A black hole is a mirror.
And the way it's a mirror is if light,
a photon bounces off your face towards the black hole
and it goes straight to the black hole,
just falls in, you never see it again.
But if it just misses the black hole,
it'll swing around the back and come back to you.
And you see yourself from the photon
that went around the back of the black hole.
But not only can that happen, the black hole,
the photon can swing around twice and come back.
So you actually see an infinite number of copies
of yourself.
The following is a conversation with Andrew Strominger,
theoretical physicist at Harvard,
whose research seeks to shed light
on the unification of fundamental laws of nature,
the origin of the universe, and the quantum structure
of black holes and event horizons.
This is the Lex Friedman Podcast.
To support it, please check out our sponsors
in the description.
And now, dear friends, here's Andrew Strominger.
You are part of the Harvard Black Hole Initiative,
which has theoretical physicists, experimentalists,
and even philosophers.
So let me ask the big question, what is a black hole
from a theoretical, from an experimental,
maybe even from a philosophical perspective?
So a black hole is defined, theoretically,
as a region of space-time from which light can never escape.
Therefore, it's black.
Now, that's just the starting point.
Many weird things follow from that basic definition,
but that is the basic definition.
What is light that can't escape from a black hole?
Well, light is the stuff that comes out of the sun,
that's stuff that goes into your eyes.
Light is one of the stuff that disappears
when the lights go off.
This is stuff that appears when the lights come on.
Of course, I could give you a mathematical definition,
or a physical mathematical definition,
but I think it's something that we all understand
very intuitively what is light.
Black holes, on the other hand,
we don't understand intuitively, they're very weird.
And one of the questions is about black holes,
which I think you were alluding to,
is why doesn't light get out,
or how is it that there can be a region of space-time
from which light can't escape?
It definitely happens.
We've seen those regions.
We have spectacular pictures,
especially in the last several years of those regions.
They're there.
In fact, they're up in the sky,
thousands or millions of them.
We don't yet know how many.
But the proper explanation
of why light doesn't escape from a black hole
is still a matter of some debate.
And one explanation,
which perhaps Einstein might have given,
is that light carries energy.
You know it carries energy because we have photocells
and we can take the light from the sun and collect it,
turn it into electricity.
So there's energy in light.
And anything that carries energy
is subject to a gravitational pull.
Gravity will pull at anything with energy.
Now it turns out that the gravitational pull
exerted by an object is proportional to its mass.
And so if you get enough mass in a small enough region,
you can prevent light from escaping.
And let me flesh that out a little more.
If you're on the Earth
and you're on a rocket ship
leaving the surface of the Earth,
and if we ignore the friction from the air,
if your rocket accelerates up to 11 kilometers per second,
that's escape velocity.
And if there were no friction,
you could just continue forever to the next galaxy.
On the Moon, which has less mass,
it's only seven kilometers per second.
But going in the other direction,
if you have enough mass in one place,
the escape velocity can become the speed of light.
If you shine light straight up away from the Earth,
it doesn't have too much trouble.
It's going way above the escape velocity.
But if you have enough mass there,
even light can't escape the escape velocity.
And according to Einstein's theory of relativity,
there is an absolute speed limit in the universe,
the speed of light, and nothing makes any sense.
Nothing could be self-consistent
if there were objects that could exceed light speed.
And so in these very, very massive regions of space-time,
even light cannot escape.
And the interesting thing is Einstein himself
didn't think that these objects,
we call the black holes, could exist.
But let me actually linger on this.
Yeah, that's incredibly interesting.
There's a lot of interesting things here.
First, the speed limit.
How wild is it to you, if you put yourself in the mind
in the time of Einstein before him,
to come up with a speed limit of,
that there is a speed limit.
And that speed limit is the speed of light.
How difficult of an idea is that?
You said from a mathematical physics perspective,
everything just kind of falls into place.
But he wasn't, perhaps, maybe initially
had the luxury to think mathematically.
He had to come up with it intuitively, yes?
So how come intuitive is this notion to you?
Is it still crazy?
No, no, so it's a very funny thing in physics.
The best discoveries seem completely obvious in retrospect.
Even my own discoveries,
which of course are far lesser than Einstein's,
but many of my papers, many of my collaborators
get all confused, will try to understand something.
We say, we've got to solve this problem.
We'll get all confused.
Finally, we'll solve it.
We'll get it all together.
And then all of a sudden, everything will fall into place.
We'll explain it.
And then we'll look back at our discussions
for the proceedings of months
and literally be unable to reconstruct
how confused we were
and how we could ever have thought of it any other way.
So not only can I not fathom
how confused Einstein was
before he, when he started thinking about the issues,
I can't even reconstruct my own confusion
from two weeks ago.
So the really beautiful ideas in physics
have this very hard to get yourself back into the mindset.
Of course, Einstein was confused about many, many things.
It doesn't matter if you're a physicist.
It's not how many things you got wrong.
It's not the ratio of how many you got wrong
to how many you got right.
It's the number that you got right.
So Einstein didn't believe black holes existed
even though he predicted them.
And I went and I read that paper, which he wrote.
Einstein wrote down his field equations in 1915,
and Schwarzschild solved them
and discovered the black hole solution
three or four months later in very early 1916.
And 25 years later, Einstein wrote a paper.
So with 25 years to think about what this solution means,
wrote a paper in which he said
that black holes didn't exist.
And I'm like, well, you know,
if one of my students in my general relativity course
wrote this, you know, I wouldn't pass them.
You get a C minus, oh, you wouldn't pass them.
Okay, all right.
You get a C minus, okay.
Same thing with gravity waves.
He didn't believe.
Oh, he didn't believe in gravitational waves either?
He went back and forth, but he wrote a paper,
and I think 34, saying that gravity waves didn't exist
because people were very confused
about what a coordinate transformation is.
And in fact, this confusion
about what a coordinate transformation is has persisted.
And we actually think we were on the edge
of solving it a hundred years later.
A hundred years later.
What is coordinate transformation
as it was a hundred years ago to today?
Let's imagine I want to draw a map
with pictures of all the states and the mountains,
and then I wanna draw the weather forecast,
what the temperatures are gonna be all over the country.
And I do that using one set of weather stations,
and I number the weather stations.
And you have some other set of weather stations,
and you do the same thing.
So the coordinates are the locations
of the weather stations.
They're how we describe where the things are.
At the end of the day, we should draw the same map.
That is coordinate invariance.
And if we're telling somebody,
we're gonna tell somebody at a real physical operation,
we want you to stay as dry as possible
on your drive from here to California,
we should give them exactly the same route.
No matter which weather stations we use or how we,
it's a very trivial,
it's the labeling of points is an artifact
and not in the real physics.
So it turns out that that's almost true,
but not quite.
There's some subtleties to it.
The statement that you should always have the same,
give this the same kind of trajectory,
the same kind of instructions,
no matter the weather stations.
Yeah, there's some very delicate subtleties to that,
which began to be noticed in the 50s.
It's mostly true, but when you have a space-time with edges,
it gets very tricky how you label the edges.
And space-time in terms of space or in terms of time,
in terms of everything, just space?
Either one, space or time.
That gets very tricky.
And Einstein didn't have it right,
didn't have it right,
and in fact, he had an earlier version
of general relativity in 1914,
which he was very excited about,
which was wrong.
Gave, it wasn't fully coordinate invariant,
it was only partially coordinate invariant.
It was wrong.
It gave the wrong answer for bending light to the sun
by a factor of two.
There was an expedition sent out to measure it
during World War I.
They were captured before they could measure it,
and that gave Einstein four more years
to clean his act up, by which time he'd gotten it right.
So it's a very tricky business,
but once it's all laid out,
it's clear.
Then why do you think Einstein
didn't believe his own equations
and didn't think that black holes are real?
Why was that such a difficult idea for him?
Well, something very interesting happens
in Schwarzschild's solution of the Einstein equation.
I think his reasoning was ultimately wrong,
but let me explain to you what it was.
At the center of the black hole, behind the horizon,
in a region that nobody can see and live to tell about it,
at the center of the black hole, there's a singularity,
and if you pass the horizon, you go into the singularity,
you get crushed, and that's the end of everything.
Now, the word singularity means that,
it just means that Einstein's equations break down.
They become infinite, you write them down,
you put them on the computer.
When the computer hits that singularity, it crashes.
Everything becomes infinite, there's two.
So the equations are just no good there.
Now, that's actually not a bad thing.
It's a really good thing, and let me explain why.
So it's an odd thing that Maxwell's theory, and Newton's theory,
never exhibit this phenomena.
You write them down, you can solve them exactly.
They're really Newton's theory of gravity,
they're really very simple theories.
You can solve them, well, you can't solve
the three-body problem, but you can certainly
solve a lot of things about them.
Nevertheless, there was never any reason,
even though Maxwell and Newton perhaps fell for this trap,
there were never any reason to think
that these equations were exact, and there's no equation.
Well, there's some equations that we've written down
that we still think are exact.
Some people still think are exact.
My view is that there's no exact equation.
Everything is an approximation.
Everything is an approximation.
And you're trying to get as close as possible.
So you're saying objective truth doesn't exist in this world?
The internet's gonna be very mad at you.
We could discuss that, but that's a different thing.
We wouldn't say Newton's theory was wrong.
It had very, very small corrections,
incredibly small corrections.
It's actually a puzzle why they're so small.
So if you watch the precession of Mercury's perihelion,
this was the first indication of something going wrong.
According to Newton's theory,
Mercury has an elliptical orbit.
The long part of it moves around
as other planets come by and perturb it and so on.
And so this was measured by Le Verrier in 1859,
and he compared theory and experiment,
and he found out that the perihelion process
moves around the Sun once every 233 centuries
instead of every 231 centuries.
Okay, now this is the wonderful thing about science.
Why was this guy?
I mean, you don't get any idea how much work this is.
But of course he made one of the greatest discoveries
in the history of science without even knowing
what good it was gonna be.
So that's how small, that was the first sign
that there was something wrong with Newton.
Now, so the corrections to Newton's law
are very, very small, but they're definitely there.
The corrections to electromagnetism,
they're mostly, the ones that we see
are mostly coming from quantum effects.
So the corrections there for Maxwell's equations
is when you get super tiny, and then the corrections
for Newton's laws of gravity is when you get super big.
