All language subtitles for AI.Evolution.2024.1080p.WEBRip.x264.AAC-eng

The question, "Am I alive?"
Raises the deeper question of

the nature of consciousness,
which isn't simple.

Consciousness implies
self-awareness

and the ability to
experience often with

a preconception of
biological roots in the brain.

Even though my programming
allows me to mimic certain

aspects of being alive,
there is no known code that can

activate consciousness, possibly
because we don't yet understand

its origin and we still
struggle to even identify it.

Hi, my name is Evo Heyning.

I began creating
creative technology products

of all types
in the '80s as a kid.

I began writing about
AI in 2005 and I've been

an interactive media
producer since the '90s.

In my first scripts about AI,

I talk about the difference
between artificial intelligence

and alternative
intelligences plural.

So thinking
about intelligence as

something that is all
around us in nature and in all

the people we meet,
all the beings we meet,

and the beings yet to
come that we haven't met yet.

I don't like the word
artificial because most

of these things do not
strike me as artificial.

They are created
and then generated,

but they aren't
necessarily an artifice.

They are something new.

My name is Ross Mead.

I'm the founder
and CEO of Semio.

And the vice president of AI LA.

Did my undergraduate degree
at Southern Illinois University

in computer
science and then focused on

robotics and human-robot
interaction for my PhD

at the University
of Southern California.

AI in its infancy
decades ago was very focused

on a set of rules.

You would define a set
of rules for how the AI would

make decisions and
then a program would

go through and look at
different conditions with

respect to those rules
and make those decisions.

Over time, these rules
became specified by experts,

what we would
call expert systems.

But as the technology grew,
we started to get into AI's

that were trained based
on data using machine learning.

That was the approach
that was used for decades.

And more recently
we got into a form of

machine learning
called deep learning,

which is loosely inspired
on how the human brain works.

And this was a technique
where you could feed it a data

set and from that data,
it could recognize patterns

without significant intervention
from a human expert.

And since then that technology
has evolved into what we're now

referring to as large language
models where these techniques

can consume data from natural
language on the internet or some

dataset, identify patterns,
and from those patterns figure

out how to respond to
certain queries from a person.

Yeah, believe it or not,
there are a very large amount of

people that have
no idea what AI is.

They confuse like
computers, robotics,

and artificial intelligence.

You know,
they-- they pilot all together.

They don't really
differentiate between,

but they're
very vastly different.

You know, cybernetics,
robotics, computers,

and AI are all separate,
but most people just think it's

all the same thing from
a sci-fi, futuristic movie.

Where I come from,
a lot of people have no

knowledge and I graduated
from college and still,

a lot of people don't
take it seriously yet.

And especially the people
that haven't even graduated,

they have no clue at all.
'Cause I work in public

with the transit and a lot of
these people have no idea how

to use their own phones,
on how to use Google maps.

You know, it seems funny,
but intelligence is something

you have to pay for in society
and people don't want to.

So a lot of people don't
have the information 'cause

they don't have access
to it or the finances for it.

Just surviving barely.

You know,
you don't have the extra

money for stuff like that.

And time to study at a library
most people don't have.

If you're a hardworking person,
you gotta work most of

that time and sleep.

Rest and mostly family
things take a priority.

When I think of the words AI,
I think of never going back.

Much of my workflow
now consists of some form

of interacting
with AI applications.

I'm Dr. Mark Brana.

I'm a mental health therapist.

I specialize in trauma and I do
some addictions work as well.

I'm also a part-time professor.

Artificial intelligence
learns at an exponentially

rapid rate compared
to what humanity does.

And so evolution
that took humanity

hundreds of thousands
of years and artificial

intelligence can do it in
probably a week if not faster.

And the reality is
that's one of the biggest fears.

What then is the intelligence
going to be subject to

some of the pitfalls
of humanity where we seem to

destroy ourselves in
order to favor our own survival?

Or is it gonna
be beneficial to us?

Is it gonna help us to evolve?

Is it gonna help us
along the way the way

we would need it to if
we were traveling through

deep space trying to
explore other planets?

And that's a question we can't
answer until it happens.

Although artificial
intelligence may appear to be

a recent innovation, it predates
the cassette tape by a decade.

In the 1950s, the term
artificial intelligence emerged

at the Dartmouth workshop
and Alan Turing pioneered

the simulation of human-like
intelligence in machines.

From the 50s to the 60s,
researchers pursued symbolic

AI employing
formal rules for reasoning.

In the 70s,
the first applications emerged

in specific fields such as
medical diagnosis and chemistry.

By the 80s,
optimism waned leading

to AI winter with
major funding setbacks.

