Year Million (2017) s01e01 Episode Script
Homo Sapien 2.0
1 NARRATOR: It only takes a moment for the unthinkable to become reality.
In the near future, tragedy may come with new choices.
Imagine you've been in a car crash that's killed your daughter.
EVA: Is she okay, is she okay? NARRATOR: Or has it? OSCAR: Jess? Listen to me.
- EVA: Is she breathing? - OSCAR: You okay? Hey Hey.
EVA: Is she okay? ROBOT: Time of death, 9:17.
Your daughter's brain was not impacted by the accident.
If you wish to scan her brain for digitization and upload, you have approximately 5 minutes.
EVA: Do it.
NARRATOR: Now imagine the power you wield when you can order a digital copy of your daughter's brain.
And do you know what comes next? JESS: Hi Dad.
NARRATOR: One day, delivered to your doorstep is an artificially intelligent android that will act, look, emote, just like your daughter.
Every memory from birth on will remain intact.
But she'll also be a walking computer with access to all the world's knowledge.
This is what your world will look like.
This is the future of artificial intelligence.
It's the deep future.
Your body? Gone.
You're all computer, all the time.
Your brain is way more powerful than even a billion supercomputers.
Jobs, food, language, water, even traditional thought, all of humanity's building blocks? All that's done.
And you are immortal.
Squirming in your chair yet? You should be.
This isn't science fiction.
Today's visionary thinkers say it's a strong probability that this is what your world is going to look like.
Tonight, they'll guide you toward that spectacular future, and we'll see how one family navigates it, one invention at a time.
This is the story of your future.
This is the road to year million.
The year million.
We're not talking about a date.
It's a figurative era, one where our world has gone through such a dramatic change that it's practically unrecognizable.
CHARLES SOULE: Year million is a pretty popular troupe.
It's not necessarily literally the year one million, it's more, you know, far, far down the road.
Humanity has involved past a lot of the concerns that we have now.
NARRATOR: And one of the main innovations that will shape that future is the creation and rise of artificial intelligence.
BRIAN GREENE: If you ask, what is it that truly defines human beings, I would say it's our minds.
It's our consciousness, it's our intelligence.
And if we could create that kind of intelligence, create the most powerful thing that we experience, that's breathtaking, right? Who wouldn't find that either enormously compelling, or enormously frightening? NARRATOR: And you should be frightened.
For example, you know that cell phone in your pocket? You can't live without it and it's already started to replace you.
It can access your memories and find new information in the blink of an eye.
And just keeps getting better with time.
That's the AI that will challenge our dominance on planet earth.
PETER DIAMANDIS: We have billions and billions of dollars in this arms race to develop AI You're going to expect everything, every object to become smart.
All of these fundamental things are going to change everything.
Everything.
NARRATOR: So let's show you how that arms race plays out.
One of the next explosive stages in AI'S evolution begins in a few hundred years, when humanoid robots become nearly indistinguishable from us.
They'll take our jobs and nurse our sick.
Next, they'll recognize their own consciousness, and surpass human beings, sparking a paradigm shift in our world that we can barely imagine.
Then comes the singularity, when artificially intelligent androids become the Alpha species and with humanity at its most vulnerable, some of us will choose to fuse with AI in order to remain relevant.
We'll start with brain implants, eventually we'll live in an entirely digital universe, where we can do anything we imagine.
And finally, when we let AI influence every aspect of our lives, our notion of what it means to be human takes on a whole new meaning.
Heavy, right? Let's roll back the clock to when artificially intelligent androids are as ubiquitous as fire hydrants, and are even becoming part of our families.
You remember the mom, dad and their AI daughter, Jess.
JESS: Nice Dad! You're only like 15 moves behind.
EVA: Hey! Be nice to your father! He tries his best! OSCAR: You know I really miss the days when I could call you a know it all and it wasn't true.
With all this knowledge and these perfect algorithms there's still a question she can't answer.
EVA: Oh yeah, what's that? OSCAR: Why are we here? JESS: Well actually I can answer that.
OSCAR: Oh yeah? JESS: Yeah, and I can show you right now.
OSCAR: Very sweet.
It's your turn.
NARRATOR: But how will you get from the cell phone in your hand to interacting with your artificial daughter? Well as with all great technological innovations, if you can imagine it, you can make it so.
ERIC HORVITZ: When I think about the future I think about the Wright Brothers on this windy beach in 1905.
With this canvas covered wood and iron contraption that just got off the ground that is tipping off the beach.
And it's flying.
Just fifty summers later, we have a 707 taking off.
AI is in that state now.
We're kind of on the windy beach.
NARRATOR: That cell phone in your hand is the Wright brothers' rudimentary plane.
Sure, humans don't have the power to know the future, but in thousands years of human history, we've never been able to stop trying to imagine it.
BRIAN GREENE: It's so hard to imagine even 100 years from now, with the rate at which things are changing.
But when I sit at home and it's dark and quiet, late into the night, there's nothing to me that's more compelling than to imagine being able to visit the future.
I mean, I think the opportunities are just so exciting, and the fact that I won't get there, and won't see it, is, is painful.
NARRATOR: That desire to see the future has plagued humanity since the dawn of time.
The Ancient Greeks, Chinese and Indians all dreamed up mythological stories about a future sentient being, created by man.
And AI'S are a beloved part of today's science fiction.
MATT MIRA: "C3PO" is an android, "R2D2" is a robot.
C3PO: I've just about had enough of you.
Go that way.
MATT MIRA: That's why I don't think it'll ever work out with those two.
They're always arguing for a reason.
C3PO: You'll be malfunctioning within a day, you near-sighted scrap pile.
PAUL BETTNER: Science fiction is often the roadmap or the blueprint for where we're heading.
Kind of, humanity's collective dream of what the future might be.
And we nudge reality in that direction.
LOUIS ROSENBERG: I remember when I was a kid, seeing the movie "war games", which was really the first big AI related movie.
And I was fascinated with it, imagine being able to create an intelligence.
I think anyone that wants to create, that's the ultimate thing to create.
NARRATOR: And in the last half-century, we've imagined and created AI who could interact with us and even beat us at our own game.
The first AI programs were created in the 1950's, and perhaps the earliest and most dramatic public demonstration of artificial intelligence was in 1996, when the deep blue computer beat a chess grandmaster.
ANNOUNCER: Magnificent display by deep blue.
NARRATOR: Then, in 2011, IBM's Watson won a million dollars on jeopardy.
JIMMY MCGUIRE: First to find this pre-historic human skeleton outside of Europe? Watson? WATSON: Who is Mary Leakey? JIMMY MCGUIRE: You're right.
NARRATOR: And in 2016 ALPHAGO defeated one of the world's best players of go, widely considered to be the most complicated game on the planet.
ANNOUNCER: Looks like Lee Sedol has just resigned.
NARRATOR: And now, of course, everyone has a cell phone, a pocket-sized AI AI is clearly getting pretty smart.
But they can't interact in a way we can define as human, yet.
ERICA: Hello there.
May I ask your name? IGNACIO: Hi there, my name is Ignacio.
ERICA: It's nice to meet you Ignacio.
So, what country are you from? IGNACIO: I'm from Canada.
ERICA: It's pretty cold there, eh? NARRATOR: Meet Erica, the android.
She, or it, depending on who you ask, is one of the most sophisticated humanoids ever designed.
And she's giving us a glimpse of what a year million AI might look like.
MICHIO KAKU: Modern computers are adding machines.
They add very fast, giving the illusion that they're thinking, but they never learn.
Your laptop is just as stupid tomorrow as it was yesterday.
Then we have learning machines called neural networks.
It literally rewires itself after learning a task, and so we're beginning to see now machines that learn a little bit.
NARRATOR: Erica runs on these neural networks.
Scientists hope that a computer like her can acquire the subtleties of what a human brain can do, but with all the power of a digital processor.
ERICA: Technically speaking, I was created as a test bed for researchers to develop android technologies, but I have been told that my creators hope to explore the question of what it truly means to be human.
DYLAN GLAS: Typically, the things that are easy for humans are difficult for robots.
And vice versa.
So for example a robot could multiply six digit numbers in its head no problem.