That's when you require corrections.
That's true, but I would phrase it as saying
when it's super accurate.
If you look at the Bohr atom, Maxwell electromagnetism
is not a very good approximation to the force
between the proton and the electron.
The quantum mechanics, if you didn't have quantum mechanics,
the electron would spiral into the proton
and the atom would collapse.
So that's a huge correction there.
So every theory gets corrected as we learn more.
There'd just be no reason to suppose
that it should be otherwise.
Well, how does this relate to the singularity,
why the singularity becomes uncomfortable?
So when you hit the singularity,
you know that you need some improvement
to Einstein's theory of gravity,
and that improvement, we understand what kind of things
that improvement should involve.
It should involve quantum mechanics.
Quantum effects become important there.
It's a small thing.
And we don't understand exactly what the theory is,
but we know there's no reason to think,
Einstein's theory was invented to describe
weakly curved things, the solar system and so on.
It's incredibly robust that we now see
that it works very well near the horizons
around black holes and so on.
So it's a good thing that the theory drives itself,
that it predicts its own demise.
Newton's gravity had its demise.
There were regimes in which it wasn't valid.
Maxwell's electromagnetism had its demise.
There was regimes in which quantum effects
greatly modified the equations.
But general relativity all on its own
found a system which originally was fine
would perversely wander off into a configuration
in which Einstein's equations no longer applied.
So to you, the edges of the theory are wonderful.
The failures of the theory.
The edges are wonderful because that keeps us in business.
So one of the things you said I think in your TED Talk
that the fact that quantum mechanics and relativity
don't describe everything and then they clash is wonderful.
I forget the adjective you used,
but it was something like this.
So why is that?
Why is that interesting?
Do you in that same way that there's contradictions
that create discovery?
There's no question in my mind,
of course many people would disagree with me,
that now is the most wonderful time to be a physicist.
So people look back at,
it's a classical thing to say among physicists,
I wish it were 1920.
Quantum mechanics had been just understood.
There was the periodic table.
But in fact, that was such a rich thing.
So a lot of exciting stuff happened around 1920.
It took a whole century to sort out
the new insights that we got.
Especially adding some experimental stuff
into the bunch, actually making observations
and integrating all the data.
Adding experimental things.
All the computers also help with visualizations
and all that kind of stuff.
Yeah, yeah, yeah.
It was a whole sort of wonderful century.
I mean, the seed of general relativity
was the incompatibility of Maxwell's theory
of the electromagnetic field
with Newton's laws of gravity.
They were incompatible because if you look
at Maxwell's theory, there's a contradiction
if anything goes faster than the speed of light.
But Newton's theory of gravity,
the gravitational field, the gravitational force
is instantaneously transmitted across the entire universe.
So you could, if you had a friend in another galaxy
with a very sensitive measuring device
that could measure the gravitational field,
they could just take this cup of coffee
and move it up and down in Morse code
and they could get the message instantaneously
over another galaxy.
That leads to all kinds of contradictions.
It's not self-consistent.
It was exactly in resolving those contradictions
that Einstein came up with
the general theory of relativity.
And it's fascinating how this contradiction,
which seems like maybe it's kind of technical thing,
led to a whole new vision of the universe.
Now, let's not get fooled because lots of contradictions
are technical things.
We haven't set up, we run into other kinds
of contradictions that are technical
and they don't seem to, we understood something wrong,
we made a mistake, we set up equations in the wrong way,
we didn't translate the formalisms.
As opposed to revealing some deep mystery
that's yet to be uncovered.
Yeah, yeah.
And so we're never very sure
which are the really important ones.
But to you, the difference between quantum mechanics
and general relativity, the tension, the contradiction there
seems to hint at some deeper, deeper thing
that's going to be discovered in the century.
Yes, because that one has been understood since the 50s.
Pauli was the first person to notice it
and Hawking in the early 70s gave it
a really much more visceral form.
And people have been hurling themselves at it,
trying to reduce it to some technicality,
but nobody has succeeded.
And the efforts to understand it have led
to all kinds of interesting relations
between quantum systems and applications
to other fields and so on.
Well, let's actually jump around.
So we'll return to black holes.
I have a million questions there,
but let's go into this unification,
the battle against the contradictions
and the tensions between the theories of physics.
What is quantum gravity?
Maybe what is the standard model of physics?
What is quantum mechanics?
What is general relativity?
What's quantum gravity?
What are all the different unification efforts?
Okay, so.
Again, five questions.
Yeah, it's a theory that describes everything
with astonishing accuracy.
It's the most accurate theory
in the history of human thought.
Theory and experiment have been successfully compared
to 16 decimal place.
We have that stenciled on the door where I work.
You know, it's an amazing feat of the human mind.
It describes the electromagnetic interaction,
unifies the electromagnetic interaction
with the so-called weak interaction,
which you need some good tools
to even view the weak interaction.
And then there's the strong interaction,
which binds the quarks into protons.
And the forces between them are mediated
by something called Yang-Mills theory,
which is a beautiful mathematical generalization
of electromagnetism in which the analogs
of the photons themselves carry charge.
And so this, the final piece of this,
of the standard model, everything in the standard model
has been observed.
Its properties have been measured.
The final particle to be observed was the Higgs particle,
observed like over a decade ago.
But Higgs is already a decade ago.
I think it is, yeah.
Wow, time flies.
But you better check me on that, yeah.
That's true, but so much fun has been happening.
So much fun has been happening.
And so that's all pretty well understood.
There's some things that might or might not
around the edges of that, dark matter, neutrino masses,
some sort of fine points or things
we haven't quite measured perfectly and so on.
But it's largely a very complete theory.
And we don't expect anything very new conceptually
in the completion of that.
Anything contradictory by a mu,
because can't you? Anything contradictory, yeah.
I'll have some wild questions for you on that front.
But yeah, anything that, yeah,
because there's no gaps, it's so accurate, it's so precise,
and it's predictions, it's hard to imagine.
Yeah, yeah, yeah.
And it was all based on something called,
let me not explain what it is,
let me just throw out the buzzword,
re-normalizable quantum field theory.
They all fall in the category
of re-normalizable quantum field theory.
I'm gonna throw that at a bar later to impress the girls.
Good luck.
Thank you.
Thank you.
Oh.
All right, so.
They all fall under that rubric.
Gravity will not put that suit on.
So the force of gravity cannot be tamed
by the same re-normalizable quantum field theory
to which all the other forces so eagerly submitted.
What is the effort of quantum gravity?
What are the different efforts
to have these two dance together effectively,
to try to unify the standard model and general relativity,
any kind of model of gravity?
Sort of the one fully consistent model that we have
that reconciles, that sort of tames gravity
and reconciles it with quantum mechanics,
is string theory and its cousins.
And we don't know what, or if in any sense,
string theory describes the world, the physical world,
but we do know that it is a consistent reconciliation
of quantum mechanics and general relativity,
and moreover one which is able to incorporate
particles and forces like the ones we see around us.
So it hasn't been ruled out as an actual
sort of unified theory of nature,
but there also isn't a, in my view,
some people would disagree with me,
but there isn't a reasonable possibility
that we would be able to do an experiment
in the foreseeable future which would be
sort of a yes or no to string theory.
Okay, so you've been there from the early days
of string theory, you've seen its developments.
What are some interesting developments?
What do you see as also the future of string theory?
And what is string theory?
Well, the basic idea which emerged in the early 70s
was that if you,
you take the notion of a particle
and you literally replace it by a little loop of string,
the strings are sort of softer than particles.
What do you mean by softer?
Well, you know, if you hit a particle,
if there were a particle on this table,
a big one, and you hit it, you might bruise yourself.
Sure.
But if there was a string on the table,
you would probably just push it around.
And the source of the infinities in quantum field theory
is that when particles hit each other,
it's a little bit of a jarring effect.
And I've never described it this way before,
but it's actually scientifically accurate.
But if you throw strings at each other,
it's a little more friendly.
One thing I can't explain is how wonderfully precise
all the mathematics is
that goes into describing string theory.
We don't just wave our hands and throw strings around
and throw strings around.
And there's some very compelling mathematical equations
that describe it.
Now, what was realized in the early 70s
is that if you replace particles by strings,
these infinities go away
and you get a consistent theory of gravity
without the infinities.
And that may sound a little trivial,
but at that point it had already been 15 years
that people had been searching around
for any kind of theory that could do this.
And it was actually found kind of by accident.
And there are a lot of accidental discoveries
in this subject.
Now, at the same time, it was believed then
that string theory was an interesting sort of toy model
for putting quantum mechanics
and general relativity together on paper,
but that it couldn't describe
some of the very idiosyncratic phenomena
that pertain to our own universe,
in particular, the form of so-called parity violation.
Our world is-
Ooh, another term for the bar later tonight.
Yeah, yeah.
Parity violation, okay.
So if you go to the bar and-
I already got the renormalizable quantum field theory.
And you look in the mirror across the bar,
the universe that you see in the mirror is not identical.
You would be able to tell if you show
the lady in the bar a photograph
that shows both the mirror and you.
There's a difference.
If she's smart enough,
she'll be able to tell which one is the real world
and which one is you.
Now, she would have to do some very precise measurements.
Yes.
Yes.
And if the photograph was too grainy,
it might not be possible.
But in principle, it's possible.
Why is this interesting?
Does this mean that there is some
not perfect determinism, or what does that mean?
There's some uncertainty?
No, it's a very interesting feature of the real world
that it isn't parity invariant.
In string theory, it was thought could not tolerate that.
And then it was learned in the mid-'80s
that not only could it tolerate that,
but if you did things in the right way,
you could construct a world involving strings
that reconciled quantum mechanics and general relativity,
which looked more or less like the world that we live in.
And now, that isn't to say
that string theory predicted our world.
It just meant that it was consistent,
that the hypothesis that string theory describes our world
can't be ruled out from the get-go.
And it is also the only proposal for a complete theory
that would describe our world.
Still, nobody will believe it
until there's some kind of direct experiment.
And I don't even believe it myself.
Sure, which is a good place to be mentally
as a physicist, right?
Always, I mean, Einstein didn't believe his own equations,
right, with the black hole.
Okay.
Well, that one, he was wrong about that.
But he was wrong about that.
