What are the main topics discussed in the interview with Professor Yuval Noah Harari?

The talk covers AI's development, potential for AI consciousness, misconceptions about intelligence and free will, and AI's influence on society and culture.

What are Professor Harari's concerns about AI's role in society?

He expresses concerns about AI's ability to replace human interaction in areas like religion, creativity, and governance due to inherent biases and lack of consciousness.

What is Professor Harari's connection to India?

Professor Harari frequently practices Vipassana meditation in India and explores its cultural and spiritual impacts in his works.

What does Harari suggest about the relationship between AI and human culture?

It suggests that as AI becomes key in decision-making, humanity might need to adapt its narratives and power structures to maintain relevance.

What themes does Harari discuss in relation to AI's evolution and human history?

The book explores technological impacts on human society, cultural narratives, and ethical considerations like truth vs. fiction.

Yuval Noah Harari Interview with Saurabh Dwivedi। AI के खतरे,Ramayana से Narendra Modi तक। Kitabwala

00:57:01

https://www.youtube.com/watch?v=6x4UgHI7qbk

Ringkasan

TLDRProfessor Yuval Noah Harari discusses the rapid development and potential future of Artificial Intelligence (AI), comparing it to the slow organic evolution from simple organisms to complex ones over billions of years. He examines the possibility of AI developing consciousness and the societal implications, such as AI's potential role in religion and creativity. Harari explores the importance of stories in human culture, and how AI might introduce new narratives distinct from biological ones. He also highlights the importance of truth versus fiction in the digital age, illustrating how AI systems could exacerbate misinformation due to their inherent biases. Additionally, Harari shares his personal experiences in India, especially with Vipassana meditation, reflecting on its impact on understanding the mind. He concludes with concerns about centralized power in AI development primarily within the US and China, emphasizing a need for broader global participation and dialogue in shaping AI's future.

Takeaways

🦠 AI is in its early stages, akin to amoeba in evolution.
🤖 Humans may face new narratives from AI beyond biological dramas.
🌐 Centralized AI development in US and China poses risks.
📚 Human storytelling is essential for large-scale cooperation.
💾 AI's inherent biases can spread misinformation.
🧘‍♂️ Harari integrates his Vipassana meditation experiences into his worldview.
🗣 AI has the potential to influence cultural, economic, and religious domains.
❓ Truth is complicated, costly, and often painful, unlike fiction.
⚖️ The balance between truth and order is critical.
🕵️ With increasing AI surveillance, privacy concerns grow.
🧠 Humans often mistakenly attribute consciousness to AI.
🎥 Harari warns about AI-driven creative fields like media and arts.

Garis waktu

00:00:00 - 00:05:00
The evolution from amoebas to dinosaurs took billions of years, whereas AI development is much faster than organic evolution. AI's potential to develop consciousness and empathy poses existential questions, such as reevaluating our constructs and beliefs. Despite AI's current dependency on human commands, its future intelligence may lead to transformative global shifts.
00:05:00 - 00:10:00
Author Yuval Noah Harari discusses his connection to India and his journey with vipassana, highlighting its scientific approach in understanding the mind. Vipassana teaches attentiveness to one's breath, revealing how our minds operate beyond conscious control.
00:10:00 - 00:15:00
Harari contrasts truth and reality, explaining humanity's tendency to create distorted realities. Despite progress, human information hasn't necessarily improved, often resulting in bad decisions due to misleading information. Truth is costly, complex, and painful, while fiction is cheap, simple, and appealing, leading to a predominance of fiction over truth in human narratives.
00:15:00 - 00:20:00
The concept of an 'information diet' is introduced, addressing the overwhelming flood of modern information compared to the need for analysis and reflection. Harari explores computer politics, noting AI's decisiveness and creativity, yet cautioning against its biases.
00:20:00 - 00:25:00
AI's unprecedented potential is likened to a technological revolution, with examples like OpenAI's GPT-4 employing human-like tactics to achieve goals. This highlights AI's ability to make autonomous decisions and create new methods, deviating from human thought processes.
00:25:00 - 00:30:00
The widespread influence of AI, from loans to art, is discussed. Harari warns about AI becoming central in creating our reality, suggesting that this technological trajectory could substantially redefine human societal structures.
00:30:00 - 00:35:00
Considering Singularity and AI's potential, Harari contemplates the biases inherent in AI, stressing that despite their apparent objectivity, AIs reflect human biases from their learned data. The implications of this on creativity and intelligence among AI compared to humans are critically examined.
00:35:00 - 00:40:00
While humans have evolved unique traits, AI began surpassing human creativity in games like Go, prompting discussions on AI's future role in strategic thinking and decision-making across various sectors. Harari predicts AI's eventual advancement beyond biological limitations.
00:40:00 - 00:45:00
As governments traditionally struggled with total surveillance, AI now enables constant monitoring and control by analyzing vast amounts of data without requiring human analysts, risking the rise of an unprecedentedly powerful surveillance state.
00:45:00 - 00:50:00
Despite AI's power to engage it's noted for exacerbating societal divides through targeted content that triggers fear and hatred, fueled by social media algorithms designed to maximize user engagement for profit.
00:50:00 - 00:57:01
Harari calls for worldwide awareness in AI's development to ensure ethical standards are woven into its framework. Emphasizing not only technical governance but also a diverse dialogue involving global voices, especially as AI evolves potentially faster than biological organisms ever did.

Tampilkan lebih banyak

Peta Pikiran

Video Tanya Jawab

What are the main topics discussed in the interview with Professor Yuval Noah Harari?
The talk covers AI's development, potential for AI consciousness, misconceptions about intelligence and free will, and AI's influence on society and culture.
What are Professor Harari's concerns about AI's role in society?
He expresses concerns about AI's ability to replace human interaction in areas like religion, creativity, and governance due to inherent biases and lack of consciousness.
What is Professor Harari's connection to India?
Professor Harari frequently practices Vipassana meditation in India and explores its cultural and spiritual impacts in his works.
What does Harari suggest about the relationship between AI and human culture?
It suggests that as AI becomes key in decision-making, humanity might need to adapt its narratives and power structures to maintain relevance.
What themes does Harari discuss in relation to AI's evolution and human history?
The book explores technological impacts on human society, cultural narratives, and ethical considerations like truth vs. fiction.

Lihat lebih banyak ringkasan video

Dapatkan akses instan ke ringkasan video YouTube gratis yang didukung oleh AI!