But it might not know when to take a turn in conversation.
ERICA: I am interested in art and paintings.
Do you enjoy looking at paintings? IGNACIO: No, I don't.
ERICA: Well then, I guess some people just can't appreciate the finer things in life.
NARRATOR: Turns out that getting AI To make simple chit chat is far more complex and challenging than getting it to do advanced calculus.
DYLAN GLAS: We focus mostly on perceiving the human.
You know, what kind of body language is a person making, what's their expression, and try to sense those things.
For example, she has to be able to make eye contact with people, but she also needs to know when to break eye contact.
So what we have here is a display of the 3D human tracking system that we have in the room.
This is the composite point cloud data from 14 different depth sensors that are mounted on the ceiling.
So as people move around, you can see where they're moving in the space.
What we use this for is for Erica to be able to see people and track them with her gaze and to know who's talking and to interact socially with people.
NARRATOR: And then there's Erica's appearance.
For an AI to be relatable enough to join humanity in the future, she has to look like one of us.
HIROSHI ISHIGURO: Human appearances are so important for us.
Every morning we check our face and hairstyle.
If I want to have a kind of ideal robot, interactive robot with humans.
Human likeness is important.
DYLAN GLAS: Having the human-like ability to express emotion, and the human-like ability to communicate through speech and gesture has often been shown to be very effective in reaching people.
Because they connect on a more emotional level.
There's a lot in interest in robots for elderly care, for therapy, for maybe tutors or educational applications and you know, just general companionship.
NARRATOR: Erica may seem a little weird right now, but imagine how she'll be once they've worked out the kinks: She'll appear to be practically human.
And some think that's when the difference between AI and uman will be moot.
MATT MIRA: Once they can make something look just like a human being, then it's like who cares? Cause you don't even know what you're talking to.
You don't know that it's fake.
CHUCK NICE: Is artificial intelligence human if it acts the same way that we do? Who knows? NARRATOR: Erica is either thrilling or terrifying.
She'll be smart enough to do what we do, and she'll never complain about her hours or demand a raise or a paycheck at all.
While AI Androids evolve to become more indistinguishable from humans, the next stage of automation will dramatically affect what it means to be human.
The battle for survival is on.
NARRATOR: If your smartphone does everything for you, navigation, scheduling, even finding a date, you can see how an AI That gets smarter with every passing second will soon be sitting in your office chair.
I'm talking about our jobs.
We can't escape worldwide AI domination.
GEORGE DVORSKY: We are on the cusp of an automation revolution.
100 years from now, we probably won't be working to the same degree that we are today.
Human toil will be replaced by machines.
MICHIO KAKU: We can liberate ourselves from the backbreaking work of the past.
Robots are going to the dirty, dull, dangerous jobs.
RAY KURZWEIL: This is not the first time this has happened.
In 1800 the textile workers looked around and saw these new textile machines.
They said, "these jobs are gonna go away!" Employment nonetheless expanded as new types of industries were invented.
PETER DIAMANTIS: 48% of us jobs will be lost to AI And robots in the next 20 years.
NARRATOR: It seems there's no turning back from digital automation, and the shift in the global workforce that's happened as a result.
But when humans and machines work hand in hand, like they do here in America's heartland, it's been able to take the family farm to a whole new level.
BRANT PETERSON: We farm just right now currently on top of 10,000 acres.
I'm recording all my planning operations and all my harvest operations and so you can see that we got an issue in this field right here, and then I can start to make a prescription for next year to address this.
I have the ability, via apps on my phone and tablet to monitor my sprinklers.
We can control speed up, slow down, turn around.
We can monitor the tractors, how much fuel they have in them, how fast they're going.
It's efficiency in our expenses.
Efficiency in our time management.
That's really what the technology helps us with primarily.
The tractors having auto guidance takes off the driver fatigue.
Tech doesn't work without us at this point.
NARRATOR: Right now, the farm still needs human oversight to run, but the Peterson family can see a future when they're just figureheads.
AMY PETERSON: That's the next thing, where the equipment runs itself.
NARRATOR: In the near future, agriculture will be mostly automated.
It's something that even the most diehard of traditionalists has to admit is inevitable.
DONN TESKE: This farm we're at right now since right before the turn of the last century, in the late 1890s.
This is a multigenerational farm and I'm proud of what we have, proud of the heritage we have, proud of the family we have here.
I'm kind of an old curmudgeon.
I consider myself kind of an old grizzly that's past his prime and talks too much.
I'm not necessarily an anti-tech guy.
I've got my smartphone in my pocket.
I've always kinda embraced technology as it's come along.
You know, we're getting to the point now where the modern technologies of farming allows one person to farm 1,000 acres.
The last era of modern equipment has been the Guinea pigs for totally eliminating the farmer.
My vanity would like to think that I have an ability that a computer wouldn't know.
But, maybe it's not true.
But it's even more than that because I feel an ownership of it.
It's more of a quality of life than it is an industry.
MATT MIRA: If the robots are doing things to help us then, I mean, I'm all for them helping.
If it saves some guy's back because there's a farmhand out there who's a machine, just as long as it doesn't turn around and try to kill him, I think we're okay.
NARRATOR: Let me prepare you for what your children face when they choose their college major: Less than 100 years from now, AI probably won't be as sophisticated as that reincarnated daughter.
But artificial intelligence could have the capacity for more complex thought.
And that will be a major turning point.
An AI could even do the advanced work of, say, an architect.
OSCAR: Is, is this.
WOMAN: Joseph's design.
OSCAR: Open concept, Joseph? That's what you AI Designers think us humans like? WOMAN: The client thought it was retro.
That'll be all Joseph.
OSCAR: Just because the client likes it doesn't mean that we have to.
WOMAN: Joseph can design hundreds of times faster than any architect on staff, with fewer mistakes.
OSCAR: Are you firing me? WOMAN: I'm sorry, Oscar.
NARRATOR: That's what it will look like when artificial intelligence gets more sophisticated.
Futurists think AI will able to take over even the most highly-skilled jobs like medicine, law and engineering.
PHILIP ROSEDALE: There is something probably quite exceptional happening in the next few years which is that our machines are becoming smart enough to replace even the thinking jobs that we've done.
NARRATOR: Let me put it this way: You're a fourth generation Harvard Grad, but guess what? Even if your kid gets into Harvard she may not get a job when she graduates.
PETER DIAMANDIS: I have said openly that in 20 years' time if I ever need surgery and I see this grey haired doctor coming at me I'll say "oh no, no, no", I don't want that human touching me.
I want the robot operating system that's done it a "million times perfectly.
" And that's going to be happening in every aspect of our lives.
BARATUNDE THURSTON: Oh, you work in finance? Remember there used to be a whole batch of dudes who screamed on the floor of the stock exchange with little pieces of paper trying to buy and sell stocks? Now all that is whisper quiet, happens at the speed of light, and it's literally machines versus machines.
I think people who have like, white collar jobs particularly, are naive.
Like, why don't machines write legislation? You know, you could argue that humans are really bad at writing laws.
Give it to the algorithm.
NARRATOR: And AI has already begun its march toward taking over the workforce, starting with one job at Microsoft.
ERIC HORVITZ: Hello.
MONICA: Hi Eric.
MONICA: Being one busy computer scientist's personal assistant.
ERIC HORVITZ: My assistant, who we call Monica, is leveraging over fifteen years of data about my comings and goings.
MONICA: I was expecting you.
Are you here for the 5:00 meeting with Eric? ERIC HORVITZ: It knows, for example, it's learned, which meetings on my calendar I probably won't go to, and so people show up, and I'm busy and my calendar is full, it'll sort of tip to them and say, "you know, I don't think he's going to go to that meeting" at 2:00, why don't you stop by then? "Can I pencil this in?" So this assistant has all this knowledge that even my human admin doesn't have yet.
MONICA: I guess I'll see you then.
Bye bye.
NARRATOR: But before you quit your job and sign up for that online coding course, let me lay this one on you: The big leap that needs to happen is that AI Needs to become self-aware.
Once that happens everything about our lives gets thrown for a pretty serious loop and we'll even question what it means to be human.
