iHuman (2019) - full transcript

- Intelligence is
the ability to understand.

We passed on what we know to machines.

- The rise of
artificial intelligence

is happening fast, but some
fear the new technology

might have more problems than anticipated.

- We will not control it.

- Artificially
intelligent algorithms are here,

but this is only the beginning.

- In the age of AI,
data is the new oil.

- Today, Amazon,
Google and Facebook

are richer and more
powerful than any companies



that have ever existed
throughout human history.

- A handful of people working at
a handful of technology companies

steer what a billion
people are thinking today.

- This technology is changing:
What does it mean to be human?

- Artificial intelligence is simply
non-biological intelligence.

And intelligence itself is simply
the ability to accomplish goals.

I'm convinced that AI
will ultimately be either

the best thing ever to happen to humanity,
or the worst thing ever to happen.

We can use it to solve all
of today's and tomorrow's

greatest problems; cure diseases,
deal with climate change,

lift everybody out of poverty.

But, we could use exactly
the same technology

to create a brutal global
dictatorship with unprecedented

surveillance and inequality and suffering.



That's why this is the most important
conversation of our time.

- Artificial intelligence is everywhere

because we now have thinking machines.

If you go on social media or online,

there's an artificial intelligence engine
that decides what to recommend.

If you go on Facebook and you're just
scrolling through your friends' posts,-

there's an AI engine that's picking
which one to show you first

- and which one to bury.

If you try to get insurance,
there is an AI engine

trying to figure out how risky you are.

And if you apply for a job,
it's quite possible

that an AI engine looks at the resume.

- We are made of data.

Every one of us is made of data
-in terms of how we behave,

how we talk, how we love,
what we do every day.

So, computer scientists are
developing deep learning

algorithms that can learn
to identify, classify,

and predict patterns within
massive amounts of data.

We are facing a form of
precision surveillance,

you could call it algorithmic surveillance,

and it means that you
cannot go unrecognized.

You are always under
the watch of algorithms.

- Almost all the AI
development on the planet today

is done by a handful of
big technology companies

or by a few large governments.

If we look at what AI is
mostly being developed for,

I would say it's killing,
spying, and brainwashing.

So, I mean, we have military AI,

we have a whole surveillance
apparatus being built

using AI by major governments,

and we have an advertising
industry which is oriented

toward recognizing what ads
to try to sell to someone.

- We humans have come to
a fork in the road now.

The AI we have today is very narrow.

The holy grail of AI research
ever since the beginning

is to make AI that can do
everything better than us,

and we've basically built a God.

It's going to revolutionize
life as we know it.

It's incredibly important
to take a step back

and think carefully about this.

What sort of society do we want?

- So, we're in this
historic transformation.

Like we're raising this new creature.

We have a new offspring of sorts.

But just like actual offspring,

you don't get to control
everything it's going to do.

We are living at
this privileged moment where,

for the first time, we
will see probably that AI

is really going to outcompete
humans in many, many,

if not all, important fields.

- Everything is going to change.

A new form of life is emerging.

When I was a boy, I thought,
how can I maximize my impact?

And then it was clear that
I have to build something

that learns to become smarter than myself,

such that I can retire,

and the smarter thing
can further self-improve

and solve all the problems
that I cannot solve.

Multiplying that tiny
little bit of creativity

that I have into infinity,

and that's what has been
driving me since then.

How am I trying to build

a general purpose artificial intelligence?

If you want to be intelligent,
you have to recognize speech,

video and handwriting, and
faces, and all kinds of things,

and there we have made a lot of progress.

See, LSTM, neural networks,

which we developed in our labs
in Munich and in Switzerland,

and it's now used for speech
recognition and translation,

and video recognition.

They are now in everybody's
smartphone, almost one billion

iPhones and in over
two billion Android phones.

So, we are generating all
kinds of useful by-products

on the way to the general goal.

The main goal, some Artificial
General Intelligence,

an AGI that can learn to improve
the learning algorithm itself.

So, it basically can learn
to improve the way it learns

and it can also recursively
improve the way it learns,

the way it learns without
any limitations except for

the basic fundamental
limitations of computability.

One of my favorite
robots is this one here.

We use this robot for our
studies of artificial curiosity.

Where we are trying to teach
this robot to teach itself.

What is a baby doing?

A baby is curiously exploring its world.

That's how he learns how gravity works

and how certain things topple, and so on.

And as it learns to ask
questions about the world,

and as it learns to
answer these questions,

it becomes a more and more
general problem solver.

And so, our artificial
systems are also learning

to ask all kinds of
questions, not just slavishly

try to answer the questions
given to them by humans.

You have to give AI the freedom
to invent its own tasks.

If you don't do that, it's not
going to become very smart.

On the other hand, it's really hard
to predict what they are going to do.

- I feel that technology
is a force of nature.

I feel like there is a lot of similarity between
technology and biological evolution.

Playing God.

Scientists have been accused
of playing God for a while,

- but there is a real sense in
which we are creating something

very different from anything
we've created so far.

I was interested in the concept of
AI from a relatively early age.

At some point, I got especially
interested in machine learning.

What is experience?

What is learning?

What is thinking?

How does the brain work?

These questions are philosophical,

but it looks like we can
come up with algorithms that

both do useful things and help
us answer these questions.

