Rizwan Virk: “The Simulation Hypothesis” | Talks at Google


[APPLAUSE] RIZWAN VIRK: Thanks, Piyush,
for that introduction. And it’s great to be
here in Mountain View. I actually live
just down the road, so it took me all about
three minutes to drive here. So today, we’re going to be
talking about these topics– video games, science
fiction, computer science, quantum physics, and mysticism. It’s quite a bit to
pack into an hour. And we need to leave
some time for Q&A. Now, you’ll notice that I included
two topics that sometimes aren’t discussed by a lot
of academics or scientists, which is science
fiction and mysticism. But it turns out
they’re actually quite relevant
and important when we talk about the
simulation hypothesis. First, most of the references we
have are from science fiction, like “The Matrix.” And that’s really the
question that we’re going to be trying
to answer today, is, do we live in the Matrix? But also, we start to get
into this question of, what is consciousness? And it’s something that science
hasn’t fully understood yet. But mystics have been studying
it for thousands of years. So although sometimes
scientists get upset at me– why are you talking about
mysticism and religion in the same breath
as quantum physics? I did that deliberately. So before we jump in, a little
bit of background on myself. As Piyush mentioned, I’m a
technical guy by background. Went to MIT undergrad many
years ago and did some– my thesis at the Media
Lab on using gameification for learning. And then I did a bunch
of enterprise software stuff back in the ’90s. And then I moved out
here about 10 years ago, ended up going to
business school. But more importantly, I got
involved in the video game industry full time. We created a game called
“Tap Fish” on mobile, which was the number one
grossing game in the iPhone app store when they first came
out with in-app purchases, and, of course, the number one
game in the Google Play Store, as well. And then made a bunch of
games based on TV shows. And really started to think
about how we build games. Then I became an investor. Invested in a bunch of
different companies. You may have even played
some of these video games, like Telltale Games
has a “Game of Thrones” game and a “Walking Dead” game. And if you haven’t used Discord,
probably your kids have. It’s a very popular
messaging app for gamers, but now being used in a
bunch of different worlds. And then a couple years
ago, I went back to MIT, to the MIT Game
Lab, which didn’t exist when I was a student. And this is the one
place that at MIT where they study video games
and the effect of video games. And I started Play Labs. But one of the important
things that I learned from some of the
faculty there was that playing games is one of
the oldest human endeavors. And if you look at how children
learn social interaction, it’s often through playing
games with each other. So playfulness is
an important topic. More than that, I submit that
most of what the technology we use today for a variety
of different applications, whether it’s AI, chat
bots, cryptocurrency– all came from video games and
entertainment in some way, whether it’s through GPUs
or through early multi-user dungeons, et cetera. So the history of video games
and the history of technology are pretty intertwined. And so that’s why I’m
going to talk a little bit about the history of video games
and how it ties to this idea that we may, in fact, be living
inside a giant video game. So I got my start with video
games a long time ago– very classic video games back
in the day of the Atari 2600– “Space Invaders” and “Pac-Man.” But what really made an
impression on my young mind was when the game started to
become a little more realistic. We’re still looking
at an 8-bit video game here, called “Pole Position.” I would always wonder what
was beyond the racetrack. You see the mountain that
looks like Mount Fuji there? And sometimes there would be
bleachers with fake people. And as I was
playing the Atari, I would wonder, what happens
to those characters when I’m not logged in? Is there a real world within
this video game where there’s actual conscious entities? And of course, with the
Atari, there weren’t. But this got more interesting
as science fiction developed. And as I got older,
I was a big fan of “Star Trek,” as many
of you may have been. And you may remember
the holodeck, which was a limited space,
but could project any kind of simulated
environment, and it would feel
real to the players of the game, who
were typically crew members of “The Enterprise.” But the episode that really
got me thinking about this took place in the holodeck. And there was a character
named Professor Moriarty. And for those of
you who have seen this will remember
Data, the android, played Sherlock Holmes
in the simulation. But this particular character
was quite intelligent. He actually figured
out that some of the people in the
holodeck were simulated, and some were real, and they
were from a world out there– “out there” being the
rest of “The Enterprise.” And so he wanted to
leave the holodeck, and he wanted to get
out of the simulation. And so this inspired me
to think about this topic for many years. And of course, the major
reference these days is “The Matrix,” where
Keanu Reeves plays a character named Neo who is
living a regular life as Mr. Anderson, working in an
office like many of us do. And he comes across
Morpheus, who’s played by Laurence
Fishburne, who is named after the
Greek god of dreams, who tells him that everything he
sees around him is an illusion. In fact, he says, you are
living in a dream world. But he also says that I
can’t tell you or show you– I can’t tell you
what the Matrix is. You have to see it for yourself. And even those who haven’t
seen the movie probably know the famous
scene– you can take the red pill or the blue pill. You take the red pill, you wake
up to see what the Matrix is. You take the blue pill, and you
just keep on living your life. Neo, of course,
takes the red pill, and he wakes up inside this pod. And he finds that he has a
connection into his neocortex– which is probably why they
called him Neo in the movie– and that everything
he thought was reality was being
beamed into his brain. And so we’ll talk a
little bit about this in the context of
current technology and where we’re going
and what this might mean. And as Piyush
mentioned, this topic has gotten popular
in the 20 years since the release
of “The Matrix.” Back then, it was just
consider science fiction. But there were a
couple of reasons. One of which is, Elon
Musk said a few years ago, at the Code Conference,
that 40 years ago, we had two squares and a dot. That was the stage
of video games. We had “Pong,” which was the
first widely available video game, made by Atari, right down
the road here in Sunnyvale. Today, we have virtual reality. We have augmented reality. And we have millions of online
players on these MMORPGs and 3D worlds. So what will happen
in a few decades? What will happen in
100 years from now? Will we get to the
point where the video is indistinguishable from reality? I would argue we’re actually
getting there very quickly. So I had my own
experience, speaking of “Pong,” a few years ago,
when I was playing a ping pong game using a virtual
reality headset. I think it was the
Vive, in this case. And I started to play this game. And you’ll notice that the
resolution isn’t that great. The opponent is there,
but it’s not even a photo realistic opponent. But what happened was that
the physics engine was so good and the reactions were so
dead on that I actually felt like I was really
