Into the Impossible With Brian Keating - The Physics Reason AI Works When It Shouldn’t

Episode Date: August 25, 2026

The same mathematics that explains why a fridge magnet sticks to your refrigerator also explains why neural networks work when they have no right to. His group did the calculation. Nobody else had. Subscribe if you want science with evidence, not speculation. Goldenfeld is a Professor of Physics at UC San Diego, a Fellow of the Royal Society, and a Member of the National Academy of Sciences who spent 36 years at the University of Illinois applying condensed matter physics to problems everyone else had given up on: why the genetic code is optimal, why early life evolved impossibly fast, and why AI works despite being overparameterized beyond anything classical statistics can explain. The thread connecting all three is one idea: that what emerges from many things together is qualitatively different from the sum of its parts. That idea explains magnetism, AI, the origin of the genetic code, and why life may be inevitable wherever the laws of physics apply. He also argues that what is happening to science right now is not a disagreement about facts but a fracture in how people decide what is true, and that is a more dangerous problem. What you’ll hear: -Why the same phase transition that explains magnetism also explains why AI works at all -Why Francis Crick concluded life must have come from outer space and what Goldenfeld found instead -What horizontal gene transfer has to do with how libraries work -Why Goldenfeld thinks Enceladus is a better bet for life than Europa -The purpose of life, stated as a thermodynamics problem -Why “different is more” is more useful than “more is different” “The impact you make is the ratio of what you do divided by what everybody else does. Minimize the denominator.” — Nigel Goldenfeld CHAPTERS 00:00 AI shouldn't work. It does. 00:56 What is a phase transition? 03:10 What the renormalization group does 08:54 Ising gave up. Wrong dimension. 13:32 Nigel almost met Ising. 40 minutes away. 26:38 AI is the best example of more is different 33:52 Bardeen won two Nobels. The transistor looked like chewing gum. 37:24 The three mysteries Crick couldn't solve 46:20 The genetic code can't evolve. And yet it did. 49:00 How early life evolved like a library 57:06 The purpose of life as a physics problem 01:00:04 Life is physics, not chemistry 01:05:56 Anti-science age. Not because of opinions. 01:13:00 20 seconds with your 20-year-old self 01:17:14 Different is more Get the transcript, fascinating bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt Have a .edu email and live in the USA? You automatically win a meteorite: https://BrianKeating.com/edu Subscribe: https://www.youtube.com/DrBrianKeating?sub_confirmation=1 Support Into the Impossible on Patreon, get my weekly M.A.G.I.C. Message, unfiltered bonus content, and live monthly Office Hours with me: https://www.patreon.com/drbriankeating Join this channel for perks, monthly Office Hours, and your name in the Member Roster at the end of every episode: https://www.youtube.com/channel/UCmXH_moPhfkqCk6S3b9RWuw/join Featured Guest: Nigel Goldenfeld website: https://guava.physics.ucsd.edu/~nigel/ Lectures on Phase Transitions and the Renormalization Group: https://www.amazon.com/dp/0201554097?lv=shuf&channelId=500&plpRedirect=mhFallbackNigel Goldenfeld on Twitter/X: https://x.com/NigelGoldenfeld My books: Losing the Nobel Prize (memoir): http://amzn.to/2sa5UpA Think Like a Nobel Prize Winner: https://a.co/d/03ezQFu Focus Like a Nobel Prize Winner: https://a.co/d/hi50U9U Galileo’s Dialogue (first-ever audiobook): https://a.co/d/iZPi9Un Twitter/X: https://x.com/BrianKeating Substack: https://briankeating.substack.com Blog: https://briankeating.com/blog Audio-only: https://briankeating.com/podcast #intotheimpossible #briankeating #NigelGoldenfeld #physics #AI #originoflife #condensedmatterphysics #podcast Learn more about your ad choices. Visit megaphone.fm/adchoices

And you're way, way into that regime where you're just fitting noise and the whole thing shouldn't work. It obviously shouldn't, and yet it does. Francis Crick said, look, there's no way that life got to this level of complexity in such a short period of time, it must have come from outer space. The genetic code is optimal in the sense of minimizing errors. If you want to know what is the purpose of life, the purpose of life is to help planets come into equilibrium. So a phase transition you know is like, for example, what happened to to a piece of metal when I cool it below a certain temperature, and I want to know, can I use

it to stick the pictures of my kids' holiday pictures on the door of my refrigerator? And the answer is yes, if it's magnetic, it'll stick with a piece of metal holding the picture up. And if it's not magnetic, it won't stick. And if you take that piece of metal that works as a magnet, fridge magnet, and you heat it up, it will stop becoming a magnet. And that's called a phase transition, as you know and maybe some of your listeners know, or viewers know. The interesting question you might ask is, as I get closer and closer to the temperature where the magnetization disappears, how does the magnetization disappear? It disappears gradually, in fact.

And if you ask how much magnetization there is, the answer is it goes like the square root of the difference between the temperature you're at and the critical temperature where the magnetization fully goes to zero. At least that's what you would expect, and that's what very generic, very persuasive, simple theory, theoretical arguments that anybody can understand. I can explain it to my class in literally, you know, 20 seconds. That's what you would predict. When you do the experiment, you find that it doesn't go like the square root of the critical temperature minus the temperature. It goes like

TC minus T to a power like 0.3, 2, 6, 5, 1, 3, 6, some weird, weird number like that. And you might say, well, it's just a,

you know, a more accurate number. Yeah. The problem is that there's no known way, or there was no known way, to account for the fact that the member's not a half. I mean, it's like,

to prove it to half, all I need to know is that magnets can be either magnetized north or magnetized south, And that's basically it. It's an argument that is so compelling, it can't possibly be wrong. And yet, in the decades from the late 1940s up to the middle of 1960s, it was discovered

that it was wrong. And it wasn't just only the magnetization. There were other thermodynamic properties like heat capacity and things like this, which I wouldn't go into, which also have a similar unaccountable behavior. And it was the fact that these numbers, of himself not particularly important. It was the fact that you couldn't even explain in principle

why they are not these simple numbers like a half and so on. And that was the reason for the puzzle. And the explanation is a truly mind-boggling explanation. But just tell you the outlines of the story, this phenomenon

was addressed by Leo Kadenoff, Ben Widom, eventually Ken Wilson. And they invented this process of looking at a physical system on different scales of energy. So, you know, you could look at matter at the scale of this room, you could hear the sound waves, you can see light bouncing off the surfaces. On the other hand, if you want to go and see that there are atoms and see that there are quarks and things like this, you need to build, you know, a machine that's, you know, put in the tunnel

17 miles long under the Swiss Alps in order to be able to see things like that. So what you can C depends on what energy you look at it at and what time scale and what length scale you look at it at. And the same thing turns out to be true of the laws of physics themselves. And Leo Kadenov was the first person who realized

that, and Ken Wilson turned it into a mathematical tool, which was called the renormalization group. In fact, it's not even a group. You ask whether it's a group. It's actually a semi-group. And the idea was this. Take a physical system and then just say, well, you know, I've got magnetic dipole moments.

They're really spins of electrons, but we'll just call them magnetic dipole moments. They're in this bit of the sample. We'll just lump them together into one effective dipole moment because, you know, in this patch over here, you know, 80% of them are pointing up, 20% are pointing down, so we'll just say, okay, it's basically just a spin pointing up. And so you sort of block things up in that way,

and then once you've done that once, you can do it again, and again, and again, and you keep on doing it. and then you ask what happens when you take that process to the infinite limit. And it's called a group, it's really a semi-group, because you can lump the spins together and then make these bigger and bigger spins, this coarse-graining, as physicists call it, but you can't go backwards. If you know the configuration of the very large scale and you say, well, what was the actual microscopic configuration of the electric dipole moment,

the electron dipole moment, there's no unique answer to that. the one that we had 80% up, 20% down. If it had been 75% up, 25% down, it would have still ended up being as regarded as a spin just pointing up. And so there's no unique answer. You can't go back. And that is why it's so profound. Because when you start looking at the laws of physics

and you say, actually what we're doing, when we're doing this, we're actually looking at the laws of physics. And I can tell you why it's the laws of physics in a minute. Then if you know what the laws are microscopically and you start saying, if I know what happens at the scale of atoms, can I work out, you know, what happens to my crystal or fluid or something like that?