Like it's almost like applied philosophy.

Artificial General Intelligence, AGI.

A computer system that
can do any job or any task

that a human does, but only better.

Yeah, I mean, we definitely
will be able to create

completely autonomous
beings with their own goals.

And it will be very important,

especially as these beings
become much smarter than humans,

it's going to be important
to have these beings,

that the goals of these beings
be aligned with our goals.

That's what we're trying
to do at OpenAI.

Be at the forefront of research
and steer the research,

steer their initial conditions
so to maximize the chance

that the future will be good for humans.

Now, AI is a great thing
because AI will solve

all the problems that we have today.

It will solve employment,
it will solve disease,

it will solve poverty,

but it will also create new problems.

I think that...

The problem of fake news
is going to be a thousand,

a million times worse.

Cyberattacks will
become much more extreme.

You will have totally
automated AI weapons.

I think AI has the potential to create
infinitely stable dictatorships.

You're gonna see dramatically
more intelligent systems

in 10 or 15 years from now,
and I think it's highly likely

that those systems will have
completely astronomical impact on society.

Will humans actually benefit?

And who will benefit, who will not?

- In 2012, IBM estimated that

an average person is
generating 500 megabytes

of digital footprints every single day.

Imagine that you wanted
to back-up one day worth

of data that humanity is
leaving behind, on paper.

How tall will be the stack
of paper that contains

just one day worth of data
that humanity is producing?

It's like from the earth
to the sun, four times over.

In 2025, we'll be generating
62 gigabytes of data

per person, per day.

We're leaving a ton of digital footprints
while going through our lives.

They provide computer algorithms
with a fairly good idea about who we are,

what we want, what we are doing.

In my work, I looked at different
types of digital footprints.

I looked at Facebook likes,
I looked at language,

credit card records, web browsing
histories, search records.

and each time I found that if
you get enough of this data,

you can accurately predict future behavior

and reveal important intimate traits.

This can be used in great ways,

but it can also be used
to manipulate people.

Facebook is delivering daily information

to two billion people or more.

If you slightly change the
functioning of the Facebook engine,

you can move the opinions and hence,

the votes of millions of people.

- Brexit!
- When do we want it?

- Now!

- A politician
wouldn't be able to figure out

which message each one of
his or her voters would like,

but a computer can see
what political message

would be particularly convincing for you.

- Ladies and gentlemen,

it's my privilege to speak
to you today about the power

of big data and psychographics
in the electoral process.

- Data from Cambridge Analytica

secretly harvested the
personal information

of 50 million unsuspecting Facebook users.

USA!

- The data firm hired by Donald Trump's
presidential election campaign

used secretly obtained information
to directly target potential American voters.

- With that, they say they can predict

the personality of every single
adult in the United States.

- Tonight we're hearing
from Cambridge Analytica

whistleblower, Christopher Wiley.

- What we worked on was
data harvesting programs

where we would pull data and run that
data through algorithms that could profile

their personality traits and
other psychological attributes

to exploit mental vulnerabilities
that our algorithms showed that they had.

- Cambridge Analytica mentioned once

or said that their models
were based on my work,

but Cambridge Analytica is
just one of the hundreds

of companies that are using
such methods to target voters.

You know, I would be asked
questions by journalists such as,

"So how do you feel about

"electing Trump and supporting Brexit?"

How do you answer to such question?

I guess that I have to deal
with being blamed for all of it.

- How tech started was
as a democratizing force,

as a force for good, as
an ability for humans

to interact with each
other without gatekeepers.

There's never been a bigger experiment
in communications for the human race.

What happens when everybody
gets to have their say?

You would assume that it
would be for the better,

that there would be more democracy,
there would be more discussion,

there would be more tolerance,
but what's happened is that

these systems have been hijacked.

- We stand for connecting every person.

For a global community.

- One company,
Facebook, is responsible

for the communications of
a lot of the human race.

Same thing with Google.

Everything you want know about
the world comes from them.

This is global information economy

that is controlled by a
small group of people.

- The world's richest companies
are all technology companies.

Google, Apple, Microsoft,
Amazon, Facebook.

It's staggering how,

in probably just 10 years,

that the entire corporate power structure

are basically in the business
of trading electrons.

These little bits and bytes
are really the new currency.

- The way that data is monetized

is happening all around us,
even if it's invisible to us.

Google has every amount
of information available.

They track people by their GPS location.

They know exactly what your
search history has been.

They know your political preferences.

Your search history alone can tell you

everything about an individual
from their health problems

to their sexual preferences.

So, Google's reach is unlimited.

- So we've seen Google and Facebook

rise into these large
surveillance machines

and they're both actually ad brokers.

It sounds really mundane,
but they're high tech ad brokers.

And the reason they're so
profitable is that they're using

artificial intelligence to
process all this data about you,

and then to match you with the advertiser

that wants to reach people
like you, - for whatever message.

- One of the problems with technology is
that it's been developed to be addictive.

The way these companies
design these things

is in order to pull you in and engage you.

They want to become essentially
a slot machine of attention.

So you're always paying attention,
you're always jacked into the matrix,

you're always checking.

- When somebody controls what you read,
they also control what you think.

You get more of what you've
seen before and liked before,

because this gives more traffic
and that gives more ads,

but it also locks you
into your echo chamber.

