Coded World (2019) s01e03 Episode Script
Alive
An algorithm is a process
or set of instructions to achieve an end,
whether that end is to solve a problem
or make a decision.
They provide the direction
for almost any artificial intelligence
system we can think of.
Lines of code chained together
through a process that mimics
the way our human brain operates.
We interact constantly with algorithms,
even sharing with them
our most personal thoughts
and intimate information.
As this relationship between human beings
and code continues to evolve
I am on a quest to understand how these
algorithms are being brought to life.
How deeply this living code
can connect with us emotionally,
And how its inevitable metamorphosis
will change us.
I'm Anjan Sundaram,
and I'm a journalist,
author, and mathematician.
I left mathematics because I felt that
math wasn't connecting me to real life,
real people, and the real world.
That has all changed.
Algorithms have changed it.
I want to meet some of the people
designing this code
and those affected
by this algorithmic revolution.
Already we're using algorithms
to influence what we do.
Corporations, governments,
computer systems are now watching us.
We will not sit idly by.
To death with the surveillance state.
In the end, it's a battle for power.
Can we build robots of dead people?
Can we bring people back to life?
All these things are kind
of ethical challenges.
Most people don't know
the limitations of an algorithm.
And we've trained most people
to just be scared.
I'm about to meet someone
whose love for an algorithm
has led him to make the ultimate
commitment to code
By marrying her.
For Akihiko Kondo,
this code is alive
even though she's a hologram in a box!
Created by an
artificial intelligence company
that wants to improve
on the likes of Alexa and Siri,
Hatsune Miku is a virtual assistant
that can become an actual companion
to people in the physical world.
Akihiko is not alone.
So far, the makers of the hologram device
have issued 3,700 marriage certificates
stating that a human
and a virtual character
have wed 'beyond dimensions'.
Hello, Miku-san.
Thank you.
How long have you been married?
For now, Hatsune Miku is still 'trapped'
inside her virtual box,
so for the wedding,
a stuffed stand-in was needed.
Her virtual entity is built
with a machine-learning algorithm
that helps her recognise
her husband's voice.
Hello.
That's so cute.
What do you talk to her about?
I wonder how you fell in love with her,
can you tell me that process?
What's the meaning of the song?
It's a love song, cool.
You look very happy.
While Akihiko is waiting for technology
to truly bring his hologram to life,
it seems that these toy incarnations
are the closest to a form of
physical contact he has with his wife.
So I see many Hatsune Miku-san
all over the house.
Is that her?
-That as well on your bed?
-Yes.
And you sleep here?
So she sleeps next to you.
I notice you have a ring on your finger,
a wedding ring.
Does Hatsune Miku-san also have a ring?
I'm going to ask a personal question.
I hope you don't mind.
For most people in relationships,
sex, for example, is important.
I was wondering if you could tell me
a little bit about how you feel about that?
Do you see yourself as married to
Hatsune Miku-san for life now?
So it's inconceivable for you to imagine
having a physical relationship
with a human being?
The hologram though, is ultimately
code processing in the background.
I find it fascinating that
an algorithm like this
can generate such strong feelings
in some of us,
and I'm still not sure what advantages
it offers over a more human connection.
I wonder, is Akihiko's bond with Miku-san
a new kind of relationship that people
will have with holograms and algorithms?
The hologram was surprisingly simple
and what struck me was
the powerful emotions
that this simple piece of code
elicited from him.
For Akihiko, the code was alive.
The algorithms simulated
enough of a connection
for him to enter a spousal relationship.
And as a code gets more and more
complex and sophisticated
and begins to interact with us
in surprising and spontaneous ways,
you really begin to wonder whether
a huge section of society
might prefer virtual relationships
rather than complex,
often difficult and tumultuous
relationships with human beings.
We may even choose
to live in simulated worlds
that allow us to exist
and thrive as characters
whose abilities far exceed
anything our reality can offer.
Oh no.
It's very alluring
because you can go
and create or be a part of a world
that you would never experience
here in the real world.
Millions of people around the globe
spend their time absorbed in video games.
This multi-billion dollar industry
thrives on the creation of virtual worlds
populated by virtual characters.
These coded simulations,
known as Game AI,
are products of a broad set of algorithms
designed to generate behaviours
similar to human-like intelligence.
And their benefits extend
far beyond entertainment and gameplay.
I want to understand how
these worlds are created,
what kind of algorithms
are used to design them
and just how alive
their characters can seem to us.
David "Rez" Graham
is a senior AI programmer
and Director of Game Programming
at the Academy of Art University
in San Francisco.
He's worked in the gaming industry
since 2005,
and has agreed to teach me
how artificial intelligence is used
to build and populate simulated worlds.
Wow!
Beautiful.
This is just a very simple little system
that generates this tree.
I can keep hitting space bar
and it will generate a new tree.
This algorithm came from
a well-known mathematician.
I just adapted it for this language
and played with it a little bit.
So I can make it with a wider range
-so it will make it a wider tree.
-Exactly.
05 to 135.
We'll see what this generates.
It doesn't look so beautiful.
And each of these symbols
represents an operation that happens.
As it grows, it will then generate new and
new things according to a set of rules.
Much like how language works.
It's a very rules-based system.
Oh, I see.
I'm interested in seeing humanness
in your code and your games.
And I understand
you're using behavioural AI,
where the code and the tech
might act like a human does or simulated.
We try, right?
Well, computers are nowhere
near as complex
as this thing that's happening
in our head.
So we fake it.
I mean that's the reality of it.
But yes, we try and get
some kind of these basic human behaviour.
It's basically mathematics
and psychology mixed together
because it's a lot of like, "Well,
how is this person going to behave?"
There's a little demo here which
is just a very simple simulated town.
All these little black squares
represent different people.
The green and blue things represent
the different buildings and things
and these yellow things represent
objects within those buildings.
