December 5, 2018 • 48 mins
There’s one last thing. Maybe the reason why we don’t see other intelligent life, maybe the reason we are in the astoundingly unique position of having to save the future of the human race, is because we are simulated human beings. It would explain a lot. (Original score by Point Lobo.)
Interviewees: Nick Bostrom, Oxford University philosopher and founder of the Future of Humanity Institute; Anders Sandberg, Oxford University philosopher; Seth Shostak, director of SETI
Learn more about your ad-choices at https://www.iheartpodcastnetwork.com
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
There is one other thing. I'm including it here is
an epilogue because the series is over, its message has
been sent. But this other thing does tie into existential
risks in some pretty interesting ways, so I wanted to
include it. I'll just leave it here and you can
decide how you feel about it. Back in two thousand three,
(00:28):
Nick Bostrom expanded on a concept that had been around
at least since the classical Greeks first recorded their thoughts
on it, the idea that what we experience as reality
isn't real the followers of Plato discussed in the Academy
at Athens, and then in the sixteen hundreds it was
taken up and examined by Renee Descartes, the Enlightenment philosopher
(00:49):
who famously identified that he was because he thought. Descartes
wondered if perhaps we exist in a reality that's actually
a dream. After considering his own experience with dreams, Descartes
decided that if we do exist in a dream, we
would never be able to tell. What he experienced in
dreams seemed like reality to him just as much as
(01:11):
reality did in waking life. Without any indication to distinguish
between the two, we would never really be able to
tell the difference. It wasn't until two thousand three, however,
when Nick Bostrom wrote his paper entitled Are You Living
in a Computer Simulation? That anyone went to the trouble
(01:32):
of formalizing the idea that what we consider reality is
an actually basement reality. Bostrom hit upon a way to
examine the nature of reality, and it's based on whether
we expect to make it through the great filter. He
called it the simulation argument, which supposes that there's a
very good chance and we are not actually real life humans,
(01:53):
that instead we are simulated humans living in a simulated universe.
So Nick Bolstrom refined this into a proper logical argument
to show that you need to kind of accept that
either we are going to go extinct really soon, or
where for some reason, our post human descember will never
(02:15):
ever run simulations, maybe because very impossible, or because they're
very very moral or have some coordination, or we are
almost certainly in a simulation. If you didn't recognize that voice,
that was Ander Sandberg, Bostrom's colleague at the Future of
Humanity Institute. Being an argument, the simulation argument doesn't provide
evidence one way or the other, which is beautiful because
(02:38):
that leads it to each person to be persuaded by
it or not. But to get into it, we first
have to agree on a couple of points. Remember when
humanities cosmic endowment came up all the way back in
episode three. That's the idea that if we are the
only intelligent life in the universe, as it seems we are,
(03:00):
then once we spread off of Earth, all the matter
and energy that we can get our hands on before
it expands out of our reach forever is ours to
put to use for whatever amazing things we can come
up with. If we managed to successfully navigate the existential
risks that are coming our way and pass through the
great Filter, then it seems likely that our descendants humans
(03:22):
in the future will use some of that cosmic endowment
to create massive computers. Perhaps we will use nanobots to
deconstruct planets to use as raw materials for those massive computers,
and we'll build Dyson's feares to capture energy at the
source of a star to power them. That our descendants
will have massive computers seems like a fairly tame prediction
(03:45):
as far as predictions about the future go. Think about
how important computing is to our civilization today. It's a
pretty good bet that as we continue moving along our
technological path, computing will grow ever more important, and should
we end up a post biological society, computers will become
even more important. They will provide the support structure for
(04:07):
our very being. So if we're agreed that if we
managed to save the world, are descendants will go on
to have vast amounts of computation available to them, we
can move on to the next point that they will
use some of that computing power to run simulations of
us their ancestors. They could have any reason to run
(04:30):
simulations of human history for fun, like the reason we
play the sims or Age of Empires as a sociological
or anthropological model, or as part of an educational exhibit
that celebrates the time that we save the future of
the entire human race and intelligent life in the universe
from near extinction. Whatever their reason, we can imagine that
(04:53):
in the future they might run what Bostrom calls ancestor simulations. Importantly,
one the beautiful things that makes the simulation argument persuasive
is that it doesn't matter when this will happen. The
argument puts no time constraints on any of this. Our
descendants can build those computers and run those simulations a
(05:13):
thousand years from now or a million. It doesn't matter,
just so long as we can agree that at some
point they will. So if we're agreed that our descendants
will at some point in the future build massive computers
and use some of that ridiculous amount of computing power
available to them to run simulations of human history, we
(05:34):
can now enter the simulation argument. It goes something like this,
Either we are living in a computer simulation run by
our descendants in the future, or we're not. If we're
not living in a simulation, then that means we are
what we tend to think. We are members of the
(05:55):
real life human race, living in basement or real reality
in the twenty first century, about three and a half
billion years on from the origin of life on Earth.
But as plainly obvious as that may seem, the simulation
argument raises a question why aren't we simulated? Bostrom identified
(06:15):
a couple of possible reasons. One is that our descendants
are fully capable of simulating us, they just choose not to.
It's possible that our ancestors won't run simulations because they're
computing power is far too precious for even a minute
fraction of it to be allocated to something potentially frivolous
like an ancestor simulation. Or maybe the humans of the
(06:38):
future are too moral to run an ancestor simulation, that
the possibility that suffering could arise in the universe they create,
like what we experience in our own universe now, is
just too much of a moral gamble for their taste.
In other words, they don't want the karmic mark against
them for creating another universe where suffering lives. Or perhaps
(06:59):
running ancestor simulations is just too hard, and of course
it's always possible, they just don't feel like it. But
take a closer look at this possibility. It shares the
same fatal flaw with arguments that look to solve the
family paradox. It presumes that not one, not one single
future human decided to build a simulation of historic humans.
(07:23):
Not a single person figured out a way to allocate
some of that vast amount of computing power to simulate
the universe and humanity in it. Not a single person
came across a scientific study that would benefit from a
model of human history, not a single person was curious
enough of the quadrillions or possibly sex to sillions of
(07:45):
humans left to come, who will populate a span of
time lasting billions of years. Not a single one of
them created an ancestor simulation. That's how it must be,
because all it takes is one for one person in
the future to run one single ancestor simulation, and the
simulation argument is activated. But there's another possibility as well.
(08:11):
Perhaps no ancestor simulations are run in the future because
there are no humans around to run them. Here's where
the simulation argument ties into existential risks if we're not
simulated humans. In a very strange way, the simulation argument
gives us a glimpse across the long expanse of time
into our future. When we look ahead, it shows us
(08:34):
that there are no humans around to run simulations of us.
There are no humans who managed to spread from Earth
and into the universe to put that cosmic endowment to use.
It tells us that we will fail that in the future.
We didn't make it through the Great Filter. But if
you think we will make it through the Great Filter,
(08:56):
and if you're not convinced that not a single one
of our descendants will run a simulation, and of humanity,
you're in luck. There's a third option that we are
simulated humans living in a simulated universe. Yes, anybody can
simply say that we are living in a simulation, just
(09:16):
like anyone can say that we are living in what
we call real reality. Neither one means anything. They're both
just assertions. But the idea that we're living in a
simulated reality actually has a leg up over the idea
that we're living in real reality thanks to probability, lovable, unpredictable, programmable.
It's the Sims. Back at the beginning of the year
(09:40):
two thousand, the software company Maxis released The Sims, a
life simulation game where you, the player, controls individual simulated people.
The Sims. As one reviewer put it, as a celebration
of the mundane Sims go about their lives, they go
to work, they clean their toilets, and yes, they die.
(10:02):
The Sims became an enormously popular game. Between two thousand
and two thousand ten, a hundred and twenty five million
copies of The SIMS and its successive editions were sold
around the world. Each time one of those people brought
their CD RAM home, loaded it into the tray, and
booted up the software. A new iteration of the SIMS
(10:22):
universe was created. It was the same universe. It followed
the same rules, the same program, It followed the same
physics and logic, and the SIMS all navigated their universe
made up of the same set of possibilities, but each
iteration was distinct and different. Each one was a discreet
version of the SIMS universe, and we can expect something
(10:45):
along the same lines with any simulations our ancestors might run.
Whether it's a scientific model for an anthropological study of history,
or whether it's the software of a popular game, whether
it's a class project. Each time that simulation is run,
a new iteration of our simulated human universe is created.