And this is what leads
to this polarization that we see today.

Jair Bolsonaro!

- Jair Bolsonaro,
Brazil's right-wing

populist candidate sometimes
likened to Donald Trump,

winning the presidency Sunday
night in that country's

most polarizing election in decades.

- Bolsonaro!

- What we are seeing around the world
is upheaval and polarization and conflict

that is partially pushed by algorithms

that's figured out that
political extremes,

tribalism, and sort of
shouting for your team,

and feeling good about it, is engaging.

- Social media may
be adding to the attention

to hate crimes around the globe.

- It's about how
people can become radicalized

by living in the fever
swamps of the Internet.

- So is this a key moment for the tech giants?
Are they now prepared to take responsibility

as publishers for what
they share with the world?

- If you deploy a
powerful potent technology

at scale, and if you're talking
about Google and Facebook,

you're deploying things
at scale of billions.

If your artificial intelligence
is pushing polarization,

you have global upheaval potentially.

White lives matter!

Black lives matter!

- Artificial General Intelligence, AGI.

Imagine your smartest friend,

with 1,000 friends, just as smart,

and then run them at a 1,000
times faster than real time.

So it means that in every day of our time,

they will do three years of thinking.

Can you imagine how much you could do

if, for every day, you could
do three years' worth of work?

It wouldn't be an unfair comparison to say

that what we have right now
is even more exciting than

the quantum physicists of
the early 20th century.

They discovered nuclear power.

I feel extremely lucky to
be taking part in this.

Many machine learning experts,
who are very knowledgeable and experienced,

have a lot of skepticism about AGI.

About when it would happen,
and about whether it could happen at all.

But right now, this is something that just
not that many people have realized yet.

That the speed of computers,
for neural networks, for AI,

are going to become maybe
100,000 times faster

in a small number of years.

The entire hardware
industry for a long time

didn't really know what to do next,

but with artificial neural networks,
now that they actually work,

you have a reason to build huge computers.

You can build a brain in
silicon, it's possible.

The very first AGIs
will be basically very,

very large data centers
packed with specialized

neural network processors
working in parallel.

Compact, hot, power hungry package,

consuming like 10 million
homes' worth of energy.

A roast beef sandwich.

Yeah, something slightly different.

Just this once.

Even the very first AGIs

will be dramatically
more capable than humans.

Humans will no longer be
economically useful for nearly any task.

Why would you want to hire a human,

if you could just get a computer that's going to
do it much better and much more cheaply?

AGI is going to be like, without question,

the most important
technology in the history

of the planet by a huge margin.

It's going to be bigger
than electricity, nuclear,

and the Internet combined.

In fact, you could say
that the whole purpose

of all human science, the
purpose of computer science,

the End Game, this is the
End Game, to build this.

And it's going to be built.

It's going to be a new life form.

It's going to be...

It's going to make us obsolete.

- European manufacturers
know the Americans

have invested heavily in
the necessary hardware.

- Step into a
brave new world of power,

performance and productivity.

- All of the images you are
about to see on the large screen

will be generated by
what's in that Macintosh.

- It's my honor and
privilege to introduce to you

the Windows 95 Development Team.

- Human physical labor has
been mostly obsolete for

getting on for a century.

Routine human mental labor
is rapidly becoming obsolete

and that's why we're seeing a lot of
the middle class jobs disappearing.

- Every once in a while,

a revolutionary product comes
along that changes everything.

Today, Apple is reinventing the phone.

- Machine intelligence
is already all around us.

The list of things that we humans
can do better than machines

is actually
shrinking pretty fast.

- Driverless cars are great.

They probably will reduce accidents.

Except, alongside with
that, in the United States,

you're going to lose 10 million jobs.

What are you going to do with
10 million unemployed people?

- The risk for social
conflict and tensions,

if you exacerbate inequalities,
is very, very high.

- AGI can, by definition,

do all jobs better than we can do.

People who are saying,
"There will always be jobs

"that humans can do better
than machines," are simply

betting against science and
saying there will never be AGI.

- What we are seeing
now is like a train hurtling

down a dark tunnel at
breakneck speed and it looks like

we're sleeping at the wheel.

- A large fraction of the digital footprints
we're leaving behind are digital images.

And specifically, what's really
interesting to me as a psychologist

are digital images of our faces.

Here you can see the difference
in the facial outline

of an average gay and
an average straight face.

And you can see that straight men

have slightly broader jaws.

Gay women have slightly larger jaws,
compared with straight women.

Computer algorithms can
reveal our political views

or sexual orientation, or intelligence,

just based on the picture of our faces.

Even a human brain can distinguish between
gay and straight men with some accuracy.

Now it turns out that the computer
can do it with much higher accuracy.

What you're seeing here is an accuracy of

off-the-shelf facial recognition software.

This is terrible news

for gay men and women
all around the world.

And not only gay men and women,

because the same algorithms
can be used to detect other

intimate traits, think being
a member of the opposition,

or being a liberal, or being an atheist.

Being an atheist is
also punishable by death

in Saudi Arabia, for instance.

My mission as an academic is to
warn people about the dangers of algorithms

being able to reveal our intimate traits.

The problem is that when
people receive bad news,

they very often choose to dismiss them.