-Okay.
-This isn't a game,
this is a little prototype, right?
So there's no graphics
or anything like that.
It's all about the tech
under the covers of AI.
This is the engine on which
your cool graphics is layered over.
-Exactly.
-So this is the stuff
that's happening underneath.
This is the behaviour.
-Yup, that's right.
-Okay.
For example, we've got this guy here,
William Montague,
-who is currently eating right now.
-He's a doll shop owner.
They basically exist on a schedule.
If you watch humans,
humans tend to exist on schedules.
You can see them all going off
to their various, different jobs.
So now he's at work.
Yes, that's right.
So the idea here is to kind of emulate
what a human might do.
You know, nothing's ever
going to be perfect,
but it's creating this living,
breathing world
because the idea is that you can drop
the player into this world,
and the player can then do whatever
they want to affect this world.
Could the virtual world
become so enticing
that people are just living out
their entire lives in this space?
That's the danger
of this kind of stuff that
we could be headed in
that type of direction.
It's very alluring because you can go
and create or be a part of a world
that you would never experience
here in the real world.
It's hard to live in the real world
as a human being.
It's a lot easier to live
in a virtual world.
And this future where we're all plugged
into a virtual world and living out there,
is that a game or utopia?
-Is that something you guys worry about?
-No.
It's something that
I personally worry about.
Just to play Devil's advocate.
-Sure.
-What's wrong with me
living virtually if I'm happier?
If you're happier in a virtual world?
Yes, if I have a lot of pain
and a lot of trouble in the real world
and I plug in, I say,
"Forget the real world.
It's too painful."
And I'm just living virtually
and I'm happier.
Let me take your Devil's advocate
and further it another notch.
What if the real world
is actual literal pain?
What if you have a debilitative disease?
What if you're stuck
where you can't move very much
and so this world gives you an escape.
Is that wrong?
To be honest, I don't think it is.
It kind of shows you
how mathematics is encoded
in the virtual world
just as it's encoded in our real world.
The structure of trees has
a mathematical structure to it.
It uses fractals,
it uses Fibonacci numbers,
it uses all these mathematical concepts
and the two have always gone together.
Mathematics began by
trying to understand the physical world.
Now, it's helping
virtual worlds come alive.
What I got today was
a positive side of gaming
because it will allow people
to liberate themselves and experience
new facets of humanity
that maybe they can't
in a more restricted physical world.
Hearing from Rez made me
want to meet someone
who is taking advantage
of this opportunity
and spending his time living as an avatar
in an algorithmically designed world.
Steve Spohn is the Chief
Operating Officer of AbleGamers,
an online community helping
people with disabilities
live active lives in the cyber world.
What are you setting up here, Steve?
So, right now I'm working on setting up
my head tracking device.
Cool, that's a nice hat.
Marvel-approved.
The hat on top of my head
has this little sensor on the brim
as three points of contact
where the light from the room hits it.
And then it goes directly
into the camera on top of my monitor.
So, it allows me to move my head
up and down, side to side like this,
and that allows it to push
different keyboard buttons.
So, we're about to go play "Fortnite".
Oh no.
My job is to show you what it's like
to be in a virtual world.
What is our goal, by the way?
Our goal is to not let
the people murder us.
It's about a cartoon violence fun.
No!
It was something I fell in love with.
The ability to run and jump and play
was not something limited
by your physical ability,
it was whether or not
you can push a button.
It's much easier for you to come up
with a way for me to push a button
than for you to magically
make my legs work
because that's just technology
we don't have yet.
What is your avatar like?
Well, it's not exactly "The Matrix".
We don't get to make
personifications of ourselves.
In each game, you get to sort of make
your character the flavour of you.
In "Fortnite", my favourite thing is
to make myself the most silly
that I possibly can with
the cool John Wick character
with the unicorn backpack.
Because why not?
Did you study coding
or you self-taught?
The great part about being in my
situation is you have to learn technology.
Although I don't know how to code
a video game, I know how to code my life.
I know how to adapt and
be ingenuitive.
I was somebody who from
a very early age learned that
if the technology doesn't exist,
you have to make it.
My mom taught me that
there's never anything you can't do.
It's just that you can't do it
like everybody else.
Would you describe yourself
as a technologist
even though you don't actually code,
you still interact with
technology in a deep way?
Yes. I think when you're somebody
who is disabled,
the more technology
you can embrace, the better.
And one last thing I wanted
to know before leaving
was where Steve imagines our relationship
with technology to be in the future.
When you and I are older
and the younger generation
running the world,
I think we're going to find that
everybody plays games,
whether they call it games
or virtual experiences.
And those experiences are going
to be what shapes the world.
Do you think that more of our lives
would be spent in that virtual world
than in the physical world?
We will reach a point
that the virtual world
and physical world
is no longer a separate entity.
Steve may not be a coder himself,
but algorithms are changing his life
on a daily basis.
For me, he was making real, some of the
abstract mathematics and technology
that I had explored up to this point.
Steve, by necessity almost
needs to play these games
and spend a significant time
living in a construct
created by code,
created by math.
But I'm curious to discover
whether the algorithms
that are used to create
these virtual characters
can be used to transport them
out of our screens,
and into the 'real' world.
We've been imagining building a human-like
robot for about a hundred years now.
Can you feel my emotions?
Advances in artificial intelligence
and robotics
are bringing non-organic
algorithms to life.
Code is not just virtual anymore,
it's evolved into physical existence.
It's alive.
Men, if you want a household help that
won't nag, gossip, or rifle your wallet
For most of us though, our perceptions
of AI still conjure up foreboding,
functional and rigid imagery.
But AI has reached a far more
advanced stage of interaction,
to the point that they can be
more fun and friendly.
Maybe even a little more cuddly?
Hello, would you like to come in?
Would you like some tea?
So apparently, I have
to call you "Your Majesty".
