🔬Top Black Holes Physicist: GPT5 can do Vibe Physics, here's what I found

Okay, so I think we're at a special time now where at least in some directions AI has become superhuman at least on certain tasks. And that's what led to

certain tasks. And that's what led to these recent papers that resolve a problem that was puzzling physicists experts in the field for over a year.

And they were unable to resolve it and AI was able to do it very quickly. So I

think that's a certain milestone that we've passed by the you guys are bringing attention to this because I I think maybe for the average person on the street who doesn't care about theoretical physics, this is not very

noticeable, but I think it's a very profound change and we've really passed some kind of a threshold. Welcome to the AI for Science podcast uh part of Lean

Space Network. Um I'm Brandon. I uh

Space Network. Um I'm Brandon. I uh

develop RNA therapeutics using AI at Atomic AI. Um I'm joined by my co-host

Atomic AI. Um I'm joined by my co-host RJ Honicky uh CTO and founder of Miro Omix. Um yeah, it's a pleasure to

Omix. Um yeah, it's a pleasure to introduce Alex Lubyansky.

Professor uh Vanderbilt University and fellow at OpenAI. He has uh for a young researcher

OpenAI. He has uh for a young researcher he has quite a storied background.

Amongst other things, he's the uh winner of the 2024 New Horizons Breakthrough Prize. It's the uh call it the Oscars

Prize. It's the uh call it the Oscars for science. I asked ChatGPT is this the

for science. I asked ChatGPT is this the most prestigious award someone of his um career could win and it it recommended a second one called the IUPAP award which

turns out to be it also won. Anyway,

right now he's uh having fun at OpenAI uh doing some really cool research of pushing the foundation of theoretical physics using uh GPT models. A pleasure

to be here.

The one message I wanted to convey is that I think we're on this trajectory which I personally find very surprising and yeah, kind of surreal but also amazing.

Where I would say a little over a year ago AI was very useful for email but not the kind of work that I do that I consider you know important theoretical physics

calculations. I thought, "Oh, that's

calculations. I thought, "Oh, that's special. It's much harder than email and

special. It's much harder than email and AI is not going to be able to do that."

And then there were a series of developments that came in rapid succession that completely changed my mind. And I can

mind. And I can walk you through some of these examples.

Specifically um in particular ChatGPT 03 was the first really strong reasoning model that could do actual math that was useful for my research and could save me a lot of time. That's when I started to really pay attention and use it a lot

more and I thought, "Wow, this is a great tool. I got to get ahead of this

great tool. I got to get ahead of this and learn how to integrate it into my workflow." Then when GPT-5 came out it

workflow." Then when GPT-5 came out it was able to reproduce one of my best papers that took me a very long time to come up with in like 30 minutes. And that's

when I really became AI pilled. I

thought, "Oh my god, this changes everything. It's the most important

everything. It's the most important discovery in my lifetime. It's going to affect everything about how we do research." And frankly, a lot of my

research." And frankly, a lot of my colleagues I would go around telling them, "This crazy thing happened. Pay

attention." And they Yeah, I was getting lots of different reactions but uh I think people weren't quite getting it.

But I talked to OpenAI.

They were also really excited and I thought I have I don't know that much about AI but I have to get in on this and to understand that this is happening and not be a part of it is a huge mistake so

I have to go to OpenAI. So I I I was on sabbatical. It was very easy to to come

sabbatical. It was very easy to to come here and I joined the company.

And then it just kept ramping up even beyond that. Um and you know, to the

beyond that. Um and you know, to the point where now I think most of my senior colleagues in physics are are aware of where things are are headed and they're all getting on board. So yeah.

Yeah, that's an awesome story. Sorry, I

was just saying like I I find it really funny that you that story because it almost remind it reminds me of a lot of different um if you people who had the same realization with Codex

starting sometime last fall especially.

Uh it just took off and a bunch of people are like even like Andrej Karpathy went from oh man, this is, you know, 20% of my work. It's kind of a nice, you know, assistant to oh crap.

What just happened?

Well, yeah, in August actually I remember when GPT-5 came out at that point I was really following AI pretty closely and I think on Twitter the reception was lukewarm. A lot of people like, "Well, we expected a lot more."

and "It's not better at writing email." And

I yeah, I remember thinking, "Well, okay, GPT-3 could write email.

How much better can it get at writing email? That's not the point." But at the

email? That's not the point." But at the science frontier the capabilities were really taking off.

Yeah, there was a lot of attention I think paid even to 03 but yeah, but but then But presumably 5 was yeah, a huge jump.

[clears throat] And I think 5.4 is also a huge jump. I I

don't know if how noticeable it is on the outside although I did hear some I saw some chatter on online. People are

running these independent benchmarks which do show this. Uh so I think it people are are realizing and also anyway in practice researchers are now all over AI using it and yeah, I'm getting

inbounds all the time because I'm the resident scientist doing physics at OpenAI and so everybody is sending me papers, chats like, "Oh my god, this happened." I got one just this week.

happened." I got one just this week.

Somebody said "Codex just wrote up a simulation of the SYK model." This is like a very technical thing in quantum mechanics and gravity and like yeah, a lot of us research groups have been trying to run this simulation and it

couldn't do it and Codex did it in 10 minutes. Just because setting it up was

minutes. Just because setting it up was so hard. Well, I think

so hard. Well, I think it partly it's because of the Venn diagram where you look at the people who have the physics knowledge and the people have the top coding skills and

maybe the overlap is not that that large although I think it's it's been growing.

But I think in this example there are a lot of really good people in physics with coding skills who be trying to simulate these things. So

I think Codex is just really good now.

Okay. Yeah. Um nice. Okay, so I think we're at a special time now where at least in some directions AI has become superhuman at least on

certain tasks and that's what led to these recent papers um which maybe we should talk about that um resolve a problem that was puzzling

physicists for experts in the field for over a year. And they were unable to resolve it and AI was able to do it very quickly. So I think that's a certain

quickly. So I think that's a certain milestone that we've passed and uh I'm glad that you guys are bringing attention to this because I I think maybe for the average person on the street who doesn't care about

theoretical physics, this is not very noticeable but I think it's a very profound change and we've really passed some kind of a threshold. Specifically

focus on the gluon paper in the physics part and we can get to the AI part later.

Okay, so um in physics there are two basic principles of nature that we think every law should respect or every

theory should respect. On the one hand, there's the principle of relativity which at some very high level declares there's an absolute law that cannot be

broken which is that you cannot transmit information faster than the speed of light.

But then there's another principle which is the uncertainty principle that underlies quantum mechanics which says that everything is a little fuzzy.

You know, your position, velocity, there's a little fuzziness to that.

And so you can see immediately at this level of description already there's a tension between these two principles. Uh

because one is an absolute law declaring you cannot go faster than the speed of light and the other one is saying, "Ah, it's a little bit fuzzy."

And this is just to give a sense of how when you try to write down these principles in mathematics the equations don't really play nicely with each other. And so it's been a real struggle

other. And so it's been a real struggle to come up with physical theories that can reconcile simultaneously both principles to describe the physical

world around us. And I would say that the great achievement of 20th century physics which is really one of the greatest triumphs in in human thought as far as I'm concerned is the elaboration

of this framework called quantum field theory which is a a general framework that can describe the physical forces of nature in a way that it accommodates both of

these principles.

And in quantum field theory which is our best theory to date I will say it gets a little bit technical but again try to keep it pretty high level.

What you're trying to compute or describe are the probabilities for certain events to occur. Because you

you're in this quantum mechanical setting, you can't say with certainty what's going to happen when you have a certain experiment but you want to predict probability distributions. And

in quantum mechanics probability distributions are obtained by squaring certain complex quantities and by complex I don't mean complicated, I mean they're not real numbers. They're

they're real plus imaginary numbers which we call quantum amplitudes.

So the goal of a theory is to predict quantum amplitudes which are these objects complex objects that square the quantum probabilities and that's the most you can say about the outcome of an experiment.

And these quantum amplitudes in particular, there's a variety of them called scattering amplitudes which describe the following scenario. Suppose

you have a bunch of particles that you throw at one another. This is what happens in particle colliders like the LHC at CERN in Geneva. You you take a bunch of

in Geneva. You you take a bunch of particles, you smash them together.

Stuff happens. They interact via the physical laws of nature.

Various processes occur and then other particles come out as a result at the end of the interaction. And so

scattering amplitude is the object that describes the the probability for a particular type of attraction. We have some particles coming in with some energies

and momenta and some other particles coming out with other energies and momenta.

And so these scattering amplitudes, they're functions of all the data describing the particles coming in and the particles coming out.

So, in general you can have arbitrarily many particles involved in an interaction.

And this is one of the hallmarks of quantum field theory that particles can be destroyed. So, you don't have the

be destroyed. So, you don't have the same number of particles at the end necessarily as you had in the beginning.

Particles can be created. Lots of things can happen.

And in general, you want to describe all the possibilities. And so you want to

the possibilities. And so you want to have an amplitude for an arbitrary number n of particles.

So, that's called an n-point amplitude.

Because there's n particles coming in and out.

And it turns out in quantum field theory that if you have a particular force and you're able to compute the n-point amplitudes, these functions

of the n parameters of the functions that square to the probabilities then you know everything about the theory.

More or less. There's always an asterisk, but it's basically the the entire content of the theory. So, wait.