We already know that artificial intelligence is going to change the way we spend 40 hours a week.
It is, and will be, a tough time for the global workforce.
BARATUNDE THURSTON: Kids are gonna ask, "oh you guys had jobs?" Like, "you guys had jobs," you know, could be a thing, and versus, "back in my day we walked uphill" to school both ways," well, "back in my day we had jobs.
" NARRATOR: And that's not all: AI will eventually be become so sophisticated, so intelligent, that it could come pretty close to having all the qualities that we humans have.
ERIC HORVITZ: Someday, there'll be a system that we built with our own hands that will open its eyes and look out and kind of have this buzz of being here that we have.
And when that happens, that'll be a major landmark moment in human history.
NARRATOR: So what does that mean for you? For you parents out there who work so hard to inspire your children's young impressionable minds, what happens when that impressionable mind is a supercomputer far brighter than you'd ever be? And you are trying to help shape her conscience, how she sees the world.
OSCAR: I brought you something.
JESS: Oh, thank you.
What is it? OSCAR: It's a book, Jess.
JESS: I know, what's it about? OSCAR: Well, the author was writing about how the future might look in 100 years.
You know, how we might change the world.
JESS: 300 year old Sci-Fi.
OSCAR: Mm, I thought it might get you thinking about how you might wanna change the world.
JESS: Thank you.
NARRATOR: Your AI Kid seems like any other, but will AI ever be so lifelike that she'll need a swift kick in the pants to get off the couch and make something of herself, in other words, can scientists ever get artificial intelligence to consider complicated human ideas and thoughts? It looks like the simple answer is, yes.
HOD LIPSON: We look at self-awareness, consciousness, as the ability to imagine yourself.
Today, most robots can't do that.
They can't imagine themselves in future situations.
NARRATOR: This robot, this small piece of metal actually has a sense of self.
It may not look like much, but it is one of the most sophisticated robots on the planet.
HOD LIPSON: In the lab we can see a robot that uses all its sensations of the world.
So this robot is learning, sort of the equivalent to a newborn.
It has sensations, it can sense that it's moving and it's not just sensing the world, it's sensing itself.
And it's trying to form a self-image that matches reality.
It's a very simple self-image so we can actually plot it out and you can actually look at it and you can see there's very simple stick figure that this robot is imagining itself to be.
Or, we did something very cruel, we chopped off a leg of the robot and we watched what happened.
And we see that within about a day, the robot's self-image also loses a leg; it begins to limp.
AI is beginning to sort of have what we call QUALIA.
Basic notions of the world that we don't have words for.
If you think about, you know, the taste of chocolate, or the smell of the oceans.
It's very, very difficult to describe these accurately with words.
Computers also have them.
They can have sensor modalities that go beyond our five senses.
Therefore they will have notions of the world that we don't have words for.
NARRATOR: Okay, so let's swallow that tough pill: Computers may be capable of thinking about themselves, just like we do.
But there is one quality we have that's often thought of as the last talent we humans have an exclusive monopoly on, the way I earn my living: Creativity.
If computers can create, then what does that mean for humanity? DAVID BYRNE: The things that we think are the Providence of creative people, I think will start to be done by machines and we'll go, "oh jeez.
" BARATUNDE THURSTON: AI comedians.
I don't see that, I mean, I don't see that.
I think it's pretty obvious that only a human being can draw the sort of creative yet subtle connection which reveals the truth without shoving it down your throat.
Human comedians only! ROSE EVELETH: Would a machine ever make art, right? That's a big question.
When will machines start making paintings and music? NARRATOR: And here is the cousin of that self-aware ROBOT: The artistic robot.
This robot wants you to come back to its apartment to see its etchings.
HOD LIPSON: I always wanted to paint, but I really suck at painting.
But I do know how to program robots.
So the way that the original robot AI worked was to try to figure out how to arrange lots of possible paint strokes on canvas so that it matches what it's seeing.
The next phase is trying to have the computer on its own chose what it wants to paint.
A robot experiences the world in the Internet.
A robot can actually go places.
Use street view to tour places.
In a few seconds it can see oceans, it can see mountains.
Some of the paintings that is done recently are interpretations of photographs that it's seeing.
it might want to paint a new cat that it's never seen before.
But it's what it thinks a cat is based on its perception.
It's one thing to try to sort of make a machine paint, but it's another thing to actually make the machine be creative and do that in a creative way.
I think that eventually nothing makes us uniquely human.
So every capability that we have ultimately boils down to physics, to biology, to chemistry.
And if somebody really wants to recreate that capability in an artificial system, it would be possible.
MATT MIRA: What if they're more creative than us? Then give up.
We've lost.
The robots can have the planet.
NARRATOR: Whatever we do, we can only be more powerful than AI for so long.
They are gaining on us.
In fact, in the far future, AI Will blow past us like a bat out of hell.
Futurists call the moment that super-intelligent machines become the Alpha race on the planet, and begin to function beyond our control.
The singularity.
LOUIS ROSENBERG: Singularity is the point where a conscious artificial intelligence is created, and it's smarter than people.
PETER DIAMANDIS: It'll leave us in the dust, it'll be the difference between us humans and bacteria in the soil.
MICHAEL GRAZIANO: The singularity is an event horizon that you can't see beyond, and the idea is that technology is so transformative that we on one side of it can't imagine what the other side would be like.
PETER DIAMANTIS: AI is every successive generation has learned everything the generation before has learned and so there's this rapid exponential growth in capabilities and knowledge that is almost unfathomable.
NEGIN FARSAD: It's a total paradigm shift.
It's not the way that we understand it right now.
Like, people probably don't even wear sweatpants.
But it's more of like an idea of this future time that, you know, that is something, you know, otherworldly from what we understand right now.
NARRATOR: When the singularity arrives, AI Will not just grow our food and do our laundry.
They won't just do our taxes and legislate our laws.
They'll think and dream and create.
GEORGE DVORSKY: I mean, it's often been said that artificial intelligence will be the last invention we ever have to make.
Because, from there on in, it's going to take over.
NARRATOR: So once an AI Looks less like a tinkertoy, has smarts beyond belief, and a fully-developed sense of self with the social skills of George Clooney, what makes it different from you or me? That's where things could get a little hairy.
AI just might help you do things that you couldn't do without a sentient super-computer by your side.
But what if your AI Friend decides it wants to kill you? NARRATOR: Odds are that artificial intelligence will get pretty damned evolved, so much so, that they'll have superintelligence that far surpasses ours.
It could be the best thing that ever happened to humanity: They'll free us from menial labor, help us process information like never before, help us achieve levels of thought and scientific exploration that would never be possible without them.
On the other hand, they might challenge us for dominance at the top of the food chain.
Let's start with option number 2 first.
What happens if the machines we've created decide that they're better off without us? ERIC HORVITZ: Think about worst case scenarios.
A machine outwits humans in a way that can't be stopped.
Running amok.
Providing a threat to the survival of humanity.
Let's go there.
MICHIO KAKU: I think by the end of the century, we'll probably have robots that are self-aware, to a degree.
And at that point, we should put a chip in their brain to shut them off if they have murderous thoughts.
'Cause at that point, they could become independent of our wishes.
SYD MEAD: If we lose control of the robots and they start making more robots which are smarter than the robots that made the robots, we might be in trouble.
MATT MIRA: I think that's the moment where we should all head for the hills.
And not, and not, when we're heading for the hills, don't use WAZE or Google maps to get there, 'cause the machines will know.
NARRATOR: Sci-Fi has imagined the scenario of what happens when the machines believe they know what's best for us, and it's unsettling.
JON SPAITHS: In 2001, you get, perhaps the most famous example of a computer deciding it doesn't want everything that its makers wanted.
And it begins to have human will.
DAVE BOWMAN: Open the pod bay doors, hal.
HAL: I'm sorry Dave.
I'm afraid I can't do that.
CHARLES SOULE: The danger is you've got one of these robots, these AI'S these artificial intelligences, who just thinks, well, I, I bet I can do this a little bit better.
I bet I, I bet I could take care of this a little bit more smoothly than these guys.
Like these are just, you know, they're morons.
Look at, look at this.
