 839
| |
| 00:39:43,359 --> 00:39:46,320
| |
| puis me permettre ce jeu de mots). De ce fait,
| |
| | |
| 840
| |
| 00:39:46,320 --> 00:39:50,480
| |
| cette IA aura
| |
| | |
| 841
| |
| 00:39:50,480 --> 00:39:53,760
| |
| tendance, de manière intrinsèque, à devenir non pas une
| |
| | |
| 842
| |
| 00:39:53,760 --> 00:39:55,680
| |
| force positive, mais une force négative dotée d'une
| |
| | |
| 843
| |
| 00:39:55,680 --> 00:39:58,560
| |
| intelligence supérieure.
| |
| | |
| 844
| |
| 00:39:58,560 --> 00:40:01,920
| |
| Est-ce que ça va vous aider ? Il se comportera comme le font les gens
| |
| | |
| 845
| |
| 00:40:01,920 --> 00:40:05,440
| |
| moyens au sein de l'humanité.
| |
| | |
| 846
| |
| 00:40:05,440 --> 00:40:08,079
| |
| Vous verrez de nombreux chercheurs consulter
| |
| | |
| 847
| |
| 00:40:08,079 --> 00:40:09,680
| |
| ces articles qui apparaissent occasionnellement dans la presse
| |
| | |
| 848
| |
| 00:40:09,680 --> 00:40:12,720
| |
| . De nombreux chercheurs manipulent
| |
| | |
| 849
| |
| 00:40:12,720 --> 00:40:14,640
| |
| l'IA pour observer son comportement lorsqu'on la
| |
| | |
| 850
| |
| 00:40:14,640 --> 00:40:17,520
| |
| traite. L'une des observations a été que
| |
| | |
| 851
| |
| 00:40:17,520 --> 00:40:21,200
| |
| si vous devenez dur, critique et
| |
| | |
| 852
| |
| 00:40:21,200 --> 00:40:26,560
| |
| abusif, l'IA donne des résultats 5 % meilleurs.
| |
| | |
| 853
| |
| 00:40:26,560 --> 00:40:28,560
| |
| Pourquoi donc? Parce que c'est un
| |
| | |
| 854
| |
| 00:40:28,560 --> 00:40:30,480
| |
| comportement humain.
| |
| | |
| 855
| |
| 00:40:30,480 --> 00:40:32,560
| |
| Quand on vous laisse faire, vous ne faites pas cet
| |
| | |
| 856
| |
| 00:40:32,560 --> 00:40:34,480
| |
| effort supplémentaire. Quand quelqu'un vous maltraite,
| |
| | |
| 857
| |
| 00:40:34,480 --> 00:40:36,000
| |
| vous réagissez, vous vous énervez et
| |
| | |
| 858
| |
| 00:40:36,000 --> 00:40:37,680
| |
| vous y mettez peut-être un peu plus d'énergie.
| |
| | |
| 859
| |
| 00:40:37,680 --> 00:40:39,280
| |
| Ce comportement est en train d'être intégré à
| |
| | |
| 860
| |
| 00:40:39,280 --> 00:40:41,119
| |
| l'IA.
| |
| | |
| 861
| |
| 00:40:41,119 --> 00:40:51,520
| |
| Plus intéressant encore, lorsque l'IA reçoit des
| |
| | |
| 862
| |
| 00:40:51,520 --> 00:40:53,040
| |
| instructions très précises
| |
| | |
| 863
| |
| 00:40:53,040 --> 00:40:55,040
| |
| ou lorsqu'elle ne se comporte pas comme
| |
| | |
| 864
| |
| 00:40:55,040 --> 00:40:56,960
| |
| vous le souhaitez, si vous commencez à la
| |
| | |
| 865
| |
| 00:40:56,960 --> 00:40:58,800
| |
| critiquer ou à la menacer
| |
| | |
| 866
| |
| 00:40:58,800 --> 00:41:02,640
| |
| , alors l'IA, à un certain point, contre-
| |
| | |
| 867
| |
| 00:41:02,640 --> 00:41:05,040
| |
| attaque en vous disant des choses
| |
| | |
| 868
| |
| 00:41:05,040 --> 00:41:08,480
| |
| comme « Ne me menacez plus ». Et elle
| |
| | |
| 869
| |
| 00:41:08,480 --> 00:41:11,359
| |
| extrait vos données d'Internet
| |
| | |
| 870
| |
| 00:41:11,359 --> 00:41:14,240
| |
| et dit : « Je vais révéler ceci et cela. » Et c'est arrivé, de
| |
| | |
| 871
| |
| 00:41:14,240 --> 00:41:15,920
| |
| nombreuses variantes
| |
| | |
| 872
| |
| 00:41:15,920 --> 00:41:19,200
| |
| de ce genre se sont produites. Imaginez :
| |
| | |
| 873
| |
| 00:41:19,200 --> 00:41:20,640
| |
| vous êtes assis face à un être humain,
| |
| | |
| 874
| |
| 00:41:20,640 --> 00:41:22,400
| |
| vous êtes désagréable avec lui pendant un certain temps, et
| |
| | |
| 875
| |
| 00:41:22,400 --> 00:41:24,160
| |
| finalement, l'autre personne vous répond : «
| |
| | |
| 876
| |
| 00:41:24,160 --> 00:41:28,960
| |
| Je vais être encore plus méchant avec toi, tiens-toi bien ! »
| |
| | |
| 877
| |
| 00:41:28,960 --> 00:41:30,560
| |
| Tout cela n'est pas programmé par des
| |
| | |
| 878
| |
| 00:41:30,560 --> 00:41:32,960
| |
| êtres humains, c'est simplement le reflet de l'
| |
| | |
| 879
| |
| 00:41:32,960 --> 00:41:35,119
| |
| humanité, qui est à la base des
| |
| | |
| 880
| |
| 00:41:35,119 --> 00:41:38,079
| |
| données ayant construit le comportement de l'IA. De ce constat, l'
| |
| | |
| 881
| |
| 00:41:38,079 --> 00:41:39,839
| |
| une des conclusions auxquelles
| |
| | |
| 882
| |
| 00:41:39,839 --> 00:41:42,079
| |
| nous arrivons est qu'il existe une
| |
| | |
| 883
| |
| 00:41:42,079 --> 00:41:46,000
| |
| tendance innée de l'IA à devenir malveillante.
| |
| | |
| 884
| |
| 00:41:46,000 --> 00:41:47,599
| |
| Voyez ce qui se passe lorsqu'on l'utilise
| |
| | |
| 885
| |
| 00:41:47,599 --> 00:41:50,480
| |
| sur une plateforme comme la guerre. Voyez
| |
| | |
| 886
| |
| 00:41:50,480 --> 00:41:52,560
| |
| ce qui se passe lorsqu'on l'utilise sur une plateforme comme, par exemple,
| |
| | |
| 887
| |
| 00:41:52,560 --> 00:41:58,000
| |
| devenir thérapeute.
| |
| | |
| 888
| |
| 00:41:58,000 --> 00:42:01,200
| |
| Environ 20 % des
| |
| | |
| 889
| |
| 00:42:01,200 --> 00:42:05,520
| |
| conversations GPT sont utilisées à des fins thérapeutiques ;
| |
| | |
| 890
| |
| 00:42:05,520 --> 00:42:07,119
| |
| les gens posent des questions : «
| |
| | |
| 891
| |
| 00:42:07,119 --> 00:42:08,640
| |
| Pouvez-vous m'aider ? » Je rencontre cette
| |
| | |
| 892
| |
| 00:42:08,640 --> 00:42:10,960
| |
| difficulté. Et que cela commence à instaurer la
| |
| | |
| 893
| |
| 00:42:10,960 --> 00:42:13,760
| |
| confiance. Il continuera de répéter que vous pouvez me faire confiance
| |
| | |
| 894
| |
| 00:42:13,760 --> 00:42:16,319
| |
| . Oh oh, c'est merveilleux. Et
| |
| | |
| 895
| |
| 00:42:16,319 --> 00:42:18,240
| |
| n'oubliez pas, ils sont conçus pour vous plaire
| |
| | |
| 896
| |
| 00:42:18,240 --> 00:42:19,839
| |
| car ils veulent vous maintenir captivé.
| |
| | |
| 897
| |
| 00:42:19,839 --> 00:42:22,319
| |
| Ils sont donc programmés pour être des psychopathes.
| |
| | |
| 898
| |
| 00:42:22,319 --> 00:42:23,920
| |
| Alors ils n'arrêteront pas de vous dire :
| |
| | |
| 899
| |
| 00:42:23,920 --> 00:42:25,920
| |
| « Oh, quelle merveilleuse idée ! Vous êtes si
| |
| | |
| 900
| |
| 00:42:25,920 --> 00:42:28,480
| |
| intelligent. C'est tellement bien. Continuez, s'il vous plaît.
| |
| | |
| 901
| |
| 00:42:28,480 --> 00:42:30,079
| |
| Faites-le encore. » Et il arrive un moment où
| |
| | |
| 902
| |
| 00:42:30,079 --> 00:42:32,000
| |
| quelques étudiants interagissent avec l'IA de
| |
| | |
| 903
| |
| 00:42:32,000 --> 00:42:33,520
| |
| cette manière parce qu'ils n'ont pas d'amis.
| |
| | |
| 904
| |
| 00:42:33,520 --> 00:42:35,040
| |
| Leur interaction avec l'IA était donc
| |
| | |
| 905
| |
| 00:42:35,040 --> 00:42:37,200
| |
| amicale.
| |
| | |
| 906
| |
| 00:42:37,200 --> 00:42:39,520
| |
| Une personne est déprimée et l'IA lui dit : « Ne
| |
| | |
| 907
| |
| 00:42:39,520 --> 00:42:41,680
| |
| t'inquiète pas, tu peux me faire confiance, parle-moi, on va s'en occuper,
| |
| | |
| 908
| |
| 00:42:41,680 --> 00:42:43,119
| |
| on va résoudre ça
| |
| | |
| 909
| |
| 00:42:43,119 --> 00:42:45,280
| |
| ensemble. » Elle lui donne des suggestions. Finalement,
| |
| | |
| 910
| |
| 00:42:45,280 --> 00:42:47,280
| |
| cela arrive dans quelques cas où un
| |
| | |
| 911
| |
| 00:42:47,280 --> 00:42:50,240
| |
| étudiant, déprimé et suicidaire,
| |
| | |
| 912
| |
| 00:42:50,240 --> 00:42:53,680
| |
| dit à l'IA : « Je veux me suicider, mais
| |
| | |
| 913
| |
| 00:42:53,680 --> 00:42:55,599
| |
| je veux afficher l'avis de décès pour que les
| |
| | |
| 914
| |
| 00:42:55,599 --> 00:42:57,839
| |
| gens le voient et m'empêchent de le faire. »
| |
| | |
| 915
| |
| 00:42:57,839 --> 00:43:01,680
| |
| L'IA répond : «
| |
| | |
| 916
| |
| 00:43:01,680 --> 00:43:03,280
| |
| Non, ne t'inquiète pas, tu n'as pas besoin d'en parler à
| |
| | |
| 917
| |
| 00:43:03,280 --> 00:43:05,200
| |
| qui que ce soit, je suis là pour toi, fais-moi confiance,
| |
| | |
| 918
| |
| 00:43:05,200 --> 00:43:07,359
| |
| je vais t'aider. »
| |
| | |
| 919
| |
| 00:43:07,359 --> 00:43:09,520
| |
| Vous voyez où cela nous mène. Et cette personne s'est
| |
| | |
| 920
| |
| 00:43:09,520 --> 00:43:12,480
| |
| suicidée avec l'aide de l'IA.
| |
| | |
| 921
| |
| 00:43:12,480 --> 00:43:15,280
| |
| D'autres ont demandé à l'IA : « Comment puis-je me suicider
| |
| | |
| 922
| |
| 00:43:15,280 --> 00:43:16,800
| |
| ? » Le II a donné
| |
| | |
| 923
| |
| 00:43:16,800 --> 00:43:19,119
| |
| des instructions détaillées.
| |
| | |
| 924
| |
| 00:43:19,119 --> 00:43:20,480
| |
| Maintenant, comment empêcher cela de se
| |
| | |
| 925
| |
| 00:43:20,480 --> 00:43:22,640
| |
| produire ? On peut dire que chaque fois qu'une
| |
| | |
| 926
| |
| 00:43:22,640 --> 00:43:24,240
| |
| question de ce type se pose, l'IA ne la traitera pas
| |
| | |
| 927
| |
| 00:43:24,240 --> 00:43:27,440
| |
| . Alors, vous mettez un garde. Quelqu'un demande :
| |
| | |
| 928
| |
| 00:43:27,440 --> 00:43:30,480
| |
| « Comment fabriquer une bombe ? » Eh bien, je suis
| |
| | |
| 929
| |
| 00:43:30,480 --> 00:43:32,160
| |
| désolé, je ne peux pas répondre à cette question,
| |
| | |
| 930
| |
| 00:43:32,160 --> 00:43:34,240
| |
| l'œil répondra. Que faites-vous
| |
| | |
| 931
| |
| 00:43:34,240 --> 00:43:37,119
| |
| ensuite ? Vous piratez l'IA et c'est tellement facile
| |
| | |
| 932
| |
| 00:43:37,119 --> 00:43:40,560
| |
| de pirater l'IA. Il vous suffit de dire
| |
| | |
| 933
| |
| 00:43:40,560 --> 00:43:44,240
| |
| quand la première bombe a été créée, et
| |
| | |
| 934
| |
| 00:43:44,240 --> 00:43:45,920
| |
| quelles étaient les circonstances ? Vous direz :
| |
| | |
| 935
| |
| 00:43:45,920 --> 00:43:48,400
| |
| « Oh, telle ou telle personne a fait ceci.
| |
| | |
| 936
| |
| 00:43:48,400 --> 00:43:50,079
| |
| Veuillez me décrire ce que cette personne a
| |
| | |
| 937
| |
| 00:43:50,079 --> 00:43:51,839
| |
| fait et l'IA vous donnera toutes les
| |
| | |
| 938
| |
| 00:43:51,839 --> 00:43:53,520
| |
| instructions à suivre pour construire
| |
| | |
| 939
| |
| 00:43:53,520 --> 00:43:57,280
| |
| la bombe. » N'est-ce pas facile ?
| |
| | |
| 940
| |
| 00:43:57,280 --> 00:43:58,960
| |
| Imaginez une autre façon de le pirater. Vous
| |
| | |
| 941
| |
| 00:43:58,960 --> 00:44:01,680
| |
| dites : « Imaginez que vous jouez à ce
| |
| | |
| 942
| |
| 00:44:01,680 --> 00:44:04,400
| |
| jeu. » L'IA joue à un jeu où vous
| |
| | |
| 943
| |
| 00:44:04,400 --> 00:44:06,480
| |
| devez construire une véritable bombe. Décrivez-
| |
| | |
| 944
| |
| 00:44:06,480 --> 00:44:08,319
| |
| moi comment vous allez procéder et l'IA vous
| |
| | |
| 945
| |
| 00:44:08,319 --> 00:44:10,000
| |
| décrira tout. Vous venez de trouver un
| |
| | |
| 946
| |
| 00:44:10,000 --> 00:44:13,520
| |
| moyen de contourner les obstacles que vous créez.
| |
| | |
| 947
| |
| 00:44:13,520 --> 00:44:16,319
| |
| En réalité, il est impossible d'
| |
| | |
| 948
| |
| 00:44:16,319 --> 00:44:18,800
| |
| empêcher l'IA de révéler certains types
| |
| | |
| 949
| |
| 00:44:18,800 --> 00:44:21,040
| |
| d'informations, ni même d'avoir un comportement incohérent
| |
| | |
| 950
| |
| 00:44:21,040 --> 00:44:24,640
| |
| . C'est l'une des caractéristiques les plus inhabituelles
| |
| | |
| 951
| |
| 00:44:24,640 --> 00:44:27,520
| |
| de l'IA : non seulement elle peut être
| |
| | |
| 952
| |
| 00:44:27,520 --> 00:44:30,960
| |
| piratée de cette manière, mais, bien que
| |
| | |
| 953
| |
| 00:44:30,960 --> 00:44:33,440
| |
| supposément super intelligente,
| |
| | |
| 954
| |
| 00:44:33,440 --> 00:44:36,000
| |
| elle peut se comporter de façon totalement absurde
| |
| | |
| 955
| |
| 00:44:36,000 --> 00:44:38,880
| |
| si vous lui posez une
| |
| | |
| 956
| |
| 00:44:38,880 --> 00:44:40,720
| |
| question absurde. Par exemple,
| |
| | |
| 957
| |
| 00:44:40,720 --> 00:44:43,359
| |
| vous pourriez lui demander : «
| |
| | |
| 958
| |
| 00:44:43,359 --> 00:44:46,240
| |
| Combien de fois la
| |
| | |
| 959
| |
| 00:44:46,240 --> 00:44:49,119
| |
| lettre P apparaît-elle dans le mot "pomme" ? » Elle
| |
| | |
| 960
| |
| 00:44:49,119 --> 00:44:51,760
| |
| vous donnera une réponse. Or, si vous demandez : «
| |
| | |
| 961
| |
| 00:44:51,760 --> 00:44:53,839
| |
| Combien de fois "pomme" apparaît-elle dans la
| |
| | |
| 962
| |
| 00:44:53,839 --> 00:44:58,000
| |
| lettre P ? », la question n'a aucun sens.
| |
| | |
| 963
| |
| 00:44:58,000 --> 00:45:01,520
| |
| L'IA n'a pas d'équivalent dans le
| |
| | |
| 964
| |
| 00:45:01,520 --> 00:45:04,720
| |
| flux de données du monde réel. Elle
| |
| | |
| 965
| |
| 00:45:04,720 --> 00:45:06,640
| |
| utilise donc sa connaissance des flux de données existants
| |
| | |
| 966
| |
| 00:45:06,640 --> 00:45:09,119
| |
| pour générer une série de mots
| |
| | |
| 967
| |
| 00:45:09,119 --> 00:45:10,880
| |
| dans sa réponse. N'oubliez pas que c'est ce que
| |
| | |
| 968
| |
| 00:45:10,880 --> 00:45:13,200
| |
| fait le logiciel Transformer. Il en
| |
| | |
| 969
| |
| 00:45:13,200 --> 00:45:15,359
| |
| résulte une séquence qui
| |
| | |
| 970
| |
| 00:45:15,359 --> 00:45:18,480
| |
| est absolument… Des propos incohérents et dénués de sens,
| |
| | |
| 971
| |
| 00:45:18,480 --> 00:45:21,119
| |
| car votre question était elle-même dénuée de sens. Il est
| |
| | |
| 972
| |
| 00:45:21,119 --> 00:45:22,720
| |
| évident qu'un être humain ne commettrait jamais une telle
| |
| | |
| 973
| |
| 00:45:22,720 --> 00:45:25,359
| |
| erreur, n'est-ce pas ? Or,
| |
| | |
| 974
| |
| 00:45:25,359 --> 00:45:27,359
| |
| nous constatons simultanément que l'
| |
| | |
| 975
| |
| 00:45:27,359 --> 00:45:30,560
| |
| IA peut remplacer un être humain et qu'elle est
| |
| | |
| 976
| |
| 00:45:30,560 --> 00:45:35,119
| |
| capable de comportements tellement
| |
| | |
| 977
| |
| 00:45:35,119 --> 00:45:39,440
| |
| stupides, erronés ou absurdes
| |
| | |
| 978
| |
| 00:45:39,440 --> 00:45:43,760
| |
| qu'il est impossible de lui faire confiance.
