Are We Living In A Simulation?

[Narrator] More than two decades after the original,

a new Matrix movie is here

and its vision of humanity trapped inside a simulation

is still relevant given the path for AI and VR

that futurists like Nick Bostrom predict.

You would have computers powerful enough

that you could generate millions and millions

of runs off all of human history.

Even if just a tiny fraction of these

technologically mature civilization's resources

were used for this purpose,

the vast majority of minds like ours

would be living in simulated worlds

rather than in the original histories.

[Narrator] 20 Years ago,

Professor Bostrom published the first draft

of his groundbreaking simulation argument, which asks,

are you living in a computer simulation?

[Neo] Right now we're inside of a computer program?

Is it really so hard to believe?

[Narrator] Not for some prominent scientists

and tech stars apparently.

So we rewatched The Matrix with Professor Bostrom,

as he explained his simulation hypothesis.

Bostrom's simulation argument doesn't resolve the issue

but it makes the case that as computing power

and progress in artificial intelligence grows,

one of the following statements must be true.

One, we will go extinct before we enter a post-human phase.

In other words, we die out before

becoming technically capable

of creating computer simulations with conscious minds in it.

Or two, we don't go extinct

but we'll be so evolved

that we won't be interested in running

these types of simulations, or three.

Is the simulation hypothesis,

that we are living in a computer simulation.

What is real?

How do you define real?

I thought especially the first Matrix movie

was pretty cool.

If the simulation argument is sound,

it would seem to be quite a revelation

about where we fit in,

into the larger structure of reality.

[Narrator] Throughout history,

from the Buddhist concept of Dharma to the Aztecs belief

that our world was a painting by the Gods,

to Plato's allegory of the cave,

we've always toyed with this idea of reality as an illusion,

as a shadow of what is truly real

I mean, people have been thinking since ancient times

about the question of how can we be sure

that we are not dreaming

or that there is not like a demon

deluding us into believing that there is an external world.

[Narrator] But it wasn't until the Wachowski's

Matrix film dropped in 1999,

that the possibility of reality run by computers

was burned into the popular imagination.

If the simulation hypothesis is true,

I don't think the best way to characterize things

would be by saying that the world that we perceive

is not real, I would rather put it it's real

but it's reality consists of being simulated in a computer.

[Narrator] In this scene from The Matrix,

viewers are shown a horrific basement or base reality.

The artificial intelligence we created

turned us into a power source to run their computers.

The matrix is a computer generated dreamworld,

built to keep us under control,

in order to change the human being into this.

[Narrator] Is this a likely outcome?

And when we consume more energy than we generate,

there are many more efficient engines already available

for converting energy,

you wouldn't want to keep a whole human organism

just to generate waste heat.

[Narrator] Okay so if the original Matrix premise

of humans as battery seems inefficient and unlikely,

then how would the simulation work?

So the way to picture that is that it's not

that there would be these kinds of organic brains

floating in tanks with some big fiber bundle plugged in

that feeds them with some kind of input

from some simulated virtual world,

but the brains themselves would be part of the simulation,

you would be conscious and in your brain

would in the simulation, would be simulated

at a sufficiently detailed level

that it would basically have

the same information processing structure.

As we previously thought

were implemented on biological neurons,

now all of those biological neurons are simulated, instead.

That information processing

could equally be implemented in silicon hardware.

This is kind of a premise of the simulation argument.

I call it substrate independence,

the idea that consciousness can be implemented,

not just on biological carbon-based substrate

but on other computational substrate as well.

[Narrator] In this scene,

viewers are introduced to a super intelligent AI program

that created the matrix.

Who are you?

I am the architect.

I created the matrix.

[Narrator] If we do live in a simulation,

who or what created our simulation?

If we are in a simulation,

then that simulation was created by

some form of super intelligence,

rather than by some human intelligence,

the technology to be able to create simulations

that are realistic and that have conscious creatures

like humans in them,

it's a very advanced kind of technology, right?

And the civilization that had developed

such advanced technology would I think

also have the technology to enhance their own intelligence

and to build very sophisticated forms

of artificial intelligence.

So by the time some civilization becomes capable

of creating these simulations,

they would have the ability to develop

greater forms of intelligence

and I think they would do so and in fact,

that's the likeliest way that they would gain the ability

to create these simulations.

So then the simulators would be super intelligent.

[Narrator] But why would a future civilization

create these worlds?

If we became capable of creating simulations,

we might create many different kinds of simulations.

We might create some simulations

that were simulations of our own past

that we tried to get as realistic as possible

based on our historical records, like historical tourism,

if you want to experience a bygone era

and you can't actually go back in time

because time machines are not possible.

And the second best might be to create

the kind of simulation of this historical epoch.

And you could learn, enter into that and experience it.

There could be many other reasons as well.

We might simulate potential alien civilizations

that we haven't encountered.

Maybe that would be very important for gaining information

about how alien civilizations

would behave in case we did run into them.

Think of all the reasons why humans have tried to create,

say fictional worlds, using whatever technologies we have,

even if it's just words on a page

that then leveraged the human imagination to conjure it up.

[Narrator] And like this scene from The Matrix

we have arrived at the moment of truth.

[Morpheus] You take the blue pill, the story ends.

You take the red pill, all I'm offering is the truth.

[Narrator] If we're all just software programs

running around in a simulated world,

why is it important to know the truth as Morpheus offers?

The world we perceive would still, in the relevant sense,

be real in at that it would, for example, matter a lot,

what happens in the simulation,

it would be a matter of great concern to us,

what experiences we will have in this simulated reality.

[Narrator] So if Morpheus offered the choice,

red pill or blue pill,

which one would Professor Bostrom choose?

Professor, red pill or blue pill?

As a philosopher, I guess you have already

chewed a little bit on the red pill, right?

Rather than the binary choice of maximal complete truth

versus staying in our current state of ignorance,

I am cautious by nature.

So if there were this third option,

I might be drawn towards that.

[Narrator] Okay, so a purple pill perhaps.

Even if we're not living in a simulation,

knowing what we know about the dangers of AI,

should we just give up on

developing machine super intelligence

before it turns against us,

like in The Matrix or Terminator

or 2001: A Space Odyssey or Westworld or Ex Machina?

I think it would be tragic

if we forever failed to do that.

I do think that transition however

will be associated with big risks,

including existential risks.

What we'll need to do is design AI systems

that are fundamentally on our side,

that share our values,

or that are aligned with human intentions.