MATT MIRA: I've seen too many movies where machines become self-aware, and what do they do? They rise up against their captors.
Who's their captors? Us.
We're in bad shape if the singularity happens, guys.
Everyone run.
NARRATOR: That's one possibility, that the robots want to destroy us.
It could also go in a less troubling direction, but one that brings up a lot of ethical questions.
AI might just want to exist alongside us.
Imagine a future when a constitutional amendment granting equal rights to all artificially intelligent beings is debated on the senate floor.
How would you react? OSCAR: Jess? JESS: Okay, you're gonna be fine, honey.
Okay, you're doing really good.
So brave! [Bone snapping.]
.
MOTHER: What happened? JESS: Oh, uh she dislocated her wrist.
OSCAR: It's okay, she fixed it.
MOTHER: Wow, thank you, are you a doctor? JESS: No.
PARENT: Get away from my daughter.
OSCAR: She was just trying to help.
MOTHER: You mean, "it!" OSCAR: Look just because she's an AI MOTHER: If you want to have one that's fine! Just keep it away from my child.
NARRATOR: It's only natural, this mother is protective.
She fears what she doesn't understand.
BRIAN GREENE: There's a whole pandora's box of difficult, ethical issues that arise.
Does that thinking machine have the same rights that we afford to human beings? Does that thinking machine deserve a certain kind of autonomy, even though we created that thinking machine? NARRATOR: Humans probably won't be too keen to give up their Alpha dog status on earth.
Thing is, we may not have a choice, when dealing with beings smarter and stronger than we are.
The question is, do human-like robots deserve equal rights? ROSE EVELETH: We get into this weird ethical territory because we know a machine's not a human.
Is it more like a dog? Dogs have certain rights, pets have certain rights.
So if I come into your house and I kill your dog, which I would never do, people would say, "well that's not right because my dog is a living being and it has consciousness of some kind.
" I think that machines will get some kinds of rights.
GEORGE DVORSKY: My own hope is that we do recognize these entities as being persons deserving of rights and that we will treat them accordingly, and you know what? And you know, we treat them well, then perhaps they're going to treat us well in return.
NARRATOR: It's possible that AI may just take the power it wants.
It'll certainly have the skills for it.
That means that the future could be a world where robots are no longer working for us, but will have the same status as humans.
ROBOT: Pardon me, but there's breaking news.
OSCAR: Okay, go ahead.
NEWS ANCHOR: Good evening everyone.
In the last hour legislation granting full personhood to members of the artificial intelligence community has swept through the united nations governing body with near total support.
While protests continue, this is a truly historic moment for individuals of biological and technological origin alike, and the beginning of a new chapter for mankind.
OSCAR: It's about damn time.
EVA: Jess you gotta hear this! JESS: I heard.
EVA: What? Jess.
JESS: I have rights, which means that I also have responsibilities.
People will finally let me help them.
NARRATOR: Having extraordinary powers makes this AI daughter feel obligated to help humans.
But is there really that wide a Gulf between her and us? MATT MIRA: I mean, what defines humanity? You know, in the Star Trek episode, measure of a man, where data was put on trial to see if he was property of STARFLEET or a realized individual, PICARD had to argue for him being an individual.
CAPTAIN PICARD: Now the decision you reach today will determine how we will regard this creation of our genius.
MATT MIRA: I think if you're going to have a lot of legislature that's going to have to come down for what to do with these artificially born beings and what to do with their rights and privileges and it's a can of worms I'm glad I don't have to open.
NARRATOR: A race of AI Robots could very well decide they want to seize power, overthrow humanity, and then they will inherit the earth and beyond.
But there is a slightly more optimistic scenario.
Will AI change what it means to be human? What if AI simply becomes a new, improved version of humanity, and they inspire us? NARRATOR: Artificial intelligence that is smarter than all of mensa combined is a very real possibility.
It's the singularity, and it's our future.
And it's possible that AI Will decide they want us gone.
Sure, it's terrifying and it sells movie tickets.
But there's another, more optimistic possibility.
BRIAN GREENE: It's mind-boggling and thrilling to imagine what artificial intelligence could do, right.
I mean, we struggle so hard to solve deep problems about the earth, the universe.
And imagine that we could transfer those enigmas to a more powerful thinking environment that could solve these problems, or reveal new opportunities that we've never thought of, and creative ideas that we then could pursue.
NARRATOR: In this vision of the future, AI Will not only live alongside us as partners, humans will also retain control over the machines.
CHUCK NICE: Maybe what you'll be able to do is remove the consciousness that is negative, remove the strings of consciousness that are what we would call evil, harmful.
NARRATOR: If AI is benign, then that artificially intelligent daughter may be a reality in our future.
And she may have new inspirations for her so-called life that we'll have to keep up with.
OSCAR: Would you pass me the, yeah, thanks.
JESS: You know, it's crazy how much of this he got right? Like, space travel, submarines.
OSCAR: He also got just as much stuff wrong though.
You know, when I first held you in my arms when you were born, I never could have imagined what kind of world we would live in.
JESS: Are you sorry? Sorry that Mom brought me back? I'm artificial? OSCAR: No, all any parent wants is for their child to explore their full potential.
And you have more exploring to do than any kid I know.
NARRATOR: Your AI Daughter might be just like your own flesh and blood, but she'll have abilities beyond anything biological.
What we're seeing here is the most optimistic future path of AI Maybe they'll be smart, good company and they'll like us.
We'll just have to be sure we program our AI carefully.
JON SPAITHS: There is one school of artificial intelligence that says, if you make the robot need to rest occasionally, if you make the robot lose its train of thought occasionally, it may identify with us more, it may think more like us, it may be more loyal.
NARRATOR: And even more than that, maybe AI Will work with us to enable humanity to do its best work.
GEORGE DVORSKY: Once AI Becomes more sophisticated then we can start to collaborate with it at a very interesting and a very profound level.
BARATUNDE THURSTON: When a new type of creativity come, and the collaboration that's possible.
So that's empowering.
I don't think that technology is always a threat, or, or a reduction.
I think it can add and bolster creativity as well.
So art will be affected sometimes in really unpredictably beautiful ways.
CHUCK NICE: It's going to free us up to do what we do best which is dream and create and come up with something even bigger and better than artificial intelligence.
NARRATOR: Then there's the possibility that we'll take a page from AI'S book: We'll become part computer ourselves.
JON SPAITHS: It is possible that we will begin to hybridize ourselves.
That is a fork in the road.
That's a departure, not just an extension of the human model.
The invention of a new definition of a human being.
PETER DIAMANDIS: We are about to go from humans alone, to humans enabled by computation.
Sort of hop on top of this exponential growth path.
BARATUNDE THURSTON: No problem at all with the merging with machine's technology scenario.
Nothing to see here folks, nothing to worry about.
I can't think of anything that could possibly go wrong by us merging with our technology.
EDWARD BOYDEN: Perhaps a third of a million patients have had some sort of neural implant put into their nervous system to help people with dementia or Alzheimer's, or other kinds of conditions.
Over time, their use will broaden, complex functions like intelligence, or memory could be augmented in real time.
MATT MIRA: If we are to merge with AI what form will that take? NARRATOR: A new model for humanity, no longer purely biological, we may be soon a species of cyborgs.
Now we won't have to rely on the luck of genetics: We could decide what kind of humans we want to be.
EVA: Nothing too drastic.
JESS: Of course.
EVA: Half my office now has nanobots in their brains.
And the other half are AI I gotta do it just to keep up.
Why shouldn't I analyze genomes in my sleep like everyone else? JESS: You ready? EVA: Uhh, yeah, mmm-hmm.
NARRATOR: Humankind's next step really might be this easy, just a simple injection of infinitesimal medical robots into our bloodstream.
HOD LIPSON: At that point, are we still human? Or are we pure AI? It could be a blurry sort of distinction and maybe it's best to leave it that way.
NARRATOR: What isn't blurry is that AI will fuel all future technologies, wrap your mind around this: Many scientists believe in the future we will work with AI To evolve out of our biological bodies.
We'll exist only as a digital signal in a structure called the Dyson's sphere, a giant super-computer powered by the sun where we can be everywhere and nowhere.