playing table tennis. And I forgot that I was in
a virtual reality headset. In fact, at the end of the
game, I put the paddle down on the table and I
leaned against the table, just like I might do after
playing a game of table tennis. Of course, the controller
fell to the floor, there was no table,
and I almost fell over. And so then, I really
started to think, it’s not so much
about the resolution. It’s about the responsiveness
of the technology that makes us believe we’re in
something when we’re actually not. And so I started to do a
thought exercise that said, could we build a Matrix? How long would it
take us to get there? And what are the
stages to get to what I call the simulation point? And that’s the point
at which we would be capable of producing
a video game that is so indistinguishable
from reality that the players would not know
that they were inside a game. And if we could get
there, that raises a bigger question, which
is about the simulation hypothesis, which is, has
someone already gotten there? Now, what most
people don’t realize is that there’s two versions
of the simulation hypothesis out there. And one of them is what
I call the NPC version. And it was put forth by a
professor named Nick Bostrom at Oxford University. And we’ll talk a little bit
about his simulation argument, which has gotten a lot of press. But he put out a
paper in 2003, “Are We Living in a Simulation?” And that got
academics and others to start taking this idea
little more seriously. But then, in that version,
we’re all simulated AI. We’re like non-player
characters, or NPCs, as we call them in
the video world. In the other version of
the simulation hypothesis, which I call the RPG version, we
exist outside of the simulation and we are playing
roles, just as I might have an avatar in “World
of Warcraft” or “Second Life.” Consciousness exists
outside the simulation. And so when you look at some
physicists like Max Planck, who is one of the founders of many
aspects of modern physics, he says he views
consciousness as fundamental and matter as derivative. In the other version,
consciousness is seen merely as a
collection of neurons and an emergent property. So matter is primary, and
consciousness is derivative. And so that tug-of-war
between these versions of the simulation
hypothesis is, I think, one of the most
interesting aspects of this question. The RPG version, by the way,
is also the Matrix version. Neo existed outside
of the simulation. Now, there were AIs or NPCs,
like the famous Agent Smith, inside the Matrix. But most of the players
actually existed outside. So let’s talk about Bostrom’s
simulation argument. Many of you may have
heard of it before. But the basic idea is, he says
that if there are civilizations in our galaxy that
get to this point– and I call it the
simulation point, which is in the middle there– that get to the point where
they can create this technology, then they are likely to create
not just one simulation. They’re likely to
create many simulations. Actually, he makes
the point that there’s three possibilities. One, civilization never
gets to that point– which I doubt and many
technologists doubt, because we think
we can get there. The second point being,
civilization gets there, but they don’t want to
make any simulations. And he calls them ancestor
simulations, or simulations of more primitive societies. That’s a possibility. The third possibility,
though, is they get there, they create these
simulations, and they create lots and lots of simulations. And so his point
was that creating a new civil– a new simulation
or a simulated universe is just a matter of spinning up
another server on Google Cloud or wherever so we can have
millions of simulated universes with billions of
beings within them. Therefore, it’s simple
statistics, Bostrom says, that if you count up
all the simulated beings in the universe and you add
up all the biological beings in the base reality, then you–
if you are a simulated being, you are more– if
you are a being, you are more likely to
be a simulated being than a real being–
simple probability. There’s way more of
these than those. And that’s what led Elon Musk to
say that the chances that we’re in base reality is
one in billions, because there are so many
more of these beings. So that’s one version
of the argument. But I wanted to delve a little
bit deeper and say, how do we get there as a civilization? And it all starts
with video games. So about a third of
the book is dedicated to how we build the technology. And if we go back
to the 1970s, text adventures where the
first place where there was a virtual world
inside the computer that could be explored. There were no graphics, but
you could give little commands like go north, go left. And you could speak with NPCs. Then we got graphical
arcade games like the ones I grew up with. And today, we’re at 3D,
massively mulitplayer online role playing games, where
there are more realistic NPCs and there are millions of people
interacting with each other– which brings up an interesting
philosophical question. If you and I are both
playing “World of Warcraft,” are we in the same space? Are we even seeing
the same world? It’s what I call
the rendered world. Of course, that’s the beginning. Stage 4 and 5 are
about virtual reality and photorealistic
augmented reality. Stage 6, though, is where it
starts to get interesting, because now we
say, at some point, we’ve got to get
rid of the glasses. I mean, even though
I forgot for a moment that I was playing ping
pong, I did in fact have the glasses on my head. And I would have
realized it eventually. And so as we start to be
able to render objects in augmented reality using light
field displays– and the idea is that you can figure out
how the light bounces off of this cup, for example. Then what is the
difference between the cup and a physical object
where light bounces off the physical object? The physical object
becomes information. And today’s 3D printers are
already showing us this. We can take any physical
object and we can encode it into a set of pixels– 3D pixels. Well, if that’s true,
then the distinction between information and
material reality starts to blur. And that’s actually one of
the key points in my book, is that we may be living
in a world of information and not a world of
physical, material objects. The more physicists open up– they look down at the molecules. They look at the atoms. They look inside the atoms. And they can’t find
physical matter. There’s nothing there
when you keep opening up the Russian nested dolls. What is there is information. But getting back to my
point about ping pong, the really critical stage is
the brain-computer interface. It’s this ability to be able
to beam into our minds a scene and then to read back
the results of how we want to react to that scene. And we’ll talk
about some analogies for how that might work. And then stage 8
is false memories. Stage 9 would be
AI that’s much more able to come across as human. And then stage 10 is
another controversial topic, but one that’s quite popular
here in Silicon Valley, which is downloadable
consciousness. Can I download my consciousness
into a silicon device, thereby living forever? And we’ll talk a
little bit about this in the context of
consciousness and mysticism later on, because
there may already be people who know how to do
this without the silicon device to upload and download
consciousness. And then we reach stage
11, the simulation point, which is that point
that I talked about. So you’ll notice that– where are we in this? We’re at about five, six stages. But we’re not that far
off from the other stages. So let’s talk a little bit