Say that at some point in the future, again, it
(11:07):
doesn't matter when, but at some point across the billions
of years, among the quadrillions of people left to come,
there's a ten year period where an ancestor simulation software
that is equally as popular as the SIMS is created
and sold. As all those future humans load their simulations,
a new iteration of our simulated universe is born, each
(11:29):
following the same guidelines, the same code, each taking radically
different courses within the same set of prescribed rules, the
same set of physics, as it were. If we're agreed
that our descendants will run ancestor simulations of us, and
that we don't have any frame of reference to distinguish
the simulation from real reality, and in the future a
(11:50):
hundred and twenty five million iterations of that simulated universe
or run, then you and I and everyone alive in
our universe has a one and one than twenty five
million chants that we are actual humans, a one and
a hundred and twenty five million chance that we are
not simulated, Or, to put it somewhat less encouragingly, there
(12:11):
is a ninety nine point chance that we are simulated
humans living in an ancestor simulation being run in the future.
Because there was one iteration of humans in real life,
there is a chance that we are members of that
genuine human race as we'd like to think of ourselves.
(12:32):
But well, here's Nick Bostro, it's a little bit as
if we ares s species. I thought we were Napoleon, right,
So there are a lot of people who have thought
that we're Napoleon. Has only been one actually Napoleon. There
have been many more people that thought that we're Napoleon.
Then actually we're in that pogum, And say, if you
think you're kind of an emperor of world historic importance,
you know, a chances are maybe you're actually a lunatic
(12:54):
in and insane asylum. Because so many simulations are running
in the future, given that we have no firm of
reference to tell us any differences between our reality and
basement reality, that utter lack of signs that we live
in a dream, as Descartes suggested, the chances are vastly
greater that we are simulated rather than real. There's just
simply been more simulated humans than real ones, and being
(13:18):
members of one group or the other, the chances state
quite plainly that it's like clear we're members of the
simulated group. You can mess with the numbers either way
to make the chances go up or down. Say that
ancestor simulation is way more popular than the sims ever was,
and an equal amount of copies are sold each decade
for a hundred years. That would mean that we had
(13:41):
one chance in over a billion of being a human,
and so on. Each new iteration of our simulated universe
lowers the chances that we are real. But like I
said before, even just one ancestor simulation run in the
future activates the simulation argument eve. And if just once
in the whole future history of the human race, only
(14:04):
one iteration of our simulated universe is run, we still
have even odds that were simulated. If we make it
through the great filter, and our descendants run ancestor simulations
of us, we have at best a fifty chance of
being real humans. I should say that if it does
(14:36):
turn out that we are simulated, our simulation could have
been built by any number of things. It's possible the
people who created us aren't future humans at all, but
some other intelligent life entirely. It could be a race
of Martian ancestors that real humans diverged from eons ago,
an intelligent machine, super intelligent mice looking for the answer
(14:58):
to the meaning of life. It's even possible that our
universe follows a more basic set of physics than what
governs the universe. Our universe is stimulated within. Perhaps there
never was a one true race of real life humans,
anymore than there was a real life race of sims.
But the thing about the simulation argument is that it
calls for the least amount of speculation, the least amount
(15:21):
of unreality. For it to be right, all it requires
is that our descendants will run simulations of humanity in
the future. It's rather elegant in its simplicity. That's what
makes it convincing. Even still, there are, as you may imagine,
a number of objections to the simulation argument. Some people
(15:44):
simply detest it on its face, dismissing it out of
hand is nothing more than material for late night conversations
in college dorms. But others engage with it seriously, and
they raise legitimate issues. The thing is, it seems that
every object action leveled against the simulation argument either imposes
unnecessary time constraints or is just a simple failure of imagination.
(16:10):
If this is the first you've heard of the simulation argument,
at this point, you might be looking around you closely
examining the edges of things, seeing if you can tell
that the world is fake. Don't bother. The simulation argument
doesn't change a single thing about our universe. The physics
are the same as ever. Being persuaded that we live
in a simulation doesn't unlock some greater understanding of our
(16:33):
reality such that we can alter it in ways that
we couldn't before. There's no pill to choose. There won't
be any flying around or walking through walls, nor should
your morality be altered. Your mother is just as real
as she was before, and she would be just as
disappointed as before if you decided to start a life
of crime because we're simulated, So why does it matter.
(16:55):
The court system is real too, and you would be
imprisoned by it just as sure as ever. And when
you explain to them that none of this is real,
that they aren't real, the car you set fire too
isn't real, that the whole universe isn't real, the mental
health system will be just as happy as ever to
lock you away in a padded room indefinitely, the walls
of which you would find you were fully enabled to
(17:17):
will yourself to walk through using your new found awareness
of the simulated basis of our universe. Our universe is
as real as it ever was. The only thing the
simulation argument can change is your perspective. One common objection
(17:37):
to the simulation argument kind of falls under this umbrella
of misunderstanding. It goes something like this, how could a
simulated banana or simulated apple, or a simulated cheeseburger possibly
provide nutrients to us. How could it sustain and nourish us,
alleviate our hunger, make us feel full. It's kind of
(17:59):
like the hard problem of consciousness, the puzzle of how
we get from neurons firing two subjective experience. Sure, you
can simulate a banana, but you're simulating a banana. It's
on a computer. And those of us who are looking
at the banana, no, we are looking at a banana
pictured on a computer. How can we account for the
experience of interacting with a banana, holding one, peeling it,
(18:22):
and needing it, of being allergic to it? How do
we get from a simulated banana to the experience of
a banana that is indistinguishable from reality. The answer is
rather simple. We're not simulating that banana or cheeseburger or whatever,
so it's not a simulation to us. It's made of
the same stuff that we are. It's no more fake
(18:44):
than you, You're no more real than it. We share
the same reality. The banana isn't a simulated banana in
our universe. It's just a banana, so it's subject to
the same rules that we are on the quantum level.
Those bananas are made up of energetic vibrations just like us.
They form particles that are held together by the fundamental forces,
(19:07):
just like us, and they form larger and more stable
structures that increase in size and complexity up to the
level of what we think of as a banana. And
when we eat that particular package of energetic vibration, it
interacts with our own energetic vibrations in such a way
that we experience taste, in the sensation of deliciousness, We
(19:28):
experience feeling full if we eat too many of them,
and our bodies break them down into smaller parts, and
we use those parts for energy, same as ever. Banana objections,
as we'll call them, are rooted in the idea that
we exist in some different reality from the banana. To
be clear, the simulation argument does not suppose that there's
(19:49):
some real version of you elsewhere. What we experience as
our universe, bananas and all, is a simulation to someone else.
To us, it is our full reality. So to answer
the banana objection, the banana can satiate and sustain us
because it's as real as we are. Other objections go
(20:10):
a little deeper. How could we possibly simulate the universe
on anything approaching a detail fine grained enough to be
indistinguishable from the original. There are a couple of answers
to this one. First, we can't. We don't have the
computing power to faithfully render the universe in that detail.
But in this objection we run into an unnecessary time constraint.
(20:34):
While we can't faithfully reproduce the universe, there's a pretty
good chance that our descendants will be able to. We
Those of us alive today are already on track to
doing this. For the last several years, humans have been
able to produce accurate simulations of quarks and glue ones
(20:54):
and the strong nuclear force that binds them together. These
models are not infinitely press size, but they're faithful enough
to reality that a simulated scientists living in the simulated
universe where the protons exist would have no way to
distinguish their unreality. They would just be protons that behaven
appear exactly like any other protons. The models that we
(21:16):
can produce today simulate an exceedingly tiny portion of our universe.
They're on the scale of femptometers, about a quadrillionth of
a meter. But the models show a proof of concept
that simulating the universe is possible, and they also show
that the human race already has a desire to simulate
the universe. Already at this early stage in modeling the universe,
(21:41):
we found that there are problems future humans will run
into when it comes to scaling up those models. It's
pretty obvious that modeling the entire universe would be a very,
very difficult undertaking, but difficulties can be worked out over time.
The problem comes when we reach impossibilities, and in two
thousand seventeen, a pair of researchers from Oxford University found
(22:05):
an impossibility that seems to show pretty conclusively that we
do not live in a simulated universe. The researchers found
that it would be physically impossible to simulate at least
one aspect of the universe, something called the quantum Hall effect,
which describes the way that electrons bounce between energy states
called a quantum leap in the presence of a magnetic
(22:28):
field and extremely low temperatures. It occupies a arcane a
corner of particle physics as there is, But the point
is the researchers found that to accurately simulate the effect.
The computational power required doubles each time you add a
new particle to the model, rather than increasing the power
needed by one unit, it grows exponentially, so within just
(22:51):
a few hundred particles, the operations per second required by
a computer to run a simulation of the quantum Hall
effect grows to a total beyond the number of atoms
there are in the universe, around ten to the eighty
second power. Not only is that number vastly more operations
than our computers can perform today, this also presents a
(23:11):
major technical issue. We need electrons to perform the computations
that create the simulated model. In the first place, electrons
are what we use to carry information in our computers,
and as we simulate more and more particles subject to
the quantum Hall effect, we will start to require a
substantial portion of the electrons that are part of those
(23:32):
ten to the eighty second power atoms in the universe.