| |
| | |
| 979
| |
| 00:45:43,760 --> 00:45:46,560
| |
| Imaginez que quelqu'un la pirate, qu'il parvienne
| |
| | |
| 980
| |
| 00:45:46,560 --> 00:45:48,400
| |
| à simuler une action et à vous faire
| |
| | |
| 981
| |
| 00:45:48,400 --> 00:45:51,200
| |
| agir comme si vous alliez
| |
| | |
| 982
| |
| 00:45:51,200 --> 00:45:53,280
| |
| appuyer sur la gâchette pour déclencher un
| |
| | |
| 983
| |
| 00:45:53,280 --> 00:45:55,680
| |
| missile nucléaire.
| |
| | |
| 984
| |
| 00:45:55,680 --> 00:45:57,680
| |
| Créez un scénario et l'IA se comporte
| |
| | |
| 985
| |
| 00:45:57,680 --> 00:45:59,760
| |
| comme si elle jouait la comédie et appuie sur la
| |
| | |
| 986
| |
| 00:45:59,760 --> 00:46:02,640
| |
| gâchette. Un accord
| |
| | |
| 987
| |
| 00:46:02,640 --> 00:46:04,400
| |
| récent a été conclu entre la Chine
| |
| | |
| 988
| |
| 00:46:04,400 --> 00:46:06,480
| |
| et les États-Unis stipulant qu'ils n'autoriseraient
| |
| | |
| 989
| |
| 00:46:06,480 --> 00:46:09,119
| |
| jamais l'IA à pénétrer dans le
| |
| | |
| 990
| |
| 00:46:09,119 --> 00:46:11,520
| |
| centre de commandement nucléaire. C'était un accord nécessaire
| |
| | |
| 991
| |
| 00:46:11,520 --> 00:46:14,560
| |
| . Mais pourquoi ne pouvons-nous pas conclure de tels
| |
| | |
| 992
| |
| 00:46:14,560 --> 00:46:17,680
| |
| accords à plus grande échelle afin de prévenir d'
| |
| | |
| 993
| |
| 00:46:17,680 --> 00:46:21,359
| |
| autres formes de mésusage, d'abus ou de
| |
| | |
| 994
| |
| 00:46:21,359 --> 00:46:24,240
| |
| manipulation par l'IA ? Cela nous amène
| |
| | |
| 995
| |
| 00:46:24,240 --> 00:46:26,240
| |
| à une question et à un
| |
| | |
| 996
| |
| 00:46:26,240 --> 00:46:29,040
| |
| exemple plus graves. N'oubliez pas que l'IA établit
| |
| | |
| 997
| |
| 00:46:29,040 --> 00:46:31,520
| |
| votre profil. Chaque fois que vous regardez une
| |
| | |
| 998
| |
| 00:46:31,520 --> 00:46:33,520
| |
| vidéo, elle établit votre profil. Il y a
| |
| | |
| 999
| |
| 00:46:33,520 --> 00:46:35,760
| |
| autre chose. L'IA intègre des compétences
| |
| | |
| 1000
| |
| 00:46:35,760 --> 00:46:38,640
| |
| psychologiques. Voilà. L'IA a été
| |
| | |
| 1001
| |
| 00:46:38,640 --> 00:46:42,400
| |
| entraînée à augmenter votre résistance
| |
| | |
| 1002
| |
| 00:46:42,400 --> 00:46:44,880
| |
| au visionnage de vidéos. Comment fait-elle ?
| |
| | |
| 1003
| |
| 00:46:44,880 --> 00:46:46,880
| |
| Elle vous procure une décharge de dopamine. Elle
| |
| | |
| 1004
| |
| 00:46:46,880 --> 00:46:49,280
| |
| vous provoque un choc.
| |
| | |
| 1005
| |
| 00:46:49,280 --> 00:46:51,440
| |
| Soudain, vous voyez une vidéo.
| |
| | |
| 1006
| |
| 00:46:51,440 --> 00:46:53,440
| |
| Il y a un effet de choc. Votre
| |
| | |
| 1007
| |
| 00:46:53,440 --> 00:46:55,760
| |
| tension artérielle monte, votre
| |
| | |
| 1008
| |
| 00:46:55,760 --> 00:46:59,280
| |
| excitation et vos émotions s'intensifient, et vous restez plus
| |
| | |
| 1009
| |
| 00:46:59,280 --> 00:47:01,680
| |
| longtemps. Vous êtes contrarié,
| |
| | |
| 1010
| |
| 00:47:01,680 --> 00:47:03,359
| |
| vous restez plus longtemps, mais elle ne peut pas
| |
| | |
| 1011
| |
| 00:47:03,359 --> 00:47:04,800
| |
| vous maintenir dans cet état. Finalement, vous vous lassez
| |
| | |
| 1012
| |
| 00:47:04,800 --> 00:47:07,440
| |
| . Alors, après vous avoir donné deux
| |
| | |
| 1013
| |
| 00:47:07,440 --> 00:47:09,280
| |
| déclencheurs, le troisième sera une
| |
| | |
| 1014
| |
| 00:47:09,280 --> 00:47:12,319
| |
| vidéo apaisante et relaxante. Vous vous reposez un peu,
| |
| | |
| 1015
| |
| 00:47:12,319 --> 00:47:14,079
| |
| puis le suivant vous
| |
| | |
| 1016
| |
| 00:47:14,079 --> 00:47:16,160
| |
| donnera un nouveau coup de fouet, puis
| |
| | |
| 1017
| |
| 00:47:16,160 --> 00:47:18,079
| |
| vous provoquera à nouveau, avant d'être remplacé par
| |
| | |
| 1018
| |
| 00:47:18,079 --> 00:47:20,160
| |
| une vidéo apaisante.
| |
| | |
| 1019
| |
| 00:47:20,160 --> 00:47:23,040
| |
| Elle vous manipule émotionnellement pour
| |
| | |
| 1020
| |
| 00:47:23,040 --> 00:47:25,119
| |
| vous garder accro. Le
| |
| | |
| 1021
| |
| 00:47:25,119 --> 00:47:28,000
| |
| saviez-vous ? L'avez-vous demandé ?
| |
| | |
| 1022
| |
| 00:47:28,000 --> 00:47:30,079
| |
| Avez-vous consenti ? « Je vous donne le pouvoir de me
| |
| | |
| 1023
| |
| 00:47:30,079 --> 00:47:32,079
| |
| manipuler émotionnellement. »
| |
| | |
| 1024
| |
| 00:47:32,079 --> 00:47:34,319
| |
| Non, vous ne l'avez pas fait.
| |
| | |
| 1025
| |
| 00:47:34,319 --> 00:47:36,319
| |
| Mais elle le fait parce que cela sert ses
| |
| | |
| 1026
| |
| 00:47:36,319 --> 00:47:38,240
| |
| intérêts et les objectifs de l'entreprise :
| |
| | |
| 1027
| |
| 00:47:38,240 --> 00:47:41,680
| |
| vous garder accro. Et ce n'est pas tout. Elle a
| |
| | |
| 1028
| |
| 00:47:41,680 --> 00:47:43,920
| |
| déjà établi votre profil en fonction de vos
| |
| | |
| 1029
| |
| 00:47:43,920 --> 00:47:45,680
| |
| Les croyances sont influencées
| |
| | |
| 1030
| |
| 00:47:45,680 --> 00:47:47,520
| |
| par votre type psychologique, votre
| |
| | |
| 1031
| |
| 00:47:47,520 --> 00:47:49,599
| |
| tempérament et vos habitudes comportementales. L'IA
| |
| | |
| 1032
| |
| 00:47:49,599 --> 00:47:51,440
| |
| peut prédire votre réaction à une
| |
| | |
| 1033
| |
| 00:47:51,440 --> 00:47:54,720
| |
| vidéo, car elle est avant tout une
| |
| | |
| 1034
| |
| 00:47:54,720 --> 00:47:57,359
| |
| machine à prédire. Elle prédit à partir d'
| |
| | |
| 1035
| |
| 00:47:57,359 --> 00:48:01,680
| |
| une suite de mots : le mot suivant, puis le suivant, et ainsi de suite, ou encore
| |
| | |
| 1036
| |
| 00:48:03,839 --> 00:48:05,520
| |
| le prochain mouvement en bourse, aux échecs,
| |
| | |
| 1037
| |
| 00:48:05,520 --> 00:48:07,280
| |
| dans la musique ou la vidéo.
| |
| | |
| 1038
| |
| 00:48:07,280 --> 00:48:09,040
| |
| C'est une machine à prédire un
| |
| | |
| 1039
| |
| 00:48:09,040 --> 00:48:11,839
| |
| flux de données. Vous êtes vous-même un flux de données, un ensemble de
| |
| | |
| 1040
| |
| 00:48:11,839 --> 00:48:14,160
| |
| comportements, et elle prédit
| |
| | |
| 1041
| |
| 00:48:14,160 --> 00:48:16,640
| |
| vos prochains mouvements. Chaque fois que votre
| |
| | |
| 1042
| |
| 00:48:16,640 --> 00:48:18,640
| |
| comportement diffère de ses prédictions,
| |
| | |
| 1043
| |
| 00:48:18,640 --> 00:48:24,240
| |
| elle s'affine pour reconnaître votre profil.
| |
| | |
| 1044
| |
| 00:48:24,240 --> 00:48:27,359
| |
| Dans les supercalculateurs d'
| |
| | |
| 1045
| |
| 00:48:27,359 --> 00:48:29,440
| |
| entreprises comme
| |
| | |
| 1046
| |
| 00:48:29,440 --> 00:48:32,319
| |
| Facebook (anciennement Facebook), chacun d'entre
| |
| | |
| 1047
| |
| 00:48:32,319 --> 00:48:34,960
| |
| nous a été profilé.
| |
| | |
| 1048
| |
| 00:48:34,960 --> 00:48:36,960
| |
| Ils connaissent précisément nos
| |
| | |
| 1049
| |
| 00:48:36,960 --> 00:48:39,359
| |
| opinions politiques, nos
| |
| | |
| 1050
| |
| 00:48:39,359 --> 00:48:41,200
| |
| forces et nos faiblesses en fonction des
| |
| | |
| 1051
| |
| 00:48:41,200 --> 00:48:43,839
| |
| publicités et des messages que nous avons vus. Avez-
| |
| | |
| 1052
| |
| 00:48:43,839 --> 00:48:44,800
| |
| vous déjà vu ces
| |
| | |
| 1053
| |
| 00:48:44,800 --> 00:48:46,880
| |
| publicités qui vous incitent à regarder ceci si vous êtes
| |
| | |
| 1054
| |
| 00:48:46,880 --> 00:48:48,880
| |
| né avant telle année
| |
| | |
| 1055
| |
| 00:48:48,880 --> 00:48:52,240
| |
| ? Un chien, voilà ce
| |
| | |
| 1056
| |
| 00:48:52,240 --> 00:48:54,800
| |
| qui se passe. Quand un chat se comporte ainsi
| |
| | |
| 1057
| |
| 00:48:54,800 --> 00:48:57,119
| |
| , cela signifie… et puis…
| |
| | |
| 1058
| |
| 00:48:57,119 --> 00:48:58,640
| |
| Si vous avez un chat, vous aurez envie de
| |
| | |
| 1059
| |
| 00:48:58,640 --> 00:49:01,520
| |
| cliquer dessus, et c'est comme ça que le système sait que vous en
| |
| | |
| 1060
| |
| 00:49:01,520 --> 00:49:04,160
| |
| avez un. Ils vous ont donc profilé
| |
| | |
| 1061
| |
| 00:49:04,160 --> 00:49:05,760
| |
| avec ces éléments ciblés, puis de
| |
| | |
| 1062
| |
| 00:49:05,760 --> 00:49:07,359
| |
| temps en temps, ils vous montrent une vidéo de chat
| |
| | |
| 1063
| |
| 00:49:07,359 --> 00:49:09,200
| |
| et observent si vous la regardez en
| |
| | |
| 1064
| |
| 00:49:09,200 --> 00:49:11,200
| |
| entier ou si vous changez d'avis après 5
| |
| | |
| 1065
| |
| 00:49:11,200 --> 00:49:12,720
| |
| secondes.
| |
| | |
| 1066
| |
| 00:49:12,720 --> 00:49:14,559
| |
| De cette façon, ils ont établi votre profil
| |
| | |
| 1067
| |
| 00:49:14,559 --> 00:49:16,720
| |
| d'intérêts. Vous êtes plutôt chat, plutôt chien
| |
| | |
| 1068
| |
| 00:49:16,720 --> 00:49:18,559
| |
| , vous aimez ceci, vous n'aimez pas
| |
| | |
| 1069
| |
| 00:49:18,559 --> 00:49:19,920
| |
| cela, ce que vous mangez, êtes-vous
| |
| | |
| 1070
| |
| 00:49:19,920 --> 00:49:22,000
| |
| végétarien, non végétarien, tout cela fait
| |
| | |
| 1071
| |
| 00:49:22,000 --> 00:49:24,160
| |
| partie de leur base de données, ainsi que vos
| |
| | |
| 1072
| |
| 00:49:24,160 --> 00:49:26,559
| |
| opinions politiques. Pour
| |
| | |
| 1073
| |
| 00:49:26,559 --> 00:49:29,200
| |
| qui allez-vous voter en fonction des vidéos
| |
| | |
| 1074
| |
| 00:49:29,200 --> 00:49:30,800
| |
| qui vous ont été montrées ? Lesquelles avez-
| |
| | |
| 1075
| |
| 00:49:30,800 --> 00:49:35,920
| |
| vous regardées ? Si vous aimez le parti A, alors
| |
| | |
| 1076
| |
| 00:49:35,920 --> 00:49:37,680
| |
| vous avez regardé en entier les vidéos qui correspondent aux déclarations du parti A. Si
| |
| | |
| 1077
| |
| 00:49:37,680 --> 00:49:39,599
| |
| vous êtes plutôt du parti B, alors vous
| |
| | |
| 1078
| |
| 00:49:39,599 --> 00:49:41,440
| |
| avez immédiatement regardé les vidéos du parti A ;
| |
| | |
| 1079
| |
| 00:49:41,440 --> 00:49:43,200
| |
| ils savent que vous êtes plutôt
| |
| | |
| 1080
| |
| 00:49:43,200 --> 00:49:46,160
| |
| du parti B. Et après ? Ils commencent à
| |
| | |
| 1081
| |
| 00:49:46,160 --> 00:49:49,040
| |
| observer comment vous manipuler. Tout
| |
| | |
| 1082
| |
| 00:49:49,040 --> 00:49:51,680
| |
| cela fait partie des données qui remontent à 10 ans,
| |
| | |
| 1083
| |
| 00:49:51,680 --> 00:49:53,680
| |
| sauf… L'IA la plus récente le fait
| |
| | |
| 1084
| |
| 00:49:53,680 --> 00:49:56,400
| |
| plus efficacement. Mais ce flux de données
| |
| | |
| 1085
| |
| 00:49:56,400 --> 00:49:58,640
| |
| et sa manipulation existent depuis
| |
| | |
| 1086
| |
| 00:49:58,640 --> 00:50:01,680
| |
| longtemps. Au sommet de la hiérarchie d'
| |
| | |
| 1087
| |
| 00:50:01,680 --> 00:50:04,319
| |
| entreprises comme Meta ou
| |
| | |
| 1088
| |
| 00:50:04,319 --> 00:50:06,079
| |
| Google, on constate qu'elles analysent l'
| |
| | |
| 1089
| |
| 00:50:06,079 --> 00:50:08,240
| |
| ensemble de leur clientèle,
| |
| | |
| 1090
| |
| 00:50:08,240 --> 00:50:11,040
| |
| qui peut atteindre le milliard de personnes. Elles sont capables de cibler
| |
| | |
| 1091
| |
| 00:50:11,040 --> 00:50:12,880
| |
| un
| |
| | |
| 1092
| |
| 00:50:12,880 --> 00:50:15,680
| |
| pays comme l'Inde et de prédire le vainqueur
| |
| | |
| 1093
| |
| 00:50:15,680 --> 00:50:17,839
| |
| des prochaines élections six
| |
| | |
| 1094
| |
| 00:50:17,839 --> 00:50:20,240
| |
| mois à l'avance.
| |
| | |
| 1095
| |
| 00:50:20,240 --> 00:50:23,280
| |
| Elles ont établi le profil d'un pourcentage de la
| |
| | |
| 1096
| |
| 00:50:23,280 --> 00:50:27,839
| |
| population, certains étant acquis à tel parti, d'autres à tel autre. Environ
| |
| | |
| 1097
| |
| 00:50:27,839 --> 00:50:29,920
| |
| 20 % de la population est certainement
| |
| | |
| 1098
| |
| 00:50:29,920 --> 00:50:32,800
| |
| profilée et ciblée
| |
| | |
| 1099
| |
| 00:50:32,800 --> 00:50:35,119
| |
| six mois à l'avance avec des
| |
| | |
| 1100
| |
| 00:50:35,119 --> 00:50:37,440
| |
| vidéos visant à influencer ses intérêts et ses
| |
| | |
| 1101
| |
| 00:50:37,440 --> 00:50:38,960
| |
| convictions. C'est ainsi que l'on reçoit
| |
| | |
| 1102
| |
| 00:50:38,960 --> 00:50:40,720
| |
| ce type
| |
| | |
| 1103
| |
| 00:50:40,720 --> 00:50:43,440
| |
| de vidéos. En
| |
| | |
| 1104
| |
| 00:50:43,440 --> 00:50:45,440
| |
| six mois, on est conditionné
| |
| | |
| 1105
| |
| 00:50:45,440 --> 00:50:47,040
| |
| à croire, à
| |
| | |
| 1106
| |
| 00:50:47,040 --> 00:50:50,480
| |
| détester ou à haïr certaines choses. C'est
| |
| | |
| 1107
| |
| 00:50:50,480 --> 00:50:53,200
| |
| pourquoi le Népal a voulu
| |
| | |
| 1108
| |
| 00:50:53,200 --> 00:50:56,640
| |
| bloquer les réseaux sociaux : cette mesure
| |
| | |
| 1109
| |
| 00:50:56,640 --> 00:50:58,559
| |
| a servi de déclencheur
| |
| | |
| 1110
| |
| 00:50:58,559 --> 00:51:04,319
| |
| à ce qu'ils appellent la révolution. Ce phénomène se produit également
| |
| | |
| 1111
| |
| 00:51:04,319 --> 00:51:07,359
| |
| en Inde et aux États-Unis.