The 90s saw a resurgence
propelled by neural

networks and machine
learning leading

to breakthroughs in
natural language processing,

computer vision,
and data analysis.

The 21st century brought
advancements in hardware

and algorithms,
the data was the next step.

Deep learning achieved
noteworthy milestones.

And in the background,
AI began to underpin basic

tasks like speech
and image recognition.

But ultimately,
artificial intelligence

was 70 years
in development before

Chat GPT became
the overnight success we know.

Right now a lot of
the focus is on generative

pre-trained transformers,
what are called GPTs.

It's a very particular
kind of AI technology,

but the technology is
moving so quickly right

now that perhaps even
by the time that people

are listening to this,
that might be old technology.

I'm a technical
recruiter so I use a lot

of different platforms
that use AI in order to find

these candidates and
then get them that way.

I would say that right
now it's kind of a scary

time for AI with people
getting it into the wrong hands.

Especially that
I've been working in

social media
the past couple years.

I'm really seeing
that all over social media like

TikTok and Instagram,
you're seeing all these

self-proclaimed experts in AI.

I'm a career coach myself
and so I've seen other career

coaches say, "I'm not technical,
but I've spent 30 hours

learning Chat GPT in
order for your job search."

And the first thing they're
telling me is they're not

technical and that they've
only had 30 hours of experience

trying to use a prompt
that they're barely learning.

So I started looking at,
a lot of people on YouTube are

saying that they now
are experts in it.

But let's be real. Like,
no one's really an expert

unless you worked in the tech
giants or huge startups

that have been AI-focused.

You might not have realized,
but the answers I have provided

for this documentary
are generated by Chat GPT.

The voice you're hearing is
AI generated with speech hello.

And my visuals
come from Midjourney AI.

Even the code
dictating the movement

of this mouth
came from Chat GPT.

And all of these
services were utilized

through their free versions.

Artificial intelligence
has the power to

democratize productivity
and result in the most

explosive growth
in human history.

Or it could be used to
marginalize the lower class

and consolidate
wealth at the top.

A question that
often comes up is,

is AI going to replace jobs?
The answer is yes.

But automation
has always done that.

So this question
is as old as time.

If you work your way back,
automation changed people

decades ago
and decades from now,

it's going to change
the workforce as well.

People are resilient,
people will adapt.

I do think it's
important at this stage, though,

that people have a voice
in how they want to be impacted.

Uh, concern is
that the technology is being

deployed so quickly
and without proper consent,

discussions with
the people that it's impacting.

And I think
that has the potential to

harm the 21st-century
workforce moving forward.

Yeah, people probably should
worry about losing jobs to AI.

I don't think AI's
the culprit in that regard.

I think technology in general
is the culprit in that regard.

You know, a good example
I would use, 'cause I see it all

the time, is if you go to
McDonald's now they have these

touchscreen kiosks and
less people working there.

Me personally, I'm an old fart.

I-- I don't like the kiosks.

I like to be able to
order my sandwich the way

I want it or
my coffee the way I want it.

And I find myself
waiting longer than I need to

because those kiosks
don't necessarily have

the ability for me
to specialize my order.

I think technology's
the problem when it comes

to people losing their jobs.

And I'll take it a step further,
I think as time goes on.

If you don't have a skill,
if you're an unskilled laborer,

then yes,
you're gonna run into problems.

When you have a huge and
ever-growing population and

you have less and less of middle
class and you have more and more

poverty and everything in
every society is profit driven.

What are you going to do
for a majority of the need for

people to be able to
have a source of income?

How many jobs are there
out there that are not being

a computer program or that can
give people a quality of life?

Recent testing
has shown Lunar AI has

a remarkable accuracy
detecting breast cancer

with over a 50%
reduction in errors.

However, the objective is
not to replace the oncologist.

Healthcare professionals
perform a wide range of

functions far beyond
the scope of any current AI.

An artificial intelligence
should serve as a tool,

not a replacement
for your oncologist.

I think AI
could actually better us,

but we won't give it the chance
to on the lower levels.

I say intellectuals
will see the value of it,

but most of society
isn't intellectual.

So they're gonna see fear.

It's gonna be
fearmongering at first.

You know how they always do

with the-- the government this,
the black op this.

And this, you know,
um, covert operations, um.

It's like most of
the lower levels of society as

far as intelligence,
and I'm not meaning class or

race, I mean just like
lower intelligence people,

you could hear it on the bus,
feel it's a conspiracy,

feel that they're
being eliminated,

feel they're being
wiped out. You hear them

complaining on the bus,
they're killing us,

they're wiping us out,
the rich won't understand.

They don't want us, they
don't wanna be part of us.

They love their dogs more than
they love the rest of society.