And as you may know, those electrons are otherwise occupied
performing other functions like making up all the matter in
the universe, which means that eventually it becomes physically impossible
to simulate the quantum Hall effect and since we can
observe the quantum Hall effect in our own universe, we
(23:55):
can reason that our universe is not a simulation. But
this presents a failure of imagination, not an impossibility. Indeed,
the researchers who made the discovery noted that they're finding
only showed that we couldn't possibly be simulated within a
traditional computer, the type that we have available to us now,
the kind that uses electrons. They also pointed out that
(24:18):
we can probably safely presume that the humans of the
future would be simulating us on better computers than what
we're using today. I've said it before, and I'll say
it again. If it turns out that the humans of
the future don't have better computers than we have now,
it would be one of the more surprising things in
this entire series. And indeed, we are working on better
(24:40):
computers already, specifically, quantum computers, the type of computer that
processes information not gleaned from the movement of electrons, but
instead from the interaction of elementary particles, which exponentially opens
up a computer's capacity to perform calculations in a single
second and meet lee transfers, rendering the quantum Hall effect
(25:02):
and a simulation from the impossible column over to the
column marked possible. We've already talked about that to accurately
simulate a human or to require simulating a mind that
is capable of subjective experience. If the hard problem of
(25:23):
consciousness isn't actually a problem after all, then by creating
an accurate model of all of the bits that make
up a human mind, the neurons and synapses, the neurochemicals,
we can imagine that they will produce subjective experience when
they interact with the matter and energy that makes up
the simulated universe. This is exactly the same concept as
(25:44):
a post biological civilization that has shed their physical form
and migrated onto computers. If you found the concept of
a post biological future for humanity plausible, then you should
have no trouble accepting the simulation argument. The only difference
between the two is that, rather than us migrating onto
computers ourselves, we will have always been there already. As
(26:10):
Nick Bostrom points out in his paper on the simulation argument,
the human brain can process about ten to the seventeenth
power operations per second. If that's the amount of processing
power it takes to experience subjective reality. Then that's about
what we would expect it would take to simulate a
human mind, and to run a simulation of all of
human history would require that amount times around a hundred
(26:32):
and seven billion, since that's the number of humans who
have been born up to this point. Although we can't
imagine that as humans die, the processing power required to
produce their subjective experience would be freed up to power
the minds of new humans as they're born. So perhaps
all you'd ever need is enough processing power to simulate
all of the minds of the humans who are alive
(26:54):
at any given point, which means that to simulate a
global population of ten bills people, you'd need a computer
capable of processing ten to seven power operations per second.
That's not including the processing power it would take to
simulate the universe, which we could expect would expand the
requirements to mind boggling numbers. There are surely a lot
(27:18):
of corners future humans can cut to create a simulation.
They wouldn't need to simulate the entire universe obviously, just
what we can detect. Then it's possible they wouldn't need
to simulate the universe at all. Rather, they could just
create a program that stimulates our minds directly to experience
the universe. Even with the shortcuts, though, there's still an
(27:39):
enormous amount of ground to cover before we get to
the point where we have the processing power of the
magnitude required to simulate a human mind, let alone all
seven and a half billion of us. Take this for example.
Back in two thousand fourteen, the Fujitsu Corporation ran a
test on its Fujitsu K, at the time, the fourth
(28:00):
fastest supercomputer in the world, ranking that is no small feat.
In the test, the Fujitsu engineers wanted to see how
long it took their K computer to process one percent
of the amount of information that a human brain can
process in one second. The K managed to process that information,
but rather than the hundredth of a second it would
(28:22):
have taken a brain, it took the computer forty minutes
to complete the task. And it's worth noting the K
required twelve point seven megawatts of electricity throughout the forty
minutes it was running this test, about the same amount
of electricity needed to power the entire city of Sacramento, California.
(28:45):
Considering the gap between where we are now and what's
required to run a respectable ancestor simulation. It might start
to seem that such simulations are indeed impossible but easy.
There we're only running up against an unnecessary time constraint. Again,
given enough time, it seems quite possible to build a
computer that could run an accurate simulation. As Boston points out,
(29:08):
a single one of those planet sized computers could perform
ten to the forty second operations per second. To run
an ancestor simulation would require just the slightest fraction of
that computer's processing power, And we can imagine that in
a far enough expanse into the future, if our descendants
have one planet sized computer, they will eventually have a
(29:30):
great many of them. And let's not forget we humans
have a long standing tradition of advancing by leaps and
bounds in computing. If you wanted to find the top
of the line in supercomputers back in you needed to
look no further than the Cray Too. Recently, scientists producing
these simulations at NASA's Aims Research Center in Mountain View, California,
(29:53):
began using a new supercomputer, the Cray Too. It was
just a little smaller than the size of a BW
beatles stood on its end, but the CRY two could
perform two hundred and fifty million operations per second, an
astounding amount. In just over thirty years later, Apple released
a computer that could perform six hundred billion operations per second.
(30:17):
Apple called There's the iPhone tent, which, by the way,
is run by a machine learning algorithm capable of self improvement.
We already share the planet with quantum computers, the very
kind of computer that could simulate a universe to a
detail down to the quantum grain. The ones we share
the planet with today are exceedingly slow compared to our
(30:39):
current traditional computers, but they are in the earliest stages
of development. It was only in the eighties that physicists
began designing quantum computers that could work in theory, and
only in the nineties when the first quantum computers were built.
We can imagine that if we continue to follow this arc,
those primitive quantum computers of today will advance as well
(31:00):
as our traditional ones have. We haven't even lived with
traditional computers for a century, yet imagine what our quantum
computers will be capable of a century from now. There are,
to be sure, numerous other objections to the simulation argument,
but most, if not all, have difficulty overcoming that non
(31:21):
limitation of time. When we give our descendants the luxury
of time, those people whose future lives we will save
by taking on our existential risks become capable of anything
that we can imagine and more, including simulating us. Back
(31:53):
in two thousand twelve, a group of physicists from Bond
University in Germany announced that they had figured out a
way to test whether we live in a simulation. If
our simulation is anything like the femtometer scale simulated models
of protons that we can produce now, then that means
it isn't an infinitely accurate copy of the real universe
(32:14):
predicted by Einstein's theory of relativity. One of the things
that relativity says is that the universe is built from
a structure made up of the three spatial dimensions and
time spacetime. Spacetime is the fabric of our universe, and
it's contiguous, one giant universe sized unit. There's no gaps
(32:35):
in it anywhere, no matter how far we zoom in.
Remember when energy fields came up in the Physics episode
and I mentioned that these energy fields are everywhere at
every point in time and space. That's what that means.
There's no place where those fields aren't because there's no
gaps in spacetime. There's no place where space doesn't exist,
(32:55):
just as there aren't any gaps in time where there's
no time. It would take an enormous computer to render
an infinitely accurate version of our universe, potentially one the
exact size of the universe, and so we might guess
that the designers of our simulation would be forced to
make a version that's just slightly less accurate than the
(33:16):
real thing, which would mean that the fabric of the
simulated universe wouldn't be infinitely smooth. There would be gaps
in its structure gept between the individual units that it's
built from. If we could only zoom in close enough
to examine it, we would find the fabric of our
simulated universe was pixelated. But we can't zoom in that
(33:36):
close to observe space time directly. So the Bond University
researchers suggested a way that we could indirectly observe it
by watching how these smallest bits of matter in the
universe ultra high energy cosmic rays travel around space. Remember
Cosmic rays are high energy particles most likely spit out
from supernova that zip around our universe and collide with
(33:59):
other particles and may or may not produce microscopic black
holes when they do. They're one of the things that
we model in particle colliders. If spacetime is contiguous in
our universe, as we think it's supposed to be, then
cosmic race should travel in every possible direction. Since there's
no gaps in spacetime, there's nothing to direct their paths.
(34:21):
There aren't any channels in the fabric of space for
them to fall into and travel along. But if spacetime
is pixelated, the very fabric of the universe would form
an infinitismally fine grid. There would be tracks that the
tiniest cosmic rays might travel along. So, say the Bond researchers,
if we managed to observe that cosmic rays follow a
(34:43):
grid pattern, or that they show a preference for traveling
at angles rather than moving in every possible direction as
they should, we might have a fairly good indication that
what we're viewing as reality is actually a simulated version
of it. Interestingly, the idea that spacetime is grainy rather
(35:05):
than smooth is starting to gain traction among physicists. The
absolutely smallest discrete measure of distance in the universe is
called the plank length. Anything smaller than it, and physics
just falls to pieces. It's difficult to get across how
small a plank length is. It's about a hundred quintillion
times smaller than a proton, which itself is about a
(35:28):
hundred thousand times smaller than an atom. In fact, if
we could zoom into the plank scale so that a
plank length was the size of an atom, the atoms
in the universe would be the size of galaxies. It's
that small. Since spacetime is supposed to be infinitely smooth,
it's weird that there's the smallest point beyond which physics
(35:50):
is just meaningless. It's possible that our math is wrong,
or that when we managed to marry relativity and the
standard model into a theory of quantum gravity, we'll find
that physics picks up again beyond the plank scale. But
some quantum physicists are beginning to wonder if perhaps there
is some non infinite bottom size to the fabric of
(36:11):
space time, and that the plank length maybe it. Perhaps
we're beginning to uncover the frame of reference they Carte
longed for to distinguish between our reality and real reality.