Well, I can do this process and cause grain like that. But I can't go backwards. If I can't go backwards, then you say to yourself, wait, I am a physicist. All I can see around me is the natural world. There's nothing in this room that, at least if I live in the 19th century, would let me know that atoms exist and things inside the atoms and so on and so forth. You're going to

crazy, right? Well, it drove Bolton to suicide. It's not just that. There isn't a way in printable. So let's suppose you're trying to work out

theories of the standard model as people were doing in the 60s and 70s. Well, you can write down all sorts of non-abillion-gaged theories that you like, and they all sound very interesting.

You have no way to know which one of them is right because they will all give you the same standard electricity and magnetism that we use in everyday life. And so, the renormalization group, this

non-un uniqueness of going down in scale and going up in scale, that's the thing that makes high energy physics, fundamental physics, if you like, hard. And it's important in condensed matter physics because you say, well, I'm going to start

at a level of description, which is I'm going to take atoms for granted and then work out what are the properties of matter. And so you can go in the way where there's any one way to do it. And that's why conduct matter physics is so successful. The other thing I would say is that this facet of the normalization group is what enables us to do physics in the first place.

So think about it this way. Suppose you're a chemist, okay? And you're trying to understand organic chemistry, chemical reactions, biochemistry or something. Let's suppose, you know, somebody knocks on the door and says, hey, we've got some really disturbing news. Somebody's just measured the radiative corrections to the mass of the top quark and the 20% difference than what we thought. Does that mean all our chemistry is wrong? Well, the answer is no. Because all of that was just lumped into one constant in the theory,

just like we were taking the spins and lumping them together into one effective spin, and that constant is the mass of the proton, the mass of the hydrogen atom, whatever. So a chemist is not worried about that. We know that the microphysics can be lumped into effective descriptions at a larger scale. all the QCD and super strings and whatever, all ends up just giving you the mass of the proton, the spin of the proton, whatever. Because we don't have to worry about that,

you can do chemistry. You can say, I don't worry about what the world is really made of. I just start with my level of description that I'm comfortable with, atoms, and then I proceed from that. And so without that, if that didn't happen, then we wouldn't have been able to do science. I want to get back to reversibility

and maybe even touched upon the origin of time, perhaps in the context of temperature. But before I get there, why didn't Ising realize this, you know, 40 years before Wilson and then Kadenoff and, etc? There's a model called the Ising Lentz model. Lens was the Isings advisor, and Lens decided it would be nice to make a simple mathematical model of a magnet.

So this was in like 1925, 1926, I think. So they didn't really know all the microphysics. So they said, well, let's just say that we have, you know, magnetic dipole moments, and we'll just say that in appropriate units, they can either point up or point down, spin up, spin down. And that model has just two variables. They can be, spin can be pus or minus one. And that's it. And then you take a lattice of all of those things, and then here's the extra ingredient. So a magnetic dipole, of course, interacts with an external magnetic field. If you apply an external magnetic field

to a dipole, it will orient with the external field. But there's another field, which is the fact that the electric dipoles exert a field that on each other, on their neighbors. And so this was one of the first models where you explicitly had a cooperative phenomenon built in. In other words, whether this spin points up or not, depends on not only whether there's an external magnetic field, but whether its neighbors are pointing up. If its neighbors are also pointing up, it's got a much hard likelihood of pointing up. If its neighbors are pointing down, it will likely point down. And so that's why this, as being, you know, it's a famous model. You know, you can model social behavior with it. You can

model the production of pistachio nuts in California orchards using it. As we often do. So that's what the model is. And it really is the drosophila of theoretical. That's the elegant. Career to Matter of physics. What did Ising do? Well, this is a fantastic, fantastic story of failure and missed opportunity. So they said, well, let's just simplify the problem, instead of having a three-dimensional material, which is what matter really is, we'll just say, we'll just do these spins in one dimension. And so then it's possible to, even if the problem is non-linear in a complicated way, you can solve it exactly. And when Ising did that, he found that it didn't have a phase transition. This phase transition of the magnet,

the fridge magnet that below a certain temperature will stick to the refrigerator door, and above that, it won't. It didn't work in this calculation. He didn't see that at all. And so they gave up. They just said, well, this model is a useless model. Okay. Now, there's a reason.

It's a very interesting thing. What they didn't know was that the behavior of matter strongly depends on dimension. And that's not obvious. And it wasn't known at that time. Now we know it. In fact, we now we use dimension as a variable. which we treat as a continuous variable and do perturbation theory in dimension and things like this.

And then Lars-on-Zaga figured out in 1944 how to solve the two-dimensional Ising model exactly. And I think it's on a par with Einstein's theory of relativity, general theory of relativity, as one of the most fantastic examples of theoretical physics I know of. It is a masterpiece, and it inspired many things, including string theory and all sorts of things like yours. There's so much to be said about that. Okay, so then we knew that there was a phase transition, and it didn't behave the way that you would have guessed, the simple square root theory that we talked about. And eventually, we discovered that, well, it's not, as a community, we discovered that being able to solve the collective behavior of matter exactly is a fool's oven. it's much better to have an approximation method

that is guaranteed to work on any problem rather than, say, some right with two-dimensional icing model, which Onzaga solved by absolutely brilliant mathematics. His solution worked, but if you apply an external magnetic field, it all goes away. It doesn't work. And if you try to do it in three dimensions, nobody knows how to do that. So these special cases are special for certain reasons, just like integral systems, solitone equations in differential equation theory.

If you can solve it exactly, it means there's something special about it. And there was, and it's now understood. And the renormalization group, why it's taught and why it's so important in graduate physics, is, well, we need to know how to solve any problem in condensed matter physics, whatever, and I say many other fields of science.

We now know we can solve them systematically to any order you like, any accuracy you like, using the renewable energy. Before I came here to UCSD, I spent 36 years at the University of Illinois, I da Baner, Champagne, and one day I discovered, and I don't remember how,

that Ising was teaching at a university, a college, a teaching college in Peoria, about 40 minutes away from the university. And I thought, holy cow. And so I wrote to him, because apparently he didn't really understand

that his name is a revolutionary, is attached to a revolutionary model in physics. and so I wrote to him and unfortunately I was a few weeks too late I could have, you know I'd been there already

maybe 15 years when this happened I cursed myself why I didn't do it earlier but I did talk to his son and exchange some correspondence with his son and explain to him various things about his father's word

that's unbelievable it's like our own office works at Chapman University you know which is less than an hour from here And he's still alive, and many, including me, consider him, you know, worthy of a Nobel Prize for his work. And inspired my late great mentor, Jim Simons, and Cien Yang. Yeah.

Just incredible lectures on face transitions and the renormalization group. And it's written by you. And it's got this lovely cover. And it's got your description. Take us through this cover. The title, Frontiers in Physics, and this beautiful cover art, Nigel. Well, the cover art is a deliberate British understatement. And when you see a cover like that,

You think to yourself, oh, I'm about to walk through a garden full of myriads of beautiful flowers, strange butterflies, and wonderful, unexpected sights. And that's actually true because one of my colleagues at the University of Illinois, after the book came out, he wasn't in this field, wrote to me and described the book in exactly those terms. Wow. And the reason is because this book is, it's still used, widely used as a graduate text in advanced statistical books. mechanics. And I wrote the book because I thought I had something unique and new to say about the renormalization group, which other people hadn't noticed. There's things in it that you won't find in any other textbook in this topic, including the fact that the renormalization group has nothing at all to do with statistical mechanics. And the first exercise in the book,

as you may remember, I don't remember exactly which order. I think one of them might be, The first problem might be to prove Pythagoras theorem using dimensional analysis. And the second one is to work out the yield of the Trinity test of the atomic bomb based on just the data from the motion of the shockwave from the photographs that were published in Life magazine. And famously Fermi did the same thing.