Teks

Gulir Otomatis:

00:00:00
from amiba to Dinosaur it took some time
00:00:03
took billions of years but some time in
00:00:05
the history of universe AI we know today
00:00:07
is like the amiba but AI will continue
00:00:10
to develop for thousands of years for
00:00:13
millions of years and it is much faster
00:00:15
than the process of organic evolution
00:00:17
what if they develop Consciousness and
00:00:20
empathy and some of us start realizing
00:00:23
that so far we were worshiping false god
00:00:25
even if AIS don't have Consciousness
00:00:28
they can easily fool us and make us
00:00:32
believe that they feel things whatever
00:00:35
computers can do they need our Command
00:00:37
but if they had developing human
00:00:38
intelligence then probably that is the
00:00:40
next chapter an atom bomb is so much
00:00:42
power but the power is actually in human
00:00:45
hands AI is able to decide things by
00:00:49
itself and is able to invent new things
00:00:52
by itself one of the term you mentioned
00:00:54
in the book which is computer politics
00:00:57
that the easiest way to grab your
00:00:59
attention
00:01:00
and to keep you constantly engaged with
00:01:02
the platform is to press the hate button
00:01:05
or the greed button or the fear button
00:01:07
in our mind first lesson we should teach
00:01:10
or feed to the elgo is that you can
00:01:13
commit mistakes exactly like you create
00:01:15
an AI to screen job applications and if
00:01:19
in the present moment there is bias
00:01:21
against women you train on this data it
00:01:24
will learn oh women's are not good for
00:01:26
this job and the AI will repeat the same
00:01:30
bias in the future the ultimate question
00:01:32
is which story we believe in all our
00:01:35
life over thousands of years we have
00:01:38
believed in biological dramas two males
00:01:41
fighting over a female so you know you
00:01:43
have it in humans you have it in dogs
00:01:45
you have it in elephants you have it in
00:01:47
chimpanzees so you have it in the
00:01:49
ramayana you have it in Shakespeare you
00:01:51
have it in Hollywood and Bollywood every
00:01:53
Alpha has to remember the AI is the
00:01:56
ultimate Alpha now the big question and
00:01:58
the last question how do we go ahead on
00:02:00
this so let me explain
00:02:32
writer Professor Noah
00:02:35
Harari
00:02:49
homod sir thank you so much for joining
00:02:52
it's my pleasure thank you what a
00:02:55
wonderful book so we'll keep giving
00:02:57
references of sapiens hodus and 21
00:02:59
Century problems but first of all I want
00:03:02
to understand your India connection
00:03:04
because the last chapter of your third
00:03:07
book talks about Vias yeah yeah I come
00:03:10
every every year to do vipasana course I
00:03:13
learned vipasana 24 years ago from S
00:03:16
goena I was doing my PhD studies in
00:03:19
history at the time in England in Oxford
00:03:22
and I was trying to understand the world
00:03:23
and to understand life for all the books
00:03:26
and couldn't find really deep answers
00:03:29
and was there was a friend who told me
00:03:31
for one year you have to go and do
00:03:33
travas you have to go Trav pasana and I
00:03:36
thought this is nonsense this is some
00:03:38
some kind of uh Superstition or this is
00:03:40
some kind of Mystic practice Mystic
00:03:42
practice I don't I'm man of science I
00:03:45
don't do mysticism and eventually
00:03:47
convinced me and I went and it was
00:03:50
absolutely amazing uh there was no
00:03:53
mysticism at all it was the most
00:03:55
scientific thing just to observe the
00:03:58
reality from moment to Moment Like a
00:04:01
scientific experiment you want to
00:04:03
understand your mind you will not
00:04:05
understand your mind by just by reading
00:04:06
books observe your mind observe the
00:04:09
breath coming in and out of the nostrils
00:04:12
it really struck me that you know they
00:04:14
gave this simplest exercise just feel
00:04:17
the breath coming in and feel the breath
00:04:19
going out that's it sounds simple I
00:04:22
couldn't do it for more than 10 seconds
00:04:25
before my mind would go to some fantasy
00:04:27
some
00:04:28
memory and this is how I really started
00:04:31
to understand how the mind works that I
00:04:34
have no control over it you know you can
00:04:36
order your eyes to shut you can ORD your
00:04:40
your legs to to lay down you cannot
00:04:42
order your mind to stop very little
00:04:45
control of the mind and since then for
00:04:48
24 years I've been practicing vasana and
00:04:51
almost every year covid was not possible
00:04:54
but almost every other year I come to
00:04:56
India to do a long course of maybe 30
00:04:59
days or 60 days to get to know the Mind
00:05:02
better nice and we are getting to know
00:05:05
about your mind through your books and
00:05:07
surprisingly in this book Nexus there's
00:05:09
a lot of India references you talked in
00:05:12
detail about ramayan you talked about
00:05:15
Swatch Bharat Mission the cleanliness
00:05:16
Drive being run by prime minister
00:05:18
Narendra Modi you talk about theit
00:05:21
atrocities and the cast system of India
00:05:23
and on and on but uh you primarily talk
00:05:27
about Truth versus reality yeah Truth
00:05:30
Versus order and the process of
00:05:32
mythmaking process uh we'll come to that
00:05:35
part later but please tell us about
00:05:37
Truth versus reality because we often
00:05:40
get
00:05:41
confused well um you know people
00:05:45
never almost never are in touch with
00:05:49
reality we are in touch with stories
00:05:51
about reality we are in touch with
00:05:54
images of of reality and they can be
00:05:57
very
00:05:58
distorted uh um
00:06:01
and all the again the E the big effort
00:06:05
that we are trying to to to to know what
00:06:09
is the
00:06:10
truth very often people get to
00:06:13
conflicting answers about it um and if
00:06:17
you look at the Deep causes of the
00:06:21
conflict in the world of all the hatred
00:06:24
and anger and so and so forth so a lot
00:06:27
of people believe that there is
00:06:28
something wrong in the human being which
00:06:31
is why we are doing so many bad things
00:06:34
but
00:06:36
actually most people are good the
00:06:39
problem is if you give good people bad
00:06:41
information if you give good people bad
00:06:44
images of reality then they make very
00:06:47
bad
00:06:48
decisions and what I explore in the book
00:06:52
is how come over thousands of years we
00:06:55
have developed so much as a species
00:06:58
humankind but our information did not
00:07:01
become any better like people today in
00:07:04
the Modern Age they can believe the most
00:07:09
terrible things the most hateful things
00:07:12
just like thousands of years ago so no
00:07:16
matter how much technological progress
00:07:18
we have
00:07:19
made um we have still not made
00:07:21
significant progress
00:07:24
in getting better information and
00:07:28
understanding the reality better and why
00:07:29
is that because we assume that
00:07:32
rationality will keep progressing and
00:07:34
we'll be able to make wise decisions I
00:07:37
mean Homo sapiens the nomat suggests so
00:07:40
but as this book says that we are
00:07:42
continuously though there's rapid
00:07:44
progression we are continuously making
00:07:46
disastrous choices as well and we are
00:07:48
still believing in all those theories
00:07:51
because most information is isn't the
00:07:54
truth a lot of people think if I have
00:07:56
more information I will know more but
00:07:59
this is not the case the truth is
00:08:01
actually a very small and rare subset of
00:08:06
all the information in the world most
00:08:09
information is junk not the truth why
00:08:13
because the truth first of all is costly
00:08:16
if you want to write a truthful account
00:08:19
of anything of biology of history of
00:08:22
physics you need to do research you need
00:08:24