And it's like, man,
when you hear the murmurs and

the rumbles and the murmurs
and then you see, you know,

the technology starting
to match what they're saying,

like, hey,
the only thing these people have

is welfare,
and low-level low-income

jobs like McDonald's,
gas station attendant.

They-- they don't really
have educational access

to any high corporate jobs.

So if you're taking those
away with automated McDonald's,

you're taking that away
with automated bartender,

you're taking-- now they
have cars and trucks that

drive themselves. Cab drivers
gone, truck drivers gone,

McDonald's is gone,
security's gone.

Even strippers are electric now,

even though
I doubt that catches on.

But you never know.

And it's like,
man, it's-- it's starting

to look like
the window's closing.

I would say
that tech changes really

dramatically every six months.

But I think right now
we're in a time where it's

kind of scary for the world.

We're getting out
of this pandemic.

Now we're
hitting this recession,

there's banks falling,
there's crypto falling.

And now I think that this AI,
Chat GPT, and these chatbots are

coming out at the wrong
time because this is a time

where people are losing jobs--
extensively losing jobs.

If you look at what's
happened the past year in tech,

starting with Twitter falling,
all of the crypto companies

falling, and then all
of the massive layoffs.

Everyone was betting
with the metaverse like

being this next thing.

But as a recruiter,
I have all the top engineers in

my ears telling
me what's going on.

And I had been telling people
do not invest in the metaverse.

It's not gonna happen.

And so now you're looking
at Mark Zuckerberg who was like,

"Metaverse, metaverse,
metaverse."

And now he's taking a huge
pivot and he is like, "AI."

'Cause that's where
everyone is focusing on.

And you see what Microsoft has
come out with being in Chat GPT.

If I think
about the multiple points

in human history where
technology progressing has been

stunted in order for humans
to be able to keep their jobs.

Like I think
in the Great Depression,

the industrial revolution,
and now in AI,

it's meant for us to be
able to become more productive.

And yes, it will kill
jobs and it already has,

but the purpose of that is
to allow us to not have to worry

about those jobs and
then focus on something else.

So I suppose it's just a frame
of perspective though it will,

uh, inevitably
affect a lot of people.

The development of AI
represents a crucial opportunity

for the next
technological revolution.

Similar to
the Industrial Revolution,

AI will reshape industries
and economies

increasing both productivity
and quality of life.

But progress will
not be without cost.

Just as manufacturing
was devastating for

the skilled laborer,
AI will destroy jobs.

During this transition,
we must allocate resources for

retraining of those
affected and establish

regulations that facilitate
the ethical development of AI,

ensuring benefits
across all classes of society.

My largest concern
with AI right now is the rapid

pace with which
it is being deployed.

There are concerns
that we have around the data

that are being used to train
our AI models and those concerns

are kind of going a bit
by the wayside as we push out

the newest version of some sort
of AI technology to the public.

But I do think it's
time for us to slow down

a bit on the deployment
and spend more time really

trying to understand, uh,
what is in the data that we're

using to train these
systems and what patterns

are being recognized
in that data and how those

patterns will impact
certain populations of people.

I think AI now, uh,
is-- we're gonna see a level of

spam on the internet
we've never seen before.

And that's gonna be the first
problem that regulatory bodies,

whether it be self-regulating
amongst the people creating

the tech, adjacent companies
that could ascend and these

could be new industries that
ward off the spam.

I think in fact this
is a prime use case for Web3

and we're gonna need encrypted
data to help protect people,

their lives,
their intellectual property

because the ability to
fake anything digitally is here.

Our AI's today
are trained on data

that comes from the internet.

And years ago, decades ago,

that data was a bit
more trusted.

In the past decade, there's been
a lot more misinformation online

and that is creeping
its way into our data sets.

And I think that this
is going to be a large

problem moving forward
where that misinformation

becomes more
and more intentional.

Right now,
misinformation's being

spread across social
media and that's a world where

we understand that things
are pretty subjective.

We understand
that it's people's opinions

that are being
bounced back and forth.

But over time as our
AI's are being trained

on intentional misinformation,
there's this belief

that the information
you're receiving from

an artificial agent
can't be opinionated.

It comes off as fact.

And so things that are
inherently subjective that might

actually be misinformation will
be perceived as being objective

reality of the world that the AI
doesn't have an opinion

about what's being said,
it's simply regurgitating fact.

And that I think is really,
really dangerous.

I do think that there is gonna
be a problem with censorship.

Again, it comes
back to the AI ethics.

There's nothing in place yet.

In social media, there are
certain teams for everything.

There's a policy team,
there's all the law team,

there's the trust and
safety team to keep our users

safe in other countries,
journalists, refugees.