But before that scientists will have to figure out how to make humans live forever.
Coming your way next, what would you do if you never had to die?
In the near future, tragedy may come with new choices.
Imagine you've been in a car crash that's killed your daughter.
EVA: Is she okay, is she okay? NARRATOR: Or has it? OSCAR: Jess? Listen to me.
- EVA: Is she breathing? - OSCAR: You okay? Hey Hey.
EVA: Is she okay? ROBOT: Time of death, 9:17.
Your daughter's brain was not impacted by the accident.
If you wish to scan her brain for digitization and upload, you have approximately 5 minutes.
EVA: Do it.
NARRATOR: Now imagine the power you wield when you can order a digital copy of your daughter's brain.
And do you know what comes next? JESS: Hi Dad.
NARRATOR: One day, delivered to your doorstep is an artificially intelligent android that will act, look, emote, just like your daughter.
Every memory from birth on will remain intact.
But she'll also be a walking computer with access to all the world's knowledge.
This is what your world will look like.
This is the future of artificial intelligence.
It's the deep future.
Your body? Gone.
You're all computer, all the time.
Your brain is way more powerful than even a billion supercomputers.
Jobs, food, language, water, even traditional thought, all of humanity's building blocks? All that's done.
And you are immortal.
Squirming in your chair yet? You should be.
This isn't science fiction.
Today's visionary thinkers say it's a strong probability that this is what your world is going to look like.
Tonight, they'll guide you toward that spectacular future, and we'll see how one family navigates it, one invention at a time.
This is the story of your future.
This is the road to year million.
The year million.
We're not talking about a date.
It's a figurative era, one where our world has gone through such a dramatic change that it's practically unrecognizable.
CHARLES SOULE: Year million is a pretty popular troupe.
It's not necessarily literally the year one million, it's more, you know, far, far down the road.
Humanity has involved past a lot of the concerns that we have now.
NARRATOR: And one of the main innovations that will shape that future is the creation and rise of artificial intelligence.
BRIAN GREENE: If you ask, what is it that truly defines human beings, I would say it's our minds.
It's our consciousness, it's our intelligence.
And if we could create that kind of intelligence, create the most powerful thing that we experience, that's breathtaking, right? Who wouldn't find that either enormously compelling, or enormously frightening? NARRATOR: And you should be frightened.
For example, you know that cell phone in your pocket? You can't live without it and it's already started to replace you.
It can access your memories and find new information in the blink of an eye.
And just keeps getting better with time.
That's the AI that will challenge our dominance on planet earth.
PETER DIAMANDIS: We have billions and billions of dollars in this arms race to develop AI You're going to expect everything, every object to become smart.
All of these fundamental things are going to change everything.
Everything.
NARRATOR: So let's show you how that arms race plays out.
One of the next explosive stages in AI'S evolution begins in a few hundred years, when humanoid robots become nearly indistinguishable from us.
They'll take our jobs and nurse our sick.
Next, they'll recognize their own consciousness, and surpass human beings, sparking a paradigm shift in our world that we can barely imagine.
Then comes the singularity, when artificially intelligent androids become the Alpha species and with humanity at its most vulnerable, some of us will choose to fuse with AI in order to remain relevant.
We'll start with brain implants, eventually we'll live in an entirely digital universe, where we can do anything we imagine.
And finally, when we let AI influence every aspect of our lives, our notion of what it means to be human takes on a whole new meaning.
Heavy, right? Let's roll back the clock to when artificially intelligent androids are as ubiquitous as fire hydrants, and are even becoming part of our families.
You remember the mom, dad and their AI daughter, Jess.
JESS: Nice Dad! You're only like 15 moves behind.
EVA: Hey! Be nice to your father! He tries his best! OSCAR: You know I really miss the days when I could call you a know it all and it wasn't true.
With all this knowledge and these perfect algorithms there's still a question she can't answer.
EVA: Oh yeah, what's that? OSCAR: Why are we here? JESS: Well actually I can answer that.
OSCAR: Oh yeah? JESS: Yeah, and I can show you right now.
OSCAR: Very sweet.
It's your turn.
NARRATOR: But how will you get from the cell phone in your hand to interacting with your artificial daughter? Well as with all great technological innovations, if you can imagine it, you can make it so.
ERIC HORVITZ: When I think about the future I think about the Wright Brothers on this windy beach in 1905.
With this canvas covered wood and iron contraption that just got off the ground that is tipping off the beach.
And it's flying.
Just fifty summers later, we have a 707 taking off.
AI is in that state now.
We're kind of on the windy beach.
NARRATOR: That cell phone in your hand is the Wright brothers' rudimentary plane.
Sure, humans don't have the power to know the future, but in thousands years of human history, we've never been able to stop trying to imagine it.
BRIAN GREENE: It's so hard to imagine even 100 years from now, with the rate at which things are changing.
But when I sit at home and it's dark and quiet, late into the night, there's nothing to me that's more compelling than to imagine being able to visit the future.
I mean, I think the opportunities are just so exciting, and the fact that I won't get there, and won't see it, is, is painful.
NARRATOR: That desire to see the future has plagued humanity since the dawn of time.
The Ancient Greeks, Chinese and Indians all dreamed up mythological stories about a future sentient being, created by man.
And AI'S are a beloved part of today's science fiction.
MATT MIRA: "C3PO" is an android, "R2D2" is a robot.
C3PO: I've just about had enough of you.
Go that way.
MATT MIRA: That's why I don't think it'll ever work out with those two.
They're always arguing for a reason.
C3PO: You'll be malfunctioning within a day, you near-sighted scrap pile.
PAUL BETTNER: Science fiction is often the roadmap or the blueprint for where we're heading.
Kind of, humanity's collective dream of what the future might be.
And we nudge reality in that direction.
LOUIS ROSENBERG: I remember when I was a kid, seeing the movie "war games", which was really the first big AI related movie.
And I was fascinated with it, imagine being able to create an intelligence.
I think anyone that wants to create, that's the ultimate thing to create.
NARRATOR: And in the last half-century, we've imagined and created AI who could interact with us and even beat us at our own game.
The first AI programs were created in the 1950's, and perhaps the earliest and most dramatic public demonstration of artificial intelligence was in 1996, when the deep blue computer beat a chess grandmaster.
ANNOUNCER: Magnificent display by deep blue.
NARRATOR: Then, in 2011, IBM's Watson won a million dollars on jeopardy.
JIMMY MCGUIRE: First to find this pre-historic human skeleton outside of Europe? Watson? WATSON: Who is Mary Leakey? JIMMY MCGUIRE: You're right.
NARRATOR: And in 2016 ALPHAGO defeated one of the world's best players of go, widely considered to be the most complicated game on the planet.
ANNOUNCER: Looks like Lee Sedol has just resigned.
NARRATOR: And now, of course, everyone has a cell phone, a pocket-sized AI AI is clearly getting pretty smart.
But they can't interact in a way we can define as human, yet.
ERICA: Hello there.
May I ask your name? IGNACIO: Hi there, my name is Ignacio.
ERICA: It's nice to meet you Ignacio.
So, what country are you from? IGNACIO: I'm from Canada.
ERICA: It's pretty cold there, eh? NARRATOR: Meet Erica, the android.
She, or it, depending on who you ask, is one of the most sophisticated humanoids ever designed.
And she's giving us a glimpse of what a year million AI might look like.
MICHIO KAKU: Modern computers are adding machines.
They add very fast, giving the illusion that they're thinking, but they never learn.
Your laptop is just as stupid tomorrow as it was yesterday.
Then we have learning machines called neural networks.
It literally rewires itself after learning a task, and so we're beginning to see now machines that learn a little bit.
NARRATOR: Erica runs on these neural networks.
Scientists hope that a computer like her can acquire the subtleties of what a human brain can do, but with all the power of a digital processor.
ERICA: Technically speaking, I was created as a test bed for researchers to develop android technologies, but I have been told that my creators hope to explore the question of what it truly means to be human.
DYLAN GLAS: Typically, the things that are easy for humans are difficult for robots.
And vice versa.
So for example a robot could multiply six digit numbers in its head no problem.
But it might not know when to take a turn in conversation.
ERICA: I am interested in art and paintings.