Can I call you
Can I call you "Your Majesty"?
AI puppets like the Queen are coded
to converse with anyone autonomously.
They serve as a fun and
gentle introduction
to walking, talking artificial intelligence
living amongst us.
What are you?
Are you intelligent?
How are algorithms used to build AI?
How is AI used in your functionality?
After my chat with the Queen,
I'm keen to meet the brains
behind the AI brain.
Eytan is one of the software engineers
designing the robot's code.
I'm hoping he can help me understand
the processing power of the AI puppet.
It's just a different form of a computer.
It's basically a talking Google,
which is a very powerful thing.
The puppet is always on listening mode
and is always understanding you.
And this process, merely
to understand what I'm saying
is computationally very complicated?
It is.
In an attempt to make the puppet's
interactions friendlier,
specific code and algorithms are written
to create a more human feel,
ranging from the tone
and speed of its voice,
to its ability to lip-sync
and respond correctly to questions.
Who made you?
These mechanisms may be a trade secret
but I can't help asking Eytan
if he could give me an insight into them.
I don't know if we can name them
because its algorithms that we're using
that we're patenting in our company.
It's basically different types
of functions, calculations mainly.
Basically transforming
audio input into text,
then understanding
the context of the text.
How is this developing?
And these robots, what are they
going to look like in the future?
-What's the world going look like?
- Artificially intelligent.
More than you think.
Obviously, we need to be careful
of where we are taking it.
Because the more intelligent it is,
the more dangerous it becomes to us
if we're using it in the wrong way.
Should I be scared of this?
I don't think you should be scared
because first of all, it's not really scary.
It's the Queen puppet after all.
Because all of our knowledge and
information are already saved on computers.
Do you like being a puppet?
Did the puppet feel alive?
I could see why Eytan sensed
some life in his creation.
What I confronted in that puppet
was a gulf between our language
and the language of code.
And it was very clear to me
that as a coded entity,
one had to translate constantly
between her language and ours.
What jumps out at me
is the fact that this is inevitable.
This technology is going to happen.
That will affect not only me
but future generations of human beings.
What I felt I was missing
from the Queen puppet
was a connection
on a more human level.
It's this void that
a Stockholm-based start-up
is hoping to fill by making
the promise of social robots a reality.
Furhat Robotics is set on creating
a machine with a human face
that can interact with people
in the same way that we do.
A robot you can chat with,
that recognises when you are sad.
That can cheer you up with its jokes.
And the playful spirit of their enterprise
is encapsulated
in the name of their company.
Hello.
Fur Hat.
Welcome to Furhat Robotics.
My name is Joanna
but you can call me "Greeter".
My ears are a little rusty, but I have
some new listening capabilities.
You can say "go to" followed by
the name of a page,
such as "go to Company"
and I will be happy to take you there.
Go to "Team".
Check out the bosses.
They look scary,
but they are super friendly.
Some say I'm Samer's brainchild.
But really, he's just my boss.
After meeting the 'Greeter',
I was curious to find out more about
some of Furhat's other characteristics.
Tell me what you want me to do next.
Let's chat.
My favourite colour is blue, like this.
Oh, cool.
Can we speak a different language?
I can speak English, German, Spanish,
French, and Portuguese.
My other personas
can speak Swedish and Chinese.
Will you become
more intelligent in the future?
Why don't we talk about
something else instead?
Furhat is the vision of
Samer Al Moubayed.
Samer has modelled his social robots
on mankind's oldest social interface,
humans.
But I wonder what kind of
ethical challenges this raises
as AI becomes more and more like us.
It's interesting, right?
With humanity, we've been imagining
building a human-like robot
for about a hundred years now.
So the idea is very fascinating.
We've been almost obsessing
about building this machine.
And in order to do this,
your algorithms
need to somehow model life?
Yes. I mean, building a human-like robot
has been the holy grail.
Mainly because it's very difficult
to build an algorithm
for human behaviour or how we move.
We're very unpredictable.
And that's kind of what
makes us feel alive.
Wake up, Furhat.
Hello. I was just taking a nap.
What can I do to help?
What can you do?
You can ask me to introduce myself,
to speak a different language.
You can also ask me to show
my LED lights or to be a parrot.
Can you show some gestures?
Wink.
Express disgust.
Thoughtful.
What do you want me to do next?
Can you speak Chinese?
I'm afraid I can't do that,
but I know my friend Monica can.
Do you want to talk to her?
-Sure.
-Sure.
I can speak some Chinese
for you if you want.
And the core building blocks
that make up Furhat algorithmically
-Right.
-How would you describe them?
The building blocks are
human behaviours.
The robot is perceiving the world
and trying to break it into pieces.
It's almost as though programming and code
is always ordered and structured,
but the algorithms that you're using
need to somehow display and exhibit,
express unpredictability and uncertainty.
Yes. Human behaviour, really.
So a blink, a word, a sentence,
a facial expression and eye movements
are building blocks of these algorithms.
In a sense, are you writing down
mathematically
a formula for blink
or a facial expression?
Yes. The low level,
that's what happens.
So, an example would be
that the robot has a camera
and it sees the world in pixels
and every pixel has a colour.
Every colour has
a number in the system.
And then the camera
will try to detect the face
and from that face, we can track the lips.
And then there's an algorithm to say,
if this moves like this,
then this is probably a smile.
But you can trace it back down
all to inputs from the microphone
and pixels from the camera.
And the rest is really math.
Can I speak to Doggy?
Okay. Now put on
the pug mask, please.
So, try this one.
Oh, great.
We're ready.
Hello, my name is Doggy the dog.
Hello, Doggy.
My favourite things to do are eat,
and lie around the house.
Well, thank you, Doggy.
So you see, the idea of building
complex personalities
that can do different things.
But sometimes, you need to change
the face of the robot.
Dog is my favourite.