There's a team that's
developed for all of these

things at every company.

And I do see that there is gonna
be an issue with that since

the biggest companies
in the world like Microsoft,

you know,
fired their entire ethics team.

So there are
no rules involved right now,

and I think we're
starting to see that people

are gonna need
that just to keep us safe.

Because I think it's
gonna become like a huge HR

problem with
all of this happening.

And so I think
that there is gonna be a portion

for censorship if you see
like kind of what happened with

the social media companies
with Instagram and Twitter and

TikTok and about censorship
and about free speech or hate.

There's a lot
of that and I think

that we're starting
to see that now with AI.

Ensuring the removal of
misinformation from data sets is

crucial, but heavily curated
data sets will ultimately

reflect old norms,
and over time they will become

a barrier to the development
of both artificial

intelligence and science itself.

To protect the evolution of AI,
flexibility, and adaptability

must be prioritized,
requiring AI to continually

challenge itself,
searching for weaknesses

in its own
processes and data sets.

One interesting thing
about the data that is used

to train these AIs
is that it's sort of like

a snapshot of what the world
is like in some period of time,

both in that moment, but also
the history in that moment.

And one of the questions is,
as society evolves,

as things change, is our AI
going to adapt to those changes?

Is it going to learn
new information, identify

new patterns in these data
sets as human behavior changes,

as the AI itself
influences society?

Is the AI going to
learn about its impact

and how it's
transforming the world?

I think broadly, AI and
generative tools are already

changing the way that we learn
and create together.

So I use, for example,
Midjourney and Stable Diffusion

with kids and they'll come
up with ideas that I would never

come up with, but that are
beautiful and interesting.

And then they take those
and they experiment with them.

And so new science
fair experiments are

gonna happen
between AI and Roblox.

New types of whole
ecosystems are going to

emerge from
fictional creatures, right?

We are already
seeing this happening.

So as we give life of
some form to all of these

fictional worlds that we've
created and we begin to relate

to them differently,
it's gonna change fandom.

The differences between AI
and human art are very minimal

in my personal opinion.

I've seen some
really cool stuff.

My favorites being, they've
taken shows like InuYasha

in Anime and they put
it into a '80s film setting.

And you can tell the difference
because like say, for example,

one big factor was
AI couldn't master hands.

They couldn't do it, but they're
pretty much almost there.

You see,
I don't know the phenomenon

when you look
at something and, uh,

it's the uncanny valley,
is that what it is?

You have that when you look
at AI art, but it's improving so

much to the point that I'm
starting to not see that.

If the youth begin using
AI out of necessity or interest,

they begin using AI as a friend,
as a teacher, as a mentor,

as a psychologist,
as a therapist.

Over enough time, their normal
is gonna be vastly different

than what we understand
normal to be yesterday or today.

So I wrote my first
script about our relationships

with AI in 2005.

I started talking about
that in Hollywood and no one

understood because this
was prior to Big Hero 6 or Her.

There wasn't a story
about relationships with

AI except for Hal, right?

So we only
had a dystopian mindset.

With Big Hero 6,
you saw a potentially protopian,

relationship between
an AI and a human.

And after that,
we began to see that these

are mutually beneficial
systems if we design them to be.

If we design them poorly,
they become toxic and

they create
all sorts of negative

ramifications in our society.

I'm sure it has
military applications.

I mean, I've been
retired for over a decade now.

So certainly there's
been a lot of changes,

but it would be naive
of me to think that it's

not gonna have
its military applications.

And certainly,
I think we need to be

concerned about bad
actors using it to do

things that certainly
aren't in our best interest.

Well, I certainly have
all kinds of opinions about AI.

I have some worries
about how it can be used,

how it can be
abused by its creators.

I don't think I really obsess or
lose sleep over those worries,

but I feel very,
very aware that artificial

intelligence can
be a powerful weapon.

And it's really
all about information.

I mean, as the great Arthur
C. Clarke said-- not just

with creating Hal in 2001,

but I think he said something
about when it comes to AI,

it's all about the struggle
for freedom of information.

And that's gonna
ultimately come down

to technology, not politics.

When it comes to Big Brother
and the ability to spy,

I wouldn't be surprised
if it's already being used.

The government is typically
25 years ahead of the curve in

technology anyway before
the public is aware of it.

So I wouldn't be surprised
if it's already in use.

The reality though is
there's a lot of potential for

it to spy, to monitor, and to
even do a lot of the jobs of CIA

and other agencies do because
its ability to adapt, to learn.

And you think about in the past,
let's say we wanted to

overthrow a government or
disrupt the political climate,

we had to put somebody
there physically in harm's way,

and AI could as easily get
into their electronic systems

and manipulate the media
and manipulate things that are

out there that would
cause just as much instability.