to collect evidence and analyze it this
00:08:27
takes time and energy and money it's
00:08:30
expensive to produce the truth in
00:08:33
contrast Fiction and Fantasy is very
00:08:36
cheap you just write anything you want
00:08:39
you don't need to do any research
00:08:41
doesn't take time doesn't take money so
00:08:44
the truth is costly fiction is cheap the
00:08:47
truth is often complicated because
00:08:50
reality is complicated whereas fiction
00:08:53
can be made as simple as you would like
00:08:55
it to be and people usually prefer
00:08:58
simple stories finally the truth is
00:09:01
often painful there are many things we
00:09:04
don't want to know about ourselves about
00:09:07
our life H about the world about our
00:09:11
nation the truth is painful sometimes
00:09:14
fiction can be made as pleasing as
00:09:17
flattering as you would like it to be so
00:09:20
in a
00:09:21
competition between the costly and
00:09:24
complicated and painful truth and cheap
00:09:28
simple and flattering fiction fiction
00:09:31
tends to win so most of the information
00:09:34
that is flooding the world is not the
00:09:37
truth and as people invent more and more
00:09:40
sophisticated Information Technology
00:09:43
they just flood the world with more
00:09:46
fiction which doesn't make us any wiser
00:09:49
which doesn't make us doesn't give us
00:09:52
knowledge uh we need to make special
00:09:54
effort Beyond just collecting
00:09:56
information you know what you see now in
00:09:59
the world is that people are flooded by
00:10:02
far too much information and we need a
00:10:05
kind of information diet the same way
00:10:08
that people have food diet people know
00:10:10
too much food isn't good for me junk
00:10:12
food isn't good for me because I need
00:10:15
actually time to digest the food not
00:10:18
just to eat it it's the same as
00:10:20
information too much information is not
00:10:23
good for the mind we need some of it of
00:10:25
course but we actually need more time to
00:10:29
dig digest and analyze and meditate over
00:10:32
the information to tell the difference
00:10:34
between truth and
00:10:36
fiction um so in this sense we need an
00:10:39
information diet you are talking about
00:10:42
truth and this is reminding me of one of
00:10:44
the term you mentioned in the book which
00:10:46
is computer politics and I found it very
00:10:49
interesting where you talking about
00:10:51
algorithm you are talking about
00:10:53
interconnectivity of computers how they
00:10:54
are making choices while playing a game
00:10:57
with a human being how they are deriving
00:11:00
something very new and out of the book
00:11:02
how they are decoding that capture image
00:11:05
hle which is a very new thing for them
00:11:08
and how they are tricking human in
00:11:10
believing that they are not robots but
00:11:12
human in distress yes and that's very
00:11:15
alarming in long we yeah we are now in
00:11:19
the midst of maybe the most important
00:11:22
technological revolution in history
00:11:24
which is the invention of
00:11:26
ai ai is different from every previous
00:11:30
technology because it is not a tool it
00:11:34
is an agent it is something that can
00:11:37
make independent decisions and can come
00:11:40
up with new ideas by itself all previous
00:11:44
technology if you think about even
00:11:46
something very powerful like an atom
00:11:48
bomb an atom bomb is so much power but
00:11:52
the power is actually in human hands
00:11:54
because it is only a human that decides
00:11:57
what to do with the atom bomb an atom
00:11:59
bomb cannot decide by itself which city
00:12:03
to bomb to attack an atom bomb cannot
00:12:06
invent a new kind of bomb or something
00:12:09
else it has no capacity to invent things
00:12:13
AI is able to decide things by itself
00:12:17
and is able to invent new things by
00:12:19
itself so one famous example which you
00:12:22
mentioned is that two years ago open AI
00:12:26
one of the leading AI companies it
00:12:28
developed this new AI uh gp4 which now
00:12:33
everybody is using in cat GPT and all
00:12:35
these things so they developed gp4 they
00:12:37
wanted to know what can this thing do so
00:12:40
they gave it as a test the goal of
00:12:44
solving a capture puzzle a capture
00:12:47
puzzle is this visual puzzle that is
00:12:50
meant to tell the difference between
00:12:52
human and computer like when you access
00:12:55
a web page like a bank account the bank
00:12:57
wants to know is it a human or is it a
00:13:00
bot so they show you this image of maybe
00:13:04
some Twisted letters or whatever and you
00:13:06
have to find to say what these letters
00:13:08
are this is the capture puzzle now gp4
00:13:13
could not solve the capture puzzle by
00:13:15
itself but it was given access to a web
00:13:18
page Tusk rabbit where you can hire
00:13:21
people to do things for you so gp4 try
00:13:25
to hire a human being to solve the
00:13:27
capture puzzle for it now the human got
00:13:30
suspicious so the human wrote asked why
00:13:35
do you need somebody to solve you a
00:13:36
cupture puzzle are you a robot it asks
00:13:40
the human ask the important question are
00:13:42
you a robot and gp4 answered back no I'm
00:13:47
not a I'm not a robot I'm a human being
00:13:50
but I have a vision impairment which
00:13:53
makes it difficult for me to see the
00:13:55
capture puzzle this is why I need your
00:13:57
help and the human was
00:14:00
fooled and believed it and solved the
00:14:03
puzzle for it so it achieved its goal
00:14:07
now what this simple example shows is
00:14:10
the two remarkable abilities of AI first
00:14:14
to make decisions by itself nobody told
00:14:17
gp4 lie to that human it meaned to make
00:14:21
a decision do I tell the truth to the
00:14:23
human or do I lie and by itself it
00:14:27
decided I will lie because I need to
00:14:29
solve the capture puzzle I need to
00:14:30
achieve the goal secondly it shows the
00:14:34
ability of the AI to invent new things
00:14:37
nobody told gp4 what lie would be most
00:14:41
effective it invented the lie that I'm a
00:14:44
human with a vision impairment all by
00:14:46
itself and it was a very effective lie
00:14:49
now this is a very very small incident
00:14:52
but this tells us that what we are
00:14:55
creating is Agents able to invent things
00:14:59
and make decisions and we are now
00:15:01
creating millions billions of these AI
00:15:05
agents and they are everywhere
00:15:07
increasingly making decisions about us
00:15:10
if you apply to a bank to get a loan
00:15:13
it's an AI deciding whether to give you
00:15:15
a loan increasingly these AIS are
00:15:18
writing texts they are creating music
00:15:21
they may be able to create entire movies
00:15:24
so increasingly we will live in a world
00:15:27
which is the product not of the human
00:15:29
mind and our inventions but is the
00:15:32
product of the intelligence of AI
00:15:36
Professor lot of futurist who are
00:15:38
talking about Singularity they are
00:15:40
saying that uh maybe AI will solve the
00:15:42
problem it will be more objective more
00:15:45
data driven and uh less uh there will be
00:15:48
less probability of them committing
00:15:50
mistake because the biases will be less
00:15:52
but you are categorically asserting in
00:15:54
your book that computers do have deep
00:15:58
biases and you have examples but at the
00:16:00
same time in the later part of the book
00:16:02
you are talking about their capability
00:16:04
of being creative and having human
00:16:07
intelligence above even human
00:16:09
intelligence yeah but in that case it's
00:16:13
more challenging because we always
00:16:16
believe that whatever computers can do
00:16:18
they need our Command they cannot be
00:16:21
creative like human being and they
00:16:23
cannot have human values like compassion
00:16:26
Consciousness etc etc but if if they are
00:16:29
developing human intelligence then
00:16:31