He just sprinkled some piece of paper. He sprinkled bits of paper and and a Taylor, a G.I. Taylor, actually did the calculation that's in the book. And he actually got into trouble because he did this calculation,

reported the results in the newspapers, and it was classified information. And so, you know, it's... So there's lots of things in this book that are very unusual, and those things have stood the test of time and have actually grown since then in importance and significance.

And another thing that's interesting about this book is that it's a bit of... about a very obscure and arcane problem, and if you want, we can get into it. We're a normalization group, the problem of critical exponents in second-order phase transitions. But it turned out that this problem completely upended our view of what physics is, what we're doing when we do physics, and the nature of scientific explanation. I want to take one more detour before we get too deep into the weeds, and that's this thing that you mentioned before, which has had to do with reversibility

and the fact that there are, there's no inject or bijection, I guess you'd say, between final state and the initial state. There are many initial states that can produce a given final state. So it's not invertible, essentially. Right.

I'm going to say, we're not talking about states. We're talking about that the variable is not time. The variable is scale. Scale. And that was the thing that Kadenov... Right. That the way, the energy scale,

at which you look at a system, is the important thing. And so I will often talk about the importance of levels of description. Now, it's a very important problem. For example, if you're a biological physicist, as I am, you might say, well, what is the right level of description to describe a biological system? Should I describe every atom in the biomolecules that are inside a cell and then inside the cells and inside the tissues and so on?

Or should I try to make a more coarse grain description? And this is not an easy question to answer. because it depends what is the question you're trying to understand. If you're trying to understand how does some particular molecule bind to some particular protein or something like this, you definitely need to understand the atomic level of description, the binding and things like this. If you're trying to understand why is it that inside a eukaryotic cell, we now know just in the last 15 years or so, that in fact the biomolecules phase separate from the rest of the, of the cell and form a membraneless compartment

inside which, God knows what happened, we were still trying to understand the function of these things, so you have these sort of phase separation processes, we understand those at a very different level of description. It has nothing to do specifically with the atoms and molecules and the specific sequences of the RNA and things like that. It's a general property. So you have to, so depending on what question you're trying to understand,

different levels of description are important. This tension is very prevalent in biology because it's not obvious what is the right? It's not as simple as saying, why I'm the chemist? So I'm just going to assume all the microscopic high-energy physics, standard model particle physics stuff is just absorbed into the mass of the proton. We don't know that you can when and where you can do that. Where the evidence. In some things complex as biology.

Interesting. You start your lectures and UCSD statistical mechanics. I can't tell if they're graduate or undergraduate, because they're graduates. They're graduates. Is you seen it on YouTube? I do watch them on YouTube. It's Chopin lover.

Is that your channel name? Shopin, Junkie. But you start the lectures, you use this famous, phrase, which I've always felt, I hate to say it, Nigel, and I know he's a hero of yours, but Philip Anderson's, the Amora is different. Yeah. I always felt that was kind of simplistic, but maybe I'm wrong.

I'm just a dull-headed experimental cosmologist. So tell me, what is the significance? Is there anything really significant about, I mean, of course, like, where does the sand grain start to become the sand pile, where do you... It's more... So tell me, what does it mean to you? Why is he a hero?

Why do you start with that in that election? Well, is the Anderson hero? Okay, many things. He won the Nobel Prize for his work on assorted electrons, but there's no field of condensed matter physics, which was left untouched by his intellect. So in condensate matter physics, he is a giant, and is in the same way that Einstein, Hawking, others. And I'd say more than bore, actually.

The more is different. The article was immensely influential. First of all, he really was the first person. I mean, there's technical ways in which more is different is important. For example, you can't have phase transitions unless you have, you take the sort of thermodynamic limit. But it's not just that when you have faced transitions, it's just that you can have previously unanticipated complex behaviors that you would never have otherwise expected based on looking

at the thing that stuff is made of. So, again, let's go back to our fridge magnet. Okay, you've got your electrons in the material, and they have magnetic dipole moments. You would have, unless you did a particular calculation, you would never know that this thing could be used to stick your kids' drawings on the door of your refrigerator. Okay, it's a cooperative effect.

It is a conspiracy of the atoms. And I actually do an experiment, which we can do right here, if you're wades to it. I'll make it. I do this. Why wouldn't I? Well, so the experiment is this, okay?

So there's the ceiling up there. Yeah. Okay? And we're going to move the ceiling. Yeah. Okay, and we're going to do it like this. Take a finger. Yeah.

To this finger. Yeah. Okay. And push. Okay. Well, my ego, I'm able to do it. Come on. Come on.

Throwing. Put some effort into it. But this is a gas, not a solid. Put a solid. It's a gas, not a solid, so we didn't move it. Right. The thing is this.

The Hamiltonian, the formula for the energy of the gas is exactly the same as the formula for the Hamiltonian of the solids. They're no different. And yet, when I take this

and I push my water bottle, you know, my fingers don't go through. All the atoms in this conspire. Does that really true? I mean, at the same level, just to be, Vanderwals versus Hooks law.

No, no, no, no. It's really true. And this is why it's important, because when we're talking about emergence, remember I said we're talking about new laws of physics. When you have a solid, there are new laws of physics. The atoms have decided that they're not just going to sit at particular sites in a checkerboard lattice that somebody can be needed out for them. They've actually conspired that they're going to keep their relative separations the same.

And because of that, they minimize them. their free energy by doing that, as a statistical mechanical description of what is happening, that means that you now have new excitations, which are, first of all, you have the rigidity, the stiffness, the emergent rigidity of a solid,

as measured by the Young's modulus and things like this, and you have the ability to transmit sound waves and other waves as well, of course. So at the level of description of the material, you now have new laws of physics. And the only thing that's changed is the temperature.

You haven't changed the interactions between the atoms. You have changed the correlations and that's the important thing. But it's a statistical property and it's not one that you can see

just looking at two atoms. Sorry, just looking at two. You have to look at the whole ensemble. So that was a thing that Anderson was very interested in

and understood the depth of its significance more than other people. And later, He wrote, like I say, in 1972. In 1985 or so, when I came to the University of Illinois, the first project I did was with my cousin, Paul Goldbart,

who's also a very well-known theoretical physicist. Will next year be the president of the American Physical Society? And we worked on this question of why rubber is solid. So everybody knows that rubber is stretchy and expand. But that's not the right question. It's stretchy, but, and, you know, the piece of, you know, with this cable here is, can bend in deformant stretchy, but it's still solid. The first question you should be asking is not why is rubber erasic, but why is it solid in the first place? Because you have a bunch of polymer molecules.

They are stapled together by crosslings, just like having a bucket of worms. Or a bucket of worms, they're all flapping around in the thermal equilibrium. You go in and do an experiment that you've never got an IRB to do, even though a worm doesn't have a back, which has staple the worms together in random, and then you find that the thing is not just connected, like a fishing net, which would just be floppy, but it's actually rigid, like a gel. And then, you know, you can tap it, and it will wobble in.

It has sound waves and things like that. And those are all emergent phenomena. It's very complicated to calculate them, because the whole thing, the polymer chains are going at random, the cross links are at random. Everything is random. It's a very hard technical problem, but it's an example of this more is different,

and the point about more is different, and the point about emergence that everybody misses is that when you talk about emergence, something emerges, what is it? And the answer is it is a particular type of rigidity, which is generalized rigidity, as Anderson called it, which basically technically comes from a certain type of response function of how does the system respond to perturbations when you poke it?

And that was the lesson of that. And it wasn't really understood in those terms. And once you understand in those terms, then that lays the groundwork for applying it to more complicated materials and more complicated systems. If I may, I'd like to give you another example of Mon. Okay. So the most stunning example of Moore is different

is something that all of you know, all of your viewers know and use, and that is AI. Okay? When you ask, as I do, how is it possible for AI to even work in the first place? Right. Okay? Let's think about this. So the first thing you would say is, let's suppose I've got a time series of data points,

you know, whatever it might be, stock prices, who knows? And you say, well, I want to make a model of that. So the first thing you do is you take your data and you say, I'm going to make a model, it's just a straight line, goes through some of those data points. But it doesn't go through all of them, the data points wiggle and twist and turn. And so the straight line, if you ask, does it fit all the data?