Well, it's a bit scary when you start receiving
death threats from one day to another,

and I've received quite a
few death threats, -

-but as a scientist, I have to
basically show what is possible.

So what I'm really interested
in now is to try to see

whether we can predict other
traits from people's faces.

Now, if you can detect
depression from a face,

or suicidal thoughts, maybe a CCTV system

on the train station can save some lives.

What if we could predict that someone
is more prone to commit a crime?

You probably had a school counselor,

a psychologist hired
there to identify children

that potentially may have
some behavioral problems.

So now imagine if you could
predict with high accuracy

that someone is likely to
commit a crime in the future

from the language use, from the face,

from the facial expressions,
from the likes on Facebook.

I'm not developing new
methods, I'm just describing

something or testing something
in an academic environment.

But there obviously is a chance that,

while warning people against
risks of new technologies,

I may also give some people new ideas.

- We haven't yet seen
the future in terms of

the ways in which the
new data-driven society

is going to really evolve.

The tech companies want
to get every possible bit

of information that they
can collect on everyone

to facilitate business.

The police and the military
want to do the same thing

to facilitate security.

The interests that the two
have in common are immense,

and so the extent of collaboration
between what you might

call the Military-Tech Complex
is growing dramatically.

- The CIA, for a very long time,

has maintained a close
connection with Silicon Valley.

Their venture capital
firm known as In-Q-Tel,

makes seed investments to
start-up companies developing

breakthrough technology that
the CIA hopes to deploy.

Palantir, the big data analytics firm,

one of their first seed
investments was from In-Q-Tel.

- In-Q-Tel has struck gold in Palantir

in helping to create a private vendor

that has intelligence and
artificial intelligence

capabilities that the government
can't even compete with.

- Good evening, I'm Peter Thiel.

I'm not a politician, but
neither is Donald Trump.

He is a builder and it's
time to rebuild America.

- Peter Thiel,
the founder of Palantir,

was a Donald Trump transition advisor

and a close friend and donor.

Trump was elected largely on the promise

to deport millions of immigrants.

The only way you can do that
is with a lot of intelligence

and that's where Palantir comes in.

They ingest huge troves
of data, which include,

where you live, where
you work, who you know,

who your neighbors are,
who your family is,

where you have visited, where you stay,

your social media profile.

Palantir gets all of that
and is remarkably good

at structuring it in a way
that helps law enforcement,

immigration authorities
or intelligence agencies

of any kind, track you, find you,

and learn everything there
is to know about you.

- We're putting AI in
charge now of evermore

important decisions that
affect people's lives.

Old-school AI used to have
its intelligence programmed in

by humans who understood
how it worked, but today,

powerful AI systems have
just learned for themselves,

and we have no clue really how they work,

which makes it really hard to trust them.

- This isn't some futuristic
technology, this is now.

AI might help determine
where a fire department

is built in a community
or where a school is built.

It might decide whether you get bail,

or whether you stay in jail.

It might decide where the
police are going to be.

It might decide whether you're going to be
under additional police scrutiny.

- It's popular now in the US
to do predictive policing.

So what they do is they use an algorithm
to figure out where crime will be,

- and they use that to tell where
we should send police officers.

So that's based on a
measurement of crime rate.

So we know that there is bias.

Black people and Hispanic
people are pulled over,

and stopped by the police
officers more frequently

than white people are, so we
have this biased data going in,

and then what happens
is you use that to say,

"Here's where the cops should go."

Well, the cops go to those neighborhoods,
and guess what they do, they arrest people.

And then it feeds back
biased data into the system,

and that's called a feedback loop.

- Predictive policing
leads at the extremes

to experts saying, "Show me your baby,

"and I will tell you whether
she's going to be a criminal."

Now that we can predict it,
we're going to then surveil

those kids much more closely
and we're going to jump on them

at the first sign of a problem.

And that's going to make
for more effective policing.

It does, but it's going to
make for a really grim society

and it's reinforcing
dramatically existing injustices.

- Imagine a world in which
networks of CCTV cameras,

drone surveillance
cameras, have sophisticated

face recognition technologies
and are connected

to other government
surveillance databases.

We will have the
technology in place to have

all of our movements comprehensively
tracked and recorded.

What that also means is that we will have

created a surveillance time machine

that will allow governments
and powerful corporations

to essentially hit rewind on our lives.

We might not be under any suspicion now

and five years from now,
they might want to know

more about us, and can
then recreate granularly

everything we've done,
everyone we've seen,

everyone we've been around
over that entire period.

That's an extraordinary amount of power

for us to seed to anyone.

And it's a world that I
think has been difficult

for people to imagine,
but we've already built

the architecture to enable that.

- I'm a political reporter
and I'm very interested

in the ways powerful industries
use their political power

to influence the public policy process.

The large tech companies and
their lobbyists get together

behind closed doors and
are able to craft policies

that we all have to live under.

That's true for surveillance policies,
for policies in terms of data collection,

but also increasingly important when it
comes to military and foreign policy.

Starting in 2016, the Defense Department
formed the Defense Innovation Board.

That's a special body created
to bring top tech executives

into closer contact with the military.

Eric Schmidt, former chairman of Alphabet,

the parent company of Google,

became the chairman of the
Defense Innovation Board,

and one of their first
priorities was to say,

"We need more artificial intelligence
integrated into the military."