about the things we still have to develop. So in brain-computer
interfaces, the first step is being able to project
a picture into the mind. And there have been
experiments since the 1950s where they used
electrical currents, and they were able to bring up
particular scenes in the mind. The harder part of it
may be to actually, as in the case of the Matrix,
to get your responses. Well, it turns out,
there’s actually a biological technology
already that does this. It happens to each
of us every night. It’s when we dream. Our dreams are like
mini simulations. We have scenes projected
in front of our minds. Some of the characters are
real, some of them are not. We react to those scenes. And we see the results. It’s the same
basic gameplay loop that we use when we
build video games. We present something. We wait for your response. I mean, today’s
video games, you’re still doing dialogue trees
and things like that. But being able to detect
people’s responses. Now, just last week, Elon
Musk unveiled neural link, which is a chip that can
go in the brain that will be able to enhance using AI. There’s a startup in Boston
called Neurable that is already trying to be able to predict
your responses so that you can actually control
video game characters. Now, this may or may not
involve an actual physical wire or a physical chip. Some of us may not want to
put a chip in our minds. There was a group at Princeton,
at the Advanced Engineering Anomalies Research Lab, where
they looked at random number generators. And they found that quantum
random number generators could be influenced by your thoughts. And so there’s a
few companies now that are trying to
commercialize this technology without necessarily requiring
it to plug into your brain– that if you change
your thoughts, it changes the color
of a lamp, for example. So we’re still in the
early stages here, but there’s a research
group at Berkeley that found that they could
get rough ideas of what you were dreaming. So these technologies
are coming. They’re still in their
infancy, but they’re not as far off as you might think. We then moved to stage 8, which
is the area of false memories. Some of you may
have seen the movie “Blade Runner,” which was based
on a novel by Philip K. Dick. So what many people don’t know
is that in 1977, Philip K. Dick was in Metz, France. And he gave a talk. And he said, we live in a
computer-generated reality, and the only clue we
have is when something in our environment changes. In fact, he really
believed this. I interviewed his wife, Tessa
Dick, as part of my research for the book. Some of you may
have seen “The Man in High Castle,” which is a
pretty popular science fiction story. He actually believed that that
was an alternate timeline that actually happened and that
the creators of the simulation decided to unwind that timeline
and re– and go forward again. So now, we’re at
the intersection of science fiction, mysticism,
and the simulation hypothesis. So let me jump forward. AI and NPCs– the
first practical AI– not many people know this–
was Claude Shannon, who was a professor at
MIT and Bell Labs, where he created a
chess-playing computer. And in that same year, in 1950,
Alan Turing came up with his idea of the imitation
game, which is now known as the Turing test–
here at Google, I’m sure you guys have
all heard of this– where the idea is that
behind one curtain, there’s a computer. And the other curtain,
there’s a person. And if you can’t
tell the difference, then it has passed the imitation
game or passed the Turing test. We seem to be
getting closer to it. We’re not quite there
yet with our NPCs. We still use dialog trees, if
you’ve played a Telltale game. It’ll say, what do you want
to say back to this character? But we’re getting there
on the video game side. So it’s just a matter
of time, I think, before we get
realistic-looking NPCs. And then we get to stage 10,
downloadable consciousness. And so the question
here is– gets back to that fundamental question. Is consciousness just
a set of information that can be downloaded and
put into a silicon device? Or is it something
else that exists somewhere else, in which case
simply copying the neurons isn’t enough? So far, we’ve had
computers which have simulated all the neurons
and the neural connections within a rat’s brain, which
is a much smaller number than the human brain. So it’s not that– it won’t
be that long before we’re able to simulate
all of the neurons and all of the neural
connections of a human brain. I think the number is
something like 10 to the 39. Once we can do that,
do we have the ability to replicate consciousness? But this brings us now to
metaphysical questions. Many of you may have heard of
the quantum teleportation done in China of a particular
particle to a satellite. But really, what
they’re teleporting is the information contained in
that particle, not the particle itself. So the question is,
if you replicate somebody’s neural
structure, are you creating a copy of that person,
or is it actually that person? So these are deeper
philosophical issues than the technology itself. But again, we’re
not that far off. So it’s very possible that we’ll
be able to get to stage 10. And once we do, we’re
at the simulation point. So if we can get to
the simulation point, who’s to say that someone
hasn’t already gotten there? And perhaps they’ve left clues
for us in our physical world, either intentionally
or unintentionally, that we’re living
inside a video game. So let’s shift and talk a
little bit about the physics. Now, I’m not going to spend
too much time on this. I’m not a physicist. But there’s some really
interesting aspects of quantum physics. And Niels Bohr one said, “those
who are not shocked when they first come across quantum theory
cannot possibly have understood it.” Physicists have gotten
to the point where they no longer want to ask the
big questions about quantum theory. They say, let’s just get
down to the calculations and figure out some
of the details. But the big question
and the big mysteries are, why do things
happen in this way? And a couple of
the big mysteries are quantum indeterminacy
and quantum entanglement. Why would we live in a universe
where these things exist? That’s the question that I
believe physicists don’t have the answer for but which the
simulation hypothesis provides a pretty good answer. So most of you may have heard
of quantum indeterminacy. This is the famous
double-slit experiment where if a particle is going
through these two slits, it starts off as a
probability wave. And then the probability
wave collapses to a single possibility
when it’s observed. So here’s an example of a wave. And probably, the better
way to understand that is Schrodinger’s infamous cat. And so common sense tells
us that Schrodinger’s cat is either alive or
dead inside this box. And Schrodinger gave
a scenario where they put some radioactive
material in the box, within the cat– with the cat, so
that within an hour, it would have a 50%
chance of being alive and 50% chance of being dead. So common sense says it’s
either alive or dead, we just don’t know, because
we haven’t opened up the box. Quantum physics tells us, no,
both of those possibilities actually exist. And it’s not until
someone actually observes it that that
possibility collapses the probability wave. And there’s a few
different theories of what causes the collapse
of the probability wave. But it seems to be
that the rule of thumb is that in quantum physics
in our physical universe, only that which gets observed
actually gets rendered– because you need an observer. Now, there’s some debate over
