Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas - AMA | March 2026

Episode Date: March 2, 2026

Welcome to the March 2026 Ask Me Anything episode of Mindscape! These monthly excursions are funded by Patreon supporters (who are also the ones asking the questions). We take questions asked by Patreons, whittle them down to a more manageable number -- based primarily on whether I have anything interesting to say about them, not whether the questions themselves are good -- and sometimes group them together if they are about a similar topic. Enjoy! Get twenty percent off your first purchase at Fast Growing Trees when using the code MINDSCAPE at checkout! #sponsored Blog post with questions and transcript: https://www.preposterousuniverse.com/podcast/2026/03/02/ama-march-2026/ Support Mindscape on Patreon.

Hello, everyone. Welcome to the March 26th, Asby Anything Edition of the Mindscape podcast. I'm your host, Sean Carroll. Various things come to mind whenever I start the intro for the AMA, because when I do an intro for a regular episode of Mindscape, I can just talk about what we're going to talk about in the episode. But the AMAs cover everything, so I sort of randomly pick things, sometimes affected by the state of the world. The state of the world right now is pretty bad in various ways, as I am. saying these words into the microphone. So instead of talking about that, I thought I would just mention one of the various things on my mind from a kind of research point of view. I don't like to talk too much about the research that I'm doing that's ongoing because when research is in progress, you never know when it's going to end up and you say one thing and then, you know, a month later it turns into something else. But we're close to finishing a paper that I'm a co-author Ron. It's actually a big team of eight people led by Fernando Rosas at Sussex. And it's kind of a review article more than an original research article, but it's about what information means, the concept

of information, because information theory, the use of information is very common. It's a hot topic in various circles, but it means different things in different contexts. So we thought it would be useful to clarify that. So under Fernando's guidance, we are identified. basically four different faces of information. The engineering face where you talk about how information is contained in messages that you communicate with and you want to do that effectively in the good old Claude Shannon way of thinking about things. But there's also a statistical face where you're not thinking about messages necessarily,

but just kind of like the theory of information to study the dependence or interdependence of many different variables that you can use to assess. the organization of systems or something like that. And then there's, of course, the thermodynamic phase, right? We have established this connection between information theory and statistical mechanics, and therefore thermodynamics. Information is a resource that we use to talk about thermodynamic processes. And finally, the ontological phase,

because there are people who think that information underlies the meaning of reality, right? the It From Bit point of view that John Wheeler famously put forward. And all these are valid ways of thinking about information. But besides this review article that we're putting the finishing touches on, I hope to make it public soon. We hope to make it public soon. It's been accepted to the journal, but we've been sharing it with people and trying to fine-tune it.

It fits in very well with the complexogenesis stuff that I've been thinking about. I did a solo podcast about this idea of how complexity comes into the universe over time. And one of the themes there is that as the universe undergoes these transitions of increasing complexity, one way you can think about that is from this thermodynamic point of view, that information is a resource and that we have a lot of information given to us by the low entropy state of the Big Bang, and that subsets of the universe learn to use information, used to learn how to leverage that resource in more and more sophisticated ways. You know, the sun in some ways uses information,

but you wouldn't even say that because it's just so trivial. Really what it's doing is it's burning fuel, right? And there's this connection between information and free energy and therefore fuel. But a living being uses information. which is still the low entropy thing. It's still the same thing that the sun uses, but it uses it in much more sophisticated way to think about and model its environment,

to predict the future, and so on. What we haven't quite done yet, although we're thinking about it, is ways to really make this connection precise to sort of label, give equations that tell you the way in which, let's say, a living being,

uses information in a more sophisticated sense than the sun does. But it makes me wonder, of course about computers and artificial intelligence. Like if you, if this theory was really good, maybe we'd be able to tell you what the next phase transition to even better complexity would be and how that would be characterized by using information in an even more sophisticated way. Now, I don't really see evidence of that in sort of the modern AI revolution, which, of course,

we're going to talk about that more in the AMA because people have questions about these things. but that's not quite the angle that people have been taking in talking about LLMs and so forth. Maybe they're better at this or that benchmark, but I don't think that the ways in which LLMs or other approaches to AIs are using information as a resource is qualitatively different than the way that human beings use it. Is there a way to use information that is just a truly different way than how human beings use it? can we build computers that kind of leverage the fact that the computer algorithms are designed in some sense, rather than just evolving organically, to be more clever, to use information to do something that human beings simply can't do,

not just can't do very well. And again, I'm not saying that modern LLMs and so forth are doing that. I think the opposite is true. But maybe we should think about that. Maybe we should think about what it would mean to have a truly different kind of information processing complex sophistication in the universe. Anyway, I have nothing more very productive to say about that. I just thought it was something to let you in on that is a little bit of a distraction from politics and foreign policy and various other things going on in the world right now. So thanks as always to the Patreon supporters of Mindscape who enable these AMAs to happen.

You could be a Patreon supporter if you go to patreon.com slash Sean M. Carroll and throw in either a dollar per episode. I think that Apple is going to be forcing us to go to a monthly system. So it'll be like five bucks a month rather than $1 per episode. So that's inflation for you. There you go. Blame the administration. No, actually that's one of the things you can't blame the administration for in this case. But the people who do that, who pitch in on Patreon, and I know not everyone,

can or once do, which is fine. Most people don't. But those who do are powering these AMA episodes and asking the questions and getting special little reflection episodes and get to ask priority questions once per their lifetime. So it's a good deal overall for relatively less money than you spend on your typical morning coffee. So with that, let's go. Svoraanga Sarma asks a priority question. Remember that priority questions are the little privilege that Patreon supporters have. Patreon supporters always get to ask the AMA questions, but they also, once in their lives, get to ask a priority question that I will promise to do my best to answer.

So, Solanga says, you've mentioned that during your time as a science advisor for Avengers Endgame. You proposed a version of time travel that the filmmakers ultimately found too elaborate or complicated for the movie, even though it was very consistent. Without breaking any NDAs, could you share the core physics of that idea? What was the specific mechanism or consequence that was too much for Hollywood? And how would it have changed the movie's logic compared to branching timelines we got? I don't think I'm breaking any NDAs here because where they went was pretty different than what I suggested. I mean, to their credit, they listened to what I said.

And what I said, you know, what you want to do when you're doing science consulting is put forward your own ideas, but also really pay attention to what they are looking for, right? I mean, you're there to serve them. It's not your movie. You're trying to help the people making the movie. And so what I really emphasized was the movie would be better if the time travel was mostly consistent. If whatever actions were taken by our heroes traveling backward in time were completely consistent with the present as they understood it. And that was mostly obeyed. It was pretty darn close.

But the other thing that we talked about was the idea that there were these separate times. timelines, right? If you, it wasn't developed super far in endgame, although it was there. The ancient one explained, I guess, to Bruce Banner that, or the Hulk, that there were these different timelines and it was important, et cetera, et cetera. The conceit, that concept was used much more explicitly in Loki, the TV series. And what I have said many times, and so this did not get implemented in the Marvel Cinematic Universe, is that, you know, if you believe in multiple timelines, you really should think that every one of them exists and has moral status, right? You know, we tend to live in one timeline, and the usual movie thing is that the timeline we live in is the one that matters. And the other ones are sort of there and slightly annoying, but ours actually matters.

And you see all sorts of terrible things going on in other timelines. But the idea that you just sort of eliminate another timeline is the most monstrous event in all of history. You're literally ending an entire universe from existing, right? Maybe that universe isn't quite as fun or cool or successful as yours, but still there's a lot of people living in that universe. and it's genocide on an unprecedented scale to just get rid of it. So what I suggested, if they wanted to sort of confront that moral reality, was, and by the way, this is completely consistent with my other stated idea that I don't care too much about what is going on on other branches of the Everettian wave function.

The difference being that I can't affect other branches of the Everettian wave function. Of course, I want people to be happy in those branches, but it doesn't stop me from saying that what I care about when I make predictions is just the probabilistic future that I might actually, that I and the people who I know might actually experience. That includes all the other Everettian branches, but I can't do anything about them once they have decohered from ours. Anyway, so what I suggested was if you want to have other timelines where other things are going on and you want to sort of resolve them in some sense without literally killing billions or. trillions or quadrillions of people, what do you have to do is have those timelines merge together. That's easy to say, but the specifics are kind of hard. If two timelines merged together, you know, these timelines only branched off, what, five years ago,

and now they're remarging. So you have people who are the same descendants in much the same way as you get in good old ever-edding quantum mechanics. They had the same past, but they had different experiences for the last five years. So what do you do about them? Is it that one set of memories get sort of overwritten on the others? Then that's just killing the people whose memories get overridden. So what I suggested was that when you merge universes, that the people who were alive before the splitting had both sets of memories still intact.

So that there was some kind of vague overlap or at least maybe one memory was dominant but the other one was still there. and accessible, and people had to deal with knowing what it was like in all of the different branches of the universe that they were in for those five years, or however much time was spent in between the splitting and the rejoining. I'm not sure how that would work. I'm not sure if it would drive people crazy, but that's what makes it fun. If the movie had been about that, it would have been a very different movie. You can imagine like a small indie movie.

We just rewatched Memento recently, one of Christopher Nolan's. first movies. And that was back when Christopher Nolan was a small indie director. And that kind of film that really took that possibility seriously, living with two pasts as a single person, would be a nice low-budget psychological thriller. Very different than what we get in endgame. And they were not ever going to do that. And I never thought they were going to do that. But, you know, it's my job to give them the options to think about these things. Rory Cochran says, if the speed of light is a universal speed limit, does the size of the universe mean it's impossible for the entire universe to embody a single brain-like consciousness

because it would simply take far too long for any kind of signal to pass across it informed thoughts. Would the first and last thought be, oh no, heat death? Yeah, definitely. This is something that is a very real limitation on fanciful imaginative scenarios in which you have galaxy or universe-sized brains. The universe isn't old enough for a brain to be the size of the universe. You know, you want there to be enough time to have some thoughts. Not only that, but you want to have some time to assemble this conscious creature, right?

So it's completely out of bounds to think that things like the universe could possibly be conscious in the same way that a human brain or a living brain in a terrestrial organism is conscious. For that matter, even the galaxy, I mean, our Milky Way galaxy, which is relatively small compared to the, universe, right? There are hundreds of billions or a trillion galaxies in the observable universe, but I vaguely remember once calculating, you know, given the speed of electrochemical transmissions in the brain and given the speed of light and comparing the size of the brain to the size of the Milky Way galaxy, you know, how, what is the subjective amount of time that the galaxy, if the galaxy of the Milky Way galaxy was a single conscious creature, that it would have had to do its thoughts. How much time has it had to think in the same way that a human brain thinks? And the

answer, I think, is like a few hours or something like that. Now, you might say to yourself, well, okay, but a few hours is pretty good. Like, I could have some thoughts over the course of an hour, but the galaxy is not arranged in the form of a brain or anything conscious like that. Even though you can have some serious thoughts over the course of a few hours, it took billions of years to evolve you, right? And so all of those timescales have to be boosted up if you think of not just a human brain, but a galaxy-sized consciousness. So again, nothing like a human consciousness could possibly be instantiated in the form of a galaxy. It's far too easy for human beings or science fiction writers to imagine things that are kind of like us human beings but are

much, much bigger or much much smaller. But in fact, the laws of physics get in the way of those kinds of thoughts. They at least make things very, very different. It's time your hard-earned money works harder for you. With the wealth front cash account, your uninvested cash can earn up to 3.95% APY. That's a 0.65% boost over an already high rate for three months, just for being a new client plus free instant withdrawals to eligible accounts. Start today at wealthfront.com. 3.3% base API via program banks has up January 30th, 2026. It is representative variable requires no minimum and is earned on funds swept to program banks, boosted up to $150,000, cash account offered by Wealthfront brokerage LLC member FINRA SIPC, not a bank.

Okay, I'm going to group two questions together. Mars and Chady says, I've read quantum fields twice and I'm still confused. At some point, you switch from the language of the wave function to the language of quantum fields. I kind of get each of them in isolation. but I'm hazy on how they inform one another. To make my question more concrete, can you illuminate how decoherence happens in the language of quantum field theory? And then S. Sanders says,

what is the relationship between the quantum fields of quantum field theory and the Everettian wave function? My limited understanding is that matter and force fields are described by vectors in actual space, whereas the configuration is described, whereas the wave function is described by wave dynamics in an abstract configuration space. So, yeah, both of these are asking a perfectly decent question about the relationship between quantum fields and the wave function.

So they're actually quite different. They're different things. They're different both sort of technically in the mathematics we use to express them, but also ontologically or metaphysically. Like, what are they? What role are they playing? And to answer the question, it's first useful to just think not about fields, but just to think about particles. Okay. In classical mechanics, you have particles. Particles are the real thing. They have locations. They have momentum. If you are very good at math or you are Laplace's demon, you can predict using the equations, what they're going to do next. The role of particles in ordinary non-relativistic particle quantum mechanics is quite different, even though we still use the language of particles. In non-relativistic point particle quantum mechanics, We have the wave function.

That's what Schrodinger wrote down just for a single particle, and then you boost it up to talk about many particles if that's what you want to do. And it's the wave function, if you're an Everettian or, for that matter, a boehmian or an objective collapse person, it's the wave function that is real. The particles are how you build the wave function. In other words, you start with these classical particles. You say, okay, the configuration space of these classical particles or a momentum space. it doesn't matter, you pick one, and then you assign complex numbers to every point in configuration space.

And so you, of course, measure quantities of the wave function or qualities of the wave function, and you see particle-like behavior. So that's why the very short motto for quantum mechanics is that when you're not looking, it's a wave, when you actually look, it's a particle. But this wave, the wave function, is nothing like a field. It's not like a field either classically or quantum mechanics. The wave function is just a different kind of thing. And you know that because the whole point of a field, like the electric field, of the gravitational field or whatever, is it's a function of space time.

At every point in space and time, there is a value of the electric field, there's a value of the magnetic field, of the gravitational field, or whatever. That's not true for the wave function. The wave function is a function of the configuration space of whatever classical system that you started to build your quantum theory on. And now if we get into slightly more careful about it, you don't even need to start with a classical theory, but you do usually. So we can still talk that way. So if you just have two particles in three dimensions, at one moment of time, the wave function lives in a six-dimensional space. Okay. It lives in the space of X-1 and X-1, where X-1 is the coordinate of particle one and X2, X2 is the coordinate of particle two, and both of those live in three-dimensional space.

So the wave function is not a function of space time. It's a function of this thing you make out of space time by having multiple copies of space at different moments in time. And you call that big thing the configuration space. If you have 10 to the 10 particles in three-dimensional space, then the wave function is a function living in some three times 10 to the 10-dimensional space, much bigger than the three-dimensional space in which you and I actually live. Okay. So now you can talk about quantum field theory. it's the same exact thing that you have. The quantum fields are just the fields that you make the wave function out of,

exactly like the particles were the things you made the wave function out of. So they're very, very different things. The quantum fields we talk about are the sort of descendants of the classical fields that we start with. You have an electric field, a magnetic field, a gravitational field, a Higgs field, etc. All of those are fields in the usual way. So the wave function is not a field. A field depends on space and time, and that's what it depends on. But the electric field, even in quantum field theory, depends on space and time, except that it doesn't have a definite value.

Just like a particle in quantum mechanics, ordinary point particle quantum mechanics doesn't have a definite position. So the values of the fields throughout all of space are analogous to the location of the particle in point particle quantum mechanics. And the quantum wave function in quantum field theory is a function of every possible value that every possible field can have at every possible point in space and time for that matter. So it's a very big space. It's very infinite dimensional, this space that the quantum wave function is a function of. Sometimes it's called a wave functional rather than a wave function just to remind you that it lives in a very big space. but it doesn't really matter that much. And what the wave function or wave functional does

is it assigns by taking its absolute value squared a probability to the question, if I were to observe the field simultaneously at every point in space. So measuring not just its value at one point, but its entire profile throughout the universe, what probability would I have of getting some specific answer? And of course the probability for any specific profile

is essentially zero, but then you have a little error bar in there and you sort of integrate and the quantum wave functional tells you the density of probability within some error bars or something like that. So they're completely different things. You can still talk about fields in quantum field theory, but you talk about them in exactly the same way that you talk about particles in ordinary quantum mechanics. They are number one, the thing you build the quantum wave function out of, and number two, they're things that show up when you do measurements in some well-defined way.

The Memes of Destruction says, my question is, if the Higgs field is the one elementary scalar field that we believe is everywhere, we get a value of 10 to the 55th times the measured vacuum energy density. I understand there are cancellations, but I'm still confused over just what is canceling out these energy contributions. Could you please help shed some light on what is going on? Well, yeah, I think that the first thing you say is a little bit, too quick, a little bit too glib. What's going on here for people who are not up on this is

in empty space, according to general relativity, Einstein's theory of general relativity, the energy density in empty space matters. Okay. In Newtonian physics or in special relativity where gravity is not there, the energy density doesn't matter in empty space, because really the only effect the energy density of empty space has is on gravity. It's the cosmological constant It has this cosmological effect of making the universe accelerate or recalapse or whatever, depending on its value. So it's important when you have quantum fields in a curved space time to figure out what the vacuum energy is, the vacuum energy density, the number of ergs per cubic centimeter or whatever you want to have. Nobody has a way of doing this. Nobody has some principled calculation that in general tells you here's the vacuum energy.

What you can do is you can estimate, given the parameters that are in your particle physics theory, what are the natural values for different contributions? You think of the vacuum energy is coming from different contributions. There's the zero point energy just from quantum fluctuations. The special thing about the Higgs field is that in addition to the zero point energy, which, by the way, is infinity, if you do a natural estimate of the size of the contribution to the vacuum energy, but then maybe you have some cut off at the plank scale or whatever, so you can make it finite, but still enormously big. The Higgs field has a non-zero value in empty space, unlike most fields, and therefore not only does it contribute to the vacuum energy through its zero point energy, it also contributes to

the energy density of empty space just through having a value. And the Higgs field has a potential, and we know a little bit about the potential energy function of the Higgs field. That's how we know that it breaks asymmetry, it gives masses to other particles and things like that. So roughly speaking, the electrons and quarks in the universe are traveling through a non-zero Higgs field in empty space, and that's affecting their properties and giving them mass. So the thing about the Higgs field, the reason why it has a non-zero value is because it's minimum energy that it can have as a function of the value of the field is not at zero. Ordinarily, when someone says I have a potential energy for a field, the first thing you think of is kind of a parabola, right?

A quadratic potential. If your field is called phi, your potential looks like phi squared. And then the field wants to sit at the minimum of its potential, which is at phi equal zero. And you can sort of naturally set that to be zero. So if the potential is literally phi squared, then the minimum is at zero energy. Of course, you have quantum corrections to that, like we just said. but we're ignoring those for right now. The Higgs field isn't like that.