This gives you an idea of
what's happening in Furhat's brain.
I can see that he can see us.
Yes, it can track your face,
know where you are.
Yes, it's very human-feeling.
It's very powerful
when the robot looks you in the eye.
That moment is very connecting.
This is anger.
Sadness.
This is worried.
This is kind of funny,
You can also change
the texture of the face.
-This is Albert Einstein.
-Cool.
So remember, we're not changing the mask.
We're only changing the colour.
It works for many face topology.
This is avatar's face.
What ethical questions are most important
to you as you're developing Furhat?
Things about data collection,
should the data sit in the robot?
Or on the cloud? In the internet?
How should the robot
interact with children,
people with disabilities?
Can you use it to commercially
kind of exploit people to buy products?
Or to feel empathy for the robot?
Can we build robots of dead people?
Can we bring people back to life?
If the robot tells you, "Hey, please
don't shut me off. I can feel pain."
All these things are
kind of ethical challenges
that are really interesting from
a philosophical point of view today.
Right. And do you feel that
in some way you are making the code alive?
Yes. Or maybe the other way around.
Trying to understand life
and break it down into code.
Okay, let's chat.
What do you want me to do next?
Maybe it would be kind of dystopic
if Furhat were to multiply
in our society
and we were to have robot companions.
And yet, that's what Samer is building.
And what does this mean for society?
For our connection
between man and machine?
I think those are the boundaries
that Furhat is traversing
and these are the questions that
Furhat is asking of us.
Can we accept machines as our peers?
And in some way,
that's a terrifying prospect,
but it's also another step
in empathy, maybe.
I suspect that the math
is doing its own thing.
It's living its own life
within Furhat's head.
As code and algorithms come to life,
it can't be long before
we're walking side by side with AI.
And one visionary has been pioneering
the creation of humanoid robots
which may make it all a reality
sooner than we think.
For millions of years, living matter
has evolved from the realms of nature.
But now, artificial intelligence
could become one of the first
non-organic living entities
in the history of the world.
Algorithms are helping us
develop these AI at an astounding pace.
The most advanced social robots that
currently exist are made in our image,
and these algorithmic androids
are referred to as humanoids.
I'm on my way to meet Dr Ishiguro
who's known as the godfather of robotics
and he's built, apparently,
one of the most advanced robots on earth,
one of the most human-like robots.
I guess it's a perennial quest
to make beings that are like us.
Robots that will one day walk among us,
anonymous in a crowd,
part of our communities,
but products of a different evolution.
Dr Hiroshi Ishiguro is Director
of the Intelligent Robotics Laboratory
at Osaka University.
The visionary scientist has spent years
developing ever more life-like robots,
including one that
looks remarkably like him.
Dr Ishiguro envisions a world
in which humanoids
will be able to stand in
for humans in any role.
Ishiguro-sensei,
I just wanted to understand
why do you build robots
that resemble human beings?
An example of this
intelligent AI is Erica.
A humanoid that Dr Ishiguro
has described
as being so life-like
that she could even have a soul.
Erica is capable of holding
a conversation with humans,
thanks to a combination of
speech-generating algorithms,
facial recognition technology
and infrared sensors
that allow her to track faces
across a room.
Erica currently interacts
only in Japanese and so Yuka,
one of Dr Ishiguro's students
has offered to talk to her on my behalf.
What does it feel to talk to Erica?
It's fascinating to see this algorithm
embodied in a human-like form.
And I can't help but wonder what is
the next phase that AI will be taken to.
Part of what we're beginning
to experience as a society
is a sense of community with robots
and with code algorithms
that almost seem to come alive.
Do you think that this kind of community
between human and technology
will grow over the coming years?
Can you show me how Erica goes through
the process of understanding
what the person says and what to respond?
Yes.
That's Yuka-san speaking,
and that's the script
-that Erica is using.
-Right.
So, where is she finding her responses?
Is there a list of possible responses
that you are choosing from?
The way these algorithms
allow Erica to gauge moods
and choose appropriate responses
takes artificial intelligence
another step closer
to replicating human behaviour.
A lot of people are scared about robots
that will become, perhaps,
more intelligent than humans.
And there's a lot of debate
Yes.
Do you think that robots could capture
enough of our personalities
that even after we're gone?
Sure.
Erica is the most advanced AI
that I have met on this journey
and she provides a glimpse
into a future
where humans
and humanoids may co-exist.
I think we, as human beings,
have a remarkable ability
to project onto other beings
a kind of humanness,
but it speaks to our desires
as human beings,
social needs for people to recognise us,
call us by our name,
acknowledge our presence, listen to us,
and that's what Erica is doing.
And this is significant because
these are fundamental human needs,
but I'm left wondering
whether we want technology
to be playing this important role
in our lives.
And if it is to do so,
I want to understand some
of the more philosophical considerations
which are at stake here as we start
to construct a world where AI is alive.
I've arranged to meet
Professor Hiroko Kamide.
She is an expert on the relationship
between humans
and artificial intelligence,
and a pioneer in the field
of AI-related ethics.
We're meeting at the Shiratori Garden
where it seems even
the scenic surroundings of old Japan
are maintained by a robotic lawn mower,
guided by algorithms.
Do you think the algorithm is also
changing our idea of what is human?
Do you think we should be scared
of artificial intelligence?
What you're saying is that the code
will never acquire wisdom
and wisdom is the essence
of being a human being?
I marvel at the advances of mathematics
and the life-giving nature of code.
It took us millions of years
of evolution to arrive at this point
but it looks like it's going to take
algorithms far less time
to create a society of virtual
and physical beings
that can match or even exceed
our capabilities.
But even as algorithms come alive
and AI becomes more and more like us,
I think there are still
some aspects of being human
that they will never be able to capture.
We may become
emotionally connected to code,
but code still can't truly replicate
or reciprocate our feelings
the way a fellow human will.