Of course, it doesn't. If you ask, does it make good predictions? The answer is, well, you know, not really, because it's too simple a model. So then you might say, look, I've got, you know, 50 data points here that I'm training my AI on. You know, why don't I just use, you know, a hundredth-order polynomial,

a much more complicated equation. That equation will fit every single data point. that you want to train the AI on. Every single point. Every single point. I've got 50 data points. I've got 100th order, polynomial.

No matter how you embedding. Well, I can find a way to, I can find, you know, a lot of data, but I can find a way to make it fit for every single data point, so there'll be no error in the way that it fits the data.

You say, this thing was going to be really great at making predictions in the future, except it's not. And it's not because I've fitted the data, but I've also fitted the noise. And so if you make predictions, they're basically contaminants.

by the noise. So then you would say, well, okay, so if I have a large error, well, I only have a linear fit, that's not going to do very well fitting a complicated data set. I've got a very complicated formula that's also not going to fit very well because it's fitted the noise. Somewhere in between those extremes, there should be a sweet spot where the two things balance out, and that should be the place where you should, you know, try to make your model. That's what AI should do. And that was the conventional thinking, okay? And so you'd say, very, very, very theater, except that's not what happens. In fact, we have, when we fit our data with modern AI, we are fitting it far more than 100 parameters. We're fitting with a trillion parameters.

And weight. Yes, and you're way, way into that regime where you're just fitting noise and the whole thing shouldn't work. That's a mystery. How is it that having such a huge number of parameters can work in principle. It obviously shouldn't, and yet it does. And that's a problem that we have a

simple, at least a very simple the simplest sort of non-trivial model of how AI works with a student here, Chan Lee. But the answer is that there's a phase transition in the statistical physics of the learning process. And that phase transition

has a rigidity, the generalized rigidity, just like the rigidity of movement of solid, which nobody knew was there because they didn't do the calculation that we did. And so we could understand the transition. We could understand... Does it have critical exponents?

Does it have the renalization? It has critical exponents. It has data collapse, all the phenomena that you'll find in my book. And the phase transition turns out to be very similar to the superconducting phase transition. Ah, what's going to be. And there's a lot that can be said about,

if you want to talk some more about that. But my point is, this is an example of more is different. It's not just that you have more things and so you can fit more things. There's actually, it's different is the important thing, not the more. But eventually. And different means that there is a, it's qualitatively different, not just, well, there's a slightly different number, you know. And it's that phase transition, the qualitative difference, that means that a material or a stochastic computer algorithm, say, stochastic gradient descent,

is used to train AI, has new behavior when you go beyond a certain point. And that's the thing that's why, you know, I think more is different is so important. Because it's not just, yeah, having more, having more money is better than having no money. Yes, I could buy a slightly better car. There's a qualitative difference that comes when you have a, and that's the message of Anderson's article. But here's when I push back with respect to you and Anderson. Here's some ice.

If I told you, this is ice that I collected at the South Pole Antarctica. He'd say, no, it's not. It's water, right? And if I had more in that actually, this was chalkfall. I filled it up to the very brim, and then it melted, and now it's this, right? So, underwent a phase transition from the South Pole of the San Diego, right? Now, if I keep putting more and more ice in there, more should be different, right? Nigelie, that's what Anderson's telling me.

But if I keep doing it, it's just going to be more of the same. So there seems to be, yes, I'm. I agree. Moore is different. There's a water molecule. It's not like this liquid in here. But when there, if I doubled the amount of, if I added more and more and more, it doesn't behave quantitatively different from this, right?

So at what point does the more start to be the same? It does. When it becomes a black hole? At what point? No, no, it is different. So the amount of water and ice that you have in there, if you measure that you've mentioned the critical exponents, like say, how the heat capacities diverting.

It's a first-order phase transition, so you don't have critical exponents. But let's suppose we were talking about, say, the magnetic transition. So, yes, you would find that as you, that there's a temperature where the, say, the divergence of the heat capacity, which you will see in an infinite system, it'll literally go to infinity, when it's a finite-sized system, it won't diverge, it will start going up, and then eventually it'll smoothly, smoothly go over. And that's important because you literally see that in, say, granular superconductors. And if you look at a machine learning, it's a neural network, in the eye. ideal case where you have an infinite number of neurons and infinite numbers of data sets

and some appropriately taken asymptotic limit, you can make a very sharp mathematical theory for that. And only in that case, mathematically, do you literally have the ability to say there's a phase transition and non-analytic behavior and so on. If you, on the other hand, make the system be finite, then the computer scientist would call this ridge regression or regularization. then in fact this infinity goes away and the behavior is different

there isn't a transition you won't be able to see that if you put more and more ice in there if you're doing this with a magnet or you're doing it with a superconductor you can do the experiment and you literally can see that only when I go to infinity

do I see the sharp base transition but you know I described it in my book if you get to within 10 to the minus 12 degrees of the critical temperature then you will start to see the fact that you don't have an infinite number of atoms

in your water water. I want to talk about a man you mentioned in your course as well, and you mentioned with great glee that the only man to win two Nobel prizes in physics was a condensed matter physicist. John Mardin. Often here it said that if it wasn't for the laws of quantum mechanics, we wouldn't have had the transistor.

I always have a little bit of problem with that. Because if you look at the first transistor that they built, you know, shockly, it was... Oh, Phil Schultzooz didn't build it. He's just poached in the photographs. Right. Yeah. I mean, the famous photograph

he's sitting down as if he built the thing. It looks like a piece of coat hanger. Green and Britain was just staring around. Like, why we here? Why is he there? Why is he there? That's actually was the reason why Bardeen left and went to Illinois. He was so, he just couldn't get on with Shockey.

Yeah, well, Shockley was a very, very troubling character. The question is, the question is he look at it. It's a piece of chewing gum. There's a coat hanger. don't get into it. There's a rocks in the middle of it, right? It's very unlikely that you'd say, hmm, this is the solution

of the shorteninger equation with Ferming levels. And do you believe that, that we look into the law? Because the reason I'm asking is people say, when we have a theory of everything, Nigel, they'll be able to look into it. And just like they did with quantum mechanics, instead of making transistors, we'll make warp drives and gravitational impellers

and multiverse, you know, teleportation devices. Well, what do you make of this? First of all, is that historically accurate? And you've seen a lot of these people You interacted with the Titans. You are one of the Titans. But tell me, Nigel, do we look into the laws of physics

to get the technology on the screen that you talk about in your course? Is that really what happens? Or do we describe it later on? Most facto by these laws that we discovered. So just with the fact of the first transistor was a big lumpy thing. I mean, we said before that everything in this room is classical. But you knew that I didn't really mean that.

I mean, look at the flowers there. They have color. The only reason that they have color is because of quantum mechanics. Well, these are made of plastic. But anyway, yes. If they were real. Oh, in that case, you gave it up. You blew my secret.

I was a real biologist, right? I can't even tell a plastic valve from a real one. But you said to them. So, yes, it was a macroscopic object, just like your iPad is. But it's operating due to laws of quantum mechanics. And so, yes, semiconductor electronics. It's not like before we understood semiconductor electronics, We could build iPads.

We could, and this thing didn't exist 15 years ago. In fact, we didn't even know enough about the liquid crystal displays, let alone the electronics to go into it and so on. I want to give you a quote from a countryman of yours of some renown who said, I'm very poorly today and very stupid, and I hate everybody and I hate everything. I'm going to write a little book for Murray on orchids, and today I hate them worse than everything,

and I hate species as well. Oh, my God. how do I hate species? Do you know who that was, that British gentleman of some renown, whose father told me, or to his father has said about him, you care nothing except for shooting dogs

and rat catching, and you will be a disgrace to yourself and to all your family. Who was that said about? To Owen? Darwin. So this man, you know, loved life. He created these ideas, and he was, he was such a fascinating character. It's reputed, and okay,

you're gonna disabuse me of this. Again, I'm a poor, experimental cosmologist, Nigel, that you have seen, and you were part of the group or team, perhaps, that is really working to maybe state the limitations of Darwinian or the restrictions on selection. So let's talk about, why is selection, why is biology? You talk about your paper with Woos, is it Woos? Cold Woos. Life is physics.