The Higgs field has this Mexican hat form, and its minimum value is not at zero. It's out at some non-zero value. Now, what we haven't mentioned yet is you're allowed when you're inventing the theory to add an arbitrary constant to the cosmological constant. You might think that there's a mechanism that picks it, whether it's supersymmetry or quantum gravity

or the anthropic principle or whatever. But if all I'm doing is inventing a theory, I can just add a constant. So when I say that the minimum value of the Higgs field, its minimum value of its potential, is not at Higgs equals zero, but Higgs equals some non-zero value, a couple hundred GEV, if you must know,

that's the expectation value of the Higgs field in empty space, that doesn't tell me what the energy is. It just tells me the difference in energy between what it would be if the Higgs field were zero and what it is when the Higgs field is non-zero. That difference in energy is enormous. It's way bigger than the observed cosmological constant. If for some reason you thought that you were supposed to follow a rule

that says the energy density of the Higgs field, when Higgs equals zero, is zero, then you would strongly predict that the actual contribution from the Higgs field, because the Higgs field is not zero, is this enormous number. And it would be negative because it's lower than zero, right? But that's okay. You can predict a large negative number also. But none of these are actual rules.

There's no rule that says that when the Higgs field value equals zero, the energy density should be zero. I can just make up a rule that says the energy density is zero at the minimum value of the Higgs field energy. I can just make up that rule. So it's not true that if the Higgs field is the only fundamental scalar, you get a certain amount of energy. There's a scale that comes in that would be sort of a natural average value, and that is indeed

much, much bigger than what we observe. This has been known for a long time, of course. Phil Anderson, who arguably was the first to come up with the idea of the Higgs field, but he didn't really develop it for various reasons. One reason was because he was a condensed matter physicist, more interested in superconductors and materials than in elementary particles. But the other reason was because he knew there'd be this energy. And he thought, well, that seems wrong.

So that's probably not what is going on. You don't get points for being too quick to come up with objections to your own theories in particle physics. So if he had just said, well, yeah, it seems like there's this huge energy, but I don't know what to do with it. So I'll just set it to zero. Then he would have invented the Higgs field. He would have won another Nobel Prize for doing that in addition to his Nobel Prize for condensed matter. So we don't know why the Higgs field scale is so much, much bigger than the actual measured vacuum energy. That's the cosmological constant problem.

Why is the total cosmological constant so small? It's not just the Higgs field contribution that is much larger by its natural value than what we actually observe. There's many different contributions. All of them are big that somehow they add up to a total value, which is small. and that seems weird. We don't know why. That's the cosmological constant problem. Maybe it's because we live in a multiverse.

Who knows? Certainly not me. Muffin says, if you were to time travel back in time to some pre-quantum and pre-relativity era of history, how significantly would you be able to shift the course of physics?

In this world, would modern physics students now be learning about Carroll's theory of generality, the Carol exclusion principle, Carol's cat, etc. Yeah, sure, I think almost any working physicist

transported back into the past would be able to become super famous just by saying things that they learned in textbooks rather than inventing themselves. Now, there's a little bit of a caveat there. If I were transported back to the 17th century, the time of Isaac Newton, I could, you know, make some suggestions

about classical mechanics. I could have, you know, preempted the Principia. Probably, you know, suggested people that they could use calculus and things like that. But it would, I don't think I would have succeeded if I had tried to introduce general relativity or quantum mechanics, there just wasn't enough empirical motivation for those ideas at the time. I would have sounded like a crackpot if I tried to push quantum mechanics on people

in the 1600s. So I do think that any working physicists could go back to any previous period of time and with their current knowledge come up with what seemed like great ideas, but the timing of exactly when they existed would affect how much impact those ideas had. T.J. McMorrow says, you managed to speak to many experts in diverse fields. Have you noticed any surprising themes or motifs that recur across several of them? So, yeah, I have some trepidation including this question. I mean, I think it's a good question, but it's a good question in principle. It's the kind of question I'm bad at answering.

I don't collect those sorts of insights in any way. And I've mentioned this before. Like, I like to think that I think and I get ideas. I make progress in understanding the world better and better, but I'm not very good at remembering when I had ideas or noticing things or anything like that. I just tend to absorb them into my background knowledge. I do think it's surprising. I don't know if this is a theme or motif, but we have noticed on the Minescape podcast that a shocking number of guests who are not professional physicists nevertheless did have an undergraduate degree in physics or some sort.

Many of them actually did undergraduate dabblings in philosophy of one way or the other. And why I think this is funny is because it's not that I'm choosing people on that basis, right? If I choose economists or novelists or whatever or biologists, I'm not looking for people because they did physics or philosophy as undergraduates, but very often they did. And I think what this is reflecting is that I am not picking people randomly either. I'm picking people who I think that I will have an interesting conversation about within the realms of biology or economics or whatever it is. And it turns out that there are common characteristics in me and those people that lead us to be interested in physics as well as whatever it is that they're an expert in or physics or philosophy or complexity or any of my favorite kinds of ideas. So I think that that's a reflection of me more than a reflection of the world or of good, ideas. The other thing I guess is, you know, I think that the people who I like, the people who I

really like enjoy talking to, and again, this is a selection bias on my part, is people who are curious, people who want to ask new questions about things, people who want to ask the big questions about how things work, people who have patience to dig into, you know, why is it this way rather than that way, right? I think that that correlates very strongly with having that philosophical cast of mind. You know, sometimes I get asked, what are the ways in which you determine whether some person who is trying to talk to you

is maybe a serious person but doesn't know a lot versus just being a crackpot? And there's a lot of people out there in both categories, so it is important to be able to distinguish between them. And the single most important thing, the single easiest way to tell the difference is are they genuinely,

curious, right? Are they actually asking questions hoping to learn something from you? Are they actually interested in what you have to say, not just in their pet little idea? Do they actually try to improve their own ideas? Do they look for the flaws in their own ideas? You know, I did once get flown out. When I was at the University of Chicago, I got flown out because a donor to the university had a friend who had a theory about physics. And so the donor said, like, you know, could we just get a visit from a physicist to talk to my friend who is, you know, has this theory of the universe. And I got picked. I was young and impressionable at the time. So I flew out the guy, the donor guy, was very nice. The friend, well, you know, it was just a crackpot. He was also

nice. He was nothing bad or mean about him. But he had a theory of confidence. cosmology. And the reason why I just instantly could tell that it was a crackpot rather than a sincere person is because he had, well, not only did he have no interest in learning about regular cosmology, he also had a conviction that he could explain everything. That's the other sign that you can get someone classified as a crackpot. They're not just explaining one important thing. They're explaining everything in the universe all at once. And anyway, the point is that just to make sure, and to try to illustrate to the donor guy what I was up against here, I would say, well, what does your theory predict about this question? And the guy would say,

oh, it predicts this. And I would say, well, okay, that's not what we actually observe in the universe. What we observe is this other thing. And the guy would instantly go, oh, no, that's right. That's what I would predict after all, right? There's a very long-winded way of saying that that's not what you want in a podcast guest, an expert in diverse fields, no matter what field you're in, you can always be trying to understand things better, trying to find the flaws in your own ideas, trying to expand your list of interesting inputs that you might have. That's what I look for,

and I think unsurprisingly, that would be a set of common themes that appear cause many, many different guests. Julio Contio says, my question is about the not so far in the future formal education. Do you think classrooms in co-location and books will be replaced by videos online, meetings, podcasts, et cetera.

I don't think we placed, no. I do think that there will be a segment that grows where you get some kind of education, some kind of formal education, that is not personal contact based, right, that is based on videos or online things or podcasts, probably not, but, you know, various pre-recorded things or, for that matter, various AI interactive things. there's a slight analogy, not a perfect analogy, but a slight analogy to the Swiss watchmaking industry, which I'm a fan of. As I've often mentioned, when I wrote my first book from Eternity to Here, my first trade book, I got interested in timekeeping and then watches, and it's a terrible thing to get interested in because it's an expensive hobby. So I try to keep that under reps, but under control, I should say.

But in Swiss watchmaking, you know, if you were wearing a wristwatch in the 1950s, you were probably wearing a Swiss mechanical watch or maybe like a Japanese mechanical watch or something like that. They were mechanical. They had little springs in them and little gears and balances. And then they invented the quartz watch. And there was what was known in the Swiss watchmaking industry as the quartz crisis. Because quartz watches are both much cheaper than mechanical watches. They're also much better at telling time. They're much more precise and accurate.

So why in the world would anyone want a mechanical watch when you can have a quartz watch, which is uniformly better on all the usual criteria that you would have? And a lot of the Swiss watchmaking companies tried to not only start making quartz watches, but literally dismantle the very specialized and hard-to-reproduce machinery they had to build some of their mechanical watches. But before that transition was complete, some segment of the Swiss watchmaking industry managed to realize they could rebrand themselves as a luxury item.

So basically, in response to the fact that nobody wanted their product, they multiplied the price of their product by a factor of 10 or 100, and then people did want it. Now, it's not quite as craven as it sounds. A good Swiss mechanical watch requires an enormous amount of human effort,

to go into it, to make it, you know. You can mass produce them, of course, and that does happen. But then once you're a luxury item, the people who are paying attention to you, notice all the subtle differences between one that is mass produced and one that has a lot of human hand craft involved in making it, et cetera. And you can charge more money because you're getting a different kind of thing. If you know an expensive Swiss watch, what does the expensive mean? You can easily get a watch for $1,000.

If you want to get a fancy watch, you're talking about $10,000, right? So this is not something that most college professors are going to do, or at least you might have one. You know, you might have the special one, but it's not going to be a major collection you're going to have. And then, of course, you can get the, you know, conspicuous consumption watches that cost $100,000. No problem whatsoever. These exist. Anyway, I'm wondering whether education won't go in a sense. similar direction where we won't eliminate colleges and personal instruction and things like that,

but it will become a luxury item. It will become less common. There will be people who get the sort of cheap and easy quartz watch AI education. There will be people who get the nice and fancy luxury human-centered education. I don't think it's a perfect analogy because I don't think that the – I still think that a lot of people will still go to college and do a more or less conventional purpose. program because it's not just about education, it's about socialization as well. Those four years you spend in college, as I talked about in the holiday message this year, are really important to your growth as a human being. So I'm hoping that that's still the most common way that people get educated. I just do think that there will be a growing segment of people who, quote unquote, get an education without actually meeting professors or other students in person. We'll see how big that. segment grows. It's not going to replace the universities, but it will be important, I think.

T-C-M-D says, given that we work almost entirely with effective field theories, what would it even mean to know that we've reached a fundamental theory? Is there a principled stopping criterion, or is fundamental just a way of saying no obvious deviations observed yet? So, for those of you who have not read my book, Quanta and Fields, an effective theory is a way pioneered by people like Ken Wilson, but also Leo Kadenov, Stephen Weinberg, and others, of saying, of admitting, look, we don't know what the theory of everything is.

So we don't know what's happening at arbitrarily high energies, arbitrarily small length scales. But guess what? We can still do quantum field theory with a cutoff. That is to say, you only include particles and fields up to a certain energy, down to a certain length scale, something like that. And miraculously,

even though there are still effects from things going on at higher energies and shorter distances, you can sum up all of those effects in their impact on the lower energy, longer wavelength, things that we actually observe. And the result is called an effective field theory. So, for better or for worse, an effective field theory lets you really describe an enormous amount about the world without having the theory of everything. So that's a good news, bad news situation. The good news is we can describe a lot.

The bad news is it's hiding from us the evidence that we would like to gather that would tell us what is the fundamental theory of everything. So to actually answer the question, I think that there's two things. Well, there's a bunch of things. I'm not going to number them because I'm making this up as I talk. I don't really plan out these answers very far ahead of time. So one thing about a fundamental theory is that it would be comprehensive. In other words, the fundamental theory would at least purport to provide an answer to every well-formed physical question that you could ask, right? Something like the standard model of particle physics doesn't do that.

There are things that go wrong at super-duper high energies. It's not just arbitrary and sort of awkward. It literally breaks down at very, very high energies. So it sort of can't be right, not to mention it doesn't include gravity, right? And gravity exists. So one thing you would want in a fundamental theory is it covers everything you already know about. And it covers it in a way that is self-contained and well-behaved in the ways that you want to use it. Now, if you – and we don't have that.

So this is not really a pressing worry that we have right here. But let's say we were to get that. Let's say we would have a theory that had no loose ends that covered everything that we knew, that accounted for dark matter and quantum gravity and black holes. and all that stuff. How would we know it was fundamental? Well, we wouldn't. You never would.

That's the other thing. That's the second point to make here. But it's not super important that you wouldn't know. So you would simultaneously say this theory that we have right now has a chance of being the fundamental theory of everything. Because it's compatible with all the data and it doesn't have any internal loopholes or anything like that. but you also have to say we would change our minds tomorrow if we got new evidence coming in. And so you can sort of decide for yourself how much credence to put on the idea that it was the true fundamental theory of everything or there was something better to be yet invented. That's okay.

Again, this is not a pressing issue since we don't have any theories right now that are 100% obviously correct in all these different ways. So we'll cross that bridge when we come to it, I guess. Marie Roskew says, going back to Mindscape episode 45 with Leonard Suskind, Lenny said about gravity that it didn't seem to be a thing to quantize because there's no recipe quantization doesn't work for gravity. It separated them that made this tension. Yeah, there's a slight, maybe I miscopied here, but there's a grammatical issue here. But, okay, it's separated those that made this tension between quantum mechanics and gravity. Suskin continued, I think the point is that they are so closely connected when we put them together, they're just too closely related. And you said, gravity is just to be quantum from the start.

I wonder what you think about that idea now. Well, I'm all in favor of that idea. This is an idea that I myself have been pursuing for quite a while now before I had that discussion with Lenny. But I think that our ways of carrying it out, mine and Lenny Susskins, are slightly different. But the underlying philosophy is quite compatible that somehow one of the things holding us back from truly understanding quantum gravity is that gravity is different from other forces of nature. And one of the ways in which it's different is that you can't just take the classical theory and quantize it. Okay. No one ever said that you had to succeed doing that.

We've been lucky, I would say, with every other theory. I mean, the other theories that we have, the standard model of particle physics, etc., all take the form of nice local field theories that you can define classically and then quantize them. And then as long as you like think carefully about renormalization and things like that and anomalies, et cetera, you can get a well-defined quantum theory. Gravity doesn't seem to be doing that. So it probably is, it at least makes sense to imagine that there is some way finding quantum gravity that starts intrinsically with a quantum theory rather than a classical theory that you then quantized.

that's not a lot of guidance though and this is why Lenny and I can sort of pursue different angles on it because there's plenty of ways you could start by having just a purely quantum theory and then finding a classical limit of it rather than finding a classical theory and then quantizing it.

So I still think that's the way to do it and you know what? It might be that everything converges at the end. It might be that, you know, doing string theory and doing quantum information-based approaches and doing Hilbert space-based approaches that all of these amount to the same thing. All of these are trajectories that get us to the same final destination.

We'll have to see. That's why we do the research. Jim asks a priority question. With the continually increasing evidence of mature structures early in the universe from the James Webb Space Telescope, I'm curious whether we're starting to explore inflation models, which can gauge the viability of a very highly chaotic inflation event where galactic filaments, superclusters, voids,

and even supermassive black holes, were baked in from the start rather than large-scale structures evolving slowly through time based on gravitational influences. So there's a couple of things going on here. The interesting cool data, evidence that we're getting from the James Webb Space Telescope is that there are both more structures in the universe

at very early times and more massive, well-developed structures at very early times than we expect it. Okay, so more of them and more massive. These are both facts that the telescope is telling us. But just to, we need to put a little bit in perspective what's going on here, okay? The cosmic microwave background, the moment of recombination that we see in our radio telescopes, tells us what was going on about 380,000 years after the Big Bang. Okay, so a relatively short period of time, cosmologically speaking.

When we look at these galaxies that are in the very early universe, we're like, oh, my goodness, look at how early that galaxy formed. We're still talking about hundreds of millions of years after the Big Bang. So we're not looking right up at the moment of recombination or anything like that. There's been hundreds of millions of years in between recombination and when we're seeing these galaxies. It is true. It's absolutely true that our best models predict that it would probably take longer than that for us to form these very, impressively large structures, but it's still a long time by any reasonable stretch of the imagination. So, you know, it's, I have to back up and talk about a meta question here because astronomers

and cosmologists would love to be super duper excited by these results. And in private, we are super duper excited by these results. But in public, it's very frustrating because there's a bunch of popular-level media reports that say, you know, new evidence from the James Webb Space Telescope calls into question the Big Bang model, which is just nonsense. It's just BS. It's not calling into question the Big Bang model. It's calling into question our theory of early galaxy formation, which is a subset, I suppose,

of the Big Bang model, but the Big Bang model itself is completely 100% robust. So we don't like to dwell on these interesting questions as much because we don't want to give people the impression that it's the actual Big Bang model. bang model that is being called into question. So like Jim says, one direction you can go in is to say that when we look at the cosmic microwave background, we see things very, very smooth at early times, but we're also looking at relatively large length scales. Okay.

We're looking at much smaller length scales now with things like the South Pole Telescope and Planck and things like that than we did when I was your age, but still, relatively large length scales. And maybe, even though things are relatively smooth on these relatively large angular scales, you can't just extrapolate down to smaller scales. Maybe there's more variation, more lumpiness on smaller scales than you would traditionally either expect or even predict on the basis of a particular inflationary model. And maybe like these larger fluctuations and density could seed galaxies or black holes or structures, superclusters or whatever. So that's absolutely possible. It's a little bit, you know, having looked at some of the models,

I mean, people absolutely do try to make models that do this. They don't seem super compelling to me, I will be very honest. They seem a little ad hoc, you know, to have such difference in the size of the fluctuations from small scales to large scales, but you can do it. And maybe, you know, there's maybe either A, the universe is a little bit ad hoc, or B, someone hasn't invented the slick and compelling way to do it yet. But also, making galaxies is tricky. It's tricky not just to do it, but to understand how it happens. There's a lot that goes into it.

There's dark matter. There's maybe early generations of stars doing things. There's all sorts of complicated physics going on. So it's also completely possible, at least from my vantage point, which is not super-duper expert in this area, that what we are missing is just a better understanding of how galaxies form. without dramatic changes in early universe cosmology. All this is great. All this is wonderful. We're learning more about the universe by collecting data and being surprised.

So that is, you know, one of the rules of thumb, when you turn on a telescope or a kind of telescope that has never, that looks at the universe in a way that it has never been looked at before, is you're always surprised. You're always discovering slightly new things. Sometimes those things are truly revolutionary. Sometimes they're just changing knobs. dials on existing theories, but you'll learn new things, and that's how we make progress.

So it's all good news from the working scientist's point of view. Patricia Paulson says, I've been wrapping my brain around decoherence versus entanglement versus wave collapse, and while making some headway, I'm wondering why the physical slits on the double slit experiment don't cause decoherence and destroy the interference pattern. Surely some of the photons, or whatever particle, strike the barrier, and wouldn't decoherence and nor entanglement with those decohered particles cause collapse. Why does a detector cause collapse but not the edges of the slits? The answer is, I like these questions where I know what the answer is.