But even as we chip away at the nature of
our own reality, if we find that we are indeed simulated,
(36:33):
we run into an issue. We still don't know anything
about the nature of real reality. All of our questions
just get kicked up to the universe that's simulating us.
There's a problem with that because we can't say for
sure that the universe we are simulated within isn't simulated itself.
So the problem can be kicked up indefinitely. Since in
(36:56):
the simulation argument, the people who are simulating us are
also humans themselves. They are, remember, the descendants of the
people we are simulations of. The same staggering probabilities against
us being real life humans applies to them as well.
Just by their membership and the human race and the
existence of ancestor simulations. The people who simulated us have
(37:18):
that same ninety nine point chance of being simulated just
because they are running the simulation of their ancestors doesn't
identify them as members of the real human race. The
simulation can run inside a simulation. When our descendants run
ancestor simulations. In the future, there will be simulated versions
(37:42):
of the universe existing within our own. But if we
are simulated, then the simulations we run will be simulations
within our simulation, and the same possibility applies to the
people running our simulation, to their universe may be simulated
as well. Perhaps it's an almost certainty that once a
(38:02):
simulated group of humans reaches some specific point, they will
begin to run ancestor simulations themselves, the ones that make
it through the great filter at least, and so ancestor
simulation spread in each direction from our own universe, simulations
within simulations within simulations, kind of like how it looks
(38:23):
inside an elevator that has mirrors on all sides. We
ner what's called an infinite regress problem. We can figure
out that our universe is simulated, but what does that
tell us about the universe the people who are simulating
us live in. Actually, this is familiar territory for physics.
One implication of cosmology is that our universe is one
(38:45):
bubble amid an infinite froth of bubbles, each their own universe.
And since our universe is an isolated system, no matter
or energy can leak out of it, and we presume
the same about those other universes, which means that we
can't glean any information about those other universes that we
rub up against in the infinite froth, and we likely
(39:06):
never will. And yet you would be hard pressed to
find a physicist who suggested throwing in the towel on
investigating our own universe when cosmology reached this conclusion. The
same holds true for the idea that we're simulated. Not
being able to know anything about the universe that simulated
us doesn't mean there's nothing to be gained from investigating
(39:26):
our own It's worth taking a step back here to
look at the implications of what it means. When we
start talking about who is simulating us, we're talking about
someone who in some respect created us. Perhaps one of
the touchiest aspects of the simulation argument is that it
(39:49):
suggests a sort of techno creationism. Although it does fall
under the umbrella of creationism, accepting the idea that we
may be simulated does to require any form of supernatural
or religious belief. If we're simulated, we can identify who
created us. It's the descendants of the people who were
simulations of future humans created us. But who exactly would
(40:16):
the software developer who wrote the code that produced us
qualify as our creator? Would the person who directly booted
up the simulation we live in? Or what if our
simulation is a scientific model, does our creator where a
lab coat and our simulation is a digital Petrie dish
of sorts? Or is our creator a far future teenager
who created us for a class project. I should remind
(40:39):
you that it's important to remember the simulation argument doesn't
prove anything about our perceived reality. It gives no evidence
one way or the other. It's a framework for viewing
the nature of our universe, which means that, like any
other argument, you can take or leave it. It is
a matter of personal belief whether the idea of being
(41:00):
created from a program written by future humans is more
or less comforting than the idea that a supernatural deity
created us, or that the world grew on the back
of a turtle, or that we're the fluke result of
an incomprehensible series of cause and effect stretching back nearly
fourteen billion years. If we did one day discover that
(41:27):
we are stimulated humans living in a simulated universe, that
would we can imagine have enormous and sweeping effects on us.
Even if you already believe that we were created, how
would you feel to learn that your creator is only human,
with all the same faults and flaws that you have.
And if you don't believe today that we were created,
(41:48):
that we are merely the result of some physical processes
that can turn dead matter into cellular life, and that
we were shaped by the forces of evolution, learning indisputably
that we were indeed created, it would be a tough
pill to swallow. But would those enormous and sweeping effects
be all bad and would they be permanent? This is
(42:09):
Ander Sandberg again. I do think that if we realize
that the universe fundamentally work differently from how we used
to believe it was, we would be shaking, But we
would also get over it, just like we have gotten
over and the hellyo some traces, and the evolution and
the Freud and so on. Most people are way more
resilient to upsets of the fundamental structure of the world
(42:32):
than we give them credit for. We humans are really
good at absorbing terrible news, whether it's a cancer diagnosis
or learning that we're facing existential risks. There's a chance
we will survive, there's a chance we won't. But do
we simply lay down and die. Humans tend not to
do that. Instead, we carry on. We seek out connections
(42:55):
and support from other humans who are also carrying on,
and we might expect the same would hold true should
we learn that we're simulated. Even despite finding that we
are the result of code, nothing about the meaning of
our lives should necessarily be lost. Well where do we
get a meaning in our own lives? And I would
argue that we're actually creating it ourselves. Even if the
(43:17):
universe was a class project to demonstrate the sociological and thing,
that doesn't mean it's the meaning of our life to
not behave socially, It's not like our loves and the
emotions that will become meaningless, and we might still have
a meaning that's very different from the reason the simulation exists.
It will probably be tough to keep in mind, especially
(43:40):
for the people who are alive at the time, when
we learn our universe is simulated, But at bottom, really
nothing much changes for us. Our understanding of our nature
only deepens in a very direct way. Learning we are
stimulated would focus our investigation into our universe it would
certainly explain a lot. The Bond University researchers who came
(44:02):
up with the cosmic ray grid test suggested that perhaps
we'd find the reason the Higgs field is in in
its lowest energy state is because of a rounding error
and the computations that produce our simulation. Maybe the measurement
problem that quantum particles seem to exist in all of
the possible forms they can take at the same time
is a shortcut to make rendering the universe more efficient.
(44:24):
The fine tuning we see would certainly be explained. The
parameters of the forces and energy that make up our
universe would have to be the way they are because
we could not be produced under other conditions. But that's correct,
whether we're simulated or not. Finding we're simulated would identify
computer code as the basis of those governing principles that
(44:45):
we already take to be true, like a biogenesis and evolution.
We wouldn't expect the understanding the truth of that would
change how those processes function, but perhaps we could learn
how to change them ourselves. Maybe learning we are simulated
would free us to abandon the notion of fate and
take control of how the universe operates to shape it
(45:07):
more to our liking. After all, isn't that what the
study of physics is? Taking control of the universe would
also focus the moral responsibility that we have in making
our universe a better place. Having the ability to reduce
suffering wherever we found it, we would be morally obligated
to do so. But then that's the case already as well.
(45:29):
The difference is that we couldn't let ourselves off the
hook by chalking that suffering up to inevitability like we
do now. This is Seth show Stack, the senior astronomer
at City who you met in the first episode on
the Family Paradox. The question that I asked the Professor
Bostrom that I think was most relevant to me was, Okay,
(45:50):
if this is all a simulation, this conversation we're having,
it's not real. It's just code. Right. If it's all
just code, do I have to live a moral existence
or kind just have fun? He thought it was better
to live a moral existence, uh, you know, And I
thought that that was kind of a testimony to how
he really felt on the matter, that you know, you
better act like it's real. The beautiful thing about the
(46:21):
simulation argument is that it doesn't change anything, only our perspective.
The world continues to operate in exactly the same way
as before. It's just that some of the mystery is
brushed away. Nothing about humanity changes, nor does anything about
where we find ourselves in the position. Where as Toby
Ord paraphrase Carl Sagan, we humans have grown powerful before
(46:44):
we've grown wise, the existential risks we face remain as
real as ever, and if we did find we are simulated,
our existential risks would only sharpen, since we would learn
of a new one that our simulation could be turned off,
which I opposed, would qualify as a natural existential risk.
(47:04):
Whether we arrived at this point where we have grown
powerful before we've grown wise, by total fluke or by predetermination,
whether we are members of the real human race or simulated,
we are in the same unique and precarious position that
we were in before. We are still entering the mouth
of the great filter, and we are still at risk
(47:26):
of our existence coming to a permanent end. If we
fail to come together as a world and learn to
use science to help us grow wiser faster, we will
surely not survive. We will accidentally wipe ourselves out with
our unwise power, and that will be the end of
humans in our universe. Does it matter whether that universe
arose from a big bang or from a system boot,
(47:49):
Whether we are real or simulated, our reality is the same.