Yes. Is that right? I mean, besides your blunder about this, a little time. Well, it could be worse. I mean, I could have said that that spherical thing that there was a cow. What relevance is physics have in biology? You make the point in your course, again, and everyone should watch your course

because it's so enjoyable and easy. It's graduate-level course, but let's be honest, you could take it as a freshman. If you're energized and you're willing to do the work, you may not get the highest grade. But you talk about how easy physicists have it compared to sociologists and what you call it

and what has been called the dismal science of economics. I just had Alvin Roth, who won the Nobel Prize in economics a few years back, talking about repugnant markets. It sure seems easier to do that. to do cosmology and try to figure out what happened, you know, 10 to the minus 30 seconds

after the Big Bang. So tell me, anyway, what does physics have to do with biology and what role do you play and perhaps overthrowing this, this irascible, you know, kind of self-loathing men named Darwin? So a lot of people did interpret our work as being against Darwin. Yeah. But that's completely wrong. Okay, same way. Okay.

The whole idea of, well, you know, I'd say it's Darwin and Alfred Russell Wallace. Wallace really was the first person who wrote the paper that was presented at the Linnian Society, and Darwin added his things to it and so on, and the correspondence between them is very interesting in a vision. But let me just say what people mistakenly are referring to. So the usual picture of evolution that people who are not necessarily deeply into biology think about is this. They say, well, you've got your genes, and then you transmit your genes to your children. and they transmit their genes to your grandchildren and so on and so forth, and the genes propagate like that.

And that is indeed what happens. But there's a very fundamental problem. I'm now going to talk about what it is that we actually did. And then we'll talk about whether it's against Darwin or not. Which is just to be crystal clear about it. So then you might ask the following question, Francis Cricketer, is another one of my compatriots. So you might ask yourself the following question.

Could the genetic code evolve? Right? So let's think, what is the genetic code? So just to review some very simple biology, you have proteins that do lots of stuff in your body. The proteins are made out of amino acid. How do you know which amino acid to put into which protein? So then you read your genome and you read sequences of nucleotide bases, which will say UCA and G, those are their sort of abbreviations for their names.

and then you read those, and then you read them in triplets, and then you take each of those triplets, and if you get you-you-you-you-you-you-get fina amymaline, and that's the amino acid that you then put that position in the protein that the ribosome is building in every cell of your body. And the map that tells you take triplets of nucleotides and convert them into one of the amino acids of life, the 20 amino acids of life, that's called the genetic code.

So it's not your genome. People always say, the genome is your genetic code, that's not true. So the question is, well, where did that map come from? There's a very interesting feature about this map. It's called the genetic code. You can write it on

a t-shirt. It's many-to-one code. Because you've got your alphabet of four letters, words are three letters long. So I've got four times, four times four. It's a 64 possible amino acids I forget. But in fact, we only use 20.

So I'd say, well, why 20? Actually, Francis Crick had an answer to that, which I can tell you, if you like. So there's obviously redundancy in this code. So then you ask, well, when did this code, when was it developed? So you go back and you do molecular, what's called molecular phylogeny. There's ways that Carl Woz was the first person to develop, to look at molecular sequences and then, you know, find what they were descended from, and therefore you can work out the evolutionary history of door-life-off. The last universal common ancestor.

That's right. He gets the last universal common ancestor where... So he coined that, or did he coin Archaia? He discovered Archaia. He discovered Archaia. So he started doing this, thinking that there's prokaryotes and eukaryotes.

And then one day he discovered that these things that are prokaryotes, they're not prokaryotes. There's something else in there. What the heck is this? Okay. And that was a methanogen that he'd...

He was doing these experiments. They were very dangerous radioactive experiments. He was doing them virtually alone for 10 years, everybody thought he was off his rocker. And his goal was to simply find a way to map out the evolutionary history of life on Earth. And he discovered a whole new domain of life that people just looked at on the microscope, say, oh, this is a very blobby thing, it must be a bacterium. It turns out have completely different evolutionary history from that. And in fact, we are descended from the archaea. We now know. He was doing this. And as you say, once you

start building these trees, you've managed to discover that you can build them all the way back to about 3.8 billion years ago, and that's the last universal common ancestor of life on Earth. And there's various converging evidences that give you that number 3.8 billion, and some people say it's even earlier, maybe 4 billion years ago.

Here's the interesting thing. How old is the Earth? Well, I want to take a segue because I forgot to give you your gift. I'm talking about magnets. Here's a magnet. And here's a magnet with some gifts on it for you. So these are pre-Earth meteorites.

These are discovered in Argentina. You know, those are yours to keep as a guest on the Into the Impossible Podcast. So the Earth is about 4.2, 4.3 billion years old. These are 4.35 billion years old. So they're quite a bit older, but they date from the pre-super, the supernova that blew up, which, by the way, was the mechanism by which was discovered by more countrymen of yours, one of whom occupied this office, Jeff Burbage, and his wife, Margaret. So I have Margaret's place. These are her photographic plates from Palomar. So we have a lot of things in common.

But, yeah, so the Earth is older than that, but not by much. I mean, life began very early. That's right. That's right. So whether it's 4.3 or some people say 4.5, something like that. As you say, it's very close. And so the thing is this, we know, because we can do the molecular pornography back to that last universe of common ancestor, that essentially the architecture of the modern

cell was already in place 3.8 billion years or so ago. So you say, so wait a minute, you're telling me that life went from nothing, 4.5, Plus of mine is billion years ago, which half that time the Earth was completely uninhabitor, all this Hadesian, and then by 3.8 billion years, you've developed the machinery for replication and... For our first ancestor, yeah. Yeah, all of that.

And somebody, you know, looking at the organisms around about that time, you would see very little, relatively little, in the sense of the global architecture of the cell, different from now. and says the question is, how is it possible for life to have evolved so quickly? So that's the first question.

And Francis Crick was very perplexed by that. Second question. Second question, why is there only one genetic code? Yeah. Okay? We call it the canonical genetic code, and there's minor variations mainly to do with stop codons,

but it's basically the same genetic code. Then there's a third one, which you probably knew that there was only one canonical genetic code for all life and earth. But the other thing you may not know is that the genetic code that we actually have is optimal in the sense that it minimizes errors of translation.

So let's suppose we were in the world of intelligent design and being deliberately provocative here. So you say, okay, Brian, okay, you know, smart guy, you know, lots of things. Design for me a good genetic code. And you would say, well, if I'm going to design a good genetic code, I know there can be lots of errors in reading and... Yeah, there's some redundancies. So 64 minus 20.

So not just the redundancy, but I should make a code so that if you get the wrong amino acid, I should make it so that the amino acid I do get is, in some appropriate biochemical way, which has to be defined, is a decent approximation to the one that I should have got. So it doesn't do too much damage so that the protein has in it the wrong amino acid, but it can still fold and do the thing that the protein is supposed to. do. And if you could create such a genetic code, you would say, well, that's going to be really, really good. That would be the one that I as intelligent designer would choose. Okay. So the genetic code, when you, when you, when you, when you analyze it, you can do this calculation. There's

many different ways you can do it. Every time you do it, you get the same qualitative answer. The genetic code is optimal in the sense of minimizing errors. Okay. It's fantastic. Okay. So there's a free fact. How on earth, Because all of those things have happened. Now, Francis Crick was very perplexed about this. Francis Crick said, look, there's no way that life could have got to this level of complexity in such a short period of time. It must have come from outer space. So eventually, he embraced the Pant-Spermia idea, which, of course, then just pushes the problem off to another way.