If you have a theory that tells you any number of particles come in and go out, then I can say I can declare anything about that system. Exactly. Then you

know everything. And importantly, these amplitudes, they're not just numbers, they're functions because the probabilities that they compute depend

on how much energy do the particles have, what are their momenta, and also a particle has something called a lot of particles like the photon, which is the particle of light has a

polarization. So, when you look at the

polarization. So, when you look at the surface of a lake and you have polarized sunglasses and you turn your head you can see more or less sunlight reflected off of the lake and

that's because a photon, which you can think of as a little particle of light, as it propagates, it carries a little arrow perpendicular to the direction of propagation, which is called the polarization.

And um this polarization has a a direction and sunglasses can selectively let in light with one polarization and not the other.

And this polarization, actually as light travels, it can rotate. It can wind. It

can do its own thing.

And in general, if it uh winds in a right-handed way, so as the particle travels, if the polarization winds to the right we call that a positive helicity

or a right-handed polarization. And if

it winds in the other direction, we call that a left-handed helicity or or negative helicity. So, in general, these

negative helicity. So, in general, these amplitudes, which are the fundamental object in quantum field theory that you want to contain all the information there is to know about physical forces,

these amplitudes depend on not just the energies and momenta, but also the the polarizations. Now, I've told you about

polarizations. Now, I've told you about how there's two basic principles of nature, relativity and quantum mechanics. They come together in this

mechanics. They come together in this framework of quantum field theory. And I

keep talking about forces.

So, there's four fundamental forces of nature.

There's electromagnetism which is responsible for basically the properties of atomic elements in the periodic table and therefore chemistry and biology and

everything that you see, touch, feel pretty much is all due to electromagnetism textures colors.

And this force is mediated by the photon which is the particle of light.

That one is the most familiar to us.

Then there's gravity which is another force that we feel very much because it keeps us to the ground.

And then there's two nuclear forces, the weak and the strong nuclear force which we don't really notice directly in our daily lives, but the weak nuclear force is responsible for radioactive decay and

other such processes. And the strong force, which is the strongest of them all is what binds the nucleus together.

So, you learn in high school that like charges repel.

But if so, then why do protons stick together inside the nucleus of the atom?

They should repel one another. And

indeed, that's the case, but if you bring them really close, then the strong force kicks in and overwhelms the relatively weaker electromagnetic force.

So, the uh strong force is mediated by the exchange of the particles of the strong force, which are called gluons, because they're what glues together the

nucleus of the atom. So, gluons

are the particles of the strong force.

And gravity is mediated by gravitons.

I I think like the gluon paper, I think was sort of the maybe the starting point for this, maybe not, but like the gluon paper had like a really like specific result, right? Like Yeah, absolutely.

result, right? Like Yeah, absolutely.

So, maybe let me just flash the the paper itself. So, we put this on the archive a

itself. So, we put this on the archive a little over a month ago now.

And uh here's the here's the paper.

Let me explain in a few sentences now that I've given a lot of background what Yeah. means. Yeah.

Yeah. means. Yeah.

So, the title says single minus gluon tree amplitudes are non-zero. This might

sound forbidding, but I think we can unpack this Yeah. for the audience.

Um So, gluons are the particles that carry the strong force.

And gluon amplitudes are functions that describe the quantum probabilities for gluons to interact via the strong force.

Now, the word tree here is a little bit of a technicality. It means we're only considering processes where no gluons are created or destroyed.

If gluons are created or destroyed, then you get loops which we can explain later, but this is just a technicality.

So, we're considering special interactions where the same gluons that come in also come out. So, for anyone who's ever fit a

out. So, for anyone who's ever fit a polynomial, you can think of trees being like a linear term and then loops can be higher order terms. Exactly.

Uh way more complicated than that, but conceptually, it's like kind of the lowest order in a series.

And so single minus, now I have to explain that. So, remember I told you

explain that. So, remember I told you earlier how particles have polarizations.

So, when you try to study gluon amplitudes, this is like a whole industry of physics. You know, it's very complicated field. People have written

complicated field. People have written thousands of papers over the decades.

So, you you always want to try to understand the simplest examples first.

That's why you start with the tree amplitudes or the the leading effects and then you worry about the loop corrections. Um

corrections. Um So, you might think that the simplest example to start with is one in which all the particles have the same helicity.

So, say they're all right-handed or that is to say they're all plus helicity particles.

It's been known for a long time that actually in that case, the amplitude is just zero.

Which means uh the interaction is forbidden and cannot happen.

That's one way to think It's just a a symmetry just explicitly forbids this.

It's So, you don't have to worry about the calculate anything. You just know.

Yeah. Just dimensional analysis. It's a

very general argument. Yeah, you don't need to do very much work. And so yeah, it's true that it's the simplest example, but it's so simple that nothing happens. So,

okay, the answer is trivial. You might

ask, what about the next level up? What

about if Oh, no. I want to understand this. You have like a bunch of gluons.

this. You have like a bunch of gluons.

They're coming into an interaction.

They're all in the same helicity. Yeah.

And then you're just saying that just can't happen. Yeah. Okay. Like

because like I can't I take my gluon gun and shoot and he takes his gluon gun and shoot and they go there and then that just can't happen. They'll just go right through each other. Oh oh, so they just They won't interact.

Ah, okay. Yeah. Okay. Yeah. Yeah. Yeah.

Yeah. That's a good clarification. Yeah.

And um now you might ask, what if one of them has the opposite helicity, but all the others have plus helicity, but one of them has

a minus helicity. So, that's what we would call single minus amplitude.

And if you look at the lecture notes and textbooks that have been written on this [clears throat] the same argument that rules out the all plus amplitudes also appears to rule out

the single minus amplitudes. They're too

simple. They can't really interact.

Nothing to see here.

Move on.

So then you might ask, okay, well, what about the next thing where there's two particles that are minus helicity and all the others if So, if there's n of them, there's n minus two

others that have positive helicity.

All right. So, these would be double minus amplitudes.

And people in the '80s studied and computed these amplitudes. They're

not zero.

And in particular there were two physicists, Parke and Taylor who found this beautiful result.

They did a lot of really hard work and computed these amplitudes. Very

technical, difficult calculation.

But at the end, you get all these terms and you have to sum them all up. And

almost all of them cancel.

And at the end, you're you're left with this very simple formula that fits in half a line.

Which is now known as the Parke-Taylor formula for these amplitudes.

And um these amplitudes are now called MHV amplitudes, which stands for maximally helicity violating because

they have the largest or so we thought possible symmetry between the plus and the minus helicity particles. That the most asymmetry. Now,

particles. That the most asymmetry. Now,

let's get to this paper which came out last month. So, this is a paper written

last month. So, this is a paper written with Alfredo Guevara, who's a post doc at the Institute for Advanced Study, David Skinner, a professor at Cambridge University

Andrew Strominger, a professor at Harvard, used to be my advisor.

And also Kevin Wheel who studied as a particle physicist in a previous life.

So, how how did this happen? Well, maybe

we'll get into how I ended up in OpenAI a little bit later, but I ended up in OpenAI started to improve the model's abilities to do physics.

The models got really really good at physics and I thought, "Okay, it's so good now we should try to solve some actual research problems at the frontier."

frontier." And I called up Andy who used to be my advisor. I said, "Hey Andy, do you want

advisor. I said, "Hey Andy, do you want to come here to the SF, visit OpenAI and we can try to solve one of your problems in physics."

in physics." And I thought, "You know, it's probably not going to work, but if it doesn't work, at least we'll figure out why it doesn't work and you know, I can do this with a different

physicist every month and eventually something will work and in the meantime we'll learn how to improve the models.

So, it's all fun and useful."

And so, Andy was the first one that I invited to do this.

And he said, "Well, I have this perfect problem that I've been thinking about with Alfredo and David for the past year."

I'll explain the the problem, but the amazing thing is that we decided to start working on it using AI a little bit before Andy was scheduled to come, like the week before. And in

fact, using ChatGPT we we solved the problem before he even got off the plane.

[laughter] Which was a huge surprise to him.

To him. Yeah, I think and to me, to be honest. We I had not expected that.

honest. We I had not expected that.

And it's really cool story. So,

Andy, David and Alfredo understood a year ago that this statement that the the single minus amplitudes, the statement that they're zero

is not exactly correct because the usual argument in the lecture notes and textbooks has a loophole.

And the loophole is that it assumes that the particles are coming from generic directions.

But in a certain regime where the particles are exactly aligned with one another, we say they're collinear, then the usual argument has a loophole and it's possible for the amplitudes to

not be zero.

But then if they're not zero, what are they? So, then suddenly these really simple amplitudes previously thought to be zero, if they're not zero, we should compute them and they should be something really nice and simple and

special.

Now I'm burying a lot of I'm sweeping a lot of details under the rug here, this this has to work in some different signature space time. You know, it it connects to lots of other things they've

been worrying about. We're not going to worry about this.

I mean, I was actually hoping at the end that we could talk about what it means to be two dimensions in space and two dimensions in time, but yeah, I mean, I think like part of this is doable. The

loophole is one about, you know, the alignment of the particles, but it's also a loophole about the space time of physics that universe we're living in and this is not really mind-bending stuff.

So, um they understood that they're not zero and they started to compute them. And

Alfredo is really, I think, the unsung hero of this story because he did a lot of really hard work to compute these things by hand. And

I'll just show you an example. So, we in the paper there's a lot of formalism.

So, here is the beginning of the definition of the general answer.