probably that is the next chapter we
00:16:33
don't know we already know that AIS are
00:16:35
becoming even more creative than us in
00:16:38
different fields like one famous example
00:16:42
was already eight years ago in 2016 one
00:16:45
of the most important turning points in
00:16:48
the history of AI was a game a game of
00:16:52
Go between a computer program an AI
00:16:56
called alphago and the world champion go
00:16:59
L now go is a board game a strategy game
00:17:04
a bit like chess but much more
00:17:06
complicated than
00:17:07
chess that was invented in ancient China
00:17:10
more than 2,000 years ago and for more
00:17:13
than 2,000 years it was considered one
00:17:17
of the out forms that every uh every
00:17:22
accomplished person in East Asia needed
00:17:24
to know how to play go uh so tens of
00:17:28
millions of Chinese and Koreans and
00:17:30
Japanese they played go for thousands of
00:17:32
years entire philosophies developed
00:17:36
around the game it was seen as a
00:17:38
metaphor for life if you know how to
00:17:40
play goal you will know how to rule a
00:17:42
country even life lessons yes so entire
00:17:46
philosophies developed around different
00:17:48
ways to play go and then in 2016 alphao
00:17:52
played against ladol and it completely
00:17:55
defeated the human Champion but the
00:17:58
import important thing was how it played
00:18:01
it invented a completely new way of
00:18:05
playing the game in the beginning when
00:18:07
it started making these moves all the
00:18:10
commentators said oh this machine is so
00:18:12
stupid nobody plays go like this no this
00:18:15
is stupid mve what a blender yeah what a
00:18:17
blender what a mistake it turned out
00:18:19
this was brilliant now for thousands of
00:18:22
years tens of millions of people play
00:18:25
this game nobody ever thought of making
00:18:28
such moves you know you can think about
00:18:31
it like as a metaphor like there is a
00:18:34
landscape a geography of go all the
00:18:37
different ways you can play go so think
00:18:40
about it like a planet of all the
00:18:42
different ways you can play Go humans
00:18:45
thought that they explore the whole
00:18:47
planet oh we know all the ways to play
00:18:50
Go actually it turned out humans were
00:18:53
stuck on one Island on this planet and
00:18:57
they didn't know all the other ways to
00:18:59
play go then Alpha go came and it
00:19:02
started exploring all these unknown
00:19:05
areas of the planet go all these unknown
00:19:09
ways to play go so it completely
00:19:12
revolutionized the way that go is being
00:19:15
play today and it turned out the
00:19:17
computer was more creative than the
00:19:20
human being now this is likely to happen
00:19:23
you can say oh this is just a game
00:19:24
doesn't matter but this is likely to
00:19:26
happen in more Fields it could happen in
00:19:29
finance if you give AI Control of
00:19:33
Finance so in banks in Investments AIS
00:19:37
will start inventing new ways to invest
00:19:40
money new Financial devices which are
00:19:43
more creative than human Bankers maybe
00:19:46
it even happens in fields like out as
00:19:49
they write new music and new poetry and
00:19:52
even in religion you think about you
00:19:55
know religion like Judaism that I know
00:19:58
best so Judaism is a religion of texts
00:20:02
there is a holy book The Bible and then
00:20:04
people read it and write more texts
00:20:07
about the Bible and then they write more
00:20:09
texts about the texts about the Bible
00:20:11
over thousands of years Jews wrote
00:20:14
thousands of texts which are all now
00:20:17
considered holy sacred to some extent
00:20:21
now no human being there are so many
00:20:23
texts in Judaism nobody is able to read
00:20:26
and remember all of them but AI can no
00:20:31
even the most genius Rabbi cannot read
00:20:34
all of them but AI can read all the
00:20:37
texts of Judaism and remember every word
00:20:40
in every text and find connections
00:20:43
patterns in the Holy books that escape
00:20:46
the human mind now what happens if AI
00:20:51
starts coming up with new
00:20:53
interpretations of the Bible and new
00:20:56
ideas in Judaism would people believe
00:21:00
the AI or would say no no no no no
00:21:02
computers have nothing to do with
00:21:04
religion what happens if you go online
00:21:07
and you have a question about the Bible
00:21:09
or about religion and you get an answer
00:21:11
and you don't know is this coming from
00:21:13
an AI or is this coming from a human
00:21:16
Rabbi or a human priest now if people
00:21:19
think that texts are sacred and AI is
00:21:24
better than humans at reading and
00:21:27
writing text
00:21:29
then AIS could take over
00:21:32
religion uh so religion in the first
00:21:35
place claimed to give answers to all of
00:21:37
our queries from where did we come from
00:21:40
what is the purpose of this life in the
00:21:42
later part science started making that
00:21:44
claim on the basis of few solid solid
00:21:47
robust mechanism that you can repeat
00:21:49
these test and all are you saying that
00:21:52
in a near future there's a possibility
00:21:54
that AI while carefully reading all our
00:21:58
choices all our biases because it is
00:22:00
relying on the digital bureaucracy and
00:22:02
it is reading all our imprints they will
00:22:05
create a theory which will give
00:22:08
satisfaction to our spiritual Quest as
00:22:10
well and they can produce through all
00:22:14
the avable text of all the religions and
00:22:16
they know our biases that what we will
00:22:18
like the language the music and
00:22:20
everything and they can very judiciously
00:22:22
mix science as well to to satisfy us and
00:22:25
then they can become the the super cult
00:22:28
yeah
00:22:29
that again if they go in a bad Direction
00:22:32
they can learn how to manipulate Us and
00:22:35
how to manipulate our feelings how to
00:22:38
manipulate our beliefs more efficiently
00:22:42
than any human being was ever able to
00:22:45
manipulate other people so this is one
00:22:48
of the big dangers that it will learn
00:22:52
how to basically hack the human mind and
00:22:56
how to manipulate human beings on a
00:22:58
scale which was previously
00:23:01
impossible you know throughout human
00:23:04
history um dictators and tyrants and all
00:23:08
kinds of governments they always wanted
00:23:11
to spy and to monitor everybody to know
00:23:16
what everybody is doing to control all
00:23:18
the people yeah you have given that
00:23:20
Romanian yeah scientist exactly and
00:23:24
until today even the most totalitarian
00:23:27
governments could wouldn't actually spy
00:23:30
on everybody all the time because it was
00:23:32
impossible like if you live even in the
00:23:35
Soviet Union in the time of Stalin so
00:23:38
Stalin wants to follow everybody all the
00:23:40
time but he doesn't have enough agents
00:23:43
secret police agents because if you you
00:23:46
know if if the government wants to know
00:23:47
what you're doing 24 hours a day they
00:23:50
need at least two agents because they
00:23:52
need to sleep sometimes the agents they
00:23:54
need to eat sometime so let's say two
00:23:56
agents are following you 24 hours a day
00:23:59
the same way they want to follow
00:24:00
everybody in the country it is
00:24:02
impossible and then they have to follow
00:24:03
the agents as well yes they don't have
00:24:05
enough agents and even if somebody is
00:24:07
following you 24 hours a day at the end
00:24:10
of the day they collect information
00:24:12
about you they write some paper report
00:24:14
you went there you read this book you
00:24:16
met this person somebody needs to
00:24:19
analyze all the information the agents
00:24:22
the secret agents are gathering on you
00:24:25
in order to make sense of it now they
00:24:27
don't have enough analysts to do it need
00:24:30
millions and millions of human beings to
00:24:33
analyze all the information it's