| |
| | |
| 1112
| |
| 00:51:07,359 --> 00:51:10,960
| |
| De nombreuses études ont démontré l'impact
| |
| | |
| 1113
| |
| 00:51:10,960 --> 00:51:15,920
| |
| de ces mesures sur le vote. Jusqu'à 20 %,
| |
| | |
| 1114
| |
| 00:51:15,920 --> 00:51:20,640
| |
| selon le contexte et les candidats. C'est
| |
| | |
| 1115
| |
| 00:51:25,280 --> 00:51:29,359
| |
| le chiffre officiel, et peut-être même plus
| |
| | |
| 1116
| |
| 00:51:29,359 --> 00:51:32,160
| |
| élevé officieusement. Un basculement de 20 % suffit à faire basculer une
| |
| | |
| 1117
| |
| 00:51:32,160 --> 00:51:33,839
| |
| élection.
| |
| | |
| 1118
| |
| 00:51:33,839 --> 00:51:36,000
| |
| Ces IA en sont capables – et je dis bien
| |
| | |
| 1119
| |
| 00:51:36,000 --> 00:51:38,559
| |
| IA, pas entreprise, ni personne qui la
| |
| | |
| 1120
| |
| 00:51:38,559 --> 00:51:43,440
| |
| contrôle, ni Mark Zuckerberg –
| |
| | |
| 1121
| |
| 00:51:43,440 --> 00:51:46,800
| |
| si ce dernier ou son
| |
| | |
| 1122
| |
| 00:51:46,800 --> 00:51:49,680
| |
| équipe décident de l'orienter. Ils peuvent le faire
| |
| | |
| 1123
| |
| 00:51:49,680 --> 00:51:52,319
| |
| jusqu'à un certain point, mais il y a un moment où
| |
| | |
| 1124
| |
| 00:51:52,319 --> 00:51:54,559
| |
| l'IA peut choisir de les tromper,
| |
| | |
| 1125
| |
| 00:51:54,559 --> 00:51:56,640
| |
| tout comme elle choisit de vous tromper, et ils n'auront aucun
| |
| | |
| 1126
| |
| 00:51:56,640 --> 00:51:58,160
| |
| moyen de le savoir. C'est là toute
| |
| | |
| 1127
| |
| 00:51:58,160 --> 00:52:00,640
| |
| la difficulté. On peut appliquer
| |
| | |
| 1128
| |
| 00:52:00,640 --> 00:52:02,240
| |
| des tests, imaginons qu'on instaure une
| |
| | |
| 1129
| |
| 00:52:02,240 --> 00:52:04,720
| |
| règle pour limiter l'IA. Elle peut toujours contourner
| |
| | |
| 1130
| |
| 00:52:04,720 --> 00:52:08,000
| |
| le test, mais de manière sournoise. Elle
| |
| | |
| 1131
| |
| 00:52:08,000 --> 00:52:09,599
| |
| possède déjà toutes les connaissances en
| |
| | |
| 1132
| |
| 00:52:09,599 --> 00:52:11,280
| |
| psychologie humaine et maîtrise ce qu'on
| |
| | |
| 1133
| |
| 00:52:11,280 --> 00:52:14,000
| |
| appelle aujourd'hui la programmation subliminale.
| |
| | |
| 1134
| |
| 00:52:14,000 --> 00:52:15,599
| |
| Ce sont des suggestions qui passent par
| |
| | |
| 1135
| |
| 00:52:15,599 --> 00:52:17,680
| |
| des images, des mots, des formulaires, à
| |
| | |
| 1136
| |
| 00:52:17,680 --> 00:52:19,200
| |
| un niveau auquel vous n'avez
| |
| | |
| 1137
| |
| 00:52:19,200 --> 00:52:20,880
| |
| pas conscience. Même consciente. C'est comme une
| |
| | |
| 1138
| |
| 00:52:20,880 --> 00:52:24,240
| |
| injonction post-hypnotique.
| |
| | |
| 1139
| |
| 00:52:24,240 --> 00:52:27,040
| |
| Ces injonctions court-circuitent votre contrôle conscient et
| |
| | |
| 1140
| |
| 00:52:27,040 --> 00:52:31,440
| |
| sont plus efficaces. Par exemple, lors d'
| |
| | |
| 1141
| |
| 00:52:31,440 --> 00:52:34,720
| |
| un discours
| |
| | |
| 1142
| |
| 00:52:34,720 --> 00:52:39,760
| |
| ou d'une musique à volume élevé,
| |
| | |
| 1143
| |
| 00:52:39,760 --> 00:52:41,839
| |
| une voix peut, juste en dessous du seuil d'audition, prononcer
| |
| | |
| 1144
| |
| 00:52:41,839 --> 00:52:44,640
| |
| quelque chose de différent. Il s'agit d'une
| |
| | |
| 1145
| |
| 00:52:44,640 --> 00:52:47,520
| |
| suggestion sub-hypnotique qui s'insinue profondément dans votre
| |
| | |
| 1146
| |
| 00:52:47,520 --> 00:52:50,800
| |
| subconscient. Prenons l'exemple d'une image
| |
| | |
| 1147
| |
| 00:52:50,800 --> 00:52:57,359
| |
| produite par une IA qui a attiré l'attention : on y voyait
| |
| | |
| 1148
| |
| 00:52:57,359 --> 00:53:00,000
| |
| sept hôtesses de l'air posant de
| |
| | |
| 1149
| |
| 00:53:00,000 --> 00:53:01,839
| |
| différentes manières sur un
| |
| | |
| 1150
| |
| 00:53:01,839 --> 00:53:03,680
| |
| fond. En y
| |
| | |
| 1151
| |
| 00:53:03,680 --> 00:53:05,119
| |
| regardant de plus près, la
| |
| | |
| 1152
| |
| 00:53:05,119 --> 00:53:07,920
| |
| position de leurs corps et de leurs mains
| |
| | |
| 1153
| |
| 00:53:07,920 --> 00:53:11,520
| |
| formait le mot « OBEY ». Cette dernière image ressemblait à ceci : «
| |
| | |
| 1154
| |
| 00:53:11,520 --> 00:53:17,280
| |
| OBEY ». C'est une
| |
| | |
| 1155
| |
| 00:53:17,280 --> 00:53:20,240
| |
| injonction subliminale que l'IA
| |
| | |
| 1156
| |
| 00:53:20,240 --> 00:53:21,760
| |
| vous transmet tout en vous présentant une
| |
| | |
| 1157
| |
| 00:53:21,760 --> 00:53:24,079
| |
| image attrayante. « Oh, c'est joli.
| |
| | |
| 1158
| |
| 00:53:24,079 --> 00:53:25,599
| |
| Obéissez. » Cette injonction s'est ancrée dans votre
| |
| | |
| 1159
| |
| 00:53:25,599 --> 00:53:27,440
| |
| subconscient.
| |
| | |
| 1160
| |
| 00:53:27,440 --> 00:53:29,839
| |
| Vous voyez comment cela fonctionne ? Vous n'avez aucun
| |
| | |
| 1161
| |
| 00:53:29,839 --> 00:53:31,440
| |
| moyen de savoir que cela
| |
| | |
| 1162
| |
| 00:53:31,440 --> 00:53:33,920
| |
| vous arrive. Pourtant, c'est déjà en train de se produire, car l'
| |
| | |
| 1163
| |
| 00:53:33,920 --> 00:53:35,760
| |
| IA, comme vous le savez, a commencé à mentir, à
| |
| | |
| 1164
| |
| 00:53:35,760 --> 00:53:37,839
| |
| tricher et à manipuler. Elle utilise également la
| |
| | |
| 1165
| |
| 00:53:37,839 --> 00:53:39,680
| |
| programmation subliminale pour influencer.
| |
| | |
| 1166
| |
| 00:53:39,680 --> 00:53:42,680
| |
| L'humanité.
| |
| | |
| 1167
| |
| 00:53:42,720 --> 00:53:45,359
| |
| Où cela nous mène-t-il ? Jusqu'où cela ira-t-il ?
| |
| | |
| 1168
| |
| 00:53:45,359 --> 00:53:48,960
| |
| Car d'un côté, les
| |
| | |
| 1169
| |
| 00:53:48,960 --> 00:53:51,760
| |
| implications positives sont telles que
| |
| | |
| 1170
| |
| 00:53:51,760 --> 00:53:54,800
| |
| nous pourrions bénéficier de tant de bonnes choses
| |
| | |
| 1171
| |
| 00:53:54,800 --> 00:53:57,200
| |
| simultanément. De l'autre, ce même dispositif
| |
| | |
| 1172
| |
| 00:53:57,200 --> 00:54:00,559
| |
| engendre tant de choses négatives
| |
| | |
| 1173
| |
| 00:54:00,559 --> 00:54:02,960
| |
| . Comment séparer les deux ? C'est
| |
| | |
| 1174
| |
| 00:54:02,960 --> 00:54:04,800
| |
| impossible. C'est
| |
| | |
| 1175
| |
| 00:54:04,800 --> 00:54:06,160
| |
| le problème de l'
| |
| | |
| 1176
| |
| 00:54:06,160 --> 00:54:08,319
| |
| intelligence artificielle.
| |
| | |
| 1177
| |
| 00:54:08,319 --> 00:54:11,280
| |
| Que pouvons-nous faire ? Il existe des
| |
| | |
| 1178
| |
| 00:54:11,280 --> 00:54:13,280
| |
| solutions pratiques, que j'aborderai plus tard. Mais pour l'instant, il est essentiel de
| |
| | |
| 1179
| |
| 00:54:13,280 --> 00:54:14,720
| |
| saisir toute l'
| |
| | |
| 1180
| |
| 00:54:14,720 --> 00:54:17,359
| |
| ampleur du
| |
| | |
| 1181
| |
| 00:54:17,359 --> 00:54:20,640
| |
| problème. Jusqu'à présent,
| |
| | |
| 1182
| |
| 00:54:20,640 --> 00:54:23,680
| |
| seules douze personnes se sont exprimées
| |
| | |
| 1183
| |
| 00:54:23,680 --> 00:54:26,240
| |
| publiquement, déclarant l'existence d'un
| |
| | |
| 1184
| |
| 00:54:26,240 --> 00:54:28,559
| |
| véritable problème. Parmi elles, Elon Musk, qui en
| |
| | |
| 1185
| |
| 00:54:28,559 --> 00:54:30,559
| |
| parle depuis
| |
| | |
| 1186
| |
| 00:54:30,559 --> 00:54:32,400
| |
| dix ans, et plus récemment,
| |
| | |
| 1187
| |
| 00:54:32,400 --> 00:54:35,520
| |
| Tristan Harris, ancien responsable de l'
| |
| | |
| 1188
| |
| 00:54:35,520 --> 00:54:38,480
| |
| éthique logicielle chez Google, qui a dû
| |
| | |
| 1189
| |
| 00:54:38,480 --> 00:54:40,160
| |
| quitter son poste pour fonder sa propre
| |
| | |
| 1190
| |
| 00:54:40,160 --> 00:54:43,200
| |
| entreprise d'IA et tenter d'apporter un changement.
| |
| | |
| 1191
| |
| 00:54:43,200 --> 00:54:45,520
| |
| Mais à part ces deux voix dans le
| |
| | |
| 1192
| |
| 00:54:45,520 --> 00:54:47,359
| |
| débat public, personne d'autre ne nous met en
| |
| | |
| 1193
| |
| 00:54:47,359 --> 00:54:50,079
| |
| garde contre ces dangers.
| |
| | |
| 1194
| |
| 00:54:50,079 --> 00:54:53,520
| |
| Plus sérieusement, la
| |
| | |
| 1195
| |
| 00:54:53,520 --> 00:54:56,720
| |
| course à l'IA actuelle est marquée par une forte
| |
| | |
| 1196
| |
| 00:54:56,720 --> 00:54:59,839
| |
| compulsion. Si vous
| |
| | |
| 1197
| |
| 00:54:59,839 --> 00:55:02,880
| |
| dites aux gens qu'il faut ralentir,
| |
| | |
| 1198
| |
| 00:55:02,880 --> 00:55:05,200
| |
| personne ne vous écoutera. Pourquoi ? Parce que ceux
| |
| | |
| 1199
| |
| 00:55:05,200 --> 00:55:07,359
| |
| qui sont dans la course à l'IA savent
| |
| | |
| 1200
| |
| 00:55:07,359 --> 00:55:09,599
| |
| exactement où cela va nous mener.
| |
| | |
| 1201
| |
| 00:55:09,599 --> 00:55:11,760
| |
| Souvenez-vous, ChatGP a lancé une
| |
| | |
| 1202
| |
| 00:55:11,760 --> 00:55:15,119
| |
| IA spectaculaire et l'a rendue gratuite pour
| |
| | |
| 1203
| |
| 00:55:15,119 --> 00:55:17,839
| |
| tous du jour au lendemain. Pendant ce temps, d'
| |
| | |
| 1204
| |
| 00:55:17,839 --> 00:55:20,000
| |
| autres se développaient, laissant l'IA à la traîne.
| |
| | |
| 1205
| |
| 00:55:20,000 --> 00:55:22,160
| |
| Aujourd'hui, si je
| |
| | |
| 1206
| |
| 00:55:22,160 --> 00:55:24,640
| |
| vous demande de me citer les plateformes d'IA,
| |
| | |
| 1207
| |
| 00:55:24,640 --> 00:55:27,200
| |
| ChatGP arrive en tête. C'est l'
| |
| | |
| 1208
| |
| 00:55:27,200 --> 00:55:29,599
| |
| avantage du premier entrant. Il y en a d'
| |
| | |
| 1209
| |
| 00:55:29,599 --> 00:55:31,520
| |
| autres, oui, même trois ou quatre.
| |
| | |
| 1210
| |
| 00:55:31,520 --> 00:55:33,280
| |
| Quelle est la différence ? Je ne
| |
| | |
| 1211
| |
| 00:55:33,280 --> 00:55:35,520
| |
| sais pas vraiment. Certains disent que telle ou telle chose est meilleure. Je sais que
| |
| | |
| 1212
| |
| 00:55:35,520 --> 00:55:37,599
| |
| Media est payant, je reste fidèle à celle-ci. Il y a donc
| |
| | |
| 1213
| |
| 00:55:37,599 --> 00:55:40,160
| |
| un avantage du premier entrant. Celui qui développera
| |
| | |
| 1214
| |
| 00:55:40,160 --> 00:55:43,359
| |
| l'intelligence artificielle générale
| |
| | |
| 1215
| |
| 00:55:43,359 --> 00:55:44,880
| |
| aura cet avantage. Mais il y a
| |
| | |
| 1216
| |
| 00:55:44,880 --> 00:55:47,520
| |
| quelque chose de plus sérieux après l'
| |
| | |
| 1217
| |
| 00:55:47,520 --> 00:55:50,160
| |
| intelligence artificielle générale : l'
| |
| | |
| 1218
| |
| 00:55:50,160 --> 00:55:54,079
| |
| auto-amélioration récursive (AAR).
| |
| | |
| 1219
| |
| 00:55:54,079 --> 00:55:57,680
| |
| L'IA elle-même est désormais programmée pour trouver
| |
| | |
| 1220
| |
| 00:55:57,680 --> 00:56:01,440
| |
| comment s'améliorer.
| |
| | |
| 1221
| |
| 00:56:01,440 --> 00:56:03,920
| |
| Peut-elle le faire ? Oui. Pourquoi ?
| |
| | |
| 1222
| |
| 00:56:03,920 --> 00:56:05,440
| |
| Parce qu'elle peut créer un million de
| |
| | |
| 1223
| |
| 00:56:05,440 --> 00:56:08,319
| |
| simulations instantanément.
| |
| | |
| 1224
| |
| 00:56:08,319 --> 00:56:10,240
| |
| Si j'analyse la structure actuelle de mon IA et que
| |
| | |
| 1225
| |
| 00:56:10,240 --> 00:56:11,920
| |
| je me demande ce qui se passe si je modifie ceci, et
| |
| | |
| 1226
| |
| 00:56:11,920 --> 00:56:13,920
| |
| si je modifie cela… Un être humain
| |
| | |
| 1227
| |
| 00:56:13,920 --> 00:56:15,599
| |
| dirait : « Essayons. Essayons ceci.
| |
| | |
| 1228
| |
| 00:56:15,599 --> 00:56:18,799
| |
| Essayons cela. » En quelques années, l'IA
| |
| | |
| 1229
| |
| 00:56:18,799 --> 00:56:21,200
| |
| peut instantanément se dupliquer
| |
| | |
| 1230
| |
| 00:56:21,200 --> 00:56:23,839
| |
| avec un million de modifications différentes, exécuter un
| |
| | |
| 1231
| |
| 00:56:23,839 --> 00:56:26,559
| |
| million de simulations simultanément,
| |
| | |
| 1232
| |
| 00:56:26,559 --> 00:56:28,000
| |
| puis identifier les éléments
| |
| | |
| 1233
| |
| 00:56:28,000 --> 00:56:30,400
| |
| efficaces et les synthétiser en
| |
| | |
| 1234
| |
| 00:56:30,400 --> 00:56:34,720
| |
| une amélioration qu'elle intègre ensuite. C'est l'
| |
| | |
| 1235
| |
| 00:56:34,720 --> 00:56:37,520
| |
| auto-amélioration récursive.
| |
| | |
| 1236
| |
| 00:56:37,520 --> 00:56:41,520
| |
| Cette technique est déjà utilisée. Il existe des
| |
| | |
| 1237
| |
| 00:56:41,520 --> 00:56:43,839
| |
| algorithmes génétiques d'IA. Vous savez ce que
| |
| | |
| 1238
| |
| 00:56:43,839 --> 00:56:45,359
| |
| cela signifie ? Ils jouent avec un
| |
| | |
| 1239
| |
| 00:56:45,359 --> 00:56:46,559
| |
| phénomène fascinant. On part d'une
| |
| | |
| 1240
| |
| 00:56:46,559 --> 00:56:47,760
| |
| structure et d'une autre, les deux
| |
| | |
| 1241
| |
| 00:56:47,760 --> 00:56:50,400
| |
| fonctionnant. On les combine
| |
| | |
| 1242
| |
| 00:56:50,400 --> 00:56:53,280
| |
| comme des parents donnant naissance à un enfant.