What? [laughter]

Yeah, it's it's pretty hard to unpack, but it starts here. Then you have to define these vertices objects V and they're complicated, involves spinor

theta functions of spinners and then you have this recursive formula. But okay,

it's a whole mess.

And concretely, if you try to unpack this definition, remember these amplitudes are a function of the number of particles involved. So,

there's a three-point amplitude where there's only three gluons in the interaction. And you know, the answer is

interaction. And you know, the answer is pretty simple. This is some function

pretty simple. This is some function that we've defined here. Not that

complicated. Then this is the four-point amplitude where now there's four particles and you can see that we go from one term to um a sum of two terms here.

But then once you get to five particles, you get a you start to get a lot more terms. There's eight of them being summed here.

And by the time you get to six terms, it explodes in your in your face. For

those people not watching this on YouTube and listening, this equation takes up a quarter of the page, is 32 terms, each of which is a product of four terms, each of which is itself

encapsulating a rather complicated formula. Yeah. So, this is super nasty

formula. Yeah. So, this is super nasty and uh that's as far as Alfredo got or Andy whatever. So, so Alfredo did it

Andy whatever. So, so Alfredo did it Look, is this just an expansion of some sort of How hard is it to do this expansion?

Very hard.

Okay. Yeah. Uh

And there's a nice graphical way to understand this in terms of Feynman diagrams. I hadn't planned to explain this, but uh there's a visual This is kind of a

visual subject. So, the math is very

visual subject. So, the math is very complicated and already back in the '40s, Richard Feynman, who's one of the pioneers of quantum field theory, came up with this very visual way to

organize our understanding of the subject. So, you can doodle these little

subject. So, you can doodle these little cartoons that represent possible interactions. And the rules of quantum

interactions. And the rules of quantum mechanics actually say that in these amplitudes where you scatter a bunch of particles, you get to fix what comes in and what comes out because that that's the question you're asking, what's the

probability for a certain interaction.

But then everything that happens in between, you don't get to choose that because the physical laws determine what happens.

And actually in quantum mechanics, you're supposed to consider all the possibilities, all the ways in which the incoming particles can interact and transform into the outgoing particles.

And you're supposed to average or sum over all the possibilities to get the final amplitude for the process as a a sum over the amplitudes for each

individual possibility for how you could get there. So, just to be clear, there's

get there. So, just to be clear, there's a There's incoming particles. They

interact. Yeah. And then there's all these different They each have their own amplitudes and then you you it's sort of like I select for this one one possibility and this one one possibility

and then I get like one possible interaction and then there's an infinite number of those for each and then I sum those infinite I suppose and I get the out Yeah. So, in principle, there are

out Yeah. So, in principle, there are infinitely many pictures to sum over.

But that's why we organize them by how complex they are. And it turns out that every time you get an interaction, every time there's a a vertex where two lines meet,

um that point interaction comes with a power of the coupling constant, which controls the strength of the interaction. And it turns out that

the interaction. And it turns out that every additional interaction makes the amplitude more suppressed. So, it it's contributes less to the final answer.

And so, you want to first consider the diagrams with the fewest possible number of interactions because they will give you most of the total final amplitude.

And then if you're trying to get a more and more refined answer, you then consider the more and more complicated cartoons with more and more interactions. And in fact, this is one

interactions. And in fact, this is one of the ways in in which the diagrams can get complicated is that they can have loops. So, for instance, here you have a particle that decays

into two particles, creating this loop because then they meet up again and disappear. So, in this interaction, you

disappear. So, in this interaction, you have intermediate particles being created and destroyed.

But whenever that happens, you get two extra vertices in your graph. So, these

diagrams are suppressed because it's less likely to happen that you get these extra felicitous interactions.

And so, you have to you don't need to worry about this as much. It's like a small correction. And of course, in

small correction. And of course, in principle, you could keep going, but you never you're never done except in very special circumstances. Or higher order

special circumstances. Or higher order powers in a polynomial or something. Or

a Taylor series. And so, to go back to the story back in the '80s with the MHV amplitudes, which I think now is a bit of a misnomer. I would call them double minus amplitude because they that's where we're going to get to in a second.

Right.

There was this heroic calculation where a lot of Feynman diagrams were summed and they were considering more and more interactions with more and more particles and every time there were more and more terms, but they

all canceled to the end always give a simple answer.

And in fact, that's what this PT term PT stands for Parke-Taylor.

These formulas are you know, they fit in a line. So, it's not that complicated,

a line. So, it's not that complicated, but it's very surprising that such a messy calculation at the end would clean up into such a simple result. And so,

what Alfredo, Andy and David did was to understand that these single minus amplitudes in the special case where some of the particles are aligned, they don't have

to be zero and then you can do this very complicated Feynman diagram expansion to get the answer, which is not zero, but the problem is if you do it this way,

well, you can represent the answer in some horrendous horrendously messy complicated way, but if you unpack it, it's extremely complicated. It's complex

in the following sense. When you

consider the end point amplitude, so the probability of n particles interacting, the number of terms in your answer, which correspond to the number of diagrams roughly that you have

to add up, it grows factorially in n, the number of particles, and factorial growth is really bad. It's super

exponential. It goes faster than an exponential. So, it blows up in your

exponential. So, it blows up in your face. This is what you're seeing here.

face. This is what you're seeing here.

And that's because roughly you have to draw all the possible cartoons and the possible combinations is a combinatorial problem and that's where the factorial behavior comes from.

But we know from the '80s that in the actually more complicated double minus case, Parke and Taylor found this miraculous simplification. And so, Andy,

miraculous simplification. And so, Andy, Alfredo and David spent the last year chasing the analog of the Parke-Taylor formula, the very simple answer that was obtained in the '80s for the double minus

amplitudes.

But now for these single minus amplitudes which they understood are not zero, but then what what are they? And

they they were getting this really complicated answer. And okay, you never

complicated answer. And okay, you never know in physics ahead of time if something will simplify. You have to believe in it to find the simplification. But because the double

simplification. But because the double minus one simplified, it felt like these should simplify too. And

we think they're important for lots of things and that these are somehow really important objects that are very fundamental. And so they should have a nice description. And so

they spent a year looking for that.

There's a funny The next line if you scroll down is something like, "We we need a simpler formula." Right.

We need a more concise formula is needed. Yeah, a more concise formula is

needed. Yeah, a more concise formula is is needed. And um

is needed. And um this is where AI comes in.

Because when I asked Andy, "Hey, do you have a a problem in your pocket that we should use AI to target?" He said, "Well, I have just the perfect thing for you."

you." We've been puzzling about this. It's

really important. It's really

interesting. It connects to all these things. And we don't know the answer.

things. And we don't know the answer.

Yeah, I mean like when I was a grad student and if I had approached something like this, I would have probably plugged it into a computer algebra system, chugged along, tried a few limiting cases, see if there's any

any like magical, you know, simplifications which happened. Um this

type of thing is, you know, something that oftentimes you see this thing like, "We need a different approach."

Exactly. Then

um before Eddie even got here, we started to play with ChatGPT and Alfredo, Andy, and I were trying different things. Lots

of different chats happening.

Going back and forth. David as well.

And the first thing that happened is that we fed the five-point amplitude into ChatGPT.

And we're like, "Can you simplify this?"

And it was like, "You know, there's a special region.

So there's an extra assumption that you can make in which this answer simplifies to this one."

So this assumption is equivalent to the you have one particle coming in and it decays into n minus one of That's one way to think about it roughly.

Okay. But we're in two time dimensions, so Yeah, it's it's complicated. But basically, you can look

complicated. But basically, you can look at what we call face space. It's the

entire space of possibilities for all the energies of incoming particles and their momenta. And there's a special

their momenta. And there's a special region in that face space where one particle has one different sign of its frequency compared to the others. And in

that region, there's a big simplification that happens that ChatGPT found. And I should say it's This was

found. And I should say it's This was the public model, but the pro version that thinks really hard. So that was that a a known fact that it it just was able to relate to the problem? Or was that

something that it put together?

With As far as I know, it put that together. Okay. It said, you know, "This

together. Okay. It said, you know, "This five-point function which is a sum of eight terms, each one of which is a product of three terms. They're all pretty complicated." It said, "Hey, like

pretty complicated." It said, "Hey, like actually this simplifies to this product of only three terms." And we stared at this thing and thought, "Wow, that's really nice. We didn't know this." It's actually In hindsight, once

this." It's actually In hindsight, once you know, you can re-derive this, but it takes a while to understand where this comes from. So I think that was a a leap

comes from. So I think that was a a leap of insight that the AI had. And I think what it did I mean, at some point said, "I wrote a Python code and I ran through

all 5,000 possibilities and I I deduced that this So it's the equivalent of running his computer algebra system, but it But it just decided to do it on its own and came up with a huge

simplification. So great. Yeah, awesome.

simplification. So great. Yeah, awesome.

This was after making the assumption?

This is after the decay one particle decay assumption.

Yeah, so it made the it figured out There was a lot of exchange. This is

very experimental. We're talking about it a lot. It figured out there's some four region in which things simplified in that region.

It said, "Okay, this thing simplifies Now ChatGPT came up with that simplification as well. Yeah, all of them. Yeah. Yeah. And then we were like,

them. Yeah. Yeah. And then we were like, "Okay, well, let's give it the six-point function which Alfredo heroically computed." And by we didn't have the

computed." And by we didn't have the seven-point function. We

seven-point function. We I don't think anybody could use the heat identity to expand it. It would be disgusting. And then ChatGPT does its

disgusting. And then ChatGPT does its little thing and then it's like, "Hop.