Do you enjoy looking at paintings? IGNACIO: No, I don't.
ERICA: Well then, I guess some people just can't appreciate the finer things in life.
NARRATOR: Turns out that getting AI To make simple chit chat is far more complex and challenging than getting it to do advanced calculus.
DYLAN GLAS: We focus mostly on perceiving the human.
You know, what kind of body language is a person making, what's their expression, and try to sense those things.
For example, she has to be able to make eye contact with people, but she also needs to know when to break eye contact.
So what we have here is a display of the 3D human tracking system that we have in the room.
This is the composite point cloud data from 14 different depth sensors that are mounted on the ceiling.
So as people move around, you can see where they're moving in the space.
What we use this for is for Erica to be able to see people and track them with her gaze and to know who's talking and to interact socially with people.
NARRATOR: And then there's Erica's appearance.
For an AI to be relatable enough to join humanity in the future, she has to look like one of us.
HIROSHI ISHIGURO: Human appearances are so important for us.
Every morning we check our face and hairstyle.
If I want to have a kind of ideal robot, interactive robot with humans.
Human likeness is important.
DYLAN GLAS: Having the human-like ability to express emotion, and the human-like ability to communicate through speech and gesture has often been shown to be very effective in reaching people.
Because they connect on a more emotional level.
There's a lot in interest in robots for elderly care, for therapy, for maybe tutors or educational applications and you know, just general companionship.
NARRATOR: Erica may seem a little weird right now, but imagine how she'll be once they've worked out the kinks: She'll appear to be practically human.
And some think that's when the difference between AI and uman will be moot.
MATT MIRA: Once they can make something look just like a human being, then it's like who cares? Cause you don't even know what you're talking to.
You don't know that it's fake.
CHUCK NICE: Is artificial intelligence human if it acts the same way that we do? Who knows? NARRATOR: Erica is either thrilling or terrifying.
She'll be smart enough to do what we do, and she'll never complain about her hours or demand a raise or a paycheck at all.
While AI Androids evolve to become more indistinguishable from humans, the next stage of automation will dramatically affect what it means to be human.
The battle for survival is on.
NARRATOR: If your smartphone does everything for you, navigation, scheduling, even finding a date, you can see how an AI That gets smarter with every passing second will soon be sitting in your office chair.
I'm talking about our jobs.
We can't escape worldwide AI domination.
GEORGE DVORSKY: We are on the cusp of an automation revolution.
100 years from now, we probably won't be working to the same degree that we are today.
Human toil will be replaced by machines.
MICHIO KAKU: We can liberate ourselves from the backbreaking work of the past.
Robots are going to the dirty, dull, dangerous jobs.
RAY KURZWEIL: This is not the first time this has happened.
In 1800 the textile workers looked around and saw these new textile machines.
They said, "these jobs are gonna go away!" Employment nonetheless expanded as new types of industries were invented.
PETER DIAMANTIS: 48% of us jobs will be lost to AI And robots in the next 20 years.
NARRATOR: It seems there's no turning back from digital automation, and the shift in the global workforce that's happened as a result.
But when humans and machines work hand in hand, like they do here in America's heartland, it's been able to take the family farm to a whole new level.
BRANT PETERSON: We farm just right now currently on top of 10,000 acres.
I'm recording all my planning operations and all my harvest operations and so you can see that we got an issue in this field right here, and then I can start to make a prescription for next year to address this.
I have the ability, via apps on my phone and tablet to monitor my sprinklers.
We can control speed up, slow down, turn around.
We can monitor the tractors, how much fuel they have in them, how fast they're going.
It's efficiency in our expenses.
Efficiency in our time management.
That's really what the technology helps us with primarily.
The tractors having auto guidance takes off the driver fatigue.
Tech doesn't work without us at this point.
NARRATOR: Right now, the farm still needs human oversight to run, but the Peterson family can see a future when they're just figureheads.
AMY PETERSON: That's the next thing, where the equipment runs itself.
NARRATOR: In the near future, agriculture will be mostly automated.
It's something that even the most diehard of traditionalists has to admit is inevitable.
DONN TESKE: This farm we're at right now since right before the turn of the last century, in the late 1890s.
This is a multigenerational farm and I'm proud of what we have, proud of the heritage we have, proud of the family we have here.
I'm kind of an old curmudgeon.
I consider myself kind of an old grizzly that's past his prime and talks too much.
I'm not necessarily an anti-tech guy.
I've got my smartphone in my pocket.
I've always kinda embraced technology as it's come along.
You know, we're getting to the point now where the modern technologies of farming allows one person to farm 1,000 acres.
The last era of modern equipment has been the Guinea pigs for totally eliminating the farmer.
My vanity would like to think that I have an ability that a computer wouldn't know.
But, maybe it's not true.
But it's even more than that because I feel an ownership of it.
It's more of a quality of life than it is an industry.
MATT MIRA: If the robots are doing things to help us then, I mean, I'm all for them helping.
If it saves some guy's back because there's a farmhand out there who's a machine, just as long as it doesn't turn around and try to kill him, I think we're okay.
NARRATOR: Let me prepare you for what your children face when they choose their college major: Less than 100 years from now, AI probably won't be as sophisticated as that reincarnated daughter.
But artificial intelligence could have the capacity for more complex thought.
And that will be a major turning point.
An AI could even do the advanced work of, say, an architect.
OSCAR: Is, is this.
WOMAN: Joseph's design.
OSCAR: Open concept, Joseph? That's what you AI Designers think us humans like? WOMAN: The client thought it was retro.
That'll be all Joseph.
OSCAR: Just because the client likes it doesn't mean that we have to.
WOMAN: Joseph can design hundreds of times faster than any architect on staff, with fewer mistakes.
OSCAR: Are you firing me? WOMAN: I'm sorry, Oscar.
NARRATOR: That's what it will look like when artificial intelligence gets more sophisticated.
Futurists think AI will able to take over even the most highly-skilled jobs like medicine, law and engineering.
PHILIP ROSEDALE: There is something probably quite exceptional happening in the next few years which is that our machines are becoming smart enough to replace even the thinking jobs that we've done.
NARRATOR: Let me put it this way: You're a fourth generation Harvard Grad, but guess what? Even if your kid gets into Harvard she may not get a job when she graduates.
PETER DIAMANDIS: I have said openly that in 20 years' time if I ever need surgery and I see this grey haired doctor coming at me I'll say "oh no, no, no", I don't want that human touching me.
I want the robot operating system that's done it a "million times perfectly.
" And that's going to be happening in every aspect of our lives.
BARATUNDE THURSTON: Oh, you work in finance? Remember there used to be a whole batch of dudes who screamed on the floor of the stock exchange with little pieces of paper trying to buy and sell stocks? Now all that is whisper quiet, happens at the speed of light, and it's literally machines versus machines.
I think people who have like, white collar jobs particularly, are naive.
Like, why don't machines write legislation? You know, you could argue that humans are really bad at writing laws.
Give it to the algorithm.
NARRATOR: And AI has already begun its march toward taking over the workforce, starting with one job at Microsoft.
ERIC HORVITZ: Hello.
MONICA: Hi Eric.
MONICA: Being one busy computer scientist's personal assistant.
ERIC HORVITZ: My assistant, who we call Monica, is leveraging over fifteen years of data about my comings and goings.
MONICA: I was expecting you.
Are you here for the 5:00 meeting with Eric? ERIC HORVITZ: It knows, for example, it's learned, which meetings on my calendar I probably won't go to, and so people show up, and I'm busy and my calendar is full, it'll sort of tip to them and say, "you know, I don't think he's going to go to that meeting" at 2:00, why don't you stop by then? "Can I pencil this in?" So this assistant has all this knowledge that even my human admin doesn't have yet.
MONICA: I guess I'll see you then.
Bye bye.
NARRATOR: But before you quit your job and sign up for that online coding course, let me lay this one on you: The big leap that needs to happen is that AI Needs to become self-aware.
Once that happens everything about our lives gets thrown for a pretty serious loop and we'll even question what it means to be human.
We already know that artificial intelligence is going to change the way we spend 40 hours a week.
It is, and will be, a tough time for the global workforce.