The answer is because, as I have said before, but in different contexts, it may or may not be perfectly obvious, interaction is not the same as entanglement. Two quantum systems entangle when they have wave functions that are superpositions of different possibilities, like the cat's awake and the cat's asleep or something like that. The observer sees the cat awake. The observer sees the cat asleep, right? Those are different possible things that could be in a superposition. And the entanglement means that different parts of the superposition of one system

interact with different parts of the superposition of the other system. So when you're the observer and you open the box and you see Schrodinger's cat, either awake or asleep, the part of you that sees it awake is different than the part of you that sees it asleep. and they're entangled because one branch only has the awake things and one branch only has the asleep things. That's very different than just having interaction. You can interact without entangling. For instance, the usual example I use, if I have a spin, so I have a spinning particle that is in a superposition of spin up and spin down. And I know that I can put that in a Stern-Gurlock experiment which splits the spins,

uses a magnetic field to move the spin in one direction or the other, depending on its direction of spinning, and you would measure the spin differently. But if you just drop the spin in a gravitational field, of course the gravitational field is interacting with the spin. It's pulling it. It's making it drop. But it interacts with the spin-up part and the spin-down part in exactly the same way.

So it doesn't become entangled with the spin. The whole spin just does its thing in the gravitational field. So there's a sort of approximation that is a very, very good approximation where the slits in the double slit experiment are kind of like the gravitational field. They're an infinitely heavy background. They're not really infinitely heavy, but from the point of view of a single electron, they might as well be infinitely heavy, right? The one electron bumping into this macroscopic barrier isn't going to move it that much. It's not going to become entangled with it. The barriers is the barrier.

It is essentially classical to a good approximation. So no entanglement happens. If you try to make entanglement happen, and you can. And people like, I think even Einstein, but certainly there's a wonderful book by Yakir Aronov, where he talks about all these puzzles in quantum mechanics. What if you make a very, very low mass barrier with two slits in it? In fact, what if you put it on wheels so that if the particle goes through in one direction,

it'll nudge the slits in that direction. You know, it'll bounce off and nudge them in one direction versus the other direction. Then, you know, you have to talk about the wave function of the barrier with the slits, right? Usually you don't have to talk about that because it's a big classical thing. But if you did have to talk about that, then you could imagine a limit in which the single electron would become entangled with the slits because they would move one way if it went through the left slit, move the other way if it went through the right slit. guess what? The interference pattern goes away. You can calculate that, just as you would expect. So I think the short answer to the question is, interaction is not entanglement. You have to think a little bit more carefully about whether entanglement is really happening. Ercon Sertelli says, ignoring practical concerns, imagine an alternative physics curriculum where students start, sorry, with math and qubits get introduced to quantum mechanics early and derived classical mechanics in the later years. Do you think the graduates would end up having a more intuitive graphic?

but fundamental physics at the cost of being worse at everyday physics, similar to how one's first language feels like second nature, whereas languages learned later rarely feel effortless even after decades of practice. I personally think that would be a tremendous disaster to try to teach physics that way, because you have to understand whenever you're teaching things or explaining things or whatever, people are not empty vessels, okay? Even young kids are not empty vessels. You're not just pouring knowledge into them. The difference between the physics example here, quantum mechanics and classical mechanics versus the languages example, is that German is not a limiting special case of English or vice versa.

They're just two separate things, okay? Quantum mechanics, on the one hand, has a classical limit, so you could, logically speaking, do quantum mechanics first and then derive classical mechanics, but also quantum mechanics is very far removed from our everyday experience. If you started talking to eight-year-olds about wave functions rather than a pendulum going back and forth, they would just look at you like you're nuts. This has no connection to anything that they've learned about. It's perfectly okay to start with special cases of true theories and build up from those special cases to more general cases.

You don't want to teach young kids false things about the world. You don't want to teach them the plum-putting model of the atom or the phlogiston theory of combustion or anything like that, because those are not limiting cases of true theories. But classical mechanics is a part of quantum mechanics, and it's much more vivid and visceral in every day. So I think it makes perfect sense to teach classical mechanics first. And then, of course, I do also think that we should be a little bit more inviting to more modern and fun kinds of physics than we are. in typical under, you know, pre-college curricula.

I think we should let the kids know that there is quantum mechanics and relativity in the Big Bang and Blackholes, but we don't have to start there. Darren Viliotti says, although I think I believe that progress is possible, as long as it's clearly defined, what are the chances that given the complex dynamical nature of human systems that the world we have is close to the best we can do, given the natural constraints? I'm not trying to be cynical or pessimistic, but sometimes I look around me and get the sense that it's just always going to feel messy, cyclical, and due to a plurality of anecdotal human's objective experiences, leading to different belief structures, what counts as progress, etc., and that, quite frankly, it's amazing we do as well as we do to manage it at all. Well, I think two things are true at once. I think that progress is absolutely possible and perfect progress is never obtained.

So to directly answer to the question, do I think that we have sort of the best we can do world right now? No, I absolutely do not. I think we can do better than the world we have right now, given the natural constraints. But it is a process. It's an inevitable process. It's like brushing your teeth or doing the dishes after you cook a meal, right? It's not something that, you know, you brush your teeth so well one day that you never have to do it again. And that's not what it's like.

Society is a complex system, like you say, and it will always be sort of relaxing to non-ideal conditions. One of the nice things that, you know, Adam Smith is famous for is pointing out that there are certain conditions under which you can actually get really good outcomes without planning them from on top, with just letting everyone pursue. their self-interests. But I know that this result, you know, this result is true if you accept that it's true in certain circumstances. It is denied by some people and raised to the status of a religion by other people. The fact is it's true in some circumstances and not true in other circumstances and you have to work at it.

You have to figure out how to stop the bad parts of either capitalism, if you want to put it that way, or democracy, if you want to put it that. way, or just relaxation of complex systems more generally. There's no rule that says in general a complex system is going to relax to a configuration that is in some sense optimal with respect to your preferences. So there will just be a give and take. The system will try to relax to what it thinks of its equilibrium, and we human beings will have to keep nudging it back toward a case where a situation where we think it's better.

it's usually hard to, well, let's put it this way. It's disastrously unsuccessful if you forget that what you're dealing with is a complex system that has kind of a natural equilibrium state that it wants to relax to. And you just say, I'm going to insist that the state be a certain way. That leads to all sorts of terrible inefficiencies and failures. And the Soviet Union and other places tried that in some way. And it really doesn't work very well. But it's also just obviously empirically wrong that if you just let the system go by itself and let it relax, everyone is happy at the end of the day. I get a lot of mileage out of talking about the podcast I did with Duran Asamoglu, who later won the Nobel Prize in Economics after the podcast, I like to point out.

Well, we talked about, you know, the outcomes historically of major transitions in technology. it's almost always true, according to Asamoglu, that after you have a major transition in technological capacity, not only do the rich get richer and the poor get poorer, but the poor get poorer not only in comparison to the rich, but even in comparison to what they were like before the transition, things really get worse for people after a major technological transition because what you've done is you've given the complex system a new way of becoming less equal,

of sapping resources from the not well off and giving them to the well off. But also, essentially every time we go through this, you then correct it, right? The people fight back a little bit and they manage to make things a little bit more equal. That never stops. That's never going to be like, okay, we've got it right now,

we just can relax. It's just not ever going to be like that. We can always try to do better than we're doing right now. I think that right now, literally right now here in 2026, you can pretty strongly make the case that we're not doing well at all at sharing the wealth equitably. And also, I think you can very easily worry that we do have technological changes that are coming super rapidly, whether it's AI or just interconnectivity through the Internet and so forth, that have enormous capacity for destabilizing the social order. And so it's going to require more work on the part of we human beings to prevent the worst from happening. So hopefully we have the smarts and the determination to get that done.

Okay, I'm going to group two questions together. One from Joan Balluda, who says, Did you see Open AI's recent new result in theoretical physics? Do you see it as a genuinely interesting from a scientific point of view or more as a case of marketing spin? And more broadly, do you think that AI systems could eventually contribute to truly deep discovery in fundamental physics. Whereas Red Links says priority question, this is something that's been bothering me lately.

This spring I will have finished the equivalent of college in my country, and after that, I plan to study astrophysics at university. Now, the thing is, I recently read that Open AI claimed that an experimental AI they developed has achieved gold medal performance in problems from the International Mathematics Olympics,

which makes me think about the claim some people make that machines will replace scientists in the near future. It even feels like the thought that scientists could become obsolete, and science education, thus becoming essentially pointless, is actually affecting my studies negatively. From what I understand, you still don't believe that current AI

is as close to thinking in the way a human does. What would be your best argument for why it is not and why getting education is still meaningful? So these are related questions, clearly, but Jones is a little bit narrower, and Red Links is a little bit broader, so let's take them in that order. There was a result by OpenAI.

No one is noticing this, but the OpenAI result is not OpenAI's result. It's a paper written by physicists. And one of the collaborators was someone at OpenAI. So Open AI put out a press release. But another collaborator was Andy Strominger, former Mindscape Guest. And what they did was, I don't know exactly what they did. They did some technical result in quantum field theory.

And they were looking for, you know, there was a conjecture and they were looking for exceptions. And some version of chat GPT was able to make a suggestion about where you would find an exception and help to do the calculations for it. It was certainly not just chat GPT doing that work. It was human beings. And indeed, my impression is that the sort of place to look for the exception that they found was suggested by Ed Witten, not by chat GPT. So it's more like collaborating with using chat GPT as a tool than actually the

LLM itself getting a brand new result in theory. theoretical physics. It is interesting. It's a perfectly good result, you know, but it's not like we're replacing scientists in any way. When you say, do I think AI systems could eventually contribute to truly deep discoveries in fundamental physics? Sure, 100%. I mean, look, that's a very low bar. Do I think AI systems could eventually contribute to truly deep discoveries in fundamental physics? Do textbooks contribute to truly deep discoveries in fundamental physics? Does Wikipedia, does Mathematica contribute to these results? There are tools that we use, and they're absolutely very useful tools, and there's nothing

wrong with that. The LLMs have this human voice, and so we feel differently about them in some sense, but it would be very strange for me to claim that they're not going to help contribute. I find them super-duper useful for helping to learn things and make simple programs and things like that. So, yeah, of course, I think that they're going to be able to contribute to making deep discoveries. Now, to Redlinks's question, you know, the example that is prodding self-examination here is that Open AI is claiming that AI can achieve gold metal performance on problems from the International Mathematics Olympics. So no one will pay you to achieve gold medal performance in problems with the International Mathematics Olympics.

Like, that's nice that you can do that. But that's a very different world than doing research at the cutting edge of mathematics. The thing about problems in the International Mathematics Olympics is they are by construction well-posed problems. They are problems that students can tackle and solve in principle. And that's just not the interesting part of scientific or mathematical research in my mind, much less philosophical research, my goodness. A huge part of research is deciding what question to ask, given what you know, given what other people are saying, like, what is the interesting place that we can make progress? How do you formulate the question correctly? And so I'm not saying that AIs won't be able to do that at some point, but I'm just saying it to be super duper careful about comparing what they can do now on some extraordinarily, you know, well-defined,

subset of problems to what real working scientists and mathematicians do all the time. So I think there's essentially zero chance that in my lifetime, scientists are going to become obsolete. Okay. The everyday workflow of scientists might change dramatically. That's absolutely possible. I mean, you know, sometimes we had John Danaher on the podcast quite a while ago. Maybe we had him on too early.

He was a philosopher who has an optimistic take on how. Now, AI and automation and robots can actually, if we want them to, help bring about a future automated utopia. Because we can remove all the busy, bad, ugly work from the need for human beings to do it. Okay. It's not necessary that we're going to end up in that kind of situation, given to what we were just saying about the need to, you know, continually maintain the system, et cetera. But it's a possible future. And so I think that there is absolutely a possible future where scientists are liberated from doing ugly, boring calculations, you know, from like analyzing data in ways that are just sort of predictable and following a pattern and things like that. Maybe that kind of scientific work is it will be automated, freeing scientists to be more imaginative and think deeply about things.

But I don't know. You know, I'm not going to claim to know. I'm not going to, there's a certain kind of person that would say, here's how it's going to be. Trust me. And I really don't know. This is a dramatically transformative technology, I think, in principle. The equilibrium that we eventually get might look quite different from the one we have now or only marginally different from the one we have now. But I don't think that scientists are going to become obsolete.

I mean, human beings are not obsolete yet. Getting an education in science helps you be a better. human being. Even if you just get an old-fashioned education in some career that is going to last, you can't guarantee that you're going to have the same career decade after decade. What you really want out of your education is preparation for all sorts of things that might happen in the future. That's why a liberal arts education that includes science as well as the humanities is the most valuable thing you can get. It prepares you to adapt to different circumstances that might come up, not to mention giving you a better idea of what should be meaningful and could be meaningful to your

own life personally. So getting education is more meaningful than ever in my mind. Jennifer Stoneman says, would a philosophy of physics question be what is a point? Like if you were discussing how empty space fields, higher entanglement of different points, would mean those points are close together in space. Well, I suppose that would be a philosophy of physics question. It's not the kind of philosophy of physics question that most people are thinking about these days. You know, there's sort of two flavors of philosophy of science.

One is you're thinking about how science gets done, right? So if any of you were undergrads or even grad students and read books by Thomas Kuhn or Carl Popper or Paul Fireob, they're talking about the process of science, how to make a theory, how to judge a theory, things like that. So that's part of what philosophy of science is. There's another part of philosophy of science that is just sort of what I've been calling natural philosophy. It's just doing science in a philosophically careful way, like having the patience to ask these questions, like what is a point? But in that world of natural philosophy, the questions you ask are not taken randomly out.

out of the set of all possible questions. They are motivated by our current knowledge of science and philosophy. So most working philosophers of physics are not asking things like what is a point. They might ask things like, what does it mean for a classical space time to emerge from a quantum wave function? They might ask like, what are the different ways in which wave functions can evolve in time, whether it's the Schrodinger equation or something else? They might ask, why is their probability entering into the equations of quantum mechanics? They might ask, does the thermodynamic arrow of time help explain other arrows of time like, you know, the psychological arrow?

The point is that all these examples I'm giving are very closely tied to the cutting edge of physics research, right? So it's philosophy of physics at the edges of physics. It's not just, you know, asking questions out of the blue, but trying to be. figure out what are the questions we need to answer to make progress in our best areas of physics. Ryan Santos says, I enjoy your musings on literature and philosophy. I'm curious if there are any myths or legends you especially connect to. Short answer is not especially, like I don't have favorite ones. I do, I love myths and legends and I even love, you know, religious works and things like that, quasi-historical things. This is why we had Emily Wilson on the podcast

talking about her translations of Homer, for example, or Shadi Barch talking about Plato in China. And, you know, I've actually, I wrote a blurb for a translation of the Tao Te Ching, which was a fun thing to try to do. I'm not an expert in different translations of it, but, you know, The Chinese philosophy is fascinating to me. We had Brian Fittenden Orden on not too long ago, Edward Slingerland on before that. The sort of storytelling style of Chinese philosophy is closer maybe to the Socratic way of doing things than a more dry, narrative-free Aristotelian way of doing things. So that's still doing philosophy in sort of a storytelling mode. But, you know, yeah, I like Gilgamesh or, you know, the Mahaburata, you know, various things.

I love all of these different stories. But I don't, I'm not an expert in any of them. I wouldn't, you know, be able to judge any of them. Like when I read Ovid's metamorphoses, it was weird. It was tough. Like I found, I don't know, for some reason, Homer, the Iliad, the Odyssey, I can just read. Or even better, listen to them on the audiobook because they're meant to be read out loud. But Ovid was just weird because there was just almost no conventional narrative structure.

Like he was telling stories. It was like, yeah, this happened and this happened and this happened. And that was kind of it. Although weirdly, you know, even the Brothers Grimm are like that. Like I recently read Philip Pullman, of all people, did a translation of Grimm's fairy tales. And it's very well translated and he makes some artistic choices. But the structure, the style of these stories is sort of not what we think of. of his good old conventional three-act narrative structure.

And so it's a little bit different than what we're used to. But anyway, this is just a rambling way of saying, I love them all. Bring on all the myths and legends. I think it's kind of fun. And modern ones as well as older ones. Nicola Ivanov says, if locality is emergent, our low energy laws look local,

even if the fundamental description is not. Do you think locality is a dynamical attractor, meaning it appears for a huge range of possible early universe Big Bang initial conditions? once you coarse grain? Or does getting a local effective field theory requires special initial conditions or a special state?

Well, I think that in the traditional way of thinking about things, the answer is no to both of these questions. In the following sense, we have some theory, string theory, or whatever it is, that's going to be the fundamental theory of everything. And there are phases you can be in,

which correspond to sort of different parts that you can be in the landscape of string theory. But all of those have their fundamental parameters varying, like the masses of particles and things like that. None of them have an option for being local or being non-local. Locality is just built in to the assumption of the theory from the start. So in the usual way of thinking about things, locality is just put in.

It's not in a dynamical attractor. It can't be. There's no freedom in the theory for locality to emerge because it's just assumed from the start. Now, interestingly, that's not exactly true if you really think that string theory is the right answer. Strings are not point particles. They don't bump into each other at a point in space time. They have some extent. So there's a little bit of non-locality.

And there were some attempts in the 90s when people were first really taking seriously the black hole information loss puzzle. to take advantage of that non-locality to help understand how black holes could emit information in their hawking radiation. You can think of these strings, which ordinarily, sure, they're not exactly a point, but they're pretty point-like from far away. You might think that for a macroscopic black hole,

which is much, much bigger than the plank scale, strings are still pretty point-like, but there's a point-of-view that says the strings seem to spread out over the horizon of the black hole and become very non-local. Anyway, nevertheless, as far as I know, that point of view didn't really go anywhere. We do have a very dramatic kind of non-locality in the holographic principle. But it's a weird kind of non-locality.

We have the bulk in the ADS-CFT picture where you have a bulk space time in N plus one dimensions, mapping onto and being an equivalent theory for a space time without gravity in N-D-Demeter. So in one dimension less, obviously, if these two theoretical descriptions are supposed to be equivalent but live or are described in different numbers of spatial dimensions, the relationship between them has to be non-local, right? If you go from the lower dimensional theory to the higher dimensional theory, you don't just go from point to point. Okay? There needs to be something much more dramatic going on. And indeed, there's some non-local mapping in the ADS-CFD correspondence, the holographic dictionary. However, despite the fact that it's wildly non-local, each individual theory is local.

You have the local theory on the boundary of ADSCFD, you have local theory in the bulk of ADSCFD. So in a very real sense, things still look pretty darn local. So there's no sense in which locality has a chance to emerge. Having said that, I have some crazy ideas myself about how maybe locality could emerge. Sorry, I need to be better about this. I'm violating a rule that I make fun of other people who are violating. Using the word emerge in two totally different senses. There's one sense of emergence, which is time independent,

which is just a simultaneous fact about two different descriptions for the same underlying physical thing going on. When I have the air in the room, and I say you can describe it in terms of molecules, or I can describe it as an emergent fluid description, that's not emerging over time, right? That's just both true at the same time. Whereas there is sort of the dynamical attractor that Nicola is asking about, which is a change over time of some physical circumstances

that lets a certain set of characteristics or properties become evident. So in that sense of emergence, in the sense of something appearing or coalescing or coming into being over time, I do have some wild ideas for how that might happen. but none of them really work out in my brain right now. And certainly it's not part of the traditional story. Dennis Banks says, I heard David Deutsch say on a podcast that as confident as we are

in inferring from the biological and paleontological evidence that Darwinian evolution occurred, we should be equally confident in inferring from quantum mechanics that we live in an Everettian multiverse. Would you ascribe to the same level of confidence? No, not quite that much level of confidence. I think I've pretty consistently said I'm at 90 to 95% confidence in Everett. I'm at a much higher level of confidence in traditional Darwinian evolution for a very simple reason.