Yep, simplifies to this." And we thought, "Whoa, okay. That is really nice." Because now instead of 32 terms,

nice." Because now instead of 32 terms, it reduces to just four terms. And it's not a sum of 32 terms. It's a product of only four terms. And then we asked ChatGPT, "Okay,

well, can you guess the general formula for all n?"

And that step by the way, I mean, you could imagine using some programming language or or symbolic manipulation software to do these reductions in certain examples. But to

tackle the general case, I don't know how to use a computer to do that. But

ChatGPT said, "Yeah, this is the answer in the general case. Boom." How long does that take?

You know, it's like using pro, it thinks for 20 minutes at a time. You go back.

But it wasn't like six days or something. No, no, no. It's just like

something. No, no, no. It's just like over the course of of several interactions. So And

interactions. So And the amazing thing is that the formula that it proposed, instead of having this factorial growth which is super exponential where the number of terms as

you consider n the number n of an increasing number n of particles, the number of terms blows up.

Here it's actually linear.

So if you double the number of particles, you only double the number of terms. It's the nicest possible behavior you could imagine. This is the equivalent, I think, of the Parke-Taylor formula for the double minus amplitudes

that was known back in the '80s.

But now for the single minus amplitudes. And this was guessed by GPT. I think it was 5.2 at the time. GPT 5.2 pro.

the time. GPT 5.2 pro.

But it it couldn't quite derive it. So I said, "Hmm, looks like

derive it. So I said, "Hmm, looks like this, but I don't know how to prove that."

that." Yeah, I think the model was not quite strong enough to prove it. But part of my work at OpenAI has been to develop stronger physics capabilities in the

models. And a lot of people have been

models. And a lot of people have been adding lots of, you know, it's not just my singular contributions. There's a lot of great research happening and it all comes together. You know, it takes a

comes together. You know, it takes a village. But we had this uh

village. But we had this uh internal model that could think for a very long time and was extra strong at physics. So we gave it the whole problem from scratch without

actually giving it this.

We just formulated the problem in a very sharp way and asked the model to solve to find the answer for the amplitude in the general case in this region. Because now we identified that

region. Because now we identified that this was the special place to look.

And it took 12 hours, which is a long time. But it came back with the same

time. But it came back with the same formula which we had not given it. So it

rediscovered the correct formula. But

this time it also found the proof that the formula is correct, the derived it.

And in fact, the remainder of the paper after we state the equation is devoted to the proof that is basically what came out of the AI. So we say the rest of this work is devoted to proving that the

conjecture is correct. There's three

steps. First, you show this. Second, you

show blah. And third, you show blah. And

then the this is basically what the AI came up with.

So now I can finally summarize the paper. The title is single minus gluon

paper. The title is single minus gluon tree amplitudes are non-zero.

So these special interactions between gluons where only one of them has a different helicity from the others, which were previously thought to never occur, actually these interactions can happen.

So the amplitudes are non-zero. That's

the main claim of the paper. I think

it's quite surprising. I think it's a like a really nice paper.

And the final result, I guess there's two results. One is understanding that

two results. One is understanding that it's not zero. That came from the humans like a year ago. But they were trying really really hard to find a simple answer for what the amplitude is.

And they were kind of stumped for a year.

They were able to get this indirect representation that's extremely complicated in terms of Feynman diagrams. But they were looking for the simple formula that is the analog of of this Parke-Taylor formula for the '80s

for the more complicated amplitudes. And

that was done with the AI. And so I I think that's a really interesting result. Yeah, totally changes the way

result. Yeah, totally changes the way you should think about where we are in physics and how AI is going to change that. You know, it's not It's a result

that. You know, it's not It's a result that top researchers in this field were thinking about for a year and then the AI solved it. So I think it's interesting. There's several things

interesting. There's several things about the story which I think people didn't understand on Twitter. If maybe

you scroll down to like equation 38 to what is it? 35 to 38. Yeah, like so I would say most even intro grad students would look at 35 to 38 and say 39 is

actually a very natural extension of this. Like that is I don't think, you

this. Like that is I don't think, you know, that surprising. I think it's interesting. I didn't know until just

interesting. I didn't know until just now that [clears throat] um you can that when you proved 39, that

was a fresh session. That was without the limiting cases. You started from scratch.

Yes. Yeah, I did it that way.

Uh because I guess it's an extra way to be confident in the answer. If the if different model independently comes up with it from scratch, then

you're not just feeding spoon-feeding the answer that you think is correct.

That's an extra confirmation. Um but

yeah, I think we thought a lot about how to put this out into the world.

And there's no perfect way to do this. I

you know, clearly we could have done a better job of communicating it. One

thing that was important to us is to not make this paper about AI. Because I

think this is a really interesting physics result. You know, people will

physics result. You know, people will keep reading this paper, I hope, for a long time.

We didn't put AI in the abstract because this is a physics result that stands on its own.

There's one paragraph really about AI where we just say the the final formula was first conjectured by GPT 5.2 pro and then proved by an internal OpenAI model.

Because, you know, that's what happened.

It's true. But we didn't really want to get into it because I think that's not the point of the paper. I

mean, it's really interesting how it happened, but the result stands on its own. And I think if you read a paper

own. And I think if you read a paper today that was written 20 years ago that used a computer to do some critical step in the argument, and it had this whole discussion of how,

"Well, I loaded MS-DOS 3.1 and it had five floppy disks and I had to swap my floppy disk after you wouldn't care, you know? You That's

not why you're reading the physics paper today. So, we didn't really want to go

today. So, we didn't really want to go into that in the paper.

And we talked a little bit about it in the the blog post that we released with the OpenAI, which is this one.

And then, I guess on Twitter there were a lot of questions and I I wrote some tweets that I think clarified it. And

there was somebody who is a physicist who wrote a great blog post like actually understanding the story.

And The Economist also put out a great article about it, which they really understood what happened. And I thought it was a great great coverage. Uh,

Science magazine also wrote about it.

Harvard, the Institute for Advanced Study put out press releases. So, I

think it it got a lot of attention, but it's kind of a subtle thing to explain. It took us an hour to go

to explain. It took us an hour to go through what happened and what what was done. So, you know, it's it's hard to

done. So, you know, it's it's hard to explain. I think it would have been kind

explain. I think it would have been kind of a distraction from the physics point of the paper to go into that. Okay,

let's talk about the physics then. Give

us a sense cuz, you know, my theoretical physics on the frontier comes from PBS Spacetime, right? Like I'm you know,

Spacetime, right? Like I'm you know, it's a great channel.

Yeah, everything that's channel, but it gives you a great high-level picture, but hard to know how this sits in the pantheon of papers that represent the

cutting edge of uh, theoretical physics.

Not exactly that. I I want to just understand like it seems like you're comparing it to this previous result that is pretty significant and like, you know, like highly cited and very

important. How does this compare to

important. How does this compare to that? Okay, you're putting me in a bit

that? Okay, you're putting me in a bit of a tough spot. I I I will say I think the result is surprising.

That's why the title is what it is, you know? Single minus amplitudes are

know? Single minus amplitudes are non-zero. And if you if you're somebody

non-zero. And if you if you're somebody who works in this field, that should catch your attention.

Ultimately, it's very hard to know in science when you release something into the world how it's going to be received and how impactful it will be. I

think the true value of a paper can only be assessed decades there is into the future based on how much future work it leads to and

what developments it it opens up. So,

let me maybe a better way of asking is, so my understanding is that previous paper kind of opened up a whole line of thinking about Yeah, I I think this is a great segue to the second paper that

came out just three weeks later. Perfect

then, let's talk about So, it got its own blog post. This is

March 4th, so I I guess two weeks ago now.

So, we're talking earlier about how there's four forces, strong force mediated by gluons, and then gravity that's mediated by gravitons,

except gluons we can produce at the LAC and we can measure their effects fairly directly.

Gravitons, we think, are also around us, being produced all the time, even as I move my hands, but we've never done an experiment that directly measures gravitons. Um,

gravitons. Um, but they're supposed to be the quantum of gravity. So, they're really

of gravity. So, they're really interesting from a theoretical standpoint. And so, going back to RJ's

standpoint. And so, going back to RJ's question earlier, what is a graviton?

There's different answers we could give.

Ultimately, the correct answer depends on what the theory of quantum gravity, which we don't know yet.

Yeah.

Um, if you just naively try to take all of your tricks from field theory that we know from the standard model, apply it to gravity, things just break down. The theory is not self-consistent.

down. The theory is not self-consistent.

There is some definition various problems. Yeah.

Just like if you took in this room there's light flowing around, there's some indivisible bit of light that you at some point can't break up into smaller bits. That's the quantum of

light. We call that the photon.

light. We call that the photon.

And the the gravitational force is mediated by the exchange of gravitational force or gravitational waves. If you try to take a

a gravitational wave and break it up into smaller and smaller pieces, at some point you get a quantum that you can't break up anymore, and that would be the graviton.

That's how we understand them. Okay, so

there's the idea being that like you can't you get to a certain point and you can't have less gravity than that. You either

have some or none.

Right, that's one way to think Yeah.

And um, so we wrote this paper which is called single minus graviton tree amplitudes are non-zero. So, it's

the same title almost, except with graviton instead of of gluon.

Uh, and that's on purpose cuz we we wanted to uh, extend the result.