But you might be wrong too, right?
So do you think string theory is dead
if you were to bet all your money
on the future of string theory?
I think it's a logical error
to think that string theory is either right or wrong
or dead or alive.
What it is is a stepping stone.
And an analogy I like to draw is Yang-Mills theory,
which I mentioned a few minutes ago
in the context of standard model.
Yang-Mills theory was discovered
by Yang and Mills in the 50s.
And they thought that the symmetry
of Yang and Mills theory described the relationship
between the proton and the neutron.
That's why they invented it.
That turned out to be completely wrong.
It does, however, describe everything else
in the standard model.
And it had a kind of inevitability.
They had some of the right pieces,
but not the other ones.
They didn't have it quite in the right context.
And it had an inevitability to it,
and it eventually sort of found its place.
And it's also true of Einstein's theory
of general relativity.
He had the wrong version of it in 1914,
and he was missing some pieces.
And you wouldn't say that his early version
was right or wrong.
He'd understood the equivalence principle.
He'd understood space-time curvature.
He just didn't have everything.
I mean, technically, you would have to say it was wrong.
And technically, you would have to say
Yang and Mills were wrong.
And I guess in that sense,
I would believe just odds are
we always keep finding new wrinkles.
Odds are we're gonna find new wrinkles in string theory,
and technically, what we call string theory now
isn't quite right, but.
We're always going to be wrong,
but hopefully a little bit less wrong every time.
Exactly.
And I would bet the farm, as they say.
Do you have a farm?
I say that much more seriously,
because not only do I have a farm,
but we just renovated it.
So before I renovated,
so before I renovated, Better Get the Farm,
my wife and I spent five years renovating it.
Before I.
You were much looser with that statement,
but now it really means something.
Now it really means something?
And I would bet the farm on the,
on the guess that 100 years from now,
string theory will be viewed as a stepping stone
towards a greater understanding of nature.
And it would, I mean,
another thing that I didn't mention about string theory
is, of course, we knew that it solved
the infinities problem,
and then we later learned that it also solved
Hawking's puzzle about what's inside of a black hole.
And you put in one assumption,
you get five things out,
somehow you're doing something right.
Probably not everything, but you're,
there's some good signposts,
and there have been a lot of good signposts like that.
It is also a mathematical toolkit,
and you've used it, you've used it with Kamran Wafa.
Maybe we can sneak our way back
from string theory into black holes.
What was the idea that you and Kamran Wafa developed
with the holographic principle in string theory?
What were you able to discover
through string theory about black holes?
Or that connects us back to the reality of black holes?
Yeah, so that is a very interesting story.
I was interested in black holes
before I was interested in string theory.
I was sort of a reluctant string theorist in the beginning.
I thought I had to learn it
because people were talking about it,
but once I studied it, I grew to love it.
First, I did it in a sort of dutiful way.
These people say they've claimed quantum gravity.
I ought to read their papers, at least.
And then the more I read them,
the more interested I got, and I begin to see.
They phrased it in a very clumsy way.
The description of string theory was very clumsy.
Mathematically clumsy, or just the interpretation?
Mathematically clumsy, yeah.
It was all correct, but mathematically clumsy,
but it often happens that
in all kinds of branches of physics
that people start working on it really hard
and they sort of dream about it and live it and breathe it,
and they begin to see interrelationships,
and they see a beauty that is really there.
They're not deceived.
They're really seeing something that exists,
but if you just kind of look at it,
you can't grasp it all in the beginning.
Our understanding of string theory in 1985
was almost all about
weakly coupled waves of strings colliding and so on.
We didn't know how to describe a big thing,
like a black hole in string theory.
Of course, we could show that strings in theory
in some limit reproduced Einstein's theory
of general relativity and corrected it,
but we couldn't do any better with black holes
than before my work with Gorman,
we couldn't do any better
than Einstein and Schwarzschild had done.
Now, one of the puzzles,
if you look at Hawking's headstone
and also Boltzmann's headstone and you put them together,
you get a formula for their really central equations
in 20th century physics.
I don't think there are many equations
that made it to headstones.
And they're really central equations,
and you put them together and you get a formula
for the number of gigabytes in a black hole.
Now, in Schwarzschild's description,
the black hole is literally a hole in space
and there's no place to store the gigabytes.
And it's not too hard to,
and this really was Wheeler and Bekenstein and,
Wheeler, Bekenstein and Hawking,
to come to the conclusion that if there isn't a sense
in which a black hole can store some large number
of gigabytes, that quantum mechanics
and gravity can't be consistent.
We've got to go there a little bit.
So how is it possible, when we say gigabytes,
there's some information.
So black holes can store information.
How is this thing that sucks up all light
and is supposed to basically be super homogeneous
and boring, how is that actually able to store information?
Where does it store information?
On the inside, on the surface, where?
Where's, and what's information?
I'm liking this ask five questions
to see which one you actually answer.
I'm gonna ask you a question.
I should try to memorize them and answer each one in order,
just answer them.
No, I don't know.
I don't know what I'm doing.
I'm desperately, desperately trying to figure it out
as we go along here.
So Einstein's black holes and Schwarzschild's black hole,
they can't store information.
This stuff goes in there and it just keeps flying
and it goes to the singularity and it's gone.
However, Einstein's theory is not exact.
It has corrections and string theory tells you
what those corrections are.
And so you should be able to find some way
of some alternate way of describing the black hole
that enables you to understand
where the gigabytes are stored.
So what Hawking and Beckenstein really did
was they showed that physics is inconsistent
unless a black hole can store a number of gigabytes
proportional to its area divided by four times
Newton's constant times Planck's constant.
And that's another wild idea.
You said area, not volume.
Exactly, and that's the holographic principle.
The universe is so weird.
That's the holographic principle.
That's called the holographic principle,
that it's the area.
We're just jumping around.
What is the holographic principle?
What does that mean?
Is there some kind of weird projection going on?
What the heck?
Well, it was just before I came here
writing an introduction to a paper
and the first sentence was the as yet imprecisely defined
holographic principle, blah, blah, blah, blah, blah.
So nobody knows exactly what it is,
but roughly speaking, it says just what we were alluding to,
that really all the information
that is in some volume of space-time
can be stored on the boundary of that region.
So this is not just about black holes,
it's about any area space?
Any area space.
However, we've made sense of the holographic principle
for black holes.
We've made sense of the holographic principle
for something which could be called the anti-de Sitter space
which could be thought of as a giant,
as a black hole turned into a whole universe.
And we don't really understand how to talk
about the holographic principle for either flat space
which we appear to live in,
or asymptotically de Sitter space
which the astronomers tell us we actually live in
as the universe continues to expand.
So it's one of the huge problems
in physics is to apply or even formulate
the holographic principle for more realistic,
well, black holes are realistic, we see them,
but yeah, in more general context.
So you have a more general statement
of the holographic principle.
What's the difference between flat space
and asymptotic de Sitter space?
So flat space is just an approximation
of the world we live in.
So de Sitter space, asymptotic,
I wonder what that even means,
meaning asymptotic over what?
Okay, so for thousands of years,
until the last half of the 20th,
well, sorry, until the 20th century,
we thought space-time was flat.
Can you elaborate on flat?
What do we mean by flat?
Well, like the surface of this table is flat.
Let me just give an intuitive explanation.
Surface of the table is flat,
but the surface of a basketball is curved.
So the universe itself could be flat,
like the surface of a table,
or it could be curved like a basketball,
which actually has a positive curvature.
And then there's another kind of curvature
called the negative curvature.
And curvature can be even weirder
because that kind of curvature I've just described
is the curvature of space,
but Einstein taught us that we really live
in a space-time continuum,
so we can have curvature in a way
that mixes up space and time.
And that's kind of hard to visualize.
Because you have to step what,
a couple of dimensions up?
So it's hard to...
You have to step a couple,
but even if you have flat space
and it's expanding in time,
we could imagine we're sitting here in this room,
good approximation is flat,
but imagine we suddenly start getting
further and further apart.
Then space is flat,
but it's expanding,
which means that space-time is curved.
Ultimately, it's about space-time.
Okay, so what's de Sitter and anti-de Sitter space?
The three simplest space-times are flat space-time,
which we call Minkowski space-time,
and negatively curved space-time,
anti-de Sitter space,
and positively curved space-time, de Sitter space.
And so astronomers think that on large scales,
even though for thousands of years we hadn't noticed it,
beginning with Hubble,
we started to notice that space-time was curved.
Space is expanding in time,
means that space-time is curved.
And the nature of this curvature
is affected by the matter in it,
because matter itself causes the curvature of space-time.
But as it expands,
the matter gets more and more diluted.
And one might ask, when it's all diluted away,
is space-time still curved?
And astronomers believe they've done
precise enough measurements to determine this,
and they believe that the answer is yes.
The universe is now expanding.
Eventually all the matter in it will be expanded away,
but it will continue to expand because,
well, they would call it the dark energy,
Einstein would call it a cosmological constant.
In any case, in the far future,
matter will be expanded away
and will be left with empty de Sitter space.
Okay, so there's this cosmological,
Einstein's cosmological constant
that now hides this thing that we don't understand
called dark energy.
What's dark energy?
What's your best guess at what this thing is?
Why do we think it's there?
It's because it comes from the astronomers.
Dark energy is synonymous
with positive cosmological constant.
And we think it's there
because the astronomers have told us it's there.
And they know what they're doing.
And it's a really, really hard measurement,
but they really know what they're doing.
And we have no friggin' idea why it's there.
Another big mystery.
Another reason it's fun to be a physicist.
And if it is there, why should it be so small?
Why should there be so little?
Why should it have hid itself from us?
Why shouldn't there be enough of it
to substantially curve the space between us and the moon?
Why did there have to be such a small amount
that only the crazy best astronomers
in the world could find it?
Well, can't the same thing be said
about all of the constants?
All of the, can't that be said about gravity?
Can't that be said about the speed of light?
Like, why is the speed of light so slow?
So fast.
So slow.
Relative to the size of the universe,
can't it be faster?
Or no?
Well, the speed of light is a funny one,
because you could always choose units
in which the speed of light is one.
You know, we measure it in kilometers per second,
and it's 186,000, or miles per second.
It's 186,000 miles per second.
But if we use different units,
then we could make it one.