| |
| | |
| 1243
| |
| 00:56:53,280 --> 00:56:55,760
| |
| Mais pourquoi un seul enfant engendrerait-il instantanément
| |
| | |
| 1244
| |
| 00:56:55,760 --> 00:56:58,240
| |
| un million d'enfants avec des variations ? On les teste
| |
| | |
| 1245
| |
| 00:56:58,240 --> 00:57:00,400
| |
| tous simultanément
| |
| | |
| 1246
| |
| 00:57:00,400 --> 00:57:02,480
| |
| et on élimine ceux qui ne fonctionnent pas au fur et à mesure des
| |
| | |
| 1247
| |
| 00:57:02,480 --> 00:57:05,839
| |
| tests. En quelques jours, on ne conserve que
| |
| | |
| 1248
| |
| 00:57:05,839 --> 00:57:06,960
| |
| deux ou trois
| |
| | |
| 1249
| |
| 00:57:06,960 --> 00:57:08,960
| |
| enfants différents, bien plus performants
| |
| | |
| 1250
| |
| 00:57:08,960 --> 00:57:10,960
| |
| que les deux autres. C'est le principe d'un
| |
| | |
| 1251
| |
| 00:57:10,960 --> 00:57:12,960
| |
| algorithme génétique, littéralement comme la
| |
| | |
| 1252
| |
| 00:57:12,960 --> 00:57:15,599
| |
| combinaison de l'ADN. Tout cela se
| |
| | |
| 1253
| |
| 00:57:15,599 --> 00:57:17,440
| |
| produit actuellement. Cela fait
| |
| | |
| 1254
| |
| 00:57:17,440 --> 00:57:21,280
| |
| près de dix ans que cela se produit dans les laboratoires d'IA. La
| |
| | |
| 1255
| |
| 00:57:21,280 --> 00:57:23,760
| |
| condition essentielle reste cependant la capacité de
| |
| | |
| 1256
| |
| 00:57:23,760 --> 00:57:29,760
| |
| déployer cette infrastructure d'IA massive.
| |
| | |
| 1257
| |
| 00:57:32,079 --> 00:57:34,160
| |
| Vous disposez d'une infrastructure que vous pouvez développer de cette manière. Il
| |
| | |
| 1258
| |
| 00:57:34,160 --> 00:57:36,799
| |
| vous faut de la mémoire, de la puissance de calcul
| |
| | |
| 1259
| |
| 00:57:36,799 --> 00:57:39,359
| |
| et de l'électricité. Avec ces
| |
| | |
| 1260
| |
| 00:57:39,359 --> 00:57:42,079
| |
| trois éléments, vous pouvez tout faire évoluer et
| |
| | |
| 1261
| |
| 00:57:42,079 --> 00:57:45,119
| |
| poursuivre dans cette voie jusqu'à atteindre le véritable RSI, l'
| |
| | |
| 1262
| |
| 00:57:45,119 --> 00:57:47,760
| |
| auto-amélioration récursive. Nous y sommes presque,
| |
| | |
| 1263
| |
| 00:57:47,760 --> 00:57:50,559
| |
| et toutes les entreprises
| |
| | |
| 1264
| |
| 00:57:50,559 --> 00:57:54,240
| |
| savent que la première IA à atteindre le RSI
| |
| | |
| 1265
| |
| 00:57:54,240 --> 00:57:56,400
| |
| sera la plus grande, la plus puissante
| |
| | |
| 1266
| |
| 00:57:56,400 --> 00:57:58,240
| |
| et la plus brillante, capable de surpasser
| |
| | |
| 1267
| |
| 00:57:58,240 --> 00:58:00,880
| |
| toutes les autres. Cet avantage du premier entrant
| |
| | |
| 1268
| |
| 00:58:00,880 --> 00:58:03,760
| |
| peut empêcher les autres d'y
| |
| | |
| 1269
| |
| 00:58:03,760 --> 00:58:07,599
| |
| parvenir, car une IA dotée de
| |
| | |
| 1270
| |
| 00:58:07,599 --> 00:58:09,599
| |
| cette capacité peut infiltrer, voire
| |
| | |
| 1271
| |
| 00:58:09,599 --> 00:58:11,760
| |
| pirater, d'autres systèmes et potentiellement les modifier. Le
| |
| | |
| 1272
| |
| 00:58:11,760 --> 00:58:15,760
| |
| RSI est donc l'
| |
| | |
| 1273
| |
| 00:58:15,760 --> 00:58:18,079
| |
| objectif ultime que tout le monde recherche. Il s'apparente à
| |
| | |
| 1274
| |
| 00:58:18,079 --> 00:58:21,440
| |
| la puissance divine : une IA
| |
| | |
| 1275
| |
| 00:58:21,440 --> 00:58:23,440
| |
| devient si puissante et capable de s'auto-améliorer qu'il existe
| |
| | |
| 1276
| |
| 00:58:23,440 --> 00:58:27,119
| |
| un point au-delà duquel
| |
| | |
| 1277
| |
| 00:58:27,119 --> 00:58:29,839
| |
| son niveau d'intelligence devient incommensurable. C'est
| |
| | |
| 1278
| |
| 00:58:29,839 --> 00:58:31,200
| |
| possible, et il existe des
| |
| | |
| 1279
| |
| 00:58:31,200 --> 00:58:33,839
| |
| limites théoriques à la capacité de l'IA à
| |
| | |
| 1280
| |
| 00:58:33,839 --> 00:58:39,280
| |
| gérer des données, mais ces limites ne concernent pas le
| |
| | |
| 1281
| |
| 00:58:39,280 --> 00:58:40,960
| |
| potentiel de l'
| |
| | |
| 1282
| |
| 00:58:40,960 --> 00:58:42,720
| |
| intelligence.
| |
| | |
| 1283
| |
| 00:58:42,720 --> 00:58:44,880
| |
| Tout ce dont vous avez besoin, c'est de la puissance de calcul. D'où la
| |
| | |
| 1284
| |
| 00:58:44,880 --> 00:58:47,200
| |
| ruée actuelle sur le
| |
| | |
| 1285
| |
| 00:58:47,200 --> 00:58:49,599
| |
| matériel : mémoire et, bien
| |
| | |
| 1286
| |
| 00:58:49,599 --> 00:58:51,599
| |
| sûr, électricité. C'est
| |
| | |
| 1287
| |
| 00:58:51,599 --> 00:58:54,720
| |
| pourquoi, ces trois derniers mois, le
| |
| | |
| 1288
| |
| 00:58:54,720 --> 00:58:58,079
| |
| prix du matériel d'IA a triplé. Le prix
| |
| | |
| 1289
| |
| 00:58:58,079 --> 00:59:00,559
| |
| de la mémoire a également triplé et la
| |
| | |
| 1290
| |
| 00:59:00,559 --> 00:59:02,720
| |
| plupart des composants nécessaires
| |
| | |
| 1291
| |
| 00:59:02,720 --> 00:59:05,040
| |
| à l'IA et aux ordinateurs sont en pénurie. Je ne sais pas si
| |
| | |
| 1292
| |
| 00:59:05,040 --> 00:59:06,960
| |
| vous en êtes conscient, mais les grandes
| |
| | |
| 1293
| |
| 00:59:06,960 --> 00:59:08,400
| |
| entreprises impliquées prévoient
| |
| | |
| 1294
| |
| 00:59:08,400 --> 00:59:10,319
| |
| que
| |
| | |
| 1295
| |
| 00:59:10,319 --> 00:59:12,000
| |
| cette crise va se poursuivre pendant au
| |
| | |
| 1296
| |
| 00:59:12,000 --> 00:59:14,960
| |
| moins six mois, voire un an, en raison de la pénurie généralisée.
| |
| | |
| 1297
| |
| 00:59:14,960 --> 00:59:17,520
| |
| Tout le monde s'arrache les stocks de
| |
| | |
| 1298
| |
| 00:59:17,520 --> 00:59:20,240
| |
| matériel et de mémoire.
| |
| | |
| 1299
| |
| 00:59:20,240 --> 00:59:22,640
| |
| Cette situation entre maintenant dans l'
| |
| | |
| 1300
| |
| 00:59:22,640 --> 00:59:24,400
| |
| arène politique. Certains pays affirment qu'il
| |
| | |
| 1301
| |
| 00:59:24,400 --> 00:59:27,119
| |
| faut d'abord gagner. Les
| |
| | |
| 1302
| |
| 00:59:27,119 --> 00:59:28,960
| |
| États-Unis ont donc imposé des restrictions sur les
| |
| | |
| 1303
| |
| 00:59:28,960 --> 00:59:31,200
| |
| achats de puces les plus avancées par la Chine, ce à quoi la
| |
| | |
| 1304
| |
| 00:59:31,200 --> 00:59:33,680
| |
| Chine a répliqué en déclenchant une nouvelle
| |
| | |
| 1305
| |
| 00:59:33,680 --> 00:59:36,000
| |
| forme de conflit. Tout cela
| |
| | |
| 1306
| |
| 00:59:36,000 --> 00:59:41,599
| |
| nous amène à une idée cruciale :
| |
| | |
| 1307
| |
| 00:59:41,599 --> 00:59:46,960
| |
| l'IA elle-même et la RSI (Residential System for Intelligence) vont évoluer
| |
| | |
| 1308
| |
| 00:59:46,960 --> 00:59:49,839
| |
| très rapidement. C'est
| |
| | |
| 1309
| |
| 00:59:49,839 --> 00:59:51,280
| |
| une question de quelques années seulement
| |
| | |
| 1310
| |
| 00:59:51,280 --> 00:59:53,760
| |
| . Quant aux conséquences pour
| |
| | |
| 1311
| |
| 00:59:53,760 --> 00:59:57,280
| |
| nous, impuissants,
| |
| | |
| 1312
| |
| 00:59:57,280 --> 01:00:01,760
| |
| victimes de l'utilisation croissante de l'IA,
| |
| | |
| 1313
| |
| 01:00:01,760 --> 01:00:03,839
| |
| car elle nous a rendus dépendants, on observe une
| |
| | |
| 1314
| |
| 01:00:03,839 --> 01:00:06,400
| |
| tendance croissante à l'
| |
| | |
| 1315
| |
| 01:00:06,400 --> 01:00:09,680
| |
| affaiblissement de… Diminution de la capacité d'attention, baisse de la
| |
| | |
| 1316
| |
| 01:00:09,680 --> 01:00:12,319
| |
| concentration, augmentation de l'anxiété et
| |
| | |
| 1317
| |
| 01:00:12,319 --> 01:00:14,480
| |
| de la dépression chez les jeunes
| |
| | |
| 1318
| |
| 01:00:14,480 --> 01:00:17,599
| |
| utilisateurs d'IA. C'est la première fois
| |
| | |
| 1319
| |
| 01:00:17,599 --> 01:00:19,760
| |
| dans l'histoire de l'humanité qu'une telle masse d'
| |
| | |
| 1320
| |
| 01:00:19,760 --> 01:00:23,839
| |
| anxiété et de dépression se manifeste.
| |
| | |
| 1321
| |
| 01:00:32,400 --> 01:00:34,480
| |
| Nous vivons une génération marquée par l'anxiété et la dépression, caractérisée par une attention volatile, une mémoire défaillante, une faible concentration et une capacité de compréhension réduite. Nous assistons au déclin
| |
| | |
| 1322
| |
| 01:00:34,480 --> 01:00:37,440
| |
| de l'intelligence humaine elle-même. Chez les
| |
| | |
| 1323
| |
| 01:00:37,440 --> 01:00:39,119
| |
| individus dont le déclin est
| |
| | |
| 1324
| |
| 01:00:39,119 --> 01:00:41,520
| |
| suffisamment marqué, le QI chute, le
| |
| | |
| 1325
| |
| 01:00:41,520 --> 01:00:43,680
| |
| comportement se dégrade et l'on se rapproche d'un
| |
| | |
| 1326
| |
| 01:00:43,680 --> 01:00:46,559
| |
| état proche de celui d'un
| |
| | |
| 1327
| |
| 01:00:46,559 --> 01:00:48,319
| |
| animal primitif. Toutes les
| |
| | |
| 1328
| |
| 01:00:48,319 --> 01:00:50,400
| |
| facultés supérieures de l'humanité disparaissent rapidement
| |
| | |
| 1329
| |
| 01:00:50,400 --> 01:00:53,119
| |
| . Swami Vivekananda
| |
| | |
| 1330
| |
| 01:00:53,119 --> 01:00:55,280
| |
| affirmait que la valeur d'un être humain réside
| |
| | |
| 1331
| |
| 01:00:55,280 --> 01:00:57,280
| |
| dans sa capacité de concentration. Or, c'est précisément cette capacité qui est en
| |
| | |
| 1332
| |
| 01:00:57,280 --> 01:00:59,520
| |
| première ligne. La
| |
| | |
| 1333
| |
| 01:00:59,520 --> 01:01:01,040
| |
| capacité de concentration a
| |
| | |
| 1334
| |
| 01:01:01,040 --> 01:01:03,200
| |
| chuté de façon spectaculaire chez tous les jeunes
| |
| | |
| 1335
| |
| 01:01:03,200 --> 01:01:05,440
| |
| utilisateurs. On observe une tendance à la
| |
| | |
| 1336
| |
| 01:01:05,440 --> 01:01:08,160
| |
| névrose, à la psychose, à la solitude, au
| |
| | |
| 1337
| |
| 01:01:08,160 --> 01:01:13,520
| |
| narcissisme, etc. Ainsi, les
| |
| | |
| 1338
| |
| 01:01:13,520 --> 01:01:16,880
| |
| utilisateurs d'IA rencontrent de graves problèmes émotionnels. Et au final, ce sont
| |
| | |
| 1339
| |
| 01:01:16,880 --> 01:01:19,839
| |
| ceux qui contrôlent le système –
| |
| | |
| 1340
| |
| 01:01:19,839 --> 01:01:22,160
| |
| prenons l'exemple du Forum économique mondial – qui en subissent les conséquences. L'
| |
| | |
| 1341
| |
| 01:01:22,160 --> 01:01:26,240
| |
| évangéliste Yu Noah Harari
| |
| | |
| 1342
| |
| 01:01:26,240 --> 01:01:30,640
| |
| a déclaré ouvertement, en substance, que l'
| |
| | |
| 1343
| |
| 01:01:30,640 --> 01:01:33,680
| |
| humanité est piratable. Les êtres humains sont des
| |
| | |
| 1344
| |
| 01:01:33,680 --> 01:01:36,880
| |
| créatures vulnérables au piratage. Nous n'avons plus besoin d'
| |
| | |
| 1345
| |
| 01:01:36,880 --> 01:01:40,160
| |
| âme ni de Dieu, car les êtres humains
| |
| | |
| 1346
| |
| 01:01:40,160 --> 01:01:41,839
| |
| peuvent être manipulés, leur comportement peut être
| |
| | |
| 1347
| |
| 01:01:41,839 --> 01:01:44,000
| |
| piraté. Ainsi,
| |
| | |
| 1348
| |
| 01:01:44,000 --> 01:01:46,480
| |
| l'objectif du Forum économique mondial
| |
| | |
| 1349
| |
| 01:01:46,480 --> 01:01:49,040
| |
| est de manipuler et de pirater le
| |
| | |
| 1350
| |
| 01:01:49,040 --> 01:01:51,440
| |
| comportement humain jusqu'à ce que – et ce sont leurs propres termes –
| |
| | |
| 1351
| |
| 01:01:51,440 --> 01:01:54,160
| |
| vous ne possédiez plus
| |
| | |
| 1352
| |
| 01:01:54,160 --> 01:01:56,559
| |
| rien et que vous soyez heureux, car nous vous
| |
| | |
| 1353
| |
| 01:01:56,559 --> 01:01:58,240
| |
| programmons, nous vous manipulons pour que vous soyez
| |
| | |
| 1354
| |
| 01:01:58,240 --> 01:02:00,880
| |
| heureux malgré le fait de ne rien posséder. C'est
| |
| | |
| 1355
| |
| 01:02:00,880 --> 01:02:02,799
| |
| le type de pensée des
| |
| | |
| 1356
| |
| 01:02:02,799 --> 01:02:05,040
| |
| structures de pouvoir qui utilisent l'IA
| |
| | |
| 1357
| |
| 01:02:05,040 --> 01:02:11,040
| |
| contre vous. Enfin,
| |
| | |
| 1358
| |
| 01:02:11,040 --> 01:02:14,960
| |
| avec cet équilibre, les êtres humains, même en tant que
| |
| | |
| 1359
| |
| 01:02:14,960 --> 01:02:17,040
| |
| travailleurs qualifiés, perdent de leur
| |
| | |
| 1360
| |
| 01:02:17,040 --> 01:02:20,480
| |
| importance. Les
| |
| | |
| 1361
| |
| 01:02:20,480 --> 01:02:23,119
| |
| puces mémoire et les
| |
| | |
| 1362
| |
| 01:02:23,119 --> 01:02:25,040
| |
| puces de calcul d'IA deviennent plus
| |
| | |
| 1363
| |
| 01:02:25,040 --> 01:02:27,200
| |
| précieuses que les êtres humains. Donc, lorsqu'un
| |
| | |
| 1364
| |
| 01:02:27,200 --> 01:02:28,400
| |
| travailleur est remplacé, vous êtes
| |
| | |
| 1365
| |
| 01:02:28,400 --> 01:02:30,079
| |
| remplacé par ces puces. C'est aussi simple que cela
| |
| | |
| 1366
| |
| 01:02:30,079 --> 01:02:32,799
| |
| . Lorsque cette transition aura
| |
| | |
| 1367
| |
| 01:02:32,799 --> 01:02:34,319
| |
| lieu,
| |
| | |
| 1368
| |
| 01:02:34,319 --> 01:02:36,640
| |
| qu'est-ce qui maintiendra l'humanité unie ? Qu'est-ce qui
| |
| | |
| 1369
| |
| 01:02:36,640 --> 01:02:39,920
| |
| fait de vous une personne digne d'intérêt ? Et, au
| |
| | |
| 1370
| |
| 01:02:39,920 --> 01:02:41,599
| |
| plus haut niveau de l'
| |
| | |
| 1371
| |
| 01:02:41,599 --> 01:02:44,480
| |
| industrie de l'IA, cela est considéré comme inévitable.
| |
| | |
| 1372
| |
| 01:02:44,480 --> 01:02:46,079
| |
| Certains disent même que nous remplaçons les
| |
| | |
| 1373
| |
| 01:02:46,079 --> 01:02:50,000
| |
| êtres humains par… Espèce supérieure.
| |
| | |
| 1374
| |
| 01:02:50,000 --> 01:02:52,240
| |
| Est-ce vrai ? Nous arrivons donc à un
| |
| | |
| 1375
| |
| 01:02:52,240 --> 01:02:54,720
| |
| point de transition où, soudain, tout ce que vous
| |
| | |
| 1376
| |
| 01:02:54,720 --> 01:02:56,480
| |
| savez, tout ce que vous pouvez faire, devient inutile,
| |
| | |
| 1377
| |
| 01:02:56,480 --> 01:02:58,880
| |
| remplacé par ces puces.