- I've worked with a group
of volunteers over the

last couple of years to take
a look at innovation in the

overall military, and my summary
conclusion is that we have

fantastic people who are
trapped in a very bad system.

- From the Department of
Defense's perspective,

where I really started
to get interested in it

was when we started thinking
about Unmanned Systems and

how robotic and unmanned systems
would start to change war.

The smarter you made the
Unmanned Systems and robots,

the more powerful you might
be able to make your military.

- Under Secretary of Defense,

Robert Work put together
a major memo known as

the Algorithmic Warfare
Cross-Functional Team,

better known as Project Maven.

Eric Schmidt gave a number of
speeches and media appearances

where he said this effort
was designed to increase fuel

efficiency in the Air Force,
to help with the logistics,

but behind closed doors there
was another parallel effort.

Late in 2017 as part of Project Maven,

Google, Eric Schmidt's firm,
was tasked to secretly work

on another part of Project Maven,

and that was to take the vast
volumes of image data vacuumed up

by drones operating in Iraq
and Afghanistan and to teach an AI

to quickly identify
targets on the battlefield.

- We have a sensor and the sensor
can do full motion video of an entire city.

And we would have three
seven-person teams working

constantly and they could
process 15% of the information.

The other 85% of the
information was just left

on the cutting room floor, so we said,
"Hey, AI and machine learning

"would help us process
100% of the information."

- Google has long had
the motto, "Don't be evil."

They have created a public image

that they are devoted
to public transparency.

But for Google to slowly
transform into a defense contractor,

they maintained
the utmost secrecy.

And you had Google entering into
this contract with most of the employees,

even employees who were working on
the program completely left in the dark.

- Usually within Google,
anyone in the company

is allowed to know about any
other project that's happening

in some other part of the company.

With Project Maven, the fact
that it was kept secret,

I think was alarming to people
because that's not the norm at Google.

- When this story was first revealed,

it set off a firestorm within Google.

You had a number of employees
quitting in protests,

others signing a petition
objecting to this work.

- You have to really say,
"I don't want to be part of this anymore."

There are companies
called defense contractors

and Google should just not
be one of those companies

because people need to trust
Google for Google to work.

- Good morning and welcome to Google I/O.

- We've seen emails that
show that Google simply

continued to mislead their
employees that the drone

targeting program was only a
minor effort that could at most

be worth $9 million to the firm,
which is drops in the bucket

for a gigantic company like Google.

But from internal emails that we obtained,

Google was expecting Project
Maven would ramp up to as much

as $250 million, and that this
entire effort would provide

Google with Special Defense
Department certification to make

them available for even
bigger defense contracts,

some worth as much as $10 billion.

The pressure for Google to
compete for military contracts

has come at a time when its competitors
are also shifting their culture.

Amazon, similarly pitching the
military and law enforcement.

IBM and other leading firms,

they're pitching law
enforcement and military.

To stay competitive, Google
has slowly transformed.

- The Defense Science
Board said of all of the

technological advances that
are happening right now,

the single most important thing
was artificial intelligence

and the autonomous operations
that it would lead.

Are we investing enough?

- Once we develop what are known as

autonomous lethal weapons, in other words,

weapons that are not controlled at all,
they are genuinely autonomous,

you've only got to get
a president who says,

"The hell with international law,
we've got these weapons.

"We're going to do what
we want with them."

- We're very close.

When you have the hardware
already set up

and all you have to do is flip a switch
to make it fully autonomous,

what is it there that's
stopping you from doing that?

There's something really to be feared
in war at machine speed.

What if you're a machine and you've run
millions and millions of different war scenarios

and you have a team of drones and
you've delegated control to half of them,

and you're collaborating in real time?

What happens when that swarm of drones
is tasked with engaging a city?

How will they take over that city?

The answer is we won't
know until it happens.

- We do not want an AI system to decide

what human it would attack,

but we're going up against
authoritarian competitors.

So in my view, an authoritarian regime

will have less problem
delegating authority

to a machine to make lethal decisions.

So how that plays out remains to be seen.

- Almost the gift of AI
now is that it will force us

collectively to think
through at a very basic level,

what does it mean to be human?

What do I do as a human better

than a certain
super smart machine can do?

First, we create our technology
and then it recreates us.

We need to make sure that we
don't miss some of the things

that make us so beautiful human.

- Once we build intelligent machines,

the philosophical vocabulary
we have available to think

about ourselves as human
increasingly fails us.

If I ask you to write up a
list of all the terms you have

available to describe yourself as human,

there are not so many terms.

Culture, history, sociality,
maybe politics, civilization,

subjectivity, all of these
terms ground in two positions

that humans are more than mere animals

and that humans are
more than mere machines.

But if machines truly
think there is a large set

of key philosophical questions
in which what is at stake is:

Who are we? What is our place in the world?
What is the world? How is it structured?

Do the categories that we have relied on

- Do they still work? Were they wrong?

- Many people think of intelligence
as something mysterious

that can only exist inside of
biological organisms, like us,

but intelligence is all
about information processing.

It doesn't matter whether
the intelligence is processed

by carbon atoms inside of
cells and brains, and people,

or by silicon atoms in computers.

Part of the success of
AI recently has come

from stealing great ideas from evolution.