whether that observer can be a machine, an AI, or it has
to be a conscious observer. And maybe if we have time, I’ll
talk about some experiments that are being done
in Southern California to try to nail that down. But it needs to be observed. Only then does it get rendered. Well, this reminded me
a lot of video games. In early video games, there
was a shared rendered world. I don’t know if anyone knows
which video game this is, but this is a game called
“King’s Quest” from the 1980s. And you’ll notice that all of
the pixels exist on memory. And as your character
moves, all you have to do is get the pixels from the
next scene and move them. So that is a pre-rendered world. And the materialistic
point of view says that we live in
a pre-rendered world. The physical world exists. It’s just a matter of us
getting– going around, moving around in it. But if you had asked
somebody in the 1980s, can you render
something like “World of Warcraft” or “Fortnite,”
a fully 3D world, they would say, no. We just don’t have
enough computing power for all of those pixels. So how did we get
from there to here? We came up with
optimization techniques. And that’s really the key. We came up with 3D
modeling and we came up with rendering engines and we
came up with physics engines. And what does a
rendering engine do? It takes the point of
view of your character and renders only that
which can be observed from your point of view. Everything else is hidden
from the point of view of the rendering engine. Now, this is the game “Doom,”
which came out in the 1990s and was one of the first
3D perspective games. But of course, the techniques
have gotten much more sophisticated since then. So the rule– golden
rule in video games is, render only that
which is observed. Which brings up the
question, is there a shared rendered world or not? Now, if you and I are
playing “Fortnite” or “World of Warcraft” or “Second
Life” or any of these games, we think we’re in
the same scene. But the reality is, there
is no shared rendered world. I’m rendering it on my laptop. You’re rendering it on
your laptop or your phone. And so physicists
have been asking the question for a while, is
there no shared rendered world? There was a Nobel Prize winning
physicist named Wigner– Eugene Wigner, who asked
this question, saying if I look at a particle and see
the collapse of the probability wave, and then Wigner’s friend
looks at it before I tell him what I observed, will
he see the same collapse of the probability
wave that I saw? And he wasn’t able to do the
experiment back in the ’60s. Well, they just did it
in England, recently, using a series of, I think,
five or six entangled particles. And they found that
Wigner and Wigner’s friend could observe different
collapses of the probability wave. So this is pretty profound. Two different people might
observe a very different rendering of the world. It’s not unlike what
we do in video games. If you’re a level
10 character and I’m a level five, on the server, we
can send you a different thing. I can put a UFO in the
scene if I want for you, and the other guy won’t see it. So we’re sending information
to a rendering device which is rendering in our
consciousness– which brings up the point of, what is
that rendering device? Is it our brains? Is it consciousness? Is it something else? John Wheeler, who we’ll
talk about in a minute, said that everything is
information theoretic in origin. He’s a pretty
well-known physicist. So it seems like quantum
indeterminacy could be an optimization technique. Same kind of techniques we
use when we build video games. What about quantum entanglement? Einstein called it spooky
action at a distance. It’s been confirmed, now. Quantum entangled
particles have been able to get pretty far
apart from each other. Theoretically, they could
get millions of miles away. But why would nature
have an instance where these particles are entangled? Why would it need to? I would posit– again, coming
at it from a computer science and video game point of view– that it’s all about
optimization and compression. Now, what does this have to
do with “Game of Thrones”? Well, the reason I can stream
“Game of Thrones” to my phone is because it doesn’t
send all the pixels. I mean, at an HD resolution,
think of how many pixels would have to be sent to my
phone and how many frames there are at 30
frames per second. So what we do is, we
take all the particles, the pixels that have the same
values, and we collapse them. We’re going back to Claude
Shannon and information theory and compression. And whenever you
compress things, you reduce the
amount of information that has to be transmitted. But you also need to decompress
that information, which means you need error
correction codes. And people– some
certain scientists are starting to find
error correction codes in the physical universe. It’s all theoretical
at this point, but there are string theorists,
for example, who find– who found error correction
codes in the equations. And people who are
simulating physical universes using a smaller number
of entangled particles are finding that error
correction codes are present, as well, there. So it’s very possible that
quantum entanglement really is yet another compression
technique or optimization technique. So these are clues that we
find in the physical world. Now, another
interpretation of this is that the probability wave
never actually collapses. What happens is that every
time there’s a decision made, we branch out into multiple
different universes. So we probably have a bunch
of computer scientists here and engineers. If you were to branch out and
create a copy of something, how would you do that? If you look at
the physical world and you look at
physical processes– creating a clone of
the sheep Dolly– you still have to grow the cell. Creating a clone of a tree– it’s a slow process
following some algorithms that we don’t fully understand. But they’re probably
fractal in nature. However, in computer science,
we just copy the information from a server. And it says, boom, we
now have another branch. If you’re branching out your
cryptocurrency blockchain like Bitcoin, for
example, you just take a copy of the
blockchain at that point. If the multi– many
worlds hypothesis is true, there must be a mechanism
that can instantly create copies of the physical
universe, which means that it’s more likely to be an
information-based reality than a physical, material
universe, in my opinion. This also begs the
question, if there are multiple parallel
universes, are they really just probability? Or do they actually exist? So when I was making
my first video games, we used to use something
called the minimax algorithm. We would look at the
current state of the board, and we would project forward. And then we would look
at all the probabilities and we would find the biggest– the best, most optimal function. And that’s the path
that we would choose. There’s a physicist named
Tom Campbell who wrote a book called “My Big TOE”– my big Theory of Everything. He believes that that’s
what nature does, is it spins off these
virtual realities as parallel universes, but
they don’t actually exist. They’re just probabilities. And then whatever
is most optimal– that’s the path that we follow. That’s how rendering happens
within the physical world. So these are all interesting. And there’s a lot more aspects
of physics which I won’t get into a ton of detail now– the delayed choice
experiment, which shows that a decision in the
future can affect the past. Do we have pixelated space–
going back to Max Planck. The smallest physical