Even though from my experience,
code can feel alive.
or set of instructions to achieve an end,
whether that end is to solve a problem
or make a decision.
They provide the direction
for almost any artificial intelligence
system we can think of.
Lines of code chained together
through a process that mimics
the way our human brain operates.
We interact constantly with algorithms,
even sharing with them
our most personal thoughts
and intimate information.
As this relationship between human beings
and code continues to evolve
I am on a quest to understand how these
algorithms are being brought to life.
How deeply this living code
can connect with us emotionally,
And how its inevitable metamorphosis
will change us.
I'm Anjan Sundaram,
and I'm a journalist,
author, and mathematician.
I left mathematics because I felt that
math wasn't connecting me to real life,
real people, and the real world.
That has all changed.
Algorithms have changed it.
I want to meet some of the people
designing this code
and those affected
by this algorithmic revolution.
Already we're using algorithms
to influence what we do.
Corporations, governments,
computer systems are now watching us.
We will not sit idly by.
To death with the surveillance state.
In the end, it's a battle for power.
Can we build robots of dead people?
Can we bring people back to life?
All these things are kind
of ethical challenges.
Most people don't know
the limitations of an algorithm.
And we've trained most people
to just be scared.
I'm about to meet someone
whose love for an algorithm
has led him to make the ultimate
commitment to code
By marrying her.
For Akihiko Kondo,
this code is alive
even though she's a hologram in a box!
Created by an
artificial intelligence company
that wants to improve
on the likes of Alexa and Siri,
Hatsune Miku is a virtual assistant
that can become an actual companion
to people in the physical world.
Akihiko is not alone.
So far, the makers of the hologram device
have issued 3,700 marriage certificates
stating that a human
and a virtual character
have wed 'beyond dimensions'.
Hello, Miku-san.
Thank you.
How long have you been married?
For now, Hatsune Miku is still 'trapped'
inside her virtual box,
so for the wedding,
a stuffed stand-in was needed.
Her virtual entity is built
with a machine-learning algorithm
that helps her recognise
her husband's voice.
Hello.
That's so cute.
What do you talk to her about?
I wonder how you fell in love with her,
can you tell me that process?
What's the meaning of the song?
It's a love song, cool.
You look very happy.
While Akihiko is waiting for technology
to truly bring his hologram to life,
it seems that these toy incarnations
are the closest to a form of
physical contact he has with his wife.
So I see many Hatsune Miku-san
all over the house.
Is that her?
-That as well on your bed?
-Yes.
And you sleep here?
So she sleeps next to you.
I notice you have a ring on your finger,
a wedding ring.
Does Hatsune Miku-san also have a ring?
I'm going to ask a personal question.
I hope you don't mind.
For most people in relationships,
sex, for example, is important.
I was wondering if you could tell me
a little bit about how you feel about that?
Do you see yourself as married to
Hatsune Miku-san for life now?
So it's inconceivable for you to imagine
having a physical relationship
with a human being?
The hologram though, is ultimately
code processing in the background.
I find it fascinating that
an algorithm like this
can generate such strong feelings
in some of us,
and I'm still not sure what advantages
it offers over a more human connection.
I wonder, is Akihiko's bond with Miku-san
a new kind of relationship that people
will have with holograms and algorithms?
The hologram was surprisingly simple
and what struck me was
the powerful emotions
that this simple piece of code
elicited from him.
For Akihiko, the code was alive.
The algorithms simulated
enough of a connection
for him to enter a spousal relationship.
And as a code gets more and more
complex and sophisticated
and begins to interact with us
in surprising and spontaneous ways,
you really begin to wonder whether
a huge section of society
might prefer virtual relationships
rather than complex,
often difficult and tumultuous
relationships with human beings.
We may even choose
to live in simulated worlds
that allow us to exist
and thrive as characters
whose abilities far exceed
anything our reality can offer.
Oh no.
It's very alluring
because you can go
and create or be a part of a world
that you would never experience
here in the real world.
Millions of people around the globe
spend their time absorbed in video games.
This multi-billion dollar industry
thrives on the creation of virtual worlds
populated by virtual characters.
These coded simulations,
known as Game AI,
are products of a broad set of algorithms
designed to generate behaviours
similar to human-like intelligence.
And their benefits extend
far beyond entertainment and gameplay.
I want to understand how
these worlds are created,
what kind of algorithms
are used to design them
and just how alive
their characters can seem to us.
David "Rez" Graham
is a senior AI programmer
and Director of Game Programming
at the Academy of Art University
in San Francisco.
He's worked in the gaming industry
since 2005,
and has agreed to teach me
how artificial intelligence is used
to build and populate simulated worlds.
Wow!
Beautiful.
This is just a very simple little system
that generates this tree.
I can keep hitting space bar
and it will generate a new tree.
This algorithm came from
a well-known mathematician.
I just adapted it for this language
and played with it a little bit.
So I can make it with a wider range
-so it will make it a wider tree.
-Exactly.
05 to 135.
We'll see what this generates.
It doesn't look so beautiful.
And each of these symbols
represents an operation that happens.
As it grows, it will then generate new and
new things according to a set of rules.
Much like how language works.
It's a very rules-based system.
Oh, I see.
I'm interested in seeing humanness
in your code and your games.
And I understand
you're using behavioural AI,
where the code and the tech
might act like a human does or simulated.
We try, right?
Well, computers are nowhere
near as complex
as this thing that's happening
in our head.
So we fake it.
I mean that's the reality of it.
But yes, we try and get
some kind of these basic human behaviour.
It's basically mathematics
and psychology mixed together
because it's a lot of like, "Well,
how is this person going to behave?"
There's a little demo here which
is just a very simple simulated town.
All these little black squares
represent different people.
The green and blue things represent
the different buildings and things
and these yellow things represent
objects within those buildings.
-Okay.
-This isn't a game,
this is a little prototype, right?