Well, I guess two simple reasons. One reason is that I do still think that there are questions that have not yet been completely answered in the context of Everettian quantum mechanics. I think it's by far the most promising view we have, but there's still some dangling threads that, you know, if you pull hard on, they might fall apart. Most obviously the probability question, which I do think is answered, but not 100% confidence. And then the structure question, why does the universe look so classical, if ever it is right? Whereas that's not true in Darwinian evolution.

There's details that certainly remain to be worked out in full glory. So, you know, I'm not someone who thinks that we've really replaced Darwinian. Darwinian evolution with, let's say, the, oh, what's it called? The new synthesis or whatever it is, where you have both Darwinian natural selection and more modern views of genetics. And likewise, we'll talk, we can talk about epigenetics and things like that that absolutely changed the story at the edges, but the fundamental story of Darwinian evolution is basically in place.

and we're going to be tweaking it in detailed ways, but we're not going to suddenly wake up tomorrow and realize that there's no common ancestor between human beings and orangutans or something like that, right? That part of the story is more or less settled in a way that Everettian quantum mechanics is not. The other thing, of course, is that there are possibly reasonable alternatives

to Everettian quantum mechanics. There's really no reasonable alternative to Darwinian evolution. So I think if you're honest, if you're being straightforward, Everett is not quite at that level of confidence yet. Sandro Stuckey says, you've recently expressed doubts

about whether computational functionalism is a good model for explaining cognition or consciousness. I'm surprised by that, but I suspect it comes down to what you mean by computational functionalism. So what do you mean by it? Is it just about inputs and outputs, or do you allow for the computational system

to have memory and learn? Well, I would certainly allow for computational systems to have memory and learn. I mean, they do. that's certainly true. And look, it's very possible that I'm misusing the phrase computational functionalism. I'm not and I do not pretend to be an expert on the various different approaches to consciousness. Net Block is an expert.

Anil Seth is an expert, people like that, but I'm not an expert. So I might be slightly mischaracterizing the term, the phrase. But I don't care about that. What I care about is what I think is true. And the point about my recent doubts is less about computational functionalism and more about putting more emphasis on the particular process by which the purportedly conscious entity is doing what it does. So one way or another, an entity, conscious or not, does have an input output map, right? You tell it some information and it gives you some response. Like that's always going to be true.

Every entity does that, conscious or not. And the original touring test, right, the original imitation game was all about that. It really said that, you know, it wasn't about consciousness, okay, to be fair, to touring, thinking was what it was about. And but it focused on the input and output that you would give to a computer program. And it suggested, and it's a perfectly reasonable suggestion, that what matters in the process of thinking is the map from inputs to outputs, right? because maybe you could have a completely different process going on inside, but if the inputs and outputs were the same, then you could make an argument that for all intents and purposes,

it's doing the same thinking. Very much like if you just wanted to do arithmetic, if you had narrowed your focus down to just adding and subtracting and multiplying and dividing in things, there are many different algorithms that will let you divide one number into another one, and it doesn't matter. It doesn't matter how you instantiate that. algorithm in terms of physical stuff.

It doesn't matter which algorithm you use. What matters is that when you take the same two numbers as input, you get the same number as output, okay? And my point, more to myself than anyone else. I don't care really if anyone else is coming along with me, but I am increasingly of the opinion that there's more to what we think of as actual consciousness. So forget about thinking, which is a broader category, but consciousness, the experience, the qualia that people who like to care about the hard problem of consciousness care about,

I'm 100% convinced that it's not dependent on stuff, on the specific atoms you use. If you mimic exactly the same processes in one kind of atom versus another, then it will be exactly as conscious. I don't think there's any weird consciousness spooky stuff that brings conscious experience to life. But I do think that the underlying processes that lead you from the input to the output are crucially important in whether or not those processes involve something that we call consciousness. And I think that's an important distinction. It's not about, you know, giving up on physicalism or anything like that. That's what I learned from talking with Anil, that you can still be 100% physicalist without being a computationalist in that sense.

Now, computational functionalism is a little bit vague, I think. It might not just be focusing on inputs and outputs because inputs and outputs of what? There might be subsystems of the conscious creature whose inputs and outputs you care about for the functional role that some subsystem in the brain or whatever or in the computer program is playing. So the inputs and outputs at that sort of more fine-grained level might be important for your definition of what is going on. And again, that's fine, and it's perfectly good argument to have. It's just not what I'm focusing on. I'm just focusing on the idea that there are many ways to get from whatever input you have to whatever output you have, not just in the creature, but even at the sub-level.

But the actual processes, especially biologically, the processes that we actually have involve metabolism and increases of entropy, using up free energy and things like that, and all sorts of things and processes going on below the surface. that you're not conscious of, but are nevertheless happening in time in a way that a computer that is just sitting there, not answering any questions, not running the program is not doing, and I think that that might matter. That's what I think is the actual opinion that I have. Antoine Chopin says, I'm reading an immense world by Ed Yong, which explains how animals

perceive the world, i.e. their umwelt. For instance, the smells experienced by dogs, UVs seen by some insects, or how bats navigate using ultrasound sonar. It's fascinating not only to learn about sensing capabilities beyond ours, but also to imagine how all animals perceive the world. My question is, do you have any insight about how Ariel and Caliban, my cats, experience the world or what their own belt could be like, acknowledging that we can only guess? Well, two things. One is, if you don't know, I did have Ed on the podcast, and we talked about exactly this stuff. It was a couple of years ago, so you can look for that if you're not familiar with it. I'm two different minds about this.

So on the one hand, I do imagine that the fact that different censoria can lead to different images of the world is super important, different umbelts, super important for how we go through life, how we experience, how we react, et cetera. You have no doubt about that. On the other hand, I also want to emphasize it's fundamentally the same world. You know, I think that when you say you have different umbelts because, you know, your sight or your hearing or your smell or whatever are different, or even if you have totally different senses like you can sense electromagnetic fields or something like that, what you have are different windows onto the same underlying world. I don't think that two real animals have constructed images of the world via their different umbelts that are completely incompatible with each other, right? They might be incomplete and might have overlaps where they need to be consistent and non-overlaps where they don't, but they can't just contradict each other in any fundamental way.

So that's my take on the importance of the Omelt as a concept. Cats, you know, cats are not that different than human beings on the scale of all the different ways that creatures can be different from each other. We are relatively close to each other. the cats certainly have better hearing than we humans do, better smell than we do, and maybe they're like sensitive to vibrations and things like that in a way that we're not. That's always hard to tell. You know, I've often mentioned that there's this thing where the cats freak out before a total solar eclipse or even a pretty good solar eclipse.

Before, like before it's happening, right? And it's not because the world has gone totally dark. I mean, they're indoor cats. They've been in the dark before. So what is going on? And the explanation that I've been told is that what's actually going on is it suddenly gets darker outside and begins to get darker. It's not all the way dark yet, but it's enough that it disturbs the animals that do live outside, right?

The insects and the birds and things like that. And because they're disturbed, they go quiet. So what actually is freaking out the cats is not that the sun is being eclipsed, but that the nature has gone quiet outside. And that sounds bad, right? So I do think that that kind of umbel can sort of lead to, you know, spooky behaviors if you don't know what is going on. I mean, Ariel, who is the more chicken of our two cats,

Caliban is gregarious. He'll come out and greet visitors and things like that. But if new people come to the house, Ariel is just going to go. under the bed and hide until they come. Unless they really hang out for enough hours that she gets impatient and then she'll come and loom at the top of the stairs and look at them. Like, why are you in my space? You should leave now.

But she has this amazing ability to know when someone comes, whether it's someone she knows or not. Like if someone who she is very familiar with, like her cat sitter, shows up at her door, shows up at our door and she's on the second floor of the house, she doesn't move. She's not disturbed by that. But if someone, you know, if there's a delivery person who comes to the door, even if it's the same noise, because we don't have a doorbell or anything like that, she freaks out and runs under the bed. So somehow she knows what kind of person it is coming to the door. And, again, I don't think that's a hugely important difference, but I do acknowledge the differences and maybe it matters a little bit.

Bearded Veteran says, I've heard physicists like Brian Green and Jan 11 and maybe even you talk about how string theory suggests that the universe has a tiny, compactified six-dimensional space at every point. But I haven't able to wrap my head around this. How can something with X dimensions exist inside something that has fewer than X dimensions? To me, it's like putting a 3D cube in a 1D dot. I've heard the garden hose analogy, but it doesn't make any clear to me. The analogy still describes a 3D object and 3D space observed by a 3D observer.

Well, I can make an effort. I can make an attempt here. It may or may not work. The point is, it's not that in string theory, so for the string theory of fissionados out there, you know the different versions of string theory might have different numbers of compactified dimensions or other complications.

Let's put aside all of that. Let's just go with the traditional story where there's a tiny compactified six-dimensional space at every point. The point is not, as it were, that that six-dimensional space is literally inside a zero-dimensional point. It's that the world, space, not space time, but just space is not three-dimensional, nor is it six-dimensional. It's nine-dimensional.

Okay. And three, sorry, three of those nine dimensions are big. They're macroscopic. They go on for a long time, maybe infinitely far, we don't know. And six of those dimensions are curled up and very tiny. So the point is that if you are made of things like electrons and protons and stuff, like that, if you are made of things that are themselves much larger in size than the size of the things that are curled up, then you just don't notice. So it's not that the six-dimensional

things are inside our three-dimensional world, it's that our world looks three-dimensional to us because it really is nine-dimensional, but six of the dimensions are so small that we can't see. The garden hose analogy is supposed to say, I mean, there's different levels of the garden hose analogies. Like the hose itself, you're right, is three-dimensional. But in the analogy, you're sort of getting rid of dimensions one at a time. So one part of the analogy or the metaphor, I should say, is that you're thinking of the surface of the hose as very thin. So you're thinking the hose itself is two-dimensional, okay? Whether or not that's a good approximation or not depends on the relative size of the thickness of the hose to the dimension that it has the diameter, I guess, that it has around.

But really, the important thing is you're imagining looking at the hose from very, very far away. Where, to a very good approximation, the hose looks one-dimensional. Okay? Because you can just see its length. You can't see that there's a little circle at every point along that length. So you're right. Really, the hose is three-dimensional. Or if I ignore the thickness of the material, it's two-dimensional.

It's like a tube. But it looks one-dimensional. If I look at it from very far away, there's just an extra dimension that is too small for me to see. And likewise, in the string theory version, space is supposed to be nine dimensional, but to us it looks three dimensional. Kyle Stevens says, what do you think is the most compelling case against liberalism? These are always weird questions. Like, I sort of, part of me appreciates the idea that we should understand the best arguments against our, perspective and to try to steal man those arguments. But then you just have to keep saying,

but I don't believe it, but I don't believe it, but I don't believe it. So take it into consideration and very pro-liberalism in the classic liberalism kind of tradition that we're talking about. Not liberalism versus conservatism, but liberalism versus authoritarianism or fascism or something like that. You know, I think that the case against liberalism is pretty straightforward. It's if you don't think that individual human beings are, equally important. That's when you would give up on liberalism. If you thought either that some human beings, there's sort of this sort of, you can go to the right politically against liberalism by thinking that some human beings are just more important than others, that maybe they're just

more talented and therefore they deserve more power, or maybe there's just some natural hierarchy of human beings where wealth and reward is appropriately distributed unequally. Or there can be a leftist version of anti-liberalism where you say that what matters are communities and societies, not individual human beings. And in both cases, you can make cases for either one of those. There would be completely different cases for those two ideas. And it's not obvious. You know, it's not obvious that those cases are wrong. You have to actually make the case for liberalism, that the locus of both responsibility and authority lies in individual human beings. Of course, the individuals come together to make collective decisions, but the government

derives its consent, its authority from the consent of the governed. That's the fundamental idea of liberal democracy. So there is a government. There are collective actions, but the only reason why that government is legitimate is because the individual human beings ultimately lend their legitimacy to it. So, you know, I'm not going to actually go through all the different arguments one could make, but one could either imagine, you've all heard arguments that some human beings are just better than others or that collections of human beings are better than individuals.

Both of those are possible cases to make. I don't think they really work myself. Luke Bailey says Tom Griffiths' work on explore, exploit tradeoffs made me wonder whether physics as a community faces similar dynamics. If the field were following something like an optimal strategy, what signals would tell us it's time to shift from refining established frameworks to exploring new foundations? Well, I think that the strategy faced by academic physics

is actually pretty good. I don't think it's perfect. So, I mean, as I've said before, I would love to sort of criticize it, but there's enough critics out there that I sort of mostly need to defend it. The way in which it falls short, I think, is that for perfectly understandable reasons, it's a little risk averse. Physics can be expensive, it can be literally expensive in the sense that you're building these billion-dollar experiments, but it's also expensive in the sense that if you're just a tiny little physics department,

who doesn't have very many faculty members, every time you hire a new faculty member, that's a huge investment from your point of view, and you don't want to mess it up. And that leads you to be a little risk averse. You don't want to hire someone who is being a little bit more speculative, you know, a little bit more out there, a little bit less mainstream. And I think that that's rational behavior from the point of view of that individual department. But when every department does that, we're left with a community that as a whole doesn't have enough experimentation and speculation in it.

So having said that, those are my criticisms. That's the sense in which I think that we're not. pursuing an optimal strategy. But otherwise, you know, I think that we do a pretty good job and part of doing a pretty good job that is not well appreciated by the critics of academia is that individual academics are trying their best to come up with good ideas and pursue them. So, you know, you're not born, let's say, a string theorist. You're not born a Catholic theologian, right? You're not born with some academic specialty. You develop an academic specialty. And if you if you think that your academic specialty is faltering or not giving you interesting results,

you can change. And people do change. And people convince each other to change. That happens all the time. So I think that, you know, there's always a back and forth between some physicists who are doing the same thing they've always done, some who are trying new ideas, some who are very much in the mainstream, some are who are trying to be more speculative. and revolutionary, some who are just trying to move incrementally, some who are trying to overturn the whole table, as it were.

And I think that's good. And I think that, you know, like it or not, part of the reason why that's even possible in a community, which is kind of conservative in some sense, is tenure. If you have tenure, you can do wild research things. The tradeoff is you don't get tenure if you're doing wild research things unless you strike gold and it really works. If you just start your career doing speculative things that are high risk, high gain, and you don't gain, then you're not going to be a long-lasting faculty member. So the urge to speculate and do crazy things is beaten out of you, and then you're given tenure

and the freedom to explore and do crazy things. So I think that we tell that it's time to shift from refining established frameworks to exploring new foundations when one strategy doesn't seem to be working and the other one does. Physics is very much a show me the money kind of situation. It's not about standing on a soapbox and saying, we've been exploring established frameworks too much. We should explore new foundations. It's about saying, here's an idea for new foundations and here's why it works.

That's the kind of thing that actually gets attention paid to. it. David Sotolongo says, in the most recent February 2025 AMA, presumably you mean February 2026, you said in the opening, despite what has been going on in the past few weeks and long before that, we are going to win. The bad people are not going to win. A little later in response to someone's question, you also said that you think that there's a greater than 1% chance that the U.S. will become an authoritarian fascist regime. My question is, how you talk another typo, how you talk about relatively low probability events in everyday conversations. Given your belief in the many worlds interpretation, see, audience, I bet you didn't think this is where we were going with this question.

Given your belief in the many worlds interpretation, wouldn't it be more accurate to say something like we're going to win in the vast majority of worlds, but not in all of them? I mean, in some sense, sure, that is something that you can always say, but it's not actually the most accurate thing that you can say. It's not really what I mean. What I mean is we're probably going to win. Remember that there is no born rule in politics. There is no idea that when I make a prediction about the probability the U.S. will become an authoritarian fascist regime. I'm not working from a perspective where I have the wave function of the United States

and I'm projecting it onto certain future evolutions and calculating a probability, right? So many worlds is just not relevant here. You know, there's an epistemic uncertainty I have that I don't know exactly what's going on in the minds of various people throughout the world and things like that. That's a different thing than the probability that comes out of a wave function that you have in quantum mechanics. There might be non-trivial branches of the wave function where really, really unlikely things happen and the world looks very different. But even if I were to say, you know, there's a 10% chance that the, U.S. will become an authoritarian regime, that doesn't mean I think that there's a 10% chance that we will end up on a branch of the wave function, and there's 90% of the branches don't have

that happening and 10% do. It was not a many-world's kind of statement. It was just a statement about my knowledge of the current political situation, et cetera. So it would be exactly the same statement if I lived in a classical world. And I think that the relevant part of the statement is what is the probability that any one version of my future self will find themselves, in one situation or the other. So you can talk that language because I think that that's perfectly adequate to the situation.

Ken Wolf says, in your solo episode on complexity in the universe, you suggested that gravity pulling more dense parts of the universe to become even denser allows for complexity to develop when concentrated matter led to interactions such as chemistry and magnetism. Have you ever considered a similar phenomenon

in the human realm? Certain regions with historical advantages have attracted further immigration and investment, leading to a cascading virtuous cycle of development at the expense of the less advantage regions. Would that not lead to a rather repugnant conclusion that achieving any sort of global equality is a lost cause?

Well, I was with you up there until the very last sentence. I mean, I think that the story you just sketched with certain regions with historical advantages, blah, blah, blah, is just true. I mean, there are positive feedback mechanisms where when you develop a little bit, it makes it easier for you to develop more and faster than you,

then you started. So we see that in the world. That's just true. But just like we were talking about before with disruptions from new technology, we can choose as a society or as a global community to try to ameliorate the effects of that inequality. So we can try to help places in the world that have not developed as much. And sometimes we do try to help. Other times we have retrenchment where we stop helping as much as we could. It's not that it's a lost cause. I mean, it depends on what you mean. You say achieving any sort of global equality,

probably achieving global equality is a lost cause in the sense that we're never going to achieve perfect global equality. That I'd be happy to say. That is very different than saying that the worst off people on Earth will get continually worse off or even remain as bad off as they are. In many ways, life for the worst off people worldwide has gotten much better over the last hundred years, right? Childhood mortality has gone down, poverty has gone down, things like that. There's no reason to think that we can't continue to do that.

