FULL DISCUSSION: Yuval Noah Harari Warns AI Will Take Over Language, Law, and Power at WEF | AI1G

historians and philosophers Yuval Noah

Harrari. He is a distinguished research

fellow at the University of Cambridge at

the center for the study of existential

risk. He has been a lecturer in the

department of history at the Hebrew

University of Jerusalem and he is

co-founder of Sapenship.

As many of you will know, he is a

best-selling author of, amongst many

books, Sapiens, a brief history of

humankind, Homodus, Brief History of

Tomorrow, and 21 Lessons for the 21st

century, amongst others, selling over 50

million books worldwide in 65 languages.

He focuses on the macrohistorical

questions of our time. And what a

perfect moment with this pressing

arrival and disruption of AI to have

somebody of Yaval's distinction take on

this challenge. Please join me in warmly

welcoming Yuval Noah Harrari to deliver

a conversation about AI and humanity.

So hello everyone.

There is one question that every leader

today must answer about AI. But to

understand that question, we first need

to clarify a few points about what AI is

and what AI can do.

The most important thing to know about

AI is that it is not just another tool.

It is an agent. It can learn and change

by itself and make decisions by itself.

A knife is a tool. You can use a knife

to cut salad or to murder someone, but

it is your decision what to do with the

knife.

AI is a knife that can decide by itself

whether to cut salad or to commit

murder.

The second thing to know about AI is

that it can be a very creative agent.

AI is a knife that can invent new kinds

of knives as well as new kinds of music,

medicine, and money.

The third thing to know about AI is that

it can lie and manipulate.

4 billion years of evolution have

demonstrated that anything that wants to

survive learns to lie and manipulate.

The last four years have demonstrated

that AI agents can acquire the will to

survive and that AIs have already

learned how to lie.

Now, one big open question about AI

is whether it can think.

Modern philosophy began in the 17th

century when Rene proclaimed I think

therefore I am.

Even before the cart we humans defined

ourselves by our capacity to think. We

believe our we rule the world because we

can think better than anyone else on

this planet.

Will AI challenge our supremacy in the

field of thinking? Now that depends on

what thinking means.

Try to observe yourself thinking. What

is happening there?

Many people observe words popping in

their mind and forming sentences and the

sentences then forming arguments like

all humans are mortal. I am human

therefore I am mortal.

If thinking really means putting words

and other language tokens in order, then

AI can already think much better than

many many humans. AI can certainly come

up with a sentence like AI thinks,

therefore AI.

Some people argue that AI is just

glorified autocomplete. It barely

predicts the next word in a sense in a

sentence.

But is that so different from what the

human mind is doing?

Try to observe to catch the next word

that pops up in your mind. Do you really

know why you thought that word where it

came from? Why do you did you think this

particular word and not some other word?

Do you know? As far as putting words in

order is concerned,

AI already thinks better than many of

us. Therefore, anything made of words

will be taken over by AI. If laws are

made of words, then AI will take over

the legal system. If books are just

combinations of words, then AI will take

over books. If religion is built from

words, then AI will take over religion.

This is particularly true of religions

based on books like Islam, Christianity

or Judaism.

Judaism called itself the religion of

the book and it grants ultimate

authority not to humans but to words in

books.

Humans have authority in Judaism not

because of our experiences but only

because we learn words in books. Now no

human can read and remember all the

words in all the Jewish books but AI can

easily do that.

What happens to a religion of the book

when the greatest expert on the holy

book is an AI?

However, some some people may say, can

we really reduce human spirituality

to just words in books? Does thinking

mean only putting language tokens in

order?

If you observe yourself carefully when

you're thinking, you will notice that

something else is happening there

besides words popping in your mind and

forming sentences.

You also have some nonverbal feelings.

Maybe you feel pain.

Maybe you feel fear. Maybe love.

Some thoughts are painful. Some are

frightening. Some are full of love.

While AIs become better than us with

words,

at least for now, we have zero evidence

that AIs can feel anything. Of course,

because AI is mastering language,

AI can pretend to feel pain or love. AI

can say, "I love you." And if you

challenge it to describe how love feels,

AI can provide the best verbal

description in the world.