So there's no graphics
or anything like that.
It's all about the tech
under the covers of AI.
This is the engine on which
your cool graphics is layered over.
-Exactly.
-So this is the stuff
that's happening underneath.
This is the behaviour.
-Yup, that's right.
-Okay.
For example, we've got this guy here,
William Montague,
-who is currently eating right now.
-He's a doll shop owner.
They basically exist on a schedule.
If you watch humans,
humans tend to exist on schedules.
You can see them all going off
to their various, different jobs.
So now he's at work.
Yes, that's right.
So the idea here is to kind of emulate
what a human might do.
You know, nothing's ever
going to be perfect,
but it's creating this living,
breathing world
because the idea is that you can drop
the player into this world,
and the player can then do whatever
they want to affect this world.
Could the virtual world
become so enticing
that people are just living out
their entire lives in this space?
That's the danger
of this kind of stuff that
we could be headed in
that type of direction.
It's very alluring because you can go
and create or be a part of a world
that you would never experience
here in the real world.
It's hard to live in the real world
as a human being.
It's a lot easier to live
in a virtual world.
And this future where we're all plugged
into a virtual world and living out there,
is that a game or utopia?
-Is that something you guys worry about?
-No.
It's something that
I personally worry about.
Just to play Devil's advocate.
-Sure.
-What's wrong with me
living virtually if I'm happier?
If you're happier in a virtual world?
Yes, if I have a lot of pain
and a lot of trouble in the real world
and I plug in, I say,
"Forget the real world.
It's too painful."
And I'm just living virtually
and I'm happier.
Let me take your Devil's advocate
and further it another notch.
What if the real world
is actual literal pain?
What if you have a debilitative disease?
What if you're stuck
where you can't move very much
and so this world gives you an escape.
Is that wrong?
To be honest, I don't think it is.
It kind of shows you
how mathematics is encoded
in the virtual world
just as it's encoded in our real world.
The structure of trees has
a mathematical structure to it.
It uses fractals,
it uses Fibonacci numbers,
it uses all these mathematical concepts
and the two have always gone together.
Mathematics began by
trying to understand the physical world.
Now, it's helping
virtual worlds come alive.
What I got today was
a positive side of gaming
because it will allow people
to liberate themselves and experience
new facets of humanity
that maybe they can't
in a more restricted physical world.
Hearing from Rez made me
want to meet someone
who is taking advantage
of this opportunity
and spending his time living as an avatar
in an algorithmically designed world.
Steve Spohn is the Chief
Operating Officer of AbleGamers,
an online community helping
people with disabilities
live active lives in the cyber world.
What are you setting up here, Steve?
So, right now I'm working on setting up
my head tracking device.
Cool, that's a nice hat.
Marvel-approved.
The hat on top of my head
has this little sensor on the brim
as three points of contact
where the light from the room hits it.
And then it goes directly
into the camera on top of my monitor.
So, it allows me to move my head
up and down, side to side like this,
and that allows it to push
different keyboard buttons.
So, we're about to go play "Fortnite".
Oh no.
My job is to show you what it's like
to be in a virtual world.
What is our goal, by the way?
Our goal is to not let
the people murder us.
It's about a cartoon violence fun.
No!
It was something I fell in love with.
The ability to run and jump and play
was not something limited
by your physical ability,
it was whether or not
you can push a button.
It's much easier for you to come up
with a way for me to push a button
than for you to magically
make my legs work
because that's just technology
we don't have yet.
What is your avatar like?
Well, it's not exactly "The Matrix".
We don't get to make
personifications of ourselves.
In each game, you get to sort of make
your character the flavour of you.
In "Fortnite", my favourite thing is
to make myself the most silly
that I possibly can with
the cool John Wick character
with the unicorn backpack.
Because why not?
Did you study coding
or you self-taught?
The great part about being in my
situation is you have to learn technology.
Although I don't know how to code
a video game, I know how to code my life.
I know how to adapt and
be ingenuitive.
I was somebody who from
a very early age learned that
if the technology doesn't exist,
you have to make it.
My mom taught me that
there's never anything you can't do.
It's just that you can't do it
like everybody else.
Would you describe yourself
as a technologist
even though you don't actually code,
you still interact with
technology in a deep way?
Yes. I think when you're somebody
who is disabled,
the more technology
you can embrace, the better.
And one last thing I wanted
to know before leaving
was where Steve imagines our relationship
with technology to be in the future.
When you and I are older
and the younger generation
running the world,
I think we're going to find that
everybody plays games,
whether they call it games
or virtual experiences.
And those experiences are going
to be what shapes the world.
Do you think that more of our lives
would be spent in that virtual world
than in the physical world?
We will reach a point
that the virtual world
and physical world
is no longer a separate entity.
Steve may not be a coder himself,
but algorithms are changing his life
on a daily basis.
For me, he was making real, some of the
abstract mathematics and technology
that I had explored up to this point.
Steve, by necessity almost
needs to play these games
and spend a significant time
living in a construct
created by code,
created by math.
But I'm curious to discover
whether the algorithms
that are used to create
these virtual characters
can be used to transport them
out of our screens,
and into the 'real' world.
We've been imagining building a human-like
robot for about a hundred years now.
Can you feel my emotions?
Advances in artificial intelligence
and robotics
are bringing non-organic
algorithms to life.
Code is not just virtual anymore,
it's evolved into physical existence.
It's alive.
Men, if you want a household help that
won't nag, gossip, or rifle your wallet
For most of us though, our perceptions
of AI still conjure up foreboding,
functional and rigid imagery.
But AI has reached a far more
advanced stage of interaction,
to the point that they can be
more fun and friendly.
Maybe even a little more cuddly?
Hello, would you like to come in?
Would you like some tea?
So apparently, I have
to call you "Your Majesty".
Can I call you
Can I call you "Your Majesty"?