It's just always a process. It's something we always need to get better at. The people who are doing very, very well can choose to help out those who are not doing as well. Whether they do or not is a different longer discussion. Randall Davis says, as a non-physicist but interested party, I'm also an avid listener of StarTalk, Neil deGrasse Tyson's podcast. He recently did an episode live with Aaron MacDonald and David Salzberg, where they talked about many Star Trek-related topics, including warp drive. Aaron cited improvements in the math regarding energy requirements of Al-Cubier warp drives as going from all the energy ever to all the energy output of a star. Is there any hope that as we learn more about the movement of the fabric of space time

and more about producing energies such as cold fusion, that there could come a time when the energy requirements at least are theoretically feasible? If so, do you suspect the universe will ensure those drives don't work some other way to ensure that causality is never violated? You know, my personal belief is that we're not ever going to get travel in time, backward in time. I think that we can speculate about it, and we don't have any proofs that it's not possible.

I'm just my personal belief is that it's unlikely. Warp drives are different. You know, warp drives do not necessarily let you go backward in time. You can absolutely imagine a warp drive that is completely compatible with causality not being violated. So I'm more open to that possibility. But number one, cold fusion is a fake. There's no cold fusion going to give us a lot of power. And number two, you know, I don't know exactly the,

refinements that Aaron McDonald is talking about. But the energy output of a star is still a lot. And I think that people don't quite appreciate when you talk about these warp drive proposals. So for those of you who don't know, Miguel Acubier, who's a good physicist in Mexico, wrote this wonderful paper back in the 90s, I think, where he showed that certain kinds of energy distributed around a spacecraft could effectively make the spacecraft look like like from the outside perspective, it was going faster than the speed of light. Okay. Now, these, it's not just you need a lot of energy to do it. You need completely implausible energy to do it, not just

that there's a lot, but that there's some negative energy as well as some positive energy. So this is very, very, very far from the realm of realism, I would say, realistic things to do. And even if you could do it, even if you could somehow manage to harness both negative and positive amounts of energy, the total amount you need is sort of astronomically large. Sure, maybe you can go from all the energy ever to all the energy output of a star. It's never going to go down to the energy output of a battery, okay? Because literally there's a minimum amount of energy you clearly need just to warp space time enough to do it.

It's not a technology problem. It's a physics problem. So let's put it this way. If you first are able to show me how you can harness all the energy output of a star, then I will start worrying about whether or not you can build warp drives in a realistic way. Jay says, what are your thoughts on the Mapemba effect? Is it real or fake? And what are your favorite explanations for the supposed observed phenomenon? Could some sort of runaway convection explain it? So for those of you don't know, the Mapemba effect is a claim that, at least under certain circumstances, which way does it go?

It's, you know, if you have water that is either cold, it's either, I forget it because this is not my area of expertise. And I'm not going to bother to look it up. Sorry, you can look it up. But it's either that you can have hot water freezing faster than cold water or that you can have cold water boiling faster than hot water. I forget which way it goes. Okay. But that's the Mapemba effect. One of those ways or the other.

And apparently it's real. At least people have claimed that it's real, experimentally. And even though I know nothing about the. the plausible explanations for it or things like that. I wanted to answer the question or address the question just because it's a wonderful example of the difference between spherical cow reasoning and dirty real world messy reasoning in physical situations. It's, you know, one's first reaction as a physicist. So let's say that the claim is that hot water freezes faster than cold water, okay? An instant physics reaction to that is that literally can't happen.

The hot water has to go through a moment in which it's cold before it can freeze, right? So however long it takes to go from cold to frozen is added to however long it takes to go from hot to cold. And therefore, the total amount of time it takes to go from hot to frozen must be larger than going from cold to frozen. There you go. I've proven it. How can you possibly wriggle out of that? Well, you can wriggle out of it. I mean, I can imagine you can regal out of it.

Again, I don't know any of the details, but there's cold water and there's cold water. There's things going on in real water that are more complicated than the simple fact about whether it's cold or whether it's hot. What are the motions of the molecules in the water? Are there large-scale motions? Is there convection or something like that? Are there little cycles inside? What about the container that the water is in? What about the immediate environment of the water?

Does it have a lot of water vapor in it? Does it not? Is that going to be different, depending on whether you started with hot water or cold water? There's just a lot of complicated things that are exactly the kinds of physics that I am bad at, but I appreciate them. I have great reverence for the people who do the hard work to figure these things out. So all I want to say is it's plausible that hot water could freeze faster than cold water because the world is very complicated, and it goes.

beyond our spherical cow approximations. Joseph Eli says, I just took home a beautiful six-year-old cat from the shelter. She's my first pet and I couldn't be happier. As our resident professional scientist cat expert, what advice do you have for new cat owners, anything you learned the hard way? So congratulations, Joseph. That is great. Especially it's great that you got a shelter cat and a six-year-old shelter cat.

That's just a wonderful thing that you're doing for that little kitty cat. I think you'll be very, very happy. I think you say you couldn't be happier. I think it will continue, which is not to say, I think, that everybody would be happy. I'm a pluralist about most things, including pet ownership. Some people won't be happy with pets at all. Some people will be happier with dogs. But if you're already happy with the cat, I predict your happiness will only go up.

You know, I think that with cats, there's not a lot of advice to give because the great thing about cats is they're pretty, they're self-starters, cats, compared to dogs. compared to dogs, for instance, dogs need a lot of attention, and that's what some people want, right? But cats are much more independent. Now, they're not perfectly independent. It's going to depend on the personality of your cat. Your cat might be the kind that wants to be in your lap

or on your shoulder or sitting next to you all the time. Or your cat might be one that wants, like, sit on the window sill and occasionally come over to say hi to you and get fed or whatever. Like, there are different cat personalities. And I think that the thing is you can't train cats. really. You can introduce them the new experiences and they can choose whether to be enthralled by them. Some cats like scratching posts, some don't, some like catnip mice, some don't, etc. Some cats like to sleep with you at night. Some don't. Some like their bellies rubbed. Some don't,

etc. The only advice I have is think of it as you learning about the cat. And the cat learning about you, not anybody training anybody else. Living with a cat is, much more like having a roommate than having a pet in a very real way. If you respect the cat, if you just give her what she needs and let everything else fall where it wants to fall, you will be super-duper happy. It's an extremely rewarding thing to do. T.K. says, would you have traded Edgecom for Janus? for those of you out there

for whom this is all Greek to you, Vijay Edgum is a basketball player for the Phil W76ers, a star rookie that they just drafted last year, who's been doing very well. And Janus Atendikumpo is a MVP-level senior basketball player for the Milwaukee Bucks who is finally, after a few years of struggling with the team,

decided that he's open to being traded to other teams. So it's a classic sports question. If you're, if you, it's weird how in the modern era, fans pretend that they're the GM, the general manager, the president of the basketball team. And it's almost as if they need to have an opinion about which players should be traded for which other ones. I've tried to wean myself off of that mindset. And I would like to root for the players who are on the team. I don't really want to try to imagine fake scenarios where I'm trading the players away. I just want to root for my team and have fun that way.

But in the thought experiment world, in the world of speculation and chit-chat, is perfectly legit to say, would you trade away a promising young player who may go on to have a multi-year all-star level career, et cetera, for a player who is already at the top of their game right now and is sort of a sure thing in some sense? I think the honest answer is no. I would probably not have made that trade, specifically for my beloved Philadelphia 76ers, because I don't think that Janus is a perfect fit for the team. I mean, the Sixers will still, despite the fact that they have Edgecom and Tyrese Maxi,

they will go as far as Joel Embed will take them. And Joelle Embed is not a good fit with Janus. They're both tall people who want to play around the basket and things like that. And Embed is going to be at his best when he's surrounded by fast-moving. perimeter players who can shoot from outside and play good defense. And Janus is many, many good things, but he's not exactly that. Shooting from outside is not his specialty. So it'll be a very clogged offensive game. It would be a great defensive team with Embed and Janice. But also, you know, Embedd is getting up there in years. So is Janus. They're both

on the late side of 30 years old, which in basketball NBA terms is on the downslide of your career. They could both play at a very high level for years to come, but they could could also break down and not be that great. Whereas V.J. Edgham is like 21 years old, and he's very bouncy and, you know, indestructible as you are at that age, and he will hopefully be playing at a high level for the next 15 years or more. So I think that the, even though it's tempting to sort of go all in for a big splash, I think that, you know, the cold-hearted calculation would be to not make that trade right now.

David Kudaverdean says, I heard from you that a black hole represents the maximum entropy possible within a fixed region of space. It's not clear to me how this is compatible with the viewpoint of a falling observer. Once they're inside the horizon, they do not appear to be in a state of maximal entropy from their own point of view. Does this mean that entropy depends on the observer's location? I think there's a couple things going on here. There's like the simple thing, which is more than enough to answer the question, but there's a slightly more complicated thing.

The entropy of a black hole is proportional to the size of its event. vent horizon, and it comes almost entirely from quantum mechanics and gravity. The entropy of a black hole is not very appreciably changed by the entropy of something you throw into it, other than the fact that the stuff you throw into it has mass and contributes to the area of the black hole. But the same thing is, just forget about black holes. Think about empty space, okay? Think about a region of the vacuum in quantum field theory.

if I take a cubic centimeter of space, a spherical region one cubic centimeter across, sorry, one centimeter across, so a little bit less than a cubic centimeter. It has entropy. It has entanglement entropy between the quantum fields inside that empty region of space and outside.

And that entropy is actually huge. It's quite large. It's one of those things which in quantum field theory naively is infinitely large, but then you have some cutoff and still very big. And you can change that entropy by putting particles in there that have entropy, but the amount you're changing it by is completely negligible compared to the empty space entropy. And the entropy of a black hole is mostly empty space entropy. It's the entropy of the quantum fields in their vacuum state near the event horizon that matters.

It's not the entropy of what you throw into it. So that's the simple answer. you know, the entropy is not that affected by what you throw in. The more sophisticated answer is, what do we mean by entropy? I don't know if that's more sophisticated or not. But, you know, Hawking did a calculation that gave a certain entropy to the black hole. But you can argue that, you know, that's the first approximation to something more specific.

And really it takes time for the black hole to equilibrate and reach that highest entropy state. So maybe the black hole's entropy isn't quite that big when you do throw something new in. And in some sense, it's going to take time for the observer to fall into the singularity and sort of have its information be scrambled. And then the black hole will finally reach its maximum entropy. But that's something that is much less understood and not super important for the specific question. Tomas Settovich says, on several occasions you have mentioned having an evil twin, Sean B. Carroll. What makes him the evil one? After all, your interests around black holes, entropy, decay,

and the quantum and chaos theory sound much closer to villainy than biology does. What kind of good person goes around calling other people evil anyway? So because I'm an ecumenical kind of guy, we did in fact have my evil twin, Sean B. Carroll, on the podcast here. You're welcome to listen to that episode. It's perfectly obvious why he's the evil twin. He's the one who has a beard. Haven't you ever seen a TV show?

It's when you have like a whole evil universe, a whole twin mirror universe where people are evil, they're the ones who have beards, right? It's the clean-shaven people who are good. So I'm just going by what Hollywood teaches me, which I presume is completely correct. And I think good people should always go around calling other people evil if they're evil. There's no reason for a good person to hide the fact that someone is evil. I wouldn't understand that motivation at all. DMI says the act of writing encyclopedia involves many interactions with fields in the environment. It seems like the environment would retain that information even if the encyclopedia was thrown into a black hole.

Maybe you could write the encyclopedia sealed in a box, but the encyclopedia would still be largely a product of entangled events during your life before the box was sealed. Unless you wrote about whether a cat in the box with you was awake or asleep. Okay, but if that information could be destroyed by throwing the box into a black hole, then it seems to imply that the information could be created by opening the box before throwing it in, am I misunderstanding something? So I got a little confused there in the question, but I think I know what the essence of it is. So let me try to address it.

I think the essence of it is up in the beginning of the question where you say, it seems like the environment would retain the information about what's going on in the encyclopedia, even if the encyclopedia itself was thrown into a black hole. And that's not true, actually. And I can tell you why we know it's not true because what the sort of assumptions here that are going on are we're imagining being able to keep perfect track of the microscopic state of the system that we're working in. So it is true that when I write on a piece of paper in various ways, I'm affecting the environment around me in ways that are more less hidden, right, more or less invisible. So I'm making noises, scratching the pencil on the paper or whatever.

There's vibrations, that some of the ink or graphite molecules are evaporating into the air. So there's different effects that I'm having on the environment. So you might say that even if I take that piece of paper, if I write on it and then I burn it, maybe if I knew everything that was going on in the universe

other than the piece of paper I burned, I can still recover what I had written. But you know that's not true because think of it going backward in time, right? you're saying I'm starting with the post-scribbling universe, which has impressions that were created by my scribbles out there in the environment, but I've burned or I've destroyed the actual piece of paper I've written on. If I play the movie backwards and I'm using completely reversible laws of physics,

then I will see me backwards in time. I'll see me writing scribbles on the piece of paper. But I could also start in the future, after the scribbling, with exactly the same environment, but a different burned piece of paper, right? I am allowed to start with that initial condition, again, running backward in time.

So I have a different piece of paper with different scribbles on it, but the same environment that I would have gotten by scribbling my initial message, I can still run that backward in time, right? And I will get something. I'll get, might be a mess. might not be anything coherent because running things backward in time is a dicey game, but there's no inconsistency there.

So the point is it can't be true that knowing just what's going on in the environment is sufficient to fix what is going on in the part of the system that you've thrown away. There might be extra assumptions you could make about entropy in the future or the past or something like that that would enable you to do a nearly accurate job in recovering what you had scribbled. For example, if I put a lot of pressure on the pad of paper as I'm writing, then probably you can look at the piece of paper beneath it and see the traces of what I've written. So that's a very direct way in which you could accurately infer what had been written from the environment. If the piece of paper under the piece of paper that I'm writing on, it counts as the

environment. But there's no general rule that that's always going to be true. You could always be fooled by trying to do that. Richard Cashdan says, in Hawking Radiation, the creative pair of particles half fall out and half drop in. So how does a black hole evaporate? Yeah, the short answer is that the particle that falls in has a negative energy. The longer answer tries to explain what it means by having a negative energy. Think about what it means for a particle to have energy.

Okay, a particle, a baseball, or whatever, some object. What is the energy? Well, you say, well, E equals MC squared, okay, the energy is mass times a speed of light squared. But that's not the full answer because that's the energy of an object that is at rest, right? If an object is moving, then it also has kinetic energy, and that also counts. And then you have to say, well, moving with respect to what, et cetera. So you need to say, in some certain reference frame, the object has a certain energy, its rest energy and its kinetic energy and whatever, other kinds of energy that it might have.

So in a black hole, a black hole is not like empty space. A black hole has a location. There is a rest frame that sort of is the rest frame of the black hole. So there's an obvious choice to make about what reference frame to talk about things in. But the extra complication is that space time is curved. So it's not the, it's not quite what you mean by a reference frame that you would talk about in special relativity. it's a way of slicing space time into hyperservices moments of fixed time. Having said that, there is kind of a natural way to do it,

to slice all of space time into moments of fixed time of different times. And you can even say there can be an observer far away from the black hole who is essentially at rest with respect to the black hole. And then that defines a certain way of middle. measuring the energy of a particle that involves both the mass of the particle, but then also its velocity in that reference frame, things like that. So it's specifically the energy as it would be measured from the perspective of a stationary observer at infinity.

Okay? So all of those words matter to saying this. And the thing is that in general relativity where spacetime is curved, there's a formula that you can plug in, it involves the word's killing vectors. There's a formula that you can use to calculate the energy of an object as seen by an observer very far away. And if the object is outside the event horizon of the black hole, that energy will always be positive. That's just a rule. But if the object is inside the black hole, the energy can be negative. It could also be positive. It could be positive or negative. There's just no more rule about it because that particle

can never come to the outside world. It's inside the black hole. And when you go through the math and look at what actually the energy of the in-falling particle is in Hawking radiation, from the point of view of the external observer and infinity,

it would count as negative. And at that moment, right at the event horizon where the two particles separate, one of them has positive energy in that sense. One of them has negative energy. The positive one goes out.

The negative one goes in. The black hole loses energy. and ultimately it can evaporate. Oa asks, should we listen to the reflection before or after the podcast proper? So this is a bit of inside baseball for the Patreon supporters, because I do a little reflection piece, like five minutes of audio after every episode, and I record them after the episode.

So I'm just sort of giving some impressions about what I thought or why I had this person on or what I thought was interesting about it, etc. To me, it makes sense to look. listen to the reflection after listening to the episode. So you can also have your own reflections about it. But you do you. You do whatever you want to do. That's completely fine. There's some disagreement about whether I should post the episode by itself, the regular episode first or the reflection first. I think the problem is that different podcast readers show you the episodes in different orders. So some people want the reflection posted first, because

because the most recent thing posted appears to them first in the queue, and they want the episode to be first in the queue. So I try to do that, but not everyone's going to be happy because different podcast readers are different. Sorry about that. For those of you who are not on Patreon, the reflections are fun. I try to make them some value added, but they're not crucial. I'm not giving away any secrets there.

Peter 42 says, We fly toward a black hole with two clocks in our spaceship. One tells our local proper time. the second tells time for a far-away observer in flat space. We have calculated that we cross the event horizon at one o'clock proper time. As we approach the event horizon, the second clock will run faster and faster, eventually running billions of years ahead of the first one, that still has not reached one o'clock.

So before we cross the event horizon, the second clock will show a time that is older than the age of our universe, and we can still turn around in principle. Does this mean that things never fall into a black hole but fall onto it instead. If so, does this mean there's no actual singularity at the center of the black hole? So I've answered questions like this in various ways.

This is an old question. This goes back, I think people were talking about this in the 1950s, right? Before they knew that there were black holes, they speculated that the Schwarzschild solution, which goes back to 1917, describe some sort of frozen star because things fall close to the event horizon,

what we now call the event horizon, but from the point of view of the external observer, you never see them. Now, the question as asked is a little bit ambiguous because, well, I'll tell you the answer. No, it does not mean we never fall into a black hole. It does not mean that there's no singularity. None of those things are true. But the ambiguity in the beginning is you say, the second clock tells the time for a faraway observer in flat space. You can't do that. That is not a well-defined thing. I mean, you can have a clock that tells you some time that you interpret as the time measured by a far-away observer, but in relativity, there's no such thing as simultaneousity. So when are you telling the time of this far-away observer? There's no meaning to that. It would depend on how you sliced space-time, as we were talking before, into moments of equal time. There are, you know, you can argue there's more and less natural ways of doing that,

and therefore you can still sort of define a clock, but someone else can define a different clock just as well. So you could, absolutely, invent a clock that in some sense read, I don't know what the best way to say it is. You would have to say something like if you were to reverse course and go back at a certain velocity, then what would the clock get the person left behind be reading? But that depends on the details of your velocity, et cetera. I think the important physics thing about this setup is what would actually happen is that if you went very, very close to the horizon, this is unrealistic because part of you would cross the horizon and part of you wouldn't.

You're not a point particle, but forget about that. If you went very, very close to the horizon and then came back, what you would notice is you are now in the far future compared to when you left, right? That's what actually happens. you should only really ever compare the times of two things in general relativity if they're located at the same point in space. If they're located far away from each other, there's just no way to do that. So what you can do is go close to the horizon but not fall in, come back out, and now you're in the future. That's just gravitational time dilation. It doesn't mean that you couldn't have crossed the event horizon from your point of view.