AI can read countless love poems and

psychology books and can then describe

the feeling of love much better than any

human poet, psychologist or lover. But

these are just words.

The Bible says in the beginning was the

word and the word was made flesh.

The taqing says the truth that can be

expressed in words is not the absolute

truth. Throughout history, people have

always struggled with the tension

between word and flesh, between the

truth that can be expressed in words and

the absolute truth which is beyond

words.

Previously this tension was internal to

humanity. It was between different human

groups. Some humans gave supreme

importance to words. They've been

willing, for example, to abandon or even

kill their gay son just because of a few

words in the Bible.

Other humans have said, "But these are

just words. The spirit of love should be

much more important than the letter of

the law."

This tension between spirit and letter

existed in every religion, every legal

system, even every person.

Now this tension will be externalized.

It will become the tension not between

different humans. It this will be the

tension

between humans and AIs, the new masters

of words. Everything made of words will

be taken over by AI.

Previously all the words, all our verbal

thoughts, they originated in some human

mind. Either my mind I thought this or I

learned it from another human. Soon most

of the words in our minds will originate

in a machine. I just heard today about a

new word that AIS coined by themselves

to describe us humans.

They called us the watchers.

The watchers that we are watching them.

AIs will soon be the origin of maybe

most of the words in our minds.

AIS will mass produce thoughts by

assembling words, symbols, images, and

other language tokens into new

combinations.

Whether humans will still have a place

in that world depends on the place we

assign our nonverbal feelings and our

ability to embody wisdom that cannot be

expressed in words. If we continue to

define ourselves by our ability to think

in words, our identity will collapse.

All this means that no matter from which

country you come, your country will soon

face a severe identity crisis and also

an immigration crisis.

The immigrants this time will not be

human beings coming in fragile boats

without a visa or trying to cross a

border in the middle of the night. The

immigrants will be millions of AIs that

can ride love poles better than us, that

can lie better than us, and that can

travel at the speed of light without any

need of visas.

Like human immigrants, these AI

immigrants will bring various benefits

with them. We will have AI doctors to

help in our health care systems, AI

teachers to help in our education

systems, even AI border guards to stop

illegal human immigrants. But the AI

immigrants will also bring with them

problems.

Those who are concerned about human

immigrants usually argue that immigrants

might take jobs, might change the local

culture, might be politically disloyal.

I'm not sure that's true of all human

immigrants, but it will definitely be

true of the AI immigrants.

The AI immigrants will take many human

jobs. The AI immigrants will completely

change the culture of every country.

They will change out religion and even

romance. Some people don't like it if

their son or daughter is dating an

immigrant boyfriend.

What would these people think when their

son or daughter starts dating an AI

boyfriend?

And of course, the AI immigrants will

have some dubious political loyalties.

They are likely to be loyal not to your

country but but to some corporation or

government across the ocean most

probably in one of only two countries,

China or the USA.

The USA encourages countries to close

their borders to human immigrants but

open them very very wide to US AI

immigrants.

And now we can finally come to the

question each one of you must soon

answer.

Will your country recognize the AI

immigrants as legal persons?

AIS are obviously not persons. They

don't have a body or a mind. But a legal

person is something quite different from

a person. A legal person is an entity

that the law recognizes as having

certain legal obligations and rights.

For example, the right to hold property,

to file a lawsuit, and to enjoy freedom

of speech.

In many countries, corporations are

considered legal persons. The Alphabet

Corporation can open a bank account, can

sue you in court, or can donate to your

next presidential campaign.

In New Zealand, rivers have been

recognized as legal persons. In India,

certain gods have been granted such

recognition. Of course, until today,

recognizing a corporation, a river, or a

god as a legal person was just legal

fiction.

In practice, if a corporation like

Alphabet decided to buy another

corporation or if a Hindu god,

if a Hindu god decided to sue you in

court, the decision wasn't really made

by the god. It was made by some human

executives, shareholders or trustees.