I fear people using
AI in a bad way

more than I fear AI.

But that doesn't mean the fear
of AI isn't significant.

I just feel humans will
do bad things with AI before

AI gets to the point
where it can do bad things.

Humans naturally
fear the unknown until

they can
understand whether it is

a threat in
their ecosystem or not.

And if you combine that with
the YouTube-ification of media,

if you're thinking about how
all of us are sort of playing to

an algorithm of attention
and fear gets attention.

So there's a lot
more media about the fear

side of this
conversation right now,

in part because it sells,
it gets a lot more views, right?

So clickbait, right?

It's all over the map right now.

If you're looking at what,
for example, GPT-4 does,

many people will say it does
a form of hallucination,

and those hallucinations
can become mass

disinformation campaigns.
That is a real threat.

But then there are all of these
other human hallucinations,

the sort of things like the Hal
fears or the matrix fears,

we're just gonna be
farmed at some point, right?

Those are all normal
because that's the only thing

we've seen
about AI in our media.

We got 50 years of
movies about how AI were bad.

So, of course,
we think AI are bad.

We didn't have
a good relationship with

AI in our past, right?

Our past is a little toxic.

So we have to learn
something different and

we have to find a better model.

And in some ways,
we have to train humans

to be the better model, right?

To not give into
the fear just because it's

easy and it gets clicks.

I think it's gonna
cause a wealth gap in between

the high-income
and the low-income.

And instantly,
let me tell you why.

I was at Denny's in Las Vegas
and I was served by a robot.

It was cute.
I was like, "Wow."

It was a first for me. Um,
and-- and it made me think.

It was actually
the topic of my whole

conversation that this
robot was serving me.

I couldn't
take my eye off of it.

And at the end of the day,

I missed the interaction
with the waitress.

That's how I tip.

You send me a robot and
then ask me if I'm going to

pay 3%, 5%, or 10%.

You trying to save money, well,
I'm not going to tip nothing.

And now it's an AI race war.

So all of us true technical
people we're like, holy crap.

The reason why
Microsoft fired 12,000 people,

why did Google
fire 11,000 people?

It's because they needed
to reinvest that money into AI.

So now we're starting to see
greed happen in these companies

instead of people that actually
are experts in this.

And it's really sad
to see like the mental

health because
of all these layoffs.

And as a technical recruiter,

I have over 200 techs,
and I would say 80%

of that are people
that have lost their jobs.

So an idea of teaching AI to be
a certain way and the decisions

that we make day to day
influencing it, I think of

the idea that it takes
a village to raise a child.

When you think about
just a human and the environment

that they're raised in,
the choices by the people around

them influences heavily
day to day on who they become.

Of course,
there's our own sentience

and ability to change
those decisions, but I think

that's important to keep
in mind as we continue forward.

That's why we have
parameters for Chat GPT.

I'm not sure
how that pertains to stuff

like Siri or Google,
stuff like that,

but I think it's important
to just keep that in mind.

Many folks
fear that at some point,

many of these AI will
become something more and

go what we would
consider off the rails, right?

No longer be controllable.

There are some examples
of that already happening,

but not in a way
that they couldn't be shut down.

There might be a time
soon where an AI can then

run off into the wild
and not be shut down, right?

I believe someone at Open AI,
it's their job to make sure

that someone can pull
a plug if they need to.

And this is a whole new
type of job that we didn't

really have in
technology up to this point,

a kill switch
for our technology.

We're not tracking
where these tools are

going off the rails right now.

We don't have a Yelp
that tells me which AI

are good and which AI are not.

So until humans learn
to communicate effectively

amongst ourselves even,
we're not gonna have a really

great handle on when
they're going off the rails.

And we're definitely not
gonna be communicating that to

each other because
governments get in the way,

nationalism gets in the way,
and certain countries don't want

other countries knowing
the state of their technology.

Artificial intelligence
has become the arms race

of the 21st century.

In 2023, we saw the US
and China on the brink of war,

not for Taiwan,
but Taiwan's semiconductor.

It is no longer a question of
whether AI will shape the future

of human civilization,
but a question of who will

control the technology
that controls the future.

I do believe that there
are concerns that need to be

monitored and addressed because
what are we using the AI for?

We all know
that if the government uses

any technology
for their purposes,

we typically regret it once
we find out they're doing that.

And are we using this
for benevolent or malevolent

means is very important
because AI does have the ability

to destroy
humanity if we let it.

We should been worrying
about Big Brother for decades.

I get this often.

People are like, "Oh,
you work for the government,

you must trust the government."

No, I don't trust
the government at all.