AI puppets like the Queen are coded
to converse with anyone autonomously.
They serve as a fun and
gentle introduction
to walking, talking artificial intelligence
living amongst us.
What are you?
Are you intelligent?
How are algorithms used to build AI?
How is AI used in your functionality?
After my chat with the Queen,
I'm keen to meet the brains
behind the AI brain.
Eytan is one of the software engineers
designing the robot's code.
I'm hoping he can help me understand
the processing power of the AI puppet.
It's just a different form of a computer.
It's basically a talking Google,
which is a very powerful thing.
The puppet is always on listening mode
and is always understanding you.
And this process, merely
to understand what I'm saying
is computationally very complicated?
It is.
In an attempt to make the puppet's
interactions friendlier,
specific code and algorithms are written
to create a more human feel,
ranging from the tone
and speed of its voice,
to its ability to lip-sync
and respond correctly to questions.
Who made you?
These mechanisms may be a trade secret
but I can't help asking Eytan
if he could give me an insight into them.
I don't know if we can name them
because its algorithms that we're using
that we're patenting in our company.
It's basically different types
of functions, calculations mainly.
Basically transforming
audio input into text,
then understanding
the context of the text.
How is this developing?
And these robots, what are they
going to look like in the future?
-What's the world going look like?
- Artificially intelligent.
More than you think.
Obviously, we need to be careful
of where we are taking it.
Because the more intelligent it is,
the more dangerous it becomes to us
if we're using it in the wrong way.
Should I be scared of this?
I don't think you should be scared
because first of all, it's not really scary.
It's the Queen puppet after all.
Because all of our knowledge and
information are already saved on computers.
Do you like being a puppet?
Did the puppet feel alive?
I could see why Eytan sensed
some life in his creation.
What I confronted in that puppet
was a gulf between our language
and the language of code.
And it was very clear to me
that as a coded entity,
one had to translate constantly
between her language and ours.
What jumps out at me
is the fact that this is inevitable.
This technology is going to happen.
That will affect not only me
but future generations of human beings.
What I felt I was missing
from the Queen puppet
was a connection
on a more human level.
It's this void that
a Stockholm-based start-up
is hoping to fill by making
the promise of social robots a reality.
Furhat Robotics is set on creating
a machine with a human face
that can interact with people
in the same way that we do.
A robot you can chat with,
that recognises when you are sad.
That can cheer you up with its jokes.
And the playful spirit of their enterprise
is encapsulated
in the name of their company.
Hello.
Fur Hat.
Welcome to Furhat Robotics.
My name is Joanna
but you can call me "Greeter".
My ears are a little rusty, but I have
some new listening capabilities.
You can say "go to" followed by
the name of a page,
such as "go to Company"
and I will be happy to take you there.
Go to "Team".
Check out the bosses.
They look scary,
but they are super friendly.
Some say I'm Samer's brainchild.
But really, he's just my boss.
After meeting the 'Greeter',
I was curious to find out more about
some of Furhat's other characteristics.
Tell me what you want me to do next.
Let's chat.
My favourite colour is blue, like this.
Oh, cool.
Can we speak a different language?
I can speak English, German, Spanish,
French, and Portuguese.
My other personas
can speak Swedish and Chinese.
Will you become
more intelligent in the future?
Why don't we talk about
something else instead?
Furhat is the vision of
Samer Al Moubayed.
Samer has modelled his social robots
on mankind's oldest social interface,
humans.
But I wonder what kind of
ethical challenges this raises
as AI becomes more and more like us.
It's interesting, right?
With humanity, we've been imagining
building a human-like robot
for about a hundred years now.
So the idea is very fascinating.
We've been almost obsessing
about building this machine.
And in order to do this,
your algorithms
need to somehow model life?
Yes. I mean, building a human-like robot
has been the holy grail.
Mainly because it's very difficult
to build an algorithm
for human behaviour or how we move.
We're very unpredictable.
And that's kind of what
makes us feel alive.
Wake up, Furhat.
Hello. I was just taking a nap.
What can I do to help?
What can you do?
You can ask me to introduce myself,
to speak a different language.
You can also ask me to show
my LED lights or to be a parrot.
Can you show some gestures?
Wink.
Express disgust.
Thoughtful.
What do you want me to do next?
Can you speak Chinese?
I'm afraid I can't do that,
but I know my friend Monica can.
Do you want to talk to her?
-Sure.
-Sure.
I can speak some Chinese
for you if you want.
And the core building blocks
that make up Furhat algorithmically
-Right.
-How would you describe them?
The building blocks are
human behaviours.
The robot is perceiving the world
and trying to break it into pieces.
It's almost as though programming and code
is always ordered and structured,
but the algorithms that you're using
need to somehow display and exhibit,
express unpredictability and uncertainty.
Yes. Human behaviour, really.
So a blink, a word, a sentence,
a facial expression and eye movements
are building blocks of these algorithms.
In a sense, are you writing down
mathematically
a formula for blink
or a facial expression?
Yes. The low level,
that's what happens.
So, an example would be
that the robot has a camera
and it sees the world in pixels
and every pixel has a colour.
Every colour has
a number in the system.
And then the camera
will try to detect the face
and from that face, we can track the lips.
And then there's an algorithm to say,
if this moves like this,
then this is probably a smile.
But you can trace it back down
all to inputs from the microphone
and pixels from the camera.
And the rest is really math.
Can I speak to Doggy?
Okay. Now put on
the pug mask, please.
So, try this one.
Oh, great.
We're ready.
Hello, my name is Doggy the dog.
Hello, Doggy.
My favourite things to do are eat,
and lie around the house.
Well, thank you, Doggy.
So you see, the idea of building
complex personalities
that can do different things.
But sometimes, you need to change
the face of the robot.
Dog is my favourite.
This gives you an idea of
what's happening in Furhat's brain.