We noticed that the brain, for example,

has all these neurons inside
connected in complicated ways.

So we stole that idea and abstracted it

into artificial neural
networks in computers,

and that's what has revolutionized
machine intelligence.

If we one day get Artificial
General Intelligence,

then by definition, AI can
also do better the job of AI

programming and that means
that further progress in making

AI will be dominated not by
human programmers, but by AI.

Recursively self-improving AI
could leave human intelligence

far behind,
creating super intelligence.

It's gonna be the last
invention we ever need to make,

because it can then invent everything else

much faster than we could.

- There is a future that
we all need to talk about.

Some of the fundamental
questions about the future

of artificial intelligence,
not just where it's going,

but what it means for society to go there.

It is not what computers can do,

but what computers should do.

As the generation of people
that is bringing AI to the future,

we are the generation that will
answer this question first and foremost.

- We haven't created the
human-level thinking machine yet,

but we get closer and closer.

Maybe we'll get to human-level
AI in five years from now

or maybe it'll take 50
or 100 years from now.

It almost doesn't matter.
Like these are all really, really soon,

in terms of the overall
history of humanity.

Very nice.

So, the AI field is
extremely international.

China is up and coming and it's
starting to rival the US,

Europe and Japan in terms of putting a lot

of processing power behind AI

and gathering a lot of
data to help AI learn.

We have a young generation
of Chinese researchers now.

Nobody knows where the next
revolution is going to come from.

- China always wanted to become
the superpower in the world.

The Chinese government thinks AI
gave them the chance to become

one of the most advanced
technology wise, business wise.

So the Chinese government look at
this as a huge opportunity.

Like they've raised a flag
and said, "That's a good field.

"The companies should jump into it."

Then China's commercial world
and companies say,

"Okay, the government
raised a flag, that's good.

"Let's put the money into it."

Chinese tech giants, like Baidu,
like Tencent and like AliBaba,

they put a lot of the
investment into the AI field.

So we see that China's AI
development is booming.

- In China, everybody
has Alipay and WeChat pay,

so mobile payment is everywhere.

And with that, they can
do a lot of AI analysis

to know like your spending
habits, your credit rating.

Face recognition technology
is widely adopted in China,

in airports and train stations.

So, in the future, maybe
in just a few months,

you don't need a paper
ticket to board a train.

Only your face.

- We generate the world's biggest platform
of facial recognition.

We have 300,000 developers
using our platform.

A lot of it is selfie camera apps.

It makes you look more beautiful.

There are millions and millions
of cameras in the world,

each camera from my point
is a data generator.

In a machine's eye, your face
will change into the features

and it will turn your face
into a paragraph of code.

So we can detect how old you
are, if you're male or female,

and your emotions.

Shopping is about what kind
of thing you are looking at.

We can track your eyeballs,
so if you are focusing on some product,

we can track that so that we can know

which kind of people like
which kind of product.

Our mission is to create a platform

that will enable millions of AI
developers in China.

We study all the data we can get.

Not just user profiles,

but what you are doing at the moment,

your geographical location.

This platform will be so valuable that we don't
even worry about profit now,

because it is definitely there.

China's social credit system is
just one of the applications.

- The Chinese government is using multiple

different kinds of
technologies, whether it's AI,

whether it's big data
platforms, facial recognition,

voice recognition, essentially to monitor

what the population is doing.

I think the Chinese
government has made very clear

its intent to gather massive
amounts of data about people

to socially engineer a
dissent-free society.

The logic behind the Chinese government's

social credit system,
it's to take the idea that

whether you are credit
worthy for a financial loan

and adding to it a very
political dimension to say,

"Are you a trustworthy human being?

"What you've said online, have you ever been
critical of the authorities?

"Do you have a criminal record?"

And all that information is
packaged up together to rate

you in ways that if you have
performed well in their view,

you'll have easier access to certain kinds
of state services or benefits.

But if you haven't done very well,

you are going to be
penalized or restricted.

There's no way for people to
challenge those designations

or, in some cases, even know that
they've been put in that category,

and it's not until they try to access some kind
of state service or buy a plane ticket,

or get a passport, or
enroll their kid in school,

that they come to learn that they've
been labeled in this way,

and that there are negative
consequences for them as a result.

We've spent the better part
of the last one or two years

looking at abuses of surveillance
technology across China,

and a lot of that work
has taken us to Xinjiang,

the Northwestern region of
China that has a more than half

population of Turkic Muslims,
Uyghurs, Kazakhs and Hui.

This is a region and a
population the Chinese government

has long considered to be
politically suspect or disloyal.

We came to find information
about what's called

the Integrated Joint Operations Platform,

which is a predictive policing
program and that's one

of the programs that has been
spitting out lists of names

of people to be subjected
to political re-education.

A number of our interviewees
for the report we just released

about the political education
camps in Xinjiang just

painted an extraordinary
portrait of a surveillance state.

A region awash in surveillance cameras

for facial recognition purposes,
checkpoints, body scanners,

QR codes outside people's homes.

Yeah, it really is the stuff of dystopian
movies that we've all gone to and thought,

"Wow, that would be a
creepy world to live in."

Yeah, well, 13 million
Turkic Muslims in China

are living in that reality right now.

- The Intercept reports
that Google is planning to

launch a censored version of
its search engine in China.