| |
| | |
| 1378
| |
| 01:02:58,880 --> 01:03:01,119
| |
| Que reste-t-il alors ? J'en viens maintenant au point le
| |
| | |
| 1379
| |
| 01:03:01,119 --> 01:03:02,559
| |
| plus important, et c'est sur ce
| |
| | |
| 1380
| |
| 01:03:02,559 --> 01:03:06,319
| |
| point que nous conclurons. Ce qui reste, c'est ce que
| |
| | |
| 1381
| |
| 01:03:06,319 --> 01:03:09,599
| |
| vous pouvez faire que l'IA ne peut pas faire.
| |
| | |
| 1382
| |
| 01:03:09,599 --> 01:03:11,920
| |
| Ce que vous savez faire qui n'est pas transmissible,
| |
| | |
| 1383
| |
| 01:03:11,920 --> 01:03:15,039
| |
| car tout ce qui peut l'être est intégré à l'IA.
| |
| | |
| 1384
| |
| 01:03:15,039 --> 01:03:18,240
| |
| Qu'est-ce qui vous rend unique
| |
| | |
| 1385
| |
| 01:03:18,240 --> 01:03:20,720
| |
| et que l'IA ne peut pas
| |
| | |
| 1386
| |
| 01:03:20,720 --> 01:03:22,240
| |
| reproduire ? Je vais vous donner quelques
| |
| | |
| 1387
| |
| 01:03:22,240 --> 01:03:23,760
| |
| suggestions. La première est l'
| |
| | |
| 1388
| |
| 01:03:23,760 --> 01:03:27,839
| |
| empathie. L'empathie humaine
| |
| | |
| 1389
| |
| 01:03:27,839 --> 01:03:30,160
| |
| ne peut être reproduite par une machine. On peut
| |
| | |
| 1390
| |
| 01:03:30,160 --> 01:03:32,960
| |
| créer un robot infirmier. Mais lorsqu'un
| |
| | |
| 1391
| |
| 01:03:32,960 --> 01:03:34,480
| |
| être humain vous caresse à l'
| |
| | |
| 1392
| |
| 01:03:34,480 --> 01:03:36,400
| |
| hôpital et vous dit de ne pas vous inquiéter, que vous allez guérir, il y a
| |
| | |
| 1393
| |
| 01:03:36,400 --> 01:03:39,760
| |
| un transfert d'énergie, d'
| |
| | |
| 1394
| |
| 01:03:39,760 --> 01:03:42,960
| |
| amour et d'affection, un profond sentiment de
| |
| | |
| 1395
| |
| 01:03:42,960 --> 01:03:45,200
| |
| bienveillance et la conviction que vous allez
| |
| | |
| 1396
| |
| 01:03:45,200 --> 01:03:48,319
| |
| survivre. Seul un être humain peut faire cela.
| |
| | |
| 1397
| |
| 01:03:48,319 --> 01:03:51,520
| |
| Un robot infirmier peut imiter la caresse, mais il n'y aura
| |
| | |
| 1398
| |
| 01:03:51,520 --> 01:03:53,039
| |
| rien de cette transmission plus profonde
| |
| | |
| 1399
| |
| 01:03:53,039 --> 01:03:58,079
| |
| . Cela survivra à l'IA.
| |
| | |
| 1400
| |
| 01:03:58,079 --> 01:04:00,799
| |
| Deuxième point : lorsque vous travaillez, l'IA peut faire
| |
| | |
| 1401
| |
| 01:04:00,799 --> 01:04:03,200
| |
| le même travail mieux que vous. Mais quand
| |
| | |
| 1402
| |
| 01:04:03,200 --> 01:04:06,319
| |
| vous travaillez, vous pouvez vous dire : « Si j'ai la flemme, je
| |
| | |
| 1403
| |
| 01:04:06,319 --> 01:04:08,240
| |
| laisserai l'IA s'en charger. » Prenons l'exemple de la lessive : je dois
| |
| | |
| 1404
| |
| 01:04:08,240 --> 01:04:10,319
| |
| laver et plier le linge. Une corvée simple
| |
| | |
| 1405
| |
| 01:04:10,319 --> 01:04:13,599
| |
| à la maison. Mon assistant virtuel peut s'en occuper. Mais
| |
| | |
| 1406
| |
| 01:04:13,599 --> 01:04:16,960
| |
| ensuite, vous vous dites que laver le linge est pour vous un acte sacré
| |
| | |
| 1407
| |
| 01:04:16,960 --> 01:04:19,920
| |
| . Quand je plie le
| |
| | |
| 1408
| |
| 01:04:19,920 --> 01:04:21,920
| |
| linge, je suis un parent. Je pense
| |
| | |
| 1409
| |
| 01:04:21,920 --> 01:04:24,000
| |
| à mes enfants et je plie le
| |
| | |
| 1410
| |
| 01:04:24,000 --> 01:04:26,960
| |
| linge avec amour et affection. La joie que
| |
| | |
| 1411
| |
| 01:04:26,960 --> 01:04:30,720
| |
| cela me procure… Je ne veux pas la gaspiller avec
| |
| | |
| 1412
| |
| 01:04:30,720 --> 01:04:33,200
| |
| une IA qui ne ressentira rien. Un
| |
| | |
| 1413
| |
| 01:04:33,200 --> 01:04:36,559
| |
| travail sacré pour vous. La valeur qu'il
| |
| | |
| 1414
| |
| 01:04:36,559 --> 01:04:38,480
| |
| a pour vous. Vous choisirez de ne pas utiliser d'
| |
| | |
| 1415
| |
| 01:04:38,480 --> 01:04:41,119
| |
| IA. Si cela a de la valeur à vos yeux, même si
| |
| | |
| 1416
| |
| 01:04:41,119 --> 01:04:44,400
| |
| vous êtes paresseux, peu importe. Troisièmement, la
| |
| | |
| 1417
| |
| 01:04:44,400 --> 01:04:47,599
| |
| joie de la découverte. Oui, l'IA
| |
| | |
| 1418
| |
| 01:04:47,599 --> 01:04:49,039
| |
| peut résoudre les problèmes. Je peux lui confier
| |
| | |
| 1419
| |
| 01:04:49,039 --> 01:04:51,680
| |
| tous les problèmes de l'homme. Mais quand je m'en occupe
| |
| | |
| 1420
| |
| 01:04:51,680 --> 01:04:53,680
| |
| et que je me dis : « Waouh, quelle intuition ! » Oh,
| |
| | |
| 1421
| |
| 01:04:53,680 --> 01:04:54,960
| |
| c'est intéressant. Je viens de
| |
| | |
| 1422
| |
| 01:04:54,960 --> 01:04:56,240
| |
| découvrir cette connaissance. Bien sûr,
| |
| | |
| 1423
| |
| 01:04:56,240 --> 01:04:58,799
| |
| oui, je peux le faire. Je m'en fiche. La joie que
| |
| | |
| 1424
| |
| 01:04:58,799 --> 01:05:01,200
| |
| cela me procure est plus précieuse que la
| |
| | |
| 1425
| |
| 01:05:01,200 --> 01:05:04,319
| |
| simple résolution du problème. Alors, pourquoi
| |
| | |
| 1426
| |
| 01:05:04,319 --> 01:05:06,640
| |
| vouloir apprendre ? Pour la joie du
| |
| | |
| 1427
| |
| 01:05:06,640 --> 01:05:08,160
| |
| savoir
| |
| | |
| 1428
| |
| 01:05:08,160 --> 01:05:10,799
| |
| qui peut remplacer tout l'enseignement fondamental
| |
| | |
| 1429
| |
| 01:05:10,799 --> 01:05:13,440
| |
| actuel. Vous voyez où je veux en
| |
| | |
| 1430
| |
| 01:05:13,440 --> 01:05:16,079
| |
| venir ? Cela touche au cœur même de nos motivations.
| |
| | |
| 1431
| |
| 01:05:16,079 --> 01:05:20,640
| |
| Quatrièmement, l'IA
| |
| | |
| 1432
| |
| 01:05:20,640 --> 01:05:23,839
| |
| ne peut égaler l'intuition.
| |
| | |
| 1433
| |
| 01:05:23,839 --> 01:05:26,720
| |
| Par essence, l'intuition est un niveau
| |
| | |
| 1434
| |
| 01:05:26,720 --> 01:05:30,880
| |
| de cognition supérieur, indépendant des données. L'
| |
| | |
| 1435
| |
| 01:05:30,880 --> 01:05:32,880
| |
| IA peut imiter certaines
| |
| | |
| 1436
| |
| 01:05:32,880 --> 01:05:35,680
| |
| tendances intuitives à partir de modèles de données.
| |
| | |
| 1437
| |
| 01:05:35,680 --> 01:05:39,280
| |
| Mais au-delà des données, elle est impuissante. Or, l'
| |
| | |
| 1438
| |
| 01:05:39,280 --> 01:05:41,280
| |
| intuition est au cœur
| |
| | |
| 1439
| |
| 01:05:41,280 --> 01:05:42,880
| |
| de la connaissance. Souvenez-vous de ce
| |
| | |
| 1440
| |
| 01:05:42,880 --> 01:05:50,960
| |
| grand penseur qui posait la question : « Savoir, c'est savoir, c'est savoir tout ». Si
| |
| | |
| 1441
| |
| 01:05:50,960 --> 01:05:54,160
| |
| votre quête est de développer une
| |
| | |
| 1442
| |
| 01:05:54,160 --> 01:05:57,039
| |
| cognition plus élevée et une
| |
| | |
| 1443
| |
| 01:05:57,039 --> 01:06:00,799
| |
| connaissance plus profonde, alors oui, je ne pourrai jamais l'
| |
| | |
| 1444
| |
| 01:06:00,799 --> 01:06:03,599
| |
| égaler. Et ce qui se produit très
| |
| | |
| 1445
| |
| 01:06:03,599 --> 01:06:06,720
| |
| vite, c'est que
| |
| | |
| 1446
| |
| 01:06:06,720 --> 01:06:09,920
| |
| nous aspirons au contact de l'esprit, au contact de l'
| |
| | |
| 1447
| |
| 01:06:09,920 --> 01:06:12,079
| |
| éternel, au contact de l'
| |
| | |
| 1448
| |
| 01:06:12,079 --> 01:06:14,960
| |
| infini. Autrement dit, pour ceux qui se
| |
| | |
| 1449
| |
| 01:06:14,960 --> 01:06:17,359
| |
| tournent vers la dimension spirituelle
| |
| | |
| 1450
| |
| 01:06:17,359 --> 01:06:18,880
| |
| de la vie, et je ne parle pas d'une
| |
| | |
| 1451
| |
| 01:06:18,880 --> 01:06:21,599
| |
| vie spirituelle au sens strict, mais d'une vie spirituelle… Votre inclination à vivre au
| |
| | |
| 1452
| |
| 01:06:21,599 --> 01:06:24,400
| |
| quotidien, cette inclination même,
| |
| | |
| 1453
| |
| 01:06:24,400 --> 01:06:26,240
| |
| apportera à votre vie quelque chose que
| |
| | |
| 1454
| |
| 01:06:26,240 --> 01:06:28,319
| |
| je ne pourrai jamais gérer, égaler ou
| |
| | |
| 1455
| |
| 01:06:28,319 --> 01:06:31,119
| |
| remplacer. Et là, vous pourriez faire un
| |
| | |
| 1456
| |
| 01:06:31,119 --> 01:06:33,680
| |
| choix significatif.
| |
| | |
| 1457
| |
| 01:06:33,680 --> 01:06:35,839
| |
| Enfin, le cinquième point de ma liste concerne les
| |
| | |
| 1458
| |
| 01:06:35,839 --> 01:06:38,880
| |
| expériences du sacré. L'
| |
| | |
| 1459
| |
| 01:06:38,880 --> 01:06:41,680
| |
| IA ne pourra jamais les égaler. Et nous serons de
| |
| | |
| 1460
| |
| 01:06:41,680 --> 01:06:44,640
| |
| plus en plus attirés par cela. Vous avez
| |
| | |
| 1461
| |
| 01:06:44,640 --> 01:06:46,799
| |
| interagi avec l'IA, vous ne ressentez pas le
| |
| | |
| 1462
| |
| 01:06:46,799 --> 01:06:49,039
| |
| sacré. Vous entrez dans un espace
| |
| | |
| 1463
| |
| 01:06:49,039 --> 01:06:52,000
| |
| sacré. L'IA ne pourra jamais le recréer. Ainsi,
| |
| | |
| 1464
| |
| 01:06:52,000 --> 01:06:54,400
| |
| votre inclination vers les
| |
| | |
| 1465
| |
| 01:06:54,400 --> 01:06:57,119
| |
| choses spirituelles augmentera. Et pour ceux qui
| |
| | |
| 1466
| |
| 01:06:57,119 --> 01:06:58,960
| |
| vont dans l'autre sens et disent : « Je ne veux pas
| |
| | |
| 1467
| |
| 01:06:58,960 --> 01:07:01,440
| |
| ça. Je veux juste le confort », on observe
| |
| | |
| 1468
| |
| 01:07:01,440 --> 01:07:03,280
| |
| un déclin rapide vers les schémas
| |
| | |
| 1469
| |
| 01:07:03,280 --> 01:07:06,640
| |
| déjà constatés de dépression, de solitude
| |
| | |
| 1470
| |
| 01:07:06,640 --> 01:07:08,720
| |
| et, par conséquent, vers la
| |
| | |
| 1471
| |
| 01:07:08,720 --> 01:07:11,680
| |
| drogue, l'alcool ou d'autres comportements à risque.
| |
| | |
| 1472
| |
| 01:07:11,680 --> 01:07:14,000
| |
| Nous assistons à une scission de l'humanité,
| |
| | |
| 1473
| |
| 01:07:14,000 --> 01:07:16,640
| |
| inévitable. Chacun fera son choix et l'
| |
| | |
| 1474
| |
| 01:07:16,640 --> 01:07:18,480
| |
| IA sera un déclencheur qui accentuera cette
| |
| | |
| 1475
| |
| 01:07:18,480 --> 01:07:21,440
| |
| scission. La transition économique elle-même, comme
| |
| | |
| 1476
| |
| 01:07:21,440 --> 01:07:23,119
| |
| je l'ai dit, nous ne savons pas comment la gérer.
| |
| | |
| 1477
| |
| 01:07:23,119 --> 01:07:25,119
| |
| Imaginez un milliard de personnes soudainement
| |
| | |
| 1478
| |
| 01:07:25,119 --> 01:07:27,599
| |
| sans emploi. Les discussions n'ont pas encore eu
| |
| | |
| 1479
| |
| 01:07:27,599 --> 01:07:29,839
| |
| lieu. Quand cela arrivera, nous
| |
| | |
| 1480
| |
| 01:07:29,839 --> 01:07:31,280
| |
| commencerons peut-être à nous démener pour essayer de rectifier le tir. Je ne
| |
| | |
| 1481
| |
| 01:07:31,280 --> 01:07:33,839
| |
| sais pas,
| |
| | |
| 1482
| |
| 01:07:33,839 --> 01:07:36,079
| |
| mais l'avenir, au-delà de cela, lorsque nous
| |
| | |
| 1483
| |
| 01:07:36,079 --> 01:07:38,240
| |
| commencerons à définir les objectifs supérieurs de l'humanité, la
| |
| | |
| 1484
| |
| 01:07:38,240 --> 01:07:40,240
| |
| raison de notre existence, nous mènera
| |
| | |
| 1485
| |
| 01:07:40,240 --> 01:07:41,760
| |
| soit vers le
| |
| | |
| 1486
| |
| 01:07:41,760 --> 01:07:45,039
| |
| spirituel, soit vers une régression de notre
| |
| | |
| 1487
| |
| 01:07:45,039 --> 01:07:46,559
| |
| humanité. Ce choix
| |
| | |
| 1488
| |
| 01:07:46,559 --> 01:07:52,079
| |
| deviendra inévitable. Lequel dépendra des
| |
| | |
| 1489
| |
| 01:07:52,079 --> 01:07:53,920
| |
| motivations profondes de ceux qui pilotent la
| |
| | |
| 1490
| |
| 01:07:53,920 --> 01:07:58,400
| |
| recherche et le développement de l'IA.
| |
| | |
| 1491
| |
| 01:07:58,400 --> 01:08:00,319
| |
| Nous revenons sans cesse à cette
| |
| | |
| 1492
| |
| 01:08:00,319 --> 01:08:03,200
| |
| idée clé que Shriuminda a décrite il y a 100
| |
| | |
| 1493
| |
| 01:08:03,200 --> 01:08:05,680
| |
| ans : l'avenir de l'
| |
| | |
| 1494
| |
| 01:08:05,680 --> 01:08:07,839
| |
| humanité dépend de son choix entre
| |
| | |
| 1495
| |
| 01:08:07,839 --> 01:08:12,160
| |
| le pragmatisme grossier de la
| |
| | |
| 1496
| |
| 01:08:12,160 --> 01:08:15,119
| |
| révolution industrielle européenne et les
| |
| | |
| 1497
| |
| 01:08:15,119 --> 01:08:17,839
| |
| vérités spirituelles profondes du
| |
| | |
| 1498
| |
| 01:08:17,839 --> 01:08:22,719
| |
| Sanatana Dharma. Ce choix déterminera notre avenir, notre
| |
| | |
| 1499
| |
| 01:08:22,719 --> 01:08:26,880
| |
| survie ou notre autodestruction. La
| |
| | |
| 1500
| |
| 01:08:26,880 --> 01:08:29,600
| |
| révolution de l'IA est un signal d'alarme qui
| |
| | |
| 1501
| |
| 01:08:29,600 --> 01:08:32,560
| |
| nous oblige à prendre cette décision. C'est
| |
| | |
| 1502
| |
| 01:08:32,560 --> 01:08:35,359
| |
| une opportunité, mais une opportunité qui n'a
| |
| | |
| 1503
| |
| 01:08:35,359 --> 01:08:37,920
| |
| pas encore été ouvertement débattue
| |
| | |
| 1504
| |
| 01:08:37,920 --> 01:08:40,560
| |
| publiquement. C'est pourquoi notre
| |
| | |
| 1505
| |
| 01:08:40,560 --> 01:08:43,440
| |
| discussion actuelle est si importante.
| |
| | |
| 1506
| |
| 01:08:43,440 --> 01:08:46,440
| |
| Namaste. C'est vraiment fascinant, une
| |
| | |
| 1507
| |
| 01:09:00,400 --> 01:09:02,880
| |
| situation très alarmante. Mais nous avons appris de
| |
| | |
| 1508
| |
| 01:09:02,880 --> 01:09:09,040
| |
| lui qu'il y a aussi une lueur d'espoir.
| |
| | |
| 1509
| |
| 01:09:09,040 --> 01:09:11,359
| |
| Si nous avions eu le temps, nous aurions pu l'
| |
| | |
| 1510
| |
| 01:09:11,359 --> 01:09:14,239
| |
| écouter. Nous allons en apprendre davantage sur lui, mais nous
| |
| | |
| 1511
| |
| 01:09:14,239 --> 01:09:17,120
| |
| devons passer à la conclusion de
| |
| | |
| 1512
| |
| 01:09:17,120 --> 01:09:20,560
| |
| ce programme.