distance we can measure is called the Planck length. Below that, we can’t measure it. Why would there need to
be a smallest distance? If nature was
continuous in material, that may not be needed. Pixelated time,
which is something we do in simulations
and computer games all the time– you
have a clock speed and you have a
certain generation. For example, if you’re
simulating fruit flies, every year, you look
at the population. But whatever speed of the
simulation, the [? set– ?] it has to be a multiple
of the clock speed. So if there is a pixelated
time, a minimum unit of time– and many physicists are
starting to think there is– then it’s very
possible that we’re in some kind of a compute
engine that’s going forward. And there’s more about
speed of light and wormholes and good stuff like that. And of course, quantum
superposition and qubits, which I talk a little bit more
about in the book, as well. So I’m going to end the
physics section here, because we’re
running out of time. But John Wheeler, who
was one of the last physicists to work
with Einstein and many of the greats of
quantum physics– he came to the conclusion
that the physical universe was really based on information. And he had a famous phrase. He called it It from
Bit, which is the idea that anything you
see that’s physical is based on information. And he says in his lifetime,
we went through three stages of physics. In the first stage,
everything was a particle. It was a material universe. In the second phase,
everything was a field– a probability field,
a quantum field. And by the third
phase, everything was a series of bits. And now that we’re looking at
quantum computers and qubits, we’re starting to
see, everything could be an on and off
structure and be built of bits– which leads us back
to this idea that we could be living in an
information-centric, simulated world around us, and those
are just some of the clues. The third big aspect
that I cover in the book is this idea of mystical
traditions and the world’s religions. And it turns out, not
just any one religion. Sometimes I get
pushback from people of a particular religion. Like, I was on a show in South
Carolina and this guy said, that’s not what
Jesus Christ said– that we’re living in a
simulated computer game, we’re sitting there playing
“Mario Brothers” or something. I said, well, let’s look
a little more deeply at what he said. But particularly,
Eastern traditions seem to– you don’t even
have to do an analogy. Buddha said, know that all
phenomena are like reflections appearing in a
very clear mirror, devoid of inherent existence. Sounds a lot like pixels. In the Hindu traditions, that’s
a picture of the god Vishnu, who’s having a dream. And we’re all part of the dream. In the Vedas, they
talk about the lila, or the grand play of life– like a stage play, where
we’re all characters, kind of like Shakespeare said–
that all the world’s a stage and we’re all characters. But what kind of a
play would it be? It’s an interactive
play where each of us can modify what’s happening
with some sense of free will. Well, that sounds a
lot like a video game. They didn’t have the
terminology video game back when they wrote the Vedas,
but if they did, that might have been a more
appropriate thing for them to say. In Buddhism and in Hinduism,
in the Eastern traditions, there’s this idea of
the Wheel of Samsara, where we go through
a life, and then we go through another life. So we play different
characters where we’re basically– the world is
basically a role playing game. And what makes this wheel turn? In those traditions
it’s the fact that we still have
unaccomplished tasks ahead of us. I’ll talk about
karma in a minute, but karma is what makes the
wheel turn in these traditions. In the Tibetan traditions, they
have this idea of dream yoga. It’s one of the six
yogas of Naropa, which is a set of
mystical practices that have been preserved
for about 1,000 years. And in dream yoga,
you learn to wake up within your
dreams– what we would call lucid dreaming today. And the idea is, if
you, inside a dream, can realize that what
you’re seeing around you is an illusion,
it’s a dream, then you can also wake
up in physical reality and realize that we are,
in fact, in a dream world and in an illusion. So you bring that consciousness. So this is one of the
yogas that they’ve taught for a period of time. And now we’ve
started to study it. And lucid dreaming has been
studied in the laboratories, as well. So we know it exists,
so– which brings us back to the age old
question that Descartes raised in, I think, the
1500s when he said, what if there’s an evil
demon that’s tricking me and everything I see around me
is just signals into my brain? I can’t be sure. The only thing I can be sure
of is that I am thinking, therefore I am. So this idea that we may
actually be in a dream world with signals being sent into
our brain has a long history. Particularly within
the Eastern traditions, we basically download
consciousness at birth from somewhere
into a physical body. And at death, the information
is uploaded or beamed back up, like in a cloud server. Now, what is it that gets
downloaded and uploaded? Now we’re back to the
transhumanist movement and folks like Ray
Kurzweil here at Google who want to be able to
download to a silicon device. Well, it turns out, in
the Eastern traditions, they’ve been talking about
uploading and downloading of consciousness for a while. One of the differences
between the Hindu and the Buddhist
traditions is in Hinduism, there’s an indestructible soul. But in the Buddhist
traditions, what gets reincarnated is
really a bag of karma. Well, what is that? A bag of information. It’s a set of things that
you have done in this life and that you have
to do in the future. So it’s really just information
being sent back and forth within– in the East traditions. Well, I started to think about
this from the point of view of a video game designer. And it turns out, you can
define a very nice questing system that basically
defines how karma is supposed to work within the Eastern
traditions– which brings up the idea, is Buddha’s endless
wheel an algorithm, actually? And is there actually a
scientific or technological basis for what these
religions have been telling us all along, where we
have a series of quests that we have to achieve? And you can read more
about that in the book. Even in the Western
religious traditions, there’s this idea of
the recording angels. And in the Islamic
traditions, in the Quran, they give them names. And they say one angel write
down every good deed you did, and one angel writes down
every bad deed you did. Now, we’re a bunch
of engineers here. If you had to implement
that, would you really have 14 billion conscious
entities sitting there writing down things in a book? No. You would probably
have AI or algorithms that are recording things. But in the Quran and in
the Western traditions, they say you have to
review what you did. So they didn’t mean
it’s literally a book that somebody is
writing this stuff down. But they say that after you die,
you have to examine your deeds and look at the
impact of those deeds. Well, this is
interesting– if you’re going to want to
talk about China, you want to talk to somebody
who’s been to China. So if we’re going to talk
about what happens after death, you want to talk to people
that have actually been dead. So I have a good friend
named Dannion Brinkley who wrote a book called
“Saved by the Light.” He was struck by
lightning and had a 20-minute
near-death experience. Turns out, there’s thousands
of these cases out there. And almost all of them
bring up this idea of a panoramic life review– he said, where there’s a