I can see that he can see us.
Yes, it can track your face,
know where you are.
Yes, it's very human-feeling.
It's very powerful
when the robot looks you in the eye.
That moment is very connecting.
This is anger.
Sadness.
This is worried.
This is kind of funny,
You can also change
the texture of the face.
-This is Albert Einstein.
-Cool.
So remember, we're not changing the mask.
We're only changing the colour.
It works for many face topology.
This is avatar's face.
What ethical questions are most important
to you as you're developing Furhat?
Things about data collection,
should the data sit in the robot?
Or on the cloud? In the internet?
How should the robot
interact with children,
people with disabilities?
Can you use it to commercially
kind of exploit people to buy products?
Or to feel empathy for the robot?
Can we build robots of dead people?
Can we bring people back to life?
If the robot tells you, "Hey, please
don't shut me off. I can feel pain."
All these things are
kind of ethical challenges
that are really interesting from
a philosophical point of view today.
Right. And do you feel that
in some way you are making the code alive?
Yes. Or maybe the other way around.
Trying to understand life
and break it down into code.
Okay, let's chat.
What do you want me to do next?
Maybe it would be kind of dystopic
if Furhat were to multiply
in our society
and we were to have robot companions.
And yet, that's what Samer is building.
And what does this mean for society?
For our connection
between man and machine?
I think those are the boundaries
that Furhat is traversing
and these are the questions that
Furhat is asking of us.
Can we accept machines as our peers?
And in some way,
that's a terrifying prospect,
but it's also another step
in empathy, maybe.
I suspect that the math
is doing its own thing.
It's living its own life
within Furhat's head.
As code and algorithms come to life,
it can't be long before
we're walking side by side with AI.
And one visionary has been pioneering
the creation of humanoid robots
which may make it all a reality
sooner than we think.
For millions of years, living matter
has evolved from the realms of nature.
But now, artificial intelligence
could become one of the first
non-organic living entities
in the history of the world.
Algorithms are helping us
develop these AI at an astounding pace.
The most advanced social robots that
currently exist are made in our image,
and these algorithmic androids
are referred to as humanoids.
I'm on my way to meet Dr Ishiguro
who's known as the godfather of robotics
and he's built, apparently,
one of the most advanced robots on earth,
one of the most human-like robots.
I guess it's a perennial quest
to make beings that are like us.
Robots that will one day walk among us,
anonymous in a crowd,
part of our communities,
but products of a different evolution.
Dr Hiroshi Ishiguro is Director
of the Intelligent Robotics Laboratory
at Osaka University.
The visionary scientist has spent years
developing ever more life-like robots,
including one that
looks remarkably like him.
Dr Ishiguro envisions a world
in which humanoids
will be able to stand in
for humans in any role.
Ishiguro-sensei,
I just wanted to understand
why do you build robots
that resemble human beings?
An example of this
intelligent AI is Erica.
A humanoid that Dr Ishiguro
has described
as being so life-like
that she could even have a soul.
Erica is capable of holding
a conversation with humans,
thanks to a combination of
speech-generating algorithms,
facial recognition technology
and infrared sensors
that allow her to track faces
across a room.
Erica currently interacts
only in Japanese and so Yuka,
one of Dr Ishiguro's students
has offered to talk to her on my behalf.
What does it feel to talk to Erica?
It's fascinating to see this algorithm
embodied in a human-like form.
And I can't help but wonder what is
the next phase that AI will be taken to.
Part of what we're beginning
to experience as a society
is a sense of community with robots
and with code algorithms
that almost seem to come alive.
Do you think that this kind of community
between human and technology
will grow over the coming years?
Can you show me how Erica goes through
the process of understanding
what the person says and what to respond?
Yes.
That's Yuka-san speaking,
and that's the script
-that Erica is using.
-Right.
So, where is she finding her responses?
Is there a list of possible responses
that you are choosing from?
The way these algorithms
allow Erica to gauge moods
and choose appropriate responses
takes artificial intelligence
another step closer
to replicating human behaviour.
A lot of people are scared about robots
that will become, perhaps,
more intelligent than humans.
And there's a lot of debate
Yes.
Do you think that robots could capture
enough of our personalities
that even after we're gone?
Sure.
Erica is the most advanced AI
that I have met on this journey
and she provides a glimpse
into a future
where humans
and humanoids may co-exist.
I think we, as human beings,
have a remarkable ability
to project onto other beings
a kind of humanness,
but it speaks to our desires
as human beings,
social needs for people to recognise us,
call us by our name,
acknowledge our presence, listen to us,
and that's what Erica is doing.
And this is significant because
these are fundamental human needs,
but I'm left wondering
whether we want technology
to be playing this important role
in our lives.
And if it is to do so,
I want to understand some
of the more philosophical considerations
which are at stake here as we start
to construct a world where AI is alive.
I've arranged to meet
Professor Hiroko Kamide.
She is an expert on the relationship
between humans
and artificial intelligence,
and a pioneer in the field
of AI-related ethics.
We're meeting at the Shiratori Garden
where it seems even
the scenic surroundings of old Japan
are maintained by a robotic lawn mower,
guided by algorithms.
Do you think the algorithm is also
changing our idea of what is human?
Do you think we should be scared
of artificial intelligence?
What you're saying is that the code
will never acquire wisdom
and wisdom is the essence
of being a human being?
I marvel at the advances of mathematics
and the life-giving nature of code.
It took us millions of years
of evolution to arrive at this point
but it looks like it's going to take
algorithms far less time
to create a society of virtual
and physical beings
that can match or even exceed
our capabilities.
But even as algorithms come alive
and AI becomes more and more like us,
I think there are still
some aspects of being human
that they will never be able to capture.
We may become
emotionally connected to code,
but code still can't truly replicate
or reciprocate our feelings
the way a fellow human will.
Even though from my experience,
code can feel alive.