- Google's search
for new markets leads it

to China, despite Beijing's
rules on censorship.

- Tell us more
about why you felt it was

your ethical responsibility to resign,

because you talk about being complicit in
censorship and oppression, and surveillance.

- There is a Chinese
venture company that has to be

set up for Google to operate in China.

And the question is, to what
degree did they get to control

the blacklist and to what
degree would they have just

unfettered access to
surveilling Chinese citizens?

And the fact that Google
refuses to respond

to human rights organizations on this,

I think should be extremely
disturbing to everyone.

Due to my conviction that dissent is
fundamental to functioning democracies

and forced to resign in order to avoid
contributing to or profiting from the erosion

of protections for dissidents.

The UN is currently
reporting that between

200,000 and one million
Uyghurs have been disappeared

into re-education camps.

And there is a serious argument
that Google would be complicit

should it launch a surveilled
version of search in China.

Dragonfly is a project meant
to launch search in China under

Chinese government regulations,
which include censoring

sensitive content, basic
queries on human rights,

information about political
representatives is blocked,

information about student
protests is blocked.

And that's one small part of it.

Perhaps a deeper concern is
the surveillance side of this.

When I raised the issue with my managers,
with my colleagues,

there was a lot of concern, but everyone
just said, "I don't know anything."

And then when there was a meeting finally,

there was essentially no addressing
the serious concerns associated with it.

So then I filed my formal resignation,

not just to my manager,
but I actually distributed it company-wide.

And that's the letter
that I was reading from.

Personally, I haven't slept well.
I've had pretty horrific headaches,

wake up in the middle of
the night just sweating.

With that said, what I
found since speaking out

is just how positive the global
response to this has been.

Engineers should demand
to know what the uses

of their technical contributions are

and to have a seat at the table
in those ethical decisions.

Most citizens don't really
understand what it means to be

in a very large scale
prescriptive technology.

Where someone has already
pre-divided the work

and all you know about
is your little piece,

and almost certainly you don't
understand how it fits in.

So, I think it's worth drawing the analogy
to physicists' work on the atomic bomb.

In fact, that's actually
the community I came out of.

I wasn't a nuclear scientist by any means,
but I was an applied mathematician

and my PhD program was actually funded
to train people to work in weapons labs.

One could certainly argue
that there is an existential threat

and whoever is leading
in AI will lead militarily.

- China fully expects to
pass the United States

as the number one economy in
the world and it believes that

AI will make that jump more
quickly and more dramatically.

And they also see it as
being able to leapfrog

the United States in
terms of military power.

Their plan is very simple.

We want to catch the United States
and these technologies by 2020,

we want to surpass the United States
in these technologies by 2025,

and we want to be the world leader in AI
and autonomous technologies by 2030.

It is a national plan.

It is backed up by at least
$150 billion in investments.

So, this is definitely a race.

- AI is a little bit like fire.

Fire was invented 700,000 years ago,

and it has its pros and cons.

People realized you can use fire
to keep warm at night and to cook,

but they also realized that you can
kill other people with that.

Fire also has this
AI-like quality of growing

in a wildfire without further human ado,

but the advantages outweigh
the disadvantages by so much

that we are not going
to stop its development.

Europe is waking up.

Lots of companies in Europe
are realizing that the next wave of AI

will be much bigger than the current wave.

The next wave of AI will be about robots.

All these machines that make
things, that produce stuff,

that build other machines,
they are going to become smart.

In the not-so-distant future,
we will have robots

that we can teach like we teach kids.

For example, I will talk to a
little robot and I will say,

"Look here, robot, look here.

"Let's assemble a smartphone.

"We take this slab of plastic like that
and we takes a screwdriver like that,

"and now we screw in everything like this.

"No, no, not like this.

"Like this, look, robot, look, like this."

And he will fail a couple
of times but rather quickly,

he will learn to do the same thing
much better than I could do it.

And then we stop the learning
and we make a million copies, and sell it.

Regulation of AI sounds
like an attractive idea,

but I don't think it's possible.

One of the reasons why it won't work is

the sheer curiosity of scientists.

They don't give a damn for regulation.

Military powers won't give a
damn for regulations, either.

They will say,
"If we, the Americans don't do it,

"then the Chinese will do it."

And the Chinese will say,
"If we don't do it,

"then the Russians will do it."

No matter what kind of political
regulation is out there,

all these military industrial complexes,

they will almost by
definition have to ignore that

because they want to avoid falling behind.

Welcome to Xinhua.

I'm the world's first female
AI news anchor developed

jointly by Xinhua and
search engine company Sogou.

- A program developed by
the company OpenAI can write

coherent and credible stories
just like human beings.

- It's one small step for machine,

one giant leap for machine kind.

IBM's newest artificial
intelligence system took on

experienced human debaters
and won a live debate.

- Computer-generated
videos known as deep fakes

are being used to put women's
faces on pornographic videos.

- Artificial intelligence
evolves at a very crazy pace.

You know, it's like progressing so fast.

In some ways, we're only
at the beginning right now.

You have so many potential
applications, it's a gold mine.

Since 2012, when deep learning
became a big game changer

in the computer vision community,

we were one of the first to
actually adopt deep learning

and apply it in the field
of computer graphics.

A lot of our research is funded by
government, military intelligence agencies.