That's the relevant, important fact. From your point of view, there's zero sense in which you naturally slow down near the event horizon, you just seem to fall in, and the singularity definitely gets formed. Armin Delennian says, in your recent talks with Ned Block and Tom Griffith's, a common theme emerged that the how of intelligence matters as much as the what. Block argued that internal mechanisms, not just behavior defined consciousness, while Griffith suggested that the specific algorithmic constraints of a mind are fundamental. to thought. This is in stark contrast to Hinton's, this is Jeffrey Hinton, the AI guy, a replacement argument, which suggests that silicon is functionally equivalent to biology. Isn't the Markovian static tensor's architecture of an LLM fundamentally different than a neuroplastic brain? Can we argue that it is a category error to call any system conscious

if its mind is physically frozen during the act of thinking? By treating consciousness as a process divorce from its substrate, don't we ignore that a subjective self requires the capacity for online permanent structural evolution where the act of experiencing is inseparable from the act of becoming? So I think I'm generally on board with what you're saying. Yes, that there are things. The mind is continually thinking. There's processes going on, as I mentioned earlier on in the AMA. And I do think that those processes, even if we are not consciously aware of the processes themselves are nevertheless constituents of what we call consciousness. They're important to consciousness. Now, I need to emphasize, I don't think that this implies you can't do

consciousness in silicon. Maybe you can, maybe you can't. In principle, I see no obstacle to doing it. You would just have to have all of those processes that are going on in a biological organism happening in the silicon, right? And it might not be simple. It might not be anything close to what we're doing right now. It might be very, very different than that, but I don't see why the actual material that you're making the minds out of really matters.

I do think that the processes matter. Martin Squibbs says, have you considered the possibility that physical reality is these sequences of our own sensations that we capture and store here in our mind? coming from our sense organs. Sensations that we interact with as human consciousnesses to formulate durations of factual past events

which we integrate into the world of our knowledge in mind. This would explain waveform collapse, with a waveform being the trace of the visual object of a particle within the formulated duration of a visual event, with each visual sensation representing a specific location in 3D space of the actual particle. So, sure. I've considered possibilities like that.

I mean, basically, that's the Copenhagen interpretation of quantum mechanics. When John Wheeler, who I've mentioned in the intro, coined the phrase, It from Bit, some people who actually didn't read his article interpreted as saying that the universe is made of information. What he's really saying is the universe is made of measurement outcomes, that what actually is reality is the set of sensations or measurements that agents or observers do in the universe, as opposed to the wave function itself being representative of reality. I've considered that. Many people have considered it. It's still quite a popular point of view among physicists and some philosophers even. I just don't think it makes any sense because I do think that there's

something called reality. I mean, if you want to, I mean, I get the temptation to say that observers measuring things brings reality into existence, but what are the observers then? I think observers are made of atoms, which are described by wave functions. So I don't know what the actual ontology is. You can't just say what exists are observers and their sensations. Like the observers have structure. They're made of things. I want to know what they're made of. To me, where the wave function describes reality, observers are part of the wave function. The thing they're observing is part of the wave function. It evolves in a very clear and known way. So I don't need to go to any of these philosophical difficulties to get reality out of them.

Ari Moody says, I was laid off over a year ago and I struggled to find work. I literally applied a thousand times, no exaggeration. I was starting to give up but finally landed a job and was with it for six months but now noticed that I'm laid off again. I'm in a really dark place because I fear having to go through at least another year or two of looking for work. I don't have much savings and I'm really struggling. I fear AI quickly eliminating careers.

I was replaced by AI twice, and I'm trying to learn it, but finding others know it so much better. My dream was to find a job in the space industry and have had no luck because of the instability in that industry. My question is, how can I get out of this funk and find the job that won't replace me in a few months? Should I pursue the space industry, or is it and science no longer viable in the current political trends? How not to give up? So I'm very sorry to hear this, Ari. I mean, this is a tough position that you're in. Two large a number of people are in positions like this.

It's kind of a bad aspect of the current moment that we're in. And I want to try to, you know, give you some encouragement here as much as I can. I don't know the details of your situation. So, as always, in these advice style questions, I'm going to, you know, it's really your job to come up with the answers to these questions. I can say things that might point you in a good direction or not, we'll have to see. But look, I think that there's two levels of this question. One is sort of a general question about how to live as a human being in trying times. Another is a more specific question about how to deal with AI and disruption in industries.

You know, it might be true that you were told that you were replaced by AI twice, but it may or may not actually be factual that that's what happened. There is this thing, oh, there's a phrase for it. I forget the phrase, but there's this thing going around. where companies do layoffs that they were planning to do anyway and they blame it on AI. Or they say, well, we're just replacing workers with AI. They were just going to replace the workers with nobody anyway. And this is an excuse that they're giving.

So I'm not sure whether certain jobs that are lost that are attributed to AI really are that. I do think that the total number of jobs is not going to change. This is always a thing where when economies are disrupted by technology, The distribution of jobs changes, but the number of things that human beings can do and will get paid to do generally doesn't change overall. Now, that's cold comfort if you're one of the ones whose career is being upended. I get that. It doesn't help any to know that there's some different thing that you're not an expert in that people are getting hired to do. But the only thing you can do there is the best you can.

Like really, that's completely useless advice, but I think that that's. the true fact about it, right? I mean, there's no once and for all general purpose strategy for dealing with these truly disruptive events. There is the higher level question about how, as a human being, you can sort of deal with all of this change and struggle and things like that. I think that the wonderful power and capacity of human beings is to extend ourselves through time, right? to remember the past and to predict and imagine the future in ways that are much more vivid and powerful and comprehensive than other species are able to do. That ability gives us enormous power, right, to plan for the future, to reason, counterfactually,

to cooperate based on the exchange of a few informational tokens in ways that, again, other species are not able to do and to reminisce and to tell stories in much more specific ways, but it's also a handicap sometimes because you can find yourself in the present moment focusing on other moments, right? Thinking too much about past failures or the prospect of future failures. And sometimes, again, even though it might not be very actionable advice, but at least the thing you should try to do is focused on the short term, right? Eventually you want to get to the long term, but sometimes you've got to focus on the short term and able to be able to get there. Maybe you have.

lost jobs or maybe you have failed to get new jobs, but maybe tomorrow you will. You just have to sort of think about the moment you're in. When you're surfing, you have to think about where the water and the wave is right now and where it's going in a very short window of time in front of you, not the whole stretch of, you know, where you started and where you might end up. And sometimes life is like that. Sometimes you're in a comfortable place where you can contemplate your retirement savings or whatever, or you're young enough that you're contemplating how to get trained for a future. But other times, you just have to live for the moment and try to improve that moment and build upon that, right? To make every day a little bit better than the last,

to keep trying to get a new job and you will often fail and sometimes you won't. And when you don't fail, that's going to be the time that matters. I once did a blog post listing all of the faculty jobs that I applied for in my academic career. It's not a thousand, but it's certainly more than 50 that I applied for and turned me down. And now it would have been larger than that. So, you know, it's failing and failing and failing over and over again is often the human condition. And that doesn't mean that you will always fail. There's a difference between failing 1% of the time and, or sorry, failing 99% of the time versus 100% of the time. There's a huge, huge difference between those two things. And it can be very depressing in every individual

instantiation, but you nevertheless got to keep going and things can change dramatically down the road. So I hope they do for you. Ben Lloyd says, sometimes I hear people say, debates are dumb. They don't change anyone's mind. They're pointless. However, I completely disagree. Sure, debates involve other skills and the person who's correct doesn't always win, but debates show ideas clashing together and oftentimes it can expose fraudulent ideas. Maybe not for everyone watching the debate, but for at least some. I think your debate with Eric Weinstein swayed the public sentiment a bit more over to the heterodoxy side, rightfully. It is a good example of debates being useful. What do you think about debates and the people who say they aren't worth it?

One reason I'm answering this question is because it gives me an opportunity. I'm not sure if I've done it before or not, but to point out, I completely goofed when I was on Pierce Morgan and I mixed up heterodoxy and orthodoxy in the heat of the moment when I was giving my first answer. My very first answer was like it was not under pressure or anything, but some reason there was a little brain glitch. And I meant to say that I was in the position of defending orthodoxy. And I said heterodoxy, which was completely 100% wrong. Sorry about that. I said the opposite of what I meant.

But as far as the question is concerned in the value of debates, I kind of agree that there is value to debates. I do think that a bunch of people don't like them precisely because they conceive of them as having a purpose that is not the purpose they actually have. If you think that what a debate is going to do is to find the truth, then you're liable to be disappointed. It's not the way that we actually, like in real academic research endeavors when we're trying to actually look for the truth, we don't have a debate between each other. We don't formalize back and forth like that. We talk and we ask questions and we admit when we're wrong and we try to learn things, right? Those are the ways to actually make progress, not to have a debate. But there are other things you might want to do in life other than find the truth.

You might want to educate people. You might want to inspire people. You might want to provoke people into thinking in new ways. You might show people by your example that people who have your side of an opinion are into that bad after all, right? There's lots of ways in which you can have a positive impact through performing in a debate. I haven't actually done that many debates over the years.

I've done a few that have gotten some attention. But they're mostly for inspiration, education, demonstration purposes, not for purposes of finding the truth. And I think if you look at them that way, they can be net positives to the world. Alex DeBrow says, you mentioned before that the scientific literature is being polluted by junkie AI written papers. I'm curious what you mean by AI written in this context. It seems there's a big range between a paper that's basically generated by AI with very little human involvement and a paper where someone uses AI to help clean up or reorganize something they've already written.

Sure, there is a big difference between those two things. I mean papers that are basically generated by AI. There are papers appearing in the literature and on the archive, etc., that have no human involvement except to submit them, I guess. Maybe not even that these days, if you've been following the growth of AI autonomous agents. But they don't even attribute the papers to a human author. They make up a name for the author of the paper because they're just lying. There's no actual value in these things at all. That's what I'm complaining about when I say junkie AI written papers. Those are just bad. They serve no purpose whatsoever and they make the world a worse place. There's, of course, an interesting question about this continuum that you point out between that and a human being,

just writing the paper all by themselves. And somewhere in between, there are papers written by humans who have consulted with AIs, used AIs to help them understand some aspect. And there's other papers that are written where people ask the AI question and then cut and pasted the answer. I think that's the cutoff for me. Like if you use AI to educate yourself, to make your brain smarter, to teach you something, and then you use that knowledge in the writing of something, that's fine. I'm all in favor of that.

I mean, maybe acknowledge it in the acknowledgments or whatever, the references or what have you. If you're letting the AI do the writing that I think you've crossed a line, and maybe for some purposes, that's okay. You can let AI do the writing. Like if you have to, my personal, my personal standard is I will let AI write something for me if I don't think anyone will ever read it. Sometimes you just like need a paragraph for something like, you know, on a webpage or a, you know, progress report or something like that. And you're pretty sure that it's just bureaucratic treadmill running. It's not actually serving any purpose. And so why not let the AI do it?

But I would never author something with my name on it where any of the sentences were written by an AI. Let me say that. If you ever read a book by me or an article by me or anything like that, everything was written by me. I might have used an AI to think about what I'm writing about, and I think that's a perfectly good use. But there are people who are cut and pasting from chat GPT or whatever into their paper or even letting the chat GPT write most of the paper. I think that, you know, that might be okay, but it has to be honest. It has to be listed as something that was generated by an AI, not by a human being. It's bad when AI's write paper with no human involvement.

It's also bad when human beings put their name as authors on things that were largely written by AI's. Both of those are misrepresenting the truth in a real way. Stan Monolov says, this has been mentioned in the past but I still have not wrapped my head around it. There are many quantum events in our bodies per second, but many branchings, many branchings happen without us observing them. How can we meaningfully talk about the total number of branching events?

Well, my philosophy about this is we shouldn't talk about the total number of branching events. It's irrelevant. This is a thing where, you know, it's one of those things that is a difference in the discourse. about many worlds between the people who are experts in it and the people who just hear the popular level discussions. In the popular level discussions, we often get a lot of questions like, where did all the worlds come from, where does the energy come from, how many worlds are there, all these questions. And the professional discourse about this has none of those questions in it. The energy question or where the worlds come from, that is just completely obvious, once you understand the equations. how many worlds are there might just not be a well-defined question.

David Wallace, former Minescape guest, has said it's like asking how many experiences you had yesterday. On the one hand, you definitely had experiences yesterday. On the other hand, assigning some fixed number to them is not very well defined. The answer might be infinity. That might be the only answer you could give, or the answer might be something a little bit, it might be finite, but it might be dependent on choices that you haven't explicitly made. So, you know, it's up to you how you define different things. But who cares?

What matters is the fraction of worlds in which a certain thing comes true versus another thing comes true, and that has been studied to death. And we think that it is the good old born rule that the probability is the wave function squared. P. Walder says, does the fact that CERN has not found any new particles at energy levels within that prescribed by the core theory rule out panpsychism as an explanation of consciousness? Or does the notion as promoted by Philip Goff that existing known particle properties like spin-in-charge are the basis of consciousness have credence? Those are two very different things that you're asking in one question. The answer to both of them is no. I mean, on the one hand,

you can't rule out panpsychism. You can rule out the idea that it's necessary. You can demonstrate that you don't need to invoke mentality everywhere as an explanatory move. in trying to come to an understanding of how consciousness works. But it may be, you know, who knows? Like, it's not something that can be ruled out. It's like ruling out things that are outside our observable horizon. You can always invent a version of panpsychism where there are mental properties that have no causal impact on the world.

How would you rule that out? It's kind of meaningless, but you can't rule it out by doing some experiment at CERN of all places. Having said that, the idea that known particle properties such as spin in charge are the basis of consciousness just makes no sense at all. Like, why those properties, you know, spin in charge do have causal impact on the world, knowing the spin of a particle. That tells me how it's going to behave in an in homogeneous magnetic field and so forth. So that's just a physical property. It's a completely different thing than consciousness. I think that what Goff is getting at is the idea of,

that there could be extra properties, not like, not spin-in-charge, not such as spin-in-charge, but like spin-in-charge in the sense that they're properties, but nevertheless are not physical properties and therefore do not affect experimental outcomes. And those mental properties,

to me, if they don't affect, if they don't affect any experimental outcomes, I can make all exactly the same predictions and get all the same level of understanding without invoking them, and so I'm going to do that. Quantum Chaos asks a, priority question. Whenever I hear people talking about candidates for dark matter, they always seem to

imply the answer will be one of the candidates, Wimps, axions, neutrinos, etc. Is there some property of dark matter that tells us it's all the same thing, or is it possible that what we're calling dark matter is some combination of multiple things, like half of it is Wimps, some black holes, some neutrinos mixed in? It's 100% possible. Yes, that is completely possible. It's just, there's no motivation for looking for that specifically. In fact, people have People have noticed that, you know, if you're doing an experiment where you're looking for the ambient dark matter that is passing through a detector, if half the dark matter is wimps and half the dark matter is axions, you can loosen the constraints on the properties of either particles because the density of either one of them is only half of what it would have been. But still, half is a pretty substantial number.

And it just seems weird that it's already hard enough to come up with dark matter candidates. You need stable particles with the right relic abundances that are electrically neutral and not strongly interacting, et cetera, et cetera. All of these properties, and they just happen by complete coincidence, have approximately the same density in the universe? Okay, yeah, it's completely possible, but there's not a lot of motivation for it. Not Gufferin says, provide cat updates. Technically, that's not a question, but an imperative, but that's okay. Although, you know, actually I was thinking about it, it's hard to do, provide cat updates. The thing about cats, one of the things about cats, if they're settled in, Ariel and Caliban, my cats are settled in now.

Puck, our previous stray cat, who we fostered for a while, has been gone for roughly a year. The cats are settled in. They're equilibrated. There's not a lot of change from day to day. Both Ariel and Caliban are approaching their ninth birthday. They're going to be nine years old very soon. They just went to the vet. They got perfectly clean bills of health.

I guess the one thing is it's very interesting because Caliban is the more extroverted cat, as I mentioned before, Ariel is more introverted. Caliban is like bigger and more robust than everything. But Caliban was the one who was really affected by when we introduced them to Puck. So we tried to make Puck a member of the household. and Puck just attacked Caliban pretty viciously, to be perfectly honest, and we knew that we had to do something different. Ariel was not attacked because she was gone. She was nowhere to be found.

Caliban, being the extrovert, was trying to make friends, and that did not go well. And one other time when we weren't even here, Puck escaped his confinement and also had an incident with Caliban. And Caliban, despite the fact that he is the extroverted happy-go-lucky cat, did not psychologically react well to being attacked by Puck. He basically went on a hunger strike. It's not the first time he's gone on a hunger strike.

He's sort of naturally a relatively chunky cat, but when he's upset, he goes on a hunger strike. So once I was on sabbatical for a few months, I was away for a few months. He went on a hunger strike then. Jennifer was taking care of him, plenty of food, but he just wasn't happy. And now when Puck and he had their incident,

he once again was not happy, even though, you know, after that, puck left and still. So for better or for worse, Caliban is very svelte right now. Like the vet is just completely in awe of how good looking Caliban is because he is not overweight at all. He's lost all of it. He's sort of rebounded to the ideal cat weight. And most indoor cats are not at the ideal cat weight. They're a little bit on the high side from that. So the vet is just amazed at how, what good shape Caliban is in. So Ariel is still pleasingly plump because she was not as psychologically affected by all

of these incidents. Otherwise, you know, they're in great shape. They love us very much. They live in a big house now. So there's plenty of room for them to each have their areas. They have their areas in both space and time. So there's certain times when Ariel will be with us, certain times when Caliban, certain times and they're both in the same room and whatever.

negotiated, you know, treaties. They don't, they don't want to be in constant contact with each other, but they don't, they don't fight or anything like that. Sometimes Caliban will chase his sister and Ariel will hide, but that's about the extent of it. Otherwise, they get along very peacefully. They share the sunbeams and everything, and they're endlessly adorable and cute. That's the cat update. As your propagation says, I'm wondering if you could entertain a shower thought that I had while taking chemistry 101. Since carbon, has four unpaired valence valence electrons. It generally always requires another atom to covalently bond. But we're told that diamond is just a crystal with carbon-carbon bonds. This begs the question. You actually mean it raises the question, but that's okay. It raises the question, what happens at the exposed face of a diamond? In my humble opinion, there is no obvious answer for what bonds the dangling electrons.

Is a diamond secretly a center of carbon surrounded by a shell of a shell of a hydrogen, but then what would happen if you grew a diamond in vacuum? So these are great questions, and I'm not the person to answer them. This is just not my kind of physics. My impression is that, yes, the outer edge of diamond is actually not carbon. Like it adheres to something else, like either hydrogen or water or oxygen. I don't know exactly what is going to be the most common thing that it adheres to. But a one atom thick layer of something might as well not be there.

for all intents and purposes, in terms of, like, what the diamond looks like when you shine light through it or whatever, you would never know. To a chemist or a materials physicist, it would be fascinating to know and to study what is happening at the surface because you do have that extra bond

that's got to go somewhere. My guess is hydrogen is mostly what it is. What would happen if you grew a diamond in vacuum? There would probably be some kind of frustration at the edge and the carbon would try to bond to other carbons with double bonds or something like that, but again, totally not what I do for a living. That would be interesting to look up.