recording of everything you did in your life. And you have to
relive that recording. But you have to live it from the
other person’s point of view. And Dannion was in
the military, and he used to literally kill
people as a sniper. He had to experience what it
was like to get the bullet. So recently, before
I wrote the book, I was involved in a startup
that did recordings of games like “League of Legends.” Some of you may have
played these games. Now, this is a 3D game
on a 2D landscape. But we would record it in 3D. And then you would put on
a virtual reality headset. And we could put the virtual
cameras anywhere we wanted to. So literally, you could play
back the character you killed or that you shot in CS: GO. And you could see it
from their point of view. So I’m not saying that what the
Western religious traditions told us is necessarily true. I’m saying if they’re true,
how would it be implemented? It would be implemented by doing
a 3D recording of everything that has happened
to you in your life. So that reminds me a lot of
the simulation hypothesis. Kids today are
recording their games and sending them on
Twitch all over the place. So that may be what’s going
on at a more cosmic scale. Now, I won’t get into
some of these, since our– we’ve talked about NDEs. But here at the Stanford
Research Institute in the ’70s, when they
were inventing the ARPANET, they were also doing
research on remote viewing where they had people
trying to figure out what was in a remote destination. And they found statistically
significant results. Not many people know
that these were going on in rooms next to each other. They’re working on the
internet and they’re investigating remote viewing. If it exists, how could it work? It’s the same way we put
a virtual camera anywhere inside the 3D world and we
can see what’s happening. It makes sense, if
it’s a simulated world. It doesn’t make any
sense if we live in a material world, at all. And so there’s a
lot of these topics. I’ll actually be giving
a talk at East West here in Mountain View,
more about the unexplained phenomenon side of it. But we want to transition
to questions here soon. But one of the big
questions that I get is, can I hack the simulation– if we’re living in a simulation. Now, you may remember the
famous scene in “The Matrix” where this young guy
was bending the spoon. And his advice to Neo was,
the thing you have to remember is, there is no spoon. Within a simulation,
it’s all information that’s being rendered. So to hack it, you
may have to figure out how to tie into that. Now, we’re very left-brained. There was also research done on
this at the Stanford Research Institute, on spoon bending,
which some people think doesn’t exist. Well, this is a picture
from the Marriott here in Santa Clara
on Friday night, where a group of 20 people
got together, took spoons from the Marriott,
and bent them all, if you zoom into that picture. Now, one of the guys
who did research on this, Jack [? Howard ?] from
Stanford Research Institute, said he went to Los
Alamos and tried this, and none of the physicists could
do it, but all of their wives could. [LAUGHTER] Isn’t that interesting? So maybe hacking the simulation
is not about a left brain quantitative approach,
but it’s about letting go and realizing something
about the underlying nature of reality. So anyhow, I want to
break for questions soon. But we get the big
questions often. Why would somebody want
to do a simulation? The answer is, why do
you play video games? For entertainment,
to have experiences that you might not be able to
have outside the simulation. I can’t ride a dragon and
kill orcs in real life. But I could do it
in a video game. So there may be something
about the experiences in the video game
that we’re having. Another big question is,
what’s outside the simulation? Elon Musk asked this
recently, as well. He said that– I think Lex Fridman,
researcher at MIT, asked him, what would you ask an AGI
if we had Artificial General Intelligence? He said, I would ask them
what’s outside the simulation. Is it aliens? Future humans? So Bostrom’s theory was that
we live inside an ancestor simulation, so they’re
future versions of us. And this is more along
the lines of what Philip K. Dick believed. Could they be time travelers? If you’re more of
a religious bent, you might think
it’s God and angels. Some people think it’s
pure consciousness, going back to Max Planck
and some of the ideas there. We don’t know the answer. But what does this mean
for me as a person? You might say, it doesn’t
mean anything to me. Well, if you’re inside
“Grand Theft Auto,” maybe your goal is
to steal automobiles. So my question
for you is, if you were a character in a video
game, what attributes– strength, intelligence,
dexterity, all those old
Dungeons and Dragons attributes– have you
brought into the game? What type of character are you? And what is your quest? What are the
challenges and things that you want to
achieve in this life? So adopting the
idea of a video game is a pretty powerful metaphor. And to take the advice
of another MIT physicist, Max Tegmark– he said, make it an
interesting simulation, because you don’t want the
simulators to shut you down. [LAUGHTER] I like to use the analogy– I like to use the analogy of
Indiana Jones films, where if at the beginning he was
just given the treasure map and it just said,
here’s the treasure. Go get it– that wouldn’t make
for a very interesting movie. It wouldn’t make for a very
interesting video game, either. You have to have a clue. And you follow the
clue to the next clue. And you have to have challenges. In fact, in the sequels to
“The Matrix,” if you remember, they said that the first
version of the Matrix was an ideal world. But humans wouldn’t accept
that as a real world because there was no strife. There was no boredom. There weren’t any challenges. And people always
ask me, well, if I was going to make
a video game, I would make myself a trillionaire
and I would do this and that. Well, that may not be
the nature of the game and that may not be the types
of quests or achievements that you’re here to achieve. So that’s the answer
to that big question. So I think we have a few
minutes for questions. In terms of where
to get the book, I like to say, support
local bookstores. So Books Inc here in
Mountain View, and East West. In New York, Strand
Bookstore, Barnes and Noble. In Boston, the MIT
coop, Harvard coop. So go to your local bookstore. And if you can’t find
it, then go to Amazon. So with that, I’d like
to open– and my website is zenentrepreneur.com. And my email is
there if you want to ask me questions afterwards. [APPLAUSE] AUDIENCE: So as a question– first of all, a
wonderful and fascinating philosophical conversation. Happy for, basically,
being able to ask you this. There’s a certain field of
machine learning research called model based reinforcement
learning where you’ll build up a simulation of your environment
and then train inside of that simulation. There is a value to having
this simulation, which is that you can generate
arbitrary amounts of data and so end up with an algorithm
that’s much more finely tuned and capable
of decision making than without that simulation. There’s also this want to
do counterfactual inference and ask, oh, well, if I hadn’t
experienced this reality, what would have happened? And in asking what
would have happened, you can do a kind
of credit assignment which says, oh, well, if I
had taken some other path, this is what the
outcome would have been. So that might be a motivation
for creating simulations. So I don’t know if you engage