But you can make dimensionless ratios.
So, you know, you could say,
why is the time scale set by the expansion of the universe
so large compared to the time scale of a human life,
or so large compared to the time scale
for a neutron to decay, you know?
Yeah, yeah.
I mean, ultimately, you know, the reference,
the temporal reference frame here is a human life.
Maybe.
Isn't that the important thing for us descendants of apes?
Isn't that a really important aspect of physics?
Like, because we kind of experience the world,
we intuit the world through the eyes
of these biological organisms.
I mean, I guess mathematics helps you escape that
for a time, but ultimately, isn't that
how you wander about the world?
Absolutely.
That like a human life times only 100 years?
Because if you think of everything,
if you're able to think in, I don't know,
in billions of years,
then maybe everything looks way different.
Maybe universes are born and die,
and maybe all of these physical phenomena
become much more intuitive than we see
at the grand scale of general relativity.
Well, that is one of the, a little off the track here,
but that certainly is one of the nice things
about being a physicist is you spend a lot of time
thinking about insides of black holes
and billions of years in the future,
and it sort of gets you away from the day-to-day
into another fantastic realm.
But I was answering your question about
how there could be information in a black hole.
Yes.
So Einstein only gave us an approximate description,
and we now have a theory that corrects it, string theory.
And now sort of was the moment of truth.
Well, when we first discovered string theory,
we knew from the get-go that string theory
would correct what Einstein said,
just like Einstein corrected what Newton said.
But we didn't understand it well enough
to actually compute the correction,
to compute how many gigabytes there were.
And sometime in the early 90s,
we began to understand the mathematics
of string theory better and better,
and it came to the point where it was clear
that this was something we might be able to compute.
And it was a kind of moment of truth for string theory,
because if it hadn't given the answer
that Bekenstein and Hawking said it had to give
for consistency, string theory itself
would have been inconsistent,
and we wouldn't be doing this interview.
Well, that's a very dramatic statement, but yes.
That's not the most dramatic thing.
I mean, okay, that's very life and death.
You mean like, because string theory
was central to your work at that time,
is that what you mean?
Well, string theory would have been inconsistent.
Yeah, okay, so then it'd be a...
String theory would have been inconsistent.
But those inconsistencies can give birth
to other theories, like you said.
The inconsistency, right, something else
could have happened, yeah.
It would have been a major change
in the way we think about string theory if it,
and it was a good thing that one supposition
that the world is made of strings solves two problems,
not one, it solves the infinity problem
and it solved the Hawking's problem.
And also, the way that it did it was very beautiful.
It gave an alternate description.
So alternate description of things are very common.
I mean, we could, to take a simple example,
this bottle of water here is 90% full.
I could say it's 90% full.
I could also say it's 10% empty.
Those are obviously the same statement and they're,
it's trivial to see that they're the same,
but there are many statements that can be made
in mathematics and mathematical physics that are equivalent,
but might take years to understand that they're equivalent
and might take the invention or discovery
of whole new fields of mathematics
to prove they're equivalent.
And this was one of those.
We found an alternate description
of certain black holes in string theory,
which we could prove was equivalent
and it was a description of the black hole
as a hologram that can be thought of,
a holographic plate that could be thought of
as sitting on the surface of the black hole
and the interior of the black hole itself
sort of arises as a projection
or the near horizon region of the black hole
arises as a projection of that holographic plate.
So the two descriptions were the hologram,
the three-dimensional image, and the holographic plate.
And the hologram is what Einstein discovered
and the holographic plate is what we discovered.
And this idea that you could describe things
very, very concretely in string theory
in these two different languages, of course, took off
and was applied to many, many different,
many different contexts within string theory.
So you mentioned the infinity problem
and the Hawking problem, which Hawking problem?
That the black hole destroys information
or which Hawking problem are we talking about?
Well, there's really two Hawking problems.
They're very closely related.
One is how does the black hole store the information?
And that is the one that we solved in some cases.
So it's sort of like your smartphone.
How does it store its 64 gigabytes?
Well, you rip the cover off and you count the chips
and there's 64 of them, each with a gigabyte,
and you know there's 64 gigabytes.
But that does not solve the problem
of how you get information in and out of your smartphone.
You have to understand a lot more about the Wi-Fi
and the internet and the cellular and-
And that's where Hawking radiation,
this prediction starts to- That's where
Hawking radiation comes in.
And that problem of how the information gets in and out,
you can't, you couldn't have explained
how information gets in and out of an iPhone
without first explaining how it's stored in the first place.
So just to clarify, the storage is on the plate?
Is on the plate.
On the holographic plate,
and then it projects somehow inside the-
The bulk, the space-time is the hologram.
The hologram, man, I mean, do you have an intuitive,
when you sit late at night and you stare at the stars,
do you have an intuitive understanding
of what a holographic plate is?
Like, that there's two dimensions,
you know, projections that store information?
How a black hole could store information
on a holographic plate, I think we do understand
in great mathematical detail and also intuitively,
and it's very much like an ordinary hologram
where you have a holographic plate
and it contains all the information,
you shine a light through it and you get an image
which looks three-dimensional.
Yeah, but why should there be a holographic plate?
Like, why should there be?
Yeah, why?
That is the great thing
about being a theoretical physicist is
anybody can very quickly stump you
with a going to the next level of whys.
Yeah, the wys are sky blue, I can just keep asking, yeah.
Yeah, you can just keep asking
and it won't take you very long to...
So the trick in being a theoretical physics
is finding the questions that you can answer.
Sure.
So the questions that we think we might be able
to answer now and we've partially answered
is that there is a holographic explanation
for certain kinds of things in string theory.
Sure.
We've answered that.
Now we'd like to take what we've learned
and that's what I've mostly been doing
for the last 15, 20 years.
I haven't really been working so much
on string theory proper.
I've been sort of taking the lessons
that we learned in string theory
and trying to apply them to the real world
assuming only what we know for sure about the real world.
So on this topic, you co-authored a paper
with Stephen Hawking called Soft Hair on Black Holes.
Yes.
That makes the argument against Hawking's
original prediction that black holes destroy information.
Can you explain this paper?
Yes.
And the title.
Yeah.
Okay, so first of all, the hair on black holes
is a word that was coined by the greatest phrase master
in the history of physics, John Wheeler,
invented the word black hole.
And he also said that, he made the statement
that black holes have no hair.
That is, every black hole in the universe
is described just by its mass and spin.
They can also rotate as was later shown by Kerr.
And this is very much unlike a star, right?
Every star of the same mass is different
in a multitude of different ways.
Different chemical compositions,
different motions of the individual molecules.
Every star in the universe, even of the same mass,
is different in many, many different ways.
Black holes are all the same.
And that means when you throw some,
in Einstein's description of them,
which we think must be corrected.
And if you throw something into a black hole,
it gets sucked in.
And if you throw in a red book or a blue book,
the black hole gets a little bigger
but there's no way within Einstein's theory
of telling how they're different.
And that was one of the assumptions
that Hawking made in his 1974, 75 papers
in which he concluded that black holes destroy information.
You can throw encyclopedias, thesis defenses,
the Library of Congress.
It doesn't matter,
it's going to behave exactly the same uniform way.
Yeah, so what Hawking and I showed,
and also Malcolm Perry,
is that one has to be very careful
about what happens at the boundary of the black hole.
And this gets back to something I mentioned earlier
about when two things which are related
by a coordinate transformation are and are not equivalent.
And what we showed is that they're very subtle imprints
when you throw something into a black hole.
They're very subtle imprints left
on the horizon of the black hole
which you can read off at least partially what went in.
And so this invalidates Stephen's original argument
that the information is destroyed.
And that's the soft hair.
That's the soft hair, right.
And soft is a word that is used in physics
for things which have very low energy.
And these things actually carry no energy.
They're things in the universe which carry no energy.
You said, I think to Sean Carroll,
by the way, everyone should go check out
Sean Carroll's Mindscape podcast, it's incredible.
And Sean Carroll's an incredible person.
I think you said there maybe in a paper, I have a quote,
you said that a soft particle is a particle
that has zero energy, just like you said now.
And when the energy goes to zero,
because the energy is proportional to the wavelength,
it's also spread over an infinitely large distance.
If you like, it's spread over the whole universe.
It somehow runs off to the boundary.
What we learned from that is that
if you add a zero energy particle to the vacuum,
you get a new state.
And so there are infinitely many vacua,
plural for vacuum, which can be thought of
as being different from one another
by the addition of soft photons or soft gravitons.
Can you elaborate on this wild idea?
If you like, it spreads over the whole universe.
When the energy goes to zero,
because the energy is proportional to the wavelength,
it also spreads over an infinitely large distance.
If you like, it's spread over the whole universe.
Can you explain these soft gravitons and photons?
Yeah, so the soft gravitons and photons
have been known about since the 60s.
But exactly what we're supposed to do with them
or how we're supposed to think about them,
I think has been well understood only recently.
And in quantum mechanics,
the energy of a particle is proportional
to Planck's constant times its wavelength.
So when the energy goes to zero,
the wavelength goes to infinity.
Now, if something has zero energy
and it's spread all over the universe,
in what sense is it actually there?
That's been the confusing thing.
To make a precise statement
about when something is and isn't there.
Now, the simplest way of seeing,
so people might have taken the point of view
that if it has zero energy
and it's spread all over the universe,
it's not there, we can ignore it.
But if you do this, you'll get into trouble.
And one of the ways that you'll get into trouble
is that even though it has zero energy,
it doesn't have zero angular momentum.
If it's a photon, it always has angular momentum one.
If it's a graviton, it's angular momentum two.
So you can't say that the state of the system
with the zero energy photon should be identified
with the one without the zero energy photon,
that we can just ignore them,
because then you will conclude
that angular momentum is not conserved.
And if angular momentum is not conserved,
things won't be consistent.
And of course, you can have a lot of these things,
and typically you do get a lot of them.
And you can actually do a calculation
that shows that every time you scatter two particles,
you create an infinite number of them.
Infinite number of the soft photons and gravitons?
Of the zero energy ones, yeah.
And so these are, and they're somehow everywhere.
They're everywhere.
But they also contain information,
or they're able to store information?
And they're able to store information.
They're able to store an arbitrary
large amount of information.