And it's the same story in the sense that it was thought that all single minus amplitudes are zero, but actually it's not true in also for for gravity.

But, gravity is a lot more complicated.

So, now if you want to compute the graviton amplitudes, it's potentially a lot harder. Do gravitons have phase the

lot harder. Do gravitons have phase the same way that the gluons do? Is that it? So, they they actually have spin two rather than spin one. It's getting into

the weeds. So, the amount of the numbers

the weeds. So, the amount of the numbers you have to use to describe them are a little bit different. They're doubled in some sense. Okay. So, their polarization

some sense. Okay. So, their polarization is more complicated.

I see.

But, this is really getting into the weeds. But, the special region in which

weeds. But, the special region in which the final answer simplifies has two labels because it's a spin two particle, whereas in the gluon case there was only one label because it was a spin one particle. So, this is like So, it's not

particle. So, this is like So, it's not the same math. Gluons and gravitons do have some spiritual similarities compared to other types of particles.

Well, yes, in the sense of they're particles of force.

Yeah, yeah. But, they're like sort of doubled. Yeah. They're sort of sort of

doubled. Yeah. They're sort of sort of doubled, yeah. Yeah, I mean, okay, this

doubled, yeah. Yeah, I mean, okay, this is I guess the the people watching this podcast probably like to geek out on this. So,

the modern definition of a particle in quantum field theory, which is our best verified framework for nature, is that particles are irreducible representations of the Poincaré group.

We just lost 90% of our audience right now. Yeah, okay. What do we Maybe we cut

now. Yeah, okay. What do we Maybe we cut this. So, there's mathematical

this. So, there's mathematical representations and um, they've all been classified. All the possibilities are

classified. All the possibilities are known by Wigner, actually, a brilliant physicist. And it turns out that the

physicist. And it turns out that the representations of possible particles are completely labeled by the mass and the spin and the charge of the particle. So,

these are the three quantum numbers.

And uh, particles of long-range forces like gravity and electromagnetism have zero mass.

They have to have integer spin.

And spin one is three of the four forces, and spin two is gravity.

And then, that's it. But, let's set that aside. The really cool thing about this

aside. The really cool thing about this paper is that well, first of all, it came out three weeks after the first one, which is really fast. And I think this is a great example of AI

accelerating science.

And in fact, we could have put this paper out three days after the first one because that's that's how fast we got the answer out of ChatGPT.

But, it took us three weeks because we wanted to check very carefully that it's correct. But, most of the time was

it's correct. But, most of the time was spent verifying the answer, not writing, which is insane, actually, if you take a step back. It it shoots If you told me a

step back. It it shoots If you told me a year ago, "Yeah, like you're going to have this AI that just does really hard calculations for you, and then most of the human effort goes to verifying the answer." I

would thought that, you know, you're crazy. So,

crazy. So, it's very surreal. And then, we also had to write it up as a nice paper, which you know, put in the citations and references. That takes some time. I also

references. That takes some time. I also

had a baby in the meantime. So,

[laughter] we said lost time there. But, we did this really fast. So, I think it's an example of accelerating science. Another

really cool thing is that for this paper, we didn't have to use an internal OpenAI model that had to think for hours. This

was all done using the publicly available GPT Pro.

In fact, we shared one of the main prompts that we used. It's If you go to the blog post extending single minus amplitudes to gravitons, and you scroll down to the text, there's a link

to one of the chats that we used. So,

you can see we used ChatGPT 5.2 Pro.

And the amazing thing about this is that we gave it the gluon paper as a seed.

And we said, "Read and understand the paper. Make sure you understand the

paper. Make sure you understand the manipulations in the appendices cuz that's where most of the hard work goes."

goes." And but it comes back and it says, "Yep, um, I understood the paper. Let me focus on the appendices. Here's what happened."

the appendices. Here's what happened."

And basically, the punchline is that GPT Pro with the gluon paper as an anchor was able to do the graviton calculation, which is really different mathematically, completely on its own

from Well, not I guess not from scratch, but from the gluon paper, but it's it's just a different thing. And it was strong enough to do it completely. So, it took this paper took the conceptual leap from

the the previous paper and and just said, "Okay, what math do I need to make that same conceptual Yeah, and it's different math. That's an

important thing to emphasize. So, in

particular, there's a crucial application of something called the directed matrix tree theorem. And

Alfredo and David, we've been thinking about these things for a very long time.

We're like, "Whoa, that's really cool.

That's surprising. We hadn't thought of that or or seen that before." That was like known math, but it it was because maybe it has such broad understanding of math and physics that it's able to say,

"Oh, this is what this is a good thing to apply in this case." Yeah, exactly.

And so, [clears throat] here it understood the paper, the gluon one, and then we said, "Okay, well, the task is to generalize this paper to the gravity case."

Here's some Here are two key changes, but otherwise manipulations should be similar. So, we tweak some things at the

similar. So, we tweak some things at the get-go. And then we said, "Good luck.

get-go. And then we said, "Good luck.

You're a brilliant theoretical physicist." So, it's like a you know, we

physicist." So, it's like a you know, we give it two paragraphs. So, we give it the Coulomb paper a couple paragraphs and say, "Good luck." Thought for 20 minutes.

And boom, it starts to think. So, it

starts at the beginning, it works through the implications. So, like

really interesting stuff. And then it says, "Here's what I would do next to turn this into the gravity paper."

Uh if you want, I can do blah. And so,

we say, "Yeah, go ahead."

[laughter] So, Another thought for 31 minutes.

Thought for 31 minutes. Yeah, so this this exchange is 110 pages.

But, I think it's hilarious. I would

describe this as vibe physics.

Because you can see it scenario goes away, does a lot of hard work. Goes lots

of equations. It's starting to do the Okay, so now you have to use this different math, you have to use these tree calculations, LLC reduction formulas. Okay, there's a lot of stuff

formulas. Okay, there's a lot of stuff happening sum over trees, concrete checks. It's starting to Yeah, well,

checks. It's starting to Yeah, well, this is one of the things I love is that it's able to uh do the same things that a human would do, which is check some basic cases as a sanity check and to get intuition.

And so, it comes back every 3 minutes and says, "Well, here's what remains to to finish the full gravity paper."

And there's a list. If you want, I can write the gravity analog. Yes.

Do that. This is the first step. Okay,

it goes back, thinks for 34 minutes.

Half collinear support, starts to do stuff. Okay, this these formulas

stuff. Okay, this these formulas actually made it in the paper in some form. This is all correct.

form. This is all correct.

There's a bunch of stuff. At the end it says, "If you want, the most the next most useful thing I can do is do this." And

we're like, "Yeah, verify this by performing the explicit check."

And it goes on and just to cut to the end, finally just we we say, "Okay, write up the paper."

the paper." And you can see the paper that it writes.

And it's very close to the final thing that we actually put on the archive. So, did it make suggestions that were not what you would have suggested as the

next steps?

It's very smart. It knows kind of where to go.

It's useful to steer it.

If you compare what it came up with with the actual paper that we put in, the intro, the abstract and introduction were written by Andy, who's an amazing

writer. And I think he gave this wider

writer. And I think he gave this wider perspective on the problem and how it fits into physics and how it connects to other things that, you know, the AI didn't do it. Just the intro it wrote was

more generic. But, okay, AI could write

more generic. But, okay, AI could write really well. We didn't really try to

really well. We didn't really try to make it. Yeah. And the other thing is

make it. Yeah. And the other thing is we added a section, this section two, which was not part of that initial exchange, is about how these graviton

amplitudes transform under certain symmetries of physics. And that's

something that we're really really interested in because we eventually want to understand quantum gravity as I mentioned earlier. And

typically the first step to uncovering a new theory is to understand what are its symmetries. That's something that gives

symmetries. That's something that gives you some kind of ground to stand on. And

in particular has been pushing this program of celestial holography, which is like a whole thing we could get into, but it's an exploration of the symmetries of quantum gravity.

And he really wanted to understand this.

And there's a separate chat. We didn't

share that one.

Where we led the AI to explaining how these answers fit into the symmetries that we know the theory should have. And

that's something that went in there.

But, actually I think from section three onwards, it's pretty much very close to to what the AI wrote. So,

I would say this is really remarkable.

It's a real solid result in quantum gravity that was done pretty much completely by an AI with human steering it and asking

kind of the right questions. But, all

the math was derived by by ChatGPT Pro, the public model you can access.

And most of the time spent was by by us the humans was like checking everything and writing it up.

And that's really wild. I mean, we're really [laughter] So, I mean, that that you as a physicist, you find yourself where a lot of coders have found themselves, where

there's a kind of a fundamental, maybe epistemological question here that if now as a physicist, like I could have done that, right? Like maybe maybe I like I needed a little more background,

but you like that a lot of it was like, "Yeah, go ahead." Right? Like take this paper, give it some prompt. You guys

obviously prompted very well, but there wasn't like maybe a undergraduate in physics could have come up with a lot of that. And so, the question is

that. And so, the question is how does the undergraduate in physics now learn when they don't have to do the hard calculation themselves? Similar to

that How does the undergraduate coder I see [clears throat] You're You're opening up many different strands of conversation, which are all super interesting. So, let's try to

super interesting. So, let's try to unpack that a little bit.

So, the most direct thing you asked is how does the next generation learn?

Yeah.

That's a really good question.

I think about this a lot. And now

that a lot of senior physicists in the field are coming to grips with these new capabilities, one of the questions that comes up very quickly is, "How do we train the next generation?"