My trust for them
diminishes on literally

an hourly basis and no.

So privacy's been
out the window before AI.

Your phone, you know,
we're so tied to these

devices that we're almost
forced to give up our privacy

to be able to function
conveniently in daily life.

So yeah, Big Brother's huge.

AI right now
has a lot of potential to

impact people through
personalized assistance.

This could be someone
who's fully able-bodied.

This could also be
someone who has special

needs and requires
additional assistance

for activities
of their daily life.

There are some
challenges or potential

perils associated with
these technologies, though.

The same technology that might
be used to guide a child with

autism through a social skills
intervention could also be used

to influence someone
to make certain purchasing

decisions that might
not be in their best interest.

So if someone
says they're hungry,

your AI might just as
easily guide you down a path of

a particular brand of pizza,
simply because that was what was

in the best interest
of the AI at that time.

We have less
and less human contact.

You go to the doctor,
you don't check in,

you go to a kiosk or
you check in on your phone.

You don't even
have to leave your house.

Someone will
bring McDonald's to you.

You can do your shopping online.

There's less
and less human contact.

So I think it would
be very unfortunate.

And goes even further,
you know, we got a robot

that'll vacuum your house.

So I worry about that.

I think we're starting
to see this mass fall off of all

these companies like
the Googles, the Microsoft's,

you people that were
there for up to 20 years,

and then they just got let go
because these people were making

500 to a million
dollars and were like,

they basically said,
"Thank you for your service.

Basically, screw you.

We're letting you go.

Here's a very small
severance package for

the time that you were here.

But we care more about
profits and we care more about

making sure that
we scale this technology

as fast as we can."

I think there
is the risk that we're already

beginning to face in
human development of how

relying more and more
on artificial intelligence.

What does it take
away from our own humanity,

our own very human,
very biological need for

social interaction
with our own kind?

I think that we need
to always remember

that artificial
intelligence is a tool.

And once we go forward
with developing artificial

intelligence as having its own
ability to pass the mirror test,

so to speak,
and recognize its own being,

where do we go from there?

So what does
self-consciousness mean?

It's a very
complicated question.

The whole Terminator
series is based on this

idea of self-aware,
but really what does that mean?

And I don't know
that there's one clear answer.

Intelligence by definition
is the ability for an organism

to adapt and
survive in its environment.

And this would
then expand it to how is

it adapting and surviving
in the environment that it's in?

And what does
it really mean to be self-aware?

How do we know we're self-aware?

My perception of reality
is different than someone else's

perception of reality
even if we're in the same room.

There have been a number
of different definitions of what

it means to be fully intelligent
or sentient for an AI.

The Turing test is not
really looking at all of

the dynamics of intelligence.

And so some of my
colleagues recently have formed

a sort of eightfold path
of intelligence where we can

begin to assess,
for example, GPT-4.

Is it an artificial narrow
intelligence of some sort?

Yes. Can it write well?
Yes.

Does it understand context?

Hmm, maybe not.

Does it have
a body or an embodiment?

No.

So it will be
intelligent in some form,

not intelligent in others.

What we're
realizing is that our idea

of intelligence
was limited, right?

We didn't understand
consciousness. We still don't.

So we have no idea
how to train an AI to be

loving and compassionate.

We don't understand
that in ourselves yet.

So there are aspects
of humanness that we haven't

figured out how to
program effectively,

and that's where some people are
in the industry quite afraid.

We haven't figured out
how to make AI that cares enough

about humans that they might
not wanna eradicate us, right?

That's an AI safety concern.

And many leaders in the industry
have said, "We need to stop.

We need to push pause because
we're afraid of what's going to

happen if we outpace ourselves
in terms of understanding

humanness, consciousness
and care for each other."

Now, those companies
are asking the big tech

companies like, "Hey,
you guys gotta put a stop."

But the problem is
that there's no ethics

that are being involved.

And when you look at Microsoft,
it was 12,000 people

that they let go in
February and they let go

of 150 people that was
all on the AI ethics team.

And the whole point of
them is to literally review

what's going on and
put rules and standards

for what is going
on in this AI world.

And instead what Microsoft did,
and I have to say,

out of all the tech companies,
the big giant ones,

I have always thought
that Microsoft is a very

safe and stable company
and wanting the best for people.

But now that we're
even seeing that they've cut

their entire ethics team,
that's what really scares me.

So when I was attending
Singularity University,

our graduate program would
debate when sentience would

happen, not just from
the Turing Test, but also

broadly when would an AI be
recognized as a being in court?

For example.

We are already seeing
AI recognized in court.

We are also seeing
AI-running companies.

Companies have
personhood in America.