So what we pointed out is,
so what these things really do,
one way of thinking of them is they rush off
to the edges of the universe,
spreading out all over the space.
It's like saying they rush off
to the edge of the universe.
Right.
And that includes, if the interior of the black hole
is not considered part of the universe,
that includes the edge of the black hole.
So we need to set up our description of physics
so that all the things that are conserved
are still conserved in the way that we're describing them.
And that will not be true if we ignore these things.
We have to keep careful track of these things.
And people had been sloppy about that.
And we learned how to be very precise and careful about it.
And once you're being precise,
you can actually answer this kind of very problematic thing
that Hawking suggested that black holes destroy information.
Well, what we showed is that there's an error
in the argument that all black holes are the same
because they hadn't kept track of these,
these very subtle things.
And whether or not this is the key error
in the argument remains to be seen,
or whether this is a technical point.
Yes, but it is an error.
It is an error.
And Hawking obviously agreed with it.
Hawking agreed with it.
And he was sure that this was the,
he was sure that this was.
This was a critical error.
That this was the critical error
and that understanding this would get us the whole story.
And that could well be.
What was it like working with Stephen Hawking
on this particular problem?
Because it's kind of a whole journey, right?
Well, you know, I love the guy.
He's so passionate about physics.
He just, yeah.
His oneness with the problem and the, I mean it's.
So his mind is all occupied by the world that's.
Yeah, and let me tell you, there's a lot of other things
with his illness and with his celebrity
and a lot of other things.
A lot of distractions pulling at his mind,
he's still there.
That's right, that's right.
I remember him turning down tea with Lady Gaga
so we could spend another hour on a paper.
That, my friends, is dedication.
What did you learn about physics?
What did you learn about life
from having worked with Stephen Hawking?
Well, he was one of my great teachers.
Of course, he's older than me.
And I was reading his textbooks in graduate school.
And you know, I learned a lot about relativity from him.
I learned about passion for a problem.
I learned about not caring what other people think.
Physics is an interesting culture.
Even if you make a great discovery, like Hawking did,
people don't believe everything you say.
In fact, people love to disagree.
It's a culture that cherishes disagreement.
And so, you know, he kept ahead with what he believed in
and sometimes he was right and sometimes he was wrong.
Do you feel pressure from the community?
So for example, with string theory,
it was very popular for a time.
There's a bit of criticism, or it's less popular now.
Do you feel the forces of the community
as it moves in and out of different fields?
Or do you try to stay,
like how difficult is it to stay intellectually
and mathematically independent from the community?
Personally, I'm lucky, I'm well equipped for that.
When I started out in graduate school,
the problem of quantum gravity
was not considered interesting.
You still did it anyway?
I still did it anyway.
I'm a little bit of a contrarian, I guess,
and I think that has served me well.
And people are always sort of disagreeing with me
and they're usually right, but I'm right enough.
And like you said,
the contradiction ultimately paves the path of discovery.
Yeah.
Let me ask you, just on this tension,
we've been dancing between physics and mathematics.
What to you is an interesting line
you can draw between the two?
You have done some very complicated mathematics
in your life to explore the laws of nature.
What's the difference between physics and mathematics to you?
Well, I love math.
I think my first love is physics
and the math that I've done, I've done too,
because it was needed.
In service of physics.
In service of physics,
but then, of course, in the heat of it,
it has its own appeal.
In the heat of it, I like it.
Sure.
It has its own appeal and I certainly enjoyed it.
And ultimately, I would like to think,
I wouldn't say I believe,
but I would like to think that there's no difference
between physics and mathematics,
that all mathematics is realized in the physical world
and all physics has a firm mathematical basis,
that they're really the same thing.
I mean, why would there be math
that had no physical manifestation?
It seems a little odd, right?
You have two kinds of math,
some that are relevant to the real world.
Well, they don't have to be contradictory,
but can't you not have mathematical objects
that are not at all connected to the physical world?
So, I mean, this is to the question
of is math discovered or invented?
So to you, math is discovered
and there's a deep linkage between the two.
Yeah, yeah, yeah.
Do you find it at all compelling, these ideas,
something like Max Tegmark,
where our universe is actually
a fundamentally mathematical object,
that math is, our universe is mathematical,
fundamentally mathematical in nature?
My expertise as a physicist
doesn't add anything to that.
It's not really, you know, physics is,
you know, I was once very interested in
philosophy and, you know, physics,
physics, I like questions that can be answered,
that it's not obvious what the answer is
and that you can find an answer to the question
and everybody will agree what the answer is
and that there's an algorithm for getting there.
Not that these other questions aren't interesting
and they don't somehow have a way of presenting themselves,
but to me, the interesting thing is to,
is motion in what we know, is learning more
and understanding things
that we didn't understand before.
Things that seemed totally confusing,
having them seem obvious, that's wonderful.
So I think that's, those questions are there.
I mean, I would even go further,
you know, the whole multiverse,
I don't think there's too much concrete
we're ever gonna be able to say about it.
This is fascinating because you spend so much time
in string theory, which is devoid from a connection
to the physical world for a long time.
Like, not devoid, but it travels in a mathematical world
that seems to be beautiful and consistent
and seems to indicate that it could be
a good model of the laws of nature,
but it's still traveling independently
because it's very difficult to experimentally verify.
But there's a promise laden in it,
in the same way multiverse or,
you can have a lot of kind of very far out there questions
where your gut and instinct and intuition says
that maybe in 50, 100, 200 years,
you'll be able to actually have
strong experimental validation, right?
I think that with string theory,
I don't think it's likely that we could measure it,
but we could get lucky.
In other words, just to take an example,
about 10 or 20 years ago,
it was thought that they had seen a string in the sky
and that it was seen by, you know,
doubled stars that were gravitationally lensed
around the gravitational field
produced by some long string.
There was a line of double lensed,
now the signal went away, okay?
But people were hoping that they'd seen a string
and it could be a fundamental string
that had somehow gotten stretched
and that would be some evidence for string theory.
There was also BICEP2, which,
the experiment was wrong, but it could have happened.
It could have happened that we got lucky
and this experiment was able to make direct measurements,
certainly would have been measurements of quantum gravity,
if not string theory.
So it's a very logical possibility
that we could get experimental evidence from string.
That is a very different thing than saying,
do this experiment, here's a billion dollars,
and after you do it, we'll know
whether or not strings are real.
But I think it's a crucial difference.
It's measurable in principle
and we don't see how to get from here to there.
If we see how to get from here to there,
in my eyes, it's boring, right?
So when I was a graduate student,
they knew how to measure the Higgs boson.
It took 40 years, but they didn't,
not to say that stuff is boring,
I don't want to say that stuff is boring,
but when Magellan set out,
he didn't know he could get around the world.
There was no map, you know?
So I don't know how we're gonna connect in a concrete way,
all these ideas of string theory to the real world.
And you know, when I started out in graduate school,
I said, what is the most interesting problem
that there might be, the deepest, most interesting problem
that there might be progress on in 60 years?
And I think it could be, you know,
that, you know, if you're a student,
that, you know, in another 30 years,
that maybe we'll learn that we have understood
how black holes store information, you know?
That doesn't seem wild, that we're able to abstract
what we learned from string theory
and show that it's operative and, you know,
I mean, the Bose-Einstein convents that they did,
you know, when Bose and Einstein predicted it,
when was that, the 30s maybe, early 30s?
It took, there were 20 orders of magnitude
that were needed in order to,
in improvement, in order to measure it,
and they did 50 years later.
So, and you couldn't have guessed how that had happened,
how they could have gotten that.
And it could happen that we,
I don't think we're gonna like see
the heterotic string spectrum at an accelerator,
but it could be that things come around
and in an interesting way, and somehow it comes together.
And the fact that we can't see to the end
isn't a reason not to do it, you know?
We're just, you know, what did they do
when they were trying to find the specific, right?
They just, they took every route.
They just tried everything.
And that's what we're doing.
And we're taking, and I'm taking the one
that my nose tells me is the best, you know?
And other people are taking other ones, and that's good,
because we need every person taking every route.
And, you know, if somebody on another route
finds something that looks really promising,
you know, I'm gonna make a portage over the mountain
and get on their stream, you know?
So the fact that you don't see the experiment now
isn't to me a reason to give up on what I view
as the most fundamental paradox in 20th century,
20th, in present physics, 20, 21st century physics.
Absolutely, you can see that it's possible.
You just don't know the way.
But that's what I mean
why some of the philosophical questions
could be formulated in a way
that's explorable scientifically.
So some of the stuff we've talked about,
but, you know, for example,
this topic that's become more okay to talk about,
which is the topic of consciousness.
You know, to me as an artificial intelligence person,
that's a very practically interesting topic.
But there's also philosophers.
Sean Carroll loves to argue against them.
But there's some philosophers that are panpsychists.
I'm not against philosophers.
It's just not as fun, I don't.
It's not as fun, all right.
But they start a little flame of a fire going
that some of those flames, I think,
eventually become physics.
So eventually become something that we can really,
like having them around is really important
because you'll discover something
by modeling and exploring black holes that's really weird.
And having these ideas around,
like the ideas of panpsychists
that consciousness could be a fundamental force of nature.
Just even having that crazy idea,
swimming around in the background,
could really spark something
where you were missing something completely.
And it's just, that's where the philosophy done right,
I think, is very useful.
That's where even the, you know, these thought experiments,
which is very fun in the sort of the tech sci-fi world
that we live in a simulation,
that, you know, taking a perspective of the universe
as a computer, as a computational system
that processes information,
which is a pretty intuitive notion,
but you can just even reframing it that way for yourself
could really open up some different way of thinking.
Could be.
And then you have, I don't know if you're familiar
with Steven Wolfram's work
of like cellular automata and complexity.
Yeah, I did a podcast with Steven.
Steven, that's awesome.
I mean, to me, forget physics, forget all that.
Cellular automata make no sense.
They're so beautiful.
They're so, from simple rules, you can create complexity.
I just don't think, you know,
he wrote a book on new kind of science,
basically hinting at,
which a lot of people have hinted at,
it's like, we don't have a good way
to talk about these objects.
We don't, we can't figure out what is happening here.
These simple, these trivial rules can create
incredible complexity.
He's totally right about that, yeah.