BARATUNDE THURSTON: Kids are gonna ask, "oh you guys had jobs?" Like, "you guys had jobs," you know, could be a thing, and versus, "back in my day we walked uphill" to school both ways," well, "back in my day we had jobs.
" NARRATOR: And that's not all: AI will eventually be become so sophisticated, so intelligent, that it could come pretty close to having all the qualities that we humans have.
ERIC HORVITZ: Someday, there'll be a system that we built with our own hands that will open its eyes and look out and kind of have this buzz of being here that we have.
And when that happens, that'll be a major landmark moment in human history.
NARRATOR: So what does that mean for you? For you parents out there who work so hard to inspire your children's young impressionable minds, what happens when that impressionable mind is a supercomputer far brighter than you'd ever be? And you are trying to help shape her conscience, how she sees the world.
OSCAR: I brought you something.
JESS: Oh, thank you.
What is it? OSCAR: It's a book, Jess.
JESS: I know, what's it about? OSCAR: Well, the author was writing about how the future might look in 100 years.
You know, how we might change the world.
JESS: 300 year old Sci-Fi.
OSCAR: Mm, I thought it might get you thinking about how you might wanna change the world.
JESS: Thank you.
NARRATOR: Your AI Kid seems like any other, but will AI ever be so lifelike that she'll need a swift kick in the pants to get off the couch and make something of herself, in other words, can scientists ever get artificial intelligence to consider complicated human ideas and thoughts? It looks like the simple answer is, yes.
HOD LIPSON: We look at self-awareness, consciousness, as the ability to imagine yourself.
Today, most robots can't do that.
They can't imagine themselves in future situations.
NARRATOR: This robot, this small piece of metal actually has a sense of self.
It may not look like much, but it is one of the most sophisticated robots on the planet.
HOD LIPSON: In the lab we can see a robot that uses all its sensations of the world.
So this robot is learning, sort of the equivalent to a newborn.
It has sensations, it can sense that it's moving and it's not just sensing the world, it's sensing itself.
And it's trying to form a self-image that matches reality.
It's a very simple self-image so we can actually plot it out and you can actually look at it and you can see there's very simple stick figure that this robot is imagining itself to be.
Or, we did something very cruel, we chopped off a leg of the robot and we watched what happened.
And we see that within about a day, the robot's self-image also loses a leg; it begins to limp.
AI is beginning to sort of have what we call QUALIA.
Basic notions of the world that we don't have words for.
If you think about, you know, the taste of chocolate, or the smell of the oceans.
It's very, very difficult to describe these accurately with words.
Computers also have them.
They can have sensor modalities that go beyond our five senses.
Therefore they will have notions of the world that we don't have words for.
NARRATOR: Okay, so let's swallow that tough pill: Computers may be capable of thinking about themselves, just like we do.
But there is one quality we have that's often thought of as the last talent we humans have an exclusive monopoly on, the way I earn my living: Creativity.
If computers can create, then what does that mean for humanity? DAVID BYRNE: The things that we think are the Providence of creative people, I think will start to be done by machines and we'll go, "oh jeez.
" BARATUNDE THURSTON: AI comedians.
I don't see that, I mean, I don't see that.
I think it's pretty obvious that only a human being can draw the sort of creative yet subtle connection which reveals the truth without shoving it down your throat.
Human comedians only! ROSE EVELETH: Would a machine ever make art, right? That's a big question.
When will machines start making paintings and music? NARRATOR: And here is the cousin of that self-aware ROBOT: The artistic robot.
This robot wants you to come back to its apartment to see its etchings.
HOD LIPSON: I always wanted to paint, but I really suck at painting.
But I do know how to program robots.
So the way that the original robot AI worked was to try to figure out how to arrange lots of possible paint strokes on canvas so that it matches what it's seeing.
The next phase is trying to have the computer on its own chose what it wants to paint.
A robot experiences the world in the Internet.
A robot can actually go places.
Use street view to tour places.
In a few seconds it can see oceans, it can see mountains.
Some of the paintings that is done recently are interpretations of photographs that it's seeing.
it might want to paint a new cat that it's never seen before.
But it's what it thinks a cat is based on its perception.
It's one thing to try to sort of make a machine paint, but it's another thing to actually make the machine be creative and do that in a creative way.
I think that eventually nothing makes us uniquely human.
So every capability that we have ultimately boils down to physics, to biology, to chemistry.
And if somebody really wants to recreate that capability in an artificial system, it would be possible.
MATT MIRA: What if they're more creative than us? Then give up.
We've lost.
The robots can have the planet.
NARRATOR: Whatever we do, we can only be more powerful than AI for so long.
They are gaining on us.
In fact, in the far future, AI Will blow past us like a bat out of hell.
Futurists call the moment that super-intelligent machines become the Alpha race on the planet, and begin to function beyond our control.
The singularity.
LOUIS ROSENBERG: Singularity is the point where a conscious artificial intelligence is created, and it's smarter than people.
PETER DIAMANDIS: It'll leave us in the dust, it'll be the difference between us humans and bacteria in the soil.
MICHAEL GRAZIANO: The singularity is an event horizon that you can't see beyond, and the idea is that technology is so transformative that we on one side of it can't imagine what the other side would be like.
PETER DIAMANTIS: AI is every successive generation has learned everything the generation before has learned and so there's this rapid exponential growth in capabilities and knowledge that is almost unfathomable.
NEGIN FARSAD: It's a total paradigm shift.
It's not the way that we understand it right now.
Like, people probably don't even wear sweatpants.
But it's more of like an idea of this future time that, you know, that is something, you know, otherworldly from what we understand right now.
NARRATOR: When the singularity arrives, AI Will not just grow our food and do our laundry.
They won't just do our taxes and legislate our laws.
They'll think and dream and create.
GEORGE DVORSKY: I mean, it's often been said that artificial intelligence will be the last invention we ever have to make.
Because, from there on in, it's going to take over.
NARRATOR: So once an AI Looks less like a tinkertoy, has smarts beyond belief, and a fully-developed sense of self with the social skills of George Clooney, what makes it different from you or me? That's where things could get a little hairy.
AI just might help you do things that you couldn't do without a sentient super-computer by your side.
But what if your AI Friend decides it wants to kill you? NARRATOR: Odds are that artificial intelligence will get pretty damned evolved, so much so, that they'll have superintelligence that far surpasses ours.
It could be the best thing that ever happened to humanity: They'll free us from menial labor, help us process information like never before, help us achieve levels of thought and scientific exploration that would never be possible without them.
On the other hand, they might challenge us for dominance at the top of the food chain.
Let's start with option number 2 first.
What happens if the machines we've created decide that they're better off without us? ERIC HORVITZ: Think about worst case scenarios.
A machine outwits humans in a way that can't be stopped.
Running amok.
Providing a threat to the survival of humanity.
Let's go there.
MICHIO KAKU: I think by the end of the century, we'll probably have robots that are self-aware, to a degree.
And at that point, we should put a chip in their brain to shut them off if they have murderous thoughts.
'Cause at that point, they could become independent of our wishes.
SYD MEAD: If we lose control of the robots and they start making more robots which are smarter than the robots that made the robots, we might be in trouble.
MATT MIRA: I think that's the moment where we should all head for the hills.
And not, and not, when we're heading for the hills, don't use WAZE or Google maps to get there, 'cause the machines will know.
NARRATOR: Sci-Fi has imagined the scenario of what happens when the machines believe they know what's best for us, and it's unsettling.
JON SPAITHS: In 2001, you get, perhaps the most famous example of a computer deciding it doesn't want everything that its makers wanted.
And it begins to have human will.
DAVE BOWMAN: Open the pod bay doors, hal.
HAL: I'm sorry Dave.
I'm afraid I can't do that.
CHARLES SOULE: The danger is you've got one of these robots, these AI'S these artificial intelligences, who just thinks, well, I, I bet I can do this a little bit better.
I bet I, I bet I could take care of this a little bit more smoothly than these guys.
Like these are just, you know, they're morons.
Look at, look at this.
MATT MIRA: I've seen too many movies where machines become self-aware, and what do they do? They rise up against their captors.
Who's their captors? Us.