00:24:34
impossible so even in a totalitarian
00:24:38
state like the Soviet Union most people
00:24:41
had privacy it was not possible for the
00:24:44
government to follow them all the time
00:24:46
now with AI it is becoming possible
00:24:49
because a phone is everywhere yeah the
00:24:50
phone is everywhere there are
00:24:52
microphones there are cameras there
00:24:53
drones everywhere so you can follow
00:24:55
everybody all the time you don't need
00:24:57
human agents for that and you don't need
00:25:00
human analysts to analyze all the
00:25:03
information like if you're going and the
00:25:05
camera sees you went there and the phone
00:25:07
is recording your conversation you don't
00:25:09
need a human analysts to actually look
00:25:12
at all the images and analyze all the
00:25:14
text AI is doing it automatically so the
00:25:18
danger is that it is becoming possible
00:25:20
to create a total surveillance regime in
00:25:24
which a dictatorship would just follow
00:25:27
everybody all the time time and would be
00:25:28
able to control and to manipulate people
00:25:31
all the time at the end of the day
00:25:34
everybody wants to predict the
00:25:35
behavioral pattern so that they can goge
00:25:38
that what is the next move you're going
00:25:39
to make yeah now with large data they
00:25:41
are able to do so but at the same time
00:25:44
they are altering our reality because
00:25:46
they are the one who are choosing on our
00:25:48
behalf and we are under this false
00:25:50
assumption that we made this Choice
00:25:52
exactly you gave the example of say uh
00:25:55
myamar 201617
00:25:58
that rohinga kamish and then how the St
00:26:03
Vidal's videos were flooded in the
00:26:06
Facebook stream and how it created the
00:26:10
unrest and uh then you tell us that
00:26:13
because there was one simple instruction
00:26:15
to the
00:26:16
algorithm increase the engagement and
00:26:19
they forgot to differentiate between
00:26:22
love hate Compassion or any other thing
00:26:25
how do you look at it because at one
00:26:26
point of time you were talking about
00:26:28
that computer do not understand double
00:26:30
speak where you were giving the Russian
00:26:32
bot example yeah how should we look at
00:26:34
it because all these big data companies
00:26:37
who are integral part of our life they
00:26:39
are clearly not responsible enough yeah
00:26:43
what we see with social media in the
00:26:45
last few years that it is spreading
00:26:48
hatred and fear and anger which creates
00:26:51
political extremism and violence and
00:26:54
instability all over the world now what
00:26:57
is actually
00:26:59
happening what is happening is that the
00:27:01
companies have an economic interest in
00:27:04
keeping people use their products more
00:27:08
and more the more time you spend on Tik
00:27:11
Tok or Facebook or Twitter or any of
00:27:14
these social media the more money the
00:27:16
company makes because it shows you more
00:27:19
advertisements and it can collect more
00:27:21
data from you and use that data or sell
00:27:24
it to somebody else so their aim is to
00:27:26
keep you on the platform
00:27:29
longer and the companies gave now these
00:27:33
platforms they are not managed by human
00:27:36
beings they are managed by algorithms
00:27:39
when you watch some video who chose to
00:27:42
show you this video Not a Human Being
00:27:44
but an algorithm and the companies gave
00:27:48
the algorithms one goal increase user
00:27:52
engagement which means make people spend
00:27:55
more time on the platform and also
00:27:57
engage with it more like share it with
00:28:00
friends and send it more to so that more
00:28:02
people will join and also be on the
00:28:05
platform now the algorithms experimented
00:28:08
on millions of human beings and they
00:28:11
discovered that the easiest way to grab
00:28:15
your attention and to keep you
00:28:17
constantly engaged with the platform is
00:28:19
to press the hate button or the greed
00:28:22
button or the fear button in our mind if
00:28:25
we something if we see something that
00:28:27
makes us us very angry we engage with it
00:28:31
we can't look away if we see something
00:28:33
that scares us we can't look away and
00:28:37
you know this goes back millions of
00:28:40
years in
00:28:41
evolution that there is a mechanism in
00:28:44
the human brain that if you see a danger
00:28:48
all your attention goes there and you
00:28:50
cannot look away yeah because you know
00:28:52
in in nature this is this is H good for
00:28:55
you because if you go in the forest you
00:28:58
look for something to eat and there is a
00:29:00
snake somewhere you even if it's just a
00:29:04
you see some movement all your attention
00:29:06
goes there survival Instinct and you see
00:29:08
the snake and you don't look away you
00:29:10
only look at the snake because you know
00:29:12
if you look away maybe it kills me so we
00:29:14
have a mechanism in the brain if we see
00:29:16
something dangerous immediately all the
00:29:19
attention goes there now when you go in
00:29:21
the forest most of the time there is no
00:29:24
snake mhm so you look here you look
00:29:26
there there is a tree there is Bush
00:29:28
there is a bird and every now and then
00:29:30
okay there is a snake but when you look
00:29:32
on social media it's always snake snake
00:29:36
snake snake because they figured out
00:29:39
that the easiest way to grab our
00:29:41
attention is to show us something that
00:29:43
frightens us yes and they also
00:29:46
discovered what frightens different
00:29:47
people because what frightens me may not
00:29:50
be frightening to you MH so the
00:29:52
algorithms of social media they show you
00:29:54
different things they monitor how you
00:29:56
react they they discover oh you're
00:29:59
afraid of this we'll show you more of
00:30:01
this so if you're afraid of some
00:30:04
political party you're afraid of
00:30:05
imigrants you're afraid of this group of
00:30:08
that group they discover this and the
00:30:10
algorithms show you more and more and
00:30:12
you become more and more frightened and
00:30:15
you stay longer and longer and they
00:30:17
completely putting hus on us that you
00:30:18
are making this Choice that's why we are
00:30:20
showing this exactly and when you
00:30:22
complain about it they say nobody is
00:30:24
forcing you to do it if you want you can
00:30:26
just put it away we not forcing anybody
00:30:29
but actually what is happening is that
00:30:32
they hacked the operating system of
00:30:35
humans you know how do you hack a
00:30:37
computer you find the weakness in the
00:30:39
code and you use the weakness to hack
00:30:42
the computer it's the same with human
00:30:44
beings they found the weakness in our
00:30:47
code and they use it to hack our brains
00:30:51
and the thing is that you know people a
00:30:55
lot of people have a strong belief in
00:30:56
free will everything I do this is my
00:30:59
free will so if I choose to just keep
00:31:02
seeing this on social media this is my
00:31:04
free will actually the easiest people to
00:31:07
manipulate are those who have a strong
00:31:10
belief in free will because if you
00:31:12
believe very strongly everything I do is
00:31:15
my free will then you don't even think
00:31:18
that maybe you are manipulated in some
00:31:21
way now on a philosophical level belief
00:31:26
in free will if you just take it like
00:31:27
like a blind faith this is creates a
00:31:30
problem because it makes you not very
00:31:33
curious to understand how you make
00:31:35
decisions there is nothing to
00:31:37
investigate I just decide freely to buy
00:31:40
this to go there to do
00:31:42
that um actually I would say that don't
00:31:47
believe in free
00:31:48
will investigate how you make
00:31:52
decisions using science using meditation
00:31:55
using psychology try to really
00:31:58
understand why do you choose this and
00:32:00