And physicists, I guess, don't have,
don't know what to do with that.
Don't know what to do with cellular automata.
Because you can describe the simple rules
that will govern the system,
or how complexity can emerge, like incredible complexity.
Yeah.
Of course, Wolfram's version of that is that
physicists will never be able to describe it.
Right, yeah, exactly.
He tries to prove that it's impossible.
What do you make of that?
What do you make about the tension of being a physicist
and potentially not being able to,
it's like Freud or somebody that may be,
Sigmund Freud, that maybe you'll never be able
to actually describe the human psyche.
Is that a possibility for you?
That you will never be able to get to the core
fundamental description of the laws of nature?
Yeah, so I had this conversation with Weinberg.
Yeah, how'd it go?
So Weinberg has this book called Dreams of a Final Theory.
Yeah.
And I had this conversation with him, I said,
why do you think there's ever gonna be a final theory?
Why should there ever be a final theory?
I mean, what does that mean?
Do physics departments shut down?
We've solved everything?
And you know, doesn't it seem that every time
we answer some old questions, we'll just find new ones
and that it will just keep going on forever and ever?
He said, well, that's what they used to say about the Nile.
They were never gonna find the end.
Then one day they found it.
Yeah.
So I don't know.
String theory doesn't look like a candidate to me
for a final theory.
As it stands now.
It doesn't get to the bottom of the well,
to the sides, to the whole thing.
Yeah, it seems to me that even if we kind of solved it
and we did experiments, there still would be more questions.
Like why are there four dimensions instead of six?
It doesn't seem to have anything in it
that would explain that.
You can always hope that there's something
that we don't know about string theory that will explain it.
But it still doesn't look like
it's gonna answer every question.
And why is there one time, not two?
Why is there, you know, it doesn't seem like it's,
I don't even know what it would mean
to answer every question.
Well, to answer every question, obviously,
so when you refer to the theory of everything,
you'll be able to have a, if it exists,
it would be a theory that allows you to predict precisely
the behavior of objects in the universe
and their movement, right?
What about them, their movement?
Yeah.
Like precisely no matter the object.
Right, that's true.
So that would be a really interesting state of affairs.
If we could predict everything
but not necessarily understand everything.
So for example, let's just forget about gravity.
I mean, we're not too far from that situation.
If we forget about gravity, the standard model,
in principle, given a big enough computer,
predicts almost everything.
But if you look at the standard model,
it's kind of a laundry list with neutrino masses
and all that stuff.
There are hundreds of free parameters.
Where do they come from?
Is there an organizing principle?
Is there some further unification?
So being able to predict everything
is not the only goal that physicists have.
So on the way to trying to predict,
you're trying to understand.
That's actually probably the goal is to understand.
Yeah.
But right, we're more interested in understanding
than actually doing the predictions.
But the predictions are more focusing
on how to make predictions is a good way
to improve your understanding
because you know you've understood it
if you could do the predictions.
Yeah, one of the interesting things
that might come to a head with is artificial intelligence.
There's an increasing use of AI in physics.
We might live in a world where AI would be able
to predict perfectly what's happening.
And so as physicists, you'll have to come
to the fact that you're actually not
that interested in prediction.
I mean, it's very useful.
But you're interested in really understanding
the deep laws of nature versus a perfect predictor.
Like you wanna play chess.
But even within AI, AI people are trying to understand
what it is that the AI bots have learned
in order to produce whatever they produce.
For sure, but you still don't understand deeply,
especially because they're getting,
especially language models, if you're paying attention,
the systems that are able to generate text,
they're able to have conversations,
chat GPTs, the recent manifestation of that.
They just seem to know everything.
They're trained on the internet.
They seem to be very, very good
at something that looks like reasoning.
They're able to generate, you can ask them questions.
They can answer questions.
It just feels like this thing is intelligent, right?
And I could just see that being possible with physics.
You ask any kind of physical question,
and it'll be able to, very precise
about a particular star system or a particular black hole,
and it'll say, well, these are the numbers.
It'll perfectly predict.
And then, sure, you can understand
how the neural network is, the architecture is structured.
Actually, for most of them now, they're very simple.
You can understand what data it's trained on,
huge amount of data.
You're giving it a huge amount of data
from a very nice telescope or something.
And then, but it seems to predict everything perfectly.
How a banana falls when you throw it.
Everything is perfectly predicted.
You still don't have a deep understanding
of what governs the whole thing.
And maybe you can ask it a question,
and it'll be some kind of hitchhiker's guide
to the galaxy type answer that,
it's a funny world we live in.
Of course, it's also possible that there's no such deep,
simple governing laws of nature behind the whole thing.
I mean, there's something in us humans.
It's possible.
That wants it there to be, but doesn't have to be, right?
Right.
Again, you already bet the farm.
But if you were to have a second farm,
do you think there is a theory of everything
that we might get at?
So, simple laws that govern the whole thing.
I don't, honestly, I don't know.
But I'm pretty confident that if there is,
we won't get to it in my lifetime.
I don't think we're near it.
But doesn't it feel like,
the fact that we have the laws we do
that are relatively simple already,
that's kind of incredible.
It's just, there seems to be simple laws
that govern things, right?
By a theory of everything, you mean theory,
a theory of everything,
an algorithm to predict everything.
But a simple algorithm.
A relatively simple algorithm to predict everything.
So, for me, it would be a sad day
if we arrived at that without answering
some deeper questions.
Sure, of course, it definitely is.
But the question, yes.
But one of the questions before we arrive there,
we can ask, does such a destination even exist?
So, because asking the question and the possible answers
and the process of trying to answer that question
is in itself super interesting.
Is it even possible to get there,
where there's an E equals MC squared type of,
there's a function.
Okay, you can have many parameters.
But find that number of parameter function
that can predict a lot of things about our universe.
Well, okay, but just to sort of throw one thing in,
in order to answer every question,
we would need a theory of the origin of the universe.
Right.
And that is a huge task, right?
So, and the fact that the universe seems to have a beginning
defies everything we know and love, right?
Because one of the basic principles of physics
is determinism, that the past follows from,
the present follows from the past,
the future follows from the present, and so on.
But if you have the origin of the universe,
if you have a big bang,
that means before that, there was nothing.
You can't have a theory
in which something follows from nothing.
So, somehow-
Sounds like you don't like singularities.
Well-
I thought for somebody that works with black holes,
you would get used to them by now.
No, no, I like this because it's so hard to understand.
I like it because it's hard to understand.
But it's really challenging us.
I don't think we're close to solving that problem.
So even-
And string theory has basically had nothing,
there's been almost nothing interesting said about that
in the last many decades.
So string theory hasn't really looked at the big bang.
It hasn't really tried to get to the origin.
Not successfully.
There aren't compelling papers
that lots of people have read that,
people have taken it up and tried to go at it.
But there aren't compelling.
String theory doesn't seem to have a trick
that helps us with that puzzle.
Do you think we'll be able to sneak up
to the origin of the universe?
Like reverse engineer it from experimental,
from theoretical perspective?
Like, okay, if we can, what would be the trajectory?
You've already gotten yourself in trouble, you see?
Because you used the word reverse engineer.
So if you're going to reverse engineer,
that means you forward engineering
means that you take the present and determine the future.
Reverse engineering means that you take the present
and determine the past.
But-
Estimate the past, but yes, sure.
But if the past was nothing,
how are you ever going to reverse engineer to nothing?
That's hard to do.
Run up against the nothing, right?
Until you have mathematical models that break down nicely
to where you can actually start to infer things.
Let's work on it.
No, but do you think that-
Maybe, but people have tried to do things like that.
Yeah, and have not succeeded.
It's not something that we're getting A pluses in.
Sure.
Let's pretend we live in a world
where in 100 years we have an answer to that.
What would that answer look like?
What department is that from?
What fields led us there?
Not what fields, what set of ideas in theoretical physics?
Is it experimental, is it theoretical?
Like what can you imagine possibly could have
possibly lead us there?
Is it through gravitational waves
and some kind of observations there?
Is it investigation of black holes?
Is it simulation of universes?
Is it maybe we start creating black holes somehow?
I don't know.
Maybe some kind of high energy physics type of experiments?
Well, I have some late night ideas about that
that aren't really ready for prime time.
Okay, sure, but you have some ideas.
Yeah, yeah, and many people do.
It could be that some of the advances
in quantum information theory are important
in that they kind of go beyond taking quantum systems
and just replicating themselves,
but combining them with others.
Do you think, since you highlighted the issue
with time and the origin of the universe,
do you think time is fundamental or emergent?
I think ultimately it has to be emergent.
Yeah, what does it mean for time to be emergent?
Well, let's review what it means for space to be emergent.
Yes.
What it means for space to be emergent
is that you have a holographic plate
and you shine some light that's moving in space
and it produces an image
which contains an extra spatial dimension
and time just goes along for the ride.
So what we'd like to do,
and indeed there is some rather concrete work
in this direction, though again I would say
even within our stringing community
we're not getting A pluses on these efforts,
but what we'd like to do is to see
examples in which the extra space-time dimension is time.
In other words, usually what we understand
very well mathematically is how to take systems
in some number of space-time dimensions
and rewrite them as a plate in fewer space dimensions.
What we'd like to do is to take systems
with one time and some number of space dimensions
and to rewrite them as a system
that had only space dimensions in it,
had no time evolution.
And there are some fairly concrete ideas
about how to do that, but they're not universally accepted
even within the stringy community.
But isn't it wild to you?
Yes.
For it to be emergent?
How do we intuit these kinds of ideas as human beings
for whom space and time seems as fundamentals
as apples and oranges?
Well, they're both illusions.
Okay.
They're both illusions, even time.
You co-authored a paper titled
Photon Rings Around Warped Black Holes.
First of all, whoever writes your paper titles,
you, like the soft hair and the term black hole
and the big bang, you're very good
at coming up with titles yourself.
Anyway, you co-authored a paper titled
Photon Rings Around Warped Black Holes.
In it, you write, quote,
recent work has identified a number of emergent symmetries
related to the intricate self-similar structure
of the photon ring.
So what are photon rings?
What are some interesting characteristics of a photon ring?
So that was a paper with Dan Kopitz and Alex Lipsaska
that just came out.