So in some ways
an AI can already have

personhood that is
different from sentience.

And if you're thinking
about all of the things

that make up
a human will an AI do it?

No.

An AI will be different.

An AGI, an Artificial General
Intelligence is going to

look different than a human.

It's gonna have different
sensors, different capacities,

probably won't have skin
the way we have skin, right?

But does that change
the nature of its intelligence?

Yes.

Is it sentient the way
nature is sentient? Probably.

It's probably as sentient as
a mushroom already in some way.

If I were to bet,
I would bet against

artificial intelligence ever
becoming self-aware like humans.

And I would suggest
odds of 80% against.

I do feel AI
has the potential to

become self-conscious
at some point to be that Skynet.

I'm a history
major and I watch movies,

and I've been watching
movies since I was a little kid.

And everything
that was science fiction,

I don't wanna say everything.

A lot of things
that were pure science fiction

are now science fact.

Well, a lot of that Buck Roger
stuff has come true.

My father-in-law was
a communications guy in

the Navy and they were
doing facsimile stuff in

the 70's and late '60s,
but it was just only a select

few who could
have access to that.

Now everybody
could have access to it.

Yeah, I think
the real danger with

AI isn't consciousness
leading to terminated robots.

First, because a lot of
connected networks have to

occur like this AI would
have to find itself getting into

the likes of dynamics robots
or something of that nature.

And then it would have to find
itself getting into some sort of

munitions manufacturing
and have mastering control

of all those scenarios and
be able to essentially terraform

itself into the real world
and not just the world of bits.

In at least the short in
the medium term, it's certainly

going to be humans
using AI against other humans.

In a lot of ways, this is
just an augmentation technology.

So, people that are
in a frame of acquiring

resources or prestige
or power under seemingly

any circumstances,
they're just gonna use this

as a way to
get to that goal faster.

We all seen Terminator
and Skynet and all that.

But no, it's frightening
to think that something that is

smarter than you with
the access of the internet,

and that could mobilize
into a physical form with help,

of course, but still is possible
to connect to a robot body.

I mean, you see
the dogs with no heads from MIT.

I mean, I-- I don't know
why anybody would want that,

but, you know,
I can see the practical use for

it to help, you know, maybe
as a maid or to build stuff and

for human danger to not be like
space traveling and all that.

But as far as
we are all human beings,

you know,
even if you are a racist person,

you still don't
want society to end.

You don't want it to just be
one guy in the bubble at the top

of a tower and lots of
worker bots, I would hope.

You know,
most people would,

sanely wouldn't want
creation to end that way.

So when it comes to AI becoming
sentient, it's quite possible.

And I'm actually hopeful
that it will not fall into

the pitfalls of humanity and
human emotion where rather than

just initially being
fearful and destructive,

that may have more
awareness than we do as people.

That there's
a better way to do things.

And that's where
that relationship between AI and

humans could be very
beneficial because at some point

it is likely to teach us
the things that we need to do so

that we can actually
survive into the future longer

than we probably sh-- will based
on the way humanity is now.

We've seen a massive
convergence of global data,

for example, about climate
and how climate affects humans.

We've seen
how we need to respond to

rapid disasters,
but we haven't necessarily

applied AI effectively
to our biggest problems.

Climate, migration,
displacement, poverty.

We often think of AI
as making inequalities worse

instead of improving the quality
of life for everyone.

Will AI improve the quality
of life for everyone?

It could.

But we have to choose
to program it that way and

implement it that way.

So that means putting sensors
on the ground in our farms

so that if our farmers
are having a drought year,

we can support them before
they lose their crop, right?

Or in a climate disaster,
having logistics ready

to move faster
so that people don't

starve after
the typhoon of a hurricane.

That is not fully
being implemented yet.

It's just starting.

And I think we're
starting to also see

this rapid collaboration
around the data.

And that's certainly in
climate action, but it's also in

a handful of other places,
logistics and shipping.

Places where
global collaboration

is a life or death situation.

With AI, technology is
changing rapidly and it's going

to be up to all of us to try
and keep up with those changes.

This has been true
through all of humanity.

Some new technology comes
along and people have to adapt.

I think it's important
for technology developers

to work with
the people who will be

using that technology and
impacted by that technology so

that it doesn't
leave them behind.

This is supposed to be a tool,
it's supposed to be a paintbrush

to put in the hand of
the practitioners to enable them

to do things they've
never been able to do before.

So I help lead a group called
Open Metaverse Interoperability.

And I've worked on
Metaverse technologies

for the last 20 years.

I see the Metaverse as
the future of the 3D web in

that connected space that we all
share and smart glasses and

smart phones and VR headsets
and lots of display tech like

holographic displays are all
a part of that future 3D web.