Because the way we were trained is by going through these, you know, these these difficult rites of passage where you have to do these really arduous calculations and this how you build confidence in your

own abilities and check test your knowledge. And it's not just about what

knowledge. And it's not just about what you're capable of doing, it's about knowing that you're capable of doing it and proving it to to yourself and building that self-confidence it is a important. And we don't have a good

important. And we don't have a good answer. This is something that

answer. This is something that academia is going to have to grapple with. So, one thing that is especially

with. So, one thing that is especially difficult is that as a professor, I have graduate students.

And the the gap between where classes take you, even graduate courses, they only go so far. They go very far, but only so far.

far. They go very far, but only so far.

And the gap between where that ends and research begins is actually huge and it's it's growing wider.

Usually as a professor, what you do is when you take on new students, you keep in your pocket a few easy problems in the sense that you know they're going to work.

Some questions that you you know in principle you could work out, not that difficult.

But, you give them to a student so that they go through the exercise of learning everything around the question, developing the technology. And then you know enough about the problem that you're sure there's an answer that you

can get there and you can advise the student in the process of discovering it.

And I think the issue is that many such problems now, I would say these models can probably crush.

These are problems that we usually take again, you know, time scale for a theoretical physics paper is 6 months to a year. That's pretty typical. So, if

a year. That's pretty typical. So, if

you tell a student, "Go away and think for 6 months about this one question.

And you have to work really hard at learn a lot of stuff around it and do lots of calculations."

Even the most determined students, would they not within the course of 6 months ever ask ChatGPT Help. Yeah, that's a little bit weird.

Help. Yeah, that's a little bit weird.

Now, it's also an opportunity because I remember that time in my graduate school career, in my second year of grad school, I took all my graduate courses my first year.

And then my second year was my first project.

And it was actually the hardest time for me in graduate school to traverse the desert from where classes take you to the research frontier. It's very hard.

And there's a lot of time spent banging your head against the wall. Like all the time you're confused, you don't understand things just cuz you need to absorb so much knowledge.

And AI can totally help you with that.

It's the best teacher and knows everything. It can unpack

everything. It can unpack any complicated fact to any desired uh level of detail.

Actually, my experience as a trained professional physicist working on my own research using GPT now is that I would say There's two key ways in which my research is is completely changed. One

is that I spend much less time being confused.

So, I'll do a calculation, get an answer, and I think, "Huh, how does this fit in with this other fact that I know? Like how do I reconcile these things in my mind? I'm

confused."

Yeah, I do that all the time. Yeah, in

research usually you you take a step, then you're like, "Ooh, you hit a roadblock, an obstacle, you're confused." And you have to think for a

confused." And you have to think for a few days. Maybe you go for a walk or you

few days. Maybe you go for a walk or you know, work on another project, come back, get a new idea. But, you spend a lot of time confused. That's the nature of

time confused. That's the nature of research. With GPT, I'm like, "Hey, I

research. With GPT, I'm like, "Hey, I just did this. I found this. How does

this mesh with this other thing?" And

then it's like, "Oh, well, you forgot this thing." Or "Oh, you didn't quite

this thing." Or "Oh, you didn't quite think about it correctly." Or "This is the standard fact." You know, and so the amount of time you spend confused just dramatically shrinks and you move so much faster. That's one of the

much faster. That's one of the accelerating effects.

The other accelerating effect is that, you know, I only have so much free time and energy.

Especially when you become a professor, you have to teach, you have students, you have grants to administer. There's a

lot of things you have to do. So, your

your free time to do think about research without distractions shrinks.

And also, uh you know, you only have so much energy to do hard calculations. And

so, what you would usually do is if you have a problem, you you know, you're at point A and you want to get to point C, you think about the route. Oh, I have to go through point B first. And actually

maybe there are multiple points and you try to plot in your mind the course that you're going to take before you actually go start do the hard work. You try to think really hard about where you're going and and to chart a course. With

AI, actually, you can launch 10 instances of chat and have each one try a different route and send it as a scout that moves very

fast into the unknown pushing outwards and you can just very quickly get some feedback to see okay these approaches are not promising these are much more promising and then if you follow them

there's a huge difference between being the first to push into the unknown versus following someone ahead of you and even if chat GPT doesn't always get everything right it just kind of having a scout that

signposts some key steps along the way that you can use to to anchor your own movement is is extremely helpful so I this is there are like two concrete ways

that AI has changed the way I work and I think if you're entering research having an assistant that can help you find your way to where you're trying to go

can be very good so I think it's it's inevitably going to change how we work how we operate and how we train students and you know part of what's exciting about my job is trying to figure

how all of this works but it's not just a job for open AI it's actually a job for every researcher and professor more generally to think about this I think the future is very bright because

we have some challenges to overcome but on balance this is such an amazing tool I think it's going to give human physicists AI superpowers because of what I just described you can do so

much more and I think actually um the kind of uh skill that is really useful to get great results out of AI is very similar to the

kind of skill that you develop as an academic collaborating with other humans this is like a collaborator and if you're a professor who's been advising students and postdocs you know a lot of

what being a professor involves is knowing for each student postdoc that you're working with exactly what question to give them so matching the problem to the person

and knowing how to give them the question in what way with how much detail what what level of detail not too much not too little and that's actually kind of how what you have to think about when you interact

with chat GPT so I think that it's a transferable skill and um uh people who are good at this are about to get AI superpowers what you just described there reminds me of several

conversations we've had on the podcast this far which keep coming up to this concept of taste um one of the things that especially you say theoretical

physics um high energy physics has maybe had a problem with I'm not sure if you want to describe it that way but it is it can be very um trendy that like there's certain things

which become in fashion because maybe right now we're in a world where we don't have the data to define new directions uh to really guide or constrain where we're going I'm curious

like how does essentially something which is superhuman and that it has basically all known physics and it's interact with a field where

at its core what oftentimes can be popular people start working on is more based upon general aesthetics or what you know the community collectively thinks is cool at

the time because I can imagine it could vibe so many different worlds um you know like for example just using client space using this sort of two sign

two spatial dimensions for this was uh already sort of a an assumption that I think is actually kind of important in some ways and does provide

feedback to our world but in the concept of you know you could have asked chat GPT to solve this problem in all sorts of number of ways

and you know maybe it could come up with all sorts of things which don't really align with maybe the useful taste as a community how do you actually deal with that like a proliferation of

really interesting results but it's actually not clear what is where the field should go I mean you're getting at the heart of what does it mean to do progress in theoretical physics and research

this is a hard question and there is a a simple answer if there were it would be research um let me say a couple of things the first one is when you go to graduate

school in physics you it's usually because you're really interested in the big questions why are there three dimensions of space

what happened at the big bang what's inside a black hole the the things these are the things that you know I was thinking about because of

sci-fi movies and books that are called you know what you realize is that actually these questions even though they're really cool and exciting

they're not really the most uh fruitful scientific questions because at any given time there's an edge of knowledge

and the role of scientists is to expand the edge of knowledge is to push into the unknown and to do that you want to find the questions that are right at the edge or

just beyond the edge but not so far that you can't grapple with them so the question of why there are three dimensions of space that's a really cool question but I don't know of anyone who said anything

really compelling about that so it's just a question that's beyond the edge so I it's not as a professional physicist I don't spend my time thinking

about this because I just don't know if any pathway to solving this question it's not useful to think about so really the

process of training as a physicist involves coming to grips with what the edge of knowledge is because that's where the interesting fruitful questions to make progress on

as a scientist that's where they live often times when you go through graduate studies you worry oh my god I have to learn about Feynman diagrams and all this math

and all these calculational methods and it's true that's a really hard thing to learn takes a lot of work but in some sense you know once you become a a professional physicist you

should feel like you can learn any tool you can pick up any tool that is needed for the task at hand and you should develop that confidence and that's what makes a competent physicist the competent physicist is one

that can learn any new mathematical tool or or you know piece of code or whatever that is needed to solve the problem at hand

and that makes you a good physicist or a competent one if you pick up the skill and in graduate school you know it's daunting you have to learn a lot but by the end you should have a lot of skills

in the toolkit and the confidence to to pick up any new one as needed the difference between a good physicist and a great physicist is knowing what is the right question to ask that is actually the hardest part of

being a scientist it's knowing what is the next fruitful question to tackle and I think AI right now is a very good physicist

in fact maybe superhuman when it comes to certain computations but it's like this extremely technically skilled graduate students

that you can give a a sharp well-posed question to it will do incredibly hard calculations correctly now and come back to you with the answer so it's it's super competent but one of

the things that it doesn't quite have yet is knowing what is the right question to ask and I think just like with humans that is actually the hardest skill to pick up that's the one that

comes last yeah I know you're not working directly on the AI so much I mean I don't know exactly how much you but do you get a sense for you know you can imagine a future where

you just do better reinforcement learning maybe you change the architecture of the model completely so that it's some other you know like

not transformer and the trajectory just keeps going like this because it's been very very rapid increase since 01 yeah of the capability reasoning

capabilities or do you get a sense that we're getting near the edge of the frontier of knowledge now so that the sort of the

the ability of the model to recombine knowledge in somewhat novel ways is you know like that's kind of and it seems like not not to disc or like not to play

down any of these results but that it seems like there was a a lot of what it did and maybe there's some not like this but a lot of what it did was recombination of known facts okay yeah yeah but do you get a sense

that that's do you have any reason to believe that will continue or if we're going to just like sort of okay we've we know how to recombine stuff really well and we can't push beyond that without

getting too philosophical I'm not sure that any of us are anything more than recombination of a stack machine working with GPT 4 on this problem to me

feels like working with a creative collaborator it did stuff that I didn't know that I I found surprising

and so I think I'm not sure there's a qualitative difference I think it's just a matter of degree okay that as we continue scaling the capabilities which is certainly happening I I don't see why it's going

to stop like we definitely have a bunch of things in the pipeline that are going to keep coming this year and you know my horizon for seeing into the future is like not not that good but beyond the year but like definitely

we're going to keep scaling up this year and I don't see any reason why it's going to stop and I think that's going to make these models display feats of insight

that look to us like real creativity I would say this already happened in this project at least you know what is creative insight is a bit in the eye of the beholder, right? I mean, AlphaGo,

right? What I want to come up with moves

right? What I want to come up with moves that were very I talked to Terry Tao a couple of weeks ago at UCLA.