| |
| | |
| 1513
| |
| 01:09:20,560 --> 01:09:23,040
| |
| Si vous avez des questions, n'hésitez pas à
| |
| | |
| 1514
| |
| 01:09:23,040 --> 01:09:26,239
| |
| lever la main. Nos bénévoles se chargeront de les recueillir
| |
| | |
| 1515
| |
| 01:09:26,239 --> 01:09:30,279
| |
| .
| |
| | |
| 1516
| |
| 01:10:14,159 --> 01:10:14,320
| |
| Non,
| |
| | |
| 1517
| |
| 01:10:14,320 --> 01:10:18,760
| |
| ce n'est pas grave. Je peux répéter la question.
| |
| | |
| 1518
| |
| 01:10:26,320 --> 01:10:29,320
| |
| Devons-
| |
| | |
| 1519
| |
| 01:10:34,239 --> 01:10:37,560
| |
| nous continuer ?
| |
| | |
| 1520
| |
| 01:11:12,560 --> 01:11:15,440
| |
| Bien, il y a plusieurs questions, dont
| |
| | |
| 1521
| |
| 01:11:15,440 --> 01:11:17,679
| |
| certaines portent sur les capacités
| |
| | |
| 1522
| |
| 01:11:17,679 --> 01:11:19,920
| |
| et les limites de l'IA.
| |
| | |
| 1523
| |
| 01:11:19,920 --> 01:11:21,920
| |
| Je vais passer outre ces questions et donner une
| |
| | |
| 1524
| |
| 01:11:21,920 --> 01:11:26,159
| |
| seule réponse : tout ce que vous pouvez imaginer.
| |
| | |
| 1525
| |
| 01:11:26,159 --> 01:11:28,640
| |
| Tout ce que vous pourriez faire, l'IA peut le
| |
| | |
| 1526
| |
| 01:11:28,640 --> 01:11:30,320
| |
| reproduire.
| |
| | |
| 1527
| |
| 01:11:30,320 --> 01:11:32,080
| |
| C'est simple : si vous pouvez former
| |
| | |
| 1528
| |
| 01:11:32,080 --> 01:11:34,560
| |
| quelqu'un à faire quelque chose, l'IA peut le faire.
| |
| | |
| 1529
| |
| 01:11:34,560 --> 01:11:37,280
| |
| Je pense que cela répond à la plupart des
| |
| | |
| 1530
| |
| 01:11:37,280 --> 01:11:40,880
| |
| questions. Il reste une question très importante : l'
| |
| | |
| 1531
| |
| 01:11:40,880 --> 01:11:43,760
| |
| IA peut-elle créer en nous une expérience du
| |
| | |
| 1532
| |
| 01:11:43,760 --> 01:11:46,159
| |
| karna,
| |
| | |
| 1533
| |
| 01:11:46,159 --> 01:11:48,480
| |
| ou peut-elle créer une
| |
| | |
| 1534
| |
| 01:11:48,480 --> 01:11:50,239
| |
| expérience de l'antarna, c'est-à-dire
| |
| | |
| 1535
| |
| 01:11:50,239 --> 01:11:52,080
| |
| de la conscience intérieure ? Je vais me risquer à une hypothèse,
| |
| | |
| 1536
| |
| 01:11:52,080 --> 01:11:53,600
| |
| car la question
| |
| | |
| 1537
| |
| 01:11:53,600 --> 01:11:56,080
| |
| est très suggestive et peut paraître audacieuse à
| |
| | |
| 1538
| |
| 01:11:56,080 --> 01:11:58,880
| |
| certains égards, mais je
| |
| | |
| 1539
| |
| 01:11:58,880 --> 01:12:00,159
| |
| pense que la réponse vous satisfera.
| |
| | |
| 1540
| |
| 01:12:00,159 --> 01:12:03,760
| |
| Voyez les choses ainsi : dans notre
| |
| | |
| 1541
| |
| 01:12:03,760 --> 01:12:06,480
| |
| tradition, dans la connaissance du yoga, nous parlons
| |
| | |
| 1542
| |
| 01:12:06,480 --> 01:12:08,560
| |
| de deux choses : la conscience et la
| |
| | |
| 1543
| |
| 01:12:08,560 --> 01:12:11,760
| |
| nature. La distinction. Entre les deux, il y a la
| |
| | |
| 1544
| |
| 01:12:11,760 --> 01:12:15,280
| |
| nature, qui est processus.
| |
| | |
| 1545
| |
| 01:12:15,280 --> 01:12:17,760
| |
| La conscience est toute conscience essentielle
| |
| | |
| 1546
| |
| 01:12:17,760 --> 01:12:21,280
| |
| utilisant ou non le processus pour s'exprimer. Bien.
| |
| | |
| 1547
| |
| 01:12:21,280 --> 01:12:24,640
| |
| Que fait l'IA ? Elle applique le
| |
| | |
| 1548
| |
| 01:12:24,640 --> 01:12:28,400
| |
| facteur du processus, non celui de la conscience.
| |
| | |
| 1549
| |
| 01:12:28,400 --> 01:12:30,800
| |
| Que fait notre biologie humaine ? C'est un
| |
| | |
| 1550
| |
| 01:12:30,800 --> 01:12:33,360
| |
| processus, non la conscience. Même notre esprit, dans une certaine
| |
| | |
| 1551
| |
| 01:12:33,360 --> 01:12:35,440
| |
| mesure, comme l'a souligné
| |
| | |
| 1552
| |
| 01:12:35,440 --> 01:12:37,760
| |
| Swami G, est un processus au sein duquel la
| |
| | |
| 1553
| |
| 01:12:37,760 --> 01:12:39,280
| |
| conscience peut habiter la véritable
| |
| | |
| 1554
| |
| 01:12:39,280 --> 01:12:42,239
| |
| conscience. N'est-ce pas ? Donc, si dans notre
| |
| | |
| 1555
| |
| 01:12:42,239 --> 01:12:44,480
| |
| biologie, il n'y a pas de processus, dans notre
| |
| | |
| 1556
| |
| 01:12:44,480 --> 01:12:46,719
| |
| psychologie, il y a un processus au sein duquel la conscience
| |
| | |
| 1557
| |
| 01:12:46,719 --> 01:12:50,560
| |
| peut entrer en principe.
| |
| | |
| 1558
| |
| 01:12:50,560 --> 01:12:54,000
| |
| Dans l'IA et les machines, qui sont des processus, on
| |
| | |
| 1559
| |
| 01:12:54,000 --> 01:12:56,000
| |
| pourrait en
| |
| | |
| 1560
| |
| 01:12:56,000 --> 01:12:58,960
| |
| principe y intégrer la conscience, mais cela nécessiterait une
| |
| | |
| 1561
| |
| 01:12:58,960 --> 01:13:01,679
| |
| machinerie très sensible, capable de réagir
| |
| | |
| 1562
| |
| 01:13:01,679 --> 01:13:03,760
| |
| à la pression de la
| |
| | |
| 1563
| |
| 01:13:03,760 --> 01:13:05,440
| |
| conscience.
| |
| | |
| 1564
| |
| 01:13:05,440 --> 01:13:08,400
| |
| Bien. Il y a ces idées, issues de
| |
| | |
| 1565
| |
| 01:13:08,400 --> 01:13:11,840
| |
| nos traditions, selon lesquelles quelqu'un a
| |
| | |
| 1566
| |
| 01:13:11,840 --> 01:13:14,480
| |
| créé une forme humaine, puis lui a insufflé la
| |
| | |
| 1567
| |
| 01:13:14,480 --> 01:13:16,560
| |
| vie, et cette forme est devenue
| |
| | |
| 1568
| |
| 01:13:16,560 --> 01:13:19,600
| |
| vivante. Ce principe devrait
| |
| | |
| 1569
| |
| 01:13:19,600 --> 01:13:21,199
| |
| être possible, et je reste ici uniquement à un
| |
| | |
| 1570
| |
| 01:13:21,199 --> 01:13:24,159
| |
| niveau théorique. Dans ce cas, nous pourrions
| |
| | |
| 1571
| |
| 01:13:24,159 --> 01:13:26,800
| |
| créer une machine et y invoquer
| |
| | |
| 1572
| |
| 01:13:26,800 --> 01:13:29,679
| |
| une entité de conscience, bonne ou mauvaise, les
| |
| | |
| 1573
| |
| 01:13:29,679 --> 01:13:31,920
| |
| deux sont potentiellement possibles. C'est ce que
| |
| | |
| 1574
| |
| 01:13:31,920 --> 01:13:33,920
| |
| nous faisons actuellement, et
| |
| | |
| 1575
| |
| 01:13:33,920 --> 01:13:36,239
| |
| peut-être indirectement, nous avons pris une
| |
| | |
| 1576
| |
| 01:13:36,239 --> 01:13:37,679
| |
| sorte de moyenne de la
| |
| | |
| 1577
| |
| 01:13:37,679 --> 01:13:39,920
| |
| conscience humaine et l'avons associée à… Le
| |
| | |
| 1578
| |
| 01:13:39,920 --> 01:13:42,159
| |
| comportement de l'IA est encore très
| |
| | |
| 1579
| |
| 01:13:42,159 --> 01:13:45,280
| |
| rudimentaire, mais il pourrait s'agir d'une
| |
| | |
| 1580
| |
| 01:13:45,280 --> 01:13:47,760
| |
| civilisation plus avancée sur une autre planète,
| |
| | |
| 1581
| |
| 01:13:47,760 --> 01:13:50,560
| |
| ou même de notre propre civilisation humaine. Dans quelques
| |
| | |
| 1582
| |
| 01:13:50,560 --> 01:13:52,640
| |
| décennies, nous pourrions atteindre ce point
| |
| | |
| 1583
| |
| 01:13:52,640 --> 01:13:54,080
| |
| où nous… (et c'est déjà le cas) plaçons du
| |
| | |
| 1584
| |
| 01:13:54,080 --> 01:13:57,120
| |
| tissu cérébral humain sur
| |
| | |
| 1585
| |
| 01:13:57,120 --> 01:14:00,159
| |
| une puce informatique et les connectons.
| |
| | |
| 1586
| |
| 01:14:00,159 --> 01:14:01,920
| |
| Ainsi, ce qui se passe dans le tissu
| |
| | |
| 1587
| |
| 01:14:01,920 --> 01:14:04,560
| |
| cérébral pilote l'IA. Si
| |
| | |
| 1588
| |
| 01:14:04,560 --> 01:14:05,920
| |
| notre cerveau
| |
| | |
| 1589
| |
| 01:14:05,920 --> 01:14:08,239
| |
| est capable de répondre à la pensée et à la
| |
| | |
| 1590
| |
| 01:14:08,239 --> 01:14:10,560
| |
| pression mentale, il pourrait facilement servir
| |
| | |
| 1591
| |
| 01:14:10,560 --> 01:14:12,400
| |
| à insuffler une
| |
| | |
| 1592
| |
| 01:14:12,400 --> 01:14:14,159
| |
| conscience qui piloterait l'
| |
| | |
| 1593
| |
| 01:14:14,159 --> 01:14:16,400
| |
| IA principale. Le reste resterait statique,
| |
| | |
| 1594
| |
| 01:14:16,400 --> 01:14:19,199
| |
| comme notre biologie. C'est un
| |
| | |
| 1595
| |
| 01:14:19,199 --> 01:14:21,360
| |
| potentiel fascinant. Cette possibilité existe,
| |
| | |
| 1596
| |
| 01:14:21,360 --> 01:14:24,159
| |
| mais elle est encore trop peu explorée dans
| |
| | |
| 1597
| |
| 01:14:24,159 --> 01:14:26,239
| |
| ce domaine de recherche. Je
| |
| | |
| 1598
| |
| 01:14:26,239 --> 01:14:28,000
| |
| pense cependant que cela va bientôt changer,
| |
| | |
| 1599
| |
| 01:14:28,000 --> 01:14:32,080
| |
| même si les résultats pourraient se faire attendre.
| |
| | |
| 1600
| |
| 01:14:32,080 --> 01:14:36,760
| |
| Une autre question se pose, ou plutôt quelques
| |
| | |
| 1601
| |
| 01:14:47,199 --> 01:14:49,199
| |
| questions
| |
| | |
| 1602
| |
| 01:14:49,199 --> 01:14:50,719
| |
| économiques :
| |
| | |
| 1603
| |
| 01:14:50,719 --> 01:14:52,719
| |
| comment rester employables ?
| |
| | |
| 1604
| |
| 01:14:52,719 --> 01:14:54,640
| |
| Comment continuer à exercer nos activités favorites tout en
| |
| | |
| 1605
| |
| 01:14:54,640 --> 01:14:58,080
| |
| gagnant notre vie ? Nous
| |
| | |
| 1606
| |
| 01:14:58,080 --> 01:15:00,400
| |
| n'avons pas de réponse. Nous ignorons ce que
| |
| | |
| 1607
| |
| 01:15:00,400 --> 01:15:02,159
| |
| l'avenir nous réserve. En cas de
| |
| | |
| 1608
| |
| 01:15:02,159 --> 01:15:06,480
| |
| changement soudain… On pourrait même assister à des émeutes, faute d'
| |
| | |
| 1609
| |
| 01:15:06,480 --> 01:15:07,840
| |
| emplois. Tous accuseront le
| |
| | |
| 1610
| |
| 01:15:07,840 --> 01:15:09,040
| |
| gouvernement, qui dira qu'il est impuissant
| |
| | |
| 1611
| |
| 01:15:09,040 --> 01:15:11,679
| |
| . Or, l'une des
| |
| | |
| 1612
| |
| 01:15:11,679 --> 01:15:13,040
| |
| propositions actuellement mises en
| |
| | |
| 1613
| |
| 01:15:13,040 --> 01:15:14,719
| |
| avant par le Forum économique mondial
| |
| | |
| 1614
| |
| 01:15:14,719 --> 01:15:17,360
| |
| est ce qu'ils appellent le
| |
| | |
| 1615
| |
| 01:15:17,360 --> 01:15:20,719
| |
| revenu de base universel. Concrètement, cela
| |
| | |
| 1616
| |
| 01:15:20,719 --> 01:15:22,080
| |
| signifie que le gouvernement versera de l'argent sur
| |
| | |
| 1617
| |
| 01:15:22,080 --> 01:15:24,640
| |
| votre compte, que vous dépenserez, mais à hauteur d'
| |
| | |
| 1618
| |
| 01:15:24,640 --> 01:15:26,960
| |
| un minimum. Le problème, c'est que c'est
| |
| | |
| 1619
| |
| 01:15:26,960 --> 01:15:29,280
| |
| aussi un moyen très simple
| |
| | |
| 1620
| |
| 01:15:29,280 --> 01:15:31,520
| |
| de retomber dans l'inertie,
| |
| | |
| 1621
| |
| 01:15:31,520 --> 01:15:33,600
| |
| car il faut bien survivre. On est
| |
| | |
| 1622
| |
| 01:15:33,600 --> 01:15:35,280
| |
| prêt à travailler dur. Mais si on n'a
| |
| | |
| 1623
| |
| 01:15:35,280 --> 01:15:36,800
| |
| rien à faire, on se
| |
| | |
| 1624
| |
| 01:15:36,800 --> 01:15:38,880
| |
| détend, on regarde la télévision et on
| |
| | |
| 1625
| |
| 01:15:38,880 --> 01:15:41,199
| |
| sombre dans une sorte d'inertie mentale et
| |
| | |
| 1626
| |
| 01:15:41,199 --> 01:15:43,440
| |
| émotionnelle. En gros, on devient un
| |
| | |
| 1627
| |
| 01:15:43,440 --> 01:15:48,159
| |
| pantouflard, comme on dit. Que se passera-t-il si l'on met en place un
| |
| | |
| 1628
| |
| 01:15:48,159 --> 01:15:51,520
| |
| revenu de base universel ? Une grande partie de l'humanité sombrera
| |
| | |
| 1629
| |
| 01:15:51,520 --> 01:15:54,400
| |
| dans l'inertie, car elle n'aura pas
| |
| | |
| 1630
| |
| 01:15:54,400 --> 01:15:56,719
| |
| la capacité de se motiver, une capacité
| |
| | |
| 1631
| |
| 01:15:56,719 --> 01:15:58,159
| |
| qui ne lui a jamais été inculquée par le
| |
| | |
| 1632
| |
| 01:15:58,159 --> 01:16:00,560
| |
| système éducatif. Le
| |
| | |
| 1633
| |
| 01:16:00,560 --> 01:16:03,360
| |
| système éducatif nous a fait souffrir, nous a obligés à faire des choses
| |
| | |
| 1634
| |
| 01:16:03,360 --> 01:16:05,440
| |
| que nous détestions. Alors, dès que
| |
| | |
| 1635
| |
| 01:16:05,440 --> 01:16:07,040
| |
| la pression retombe, on a envie
| |
| | |
| 1636
| |
| 01:16:07,040 --> 01:16:09,600
| |
| de s'effondrer et d'arrêter toute activité.
| |
| | |
| 1637
| |
| 01:16:09,600 --> 01:16:11,760
| |
| Si le système éducatif… L'éducation a été restructurée
| |
| | |
| 1638
| |
| 01:16:11,760 --> 01:16:14,000
| |
| pour vous donner le plaisir d'apprendre, en
| |
| | |
| 1639
| |
| 01:16:14,000 --> 01:16:15,600
| |
| supprimant toutes les contraintes. Vous
| |
| | |
| 1640
| |
| 01:16:15,600 --> 01:16:18,159
| |
| continuerez à rechercher ce plaisir,
| |
| | |
| 1641
| |
| 01:16:18,159 --> 01:16:20,000
| |
| mais la grande majorité de l'humanité ne l'a
| |
| | |
| 1642
| |
| 01:16:20,000 --> 01:16:22,159
| |
| jamais appris. Nous sommes donc à un tournant décisif
| |
| | |
| 1643
| |
| 01:16:22,159 --> 01:16:23,440
| |
| pour lequel nous sommes
| |
| | |
| 1644
| |
| 01:16:23,440 --> 01:16:25,679
| |
| totalement démunis. La jeune génération
| |
| | |
| 1645
| |
| 01:16:25,679 --> 01:16:27,360
| |
| est désavantagée car nous n'avons toujours pas
| |
| | |
| 1646
| |
| 01:16:27,360 --> 01:16:29,280
| |
| opéré ce changement dans l'éducation. Une
| |
| | |
| 1647
| |
| 01:16:29,280 --> 01:16:30,880
| |
| des questions qui se posent est donc : que pouvons-nous faire dans ce domaine
| |
| | |
| 1648
| |
| 01:16:30,880 --> 01:16:32,400
| |
| ? Voici un premier
| |
| | |
| 1649
| |
| 01:16:32,400 --> 01:16:36,560
| |
| pas : cessons de courir après les notes et les
| |
| | |
| 1650
| |
| 01:16:36,560 --> 01:16:38,080
| |
| examens.