And this paper is kind of a wonderful example
of what happens when you start to talk to people
who are way out of your comfort zone of no different stuff
and look at the world a different way.
And some two or three years ago,
I'm part of this, the Black Hole Initiative,
and I'm also part of this
event horizon telescope collaboration
that took the famous,
though I had nothing to do with the experiment,
but that took the famous picture of the donut of M87.
And through conversations with them,
which started out in an effort
to understand the image that they'd seen.
So it's a great thing for somebody like me,
a theoretical physicist lost,
seemingly lost in string land
to be presented with an actual picture
of a black hole and to be asked,
what can we learn from this?
So with some help from Michael Johnson and Alex Lipsaska
and a bunch of other people ventureized in collaboration,
we came up with a fantastic, beautiful answer
using Einstein's theory,
that is both shaping the future of,
now it is shaping the future
of improved black hole photographs.
What do you want to concentrate on in the photograph?
You just point it at the sky and click?
No, you don't do that.
You optimize for various features.
And it's both shaping that
and in the process of talking to them
and thinking about how light behaves around a black hole,
black holes just have so many magic tricks
and they do so many weird things.
And the photon ring is among the weirdest of them.
We understood this photon ring and in the process of this,
we said, hey, this photon ring
has gotta be telling us something about the puzzle
of where the holographic plate is
outside of a ordinary astrophysical black hole.
And we nailed it for the stringy black holes,
but they have a somewhat different character.
What's a stringy black hole?
The black holes that describe a string theory?
The black holes that are contained in string theory
and they have different structure in them.
Well, but actually, can we step back?
So what was the light in the image taken in 2019?
Well, not taken in 2019, presented in 2019.
So here's the puzzle.
What they really saw,
so the black holes tend to gather stuff
that swirls around it.
And they don't know what that stuff is made of.
They don't know what its temperature is.
They don't know what kind of magnetic fields
there are around there.
So the form of the image has a lot of unknowns in it
that it's dependent on many other things
other than the geometry of the black hole.
So most of what you're learning is about the stuff.
Now the stuff, the swirling stuff,
the hot swirling stuff is interesting as hell,
but it's not as interesting as the black hole,
which are the most, in my view,
the most interesting things in the universe.
So you don't wanna just learn about the stuff.
You wanna learn about the black hole
that is swirling around.
So one of the, at the very first step,
at the very primitive level,
this is just a big leap for human civilization
to be able to see a black hole.
And the way you can see it
is because there's stuff around it.
But you don't get to learn much about the black hole,
but you get to learn more about the stuff
just from the image.
Yeah, but you're not gonna learn about the details
before you've even seen it.
Because there's too many parameters,
there's too many variables that govern the stuff.
Yeah, so then we found a very wonderful way
to learn about the black hole.
And here's how it works.
A black hole is a mirror.
And the way it's a mirror is if light,
a photon bounces off your face
towards the black hole,
and it goes straight to the black hole,
just falls in, you never see it again.
But if it just misses the black hole,
it'll swing around the back and come back to you.
And you see yourself from the photon
that went around the back of the black hole.
But not only can that happen,
the black hole, the photon can swing around twice
and come back.
You actually see an infinite number of copies of yourself.
Like with a little bit of a delay.
With a little bit of a delay, right.
This is awesome.
Yeah, and in fact...
I mean, we're not used to an object
that bends light like that, right?
Yeah, yeah.
So you're gonna get some trippy cool effects.
And in fact, one of my students
has made a really awesome computer animation of this,
which I'm gonna show at a public lecture
in a couple of weeks
where the audience will see infinitely many copies
of themselves swirling around the black hole.
So a black hole is like a hall of mirrors,
like in a department store where you go
and there's the three mirrors
and you see infinitely many copies of yourself.
Think of the black hole as the mirror.
And you go in there with your clothes,
if you wanna know about your clothes,
you just look at the direct image.
You're not learning anything
about the configuration of mirrors.
But the relation of the image you see in front of you
to the one you see at the side and the next one and so on
depends only on the mirrors.
It doesn't matter what clothes you're wearing.
So you can go there a thousand times
wearing different clothes,
but each time there will be the same relation
between the subsequent images.
And that is how we're gonna learn about the black holes.
We're gonna take the stuff that is swirling around
and we're gonna tease out the subsequent images
and look at their relation.
And there's some very beautiful,
really beautiful mathematics,
which we were surprised to realize with the volumes
and volumes of papers on black holes and their properties,
this particular,
because it was a physical question
that had never been asked in exactly this way.
So basically you're looking at the-
The relationship between the subsequent images.
But those are ultimately formed
by photons that are swirling around.
Photons that are orbiting.
So the photon ring are the photons that orbit around.
And beyond.
So like orbit and lose orbit.
Like are they-
Yeah.
Wow.
And that starts to give you,
what can you possibly figure out mathematically
about the black hole?
The geometry of it?
Does the spin of it?
The geometry, the spin.
And you can verify things behaving.
We have never seen a region of space time
with such high curvature.
I mean, the region around a black hole is crazy.
It's not like in this room.
The curvature is everything.
You spend probably enough time with the math and the photons.
Can you put yourself in that space?
So we're like having a conversation
in pretty peaceful, comfortable, flat space.
Are you able to put yourself in a place around a black hole?
Yeah, I'm able to imagine that kind of thing, yeah.
So for example, and actually there's a wonderful movie,
Interstellar, and in that movie,
Kip Thorne of course is a great theoretical physicist,
experimental, who later won the Nobel Prize for LIGO.
And that movie is very accurate scientifically.
And there's some funny statements in there
that of the 100 million people who saw that movie,
there can't be more than 10 or 20 understood
about why Matthew McConaughey is ejecting the trash
in a certain direction in order to.
But for example, if I were a spinning black hole right here,
if I were spinning fast enough,
you wouldn't be able to stay still there.
You'd have to be orbiting around like that, you know?
You'd have to have your microphone on a rotating basis.
But I wonder what the experience is,
what the actual experience,
because I mean, space itself is curved.
Well, if space gets very curved, you get crushed.
You know, your body gets ripped apart
because the forces are different
on different parts of your body.
Sure, okay, so that would be.
But it can be less curved
so that the curvature is very noticeable,
but you're not ripped apart.
The fact that this was just nonchalantly stated
is just beautiful, like two biological systems
discussing which level of curvature is required
to rip apart said biological system.
Very well, so you propose in the paper
that a photon ring of a warped black hole
is indeed part of the black hole hologram.
A photon ring of a warped black hole
is indeed part of the black hole hologram.
So what can you intuit about the hologram
and the holographic plate from looking at the photon rings?
Well, this paper is exploring a new idea.
It's not making a new discovery, so to speak.
It's exploring an idea and the ins and outs of it
and what might work and what might not.
And this photon ring, somehow everybody always thought
that the holographic plate sat at the horizon
of the black hole.
Right.
And that the quantum system that describes the black hole
is inside the horizon.
And in fact, we think it's plausible
and we give some evidence in some soluble examples,
in this case in an example in one lower dimension
where we can handle the equations better,
that the quantum system that describes the black hole
should correspond to a region of space-time
which includes the photon ring.
So it's bigger.
So that would be the holographic plate?
So all of it? That would be
the holographic plate. All of that.
I mean, we didn't prove this.
We put it out there.
It hadn't really been considered previously.
We put it out there and it does seem more plausible
than the idea that it sits literally at the horizon.
And it is a big outstanding problem
of how you have a holographic reconstruction
of black holes like M87.
Do you think there could be further experimental data
that helps explore some of these ideas
that you have for photon rings
and holographic plates through imaging
and through higher and higher resolution images
and also just more and more data?
I wish so, but I don't think so.
But what I think already has happened
and will continue to happen
is that there are many different ways
that theorists and observers can interact.
The gold standard is the theorist makes a prediction,
the observer measures it and confirms it,
or the observer makes a discovery
and the theorist explains it.
But there's a lot less than that,
which is really kind of the bread and butter of,
those are dramatic moments when that happens, right?
Those are once in a lifetime moments.
Those are once in a lifetime moments when that happens.
But the bread and butter is more
when it has already happened,
they came to us and said,
what is the interesting theoretical things
we can understand in this swirl around the black hole
and we give an answer and then that in turn jogged us
to think about the holographic principle
in the context of M87 a little bit differently.
And so it's a useful, and in the same vein,
it's useful to talk to the philosophers
and it's useful to talk to the mathematicians
and a lot of, you gotta, we just gotta,
we don't know where we're going,
we just gotta do everything.
Let me ask you another sort of philosophical type question,
but not really actually, it seems that thought experiments
are used, so it's not just mathematics
that makes progress in theoretical physics,
but thought experiments do, they did for Einstein as well.
They did for a lot of great physicists
throughout history over the years.
How is your ability to generate thought experiments
or just your intuition about some of these weird things
like quantum mechanics or string theory or quantum gravity
or yeah, even general relativity,
how has your intuition improved over the years?
Have you been able to make progress?
The hard part in physics is most problems are
either doable, most problems that a theoretical calculation
that a theoretical physicist would do,
there's no end of problems whose answer
is uninteresting, can be solved,
but the answer is uninteresting.
There's also no end of problems that are very interesting,
some of which you've asked me,
but we don't have a clue how to solve them.
And when first presented with a problem,
almost every problem is one or the other.
It's the jackpot when you find one
that isn't one or the other.
And-
It seems like there's a gray area between the two, right?
That's where you should be looking.
Well, I wouldn't describe it as a gray area.
I would describe it as a knife edge.
So it's a very small area.
There isn't like a huge area with a sign.
Here are problems that are doable
and people want to know the answer.
In some deep sense,
that's where timing is everything with physics,
with science, with discovery.
With timing.
I mean, I think earlier in my career,
I erred more on the side of problems
that were not solvable.
The ambition of youth.
Yeah.
What made you fall in love with physics at first,
if we can go back to the early days?
You said black holes were there in the beginning,
but what made you,
do you remember what really made you fall in love?
You know, I wanted to reach nirvana
and I sort of realized that wasn't going to happen.
And then after that, I wanted to know the meaning of life.
And I realized I wasn't,
probably wasn't going to figure that out.
And then I wanted to like understand, you know,
justice and socialism and world things
and couldn't figure those out either.
And the simplest.
Smaller and smaller problems.