Right. The origin of life on Earth is solved, but not the origin of life in terms. Exactly, exactly. But in fact, there's more to the problem than that. There's these other two facts that I've talked about. Francis Crick was also very perturbed because, as he argued in 1968, the genetic code

can't possibly be something that evolves. Right? Because we're going to think about this way. Suppose it does evolve. So think about this. Think about we're doing this experiment, okay? I'm communicating to you in code. Okay? And I write down my coded message and you get the coded message. You use a code book

to translate the message. So that works fine. Let's suppose halfway along, and us doing this and we're in several continents or something like this, I unilaterally decide I'm going to use a different code. Suddenly all my messages are going to stop making sense. You won't be able to

interpret them. So the code book is the genetic code. It literally is to tell you how to translate from the message that is in the DNA and the RNA into the protein that you ultimately going to produce. And so obviously if you evolve the

which means change it midstream, then it won't, then there's a whole, then you'll start getting the wrong proteins and then everything will die out. Okay, so it can't evolve. What we did in this paper was we figured out how to solve all of these three problems. We figured out why the genetic code is unique, why it is optimal, and why it evolved so quickly. And in fact that it really did evolve. Obviously, that the fact is it's optimal, which Francis Quick didn't know, the fact is it's optimal, either you think that it was

intelligently designed or it evolved under selection. So in what sense is this a canonical critique of Darwin? I mean, why do people didn't say that?

I'm not saying, I believe. I'll tell you why. It isn't. I'll tell you why. Because what we discovered was that, indeed, these things would not have happened if you had just,

were you just using the vertical evolution that we talked to at the beginning. You give your genes to your children, they give their genes to your grandchildren, children, and so many as well. If that was the process operative at the dawn of life, it wouldn't have happened this way. But, in fact, what happened was horizontal gene transfer.

Mainly, that genes can be transferred between organisms that are not related. For example, let's suppose we could do this, okay? So let's suppose you decide that you want to learn, you know, renormalization group theory from my book. So you could slog through my book and go to my classes and so on, but wouldn't it be easier if I could just pop out the gene that enables you to do Feynman diagrams and 4 minus epsilon dimensions, let's suppose the word gene for an arbitrage corpse never, and I just give you the DNA and you just take that DNA, put it into the end up to end up and great, I know how to solve the, you know, I can solve fine one diagrams and four minus epsilon dimensions. Whoopi, I can compute critical. It doesn't happen like that

for us, but it does happen in the world of microbes. That's how antibiotic resistance, for example, is transmitted so rapidly. And the reason it happens so rapidly is because when you're transmitting genes in this particular way, you're using a network effect. I can distribute my genes, not just to you, or not just to my one or two children. I've got two children, exactly, two children's not one or two. You can distribute them to hundreds, a thousand. That's how libraries work. Libraries do this. It's a Lamarckian mode of evolution, but it's not, it's still evolution. In other words, only the books that are actually good end up in the library. Only the right physics books.

The physics books will tell you the actual right physics, the story books that are actually really entertaining. You know, so you have a network process, a horizontal gene transfer process, which is different from your traditional view of how genes are transmitted vertically. I have to interrupt. I have to. It seems to me you're like taking a PowerPoint file on a modern, you know, SSD drive, and then putting it into a Windows 95.

computer from 1995 and somehow it's working? Like, how is that even possible? You just get gibberish. You'd get, you know, these glyphs and how is that even possible? So you have to ask what happened at the dawn of life. At the dawn of life, the genotype phenotype distinction had not yet really been clear. The organisms were very porous. They underwent endosimbiosis.

That means that they would absorb one another. And then the stuff inside that, hey, I can take all the stuff. And we know that that's where our mitochondria come from. That's where chloroplasts come from in the plastic flounds and things like this. So life did that and life transmits exchange to genes in that way. And today, organisms do this. I mean, if you sequence the drosophila genome, the fruitfly genome, okay? The Izing model, Valjean, if you will.

It's like Harvard is the UCSD of the East Coast. You will find in it the genome of Wolbachia. It is a parasite, a microbe, a bacterial parasite of drosophila, and it has inserted its whole genome, actually multiple times, into the drosophila gene. And there's many other examples of holismal gene transfer. If you look at the phylogeny of flowering plants, endosperms, very, very complicated. It's not like a family tree. It's a network.

And what we discovered was that the early life evolved. through this, through a network effect, which was called that state of life the progenote. I don't know why, but he did. And then there was a phase transition to an era of vertically dominated evolution. And when I'm talking about what is evolving, we're tracking the genes, specifically the genes that code for the architecture of the cell, the fundamental cellular processes, such as translation, replication, and so on. So that's how we define species today.

And so today we build the tree of life, but there's nothing mandatory that says it should be a tree. And in fact, prior to the last universal common ancestor, it was a network. And as there was a network, it evolves faster. But it is still doing Darwinian evolution, or while it's Darwinian evolution, it's still survival of the fittest or all of that, you know, however you interpret that. It's a complex argument in and of itself. But basically, it's fundamentally, we're not saying anything a difference about. But we're just talking about what is called the mode of evolution. One question about the network.

Does it exhibit things like Risham's law, but the network law, that is the scaling goes geometric, and the reason that it's so fissoned is because of this network dynamics that, you know, lately we've learned about with social graphs, but in fact, we can understand it maybe how successful it is via network theory rather than, you know, just pure genes. Yes. I haven't personally done a network analysis of what kind of network you get from this specific process. I mean, I think the more interesting thing is that there is a network effect. And the thing that Crick had missed and other people had missed was that I can explain by a kind of analogy.

Let's suppose that I drive over to your house and the wheel comes off my Toyota Corolla while I'm there. And I say, well, that's too bad. you say, well, you know, I've got a Tesla in my garage show. I don't know if you have, or whatever you have, you know. But you just take the wheel of that. Well, obviously, that's not going to work, okay? But it was we were doing that, say, 120 years ago, right? It was the dawn of the age of automobiles.

So I drive over to your house with my jallopee, whose top speed is like 20 miles an hour or something like this, and the wheel comes off and it's broken and so on. And you say, well, look, I've got a bicycle. My mod, I got a bicycle. I got a bicycle. So take the wheel off my bicycle and stick it on it. So I just goes out, take the screwdriver out and screw it on, and I'm good to go. And you can do that because the very early primitive forms of an automobile are very, you know, you can just swap things in and out.

The technology is not very advanced. It's not. It's not sick. Right. Precision and things like this. In the early days of living systems, they were very simple. And so they could tolerate ambiguity in the proteins that they use. As they became more and more complex, then, you know, you really have to have just the right protein to fold in just the right way.

to be able to make the thing that goes into your neurons, or something like this. And so what we realized was that you can build a dynamical systems model of the co-evolution, of the complexity of the organisms, along with the evolution of the genetic code. And so you find that then through this network effect, it evolves very rapidly, and eventually gets the point where it shuts off the network effect, and then transitions to the vertical evolution. we're in right now. So does that make you more or less sanguine, you know,

getting back to Fermi's question to our late, great colleague, Herb York, you know, where are they, where are the aliens? Are you more, I mean, knowing this level of, you know, kind of punctuated equilibrium, you also quote, Gould and the fact that, you know, we don't actually have that many more genes or, you know, anything productive compared to a worm, you know, seem to like fewer, it's pretty much the same. And if you really wanted to punch a hole in more is different,

you would say, well, that can't be true because of the same number of genes, it was as Gould wrote in his year 2000, whatever it was, New York Times op-ed piece, there's many more interactions between the units than you have in C. Elegans is so simple that we know every single... Right.

...neuron, anyway, another new one. Every new audience is mapped. Yeah. It's not the case for us. Yeah, exactly. Right. If we spray C. elegans throughout, you know, on the planet, Mars, you know, it's different than spraying koala bears on there.

Where does this leave us? Tardigrades. Well, they're already there. I mean, there's human poop on Mars right now. I guarantee it because the astronauts are spraying out, you know, they bent it out to the space and then eventually gets to, you know, I have a piece of the moon here. You know, this is a meteorite from the moon, so stuff is striking. I think it's such a gradual on the moon.

Yeah, exactly. So I'm sure they're on Mars. But tell me, Nigel, does this make you more or less anguine about life elsewhere in the universe? forgetting or pausing for now the origin of life, generally, but just origin of life specifically on other solar systems, in other solar systems. I tend to believe that life is the inevitable consequence of the laws of physics, which we understand imperfectly.

And I say physics, not chemistry, because I don't think that life is restricted to particular chemistries. It could be silicon-based or could be a different genome. Well, I have a question for you about that. which would you be the ideal person to answer. I do think that it is a physical process, and I can even say a little bit more about why I think that.