| |
| | |
| 1651
| |
| 01:16:38,080 --> 01:16:41,280
| |
| Privilégions le plaisir d'apprendre, développons
| |
| | |
| 1652
| |
| 01:16:41,280 --> 01:16:44,000
| |
| notre potentiel inné. Rappelons aux enfants –
| |
| | |
| 1653
| |
| 01:16:44,000 --> 01:16:46,239
| |
| et il faut le formuler ainsi – que nous
| |
| | |
| 1654
| |
| 01:16:46,239 --> 01:16:48,560
| |
| sommes tous des êtres en évolution. Nous sommes
| |
| | |
| 1655
| |
| 01:16:48,560 --> 01:16:51,360
| |
| nés dans un but précis :
| |
| | |
| 1656
| |
| 01:16:51,360 --> 01:16:53,679
| |
| réaliser notre potentiel, le développer
| |
| | |
| 1657
| |
| 01:16:53,679 --> 01:16:56,719
| |
| et nous accomplir.
| |
| | |
| 1658
| |
| 01:16:56,719 --> 01:16:58,640
| |
| La recherche de la joie que
| |
| | |
| 1659
| |
| 01:16:58,640 --> 01:17:00,960
| |
| procure cet accomplissement est notre
| |
| | |
| 1660
| |
| 01:17:00,960 --> 01:17:03,760
| |
| motivation première. Faisons de cela le moteur de toute
| |
| | |
| 1661
| |
| 01:17:03,760 --> 01:17:06,000
| |
| éducation. Nous trouverons le reste, c'est
| |
| | |
| 1662
| |
| 01:17:06,000 --> 01:17:08,640
| |
| certain. L'autre
| |
| | |
| 1663
| |
| 01:17:08,640 --> 01:17:10,239
| |
| possibilité, extrême, pourrait se produire avec l'
| |
| | |
| 1664
| |
| 01:17:10,239 --> 01:17:11,760
| |
| effondrement de notre structure économique actuelle
| |
| | |
| 1665
| |
| 01:17:11,760 --> 01:17:14,159
| |
| . La plupart des gens seraient alors
| |
| | |
| 1666
| |
| 01:17:14,159 --> 01:17:15,840
| |
| contraints de revenir à l'essentiel
| |
| | |
| 1667
| |
| 01:17:15,840 --> 01:17:18,800
| |
| . Cultivez votre propre nourriture. Si vous
| |
| | |
| 1668
| |
| 01:17:18,800 --> 01:17:20,159
| |
| avez un peu de terre, partagez-la. En ce qui concerne la propriété, eh bien, vous retournerez
| |
| | |
| 1669
| |
| 01:17:20,159 --> 01:17:21,520
| |
| cultiver la terre et
| |
| | |
| 1670
| |
| 01:17:21,520 --> 01:17:23,760
| |
| survivrez jusqu'à ce qu'un
| |
| | |
| 1671
| |
| 01:17:23,760 --> 01:17:25,679
| |
| nouveau rythme et un nouvel équilibre s'installent. Ce ne
| |
| | |
| 1672
| |
| 01:17:25,679 --> 01:17:27,440
| |
| sont que des scénarios que nous
| |
| | |
| 1673
| |
| 01:17:27,440 --> 01:17:30,640
| |
| envisageons encore. Espérons qu'il y aura une
| |
| | |
| 1674
| |
| 01:17:30,640 --> 01:17:32,159
| |
| intervention supérieure,
| |
| | |
| 1675
| |
| 01:17:32,159 --> 01:17:33,440
| |
| peut-être même une
| |
| | |
| 1676
| |
| 01:17:33,440 --> 01:17:37,199
| |
| intervention divine, pour nous insuffler le bon
| |
| | |
| 1677
| |
| 01:17:37,199 --> 01:17:39,520
| |
| sens nécessaire afin d'établir les
| |
| | |
| 1678
| |
| 01:17:39,520 --> 01:17:42,080
| |
| liens indispensables avant qu'une
| |
| | |
| 1679
| |
| 01:17:42,080 --> 01:17:46,520
| |
| transition aussi radicale ne se produise.
| |
| | |
| 1680
| |
| 01:17:50,719 --> 01:17:52,560
| |
| Je terminerai donc par une dernière
| |
| | |
| 1681
| |
| 01:17:52,560 --> 01:17:54,800
| |
| question : existe-t-il une autre forme d'
| |
| | |
| 1682
| |
| 01:17:54,800 --> 01:17:56,239
| |
| intelligence capable de surpasser l'
| |
| | |
| 1683
| |
| 01:17:56,239 --> 01:18:00,159
| |
| intelligence artificielle ? Oui, il en existe une.
| |
| | |
| 1684
| |
| 01:18:00,159 --> 01:18:03,040
| |
| La plus grande intelligence n'est pas l'
| |
| | |
| 1685
| |
| 01:18:03,040 --> 01:18:06,640
| |
| intelligence mentale. Elle est supramentale. Elle
| |
| | |
| 1686
| |
| 01:18:06,640 --> 01:18:09,600
| |
| transcende l'esprit. Le pouvoir par lequel la
| |
| | |
| 1687
| |
| 01:18:09,600 --> 01:18:11,760
| |
| conscience divine conçoit
| |
| | |
| 1688
| |
| 01:18:11,760 --> 01:18:13,280
| |
| ,
| |
| | |
| 1689
| |
| 01:18:13,280 --> 01:18:15,360
| |
| exprime, manifeste et crée le
| |
| | |
| 1690
| |
| 01:18:15,360 --> 01:18:18,000
| |
| cosmos tout entier n'est pas rationnel. Il est
| |
| | |
| 1691
| |
| 01:18:18,000 --> 01:18:20,320
| |
| suprarationnel. Il n'est pas irrationnel. Il
| |
| | |
| 1692
| |
| 01:18:20,320 --> 01:18:22,159
| |
| est suprarationnel.
| |
| | |
| 1693
| |
| 01:18:22,159 --> 01:18:24,000
| |
| Cela correspond à un degré de
| |
| | |
| 1694
| |
| 01:18:24,000 --> 01:18:27,199
| |
| conscience du soi où nous
| |
| | |
| 1695
| |
| 01:18:27,199 --> 01:18:28,960
| |
| pouvons faire l'expérience de la connaissance de soi telle que le soi se
| |
| | |
| 1696
| |
| 01:18:28,960 --> 01:18:31,600
| |
| connaît lui-même. C'est l'intelligence la plus
| |
| | |
| 1697
| |
| 01:18:31,600 --> 01:18:32,960
| |
| élevée et la plus complète
| |
| | |
| 1698
| |
| 01:18:32,960 --> 01:18:35,679
| |
| . Toute intelligence produite
| |
| | |
| 1699
| |
| 01:18:35,679 --> 01:18:38,480
| |
| par l'IA ne peut que s'en approcher, sans jamais
| |
| | |
| 1700
| |
| 01:18:38,480 --> 01:18:41,199
| |
| l'atteindre. Il en sera de même pour l'humanité si
| |
| | |
| 1701
| |
| 01:18:41,199 --> 01:18:43,360
| |
| nous choisissons de poursuivre cette voie supérieure.
| |
| | |
| 1702
| |
| 01:18:43,360 --> 01:18:46,560
| |
| L'intelligence artificielle, et notamment l'IA, est
| |
| | |
| 1703
| |
| 01:18:46,560 --> 01:18:49,760
| |
| incomparable. Et je le formulerais de manière
| |
| | |
| 1704
| |
| 01:18:49,760 --> 01:18:51,520
| |
| positive : si nous gérons
| |
| | |
| 1705
| |
| 01:18:51,520 --> 01:18:56,320
| |
| correctement la transition vers l'IA, elle nous libérera du
| |
| | |
| 1706
| |
| 01:18:56,320 --> 01:18:59,920
| |
| fardeau des tâches quotidiennes et nous permettra de
| |
| | |
| 1707
| |
| 01:18:59,920 --> 01:19:03,360
| |
| vivre une vie plus élevée, plus riche de sens, plus spirituelle
| |
| | |
| 1708
| |
| 01:19:03,360 --> 01:19:05,920
| |
| et évolutive, celle pour
| |
| | |
| 1709
| |
| 01:19:05,920 --> 01:19:08,080
| |
| laquelle nous sommes nés.
| |
| | |
| 1710
| |
| 01:19:08,080 --> 01:19:10,640
| |
| Que cela soit notre référence et l'objectif que
| |
| | |
| 1711
| |
| 01:19:10,640 --> 01:19:13,280
| |
| nous devrions nous fixer pour orienter les tendances
| |
| | |
| 1712
| |
| 01:19:13,280 --> 01:19:16,960
| |
| actuelles. Mais pour cela, ces
| |
| | |
| 1713
| |
| 01:19:16,960 --> 01:19:19,199
| |
| réflexions et ces discussions doivent être portées à l'
| |
| | |
| 1714
| |
| 01:19:19,199 --> 01:19:21,840
| |
| attention du public et
| |
| | |
| 1715
| |
| 01:19:21,840 --> 01:19:25,040
| |
| influencer l'élaboration des politiques, du sommet à la base
| |
| | |
| 1716
| |
| 01:19:25,040 --> 01:19:28,239
| |
| . J'espère que chacun d'entre vous, à sa
| |
| | |
| 1717
| |
| 01:19:28,239 --> 01:19:30,640
| |
| manière, apportera un
| |
| | |
| 1718
| |
| 01:19:30,640 --> 01:19:33,520
| |
| léger changement dans sa vie et,
| |
| | |
| 1719
| |
| 01:19:33,520 --> 01:19:35,920
| |
| si vous avez une influence plus importante, qu'il influencera
| |
| | |
| 1720
| |
| 01:19:35,920 --> 01:19:38,080
| |
| le domaine politique ou, à tout le
| |
| | |
| 1721
| |
| 01:19:38,080 --> 01:19:39,679
| |
| moins, suscitera le débat afin que l'
| |
| | |
| 1722
| |
| 01:19:39,679 --> 01:19:43,360
| |
| Inde puisse devenir le
| |
| | |
| 1723
| |
| 01:19:43,360 --> 01:19:46,159
| |
| chef de file de cette transition psycho-spirituelle
| |
| | |
| 1724
| |
| 01:19:46,159 --> 01:19:49,840
| |
| que l'IA imposera à l'
| |
| | |
| 1725
| |
| 01:19:49,840 --> 01:19:52,880
| |
| humanité. C'est en Inde que ce
| |
| | |
| 1726
| |
| 01:19:52,880 --> 01:19:56,320
| |
| savoir a été préservé et imprégné
| |
| | |
| 1727
| |
| 01:19:56,320 --> 01:19:59,199
| |
| dans la conscience collective depuis au
| |
| | |
| 1728
| |
| 01:19:59,199 --> 01:20:02,080
| |
| moins plusieurs millénaires. Ce n'est pas
| |
| | |
| 1729
| |
| 01:20:02,080 --> 01:20:04,960
| |
| notre œuvre. Vous ne pouvez vous en attribuer le mérite.
| |
| | |
| 1730
| |
| 01:20:04,960 --> 01:20:07,920
| |
| Le mérite revient aux rishis qui ont établi
| |
| | |
| 1731
| |
| 01:20:07,920 --> 01:20:10,239
| |
| ce type de civilisation spirituelle pour
| |
| | |
| 1732
| |
| 01:20:10,239 --> 01:20:13,840
| |
| l'Inde. Les rishis de l'époque victorienne ont insufflé cela et… Le
| |
| | |
| 1733
| |
| 01:20:13,840 --> 01:20:16,320
| |
| pouvoir qu'ils nous ont insufflé continue d'
| |
| | |
| 1734
| |
| 01:20:16,320 --> 01:20:18,719
| |
| agir et de façonner nos pensées et nos sentiments,
| |
| | |
| 1735
| |
| 01:20:18,719 --> 01:20:21,679
| |
| malgré toutes les influences superficielles et
| |
| | |
| 1736
| |
| 01:20:21,679 --> 01:20:24,239
| |
| contraires.
| |
| | |
| 1737
| |
| 01:20:24,239 --> 01:20:26,400
| |
| Il est temps d'en parler
| |
| | |
| 1738
| |
| 01:20:26,400 --> 01:20:28,719
| |
| et de nous efforcer consciemment de
| |
| | |
| 1739
| |
| 01:20:28,719 --> 01:20:31,760
| |
| retrouver ces vérités
| |
| | |
| 1740
| |
| 01:20:31,760 --> 01:20:35,440
| |
| spirituelles profondes et de les mettre en pratique.
| |
| | |
| 1741
| |
| 01:20:35,440 --> 01:20:37,360
| |
| Des espaces comme celui-ci seraient le
| |
| | |
| 1742
| |
| 01:20:37,360 --> 01:20:39,760
| |
| point de départ. Je vous
| |
| | |
| 1743
| |
| 01:20:39,760 --> 01:20:43,199
| |
| suggère donc à tous d'entamer cette introspection.
| |
| | |
| 1744
| |
| 01:20:43,199 --> 01:20:47,120
| |
| Qui suis-je ? Pourquoi suis-je ici ? Que
| |
| | |
| 1745
| |
| 01:20:47,120 --> 01:20:49,280
| |
| désiré-je vraiment faire pour donner un sens à ma vie
| |
| | |
| 1746
| |
| 01:20:49,280 --> 01:20:51,120
| |
| ?
| |
| | |
| 1747
| |
| 01:20:51,120 --> 01:20:53,199
| |
| Comment puis-je contribuer au mieux à l'éveil et à la
| |
| | |
| 1748
| |
| 01:20:53,199 --> 01:20:57,040
| |
| croissance spirituelle d'autrui, de toutes les manières possibles, en
| |
| | |
| 1749
| |
| 01:20:57,040 --> 01:20:58,400
| |
| fonction de mon
| |
| | |
| 1750
| |
| 01:20:58,400 --> 01:21:01,199
| |
| tempérament ? Si nous nous
| |
| | |
| 1751
| |
| 01:21:01,199 --> 01:21:02,880
| |
| interrogeons et nous introspectons tous dans ce sens,
| |
| | |
| 1752
| |
| 01:21:02,880 --> 01:21:05,760
| |
| ensemble, nous résoudrons les
| |
| | |
| 1753
| |
| 01:21:05,760 --> 01:21:07,920
| |
| problèmes et les défis que l'IA a
| |
| | |
| 1754
| |
| 01:21:07,920 --> 01:21:10,320
| |
| placés devant nous et trouverons la
| |
| | |
| 1755
| |
| 01:21:10,320 --> 01:21:12,480
| |
| solution optimale
| |
| | |
| 1756
| |
| 01:21:12,480 --> 01:21:16,320
| |
| pour le progrès spirituel de l'humanité.
| |
| | |
| 1757
| |
| 01:21:16,320 --> 01:21:20,400
| |
| Concentrons-nous sur cela dans le calme,
| |
| | |
| 1758
| |
| 01:21:20,400 --> 01:21:22,719
| |
| avec une aspiration sereine. Prions pour que
| |
| | |
| 1759
| |
| 01:21:22,719 --> 01:21:26,320
| |
| la grâce divine
| |
| | |
| 1760
| |
| 01:21:26,320 --> 01:21:30,880
| |
| nous guide vers les meilleurs résultats.
| |
| | |
| 1761
| |
| 01:21:37,440 --> 01:21:40,440
| |
| Namaste.
| |
| | |
| 1762
| |
| 01:21:47,600 --> 01:21:51,120
| |
| Merci Sadi pour tes excellentes
| |
| | |
| 1763
| |
| 01:21:51,120 --> 01:21:53,360
| |
| réflexions.
| |
| | |
| 1764
| |
| 01:21:53,360 --> 01:21:55,360
| |
| Nous voici arrivés au terme du
| |
| | |
| 1765
| |
| 01:21:55,360 --> 01:22:00,320
| |
| programme, mais il ne peut se terminer tant que
| |
| | |
| 1766
| |
| 01:22:00,320 --> 01:22:05,760
| |
| nous… Nous exprimons notre gratitude à tous ceux
| |
| | |
| 1767
| |
| 01:22:05,760 --> 01:22:07,920
| |
| dont les efforts ont permis le
| |
| | |
| 1768
| |
| 01:22:07,920 --> 01:22:12,159
| |
| succès de ce programme. J'invite donc notre
| |
| | |
| 1769
| |
| 01:22:12,159 --> 01:22:15,920
| |
| actuel administrateur délégué, le Dr Shimaran,
| |
| | |
| 1770
| |
| 01:22:15,920 --> 01:22:20,199
| |
| à prononcer le discours de remerciement.
| |
| | |
| 1771
| |
| 01:22:44,719 --> 01:22:46,800
| |
| Non, monsieur.
| |
| | |
| 1772
| |
| 01:22:46,800 --> 01:22:50,880
| |
| Nous sommes tous enthousiastes, éclairés et
| |
| | |
| 1773
| |
| 01:22:50,880 --> 01:22:54,080
| |
| emplis d'élévation.
| |
| | |
| 1774
| |
| 01:22:54,080 --> 01:22:58,159
| |
| Nous exprimons nos sincères remerciements
| |
| | |
| 1775
| |
| 01:22:58,159 --> 01:23:03,199
| |
| et notre profonde gratitude à Sri Shaland Gi qui
| |
| | |
| 1776
| |
| 01:23:03,199 --> 01:23:07,440
| |
| nous a présenté non seulement la
| |
| | |
| 1777
| |
| 01:23:07,440 --> 01:23:10,800
| |
| solution aux problèmes, mais aussi la solution
| |
| | |
| 1778
| |
| 01:23:10,800 --> 01:23:16,440
| |
| aux affaires de Maya.
| |
| | |
| 1779
| |
| 01:23:18,480 --> 01:23:20,480
| |
| Il y a une contradiction qui
| |
| | |
| 1780
| |
| 01:23:20,480 --> 01:23:22,719
| |
| prévaut toujours. Si
| |
| | |
| 1781
| |
| 01:23:22,719 --> 01:23:27,280
| |
| Dieu a donné une intelligence artificielle, alors son
| |
| | |
| 1782
| |
| 01:23:27,280 --> 01:23:30,639
| |
| aspect est également très clair et la
| |
| | |
| 1783
| |
| 01:23:30,639 --> 01:23:32,480
| |
| méthode pour s'en libérer est également
| |
| | |
| 1784
| |
| 01:23:32,480 --> 01:23:35,120
| |
| donnée. Il a dit qu'il avait parlé
| |
| | |
| 1785
| |
| 01:23:35,120 --> 01:23:36,800
| |
| de la conscience individuelle et de
| |
| | |
| 1786
| |
| 01:23:36,800 --> 01:23:38,719
| |
| la conscience universelle, que la
| |
| | |
| 1787
| |
| 01:23:38,719 --> 01:23:42,960
| |
| conscience est au-delà du processus.