We had an OpenAI event with IPAM, which is this Institute of Mathematics there.

And I talked to Terry Tao, and he said that in his view, all of the proofs that he's seen AI come up with in math, even the ones that at first seemed creative and surprising, later on were tracked

back down and found to have really pulled facts out of some obscure reference. So,

reference. So, I don't want to put words in his mouth, but my understanding was that, you know, Terry Tao has not yet been impressed by a creative move in math. But, you know, Terry Tao is a unique individual.

I've been impressed. I consider myself, you know, my bar is is lower. And I I think as we keep scaling this up, I I can't go into the details, but there's a lot of effort at OpenAI. There's a lot

of really smart, hard-working people that are pushing very hard to to take this next step. And I I think it's going to come eventually. I mean, just look at the

eventually. I mean, just look at the trajectory that we're on. So, a year ago, I was black hole physicist in academia, not really paying too much attention to AI. I thought, yeah, it's cool for

AI. I thought, yeah, it's cool for emails, but it's not going to do what I do, which is special.

03, which was the really first really strong reasoning model, came out and was able to do a calculation for me that would have taken

me days. It did it 11 minutes.

me days. It did it 11 minutes.

And I thought, wow. That was shocking to me, and I can We could go into the the details if we have time. I could show the example, cuz I Yeah. I saved it. It was really

Yeah. I saved it. It was really surprising to me. And then I thought, okay, I got to really start using this tool. There's no other software that can

tool. There's no other software that can do this kind of calculation as far as I know. It's

really surprising and really cool. And

then GPT-5 came out 6 months later, and that was able to reproduce one of my hardest calculations, which I think the number of people in the world that could do that you could count on your hands.

So, when you say reproduce, you mean this has been published or not published? It was a secret or internal

published? It was a secret or internal or So, last summer, in June, I put out this paper, which I really like, in which I found It's called Why is there no love in black holes?

Yeah, and love is actually a technical term. It refers to Augustus Love, a

term. It refers to Augustus Love, a British mathematician who studied the tides.

So, when you have an object like the moon going around the Earth, it exerts tidal forces on the oceans.

And so, when you can measure the tidal response, say of the Earth and its oceans to the moon, via some coefficients that encode the strength of the tidal response, and

these are called Love numbers, uh in reference to Augustus Love. But,

famously, black holes do not experience tides. So, they have no love.

tides. So, they have no love.

And there's been a resurgence of interest in this fact in the last 5 years because people understood that this can be connected to symmetry principle. So, in physics, whenever

principle. So, in physics, whenever something is zero, like, why should black holes never experience tides? That's surprising.

experience tides? That's surprising.

Well, often times, the answer is because there's a symmetry principle at work that forbids the existence of tides, that protects the structure of the black hole. And so,

I found these new symmetries. So, these

are differential operators that act on solutions to this equation that describe perturbations of a black hole.

And these generators are symmetries because if you act on the solution to this equation, you get a new solution.

You know, I thought this was very beautiful. I liked it very much, and it

beautiful. I liked it very much, and it came out in June on the archive.

And in August, GPT-5 came out. And the

cutoff date for its training set precedes the release of this paper. So,

GPT did not see this paper in training.

And when it came out, I was like, okay, I'm going to got to meet Mark Chen, who's chief research officer at OpenAI, and he said, give GPT Pro a really hard problem.

Let's see how good it is. And I was like, You want a hard problem?

I'll give you I'll give you a hard problem. I just solved this problem, and

problem. I just solved this problem, and I wrote a paper. I was very excited about it. I thought, this is really deep

about it. I thought, this is really deep and cool, and um I gave GPT the equation here, and I said, what are the symmetries? I didn't

tell it that there are symmetries, because by default, they go via suit should be that there aren't any. And

thought for 5 minutes and said, yeah, there are no symmetries, which is what usually happens. And that was wrong. And

usually happens. And that was wrong. And

Mark Chen was visibly crestfallen. He's

like, oh, uh well, okay, what if you give it an easier question? And so, then I gave it the

question? And so, then I gave it the same question, but not for a black hole space time, but for an empty flat space time, which is a simpler problem, but that's actually how I approached this problem myself. You

know, you warm up on the easier question first. So, I gave it the flat space

first. So, I gave it the flat space question, which uh is in this paper.

Also, so it's it's uh this equation, which looks much simpler. Then, this

also has three symmetry generators, which are shown here. This is not new.

It's These equations have been studied for 200 years. Everything in flat space has been known forever. And GPT-5 Pro thought for um it was like 9 minutes, and it came up

with the answer, very beautiful answer, perfectly structured, perfectly correct.

Actually, at the time, I I also tried the other models from our competitors, and none of them could could get this at the time. So, GPT Pro is really ahead,

the time. So, GPT Pro is really ahead, and I think it continues to be the best model for this kind of mathematical physics work.

And um Mark Chen was like, okay, well, this is great, but now that it's done the warmer problem, in the same instance of chat, try the full problem again. Now that

it's been primed, I thought, okay, why not? And so, I gave it the the same

not? And so, I gave it the the same question as before. What are the symmetries of this equation? Now, the

full black hole problem. And this time it thought for 18 minutes, which I'd never seen before, and it came up with the answer. So,

basically, in under 30 minutes, with one hint, which is the obvious warmer problem to prime the model on first, it completely solved this problem, which, you know, is one of the

nicest calculations that I've ever done.

And uh that really blew my mind. That

was my move Move 37, the Evan moment. Yeah,

that's how we call it in the AI world.

And um once I saw that, I thought, okay, we're on this crazy trajectory where, you know, 18 months ago was not useful. A

year ago, it could do really hard calculations that would take me days. Eight months

ago, it could reproduce some of my best work in like under 30 minutes.

And then, now, last in the last month, it solved these questions that we've discussed at length, which the world experts had spent a year thinking about

without being able to get to the answer.

So, you know, I think it's just going to keep getting better. Where are we going to be in 6 months or a year?

I don't see any reason why it would stop. And I I think we're going to be

stop. And I I think we're going to be uh having a very exciting year.

[laughter] Okay. Uh going back to this like these

Okay. Uh going back to this like these thoughts about scientific discovery and what can these models do versus just being very good at superhuman at solving physics.

Um the people keep asking this question.

Hypothetically, could we train a version of ChatGPT where it's never seen anything post, you know, 1904, and could it rediscover relativity?

I think that there's a very analogous like question we could ask right here, which is new conceptual result about uh single minus gluon amplitudes, which was

sparked by human insight. And, you know, there were some certain very specific assumptions which went in here, like understanding working in Kerr space time is something that people have been

thinking about and has some useful like transferable insight. Um people have

transferable insight. Um people have been thinking about, you know, maximally helicity violating amplitudes for quite some time. Have you ever tried kind of

some time. Have you ever tried kind of using a model right before the cutoff date of this and asked, given a Kerr metric, is there anything

interesting with regards to helicity violation? Or maybe turning around

violation? Or maybe turning around saying, you know, it's long been thought or it's long been known that with exception of some set of measure zero due to Witten,

there's no, you know, single minus nonzero amplitudes. Have you tried

nonzero amplitudes. Have you tried either of these directions and asked it to like discover a new insight, push the boundary as you were just talking about, and make a leap, in addition to not just

solve a problem that you can give it, but actually, could you get that intuition?

Yes.

Yeah.

You you have tried this? Not exactly the counterfactual version that you're describing. I personally haven't done

describing. I personally haven't done that, but pushing the models at the frontier to try to make this type of leap is something that we're very

focused on. Yeah, and I think um

focused on. Yeah, and I think um Well okay.

I don't want to talk about the internal research we're doing, but I can say something publicly, I think, which is you can take this page of this paper, and you can feed it to

ChatGPT Pro, say, like the best model we have out right now, and you can ask it, what should I do next? Give me the top three

follow-up questions to ask based on this paper. I've done this experiment, and

paper. I've done this experiment, and the top three questions it comes up with are like my top three questions for what I should do next. So, I think the models are smart enough now and have enough

background knowledge that, you know, for this paper, I'd say GPT is about as good at as me at finding the next thing to ask. And so, that's really interesting, and it opens up a

lot of uh So, can you just, you know, do What is What is the name of the loop that the agent loop they were is about where you just say like okay, what's the next question go ahead and solve that what's the next question

going in. So I I guess in this goes back

going in. So I I guess in this goes back to the the question I was asking before which is if you do that and probably has been you've tried it or something that OpenAI has tried that like I presume you get to

some plateau, right? Where you like you're not pushing the boundary of knowledge anymore or is it just like the plateau is money and if you had more money you could go further?