There is one other thing. I'm including it here is
an epilogue because the series is over, its message has
been sent. But this other thing does tie into existential
risks in some pretty interesting ways, so I wanted to
include it. I'll just leave it here and you can
decide how you feel about it. Back in two thousand three,
(00:28):
Nick Bostrom expanded on a concept that had been around
at least since the classical Greeks first recorded their thoughts
on it, the idea that what we experience as reality
isn't real the followers of Plato discussed in the Academy
at Athens, and then in the sixteen hundreds it was
taken up and examined by Renee Descartes, the Enlightenment philosopher
(00:49):
who famously identified that he was because he thought. Descartes
wondered if perhaps we exist in a reality that's actually
a dream. After considering his own experience with dreams, Descartes
decided that if we do exist in a dream, we
would never be able to tell. What he experienced in
dreams seemed like reality to him just as much as
(01:11):
reality did in waking life. Without any indication to distinguish
between the two, we would never really be able to
tell the difference. It wasn't until two thousand three, however,
when Nick Bostrom wrote his paper entitled Are You Living
in a Computer Simulation? That anyone went to the trouble
(01:32):
of formalizing the idea that what we consider reality is
an actually basement reality. Bostrom hit upon a way to
examine the nature of reality, and it's based on whether
we expect to make it through the great filter. He
called it the simulation argument, which supposes that there's a
very good chance and we are not actually real life humans,
(01:53):
that instead we are simulated humans living in a simulated universe.
So Nick Bolstrom refined this into a proper logical argument
to show that you need to kind of accept that
either we are going to go extinct really soon, or
where for some reason, our post human descember will never
(02:15):
ever run simulations, maybe because very impossible, or because they're
very very moral or have some coordination, or we are
almost certainly in a simulation. If you didn't recognize that voice,
that was Ander Sandberg, Bostrom's colleague at the Future of
Humanity Institute. Being an argument, the simulation argument doesn't provide
evidence one way or the other, which is beautiful because
(02:38):
that leads it to each person to be persuaded by
it or not. But to get into it, we first
have to agree on a couple of points. Remember when
humanities cosmic endowment came up all the way back in
episode three. That's the idea that if we are the
only intelligent life in the universe, as it seems we are,
(03:00):
then once we spread off of Earth, all the matter
and energy that we can get our hands on before
it expands out of our reach forever is ours to
put to use for whatever amazing things we can come
up with. If we managed to successfully navigate the existential
risks that are coming our way and pass through the
great Filter, then it seems likely that our descendants humans
(03:22):
in the future will use some of that cosmic endowment
to create massive computers. Perhaps we will use nanobots to
deconstruct planets to use as raw materials for those massive computers,
and we'll build Dyson's feares to capture energy at the
source of a star to power them. That our descendants
will have massive computers seems like a fairly tame prediction
(03:45):
as far as predictions about the future go. Think about
how important computing is to our civilization today. It's a
pretty good bet that as we continue moving along our
technological path, computing will grow ever more important, and should
we end up a post biological society, computers will become
even more important. They will provide the support structure for
(04:07):
our very being. So if we're agreed that if we
managed to save the world, are descendants will go on
to have vast amounts of computation available to them, we
can move on to the next point that they will
use some of that computing power to run simulations of
us their ancestors. They could have any reason to run
(04:30):
simulations of human history for fun, like the reason we
play the sims or Age of Empires as a sociological
or anthropological model, or as part of an educational exhibit
that celebrates the time that we save the future of
the entire human race and intelligent life in the universe
from near extinction. Whatever their reason, we can imagine that
(04:53):
in the future they might run what Bostrom calls ancestor simulations. Importantly,
one the beautiful things that makes the simulation argument persuasive
is that it doesn't matter when this will happen. The
argument puts no time constraints on any of this. Our
descendants can build those computers and run those simulations a
(05:13):
thousand years from now or a million. It doesn't matter,
just so long as we can agree that at some
point they will. So if we're agreed that our descendants
will at some point in the future build massive computers
and use some of that ridiculous amount of computing power
available to them to run simulations of human history, we
(05:34):
can now enter the simulation argument. It goes something like this,
Either we are living in a computer simulation run by
our descendants in the future, or we're not. If we're
not living in a simulation, then that means we are
what we tend to think. We are members of the
(05:55):
real life human race, living in basement or real reality
in the twenty first century, about three and a half
billion years on from the origin of life on Earth.
But as plainly obvious as that may seem, the simulation
argument raises a question why aren't we simulated? Bostrom identified
(06:15):
a couple of possible reasons. One is that our descendants
are fully capable of simulating us, they just choose not to.
It's possible that our ancestors won't run simulations because they're
computing power is far too precious for even a minute
fraction of it to be allocated to something potentially frivolous
like an ancestor simulation. Or maybe the humans of the
(06:38):
future are too moral to run an ancestor simulation, that
the possibility that suffering could arise in the universe they create,
like what we experience in our own universe now, is
just too much of a moral gamble for their taste.
In other words, they don't want the karmic mark against
them for creating another universe where suffering lives. Or perhaps
(06:59):
running ancestor simulations is just too hard, and of course
it's always possible, they just don't feel like it. But
take a closer look at this possibility. It shares the
same fatal flaw with arguments that look to solve the
family paradox. It presumes that not one, not one single
future human decided to build a simulation of historic humans.
(07:23):
Not a single person figured out a way to allocate
some of that vast amount of computing power to simulate
the universe and humanity in it. Not a single person
came across a scientific study that would benefit from a
model of human history, not a single person was curious
enough of the quadrillions or possibly sex to sillions of
(07:45):
humans left to come, who will populate a span of
time lasting billions of years. Not a single one of
them created an ancestor simulation. That's how it must be,
because all it takes is one for one person in
the future to run one single ancestor simulation, and the
simulation argument is activated. But there's another possibility as well.
(08:11):
Perhaps no ancestor simulations are run in the future because
there are no humans around to run them. Here's where
the simulation argument ties into existential risks if we're not
simulated humans. In a very strange way, the simulation argument
gives us a glimpse across the long expanse of time
into our future. When we look ahead, it shows us
(08:34):
that there are no humans around to run simulations of us.
There are no humans who managed to spread from Earth
and into the universe to put that cosmic endowment to use.
It tells us that we will fail that in the future.
We didn't make it through the Great Filter. But if
you think we will make it through the Great Filter,
(08:56):
and if you're not convinced that not a single one
of our descendants will run a simulation, and of humanity,
you're in luck. There's a third option that we are
simulated humans living in a simulated universe. Yes, anybody can
simply say that we are living in a simulation, just
(09:16):
like anyone can say that we are living in what
we call real reality. Neither one means anything. They're both
just assertions. But the idea that we're living in a
simulated reality actually has a leg up over the idea
that we're living in real reality thanks to probability, lovable, unpredictable, programmable.
It's the Sims. Back at the beginning of the year
(09:40):
two thousand, the software company Maxis released The Sims, a
life simulation game where you, the player, controls individual simulated people.
The Sims. As one reviewer put it, as a celebration
of the mundane Sims go about their lives, they go
to work, they clean their toilets, and yes, they die.
(10:02):
The Sims became an enormously popular game. Between two thousand
and two thousand ten, a hundred and twenty five million
copies of The SIMS and its successive editions were sold
around the world. Each time one of those people brought
their CD RAM home, loaded it into the tray, and
booted up the software. A new iteration of the SIMS
(10:22):
universe was created. It was the same universe. It followed
the same rules, the same program, It followed the same
physics and logic, and the SIMS all navigated their universe
made up of the same set of possibilities, but each
iteration was distinct and different. Each one was a discreet
version of the SIMS universe, and we can expect something
(10:45):
along the same lines with any simulations our ancestors might run.
Whether it's a scientific model for an anthropological study of history,
or whether it's the software of a popular game, whether
it's a class project. Each time that simulation is run,
a new iteration of our simulated human universe is created.
Say that at some point in the future, again, it
(11:07):
doesn't matter when, but at some point across the billions
of years, among the quadrillions of people left to come,
there's a ten year period where an ancestor simulation software
that is equally as popular as the SIMS is created
and sold. As all those future humans load their simulations,
a new iteration of our simulated universe is born, each
(11:29):
following the same guidelines, the same code, each taking radically
different courses within the same set of prescribed rules, the
same set of physics, as it were. If we're agreed
that our descendants will run ancestor simulations of us, and
that we don't have any frame of reference to distinguish
the simulation from real reality, and in the future a
(11:50):
hundred and twenty five million iterations of that simulated universe
or run, then you and I and everyone alive in
our universe has a one and one than twenty five
million chants that we are actual humans, a one and
a hundred and twenty five million chance that we are
not simulated, Or, to put it somewhat less encouragingly, there
(12:11):
is a ninety nine point chance that we are simulated
humans living in an ancestor simulation being run in the future.