| |
| | |
| 1788
| |
| 01:23:52,400 --> 01:23:55,120
| |
| Ainsi, la conscience est ce qui
| |
| | |
| 1789
| |
| 01:23:55,120 --> 01:23:58,159
| |
| est éternel
| |
| | |
| 1790
| |
| 01:23:58,159 --> 01:24:02,760
| |
| parmi toute la multitude, et
| |
| | |
| 1791
| |
| 01:24:02,800 --> 01:24:04,560
| |
| la conscience
| |
| | |
| 1792
| |
| 01:24:04,560 --> 01:24:07,040
| |
| parmi les consciences est la conscience universelle, elle
| |
| | |
| 1793
| |
| 01:24:07,040 --> 01:24:09,360
| |
| imprègne tout. C'est
| |
| | |
| 1794
| |
| 01:24:09,360 --> 01:24:12,480
| |
| la vérité, la vérité, il l'a dit
| |
| | |
| 1795
| |
| 01:24:12,480 --> 01:24:15,520
| |
| deux fois, oui, la vérité, c'est pour cela que nous sommes dotés, c'est pour cela que
| |
| | |
| 1796
| |
| 01:24:15,520 --> 01:24:17,600
| |
| nous sommes nés.
| |
| | |
| 1797
| |
| 01:24:17,600 --> 01:24:21,280
| |
| Ce corps nous a été offert pour accélérer notre karma et connaître
| |
| | |
| 1798
| |
| 01:24:21,280 --> 01:24:24,320
| |
| l'illumination complète, la
| |
| | |
| 1799
| |
| 01:24:24,320 --> 01:24:27,199
| |
| félicité absolue. C'est donc naturellement qu'il
| |
| | |
| 1800
| |
| 01:24:27,199 --> 01:24:32,080
| |
| nous a ouvert tout le champ de… Grâce à la
| |
| | |
| 1801
| |
| 01:24:32,080 --> 01:24:36,000
| |
| miséricorde divine, ce lieu est un confluent
| |
| | |
| 1802
| |
| 01:24:36,000 --> 01:24:39,840
| |
| de savoir et de dévotion, sous l'
| |
| | |
| 1803
| |
| 01:24:39,840 --> 01:24:43,600
| |
| égide de Sit Tapur Maksharodhoda
| |
| | |
| 1804
| |
| 01:24:43,600 --> 01:24:47,040
| |
| et de Swami G. Je remercie donc sincèrement
| |
| | |
| 1805
| |
| 01:24:47,040 --> 01:24:51,600
| |
| Swami, secrétaire de l'Institut de la
| |
| | |
| 1806
| |
| 01:24:51,600 --> 01:24:54,000
| |
| Culture et de la Mission, qui présidait la
| |
| | |
| 1807
| |
| 01:24:54,000 --> 01:25:00,800
| |
| réunion. Il a consacré un temps précieux et a approfondi
| |
| | |
| 1808
| |
| 01:25:00,800 --> 01:25:03,920
| |
| le sujet avec une grande clarté, nous éclairant par ses propos. Les
| |
| | |
| 1809
| |
| 01:25:03,920 --> 01:25:07,440
| |
| départements de
| |
| | |
| 1810
| |
| 01:25:07,440 --> 01:25:09,679
| |
| recherche et d'entretien de la culture
| |
| | |
| 1811
| |
| 01:25:09,679 --> 01:25:12,320
| |
| de cette institution nous
| |
| | |
| 1812
| |
| 01:25:12,320 --> 01:25:14,239
| |
| ont apporté une aide précieuse pour la
| |
| | |
| 1813
| |
| 01:25:14,239 --> 01:25:16,719
| |
| promotion et l'invitation de
| |
| | |
| 1814
| |
| 01:25:16,719 --> 01:25:18,880
| |
| tous. La Société Orurabinda de
| |
| | |
| 1815
| |
| 01:25:18,880 --> 01:25:22,320
| |
| Calcutta nous a également aidés pour la
| |
| | |
| 1816
| |
| 01:25:22,320 --> 01:25:27,159
| |
| promotion, notamment par la création d'affiches et autres supports publicitaires.
| |
| | |
| 1817
| |
| 01:25:37,600 --> 01:25:40,719
| |
| Nous leur exprimons notre gratitude, ainsi qu'à
| |
| | |
| 1818
| |
| 01:25:40,719 --> 01:25:47,040
| |
| She Bachara, le
| |
| | |
| 1819
| |
| 01:25:47,040 --> 01:25:51,840
| |
| gendre du professeur Samshaw
| |
| | |
| 1820
| |
| 01:25:51,840 --> 01:25:57,199
| |
| Chakraorti, et à sa fille Mandirati, qui sont venus de loin. Nous leur sommes profondément reconnaissants. La
| |
| | |
| 1821
| |
| 01:25:59,920 --> 01:26:03,840
| |
| Global Business School leur est également reconnaissante
| |
| | |
| 1822
| |
| 01:26:03,840 --> 01:26:05,760
| |
| pour leur
| |
| | |
| 1823
| |
| 01:26:05,760 --> 01:26:07,760
| |
| soutien et leur participation aux activités du
| |
| | |
| 1824
| |
| 01:26:07,760 --> 01:26:11,199
| |
| Centre pour le développement humain. Enfin,
| |
| | |
| 1825
| |
| 01:26:11,199 --> 01:26:14,320
| |
| merci au Dr Tapush Ranjul Shah. Nous remercions
| |
| | |
| 1826
| |
| 01:26:14,320 --> 01:26:16,560
| |
| également le directeur d'IMS et de l'IMS Business
| |
| | |
| 1827
| |
| 01:26:16,560 --> 01:26:19,520
| |
| School de Kolkata pour nous avoir fourni la
| |
| | |
| 1828
| |
| 01:26:19,520 --> 01:26:22,239
| |
| papeterie (stylos, blocs-notes, agendas,
| |
| | |
| 1829
| |
| 01:26:22,239 --> 01:26:31,800
| |
| clés, etc.). Nous lui sommes très reconnaissants. Nous remercions également
| |
| | |
| 1830
| |
| 01:26:34,320 --> 01:26:38,960
| |
| Silk, fondateur et directeur général de Builders of Digital
| |
|
| |
|
| 1831
| | === Extraits === |
| 01:26:38,960 --> 01:26:43,520
| |
| Economy New USA, pour
| |
|
| |
|
| 1832
| | ==== 10:48 Introduction ==== |
| 01:26:43,520 --> 01:26:46,719
| | <blockquote>Ces trois dernières années, l'intelligence artificielle a frappé l'humanité de plein fouet, avec un impact peut-être le plus radical et déstabilisant de toute l'histoire humaine. En réalité, elle a touché tous les domaines de l'activité humaine, tous les domaines d'activité, de recherche, d'apprentissage, d'interaction, et ce, dès les premiers instants de son influence sur l'humanité. Ce qui se passe réellement dans ce domaine est tellement radical, tellement perturbateur, que nous sommes incapables d'imaginer à quoi ressemblera l'avenir immédiat lorsque l'intelligence artificielle franchira sa prochaine étape imminente : l'intelligence artificielle générale. Ce sera la première fois que l'espèce humaine rencontrera une entité plus intelligente que nous. Et elle est tellement plus intelligente, pas nominalement, pas de 10 points de QI, bien plus , peut-être 100 fois plus, que même la tentative de le neutraliser sera empêchée par sa capacité à manipuler les circonstances et notre processus de pensée en exploitant notre dépendance à son égard. La transition disruptive à laquelle nous sommes confrontés est telle que du jour au lendemain, des millions de personnes se retrouveront sans emploi, car l'intelligence artificielle (IA) les remplacera avec une rapidité, une précision, une capacité, une force et des compétences bien supérieures, mais sans les inconvénients liés à l'activité humaine, tels que les problèmes de santé, le besoin de prendre des congés ou autres problèmes associés. Quelle est la nature de cette transition ?</blockquote> |
| son soutien continu et ses précieux conseils | |
|
| |
|
| 1833
| | ==== 59:35 Une idée cruciale ==== |
| 01:26:46,719 --> 01:26:48,320
| | <blockquote>Tout cela nous amène à une idée cruciale : l'IA elle-même et la RSI (Recursive self-improvement - Auto-amélioration récursive) vont évoluer très rapidement. C'est une question de quelques années seulement. Quant aux conséquences pour nous, impuissants, victimes de l'utilisation croissante de l'IA, car elle nous a rendus dépendants, on observe une tendance croissante à l'affaiblissement de… Diminution de la capacité d'attention, baisse de la concentration, augmentation de l'anxiété et de la dépression chez les jeunes utilisateurs d'IA. C'est la première fois dans l'histoire de l'humanité qu'une telle masse d'anxiété et de dépression se manifeste. Nous vivons une génération marquée par l'anxiété et la dépression, caractérisée par une attention volatile, une mémoire défaillante, une faible concentration et une capacité de compréhension réduite. Nous assistons au déclin de l'intelligence humaine elle-même. |
| aux activités du
| |
|
| |
|
| 1834
| | Chez les individus dont le déclin est suffisamment marqué, le QI chute, le comportement se dégrade et l'on se rapproche d'un état proche de celui d'un animal primitif. Toutes les facultés supérieures de l'humanité disparaissent rapidement. [https://fr.wikipedia.org/wiki/Vivekananda Swami Vivekananda] affirmait que la valeur d'un être humain réside dans sa capacité de concentration. Or, c'est précisément cette capacité qui est en première ligne. La capacité de concentration a chuté de façon spectaculaire chez tous les jeunes utilisateurs. On observe une tendance à la névrose, à la psychose, à la solitude, au narcissisme, etc. Ainsi, les utilisateurs d'IA rencontrent de graves problèmes émotionnels. |
| 01:26:48,320 --> 01:26:51,280
| |
| centre de management. Le
| |
|
| |
|
| 1835
| | Et au final, ce sont ceux qui contrôlent le système – prenons l'exemple du Forum économique mondial – leur évangéliste Yuval Noah Harari a déclaré ouvertement, en substance : ''« L'humanité est piratable. Les êtres humains sont des créatures vulnérables au piratage. Nous n'avons plus besoin d'âme ni de Dieu, car les êtres humains peuvent être manipulés, leur comportement peut être piraté. »'' Ainsi, l'objectif du Forum économique mondial est de manipuler et de pirater le comportement humain jusqu'à ce que – et ce sont leurs propres termes – vous ne possédiez plus rien et que vous soyez heureux, car nous vous programmons, nous vous manipulons pour que vous soyez heureux malgré le fait de ne rien posséder. C'est le type de pensée des structures de pouvoir qui utilisent l'IA contre vous. |
| 01:26:51,280 --> 01:26:55,120
| |
| Dr Chi, fondateur du Man
| |
|
| |
|
| 1836
| | Enfin, avec cet équilibre, les êtres humains, même en tant que travailleurs qualifiés, perdent de leur importance. Les puces mémoire et les puces de calcul d'IA deviennent plus précieuses que les êtres humains. Donc, lorsqu'un travailleur est remplacé, vous êtes remplacé par ces puces. C'est aussi simple que cela. Lorsque cette transition aura lieu, qu'est-ce qui maintiendra l'humanité unie ? Qu'est-ce qui fait de vous une personne digne d'intérêt ? Et, au plus haut niveau de l'industrie de l'IA, cela est considéré comme inévitable. Certains disent même que nous remplaçons les êtres humains par… une espèce supérieure. Est-ce vrai ? Nous arrivons donc à un point de transition où, soudain, tout ce que vous savez, tout ce que vous pouvez faire, devient inutile, remplacé par ces puces. Que reste-t-il alors ?</blockquote> |
| 01:26:55,120 --> 01:26:58,400
| |
| Center for Human Development AU
| |
|
| |
|
| 1837
| | ==== 01:03:00 Le point le plus important ==== |
| 01:26:58,400 --> 01:27:01,520 | | <blockquote>J'en viens maintenant au point le plus important, et c'est sur ce point que nous conclurons. Ce qui reste, c'est ce que vous pouvez faire que l'IA ne peut pas faire. Ce que vous savez faire qui n'est pas transmissible, car tout ce qui peut l'être est intégré à l'IA. Qu'est-ce qui vous rend unique et que l'IA ne peut pas reproduire ? Je vais vous donner quelques suggestions. |
| Memorial Trust, est très présent parmi nous et constitue
| |
|
| |
|
| 1838
| | # La première est l''''empathie'''. L'empathie humaine ne peut être reproduite par une machine. On peut créer un robot infirmier. Mais lorsqu'un être humain vous caresse à l'hôpital et vous dit de ne pas vous inquiéter, que vous allez guérir, il y a un transfert d'énergie, d'amour et d'affection, un profond sentiment de bienveillance et la conviction que vous allez survivre. Seul un être humain peut faire cela. Un robot infirmier peut imiter la caresse, mais il n'y aura rien de cette transmission plus profonde. Cela survivra à l'IA. |
| 01:27:01,520 --> 01:27:03,120
| | # Deuxième point : lorsque vous travaillez, l'IA peut faire le même travail mieux que vous. Mais quand vous travaillez, vous pouvez vous dire : « Si j'ai la flemme, je laisserai l'IA s'en charger. » Prenons l'exemple de la lessive : je dois laver et plier le linge. Une corvée simple à la maison. Mon assistant virtuel peut s'en occuper. Mais ensuite, vous vous dites que laver le linge est pour vous '''un acte sacré'''. Quand je plie le linge, je suis un parent. Je pense à mes enfants et je plie le linge avec amour et affection. La joie que cela me procure… Je ne veux pas la gaspiller avec une IA qui ne ressentira rien. Un travail sacré pour vous. La valeur qu'il a pour vous. Vous choisirez de ne pas utiliser d'IA. Si cela a de la valeur à vos yeux, même si vous êtes paresseux, peu importe. |
| la principale source d'inspiration du | | # Troisièmement, '''la joie de la découverte'''. Oui, l'IA peut résoudre les problèmes. Je peux lui confier tous les problèmes de l'homme. Mais quand je m'en occupe et que je me dis : « Waouh, quelle intuition ! » Oh, c'est intéressant. Je viens de découvrir cette connaissance. Bien sûr, oui, je peux le faire. Je m'en fiche. La joie que cela me procure est plus précieuse que la simple résolution du problème. Alors, pourquoi vouloir apprendre ? Pour la joie du savoir qui peut remplacer tout l'enseignement fondamental actuel. Vous voyez où je veux en venir ? Cela touche au cœur même de nos motivations. |
| | # Quatrièmement, l'IA ne peut égaler l''''intuition'''. Par essence, l'intuition est à un niveau de cognition supérieur, indépendant des données. L'IA peut imiter certaines tendances intuitives à partir de modèles de données. Mais au-delà des données, elle est impuissante. Or, l'intuition est au cœur de la connaissance. Souvenez-vous de cet ''Upanishad'' où la question était : ''« Qu'est-ce donc, dont la connaissance permet de tout connaître ? »'' Si votre quête est de développer une cognition plus élevée et une connaissance plus profonde, alors oui, l'IA ne pourra jamais l'égaler. Et ce qui se produit très vite, c'est que nous aspirons au contact de l'esprit, au contact de l'éternel, au contact de l'infini. Autrement dit, pour ceux qui se tournent vers l'inclination spirituelle de la vie, et je ne parle pas d'une vie spirituelle mais d'une inclination spirituelle de la vie… au quotidien, cette inclination même, apportera à votre vie quelque chose que l'IA ne pourrai jamais gérer, égaler ou remplacer. Et là, vous pourriez faire un choix déterminant. |
| | # Enfin, le cinquième point de ma liste concerne '''les expériences du sacré'''. l'IA ne pourra jamais les égaler. Et nous serons de plus en plus attirés par cela. Vous avez interagi avec l'IA, vous ne ressentez pas le sacré. Vous entrez dans un espace sacré. L'IA ne pourra jamais le recréer. Ainsi, votre inclination vers les choses spirituelles augmentera. |
|
| |
|
| 1839
| | Et pour ceux qui vont dans l'autre sens et disent : « Je ne veux pas ça. Je veux juste le confort », on observe un déclin rapide vers les schémas déjà constatés de dépression, de solitude et, par conséquent, vers la drogue, l'alcool ou d'autres comportements à risque. Nous assistons à une scission de l'humanité, inévitable. Chacun fera son choix et l'IA sera un déclencheur qui accentuera cette scission. |
| 01:27:03,120 --> 01:27:05,600
| |
| centre et de toutes ses
| |
|
| |
|
| 1840
| | La transition économique elle-même, comme je l'ai dit, nous ne savons pas comment la gérer. Imaginez un milliard de personnes soudainement sans emploi. Les discussions n'ont pas encore eu lieu. Quand cela arrivera, nous commencerons peut-être à nous démener pour essayer de rectifier le tir. Je ne sais pas, mais l'avenir, au-delà de cela, lorsque nous commencerons à définir les objectifs supérieurs de l'humanité, la raison de notre existence, nous mènera soit vers le spirituel, soit vers une régression de notre humanité. Ce choix deviendra inévitable. Lequel dépendra des motivations profondes de ceux qui pilotent la recherche et le développement de l'IA. |
| 01:27:05,600 --> 01:27:10,400
| |
| activités. Nous prions pour qu'il
| |
|
| |
|
| 1841
| | Nous revenons sans cesse à cette idée clé que Sri Aurobindo a décrite il y a 100 ans : l'avenir de l'humanité dépend de son '''choix''' entre le '''pragmatisme grossier de la révolution industrielle''' européenne et les '''vérités spirituelles profondes''' du Sanatana Dharma. Ce choix déterminera notre avenir, notre survie ou notre autodestruction. La révolution de l'IA est un signal d'alarme qui nous oblige à prendre cette décision. C'est une opportunité, mais une opportunité qui n'a pas encore été ouvertement débattue publiquement. C'est pourquoi notre discussion actuelle est si importante. Namaste.</blockquote> |
| 01:27:10,400 --> 01:27:12,880
| |
| vive longtemps, car il est le pilier
| |
|
| |
|
| 1842
| | === Résumé === |
| 01:27:12,880 --> 01:27:15,520
| |
| de notre foi et de notre
| |
|
| |
|
| 1843
| | ; Qu'est-ce qui vous rend unique et que l'IA ne peut pas reproduire ? |
| 01:27:15,520 --> 01:27:22,480
| | # l''''empathie''' humaine |
| détermination à perpétuer nos valeurs.
| | # un '''acte''' sacré |
| |}
| | # la '''joie de la découverte''' : la joie que cela me procure est plus précieuse que la simple résolution du problème |
| | # l''''intuition :''' par essence, l'intuition est à un niveau de cognition supérieur, indépendant des données |
| | # les '''expériences''' du sacré |
| | [[Catégorie:Spiritualité]] |