I would say that if you want to know what is the purpose of life, what is the meaning of life, if you like, what is the purpose of life? The purpose of life is to help planets come into equilibrium. How I say? So think about a planet. A planet after it's formed, has a huge variety of chemical potential redox. gradients in its environment, and those gradients will eventually relax and homogenize as they show from second morphemodynamics and all sorts of other good reasons, and that happens.

And what life does is life uses information to find new pathways to short circuit, if you will, those chemical potential gradients, and use the energy to power life. And that's how ecosystems work. ecosystems compete with abiotic processes to literally take chemical potential differences and use the flow of energy in them to make

living things and those living things are powered by this chemical potential gradient. So life uses the information just in the way that I was saying with a horizontal gene transfer, that's one very fast way of searching a space and finding

new ways to solve the problems that emerge, the organizational problem that emerge, and we know that that happened. By the way, there's lots of supporting evidence for our horizontal gene transfer theory, and there's a recent paper that just came out in the journal Astrobiology, which is a sort of review.

It's not one that I wrote it with other people, but it's, you're looking back on that, and there's even other data which supports this theory. But the point is, that's what living systems do, and there's nothing special about doing it on Earth, as opposed to Enceladus, which would be my favorite place. where I used to direct NASA Astrobiology Institute.

And I tried very hard before NASA disbanded the whole N.A.I. program, sadly, to persuade anybody who would listen, that the place we should go, not Europa, we should listen to 2000 World of Space and Sea, and give Europa a miss. Go to Ensardos, because there you've got a much better chance. And we already know from the Cassini mission that, you know, you can sample already the water. Yeah, it's there. And it actually looks like alkaline hydrophothermal vent.

So there's all sorts of interesting astrobiology that could be done. That sounds amazing because it seems to me it's closer to answering Schrodinger's question than Fermi's question, let alone that. It's. It's. It's. Carl Wos and I wrote a review article called Life is Physics. And the reason we wrote it like that is the following. First of all, Schrodinger wrote this book called What is Life, which, of course, everybody is inspired by.

And then the other reason is that when I go to astrobiology conferences, you know, on the first day, somebody will stand up and say, well, life is chemistry. But I don't agree any more than I think that a computer is, you know, if you ask me what a computer was in Victorian England, I'd say, well, it's Babbage's machines built out of cogwheels and springs and levers, and you sort of turn things like this and we'll compute. And then love this. And then you, that's right.

Exactly. He's a loveless. And then you go and ask, you know, an engineer trying to figure out how to design hydrogen bombs at the Institute for Advanced Study in the 1950s. And he'll say, well, it's John von Leumann's building it's in that shed

over there it's lots of thermionic valves and relays and that's what the computer is and now you ask

Nazan the answer you ask me you know 20 years ago you or anybody

you say well it's my iPad my windows my MacBook or my my Windows computer

or somebody today what is if it was in my phone or my glasses or something

you know there's a difference between the substrate in which something is made and you know

what it what it actually is. And so when we think about trying to understand the fundamentals of living systems, of course, if you want to know how to make somebody better because they're ill for some disease, well, you better understand something about biochemistry, for sure. But if you want to understand, you know, why is their life in the universe, why does the phenomenon of life even exist?

That is a fantastically profound and interesting question. And we don't, truthfully speaking, I don't feel that we know the answer to it. I think we make steps towards the answer, but I think the answer is it is a physical process. It can be realized in certain types of atoms and so on. But here's a question for you. Could life exist three minutes after the Big Bang? Depends what you call life.

I think the universe did transfer through a period of time when water was liquid. The C&B was once at 300 Kelvin, right? So there's no, you know, that wasn't very that soon after the Big Bang. I mean, I think in terms of atoms forming in 380,000 years. Right. So it's like very, very implausibly, but perhaps, as Deutsch says, you know, if it doesn't violate the laws of physics, perhaps. So let me tell you a science question story. It's not meant to be real.

It's meant to be a thought experiment which is meant to raise your consciousness. Okay. So I'm going to make the following claim, which I am saying, this is not a scientific statement. It's a thought experiment. Okay. That you have at that early stage of the universe, we're way above the physics that applies to the standard model of particle physics as we know it now. And you've got some, I don't know, non-a-bellion gauge fields or some, God-know-know-know-gauge group, or some strings or something like that, and they have non-a-billion flux tubes that go between whatever the quark-like excitations of this thing.

And those things are non-abillion, so they can wrap around, they can store information just like we want to use non-abillion anions to build quantum computers and store information. And so you could store information in this way, and then you could have the chemistry of these objects. And so you could imagine you could build a self-organized object that are built out of, you know, non-Abelian flux tubes. The thing is that they're on the scale of like 10 to the minus 20 meters. And an energy scale of, you know, turn to 100 GV or something like that. And they last for 10 to the minus... Plank times, yeah. Plank times.

Plank time or something like that. But you could imagine that. And you could imagine those being sitting around in their monobillion gauge theory bar, drinking beer and saying, you know, what, do you think life could exist, you know, for 13, 14 billion years after the Big Bang? And they'd say, oh, come on, don't be so stupid. I mean, they'd have to be absolutely enormous. The scales would be enormous.

And the time scales, don't even ask you all the time scales, they'd be just huge. And by the energy scales would be pathetic, you know, what could they do, right? There's a completely ridiculous suggestion, of course, sounds. you know so I think it says that you know when you think about what is the physics of life the processes that are involved

in creating the phenomenon of life you know they're on a sort of logarithmic scale of energy and time and space and complexity and so on and so forth and when you talk about life we usually mean life like us but if you want to ask about life that's not like us well why not and I certainly think that

you're going to find microbial life long before you find dolphins with iPhone And the dolphins are kind of the embodiment of the more is different to me. I mean, I'm sure the listeners can determine for themselves the vast, you know, kind of depth and breath that Nigel engages in. But you're also a citizen scientist in the kind of tradition of our late-grade colleagues like Herb York and Roger Ravell and many others throughout history. And throughout different continents that you've lived on and you've experienced this. But now, you know, I get the sense. For someone as cheerful as you are, I consider you a very optimistic pessimist.

And you're seeing things, and you and I have spoken offline about the kind of precariousness of the age that we live in. I want to ask you, first of all, is it our fault? You know, Catalan Carrico, co-inventor of the COVID vaccine, sat in that chair. And she told me that, you know, we sort of have this inflated view of scientists, and actually we're quite egotistical. And, you know, she went through kind of the negative side of academia. You know, how much of it are scientists to blame ourselves?

I don't want to make it sound like we are even a very large fraction of the blame. But you hear nowadays, and we're talking now, if Fauci is going in front of Rand Paul and there's this big theatrics, I think it's all nothing, Berger. I don't think anything is going to happen. But I do feel like we're living in an anti-science age, but not because of the reason that everyone has their truth. I don't care what you believe. I don't care what you do in your private life. But if we have different epistemologies, that's very dangerous, right? If you want to have different ways of discovering what is true and characterizing what is true, like, you may believe that 9-11 was an inside job and that fire doesn't melt steel and whatever.

And I may believe that, no, no, no, actually, it didn't have to melt. You know, so we have different ways. At least we have the same epistemology. We're using science in some way. I'm not saying that those people aren't crackpots. But, Nigel, are we living in an age that's not like post-truth? It's just, it's relative truth. It's renormalized truth where you can believe whatever you want, Nigel.