with that at all in the book or what your thoughts are on
this sort of access of machine learning research. RIZWAN VIRK: Not so much the
machine learning research, but this idea that
you may want to look at alternate versions of
what might have happened and find a way to score those. I think that may be
a fundamental part of this view of
parallel realities and what they actually are. It gets back to
this question of, why are we doing a simulation? If I’m simulating
fruit flies, it’s to see how the population
might evolve over time. If I’m simulating “Sim City”
or “Sid Meier’s Civilization,” what is it that I’m
trying to figure out? Some people think that
we’re trying to figure out if we’re going to
destroy the planet or if we’re going to destroy
ourselves in this simulation. I don’t think there’s
necessarily just one purpose. And there’s not just
one way of scoring what the different alternatives
might be, but really, that each of us has
our own versions of what we’re here to do. And each of us is
doing something similar to what you talked
about there with the machine learning, which
is that each of us is projecting into the future,
figuring out what might happen, and then coming back to the
present– which gets back to this idea in quantum
physics of the delayed choice experiment, where some
observation of the future is actually affecting the past. So it’s a very high level answer
to what you’re asking there, but I think that general
approach may be what’s at work in the simulation. AUDIENCE: Before we jump
to another live question, there’s a question from
the live audiences. This person asks if
there’s any possibility to wake up from the
simulation like Neo did, and how can he achieve that? [LAUGHTER] RIZWAN VIRK: That’s
a great question. And I would submit that
most religions today happened because somebody
woke up from the simulation. They peeked outside
the simulation. They saw what was out there. And they came back and
tried to describe it. And then somebody built a
whole religion around it. But it’s kind of like the three
blind men with the elephant. They’re all feeling
the elephant, but they’re feeling different
parts of the elephant. One thinks it’s like a snake
because they have the trunk. One thinks it’s like a tree
because they’re feeling the legs of the elephant. One thinks it’s like a house. So I think there are ways to
peek outside the simulation, but I think they may have to
do more with our consciousness and realizing that there may
be more going on than what we’re seeing around us. And so that’s what
yogis and these guys have dedicated thousands
of years of techniques to– basically, peeking
outside the simulation. Now, the other
question might be, can we do it more from a
scientific point of view? And I think a
related question is, if the simulation hypothesis
isn’t falsifiable, then why should we
take it seriously? But I would say that just
because something can’t be proven wrong doesn’t mean
that you can’t find evidence for it. 200 years ago, the idea of
meteorites was unfalsifiable. Scientists said,
that’s ridiculous. There’s no way that
that’s happening. They just didn’t believe it. There was a belief
system that said, these things cannot be here,
therefore, they don’t exist– contrary to some evidence
that people had seen things falling from the sky. They said, ah, you
imagined the whole thing. But eventually, they did find
some evidence for meteorites. And today, we actually
accept that meteorites do exist and things
do fall from the sky. But we needed a
different model of how the universe worked before we
were able to get to that point. And so I think you
can find evidence. And in the book, I go into
some of the experiments that people have done to
try to figure out if we have a pixelated reality, et cetera. And so there may be a more
scientific way of doing it. But I would say, explore
consciousness is the way to wake up from the game. AUDIENCE: So I really
enjoyed the parallels you drew between aspects
of a reality and video game software, like the pixelization
of the world with Planck’s constant and how these
quantum phenomena might be forms of compression. So what I’m wondering is, we
all know that software has bugs and that they’re unavoidable. So what is the analogy– what
are bugs in the simulation? What do they look like? Can we detect them? RIZWAN VIRK: Great question. So if you remember, in
the movie “The Matrix,” they coined the phrase,
which is used pretty often now, “a glitch in the Matrix.” And if you remember, it was the
cat that was going by the door. And then he looked again, and
it was the same scene replayed. It was a glitch in the software. So the question is, are there
glitches in the physical Matrix around us? Now, I mentioned Philip K. Dick. He wrote a novel called
“The Adjustment Team”– a short story which became
“The Adjustment Bureau,” which is a big Blockbuster
movie with Matt Damon and Emily Blunt. But he believed that it
was experiences of deja vu just like the glitches in
the Matrix in the movie. And he– at one
point, he said, I used to have a light switch
here, in my bathroom. Now I have this chain. What the heck changed? He thought that somebody had
played with the simulation and had actually changed little
things in the physical realm. So he believed in big
changes and little things. And so a lot of the unexplained
phenomenon that I talked about– but I didn’t really
get into detail today– that a lot of modern
science dismisses could, in fact, be
glitches in the Matrix, particularly around things
like feelings of deja vu and precognitive dreaming. There’s a lot of
literature on this stuff that, because our belief
system doesn’t allow it, most scientists won’t
even look at it. If you’ve ever heard
of a scientist who won’t look at data, that
happens in this realm, but not in other realms. And so I think that–
that’s why I included the whole area of mysticism
and religions in the book, because I believe there’s a
lot of glitches in the Matrix, but we have been
conditioned for a long time to not look at those
as anything real. AUDIENCE: This one says, the
general simulation hypothesis is inconsequential
and unfalsifiable, because if it were
somehow proven true, it would make no difference
within observable reality, and one can always claim
that reality is a simulation. Is your preferred version
of the simulation hypothesis either consequential
or falsifiable? RIZWAN VIRK: So getting
back to what I just said a few moments ago,
just because you can’t prove it’s not true doesn’t mean
that you can’t find evidence that it is true. And I was just down
at a university down in Southern
California where we’re running some experiments
to try to show evidence that, in fact,
consciousness is needed to collapse the probability
wave, not just a device. Because the device could be
part of the simulation, as well. So if– and depending
on what we find, if in fact you need a conscious
person as opposed to a device, that brings up– back to my preferred version
of the simulation hypothesis, which is that– the RPG version, that we
are conscious entities who are playing characters. Now, is it consequential? I’ll get back to the– to
my last point in this talk. If you were playing
a particular game like “Eve Online,”
your goal may be to build spaceships
and build an empire– intergalactic. Knowing that you’re
in a simulation might alter your
behavior, because you’ll have a different set of
quests and achievements. That’s why I like the video
game analogy more than just a straight simulated
environment itself. AUDIENCE: We want to thank
you so much for being here. RIZWAN VIRK: Thanks, everybody. [APPLAUSE]