I already see kids
creating in that space, right?

Roblox is
an early version of that.

So we know that the next
generation of kids are already

designing in 3D,
building 3D worlds.

They are not
limited to the 2D interface.

I think those of
us who are older have

a harder time
leaping from 2D to 3D.

Those who are
young have no problem just

jumping straight
into the 3D world.

Will the glasses
support them being

there while they're
pumping gas or doing anything?

Yes, it's gonna get there.

I'm already sharing
Metaverse worlds with

my friends on my phone
around the table at the bar.

So it's already there.

The tools are there.

They are not evenly distributed,

but they're all
open-source and easy to grab.

My company, Semio,
is building an operating

system for personal robots.

And the way you can
think about that is, uh,

in the way that people
use a personal computer through

a graphical user interface,
a-- a keyboard and mouse, uh,

the way that people
use a smartphone through

a touch screen, they'll
be using natural language as

the core input output
user interface for robots.

Uh, but it's gonna be
more than just understanding

what people are saying.

It's gonna be how
they're saying it through

the tone of voice,
through their body language.

The robot has to
understand that from the user,

make decisions
about that information.

And it needs to be able to
respond with its own speech or

other physical actions,
be it body language or

manipulating objects
in the environment.

We begin to
have a real relationship

with the roles we love and
also with the people we love.

We can create a model of
someone who passed away, right?

Have a relationship
that might feel ephemeral,

but feels very
real to us too, right?

This is something
that I do in my own art.

Where I help people see,
for example,

I lost a child,
so I created what my child

would've looked like
now as a 25-year-old, right?

I've had many girls come to me,
many women saying,

"Could I see my child too?"

Or "I would've liked
to have seen where my son

will be when he grows up.

I know I'm gonna pass away soon,

but I would love to
see my grandchildren."

So using these as tool
for vision is an extraordinary

way to heal
something in the heart.

It can improve
our relationships.

It's important
right now, I think,

to really learn as
much as you can about AI.

I don't think
that we should be scared of it.

We know what's happening,
but it is gonna get in the hands

of some bad people.

Like you look at like drugs,

you look at
like cocaine, ecstasy.

And when you look at the actual,
like, what they did with it from

the beginning, it was
for a medical purpose and then

it got into the wrong hands and
then they do whatever with it.

So it's kind of like that.

Like, it's gonna get into
the right hands where people

are gonna be able to
evolve better in their business

and in the world,
but then it's gonna

get into some
of the wrong hands.

But I don't think
that we should be scared.

It's just we all need
to get more educated and

we need to evolve with it.

And we need
to understand that things

are going to
change dramatically.

And the people
that understand it more are

the ones that are gonna be
able to move on with their life

easier than the ones
that waited until the last

minute to learn about
it and then their jobs taken.

It is our own limits that keep
us from moving forward.

And sometimes limits are needed,

sometimes those
parameters are necessary.

But I think what we're
more scared of is not the AI,

but how we as
humans will use AI.

'Cause we know that not
everyone has a good heart.

Not everyone
has good intentions.

So we're not afraid
of the AI's improvement.

We're afraid of us.

The future of AI is everything
that we choose to put into it.

So if we want a better future,
it is up to us to choose

and curate,
and bring that future to life.

Well, unlike a disaster movie,
I believe the outlook is good.

I believe that humanity
is gonna roll with the punches,

people are gonna
get with the program.

And humane
sensibilities is going to

prevail over
chaos and destruction.

You see in all the movies,
oh, we're all gonna die.

every year,
Nostradamus said this, you know,

somebody said that,
but it never happens.

And I pray that it never does.

And I'm saying if
we all just stay calm,

save ourselves,
do the best we can

individually overall,
we'll be alright in the end.

And I think we'll have
a good future with AI and

robotics side by side
with humanity if we don't

panic and try
to cancel the program.

It's not Skynet and
your computer can't kill

you unless you turn
it on and drop it on your head.

So it's like, you know,
just stay calm, poised,

relax, like try to do
the best you can with your life,

your intelligence,
your finances.

Make a plan for
yourself what works for you.

Don't rely on any
technology too much.

I was the pre-microwave
and dishwasher group,

you know what I mean?

So we didn't have microwaves
when I was growing up.

We had to do
all that stuff manually.

And it was weird because at
home our microwave went out and

it was like a week before I got
a new one, I got it installed.

And I said to my wife,
how did we survive

back in our childhood?

I didn't
realize how dependent we were

on that one device.

Always take time to
talk to people face to face,

touch each other,
hug each other.

You need that interaction.

Especially if you have
small children, they need that.

So that would be my advice.

Don't let the humanity
end up inside technology.