I'm going to believe that, you know, that there's something that, you know, that there's something that. various conspiracy and I have different ways of getting to my truth. What do you make of this age? And who's to blame and what can we do? Sorry to wrap three questions and one. But what's going on here? I would say that the, you know, I've been very outspoken and active in trying to defend science in the United States of the last year and a half or so. I would say longer just to give you, just to give credit where I don't think people realize the role that you played in the 2020 COVID, kind of a pandemic and bring quantitative. I mean, you were really at the forefront of being data-driven

and predictive using a lot of your models, which we didn't have time to talk about today, but we'll do a part too. And I think that's remarkable. So I don't want to say it's just Trump. I do think that there are other factors at Bay, but he certainly plays a role as unique now. Yes, but I think that's the biggest threat

that I'd like to talk about. Most of my activity has been really to defend the public interest because I do feel that what is happening to science, and it's happening to science, is not being done in good faith. As I am a member of the National Academy of Sciences, a fellow world's society and so on, I feel that it is the job of scientists to speak out

and to try to work with Congress, in the case of the United States and so on, in order to make sure that they have the best interests, the best information. And that's how the National Academy of Sciences does sound it. by limkingly in 1863 during the Civil War. And so a lot of the work I've been doing, you know,

it's been work that the Academy itself could not do on its own for various reasons, which I wouldn't go into, but is now becoming actually in the way that it possibly can, much more visible in the public eye, so much so that Donald Trump has literally is seeing quotes on social media about, you know, defunding the National Accountancy of Sciences. Beliefly, yeah.

Beliefly. Just after the World Cup or during the World Cup. So I won't say too much about what I've been doing, but I've been doing, but I want to talk about why I think it's important. And the reason is this. It's not that I feel well, I want my money, I want my lab, I want the money coming in that supports the research. That's not the important thing.

The reason we do science is because it is in the public interest. If people like me, I'm actually working on cancer at the moment, trying to understand how cancer works. I'm very excited about some work that we've done. A $11 million grant proposal to NIH from here, which will probably never get funded. But the reason is important. Jay Panacharya, I was the director, is a friend, and he sat in that chair too, so maybe we can talk. Well, just get me going. Anyway, but the point is, it's not a question of, is my personal hobby fund out?

The question is, we're doing this for the public good. All the technology that we have, you know, that we hold in our hands, our silicon security blankets and the medicines that will help us live longer and all those things came from the scientific process. And I believe, as many do, that this is in peril at the present time in this country. And it is our duty. and it's the duty of people who engage the public, like yourself, to make those things crystal clear to people so that people understand what is happening, who stands to benefit from this,

and why it is not in the public interest? Is the scientific system perfect? No, no system is. Scientists are being attacked in the media. If you poll people and ask who are the most trustworthy people, politicians are right at the bottom. scientists and teachers are right at the top. You shouldn't be weaponizing the inevitable flaws in the system, like peer review, is it, is peer review broken? Is it really true that we've stopped innovating in science, which I think is complete nonsense? But that's the rationale that's being used by Michael Kratzios in particular, the director of the Office of Science and Technology Policy in the White House, for saying we need to take science away from the universities, putting more into industry and so on, towards an investigation. individual scientists, right?

Yeah, none of which makes any sense. Or benefit AI. AI is science now, according to all the David Sachs's and the advisors to the president. It's very troubling. So what is alarming is that these are interesting discussions, but they're not being held in isolation by disinterested parties arguing in good faith.

And I think that to me is a problem more than someone who had difficulty getting tenure and it ends up winning a Nobel Prize and so on. Yes, we know that there's examples like that and I've been, I was actually very lucky spending the first half of my career,

the first 36 years, at the University of Illinois, where I was essentially working on the lunatic fringe end of condensed matter physics, but people like twice Nobel Prize winner in physics, John Bardeen, said, okay, I, you know,

I support what you're doing. gave you the encouragement and, you know, yes, go ahead and do this. You know, I was working on high-temperature superconductivity, and I had a view on it that was completely not shared by anybody else in the community for five or six years until eventually we could prove that it was correct, and they named the D-Wave as a D-Wave on nature of the superconductivity. John Bardeen was the first person who said, that is wonderful, and he gave me moral support and told people, we should follow that.

So I do know what it's like to be an outsider. Yeah. But I've been very lucky that I've been able to do enough things that are sort of mainstream, as it were, that even though I don't stay in my lane, I've been supported. And it's not true of everybody. I think if I'd been at another university, we wouldn't be sitting here talking now my career would have been very different. I think you're absolutely right. And to use a phrase from the namesake generator of this podcast and the... the namesake generator of the word podcast in general, Arthur C. Clark,

he's any sufficiently advanced technology is indistinguishable from magic. I want to ask you two questions, kind of as we close, that are prompted by him. And that's the first one. What is sort of the most magical? I mean, we talked about so many marvelous things today. And literally, we've scratched the surface. I feel like you and I could talk for hours, and hopefully we'll get another chance. You're the second condensed matter physicist from UCSD physics department after Jorge

Hirsch to come on, and he's been a two-time guest, so you have to be a multiple-time guess. Can't let Jorge have all the fun. But Nigel, tell me what is the most magical? If you could put something on a monolith and launch it into space for four billion years, what would it be? What would encapsulate, as Feynman said, the most information in the fewest lines of text or code? Well, Feynman's answer to that was that atoms exist. And I guess my answer would be more is different.

Because it's not enough just to know that atoms exist. Very good. Everything that we've talked about are emergent properties of different levels of description and so on. And I would say, you described me as a condensed matter of physicist, and that's where my intellectual roots are, but I work in, you know, astrobiology and evolutionary and food mechanics and all sorts of other things. But I think it's the recognition that there are emergent phenomena,

which I think is not a philosophically obvious thing. And if you don't know that, then many things in the universe are far more perplexing, then they would seem to be. So I'll give you an example. Humans try to figure out how the world works, and so we came up with one answer, oh, there must be a God that makes everything do the things that it does.

I'm a practicing Jewish atheist, okay? And I don't believe in God, but I do think that you see in, you know, what you see in society and in the world around you, phenomena that are seemingly inexplicable, the hidden hand, as Adam Smith called it, about gap economics. But you see the same thing in all aspects of human life. I'd say, well, that's really God. I'd say, well, it's an emergent aspect of things. And I think this sort of motif really, there's more of this difference as you brought it up.

That really does have many, many ramifications beyond the most trivial ones. So I think that would be, that would be my answer. That would be that. Okay, last question. Anybody who's sophisticated, you have to be able to read it. Whether they know it. That's right. Could he see this monolith or no?

No, all the sassel, I should send it off into the space. So they get two finalists. That's even more exciting. Arthur C. Clarke said the only way of knowing the limits of the possible is to go beyond them into the impossible. That's the namesake of this podcast. They didn't give her of the podcast. If you had 20 seconds with a 20-year-old Nigel Golden-filled, what would you tell him? What would you tell him to give him the courage to do what you've done, which is to be a remarkable scientist, but a citizen scientist as well?

I would just say that you can do this. I don't think it's true that you have to be a genius to do good science. It might help. It may not help. It's not obvious that it does. It depends on how you approach things. You know, Einstein, who I think was, could have won seven Nobel Prize,

I can listen for you. I don't know that he was smarter than anybody else, but I think he had a better algorithm and a better approach. And so I think the, you know, the question I always ask myself and I ask other people this, I ask other scientists I meet,

how do you choose the problems he work on? Yeah. What is the way you decide what to work on and what not to work on? That's right. And so I think that's what I would. That's a matter of taste, yeah. It's not a matter of taste. I don't agree with that. It's a matter of how you can make the biggest impact and increase the likelihood of making discoverers. If I could, yeah, more than 20 seconds, I would say it like this. The impact you make is the ratio of what you do, derived by what everybody else does. And the usual algorithm that people have is, well, I'll try to maximize the numerator, you know, but that's limited by things like, you know, how much funding you have, which university are, how smart you are, how smart you are, all.

your family circumstances, a million other things. But the better strategy is to minimize the denominator. Don't work on something. I don't work on anything if I think that if I didn't do it, somebody else would do it

a few weeks later. You said, yeah. You only work on, that's my philosophy of writing books. I only write books that only I could write. That's right. When I talk to students,

I often try to give when you ask me for advice, I tell them, you know, do something different. It's not that more is different. It's the different as more. That's beautiful. We just said the title of this episode.

Roger Goldenfeld is so proud and happy to have you as a colleague. And I have a question about the Izing model applied to cosmology, which I'm going to run by you on the Blackboard outside. But Nigel, thank you so much for joining us. Thank you very much for your interesting questions and for having me on your show. Hopefully this will be part one of many, of more, many more. Thank you.