Because there was one iteration of humans in real life,
there is a chance that we are members of that
genuine human race as we'd like to think of ourselves.
(12:32):
But well, here's Nick Bostro, it's a little bit as
if we ares s species. I thought we were Napoleon, right,
So there are a lot of people who have thought
that we're Napoleon. Has only been one actually Napoleon. There
have been many more people that thought that we're Napoleon.
Then actually we're in that pogum, And say, if you
think you're kind of an emperor of world historic importance,
you know, a chances are maybe you're actually a lunatic
(12:54):
in and insane asylum. Because so many simulations are running
in the future, given that we have no firm of
reference to tell us any differences between our reality and
basement reality, that utter lack of signs that we live
in a dream, as Descartes suggested, the chances are vastly
greater that we are simulated rather than real. There's just
simply been more simulated humans than real ones, and being
(13:18):
members of one group or the other, the chances state
quite plainly that it's like clear we're members of the
simulated group. You can mess with the numbers either way
to make the chances go up or down. Say that
ancestor simulation is way more popular than the sims ever was,
and an equal amount of copies are sold each decade
for a hundred years. That would mean that we had
(13:41):
one chance in over a billion of being a human,
and so on. Each new iteration of our simulated universe
lowers the chances that we are real. But like I
said before, even just one ancestor simulation run in the
future activates the simulation argument eve. And if just once
in the whole future history of the human race, only
(14:04):
one iteration of our simulated universe is run, we still
have even odds that were simulated. If we make it
through the great filter, and our descendants run ancestor simulations
of us, we have at best a fifty chance of
being real humans. I should say that if it does
(14:36):
turn out that we are simulated, our simulation could have
been built by any number of things. It's possible the
people who created us aren't future humans at all, but
some other intelligent life entirely. It could be a race
of Martian ancestors that real humans diverged from eons ago,
an intelligent machine, super intelligent mice looking for the answer
(14:58):
to the meaning of life. It's even possible that our
universe follows a more basic set of physics than what
governs the universe. Our universe is stimulated within. Perhaps there
never was a one true race of real life humans,
anymore than there was a real life race of sims.
But the thing about the simulation argument is that it
calls for the least amount of speculation, the least amount
(15:21):
of unreality. For it to be right, all it requires
is that our descendants will run simulations of humanity in
the future. It's rather elegant in its simplicity. That's what
makes it convincing. Even still, there are, as you may imagine,
a number of objections to the simulation argument. Some people
(15:44):
simply detest it on its face, dismissing it out of
hand is nothing more than material for late night conversations
in college dorms. But others engage with it seriously, and
they raise legitimate issues. The thing is, it seems that
every object action leveled against the simulation argument either imposes
unnecessary time constraints or is just a simple failure of imagination.
(16:10):
If this is the first you've heard of the simulation argument,
at this point, you might be looking around you closely
examining the edges of things, seeing if you can tell
that the world is fake. Don't bother. The simulation argument
doesn't change a single thing about our universe. The physics
are the same as ever. Being persuaded that we live
in a simulation doesn't unlock some greater understanding of our
(16:33):
reality such that we can alter it in ways that
we couldn't before. There's no pill to choose. There won't
be any flying around or walking through walls, nor should
your morality be altered. Your mother is just as real
as she was before, and she would be just as
disappointed as before if you decided to start a life
of crime because we're simulated, So why does it matter.
(16:55):
The court system is real too, and you would be
imprisoned by it just as sure as ever. And when
you explain to them that none of this is real,
that they aren't real, the car you set fire too
isn't real, that the whole universe isn't real, the mental
health system will be just as happy as ever to
lock you away in a padded room indefinitely, the walls
of which you would find you were fully enabled to
(17:17):
will yourself to walk through using your new found awareness
of the simulated basis of our universe. Our universe is
as real as it ever was. The only thing the
simulation argument can change is your perspective. One common objection
(17:37):
to the simulation argument kind of falls under this umbrella
of misunderstanding. It goes something like this, how could a
simulated banana or simulated apple, or a simulated cheeseburger possibly
provide nutrients to us. How could it sustain and nourish us,
alleviate our hunger, make us feel full. It's kind of
(17:59):
like the hard problem of consciousness, the puzzle of how
we get from neurons firing two subjective experience. Sure, you
can simulate a banana, but you're simulating a banana. It's
on a computer. And those of us who are looking
at the banana, no, we are looking at a banana
pictured on a computer. How can we account for the
experience of interacting with a banana, holding one, peeling it,
(18:22):
and needing it, of being allergic to it? How do
we get from a simulated banana to the experience of
a banana that is indistinguishable from reality. The answer is
rather simple. We're not simulating that banana or cheeseburger or whatever,
so it's not a simulation to us. It's made of
the same stuff that we are. It's no more fake
(18:44):
than you, You're no more real than it. We share
the same reality. The banana isn't a simulated banana in
our universe. It's just a banana, so it's subject to
the same rules that we are on the quantum level.
Those bananas are made up of energetic vibrations just like us.
They form particles that are held together by the fundamental forces,
(19:07):
just like us, and they form larger and more stable
structures that increase in size and complexity up to the
level of what we think of as a banana. And
when we eat that particular package of energetic vibration, it
interacts with our own energetic vibrations in such a way
that we experience taste, in the sensation of deliciousness, We
(19:28):
experience feeling full if we eat too many of them,
and our bodies break them down into smaller parts, and
we use those parts for energy, same as ever. Banana objections,
as we'll call them, are rooted in the idea that
we exist in some different reality from the banana. To
be clear, the simulation argument does not suppose that there's
(19:49):
some real version of you elsewhere. What we experience as
our universe, bananas and all, is a simulation to someone else.
To us, it is our full reality. So to answer
the banana objection, the banana can satiate and sustain us
because it's as real as we are. Other objections go
(20:10):
a little deeper. How could we possibly simulate the universe
on anything approaching a detail fine grained enough to be
indistinguishable from the original. There are a couple of answers
to this one. First, we can't. We don't have the
computing power to faithfully render the universe in that detail.
But in this objection we run into an unnecessary time constraint.
(20:34):
While we can't faithfully reproduce the universe, there's a pretty
good chance that our descendants will be able to. We
Those of us alive today are already on track to
doing this. For the last several years, humans have been
able to produce accurate simulations of quarks and glue ones
(20:54):
and the strong nuclear force that binds them together. These
models are not infinitely press size, but they're faithful enough
to reality that a simulated scientists living in the simulated
universe where the protons exist would have no way to
distinguish their unreality. They would just be protons that behaven
appear exactly like any other protons. The models that we
(21:16):
can produce today simulate an exceedingly tiny portion of our universe.
They're on the scale of femptometers, about a quadrillionth of
a meter. But the models show a proof of concept
that simulating the universe is possible, and they also show
that the human race already has a desire to simulate
the universe. Already at this early stage in modeling the universe,
(21:41):
we found that there are problems future humans will run
into when it comes to scaling up those models. It's
pretty obvious that modeling the entire universe would be a very,
very difficult undertaking, but difficulties can be worked out over time.
The problem comes when we reach impossibilities, and in two
thousand seventeen, a pair of researchers from Oxford University found
(22:05):
an impossibility that seems to show pretty conclusively that we
do not live in a simulated universe. The researchers found
that it would be physically impossible to simulate at least
one aspect of the universe, something called the quantum Hall effect,
which describes the way that electrons bounce between energy states
called a quantum leap in the presence of a magnetic
(22:28):
field and extremely low temperatures. It occupies a arcane a
corner of particle physics as there is, But the point
is the researchers found that to accurately simulate the effect.
The computational power required doubles each time you add a
new particle to the model, rather than increasing the power
needed by one unit, it grows exponentially, so within just
(22:51):
a few hundred particles, the operations per second required by
a computer to run a simulation of the quantum Hall
effect grows to a total beyond the number of atoms
there are in the universe, around ten to the eighty
second power. Not only is that number vastly more operations
than our computers can perform today, this also presents a
(23:11):
major technical issue. We need electrons to perform the computations
that create the simulated model. In the first place, electrons
are what we use to carry information in our computers,
and as we simulate more and more particles subject to
the quantum Hall effect, we will start to require a
substantial portion of the electrons that are part of those
(23:32):
ten to the eighty second power atoms in the universe.