Enrique Ariola says, I understand why the laws of physics are time symmetric. Would you please use radioactive decay to illustrate how it's also applicable to those processes as well? And if entropy is anything to do with why we observe half-lives of different radioactive elements and not an opposite process where elements randomly fuse resulting in heavier elements. So, of course, we do.

observe the reverse process, it's just less likely. When a heavy element decays into lighter elements, that's fission. The opposite is called fusion, right? And we observe it all the time. Now we usually, I mean, observe is a strong thing. It's happening at the center of the sun all the time, okay? So you usually need a high density of particles to make something like that happen and also a lot of energy. All by themselves, you will indeed get particles decaying more than particles fusing together. And that is just a matter of counting the number of states you can be in. If you have a single particle sitting there all by itself, there's basically unique state up to it can have different positions of momentum

and things like that. But there's basically one way to arrange it. If that particle decays like a neutron decaying into a proton, electron, and antineutrino, there's many arrangements of the proton, electron, antineutrino that still have the same. center of mass, momentum, and things like that. So there's more ways for the system to be a proton, an electron, and an antineutrino than there are for it to be a neutron. So just like entropy increases, heavy particles tend to decay into lighter ones. The result has higher entropy.

Now, does it affect the rate of decay? A little bit it does. The technical term, if you're going to look up in your quantum field theory book, how to calculate the lifetime of an unstable particle is the phase space volume that is allowed for the decay. So when a neutron decays into proton, electron, and antineutrino, you can sort of divvy up the momentum a little bit differently while conserving the overall momentum between the three particles. If you're only decaying into two particles, if one particle decays into two, there's a unique answer for the total energy that goes into each particle.

each of the two decay products because they're conserved, but there's still ambiguity because they can go back to back in all sorts of different directions, right? Once you have three particles, you have more ambiguity than that. So there's more phase space for the particles to end up in. And indeed, one of the reasons why neutrons have a relatively long lifetime, minutes. Minutes is a huge lifetime compared to most particle physics processes.

one of the reasons for that is because there isn't that much phase-space volume. And the reason for that is because protons are nearly the same mass as neutrons. There's not a lot of wiggle room. If the neutron could somehow decay into two electrons and two positrons, it can't do that because it would violate baryon number, but let's say it could. Those particles have a total mass much less than the original neutron. So there's a lot of room for that to happen, and it would happen very rapidly. It doesn't because varian number is conserved, and so that does play a role in calculating the lifetime.

The other thing that plays a role is the strength of the interaction itself. A strong interaction decay will generally happen faster than a decay mediated by the weak interactions. Okay, Leo says, I enjoyed seeing you on Alex O'Connor's podcast recently. As a fan of both of yours, I found that there seemed to be two schools of thought in philosophy. today. With people like yourself and the late Dan Dennett representing a more practical, scientifically-minded approach, and people like Philip Goff and Thomas Nagel representing an approach that seems to rely more on human-level concepts, like purpose and qualia, to explain more fundamental aspects of reality. Have you noticed this as well? And if so, why does it seem to be so difficult to get philosophers to take modern science seriously when constructing their ontologies?

I think there definitely are different schools of thought. I mean, I think that what you're pointing at is simply physicalists and non-physicalists about the fundamental ontology of reality. Do you need more than the physical world to explain what is going on? It's not necessarily, strictly speaking, a matter of taking science seriously. You can take science seriously and not be a physicalist. But I do think that there's sort of a natural connection, a natural correlation, let's say, between taking science seriously intending to be a physicalist. But not necessarily for the reasons you think.

I think that the more you understand science, the more you appreciate its unforgiving nature. You know, if you don't know much about physics or biology or whatever, you might say, well, sure, the theories say this, but what about this other possibility? Whereas the well-trained scientist will say, well, that possibility can't work for the following three reasons. You know, a really good scientific theory is actually not that flexible, right? it's not that floppy. It's actually pretty rigid in where it starts from and where it goes. And I think that both Goff and Nagel in different ways are pretty happy to imagine sort of modifying theories in vague ways. Like how hard can it be?

I think the answer is very hard. Thomas Nagel, for example, you know, who is one of the leading philosophers of modern times, a super smart guy, there's no doubt about that. He wrote a little book called Mind and Cosmos, which is not his best. book. But in it he, you know, it's his version of my book, The Big Picture. But in my book, the big picture, it's very long. Mine and Cosmos is very short. And also it takes the opposite tack. Like my, my angle in the big picture was to fit everything into a physicalist framework. And Nagel's goal in Mining Cosmos was to reject a physicalist framework. But it's just not very good.

And because he says things like, you know, yeah, biology, you know, evolution, natural selection, section, Darwin, it all seems to work pretty well, but I just don't really think it can get all the way to explaining the marvelous complexity of a contemporary human being or other living organisms. That's it. I mean, it's just sort of like a vague feeling, right? It's not quantitative. You know, he says, I'm not an expert on these things.

That's just my feeling. I do think that the more you become an expert in it, the more you appreciate how hard it is to deviate from the strict boundary. of the physical theories that we have. Alexander Knurkel says, I'm still baffled how wave function collapse in the Copenhagen interpretation does not completely destroy time reversal symmetry

since they collapse in a non-unitary way but never uncollapse. Why doesn't it? It does. There you go. That's the answer. There's no time reversal invariance in the Copenhagen interpretation

of quantum mechanics. Now, you might wonder why some physicists who would profess to believe, in the Copenhagen interpretation would nevertheless say that, you know, the underlying laws of physics are time reversal and variant. It's because they're not including measurement when they talk about the fundamental laws of physics. When particle physicists in particular talk about measurement or sorry, talk about dynamics or Lagrangians or Hamiltonians or whatever, they're only talking

about the evolution of the system when it is not being observed. That's all they're talking about. So when they say locality, a very big sort of vocabulary difference is that if you asked a particle physicist, is known physics local, they would say, yes, of course it is. And they would point you to the Lagrangian for the standard model of particle physics, which is 100% local. If you asked a person working in the foundations of quantum mechanics, especially a philosopher, is known physics local, they would say, of course not. We all know that. We've known that since John Bell and his theorem. And they're thinking about quantum measurement. They're not thinking about the Lagrangian of the theory.

So they're both right. They just both mean different things. Kyle Koppasares says, you've mentioned that you knew you wanted to do theoretical physics from a young age. Were you self-studying out of college-level textbooks before going off to college and trying to seriously answer unsolved problems in physics in your pre-college era? So no, I was not one of those people. I know what you mean. There are people who, you know, are not only. only clear about their goals, but take the initiative and are, you know, very enthusiastic and

energetic and are going to move ahead of the curriculum they're being exposed to. I was just not that person. And I think I say that as an apology because I wish I were. Like now, if I knew then what I know now, I would totally have been that person. I would absolutely have educated myself and taken the initiative and moved ahead. But I wasn't. I was. just a rule follower. I took the courses that I was offered and, you know, I took as many courses as I could, but I didn't, you know, just, I could have just bought the quantum field theory textbook and read it, right? But that never, I don't know, that option wasn't there. And I think that, you know, part of that is my personality. I'm not a, you know, rock the boat

kind of person. But another part of it is the difference between growing up in an academic environment or a privileged environment and not, right? I didn't know anybody who was academic or, you know, some of the adults I knew had gone to college. Some did not, but none of them really went to, like, get a PhD or anything like that. So I had really no conception of what it meant to be an academic, to be a professor, and how to do it and how to succeed at it. So I just kind of went along with what I was told I was able to do. And I regret this very, very much. And I much now, just think how smart I would be right now if I really started early enough to learn a lot more, a lot earlier on when my brain was just growing and ready to learn new things.

P.T. Milo says, in your interview with Thomas Hurtog, you asked whether his top-down cosmology can show that an otherwise empty universe with a Milky Way was less likely than our universe, which I saw as a version of the Boltzman brain question. He primarily responded by referencing papers in progress on this issue, although he did discuss this in the book. He has since come out with papers that argue from a holographic perspective that the full universe Big Bang complexity represents a saddle point phase transition given the top-down starting point, making it more likely than seemingly simpler Boltzmann-type stories. I'm not sure if you've kept up with his argument, so my question would be more general. Isn't he still begging the counterfactual

question here, irrespective of phase transitions in saddle points? Meaning, is he genuinely pointing out an alternative to the measure and typicality problem, or is he just conditioning on a very large slice of what we observe today, not just the Milky Way, but evidence of the microwave background, et cetera, and therefore merely claiming that such a highly specified result is unlikely to be Boltzmann-esque while simply refusing to condition on simple counterfactuals such as a lone Milky Way without other features of our universe. Well, I'm trying to answer this question. I won't answer it very well because I'm not actually familiar with these new papers of Tomas. I mean, he's a smart guy. He's worth paying attention to. So I would, uh,

assume that these papers are saying something interesting and correct in their domain, but I do think that there is always a danger in this game where you're trying to come up with a cosmological explanation of our universe. The problem being that we know a lot about what our universe is. So it's not like you're making the theory first and then making the prediction. You know what the prediction is you want to make. And this is a very, very common problem in early universe cosmology and quantum cosmology, where people find the answer they want and then they stop. So for a little bit of background, for those of you who don't know really what's going on here, the idea is if what you want is to explain a cosmology with us in it,

with people like you and me, what exactly do you need? What exactly do you need to have you and me in the universe? And you might think, well, we arose from a big bang where galaxies formed and stars formed in planets and eventually life and eventually us. We think that that's actually how it happened indeed, but you don't need that just to get us. In fact, you could just have a much, much smaller collection of matter come together to form, let's just say, the solar system, okay? And you might say, well, but the solar system needs heavier elements, so that's okay, put the heavier elements in the initial conditions. And you can just put that together, make a little solar system, make us here without any other stars or galaxies in the universe.

and that might seem like a weird universe, but simply in terms of the counting of how many universes we can imagine that look that way versus the actual universe we live in, there are a lot more universes, and this is a rigorous counting that you can do at the level of the equations, a lot more universes that look just like a little tiny solar system coming into existence than the trillion galaxies that we see in our universe. So why in the world would you go through this phase of the hot big bang, which is very low entropy, very apparently finely tuned, et cetera?

That's the challenge for cosmology. And very few people actually try to meet this challenge. And so at least Tomas is trying to meet the challenge, whether or not he's cheating by excluding some even more likelier scenarios than what he claims his theory predicts, that I really don't know. I do think, you know, it's something to try to do, try to find something special, something about our actual universe with the very, very low entropy beginning that somehow makes it better than the fact that it's just a very tiny, tiny, tiny fraction of all the possible universes. Maybe there's some other reason why it pops up. And I think what Tomas is trying to say is that if indeed you didn't have any need for any observance. observers at all, then you would just live in empty space. But once you conditionalize on having observers, in his theory, not in just the generally

counting over all possible universes theory, in his theory, he's claiming that the way to get people like us is through the Big Bang. Whether or not that actually works in his theory, I am not qualified the judge right now. Sean Miller says, if physicalism up here above the Planck scale is necessarily coarse-grained, What makes some coarse grainings objectively better than others for describing agency? In particular, do you think that there is a fact of the matter in the microphysics that picks out boundaries and macro states with causal autonomy,

the kind of partitions where mapping, control, and error correction become real, or is that selection always observer chosen? Well, I think observers choose how to... There's an old question. Sorry, let me back up. This is an old question that appears in many different contexts, not just agency, but just any context of emergence. when you have the emergence of fluid descriptions from kinetic theory, atomic theory, molecules, and things like that, who chooses what the coarse graining is, right?

And as I like to say, Laplace's demon doesn't know about the coarse graining. Laplace's demon just works with the microscopic constituents from the start. We human beings who are not Laplace's demon have much less information, and therefore we work with coarse graining's. So in that sense, the course gratings are observer chosen. different observer, one with perfect information, wouldn't coarse grain at all. And I think that people misunderstand this claim, even though it's a true claim. They misunderstand it to think that the coarse grains are somehow arbitrary, like they're just made up by human beings. But the existence of the fluid description of the gas around you, of the air around you, isn't

dependent on the existence of human beings. It's not like there were lots of different ways to coarse-grained the underlying microscopic description so as to get a good higher-level emergent description. It turns out that there are very few ways of course-graining the microscopic things to give you an autonomous higher-level description. The thing about temperature and pressure and other fluid dynamics variables is that when you know those, you have an enormous amount of information, not quantitatively information, an enormous amount of useful information, even though it's a relatively small number of actual physical quantities, that you can use to talk about what's going to happen next, right?

You can predict the weather. If you had arbitrarily coarse-grained versions of the atomic structure of the atmosphere, you would not be able to predict the weather at all. So I think the same thing is true with agency. I think that if you talk about, you know, human beings or organisms or organs or cells or whatever, these are all higher-level coarse-grained things, but they're coarse-grained for that reason. There's a reason why you pick the course-graining

to be those particular maps from micro to macro rather than something else. And the reason is that leaves you with enough information to make real predictions, to have real understanding. So observers use the existence of these useful course graining, but it's not just that they are arbitrarily picking them. Okay, I'm going to group together two questions that are related to each other.

One is from Michael Bright, who says, my understanding is that causality does not hold in quantum mechanics because individual actions at the subatomic world are random, probabilistic. For example, we don't know the moment one atom will decay. But I've always had a difficult time squaring this with probability theory in which the occurrence of one individual event is irrelevant. What matters is having a statistically significant sample size of random events and then the bell curve of their collective behavior is what matters.

Borrowing from the radioactive decay example, isn't this exactly why carbon dating works so well? Surely in the universe, events are comprised of enough particles to make their probability distribution the thing that matters. Or am I missing something slightly more profound, such as that perhaps the cut between the quantum world and macroscopic world occurs somewhere right around the moment when you actually do have a statistically significant sample of particles at play in an event?

And then Scott Collins has a much shorter question, simply, does causality require time? So I think both of these questions are about the relationship of the idea of causality or causation to microscopic physics, to quantum mechanics or to time or whatever. So guess what? It depends on what you mean by causality, causation, or whatever is the vocabulary word you're choosing to use. I would not say that causality does not hold in quantum mechanics. I don't think that's the right way to think about it. But to explain the right way to think about it, you do need to think about the word causality.

What do you mean? So there's sort of an old-fashioned sense, an intuitive sense, a folk physics sense of what do you mean by causality, which is that event A leads to event B in the sense that when A happens, B happens. If A hadn't happened, B would not have happened, right? That's what it means to say that A caused B. I was late for work because there was an accident on the road when I was driving in, right? If the accident hadn't been there, I would not have been late. That notion of causality, number one, does require time to answer Scott's question. Number two, is nowhere to be found in any theory of microscopic physics, classical physics or quantum physics, right?

Because the laws of physics are reversible. The folk version of causality has an arrow of time. causes always precede the effects. Whereas in classical physics, in Newtonian physics, the information you need to predict one state from another doesn't have an arrow of time. You can predict what's happening at T not from what is happening at T0 minus 1 or from what's happening at T0 plus 1 works equally well. Now, quantum mechanics make things more complicated. But I would say that therefore even in classical mechanics, that form of causality doesn't exist. It's not that everything has gone crazy.

It's not that anything goes, right? There's still laws of physics, but the laws of physics don't take the form of cause and effect relations. They take the form of differential equations or some perfect mapping from one moment of time to another. Mostly the same thing happens in quantum mechanics. In quantum mechanics, you have the Schrodinger equation. It's just like Newton's laws. It maps things, states at one moment of time to the other moments of time, either forward or backward, equally well. Now you have, of course, the additional thing in quantum mechanics that there are measurements, and measurements collapse the wave function from the point of view of observers, and there we have the measurement problem.

What's really going on? People disagree. If you're in Everettian, you think that there's no true collapse. It's just apparent to people living on individual branches, and therefore there's no causality or arrow of time in the fundamental physics in Everettian quantum mechanics any more than there is in Newtonian mechanics. But let's put aside the difficult questions of the measurement problem, et cetera, and just think about, you know, the Copenhagen interpretation or textbook quantum mechanics where there are collapses of the wave function and they are truly probabilistic. That doesn't still mean that there's no cause and effect, you know. As Michael's question implies, I guess, it's just that the cause and effect relationships, the relationships, I shouldn't say the cause and effect. cause and effect relationships. The relationships between the state at one moment and the state at another moment when you've done a measurement have a stochastic element to them. They still exist. I mean, maybe the probability of getting a certain spin measurement is 0.99999, right? But maybe it's one for that matter. Or maybe like when you do many, many measurements, even if it's 50-50,

if you do many measurements, you'll get a probability distribution. All of these are perfectly well-defined quantitative statements you can make. I wouldn't call them cause and effect. but there's still relationships, there's still laws of physics. It's just that there's a probabilistic element to them. So I think you just have to, you know, get over the idea that if you have a relationship, it has to be called a causal relationship. That was fine, 2,500 years ago, but nowadays we have to think a little bit more carefully about what the kinds of relationships that we have are. Finally, to go back to Scott's questions about whether or not causality requires time, again, depends on what you mean. Usually, like I said, we have a lot.

an idea of causes and effects as unfolding in time, but not always, you know, NERTER's theorem says that you have symmetries giving rise to conserved quantities, like the conservation of energy comes from the symmetry of time translation and variance. The laws of physics are the same at every moment in time. So that's true at every moment. There's nothing different from moment to moment. Energy is just conserved because laws of physics are time translation and variant. That's an eternal, timeless fact. But do you want to call it a cause? Do you want to say that energy is conserved because the laws of physics for time translation invariant?

You could. I mean, that's a thing that you could say if you know what you're doing. Because it's certainly true that if the laws of physics were not time translation invariant, then energy would not be conserved. If the gravitational constant were varying with time, then the energy of two systems moving around each other in orbits would change over time. So that's a different kind of cause, but it's the kind of thing that people still nevertheless attach those words to. So whether you want to attach those words or not, that's up to you. You just have to, again, keep your wits about, you know what you're talking about.