18 thoughts on “Rizwan Virk: “The Simulation Hypothesis” | Talks at Google”

  1. 8:25 – No. Bostrum's 2003 paper does NOT assert we may be npc's in a simulation.
    Readers should observe this rather easily, at https://www.simulation-argument.com/
    How can you get something this simple and fundamental, that wrong?
    Using your terms, he posits primarily rpgs within a simulation.

    10:20 – No. Bostrum doesn't "make the point" that civilization never gets to the point of creating world simulations, that is merely one of his three conditions at the head of his paper:

    "This paper argues that at least one of the following propositions is
    true: (1) the human species is very likely to go extinct before reaching
    a “posthuman” stage; (2) any posthuman civilization is extremely
    unlikely to run a significant number of simulations of their
    evolutionary history (or variations thereof); (3) we are almost
    certainly living in a computer simulation. It follows that the belief
    that there is a significant chance that we will one day become
    posthumans who run ancestor-simulations is false, unless we are
    currently living in a simulation. A number of other consequences of this
    result are also discussed."

  2. I organized this edition of "Talks at Google". I asked this question at 44:30 on behalf of a dear friend. I work at Google in the Bay Area.

    I had the privilege to meet Riz in person that day and get to know him a little. I felt that he embodies the truth. In other words, his behavior speaks more to his understanding of reality than does his knowledge. He was one of the most pleasant and peaceful person I've come across and this was especially surprising to me considering the responsibility he must have running a department at MIT while writing a book in parallel and yes, he's also produced movies.

    However, do you believe that he could be right about this? Could be be dreaming our reality? What does it mean that our reality is just a simulation? What did the Buddha mean when he said that the self is an illusion. How can that be? It's madness.

    If madness doesn't scare you, then you should try to perform this simple experiment with your hand eye coordination @ madness.kotuhal.com 🙂

    I hope you enjoy the talk.

    Wake up, Neo…

  3. Oh, darn. I was hoping that Rizwan Virk was not going to go down the heavily worn path of pseudoscience. His idea(s) that it is the observer which is needed for reality to manifest is an attempt to fill a knowledge gap with a heap of fantastical speculation. I am not saying that he ought to stop speculating. I do hope that Rizwan can get back on the path of exploring what might be a testable hypothesis about a simulation hypothesis.

    I am in awe that great minds are getting trapped by what appears to be the descriptor "observer". Look, I do not know anything. But, I do know that making the claim that an observer is necessary moves us from the application of science and of acts which add to our body of knowledge to that of a sci-fi/fantasy steeped speculation.

    If anything, I do like Rizwan's passion for what he is speculating about. May Rizwan work together with those who are our pioneering minds and who are quantum physicists without being sideswiped by all of that unnecessary woo. The universe is exceptionally fascinating and breathtaking without there being any need to add fantastical speculations to our experience of awe and of wonder about the universe. I sound like a curmudgeon. Forgive me.

  4. I believe the whole issue of video games, other than for mere entertainment, is at best a distraction from developing a better understanding of the quantum world and reality. Studying dreams and consciousness are much more worthy pursuits than delving deeper into video games. Think about a basic dream that people may experience on any given night: the brain apparently creates an entire world that, unless it's a lucid dream, cannot be distinguished from reality itself. then we do it again, and again. And most importantly, more often than not, it's a believable reality, at least until we perform some "impossible" task that wakes us up. Science:Fiction is akin to Truth:Religion: if you have the former, you don't need the latter. And games are, well, games! – j q t –

  5. Quoth ]03:30+035s[ What happens to those characters when I'm not logged in
    Insightful ]42:00+090s[ Great first questions regarding ML and dare I say recursive merit system that determines if right action by assessing all outcomes and ranking them. Sounds exactly like what Sussman and RMS were playing with idea wise back in the 70's on the fastest and most efficient learning. Force decision points to gain more information and data which provides outcome assessments retroactively.

    This is why I fear spawning too many minecraft worlds. They are massive and should be attended to.

  6. When he was talking about the rendered objects in the simulation, all I could think of was a recent popular music album by Billie Eilish called "When we all fall asleep, where do we go?" In that album, the song "bury a friend" asks this very question.

  7. would this help explain why massive objects affect time? the more mass involved the "longer" itd take the simulation to process all them bits, slowing time relative to other regions

  8. Isn't the claim at 18:55 about how we've simulated all the neurons and connections in a rat's brain totally bogus? Maybe every TYPE of neuron and TYPE of connection? Or what is he referring to?

Leave a Reply

Your email address will not be published. Required fields are marked *