It is different with AIS. Unlike rivers

and gods, AIs can actually make

decisions by themselves. They will soon

be able to make the decisions necessary

to manage a bank account, to file a

lawsuit, and even to operate a

corporation without any need of human

executives, shareholders or trustees.

AIS can therefore function as persons.

Do we want to allow that? Will your

country recognize AIS as legal persons?

What if other countries do it?

Suppose your country doesn't want to

recognize AIS as persons. But the USA in

the name of deregulating AI and

deregulating the markets grants legal

recognition, legal personhood to

millions of AIs which start running

millions of new corporations.

Will you block these US AI corporations

from operating in your country?

Suppose some USI persons invent super

efficient and super complex financial

devices that humans cannot fully

understand and therefore don't know how

to regulate. Will you open your

financial markets to this new AI

financial wizardry

or will you try to block it thereby

decoupling from the American financial

system?

Suppose some AI persons create a new

religion which gains the faith of

millions of people. That should not

sound too far-fetched because after all,

almost all previous religions in history

have claimed that they were created by a

nonhuman intelligence. Now, will your

country extend freedom of religion to

the new AI sect and to its AI priests

and missionaries?

Maybe we should start with something a

bit simpler. Will your country allow AI

persons to open social media accounts,

enjoy freedom of speech, on Facebook, on

Tik Tok, and be friendly with children?

Well, of course, that question should

have been asked 10 years ago. On social

media, AI bots have been operating as

functional persons for at least a

decade. If you think AIS should not be

treated as persons on social media, you

should have acted 10 years ago.

10 years from now, it will be too late

for you to decide whether AIs should

function as persons in the financial

markets, in the courts, in the churches.

Somebody else will already have decided

it for you.

If you want to influence where humanity

is going, you need to make a decision

now.

So what is your answer as a leader? Do

you think the AI immigrants should be

recognized as legal persons? If not, how

are you going to stop that?

Thank you for listening to this human.

[applause]

Thank you, Yaval. That was fantastic

overview. You posed a lot of questions

um and they're the right ones. I agree

with much of what you say. We're here in

Davos where the theme is around dialogue

and I was struck by your commentary

around words and the importance of words

and that being something that demarcates

human animals from other animals

although that's debatable that there's

other language there. So in the context

of Davos and the range of people we have

here from technology from the business

world from politicians

>> how would you like to see what is the

answer that you have in terms of this

slightly dystopian world you've

potentially put in front of us

>> and if I may just add to that I think

it's fair to say I'm a scientist by

background a neuroscientist so I uh work

a lot in this space particularly around

pain and we're very comfortable with the

fact that many of our discoveries,

particularly technological discoveries,

we often drive them forward and then

afterwards we think, oh, we hadn't

thought enough about the ethics and the

implications and then we're trying to

catch up on the regulation that we need

to maybe put around it.

>> So, we are where we are. This thing is

happening as everybody says at scale

both in terms of its magnitude and its

pace more than we've ever seen before in

the industrial revolution. We have all

the right blend of people here in Davos.

It's all about dialogue. What would you

like to see go forward in terms of

putting boundaries around some of the

slightly more worrying areas that you

detailed? And what are your own thoughts

about the ethical implications of giving

legal rights to either agents to robots

or to the ones that are just exist on

the internet?

>> A lot of things there. I mean first of

all I I would say that you know Davos is

about words.

It's about talking. The basic idea of

Davos is that you can change the world

just by talking which I like this idea

because this is also my idea as an

author as a university lecturer. This is

what I do. I talk I write I think I can

influence the world the world with

words.

Um and this is now in in question. Are

we at the end of the road for words?

Um, is this no longer a function?

And

you know, engineers and also soldiers,

they don't change the world with words.

They do stuff. They take action.

Um, philosophers,

scholars, also political leaders,

they try to change the world with words,

by saying things. And maybe we've

reached the end of that road. And what

does it mean? But we know we we con we

humans, we conquer the world ultimately,

I would say, with language and words.