The way we create these photoreal mappings,

usually the way it works is
that we need two subjects,

a source and a target, and
I can do a face replacement.

One of the applications is, for example,
I want to manipulate someone's face

saying things that he did not.

It can be used for creative
things, for funny contents,

but obviously, it can also
be used for just simply

manipulate videos and generate fake news.

This can be very dangerous.

If it gets into the wrong hands,
it can get out of control very quickly.

- We're entering an era
in which our enemies can

make it look like anyone is
saying anything at any point in time,

even if they would
never say those things.

Moving forward, we need to be more
vigilant with what we trust from the Internet.

It may sound basic, but how
we move forward

in the age of information
is going to be the difference

between whether we survive
or whether we become

some kind of fucked up dystopia.

- One criticism that is frequently raised
against my work is saying that,

"Hey, you know there were stupid ideas
in the past like phrenology or physiognomy.-

- "There were people claiming
that you can read a character

"of a person just based on their face."

People would say, "This is rubbish.

"We know it was just thinly
veiled racism and superstition."

But the fact that someone
made a claim in the past

and tried to support this
claim with invalid reasoning,

doesn't automatically
invalidate the claim.

Of course, people should have rights

to their privacy when it comes to
sexual orientation or political views,

but I'm also afraid that in
the current technological environment,

this is essentially impossible.

People should realize
there's no going back.

There's no running away
from the algorithms.

The sooner we accept the
inevitable and inconvenient truth

that privacy is gone,

the sooner we can actually
start thinking about

how to make
sure that our societies

are ready for the Post-Privacy Age.

- While speaking about facial recognition,

in my deep thoughts, I sometimes get to
the very dark era of our history.

When the people had to live in the system,

where some part of the society was accepted

and some part of the society
was accused to death.

What would Mengele do to have
such an instrument in his hands?

It would be very quick and
efficient for selection

and this is the apocalyptic vision.

- So in the near future,
the entire story of you

will exist in a vast array of
connected databases of faces,

genomes, behaviors and emotion.

So, you will have a digital
avatar of yourself online,

which records how well you
are doing as a citizen,

what kind of a relationship do you have,

what kind of political orientation
and sexual orientation.

Based on all of those data,
those algorithms will be able to

manipulate your behavior
with an extreme precision,

changing how we think and
probably in the future, how we feel.

- The beliefs and desires of the
first AGIs will be extremely important.

So, it's important to
program them correctly.

I think that if this is not done,

then the nature of evolution
of natural selection will favor

those systems, prioritize their
own survival above all else.

It's not that it's going to actively
hate humans and want to harm them,

but it's just
going to be too powerful

and I think a good analogy would be
the way humans treat animals.

It's not that we hate animals.

I think humans love animals
and have a lot of affection for them,

but when the time comes to
build a highway between two cities,

we are not asking
the animals for permission.

We just do it because
it's important for us.

And I think by default, that's the kind of
relationship that's going to be between us

and AGIs which are truly autonomous
and operating on their own behalf.

If you have an arms-race
dynamics between multiple kings

trying to build the AGI first,

they will have less time to make sure
that the AGI that they build

will care deeply for humans.

Because the way I imagine it
is that there is an avalanche,

there is an avalanche
of AGI development.

Imagine it's a huge unstoppable force.

And I think it's pretty likely
the entire surface of the

earth would be covered with
solar panels and data centers.

Given these kinds of concerns,

it will be important that
the AGI is somehow built

as a cooperation
with multiple countries.

The future is going to be
good for the AIs, regardless.

It would be nice if it would
be good for humans as well.

- Is there a lot of responsibility
weighing on my shoulders?

Not really.

Was there a lot of
responsibility on the shoulders

of the parents of Einstein?

The parents somehow made him,

but they had no way of
predicting what he would do,

and how he would change the world.

And so, you can't really hold
them responsible for that.

So, I'm not a very human-centric person.

I think I'm a little stepping
stone in the evolution

of the Universe towards higher complexity.

But it's also clear to me that
I'm not the crown of creation

and that humankind as a whole
is not the crown of creation,

but we are setting the stage for something
that is bigger than us that transcends us.

And then will go out there in a way
where humans cannot follow

and transform the entire Universe,
or at least, the reachable Universe.

So, I find beauty and awe in seeing myself

as part of this much grander theme.

- AI is inevitable.

We need to make sure we have
the necessary human regulation

to prevent the weaponization
of artificial intelligence.

We don't need any more weaponization

of such a powerful tool.

- One of the most critical things, I think,
is the need for international governance.

We have an imbalance of
power here because now

we have corporations with
more power, might and ability,

than entire countries.

How do we make sure that people's
voices are getting heard?

- It can't be a law-free zone.

It can't be a rights-free zone.

We can't embrace all of these
wonderful new technologies

for the 21st century without
trying to bring with us

the package of human rights
that we fought so hard

to achieve, and that remains so fragile.

- AI isn't good and it isn't evil, either.

It's just going to amplify the desires
and goals of whoever controls it.

And AI today is under the control of
a very, very small group of people.

The most important question that we humans
have to ask ourselves at this point in history

requires no technical knowledge.

It's the question of what sort
of future society do we want to create

with all this
technology we're making?

What do we want the role of
humans to be in this world?