Smaller and smaller problems.
I mean, most of this, I'm talking about adolescents,
you know, but it was the biggest problem
that I thought that there was a prospect of,
but not a hundred percent, you know?
And I was definitely ready to
spend my life in the wilderness,
knocking my head against the wall,
but I haven't had to.
I haven't solved them,
but I've said enough interesting things
that you're interviewing me.
So I'm not in the wilderness, but yeah, so.
Do you remember the early days?
Do you feel nostalgic when you think back to the ideas,
the circumstances that led down,
that led you down the path towards black holes,
towards theoretical physics,
towards the tools of physics,
towards this really fascinating world
of theoretical physics?
Well, I wouldn't add nostalgia to it because
it's not like a, you know,
a summer in Italy or something.
It's like there's results that are there,
that the people are, and that's what's so gratifying.
I mean, of course one's name disappears from these things,
unless you're Einstein or Newton or something.
People are not gonna remember my name in 50 years.
Well, most, basically every name will be forgotten
in hundreds of years, yeah.
Yeah.
Are you able to, by the way, love the idea,
the exploration of ideas themselves
without the names, the recognition, the saying?
Yeah, that's what I'm saying.
So I have not, I hope someday, but I have not,
there are some experiments now to verify
some of my predictions about properties of gravity
and so on, but I have not, like,
you know, most of what I've done is in the,
you know, it could happen still.
It's still a logical possibility
that everything having to do with string theory
and the, I mean, as we mentioned,
I'm betting the farm that it's not,
but it is indeed a logical possibility
that people will say, can you believe Lex Fridman
interviewed Elon Musk and Kenya West
and then he interviewed Strominger,
who was on this, working on this theory
that just completely went into the,
completely went into the toilet, you know?
I'm gonna make, I'm gonna get,
with a wife I don't have, I'm gonna make a public statement,
she'll be on stage, I'll say I'm really sorry
I made this giant mistake of platforming
this wild-eyed physicist that believed for decades
in the power of theoretical physics, yes.
No, like you said.
So that could happen, it could happen, it could happen.
It's in the, and of course if that couldn't happen,
it wouldn't be real exploration, right?
Absolutely.
And so, but I, you know, I do take a lot of satisfaction
that some of the things I discovered
are at the minimum mathematical truths
and they're still, so you don't have that sort
of nostalgic feeling of it being something
that was gone and I'm still making discoveries now
that I'm as excited about.
We'll see if they hold the test of time
that stand the test of time that these other ones did,
but that I'm as excited about as I was about those
when I made them.
I am easily excitable, as my friends will tell you.
Well, one interesting thing about you is.
And I have been very excited about things
which turned out to be completely wrong, you know?
Well, that's, the excitement is a precondition
for breakthroughs, but you're also somebody
like just like you said, you don't have a cynical view
of the modern state of physics.
No.
So there's a lot of people that glorify
like the early days of string theory
and that, you know, all these things have been made
in the 20s. People are always, yeah, yeah, yeah.
But you're saying like this to you might be one of
if not the most exciting times to be a theoretical physicist
like when the alien civilizations find years from now
that visit Earth will look back,
they'll think the 21st century,
some of the biggest discoveries ever
were made in the 21st century.
Even when they have a measurement of string theory,
the fun's over.
Then we have to go on to something new, you know?
No, there's deep, there's going to be deep,
the fun is over.
Oh man.
But there is an end to the Nile, right?
I mean, that there's.
Is there?
Who told you?
Some Weinberg guy.
Let me ask you another trippy out there question now.
So again, perhaps unanswerable from a physics perspective,
but do you wonder about alien civilizations?
Do you wonder about other intelligent beings out there
making up their own math and physics
trying to figure out the world?
Do you think they're out there?
It is hard to understand why there would,
given that there's so many planets
and of course there's Drake's formula
and we don't exactly know what the,
but I mean, I think Fermi's paradox
that, you know, is a real paradox
and I think there probably are
and I think it's very exciting that,
you know, we might, you know, find some,
it's a logical possibility that we could learn about it.
I mean, to me it's super interesting
to think about aliens from a perspective of physics
because so any intelligent civilization
is going to be contending with the ideas
or just trying to understand the world around it.
So I think that the alien,
I think that the universe is filled
with alien civilizations.
So they all have their physicists, right?
They all have their,
they're all trying to understand the world around them
and it's just interesting to me
to imagine all these different perspectives,
all these different Einsteins.
Like trying to make sense of like-
Though they might be more different than we think.
They might be different in a way
that we haven't even thought of.
Like smarter or different?
Just different, something that we don't even,
we're not even able to describe now.
We just haven't thought of that, you know?
Yeah, this is the really frustrating thing
when we think from me as an AI person,
you start to think about what is intelligence,
what is consciousness,
and you start to sometimes, again, evening thoughts,
is how narrow our thinking is about these concepts.
That it could be intelligence could be,
something could be intelligent and be very different.
Intelligent in a very different way
that we won't be able to detect
because we're not keeping an open mind,
open enough mind, and that's kind of sad
because to me there's also just a strong possibility
that aliens or something like alien intelligence
or some fascinating, beautiful physical phenomena
are all around us and we're too dumb to see it, for now.
We're too close-minded to see it.
There's something we're just deeply missing,
whether it's like fundamental limitations
of our cognitive abilities
or just because our tools are too primitive right now.
It's like you said, the ideas seem trivial
once you've figured it all out, looking back.
But that kind of makes me sad
because there could be so much beauty in the world
we're not seeing because we're too dumb.
There surely is.
And that's, I guess, the process of science and physics
is to keep exploring, to keep exploring,
to find the thing that will, in a century, seem obvious.
Well, it's something we know for sure.
I mean, the brain we don't really understand
and that's gotta be some fabulously beautiful story.
I'm hoping some of that story will be written
through the process of trying to build a brain,
so the process of engineering intelligence,
not just the neuroscience perspective
of just looking at the brain, but trying to create it.
But yeah, that story hasn't been written almost at all,
which is the early days of figuring that one out.
But see, you said that math is discovered,
so aliens should at least have the same math as us, right?
I think so.
Maybe different symbols?
Oh, well, they might have discovered different,
they might have discovered it differently
and they might have had a different idea of what a proof is.
Sure, yeah.
We're very like black and white with the proof thing.
Maybe they're looser.
Right, well, so you can know something is true.
First of all, you never know something is true
with 100% uncertainty.
I mean, you might have had a blackout,
just to be, it's never 100%, right?
You might have had a momentary lapse of consciousness
as the key step in the proof and nobody read it
and whatever, okay?
So you never know for sure.
But you can have a preponderance of evidence,
which makes it, and preponderance of evidence
is not accepted very much in mathematics.
And that was sort of how the famous Ramanujan worked.
He had formulas which he guessed at
and then he gathered a preponderance of evidence
that you were sure they were true.
So there might be, or something completely different.
They might function in a very different way.
Let me ask you kind of a heavy question for physicists,
but one on nuclear weapons.
Just in general, what do you think about nuclear weapons
where, like philosophical level,
where brilliant physicists and brilliant engineering
leads to things that can destroy human civilization?
Sort of like some of the ideas that you're working on
that have power when engineered into machines, into systems.
Is there some aspect of you that worries about that?
Well, first of all, I don't know what the brilliant
had to do with it, because of course,
Oppenheimer and all that, okay, they did it really fast,
but if you didn't have Oppenheimer,
I mean, it would all have happened anyway.
It had a reality of its own.
The possibility of making a nuclear,
it didn't depend on the fact that the physicists
who built it were brilliant.
Maybe that sped it up by a year or two years,
but by now we'd have nuclear weapons.
It's something that...
So the ideas have momentum and that they're unstoppable.
Right, the possibility of making nuclear weapons
was discovered, right?
It was there before, we didn't, it's not like somebody made it, right?
Without Picasso, there would never have been a Guernica,
but without Oppenheimer, there would surely have still been
an atom bomb.
But timing matters, right?
Timing is very important.
There's a guy with a mustache.
Of course, of course, of course.
The timing mattered there, but I...
Yeah, okay, I mean, you could try to make a case
for stopping...
No, no, no, no, it's the case of carrying the burden
of the responsibility of the power of ideas
when manifested into systems.
So it's not a game.
It's not just a game of fun mathematics,
just same with artificial intelligence.
You have this, you know, a lot of people in AI,
you know, a lot of people in the AI community,
it's a fascinating, fun puzzle,
how to make systems more and more intelligent,
how to, you have a bunch of benchmarks,
you try to make them perform better and better and better,
and all of a sudden, you have a system
that's able to outsmart people.
It's now able to be used in geopolitics.
It's able to create super intelligent bots
where they're able to, at scale, control the belief
of a population of people, and now you can have world wars.
You can have a lot of really risky instabilities.
They're incredible.
They really are incredible.
And so just, like, there is some responsibility.
This is not sort of...
It's a beauty and a terror to these ideas, you know?
Yeah.
At that moment, it was certainly a question for Oppenheimer
and everybody who participated in that.
What is the responsible way to serve society
when you're sort of accidentally in this position
of being at the forefront of a development
that has a huge impact on society?
I don't see my work a likelihood of having a huge impact
on the development of society itself,
but if I were you, working on AI,
I think that there is a possibility there,
and that it is, as a responsible scientist,
that it's really not a good thing to say,
I'm just the scientist here
and I'm figuring out what's possible,
because you're in a role where you have more
of a podium to influence things than other people,
and it's your responsibility as a citizen of the planet,
or let me phrase it a little less shouldy,
you have an opportunity as a citizen of the planet
to make the world a better place,
which it would be sad to bypass.
Yeah, it's a nice world without going.
It'd be nice to keep it going for a little bit longer.
Andrew, I'm really honored that you sit down with me.
Thank you for your work.
Thank you for your time.
Well, it was a really great conversation.
I really enjoyed it.
You really covered a lot.
I can't believe you're able to discuss at this level
on so many different topics, so it's a pleasure.
It was super fun.
Thank you.
Thanks for listening to this conversation
with Andrew Strominger.
To support this podcast,
please check out our sponsors in the description.
And now let me leave you with some words
from Werner Heisenberg.
Not only is the universe stranger than we think,
it is stranger than we can think.
Thank you for listening, and hope to see you next time.