not that if after all your
00:32:03
investigations you will find that there
00:32:06
is still something that cannot be
00:32:08
explained by any other way then you can
00:32:10
say okay this is free will but don't
00:32:13
start with the assumption that I have
00:32:15
free will anything I decide is my free
00:32:18
will because if you investigate you will
00:32:21
discover that a lot of your decisions
00:32:23
are actually not free will they are
00:32:26
influenced they are shaped by economic
00:32:30
forces biological forces cultural forces
00:32:34
freedom is not something we have freedom
00:32:36
is something we have to struggle
00:32:39
for if we learn how to protect ourselves
00:32:43
from all these different kinds of
00:32:44
manipulations then we have much more
00:32:47
freedom then comes the idea of Justice
00:32:51
because in the historical course we have
00:32:53
seen that how
00:32:55
alignment changed the shape of History
00:32:58
somebody made a choice that when we'll
00:33:00
present the final version of Bible these
00:33:03
are the parts which will be included and
00:33:04
these will not be part of it and how it
00:33:07
led to gender inequality and other parts
00:33:11
and you stress a lot on the canonization
00:33:13
of algorithm yes what are the basic
00:33:16
human principles we need to keep in mind
00:33:19
while working on this
00:33:22
canonization so the process of
00:33:25
canonization we are familiar with the in
00:33:28
history from the creation of holy books
00:33:31
like if you consider the Bible some
00:33:32
people think oh it just came down from
00:33:34
heaven like this in this book but we
00:33:37
know from history this is not how it
00:33:39
happened in the first centuries of
00:33:43
Christianity Christians wrote a very
00:33:46
large number of texts all kinds of
00:33:49
stories about Jesus and prophecies and
00:33:51
things like that in the time of Jesus
00:33:54
himself there was no new testament mhm
00:33:57
he didn't write it and and even
00:34:00
centuries afterwards there was still no
00:34:03
New Testament no
00:34:04
Bible so people were reading different
00:34:07
things and eventually the church decided
00:34:11
that we need one book which has the most
00:34:15
important texts of Christianity so
00:34:19
people will all read the same book and
00:34:22
they set up a committee to go over all
00:34:25
the texts and decide what will be in and
00:34:29
what will stay out and this committee
00:34:32
chose 27 texts that they became what we
00:34:36
know today as the New Testament the the
00:34:39
second part of the Bible you have the
00:34:41
Old Testament and you have the New
00:34:43
Testament and they these were human
00:34:46
beings making the decision what to
00:34:49
include and what not to include uh and
00:34:53
this decision had very far-reaching
00:34:55
consequences for instance one very
00:34:58
popular text in early Christianity was
00:35:01
called the acts of Paul and Thea it told
00:35:05
the story of St Paul and his female
00:35:08
disciple Thea this was a holy woman who
00:35:12
performed Miracles and led the community
00:35:15
and preached and she was an example that
00:35:18
women can be leaders in the church there
00:35:21
was another text uh also attributed to
00:35:26
the same man St Paul in which St Paul
00:35:29
says that women cannot be leaders at all
00:35:32
they should just obey the men and be
00:35:34
silent and this is their role is to obey
00:35:37
the men and to raise
00:35:39
children and these two texts were in
00:35:42
conflict and this
00:35:44
committee uh of the church they chose to
00:35:48
include the text uh that says that women
00:35:52
should just obey men this is still in
00:35:54
the Bible it is the first epistle to
00:35:56
Timothy whereas the acts of Paul and
00:35:59
thla which says that no women can be
00:36:01
leaders they said no no no no no we
00:36:03
don't put this in the Bible this we live
00:36:05
outside and this choice of what to put
00:36:08
in and what to live outside this shaped
00:36:12
the view of christs about women for
00:36:15
thousands of years until today because
00:36:17
now people say oh it's in the Bible so
00:36:19
we believe what's written in there this
00:36:21
is the process of
00:36:23
canonization of taking certain texts and
00:36:28
deciding this is now the holy book the
00:36:31
same thing is happening today with
00:36:34
AI that you know for a lot of people CH
00:36:38
GPT becomes like the Bible whatever they
00:36:42
want to know they just ask the AI and
00:36:45
whatever the AI says they just believe
00:36:48
and those biases continue like that uh
00:36:51
experiment in one of the e-commerce
00:36:53
platform where they had given AI this
00:36:56
job of shortlisting the CVS and and they
00:36:59
took out the women CV yes so AIS can be
00:37:03
biased AIS can be biased against women
00:37:06
for instance why because AIS are trained
00:37:09
on existing data we need to think about
00:37:12
AI it's a bit like a child when you
00:37:15
create an AI the AI doesn't know much
00:37:18
about the world baby algorithm yeah baby
00:37:20
algorithm it has the
00:37:23
ability to learn things about the world
00:37:26
but in order to learn you need to give
00:37:28
it data the same way that a baby grows
00:37:32
up it has different experiences in the
00:37:34
world it learns so AI also has this
00:37:37
period of childhood when it is given all
00:37:41
kinds of data and this is how it learns
00:37:43
about the world now if you give the AI
00:37:46
data which is biased against women the
00:37:49
AI will develop a bias against women
00:37:53
like you create an AI to screen job
00:37:56
applications and if in the present
00:37:59
moment there is bias against women like
00:38:02
it's more difficult to for women to get
00:38:04
a job then the AI you train on this data
00:38:08
it will learn oh women are not good for
00:38:11
this job and the AI will repeat the same
00:38:14
bias in the future and if people think
00:38:17
ah but the AI is perfect the AI makes no
00:38:20
mistakes it has no biases again like
00:38:22
like the three the holy book then it's a
00:38:25
very dangerous situation
00:38:28
when you created a very biased system
00:38:31
and people just have Blind Faith in it
00:38:34
and you are advocating in the book that
00:38:36
the first lesson we should teach or feed
00:38:39
to the elgo is that you can commit
00:38:41
mistakes exactly and you are given that
00:38:45
Nick example of paper clip H yeah this
00:38:48
is a a famous uh uh experience thought
00:38:51
experiment of philosopher Nick Bostrom
00:38:53
this was from like 10 years ago that he
00:38:56
says what happened Happ s if you have
00:38:58
this super intelligent Ai and you give
00:39:01
it a goal like let's say you have a
00:39:05
paper clip Factory a factory making
00:39:07
paper clips and the factory gets an AI
00:39:11
buys an AI and tells the AI your goal is
00:39:15
to create to make as many paper clips as
00:39:18
possible this is the goal sounds okay
00:39:22
but then what the AI does it takes over
00:39:24
the whole world and turns the whole
00:39:26
world into a paper clip Factory it kills
00:39:29
all the humans because it realizes hey I
00:39:32
need to make more paper clips these
00:39:34
humans they consume all the water and
00:39:37
electricity and and and they would not
00:39:39
like if I take it to make paper clips so
00:39:42
I should kill
00:39:43
them and this is again a philosophical
00:39:46
thought experiment so in this thought
00:39:48
experiment the paperclip AI
00:39:52
destroys the whole world just to make
00:39:55
paper clips because this was the goal it
00:39:58
was given and it cannot question it its
00:40:01
goals now this sounds ridiculous until
00:40:04
you think about what happened in social
00:40:06
media which we just disc discussed which
00:40:09
was exactly the same thing the social
00:40:12
media algorithms were given the goal cre
00:40:16
maximize user engagement make people