And as you may know, those electrons are otherwise occupied
performing other functions like making up all the matter in
the universe, which means that eventually it becomes physically impossible
to simulate the quantum Hall effect and since we can
observe the quantum Hall effect in our own universe, we
(23:55):
can reason that our universe is not a simulation. But
this presents a failure of imagination, not an impossibility. Indeed,
the researchers who made the discovery noted that they're finding
only showed that we couldn't possibly be simulated within a
traditional computer, the type that we have available to us now,
the kind that uses electrons. They also pointed out that
(24:18):
we can probably safely presume that the humans of the
future would be simulating us on better computers than what
we're using today. I've said it before, and I'll say
it again. If it turns out that the humans of
the future don't have better computers than we have now,
it would be one of the more surprising things in
this entire series. And indeed, we are working on better
(24:40):
computers already, specifically, quantum computers, the type of computer that
processes information not gleaned from the movement of electrons, but
instead from the interaction of elementary particles, which exponentially opens
up a computer's capacity to perform calculations in a single
second and meet lee transfers, rendering the quantum Hall effect
(25:02):
and a simulation from the impossible column over to the
column marked possible. We've already talked about that to accurately
simulate a human or to require simulating a mind that
is capable of subjective experience. If the hard problem of
(25:23):
consciousness isn't actually a problem after all, then by creating
an accurate model of all of the bits that make
up a human mind, the neurons and synapses, the neurochemicals,
we can imagine that they will produce subjective experience when
they interact with the matter and energy that makes up
the simulated universe. This is exactly the same concept as
(25:44):
a post biological civilization that has shed their physical form
and migrated onto computers. If you found the concept of
a post biological future for humanity plausible, then you should
have no trouble accepting the simulation argument. The only difference
between the two is that, rather than us migrating onto
computers ourselves, we will have always been there already. As
(26:10):
Nick Bostrom points out in his paper on the simulation argument,
the human brain can process about ten to the seventeenth
power operations per second. If that's the amount of processing
power it takes to experience subjective reality. Then that's about
what we would expect it would take to simulate a
human mind, and to run a simulation of all of
human history would require that amount times around a hundred
(26:32):
and seven billion, since that's the number of humans who
have been born up to this point. Although we can't
imagine that as humans die, the processing power required to
produce their subjective experience would be freed up to power
the minds of new humans as they're born. So perhaps
all you'd ever need is enough processing power to simulate
all of the minds of the humans who are alive
(26:54):
at any given point, which means that to simulate a
global population of ten bills people, you'd need a computer
capable of processing ten to seven power operations per second.
That's not including the processing power it would take to
simulate the universe, which we could expect would expand the
requirements to mind boggling numbers. There are surely a lot
(27:18):
of corners future humans can cut to create a simulation.
They wouldn't need to simulate the entire universe obviously, just
what we can detect. Then it's possible they wouldn't need
to simulate the universe at all. Rather, they could just
create a program that stimulates our minds directly to experience
the universe. Even with the shortcuts, though, there's still an
(27:39):
enormous amount of ground to cover before we get to
the point where we have the processing power of the
magnitude required to simulate a human mind, let alone all
seven and a half billion of us. Take this for example.
Back in two thousand fourteen, the Fujitsu Corporation ran a
test on its Fujitsu K, at the time, the fourth
(28:00):
fastest supercomputer in the world, ranking that is no small feat.
In the test, the Fujitsu engineers wanted to see how
long it took their K computer to process one percent
of the amount of information that a human brain can
process in one second. The K managed to process that information,
but rather than the hundredth of a second it would
(28:22):
have taken a brain, it took the computer forty minutes
to complete the task. And it's worth noting the K
required twelve point seven megawatts of electricity throughout the forty
minutes it was running this test, about the same amount
of electricity needed to power the entire city of Sacramento, California.
(28:45):
Considering the gap between where we are now and what's
required to run a respectable ancestor simulation. It might start
to seem that such simulations are indeed impossible but easy.
There we're only running up against an unnecessary time constraint. Again,
given enough time, it seems quite possible to build a
computer that could run an accurate simulation. As Boston points out,
(29:08):
a single one of those planet sized computers could perform
ten to the forty second operations per second. To run
an ancestor simulation would require just the slightest fraction of
that computer's processing power, And we can imagine that in
a far enough expanse into the future, if our descendants
have one planet sized computer, they will eventually have a
(29:30):
great many of them. And let's not forget we humans
have a long standing tradition of advancing by leaps and
bounds in computing. If you wanted to find the top
of the line in supercomputers back in you needed to
look no further than the Cray Too. Recently, scientists producing
these simulations at NASA's Aims Research Center in Mountain View, California,
(29:53):
began using a new supercomputer, the Cray Too. It was
just a little smaller than the size of a BW
beatles stood on its end, but the CRY two could
perform two hundred and fifty million operations per second, an
astounding amount. In just over thirty years later, Apple released
a computer that could perform six hundred billion operations per second.
(30:17):
Apple called There's the iPhone tent, which, by the way,
is run by a machine learning algorithm capable of self improvement.
We already share the planet with quantum computers, the very
kind of computer that could simulate a universe to a
detail down to the quantum grain. The ones we share
the planet with today are exceedingly slow compared to our
(30:39):
current traditional computers, but they are in the earliest stages
of development. It was only in the eighties that physicists
began designing quantum computers that could work in theory, and
only in the nineties when the first quantum computers were built.
We can imagine that if we continue to follow this arc,
those primitive quantum computers of today will advance as well
(31:00):
as our traditional ones have. We haven't even lived with
traditional computers for a century, yet imagine what our quantum
computers will be capable of a century from now. There are,
to be sure, numerous other objections to the simulation argument,
but most, if not all, have difficulty overcoming that non
(31:21):
limitation of time. When we give our descendants the luxury
of time, those people whose future lives we will save
by taking on our existential risks become capable of anything
that we can imagine and more, including simulating us. Back
(31:53):
in two thousand twelve, a group of physicists from Bond
University in Germany announced that they had figured out a
way to test whether we live in a simulation. If
our simulation is anything like the femtometer scale simulated models
of protons that we can produce now, then that means
it isn't an infinitely accurate copy of the real universe
(32:14):
predicted by Einstein's theory of relativity. One of the things
that relativity says is that the universe is built from
a structure made up of the three spatial dimensions and
time spacetime. Spacetime is the fabric of our universe, and
it's contiguous, one giant universe sized unit. There's no gaps
(32:35):
in it anywhere, no matter how far we zoom in.
Remember when energy fields came up in the Physics episode
and I mentioned that these energy fields are everywhere at
every point in time and space. That's what that means.
There's no place where those fields aren't because there's no
gaps in spacetime. There's no place where space doesn't exist,
(32:55):
just as there aren't any gaps in time where there's
no time. It would take an enormous computer to render
an infinitely accurate version of our universe, potentially one the
exact size of the universe, and so we might guess
that the designers of our simulation would be forced to
make a version that's just slightly less accurate than the
(33:16):
real thing, which would mean that the fabric of the
simulated universe wouldn't be infinitely smooth. There would be gaps
in its structure gept between the individual units that it's
built from. If we could only zoom in close enough
to examine it, we would find the fabric of our
simulated universe was pixelated. But we can't zoom in that
(33:36):
close to observe space time directly. So the Bond University
researchers suggested a way that we could indirectly observe it
by watching how these smallest bits of matter in the
universe ultra high energy cosmic rays travel around space. Remember
Cosmic rays are high energy particles most likely spit out
from supernova that zip around our universe and collide with
(33:59):
other particles and may or may not produce microscopic black
holes when they do. They're one of the things that
we model in particle colliders. If spacetime is contiguous in
our universe, as we think it's supposed to be, then
cosmic race should travel in every possible direction. Since there's
no gaps in spacetime, there's nothing to direct their paths.
(34:21):
There aren't any channels in the fabric of space for
them to fall into and travel along. But if spacetime
is pixelated, the very fabric of the universe would form
an infinitismally fine grid. There would be tracks that the
tiniest cosmic rays might travel along. So, say the Bond researchers,
if we managed to observe that cosmic rays follow a
(34:43):
grid pattern, or that they show a preference for traveling
at angles rather than moving in every possible direction as
they should, we might have a fairly good indication that
what we're viewing as reality is actually a simulated version
of it. Interestingly, the idea that spacetime is grainy rather
(35:05):
than smooth is starting to gain traction among physicists. The
absolutely smallest discrete measure of distance in the universe is
called the plank length. Anything smaller than it, and physics
just falls to pieces. It's difficult to get across how
small a plank length is. It's about a hundred quintillion
times smaller than a proton, which itself is about a
(35:28):
hundred thousand times smaller than an atom. In fact, if
we could zoom into the plank scale so that a
plank length was the size of an atom, the atoms
in the universe would be the size of galaxies. It's
that small. Since spacetime is supposed to be infinitely smooth,
it's weird that there's the smallest point beyond which physics
(35:50):
is just meaningless. It's possible that our math is wrong,
or that when we managed to marry relativity and the
standard model into a theory of quantum gravity, we'll find
that physics picks up again beyond the plank scale. But
some quantum physicists are beginning to wonder if perhaps there
is some non infinite bottom size to the fabric of
(36:11):
space time, and that the plank length maybe it. Perhaps
we're beginning to uncover the frame of reference they Carte
longed for to distinguish between our reality and real reality.