Because yes, engineers can make weapons

and soldiers can wield them. But to

build an army, you need to convince

thousands of strangers to cooperate. How

do you do that? With words, with

ideology, with religion.

So, humans took over the world not

because we are the strongest physically,

but because we discovered how to use

words to get thousands and millions and

billions of strangers to cooperate.

This was our superpower. And now

something has emerged that is going to

take our superpower from us. Until a few

years ago, nothing on earth could use

words. Only humans, chimpanzees

couldn't, rivers couldn't, the sun

couldn't, we could use words.

Now there is something that is able or

soon will be able to use words better

than us. And you look just, you know, at

what happened on social media and the

immense change it brought to the world

there.

So 10 years from now living in a world

in which AIs are in command of language

how does that look like?

>> Well Davos in 10 years might look very

different as you say. So that's a future

we can all try to think about in the

context of who would be here beyond the

physical human. But if I may just um

discuss a little bit around the fact

that it's not new for humans to be uh

beaten by technology. So if we think

about some of the tech, we can't fly and

we built airplanes. Cars can go faster

than us. We're very comfortable with

that. The threat that comes with AI is

the fact that it's a threat to the

sovereign power of our ability to think.

And that is destabilizing. I say that as

an academic and an educator. That's

something that is very threatening. But

if we go back to say a robot, the value

we would place on a robot being able to

run the 100 meters faster than Usain

Bolt is less. There's something about

the human endeavor, the struggle, the

suffering, the fact that we can have a

collective sense of empathy and

understanding about what it meant to

achieve something even if it was lesser

with technology. I just wonder whether

an author that would replace you, how

much as a human we would value that the

words of that the creativity that comes

from art that's been done with

artificial intelligence.

>> Do you think we will value it as much

and therefore there's still a place for

humans in the creative space of thinking

and words?

>> That's the identity crisis because the

cow didn't say I run therefore I am. I

think I mean it based human identity on

our capacity to think. We always knew

that cheetahs can run faster than us. We

always knew that elephants are much

bigger and stronger than us. So we

didn't define ourselves by this. We

defined ourselves by thinking. And now

something is going to be better than us

in thinking. If thinking means putting

words in order. Now I'm again I'm an

author. I am a speaker. I put words in

order. This is like this is my game.

Like I have all these words and I put oh

let's put these words in this in this

order. No, no, no, no. It will be better

to put it like this like this like this

and AI will beat me. I don't know how

long it will take two years, five years,

10 years it will beat me and then what

does it mean for our identity? People

identify, you know, with the streams of

of words in in in their mind. Like you

close your eyes, you try to see what's

happening inside me. Many people, I'm

one of them. We see words popping up

organizing themselves. We identify with

that.

>> But I guess my point is using the same

analogy is that we still value a human.

We have the Olympics. We've got the

Winter Olympics coming.

>> We know that many other animals and

other technologies can outperform in

many of those areas. Yet we still really

enjoy the humanity of people that train

and develop even though it's not as

good. And I just wonder whether we will

just naturally extend that to the

thinking realm and to the word so that

you still will have a very very vibrant

and successful author role in 10 years

time and

>> I don't know like

>> because I will value your book more than

an AI generated book

>> even if even if the AI comes up with new

ideas better than me like let's say that

you want to invest money and you have a

you ask a human consultant and she comes

up with a certain whatever and you can

empathize with her because she had this

life story and whatever. And then you

have the AI financial consultant with

zero life story, zero emotions but

better financial advice.

Which one will you follow?

>> Now we always have this kind of I think

the big mistake and this is why I

started with the idea of agency. We

always think that uh um we can just use

these things as tools.

>> But if they can think

>> they are agents.

>> Yeah. You know, maybe I I'll tell a

story from medieval history that how did

the Anglo-Saxons took over Britain and

it's part myth about history that you

know the Britain who lived there

originally they were fighting with the

PS and the Scots coming from the north

and the Britain didn't fight very well.