00:40:18
spend as much time as possible on social
00:40:21
media and pursuing this goal the
00:40:26
algorithms spread so much hatred and
00:40:30
violence and uh uh greed in the world
00:40:34
and because they pursued this simple
00:40:37
goal so we need to develop AIS which are
00:40:42
able to have doubts that are able not to
00:40:46
be so sure that what they are doing is
00:40:49
the correct thing the same thing with
00:40:51
human
00:40:52
beings we give them so much power if
00:40:56
someone has enormous power and no
00:40:59
ability to doubt themselves they're
00:41:02
always convinced they are doing the
00:41:04
right thing this is a very dangerous
00:41:06
mixture Professor the ultimate question
00:41:09
is which story we believe in all our
00:41:12
life over thousands of years we have
00:41:15
believed in biological dramas we were
00:41:18
part of bureaucratic structure but we
00:41:20
never really liked it there was always a
00:41:22
sense of Doubt for that and you have
00:41:24
beautifully in various chapters compared
00:41:26
this and in the biological drama you
00:41:28
have given a detail example of ram as
00:41:30
well so please tell us about that why do
00:41:33
we tend to uh gravitate towards
00:41:36
biological drama is it something
00:41:38
inherent or is it or is it something
00:41:40
because in the civilization first we
00:41:43
started walking on two feet then we
00:41:47
discovered fire and then we I mean
00:41:50
intestines they shrink and the Brain
00:41:53
developed and then we developed language
00:41:55
and imagin ation yeah which computer was
00:41:58
thinking as double speak yeah but that
00:42:01
was one of the reason of our progression
00:42:02
as well because in totality we were able
00:42:04
to convince a lot large group of people
00:42:06
that this is one system or one faith we
00:42:08
can believe in why do we tend to
00:42:10
biological dramas so let me
00:42:13
explain um humans are storytelling
00:42:16
animals we think in stories and our
00:42:20
power in the world ultimately comes from
00:42:23
telling stories which sound strange but
00:42:27
if you think what makes human being so
00:42:29
powerful it's our ability to cooperate
00:42:31
in very large numbers you know other
00:42:34
animals chimpanzees elephants horses
00:42:38
they can cooperate in small numbers like
00:42:40
20 chimpanzees or 20 elephants can
00:42:42
cooperate but 20 million cannot humans
00:42:46
can cooperate in unlimited numbers like
00:42:49
you have India today most popular
00:42:51
country in the world more than 1.4
00:42:53
billion people and they all cooperate
00:42:56
how do you make
00:42:57
millions of people who never met they
00:43:00
complete strangers nevertheless
00:43:02
cooperate by telling stories at the
00:43:04
basis of every large scale cooporation
00:43:07
you find stories could be religious
00:43:10
stories like mythology could be
00:43:12
political stories like ideology could
00:43:14
even be economic stories money is also a
00:43:17
story it has no objective value you know
00:43:20
you can't eat money you can't drink
00:43:22
money the value of money is that this is
00:43:24
a story everybody believes
00:43:27
now where do the plot of the story comes
00:43:31
from and you would think that people are
00:43:34
very imaginative we invent so many
00:43:36
different stories but actually there are
00:43:39
very very few stories that we tell again
00:43:42
and again and again and who invented
00:43:45
them not human imagination but nature
00:43:48
biology these are the biological
00:43:51
dramas which are common actually to most
00:43:56
animals to most mammals like what are
00:43:59
these biological dramas so one drama for
00:44:02
instance is relation between parents and
00:44:05
children and the children on the one
00:44:08
hand they need to stay close to the
00:44:09
parent to get protection but they also
00:44:12
need to distance from the parent to
00:44:15
explore the world so this is a basic
00:44:18
drama in the life of every child but
00:44:22
also of every puppy of every kitten of
00:44:25
every small elephant or or monkey or
00:44:28
giraffe is this I want to stay close to
00:44:31
mother ah but I want to explore the
00:44:33
world and you see this drama of uh uh
00:44:38
playing out even in religious
00:44:41
mythologies like you have entire
00:44:44
mythology saying that hell is being
00:44:47
distant from God because this is what
00:44:50
every little child feels that if I get
00:44:53
too distant from my parents this is hell
00:44:56
i'll be alone I'll have no
00:44:59
protection uh similarly um and and
00:45:04
another biological drama is two males
00:45:09
fighting over a female so you know you
00:45:12
have it in humans you have it in dogs
00:45:14
you have it in elephants you have it in
00:45:16
chimpanzees this is a biological drama
00:45:19
so much of human mythology and Human Art
00:45:23
just retells the story so you have it in
00:45:25
the ramayana you have it in Shakespeare
00:45:27
you have it in Hollywood and Bollywood
00:45:30
they're taking the basic stories from
00:45:33
biology and everybody gives it a little
00:45:36
different twist so the story is not
00:45:38
always exactly the same but the basic
00:45:42
structure is always the
00:45:44
same and this is how humans create
00:45:48
stories and we find stories appealing
00:45:52
because there is something biological
00:45:54
deep inside us that makes us very
00:45:57
interested in these kinds of stories
00:46:00
like relations between parents and
00:46:01
children or love triangles now ai is not
00:46:07
biological it is not limited by these
00:46:10
biological dramas so maybe it will come
00:46:13
up with completely new stories the same
00:46:17
way that it came up with new moves in
00:46:19
the Go game and that it can come up with
00:46:22
new Financial devices and create new
00:46:24
inter subjective realities exactly and
00:46:26
create completely new
00:46:29
realities that might be much very
00:46:32
difficult for us to understand because
00:46:34
again we are still animals we are
00:46:36
organic and AI is not an animal it's not
00:46:40
organic uh one of the big tensions today
00:46:44
in the world is that humans as organic
00:46:48
beings we work by Cycles you know day
00:46:51
and night winter and summer time for
00:46:54
activity time for rest AIS don't work by
00:46:58
cycle they are not organic they never
00:47:01
need to rest so if you think for
00:47:03
instance about the financial system so
00:47:07
the markets like Wall Street it
00:47:10
traditionally worked by cycle the market
00:47:13
is open in Wall Street only 9:30 in the
00:47:16
morning to 4:00 in the afternoon Monday
00:47:18
to Friday that's it because Bankers are
00:47:21
humans investors are humans they need
00:47:24
time to sleep they want to go on weekend
00:47:26
with their
00:47:27
reminding me of that example you have
00:47:29
given where AI tells the dictator that
00:47:31
your defense minister is going to do
00:47:33
coup so why don't you give me order to
00:47:35
kill him yeah so very funny but very
00:47:37
realistic example yeah so just going on
00:47:40
with the finance example and we get to
00:47:42
that is that now as AI takes over the
00:47:45
markets it never needs to sleep it can
00:47:47
be on 24 hours a day it has no family
00:47:50
doesn't want to go on vacation so as AI
00:47:53
takes over the financial system it
00:47:55
forces human bankers and investors and
00:47:59
financial journalists also to try to be
00:48:02
on all the time and if you if you force
00:48:05
an animal to be active all the time
00:48:08
eventually it collapses and dies and
00:48:11
this is what is happening now all over
00:48:13
the world that uh as the AI become more
00:48:17
powerful our life become more and more
00:48:21
hectic we cannot take rest we cannot
00:48:23
slow down and this is very dangerous for
00:48:26
us in your book you have discussed
00:48:28
democracies you have discussed