But even as we chip away at the nature of
our own reality, if we find that we are indeed simulated,
(36:33):
we run into an issue. We still don't know anything
about the nature of real reality. All of our questions
just get kicked up to the universe that's simulating us.
There's a problem with that because we can't say for
sure that the universe we are simulated within isn't simulated itself.
So the problem can be kicked up indefinitely. Since in
(36:56):
the simulation argument, the people who are simulating us are
also humans themselves. They are, remember, the descendants of the
people we are simulations of. The same staggering probabilities against
us being real life humans applies to them as well.
Just by their membership and the human race and the
existence of ancestor simulations. The people who simulated us have
(37:18):
that same ninety nine point chance of being simulated just
because they are running the simulation of their ancestors doesn't
identify them as members of the real human race. The
simulation can run inside a simulation. When our descendants run
ancestor simulations. In the future, there will be simulated versions
(37:42):
of the universe existing within our own. But if we
are simulated, then the simulations we run will be simulations
within our simulation, and the same possibility applies to the
people running our simulation, to their universe may be simulated
as well. Perhaps it's an almost certainty that once a
(38:02):
simulated group of humans reaches some specific point, they will
begin to run ancestor simulations themselves, the ones that make
it through the great filter at least, and so ancestor
simulation spread in each direction from our own universe, simulations
within simulations within simulations, kind of like how it looks
(38:23):
inside an elevator that has mirrors on all sides. We
ner what's called an infinite regress problem. We can figure
out that our universe is simulated, but what does that
tell us about the universe the people who are simulating
us live in. Actually, this is familiar territory for physics.
One implication of cosmology is that our universe is one
(38:45):
bubble amid an infinite froth of bubbles, each their own universe.
And since our universe is an isolated system, no matter
or energy can leak out of it, and we presume
the same about those other universes, which means that we
can't glean any information about those other universes that we
rub up against in the infinite froth, and we likely
(39:06):
never will. And yet you would be hard pressed to
find a physicist who suggested throwing in the towel on
investigating our own universe when cosmology reached this conclusion. The
same holds true for the idea that we're simulated. Not
being able to know anything about the universe that simulated
us doesn't mean there's nothing to be gained from investigating
(39:26):
our own It's worth taking a step back here to
look at the implications of what it means. When we
start talking about who is simulating us, we're talking about
someone who in some respect created us. Perhaps one of
the touchiest aspects of the simulation argument is that it
(39:49):
suggests a sort of techno creationism. Although it does fall
under the umbrella of creationism, accepting the idea that we
may be simulated does to require any form of supernatural
or religious belief. If we're simulated, we can identify who
created us. It's the descendants of the people who were
simulations of future humans created us. But who exactly would
(40:16):
the software developer who wrote the code that produced us
qualify as our creator? Would the person who directly booted
up the simulation we live in? Or what if our
simulation is a scientific model, does our creator where a
lab coat and our simulation is a digital Petrie dish
of sorts? Or is our creator a far future teenager
who created us for a class project. I should remind
(40:39):
you that it's important to remember the simulation argument doesn't
prove anything about our perceived reality. It gives no evidence
one way or the other. It's a framework for viewing
the nature of our universe, which means that, like any
other argument, you can take or leave it. It is
a matter of personal belief whether the idea of being
(41:00):
created from a program written by future humans is more
or less comforting than the idea that a supernatural deity
created us, or that the world grew on the back
of a turtle, or that we're the fluke result of
an incomprehensible series of cause and effect stretching back nearly
fourteen billion years. If we did one day discover that
(41:27):
we are stimulated humans living in a simulated universe, that
would we can imagine have enormous and sweeping effects on us.
Even if you already believe that we were created, how
would you feel to learn that your creator is only human,
with all the same faults and flaws that you have.
And if you don't believe today that we were created,
(41:48):
that we are merely the result of some physical processes
that can turn dead matter into cellular life, and that
we were shaped by the forces of evolution, learning indisputably
that we were indeed created, it would be a tough
pill to swallow. But would those enormous and sweeping effects
be all bad and would they be permanent? This is
(42:09):
Ander Sandberg again. I do think that if we realize
that the universe fundamentally work differently from how we used
to believe it was, we would be shaking, But we
would also get over it, just like we have gotten
over and the hellyo some traces, and the evolution and
the Freud and so on. Most people are way more
resilient to upsets of the fundamental structure of the world
(42:32):
than we give them credit for. We humans are really
good at absorbing terrible news, whether it's a cancer diagnosis
or learning that we're facing existential risks. There's a chance
we will survive, there's a chance we won't. But do
we simply lay down and die. Humans tend not to
do that. Instead, we carry on. We seek out connections
(42:55):
and support from other humans who are also carrying on,
and we might expect the same would hold true should
we learn that we're simulated. Even despite finding that we
are the result of code, nothing about the meaning of
our lives should necessarily be lost. Well where do we
get a meaning in our own lives? And I would
argue that we're actually creating it ourselves. Even if the
(43:17):
universe was a class project to demonstrate the sociological and thing,
that doesn't mean it's the meaning of our life to
not behave socially, It's not like our loves and the
emotions that will become meaningless, and we might still have
a meaning that's very different from the reason the simulation exists.
It will probably be tough to keep in mind, especially
(43:40):
for the people who are alive at the time, when
we learn our universe is simulated, But at bottom, really
nothing much changes for us. Our understanding of our nature
only deepens in a very direct way. Learning we are
stimulated would focus our investigation into our universe it would
certainly explain a lot. The Bond University researchers who came
(44:02):
up with the cosmic ray grid test suggested that perhaps
we'd find the reason the Higgs field is in in
its lowest energy state is because of a rounding error
and the computations that produce our simulation. Maybe the measurement
problem that quantum particles seem to exist in all of
the possible forms they can take at the same time
is a shortcut to make rendering the universe more efficient.
(44:24):
The fine tuning we see would certainly be explained. The
parameters of the forces and energy that make up our
universe would have to be the way they are because
we could not be produced under other conditions. But that's correct,
whether we're simulated or not. Finding we're simulated would identify
computer code as the basis of those governing principles that
(44:45):
we already take to be true, like a biogenesis and evolution.
We wouldn't expect the understanding the truth of that would
change how those processes function, but perhaps we could learn
how to change them ourselves. Maybe learning we are simulated
would free us to abandon the notion of fate and
take control of how the universe operates to shape it
(45:07):
more to our liking. After all, isn't that what the
study of physics is? Taking control of the universe would
also focus the moral responsibility that we have in making
our universe a better place. Having the ability to reduce
suffering wherever we found it, we would be morally obligated
to do so. But then that's the case already as well.
(45:29):
The difference is that we couldn't let ourselves off the
hook by chalking that suffering up to inevitability like we
do now. This is Seth show Stack, the senior astronomer
at City who you met in the first episode on
the Family Paradox. The question that I asked the Professor
Bostrom that I think was most relevant to me was, Okay,
(45:50):
if this is all a simulation, this conversation we're having,
it's not real. It's just code. Right. If it's all
just code, do I have to live a moral existence
or kind just have fun? He thought it was better
to live a moral existence, uh, you know, And I
thought that that was kind of a testimony to how
he really felt on the matter, that you know, you
better act like it's real. The beautiful thing about the
(46:21):
simulation argument is that it doesn't change anything, only our perspective.
The world continues to operate in exactly the same way
as before. It's just that some of the mystery is
brushed away. Nothing about humanity changes, nor does anything about
where we find ourselves in the position. Where as Toby
Ord paraphrase Carl Sagan, we humans have grown powerful before
(46:44):
we've grown wise, the existential risks we face remain as
real as ever, and if we did find we are simulated,
our existential risks would only sharpen, since we would learn
of a new one that our simulation could be turned off,
which I opposed, would qualify as a natural existential risk.
(47:04):
Whether we arrived at this point where we have grown
powerful before we've grown wise, by total fluke or by predetermination,
whether we are members of the real human race or simulated,
we are in the same unique and precarious position that
we were in before. We are still entering the mouth
of the great filter, and we are still at risk
(47:26):
of our existence coming to a permanent end. If we
fail to come together as a world and learn to
use science to help us grow wiser faster, we will
surely not survive. We will accidentally wipe ourselves out with
our unwise power, and that will be the end of
humans in our universe. Does it matter whether that universe
arose from a big bang or from a system boot,
(47:49):
Whether we are real or simulated, our reality is the same.
The End Of The World with Josh Clark News
Host
Josh Clark
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com