Yeah, just to be very explicit because I haven't said this quite out loud.

I think we now have models that can really turn out papers that are as good as human written papers. In fact, um

this is a bit of a problem because when a professional physicist uses this tool and they steer the model and they check the answer they can get amazing results,

but there are also people that uh feed it kind of wrong questions that go off the deep end that yeah, they submit that to archive and this is a problem

that the academic community is dealing is try come to grips with now which is this problem of AI slop it for science.

This is something we have to figure out.

But you know, I would say that with proper steering you could probably turn out a paper a day now, you know, just I don't know it's like give the question

to chat GPT it'll solve it if it's not that hard of a question or it's a similar calculation to stuff that's already been done. It can totally do it in 30 minutes and then you could say write it up as a paper and you can send

it to archive. Okay.

So I think that we're already in this moment we've we've passed that threshold. This is like the new reality.

threshold. This is like the new reality.

And more and more people are catching on to this all the time and so some of them are doing this and this is why the archive is now inundated with with submissions.

So what's the correct response to this?

I think we put out these two papers in very fast succession.

We could spend the rest of the year writing 30 more papers like this.

I don't think that's what we should be doing. Um

doing. Um instead, I think now that we have this new tool that gives us AI superpowers, I think we should just raise the bar for what it means to write a good paper.

Like we should aim higher basically.

One thing that I'm excited about is that I think these uh single minus amplitudes papers they open the way now to a whole direction of research which I think is a line of

attack on really interesting questions in quantum gravity. To go back to the start of the the session this is the missing piece of the puzzle of fundamental theoretical physics.

And I think we have a pretty clear line of attack through series of questions all of which I think will be amenable to solution with AI.

And so I think you know, I'm excited to spend a good part of of this year trying to follow this path but really solve harder and harder problems.

And you know, this is this is a this paper gave an answer to a question that had stumped Andy, Alfredo and David who are experts in this for a year.

But we haven't seen an AI yet solve a question that stumps an entire community of physicists for decades. That hasn't happened yet.

decades. That hasn't happened yet.

But I think given the trajectory that we're on at some point hopefully not too far in the future we should see that.

And so I think that's the exciting thing to try to go towards which is pushing the envelope of what can be done. We

wanted to start asking all I guess question which is if you could remove one bottleneck for your domain. So in this case maybe it's AI

domain. So in this case maybe it's AI for physics or maybe it's physics or maybe it's mostly AI. But if you could remove one bottleneck for for your domain what would that be and why?

Well, off the top of my head, you know, I spend so much of my time writing papers and the way I think now is so far from papers it just feels like not the right

way somehow to store and communicate knowledge.

I think an extreme version of this which makes the problem more apparent is math.

Especially certain parts of math where papers are very terse and they take four pages. I had this experience when I was

pages. I had this experience when I was learning algebraic geometry in graduate school.

Go to a mathematician and say what's going on in this like four-page paper? It's just like very

four-page paper? It's just like very terse notation.

And he said oh, forget what's in the paper and he took me to the blackboard and he started to draw pictures. You

know, he's like this is how you should think about it. And then I was like oh wow, this is amazing. But like none of that is in the paper. And you know, mathematicians I think have this cultural norm that they kind

of hide the messy work and they will write these beautiful short pristine papers. It depends on the subfield, but

papers. It depends on the subfield, but often times that's the case and the way they actually think about the subject is a living breathing entity is very different from the way it which is

recorded in papers. Some of that is also true for physics. You know, I love doing calculations, coming up with questions, finding the answer.

And then I would say the huge bottleneck is writing it up.

So um somehow it feels like papers are not quite the way of the future or at least the way that we currently operate with I write it up

send it to a journal it takes six months. I I don't know it's just like

months. I I don't know it's just like why are we doing all of this?

Feels like maybe there should be something better. I mean, you could you

something better. I mean, you could you know, if you want to understand this paper one thing you can do is upload it into chat GPT and ask it to explain it

to you and you can keep unfolding the complexity into more and more detailed explanations.

And so if we move it to a world where we use AI to do the calculation get the result then we have the step condense it into a paper and then you know, I send the

paper to Brandon and he puts it back into an AI that will I mean let's get it. Like why are we why are we doing

it. Like why are we why are we doing this?

Yes. Right? That's a little bit funny. So I feel like you know, if you

funny. So I feel like you know, if you ask me would I be confident that in 20 years we'll have these sort of like static documents in which we publish our results as papers, I would think not.

Like that doesn't seem like the best thing we could be doing.

Maybe some kind of interactive paper which lives in some LLM.

Maybe your whole paper is some chat GPT page.

And you know, you there's a chat bot attached to the paper and you can say explain the big picture and like zoom into this fact.

I think we're going to head in that direction. That would be a cool thing to

direction. That would be a cool thing to see. Writing a paper though is a useful

see. Writing a paper though is a useful exercise cuz it forces you to condense your thoughts and make them really clear. So I'm not saying

clear. So I'm not saying it's a bad thing to do in general, but just the way we do it is is very slow.

But that's that's the first thing that came to mind. Maybe another answer is in this project the graviton paper we got

to a draft of a paper extremely fast and then we spent most of our time checking the answer. So I think

the answer. So I think that will effectively be a big like maybe the next big bottleneck. And

bottleneck. And that is one of the things that the models I would say if you ask me what are what is missing in in the models?

Like what can we really improve for scientific research?

I think we've we've kind of touched on on the two big things already, but just to spell them out.

One is creativity and the spark of invention and really taking the the next step. I think that will come as we scale up the intelligence. Well, we'll see, but I I

intelligence. Well, we'll see, but I I don't know that there's something missing inherently. I think it's just

missing inherently. I think it's just it's starting to to make these leaps for me.

But maybe we should encourage the models to to try to make bigger leaps because large language models after all they're trained to give you the middle of the road answer. Like if you ask an AI like

road answer. Like if you ask an AI like chat GPT write me an email about blah.

You want it to give you kind of the expected answer not to sample from the tails.

Like a wacky email, you know, you kind of want it to give you a a reasonable thing. So for most tasks you want that,

thing. So for most tasks you want that, but for science research sometimes you want the idea that comes out of left field the thinking outside the box or or really sampling far out of the distribution.

And that's something we could do in principle, but that's not how the models are you know, we're not really favoring that.

So we might have to do tweaks of this kind to make the models be able to take bigger leaps. And then the second thing is

leaps. And then the second thing is verification because we're now in this new regime where the models are so capable that for very hard computations at the frontier of

knowledge they they can just do the whole thing, but you know, is it correct? In this case it was correct.

correct? In this case it was correct.

You know, sometimes I get emails from people saying oh, I did this really long calculation, but there was a mistake somewhere disappointing.

Okay, I mean the calculations are getting more and more complicated longer and longer, but yeah, sometimes they mess up and so I think improving verification or even just having the model

indicate more directly how confident it is in its answer cuz I think they're smart enough to know yeah, whether very confident in the answer versus when they're just kind of guessing in some step and getting the AI

to be more explicit about that is I think a way to improve them for research and that verification step I think is going to become maybe a bigger

bottleneck this year. Yeah, Korean Hong from Axiom would agree with you emphatically. She

emphatically. She formal verification is their thing, right? Yeah, I think

right? Yeah, I think you know, it's interesting a year ago I would have said it's super important to have formal verification.

Then the models got so smart that I thought well, you know, if Brandon and I talk about a mathematical proof and we go over it, we're not going to formalize it in

set theoretic notation or you know, we don't think the way lean, which is this language for formal verification, reasons.

We reason through the proof in natural language. We use words. And so if a

language. We use words. And so if a model is really smart enough, then it should be able to do the same thing. And that's we've been saying this

thing. And that's we've been saying this huge increase in capability for mathematical reasoning and development of proof using natural language.

So then it maybe for a while it's looked like that that wasn't the thing to really focus on.

But now that we're in this regime where you can just get chat GPT to tackle thousands of questions at the same time and it will return proofs for a

significant fraction of them.

Now actually the onus is back on the humans to verify all the outputs.

And so yeah, if as that becomes a bottleneck, I think formalizing math and automating verification will become more valuable it looks like to me and that's something we're thinking a lot about as well.

Thanks. What do you want the audience to take away from today? Like is there one message that you want them to leave with? Yeah, I think it's important to

with? Yeah, I think it's important to get the word out that the models that we're developing at Open AI are becoming really capable in

scientific research.

I myself was a bit of an AI skeptic a year plus ago because I thought the models are very good at writing tasks but not mathematical tasks. That changed

with O3, the first strong reasoning models.

And then GPT-5 was able to do some of the hardest calculations that I can do and reproduce them correctly. And now

recently in the past month we've seen models solve open questions in theoretical physics.

And now they're solving problems in quantum gravity and quantum field theory. So if you just extrapolate that

theory. So if you just extrapolate that into the future, imagine where we're going to be in 6 months or a year, I think it's kind of surreal to live through this time,

but it's really happening. It's It's

really amazing. And I think we're going to see a lot of big changes happening in research. So I that's

research. So I that's Yeah, pay attention to this space. Let's

stay tuned. That's awesome. So we go so much for taking the time. This is like I I learned a lot from our discussion and and I'm going to definitely keep up with what you're what

you're up to. Thank you. It's been great to be here. Thank you. Thank you.