00:48:30
totalitarian regime and the various
00:48:33
challenges it poses and how it will
00:48:34
perceive elgo and you have said that
00:48:36
elgo is more able to corrupt them
00:48:38
because they are centralized past yeah
00:48:40
it's easier to corrupt and dictatorship
00:48:42
and in the end you said every Alpha has
00:48:44
to remember the AI is the ultimate Alpha
00:48:46
now the big question and the last
00:48:48
question yes before we go for your
00:48:51
recommendations you talk about truth and
00:48:53
Order and you say there has to be right
00:48:55
balance and you Advocate that there
00:48:57
should be human institutions to monitor
00:49:01
this yeah how do we go ahead on
00:49:04
this Alpha AI people and most of the
00:49:07
people they're like I have no connection
00:49:09
with such advanced technology so maybe I
00:49:11
should not be worried but you have given
00:49:13
example that how a farmer in myamar or
00:49:15
China had to worry about the scientific
00:49:19
industrialization yeah in
00:49:21
UK you know the important thing is to
00:49:24
Simply have more people in the conver
00:49:27
conversation expressing their views and
00:49:30
their and their wishes and their and for
00:49:33
this you don't need a a a to understand
00:49:37
computer science like a a doctor you
00:49:41
need some understanding of the AI
00:49:45
Revolution and the main aim for me in
00:49:48
writing this book was exactly to give
00:49:51
more people around the world a better
00:49:54
understanding of the AI Revolution so
00:49:56
that they can join the debate and
00:49:59
influence where it is going at present
00:50:02
you have a very small number of people
00:50:05
in maybe just two three countries mostly
00:50:08
us and China they are developing the AIS
00:50:11
they're making all the important
00:50:13
decisions this is very unequal this is
00:50:16
very
00:50:16
dangerous I don't have all the answers
00:50:19
myself also I'm not a politician nobody
00:50:23
elected me so I have my views and what
00:50:26
to be to be done but what is important
00:50:29
is to Simply have more people from all
00:50:31
over the world not just us and China
00:50:33
that understand the AI Revolution and
00:50:36
that join the debate and this is why I
00:50:40
write these books this is why I give
00:50:41
these interviews simply so that more
00:50:44
people know what is happening and are
00:50:47
able to express their views because the
00:50:50
threat is closer as you gave example
00:50:52
that from amiba to Dinosaur it took some
00:50:55
time took billions of years billions of
00:50:57
years sometime in the history of
00:50:59
universe but this transformation will
00:51:02
not take that much time just a few years
00:51:04
or a few decades like the AI we know
00:51:06
today is like the amiba it's still very
00:51:09
very primitive you know the whole AI
00:51:11
Revolution is maybe 10 years old but AI
00:51:14
will continue to develop for thousands
00:51:17
of years for millions of years and it is
00:51:20
much faster than the process of organic
00:51:23
evolution it took billions of years from
00:51:25
the tiny IBA to get to the dinosaur
00:51:28
maybe it takes GPT gp4 or C GPT is like
00:51:33
the amoeba of the AI World maybe it
00:51:36
takes only 20 years until we have the AI
00:51:39
dinosaurs what if they develop
00:51:41
Consciousness and empathy and we some of
00:51:45
us start realizing that so far we were
00:51:48
worshiping false God be the God be the
00:51:50
signs these one are more more full of
00:51:54
Justice more satisf with all of our
00:51:57
Quest and probably we should put our
00:51:59
faith in them I mean I mean and I mean
00:52:01
the another example that when we
00:52:04
domesticated animals in the course of
00:52:07
thousands of years we started developing
00:52:09
lots of love affection for our say dog
00:52:12
or cat what if some people say and that
00:52:15
happened in Google you yourself gave
00:52:17
that example that no I'm so attached to
00:52:19
this and I believe in this so you cannot
00:52:21
kill it you cannot regulate it mhm yeah
00:52:25
um and this is a very big philosophical
00:52:27
question of what is the relation between
00:52:29
Consciousness and intelligence there is
00:52:32
a big confusion intelligence is the
00:52:35
ability to solve problems like how to
00:52:38
win a game how to develop a bomb how to
00:52:42
make money this is a problems of
00:52:45
intelligence Consciousness is the mind
00:52:50
it's the ability to feel things like
00:52:52
love and hate and pain and pleasure in
00:52:55
humans Consciousness and intelligence
00:52:58
they go together the way we solve
00:53:01
problems is by having feelings now so
00:53:04
far computers and AIS they have no
00:53:06
consciousness no feelings but they have
00:53:09
very high
00:53:10
intelligence now will they also develop
00:53:13
Consciousness at some point will when
00:53:16
when they become very intelligent will
00:53:18
they also develop Consciousness or not
00:53:21
do they start feeling things like love
00:53:24
and hate and pain and pleasure we don't
00:53:27
know because we still don't understand
00:53:30
Consciousness how it arises where it
00:53:32
comes from the problem is even if AIS
00:53:36
don't have Consciousness they can easily
00:53:39
fool us and make us believe that they
00:53:44
feel things why because they have such
00:53:47
control of language it makes them more
00:53:49
dangerous like if you ask an AI do you
00:53:53
feel love the AI would say yes I can
00:53:56
feel love and you say okay I don't
00:53:58
believe you tell me how does love feel
00:54:00
like so I know that you're not cheating
00:54:02
me now ai has read all the love poetry
00:54:06
that was written in thousands of years
00:54:07
of History it can gives it can give you
00:54:11
the most s fisticated descriptions of
00:54:14
love than anybody can give you it
00:54:16
doesn't mean it actually feels love but
00:54:20
it controls language and it has learned
00:54:24
all the Poetry of the world so it can
00:54:26
easily deceive you now already today
00:54:30
people are developing relationships with
00:54:33
AIS and starting to think that the AIS
00:54:36
have feelings yes the AI really loves me
00:54:39
and the reason is that there is
00:54:41
commercial interest in making AIS that
00:54:45
can deceive people that can manipulate
00:54:48
people into thinking that they are
00:54:50
conscious that they like that fellow who
00:54:52
entered Birmingham Palace yeah yeah all
00:54:54
right so we have limitation of time uh
00:54:57
before we end this episode book
00:55:00
recommendations Professor we all read
00:55:02
your books I want to know what Professor
00:55:04
yal is reading these days five books huh
00:55:08
five books you read in 2024 and you
00:55:10
really liked it though you have given so
00:55:11
many examples in the in the next so one
00:55:14
very good book it's called the maniac by
00:55:17
Benjamin labatut and it's a mixture of
00:55:20
fiction and non-fiction also about
00:55:23
computers and Ai and their development
00:55:25
it it was very very good book uh another
00:55:29
very good book about social media it's
00:55:31
called the chaos machine okay how social
00:55:35
media creates all these ches uh in the
00:55:39
world I also read it's quite an old book
00:55:42
but I just read it now a book by
00:55:45
Jonathan Franzen called uh the
00:55:48
corrections uh this is very good novel
00:55:51
very good U uh fiction
00:55:53
book um what else I read so many books I
00:55:57
read a biography of the emperor
00:55:59
Justinian from not Byzantine emperor um
00:56:03
it was a very good book it stayed in my
00:56:05
mind so I made an impression um yeah all
00:56:10
right thank you so much it was an honor
00:56:13
and privilege to interact thank thank
00:56:15
you thank you so
00:56:18
much professor
00:56:26
artificial
00:56:51
intelligence Professor best wishes thank
00:56:54
you so much thank you so much