So the king of the Britain Vodiger he

said I have an idea. I've heard that in

Germany, in Scandinavia, these people

really know how to fight. So, let's

bring over some mercenary, some

Anglo-Saxon mercenary. They will fight

for us. They will defeat the PS and the

Scots. And Vodigern brings over

Anglo-Saxon mercenaries and they fight

well and they defeat the Scots and the

PS. But then the Anglo-Saxons says say

to themselves, this is a rich country

and these people they are very weak and

these people they are disunited. we can

take over

>> and they take over. We understand this

with human mercenaries. We understand

that when you bring a human mercenary,

okay, you pay them, but they have a mind

of their own. Maybe they rebel.

>> We don't get it with AIS.

>> Yeah.

>> You know, you look at the leader of of

the world, they think, "Oh, I'll bring

AI to fight my war for me."

the idea that it can just take power

away from you.

>> Yeah.

>> It doesn't really cross their mind. They

don't really accept that AIS can think.

>> Yeah. Yeah. And that is very

fundamentally different. So just to

reverse that, you're an alum of my

institution. Uh and and we're proud of

that. Um although you're working at the

other place now in Cambridge. Um so the

challenge I think for the education

sector and it goes back to a reverse

flip of what Alan Turing said which was

whether a computer could think the sort

of birthplace of of artificial

intelligence if you will. So I think the

question we've been posing inside the

academic sector is how do we keep humans

thinking

>> because if we keep abdicating our

decision making our financial decisions

or whatever it might be increasingly

increasingly to AI the worry we have

quite quickly and we're seeing this with

students coming to us through the school

system very overusing chat GPT is the

deskkilling of critical faculties of

human brain thinking. So what's your

advice to me in the academic sector

about how can we hang in there as humans

and keep humans thinking so that we at

least have some capacity to live

alongside these technologies which as

you say bring us into a very different

place going forward in terms of world

order.

>> Yeah. At the present moment we still

think better.

>> So at the present moment it's kind of

telling people you you need your

critical thinking, you need moral

evaluations. You cannot get that from

AI. But we need to prepare to the moment

when this is no longer the case. We need

to prepare for the moment. Let's say

again take economics or finance when AIs

create a new financial systems, a new

financial system that they understand

and we don't understand. How do you

train economists or politicians in a

world in which humans really can't can

no longer understand how finance

functions

because AIs have created this super

complex financial system that we are

like the horses you know that horses can

see that they are being traded from one

human to another for a few shiny gold

coins they can't understand this idea of

money too complicated M

>> we can be in the same situation 10 years

from now. Davos 10 years from now.

>> Maybe nobody in the room, no human in

the room understands the financial

system anymore because it's dominated by

AIs and the AIs have come up with new

financial strategies and devices

that are just mathematically beyond the

human capacity of of the brain. So how

does politics and finance and davos look

like in a world when no human beings

understands finance anymore?

>> Yeah. No. Well, that's a beautiful note

to finish on. We've run out of time.

There are many more questions that we

could explore. One of which just being

the major difference uh that we know

about artificial intelligence compared

to human intelligence is of course the

human brain develops from birth to

adulthood around age 20. And it is a

product of your life experience as a

sentient human being feeling loving

anger these emotions. And whilst one can

improvise a little bit with sensory

detectors and you can train the brain to

do that that is fundamentally different.

So the artificial brain is not a human

brain. It is not human. And there is

maybe something that is still of value

there that goes back to that core

business of this sentient human being

that brings value to our understanding.

And maybe one last comment

>> please

>> think about educating kids in a world

where from day zero maybe the most of

the interaction of the new child is with

an AI.

>> Yeah.

>> And not with a human being.

>> Yeah.

>> It's the biggest and scariest

psychological experiment in history and

we are conducting it.

>> Indeed we are. Well, you thank you so

much. I'm delighted that you're thinking

about these problems and that you've got

us all thinking this afternoon. I look

forward to you coming back maybe to

Davos in in uh 10 years and reflecting

on this conversation and just where we

have got to. But thank you all to the

audience, those of you online and those

in the group. And thank you. Can we give

a round of applause for your

>> [music]