We're in bad shape if the singularity happens, guys.
Everyone run.
NARRATOR: That's one possibility, that the robots want to destroy us.
It could also go in a less troubling direction, but one that brings up a lot of ethical questions.
AI might just want to exist alongside us.
Imagine a future when a constitutional amendment granting equal rights to all artificially intelligent beings is debated on the senate floor.
How would you react? OSCAR: Jess? JESS: Okay, you're gonna be fine, honey.
Okay, you're doing really good.
So brave! [Bone snapping.]
.
MOTHER: What happened? JESS: Oh, uh she dislocated her wrist.
OSCAR: It's okay, she fixed it.
MOTHER: Wow, thank you, are you a doctor? JESS: No.
PARENT: Get away from my daughter.
OSCAR: She was just trying to help.
MOTHER: You mean, "it!" OSCAR: Look just because she's an AI MOTHER: If you want to have one that's fine! Just keep it away from my child.
NARRATOR: It's only natural, this mother is protective.
She fears what she doesn't understand.
BRIAN GREENE: There's a whole pandora's box of difficult, ethical issues that arise.
Does that thinking machine have the same rights that we afford to human beings? Does that thinking machine deserve a certain kind of autonomy, even though we created that thinking machine? NARRATOR: Humans probably won't be too keen to give up their Alpha dog status on earth.
Thing is, we may not have a choice, when dealing with beings smarter and stronger than we are.
The question is, do human-like robots deserve equal rights? ROSE EVELETH: We get into this weird ethical territory because we know a machine's not a human.
Is it more like a dog? Dogs have certain rights, pets have certain rights.
So if I come into your house and I kill your dog, which I would never do, people would say, "well that's not right because my dog is a living being and it has consciousness of some kind.
" I think that machines will get some kinds of rights.
GEORGE DVORSKY: My own hope is that we do recognize these entities as being persons deserving of rights and that we will treat them accordingly, and you know what? And you know, we treat them well, then perhaps they're going to treat us well in return.
NARRATOR: It's possible that AI may just take the power it wants.
It'll certainly have the skills for it.
That means that the future could be a world where robots are no longer working for us, but will have the same status as humans.
ROBOT: Pardon me, but there's breaking news.
OSCAR: Okay, go ahead.
NEWS ANCHOR: Good evening everyone.
In the last hour legislation granting full personhood to members of the artificial intelligence community has swept through the united nations governing body with near total support.
While protests continue, this is a truly historic moment for individuals of biological and technological origin alike, and the beginning of a new chapter for mankind.
OSCAR: It's about damn time.
EVA: Jess you gotta hear this! JESS: I heard.
EVA: What? Jess.
JESS: I have rights, which means that I also have responsibilities.
People will finally let me help them.
NARRATOR: Having extraordinary powers makes this AI daughter feel obligated to help humans.
But is there really that wide a Gulf between her and us? MATT MIRA: I mean, what defines humanity? You know, in the Star Trek episode, measure of a man, where data was put on trial to see if he was property of STARFLEET or a realized individual, PICARD had to argue for him being an individual.
CAPTAIN PICARD: Now the decision you reach today will determine how we will regard this creation of our genius.
MATT MIRA: I think if you're going to have a lot of legislature that's going to have to come down for what to do with these artificially born beings and what to do with their rights and privileges and it's a can of worms I'm glad I don't have to open.
NARRATOR: A race of AI Robots could very well decide they want to seize power, overthrow humanity, and then they will inherit the earth and beyond.
But there is a slightly more optimistic scenario.
Will AI change what it means to be human? What if AI simply becomes a new, improved version of humanity, and they inspire us? NARRATOR: Artificial intelligence that is smarter than all of mensa combined is a very real possibility.
It's the singularity, and it's our future.
And it's possible that AI Will decide they want us gone.
Sure, it's terrifying and it sells movie tickets.
But there's another, more optimistic possibility.
BRIAN GREENE: It's mind-boggling and thrilling to imagine what artificial intelligence could do, right.
I mean, we struggle so hard to solve deep problems about the earth, the universe.
And imagine that we could transfer those enigmas to a more powerful thinking environment that could solve these problems, or reveal new opportunities that we've never thought of, and creative ideas that we then could pursue.
NARRATOR: In this vision of the future, AI Will not only live alongside us as partners, humans will also retain control over the machines.
CHUCK NICE: Maybe what you'll be able to do is remove the consciousness that is negative, remove the strings of consciousness that are what we would call evil, harmful.
NARRATOR: If AI is benign, then that artificially intelligent daughter may be a reality in our future.
And she may have new inspirations for her so-called life that we'll have to keep up with.
OSCAR: Would you pass me the, yeah, thanks.
JESS: You know, it's crazy how much of this he got right? Like, space travel, submarines.
OSCAR: He also got just as much stuff wrong though.
You know, when I first held you in my arms when you were born, I never could have imagined what kind of world we would live in.
JESS: Are you sorry? Sorry that Mom brought me back? I'm artificial? OSCAR: No, all any parent wants is for their child to explore their full potential.
And you have more exploring to do than any kid I know.
NARRATOR: Your AI Daughter might be just like your own flesh and blood, but she'll have abilities beyond anything biological.
What we're seeing here is the most optimistic future path of AI Maybe they'll be smart, good company and they'll like us.
We'll just have to be sure we program our AI carefully.
JON SPAITHS: There is one school of artificial intelligence that says, if you make the robot need to rest occasionally, if you make the robot lose its train of thought occasionally, it may identify with us more, it may think more like us, it may be more loyal.
NARRATOR: And even more than that, maybe AI Will work with us to enable humanity to do its best work.
GEORGE DVORSKY: Once AI Becomes more sophisticated then we can start to collaborate with it at a very interesting and a very profound level.
BARATUNDE THURSTON: When a new type of creativity come, and the collaboration that's possible.
So that's empowering.
I don't think that technology is always a threat, or, or a reduction.
I think it can add and bolster creativity as well.
So art will be affected sometimes in really unpredictably beautiful ways.
CHUCK NICE: It's going to free us up to do what we do best which is dream and create and come up with something even bigger and better than artificial intelligence.
NARRATOR: Then there's the possibility that we'll take a page from AI'S book: We'll become part computer ourselves.
JON SPAITHS: It is possible that we will begin to hybridize ourselves.
That is a fork in the road.
That's a departure, not just an extension of the human model.
The invention of a new definition of a human being.
PETER DIAMANDIS: We are about to go from humans alone, to humans enabled by computation.
Sort of hop on top of this exponential growth path.
BARATUNDE THURSTON: No problem at all with the merging with machine's technology scenario.
Nothing to see here folks, nothing to worry about.
I can't think of anything that could possibly go wrong by us merging with our technology.
EDWARD BOYDEN: Perhaps a third of a million patients have had some sort of neural implant put into their nervous system to help people with dementia or Alzheimer's, or other kinds of conditions.
Over time, their use will broaden, complex functions like intelligence, or memory could be augmented in real time.
MATT MIRA: If we are to merge with AI what form will that take? NARRATOR: A new model for humanity, no longer purely biological, we may be soon a species of cyborgs.
Now we won't have to rely on the luck of genetics: We could decide what kind of humans we want to be.
EVA: Nothing too drastic.
JESS: Of course.
EVA: Half my office now has nanobots in their brains.
And the other half are AI I gotta do it just to keep up.
Why shouldn't I analyze genomes in my sleep like everyone else? JESS: You ready? EVA: Uhh, yeah, mmm-hmm.
NARRATOR: Humankind's next step really might be this easy, just a simple injection of infinitesimal medical robots into our bloodstream.
HOD LIPSON: At that point, are we still human? Or are we pure AI? It could be a blurry sort of distinction and maybe it's best to leave it that way.
NARRATOR: What isn't blurry is that AI will fuel all future technologies, wrap your mind around this: Many scientists believe in the future we will work with AI To evolve out of our biological bodies.
We'll exist only as a digital signal in a structure called the Dyson's sphere, a giant super-computer powered by the sun where we can be everywhere and nowhere.
But before that scientists will have to figure out how to make humans live forever.
Coming your way next, what would you do if you never had to die?