Michael Long says, in a mathematician's apology, G.H. Hardy talks about the elegance of mathematics, which he demonstrates by proving through infinite descent that the square root of two is irrational. Have there been times in your career when doing the math instantiated some physical reality in a way that seemed particularly artful or surprising. I mean, not in the really deep sense, I would say. Most of the papers that I've been a co-author on or an author on are either more conceptual or we do some calculation that was pretty well, you know, pretty easy to know what to do, right? If you're asking, okay, how does a polarization state of a electromagnetic wave change if you

violate Lorentzen variance, you just do this. the calculation. You just crank through it, right? If you want to know what is the relic abundance of this dark matter candidate, you just crank through it. It's not like you're seeing some deep mathematics there. Probably the closest. I mean, I have done more advanced mathematics. I wrote papers on, you know, free variable loop equations in 2D Euclidean quantum gravity and papers on calculating homotopy groups, vacuum manifolds and spontaneously broken symmetry theories, things like that. So I've done that. But it's still, you know, known mathematical. wasn't surprising, I guess, whatever we figured out. The kind of thing, the closest I could come

was I wrote a paper very early on when I was still a grad student, or maybe we finished it when I was a postdoc. We were very slow on time machines in two plus one dimensional gravity. The idea is in two plus one dimensional. So three space time dimensions. So space is just two dimensional. So it's flat land, but it's gravity. So it's not really flatland. You have particles. And in fact, Einstein's equations in three dimensions are very special. The amount of space-time curvature, unlike in four dimensions, four spacetime dimensions, you have space-time curvature caused by the Earth or the Sun or some other gravitational field, some of the gravitational source, I should say, but the curvature spreads out from that source. So the Sun creates a gravitational field

around it and then the Earth orbits in it. In three-dimensional gravity, the curvature of space-time only exists where the source is located. In between the particles or whatever, it's completely flat space-time. So that makes it much easier to analyze what is going on. And we looked at, this is with Alan Gooth and Edie Fari and Ken Olam. We looked at trying to make close-time-like curves in two-plus-one dimensional gravity, inspired by an idea of Richard Gott. And it turns out that you needed to like parallel transport some vectors around some curves, which is fine. You know, physicists know how to do that. And the resulting transformation of the vector is a Lorentz transformation.

That is to say, it is some combination of a rotation and a boost. You rotate the vector in space and then you change its reference frame in space time. And it turns out that the set of all possible Lorentz transformations in three dimensions, technically S. S-O-2 comma 1. So this is a little bit of math, for those of you who care about math. Every Lee group, every continuous smooth symmetry group, in other words, not like a discrete group like Z2, where you just like take a reflection and then reflect back, but a Lee group is a smooth transformation, like a rotation or a boost.

The Lorentz group is a Lee group. Every such group can be thought of itself as a manifold with a metric. So there's a little bit tricky. because the group transformations are manifold with the metric, and you're applying them to a different manifold with a different metric, namely the actual space time with particles in it that you care about, okay? But the set of all Lorenz transformations is itself a space time. Sorry, yeah, is itself a manifold with a metric? And in 2 plus 1 dimensional gravity, that manifold with metric is three-dimensional, and it looks like a three-dimensional space time. In fact, what it looks like is anti-de-sitter space.

This is back in like 1992, we were thinking about this. So before ADS-C-F-T was a thing. So we were thinking and drawing conformal diagrams for antide-sitter space long before ADS-CFT became a thing. So we were thinking about a totally different thing, a much less interesting thing, whether or not you could make time machines in 2-1-dimensional gravity. Anyway, I had an insight that we were proving that certain,

collections of particles in 2 plus 1 dimensional gravity could not cause closed time-like curves. And we phrased that as a restriction on what possible Lorentz transformations you could get from parallel transporting vectors around closed loops, halonamies, as they are known. And I was able to show that we had a very, very complicated proof, that certain things couldn't happen. And I was able to completely replace our complicated proof with a single diagram. A well-chosen plot of what was happening in antideocidder space was enough to prove our proof, even though we were not in anti-de-sitter space, but the set of Lorentz transformations looks like has the shape of antideciter space, and you can make statements about possible geodesics in antide-sitter space that corresponded exactly to our very long, complicated algebraic proof that you could not build time machines in 2-1-1-dimensional graphics.

So, okay, for most of you, that made no sense what I just said, but it's just a little, is it trying to answer the question, like, has there ever been a moment where math met upon to some physical reality in a way that seemed particularly artful or surprising? I think for me, that's the closest I've ever come. Gregory Kusnik says, in your conversation with Ned Block, you both emphatically agreed that our cells do experience the passage of time and conscious experience is intrinsically temporal. So now I'm confused because the episode with Doris Sal left me convinced that eternalism is correct, and the continuity of consciousness and our apparent experience of time passing are illusions born of momentary experiential snapshots

coexisting in short-term memory. Can you unconfuse me? Well, you notice that if you actually listen to my own words, I almost never use the word illusion in these contexts. I know other people do. They use it a lot. There's a whole school of illusionism. about consciousness. But they don't actually think consciousness is an illusion. They don't really think that. They just have chosen the word illusionism, sadly. I am not saying that the passage of time is an illusion. I think it is an emergent effect.

I think those are different things. We have the impression that time passes, which means in a very carefully phrased way, that we describe the world in a higher level folk physics kind of understanding that includes the concept of the passage of time. And you can derive that from underlying more fundamental laws where you're an eternalist and time doesn't look like it's passing. But just because it's higher level emergent doesn't mean an solution. It's perfectly fine to talk about the passage of time in exactly the same way.

It's perfectly fine to talk about tables and chairs. You just have to realize they're not there in the fundamental laws of physics. That's all. So I have no problem saying that cells experience the passage of time. cells are constantly taking in sensory data, and sensory data might be a slightly exaggeration, but data from their external environment, they're churning out ATP and things like that. Time is passing. Things are happening from the point of view of the cells. Elias says, you say, inevitable heat death of the universe. But I've been rereading from eternity to hear, so maybe there is solace in the continued creation of baby universes. There are so many AMA questions I could ask inspired by that book,

but here's one that is inspired by this pregnant decider scenario, i.e. empty space giving birth to baby universes. You write that the baby universes are created via quantum fluctuations. Can you make that a little more precise? Normally, I think of these quantum fluctuation explanations as dependent on a measurement slash decoherence, which requires an external observer or environment. How does that work here?

Well, I think you're right to be, you know, very leery of the first. phrase quantum fluctuations. They mean different things. That phrase means different things to different people. I wrote a whole paper about that with Jason Pollock and Kim Boddy about quantum fluctuations into sitter space. So sometimes when people say quantum fluctuations, they just mean the fact that quantum

mechanics is not the same as classical mechanics, right? They talk about quantum fluctuations inside a proton, even though the proton is perfectly stationary. There's nothing actually time dependent inside the quantum state of the proton. And you're right that in a more careful reading, what they really mean is that if you were to measure over and over again, you would get different results depending on when you did the measurement. That would, you would actually see fluctuating measurement outcomes. But there is, you know, a sort of intermediate case.

Like think of the decay of a particle like we were talking about before, the decay of a neutron, into a proton, electron, antineutrino. What's really going on is that the neutron has a, a quantum state, a wave function, and that quantum state is in a superposition of having decayed and having not decayed. So it's in a superposition of proton plus electron plus antineutrino and neutron, right? And over time, the neutron part of the superposition decreases, and the proton, electron, antineutrino part of the superposition increases.

You can, that's all true. That's all just the Schroeder equation. And then you can add to that the statement that if you were to observe it, you would only see one or the other. You don't see the superposition. That's because you get definite measurement outcomes in quantum mechanics. So you would either see the neutron and it looks like it hasn't decayed yet, or you see that it has decayed and you see the proton electron antinutrino. It's all exactly the same story for baby universes out of decider space. There's a wave function, really, and we didn't talk about that in our paper that we wrote, or I didn't talk about it that much in,

from eternity to here because it doesn't add that much to the discussion, honestly. But really, there is a wave function that is a superposition of one big universe and some number of tiny baby universes, zero, infinity, who knows. And you can talk about the evolution of that wave function over time. But if you were to observe it, which you could do. If you're a person inside one of those universes, you would say, where am I? What does it look like? Oh, I'm in a baby universe. Okay, I have now collapsed the wave function from my point of view.

I am on a branch where that baby universe is real. So you don't need an external observer or environment, just like when we look at the cosmic microwave background. We see tiny fluctuations in temperature from point to point. That is because there was a quantum state. If you believe in inflation, that's because there was a quantum state that was measured effectively by decoherence. Even though there weren't any measurers around,

there aren't any observers around. The whole wonderful thing about Everett is that you don't need an external observer or environment. The wave function of a system can just branch all by itself. And that's what happens in our early universe, leaving us with a particular pattern of temperature and density fluctuations. And that's what you see in the CMB, that baby universe is from decider space work in exactly the same way. Red Antinov says, near the end of your discussion with Rachel Powell, you briefly touched on the long-term future for humanity and life on Earth. As the only species on the planet capable of recognizing that life on Earth is likely to become extinct in a few billion years,

do you feel we have a moral obligation to consider ways of ensuring its survival? For example, either by seeding potentially habitable satellites of the gas giants, or even a scattershot of life's ingredients to systems with potentially habitable planets. If yes, would you deem it a type of categorical imperative that any alien civilization would also answer in the affirmative? Well, I'll be honest, I don't like any of this talk of moral obligations and moral imperatives. I think that that's sort of a dangerous way to think about morality. I think that you can make arguments that certain actions are more or less moral than other actions. But once you talk about imperatives and obligations, you run into trouble because you're probably not thinking carefully about possible exceptions and things like that.

So I think the question is, you know, would it be morally right to scatter life around the solar system or the galaxy or whatever? So I can see a case for doing that. You know, by the way, none of this is very short term. This is a worry that we have right now. We don't have the technology to do that. So this is pure thought experiment speculation. But the argument for doing that is just like you say, you know, if we're the only living creatures around, it might be nice to give other living. creatures in the future an opportunity to live and grow and evolve and whatever, and so maybe

we can see the galaxy. On the other hand, maybe there is life elsewhere in the galaxy, and if we start sending our life out there, maybe we will compete with that life, and maybe even we will win. Maybe our cells that we use to spread life to other planets will destroy the life that's already there, or trying to find its foothold or something like that. I have no idea what the answer to that is. So I don't see any imperative to start, you know, scattering our seed around the galaxy. You know, 10 to the 100 billion, 10 to the 100 years from now, none of us is going to be around one way or the other. So these are all matters of degree, not matters of absolute difference. Steve Bonner says, given that neutrinos have mass, will they become

slower and slower as we approach the likely heat death of the universe and approach zero velocity? Well, I mean, I hope you know that when you ask a question like this, you have to be a little bit careful. Velocity with respect to what? In relativity, there's no such thing as just having a velocity, right? You have a velocity in a reference frame or with respect to some other object. So maybe what you mean is that, in fact, there's a lot of neutrinos in the universe. There's a neutrino background, just like there is a photon background in the cosmic microwave background. and neutrinos in the neutrino background have a relative velocity

that you can actually measure in terms of their effective temperature, right? A high temperature means that they're moving relatively fast, a low temperature means they're moving very slowly, and temperature you can just measure. You can measure the temperature of the neutrino background in its rest frame. That's a well-defined number, and that number is going down. And that number goes down in exactly the same way, almost exactly the same way, as the temperature goes down for the photon.

in the cosmic microwave background. They cool off. As the universe gets bigger, the wavelengths get stretched. Even though neutrinos are matter particles, not radiation particles, their momentum decreases. Their relative momentum with respect to each other decreases as the universe expands so their temperature goes down. So in that well-defined sense, yes, neutrinos do indeed move slower and slower with respect to the rest of the neutrino background, and they will eventually approach zero velocity. Stephen Pilling says, I was rereading from eternity to hear. I love it that people are rereading from eternity to here.

There's a lot of good stuff in that book. I recommend that people read it. And I was pondering on the asymmetry of past and future. Just subjectively, it seems that the greatest asymmetry is that a past event has a potentially lasting effect on me, whereas a future event does not. Thus, if I am hit by a car yesterday, I have a broken leg today, but if I am hit next week, I have no ill effects today. This does not seem intuitively completely attributable to entropy increase, although it is clear that without entropy increasing, I would not exist at all to be injured.

Do you have any thoughts on this maybe? I guess my intuition is letting me down again. Well, I think this is an important question that people write papers about, write books about. David Albert and David Wallace, other people that we've had here on the podcast, Carlo Rovelli, have written technical papers trying to connect the thermodynamic arrow of time, the increase of entropy, to the causal asymmetry of time, the fact that causes precede effects. I have something that I've given talks on. I've never quite written it up into a paper, but I have hopes to do that. The general consensus is indeed that the reason why causes precede effects rather than come after them is because entropy is increasing. The details are tricky and remain to be worked out, and people don't agree on where to put

the emphasis here and there. But let me just clarify a little bit about. the example you use. So you say, if I'm hit by a car yesterday, I have a broken leg today, if I'm hit next week, I have no ill effects today. That's true, but it's not quite the way to phrase the question. If you have a broken, if you hit by a car yesterday, you have a broken leg today, and tomorrow you might have a cast on your leg, right? I think the real question is to say, if I know I will have a cast on my leg tomorrow, does that mean that I have a broken like today. So actually run the causal arrow backward with exactly the correct setting of events and just ask, was it necessary? Did it follow? Necessary might be a little bit too strong, but you know what I mean.

Under the right thought experiment parameters, if you're hit by a car yesterday, you have a broken light today, you have a cast on tomorrow. If I don't know that someone has a broken light today and was hit by a car yesterday, but I do know they have a cast on their leg tomorrow. Can I infer that they broke their leg two days ago or whatever? And the answer is no. Like if you go through, well, the answer, it depends. Let's be super duper careful about it because it's a super duper tricky thing. If all you knew were the laws of physics and you say, okay, someone has a cast on their leg

at a certain day, does that imply that two days prior to that they had a car accident or or some number of days. The details don't matter. Okay. The basic idea is what matters here. No, just because you have a cast on your leg doesn't mean that you were necessarily hit by a car, right? You could have just put a cast on your leg or something like that. But then you might say, well, okay, but why would I do that? Why would I just put a cast on my leg if I weren't hit by a car? And then you're asking a slightly different question because you were just saying, if all I know is laws of physics and I have a certain boundary condition,

at this time I have a cast on my leg, what can I infer about the past? The answer is very little. But then you can say, but I know more than that. Like I know the fact that there's certain reasons why things happen. Those reasons why things happen are ultimately because entropy is increasing. So the sort of shortest but not especially transparent way of saying it is in the space of all possible evolutions of the universe that include you on a certain day having a cast on your life. leg. Most of those do not involve you having a car accident or some other broken leg before that in

time. But if you have not only everything we know about the universe and the fact that you have a cast on your leg a certain day and the low entropy big bang, then you can infer that of all the trajectories of the universe that connect the low entropy big bang to you on a certain day having a cast on your leg, most of them involve you breaking your legament, or whatever it is that maybe you just sprained a ligament, who knows, but it involved you needing to have a cast on your leg. So that's the kind of reasoning that goes into understanding why causes precede effects. You need to say, well, what would the implications be of some macroscopic information about the universe at one moment in time? What does it imply about moving forward and backward in time?

how does it change when you conditionalize on saying there is a boundary condition at one end of time, which is very, very low entropy near what we call the Big Bang? And the project is to establish that putting all those ingredients in gives you a relationship that says the things that we call causes precede the things that we call effects. And that's what people are trying to do. I think it's true. I just don't know if we have completely settled why it's true yet. Okay, I'm going to group a bunch of questions together. I think this is Quantum Magazine's fault that I'm getting a bunch of questions on the same thing here.

Tyler Whitmer says, I read an article in Quanta about Voichek-Zurek's new book, Decoherence and Quantum Darwinism, and his ideas about bridging the quantum and the classical via decoherence. Made me curious to get your take, do you have one? David says, I'm starting to read Voichek-Zurek's recent book on quantum Darwinism. It may take a while.

I understand decoherence as a process that makes the quantum system appear classical, but quantum Darwinism is new to me. Is this somehow trying to explain why a particular classical appearing world gets singled out? Some framing would be useful. Is this just many worlds are an enhancement or opposed to it? And finally, Elif Erdiken says, I recently read a quantum article about X interpretation of quantum mechanics, which claims to rely only on the Schrodinger equation without adding wave function collapse.

Zerx framework also invokes entanglement and decoherence. What I'm confused about is this. When I hear just follow the math, along with entanglement and decoherence, that sounds very much like the root that leads to many worlds. Yet Zurich's interpretation is not presented as many worlds. If both approaches rely on unitary shorteninger revolution plus decoherence, where exactly do they diverge? What assumption or step is different?

So let's go backwards. Voichek-Zurek is a super productive physicist, a very famous guy, as his, to me, his most important work has been done in the foundations of quantum mechanics, in understanding decoherence, quantum computing, things like that. He's a pioneer of many, many ideas that we now take for granted in understanding how decoherence branches the wave function of the universe. Now, for whatever reason, I've met Vojcheck, I think that's how you pronounce his name, Voj-T-V-E-C-E-C-H. It's spelled W-O-J-C-I-E-C-H. And I've asked about this and we've emailed back and forth

and he is for some reason reluctant to say that what he's doing is many worlds. I agree with Elith here. I think that it's just many worlds. Like you don't have any other equations. You have wave functions. They evolve with time.

Vojchek wants to place some interpretation on it where only one world is real at a time without changing any of the equations in any way, which I think is cheating a little bit. But, you know, it's okay. Like, at the technical level of the equations, what he's done is super-duper useful, I would say.

Chip Siemens and I leaned on some of his equations when we did our paper deriving the Bourne Rule in many worlds, even though Vojek wouldn't say that it is many worlds. I think that it is. And so I think that's kind of useful. Now, the other questions have to do with this specific idea of quantum Darwinism. And I think that, yeah, if you haven't heard about quantum Darwinism, it's one of those things, I feel very, very sympathetic to Vojtzek because he is realizing that there was a problem that people never talk about and he is solving it. And I know from experience that in physics, no one cares when you solve a problem that people didn't think they had already.

I guess that's not exactly true. Like there's different ways that people can get excited about things. inflationary cosmology is an example of a solution to a problem that a couple of people knew about. Jim Peoples and Bob Dickie wrote papers about the flatness and the horizon problems and inflation solved them. But most of the working cosmologists out there paid no attention whatsoever to these problems until inflation came around as a solution. And then they started talking as if these problems are super duper important. So quantum Darwinism is, as I understand it,

which is like a medium level. I wouldn't say a super high level, but nor a super low level. It's an answer to the question, okay, I have observed some quantum system. I am in a branch of the wave function. I see that the spin is up, okay? But you, you're another scientist here in the same laboratory as me.

You also observe the quantum system. You're going to branch, right? There's going to be a version of you that sees it up, a version of you that sees it down. when I talk to you, are we going to say the same thing? Are we living on the same branch? Like, the equations that say that I live on a branch are pretty easy. The equations that say that you live on a branch are pretty easy.

The equations that say we live on the same branch are a little bit trickier. And that's what quantum Darwinism addresses in some sense. It makes the case. And again, it's all true. I think it's all good and useful. It just no one is worried about this traditionally. it makes the case that in a typical decoherence event, not only is there information about the measurement outcome encoded in the environment,

but it's encoded redundantly in the environment. This is like a fancy way of saying that when you literally look at the outcome of your spin measurement, there's a dot on a screen that says the spin has gone up or the spin has gone down. Okay. the light from that dot moves out in all directions from the dot. So there's a bit of the environment where the light is moving to the left and moving to the right, moving up and moving down. And all of these are sort of sending out the same signal for what the result was.

And so decoherence isn't just a single point-like fact. It's something that spreads out through space in a very definite way. And that is what leads us to say that we live on consistent, coherent branches of the wave function. And the Darwinism thing is that he has to make arguments that it is this kind of structure that survives, right? It's a survival of the fittest kind of thing. I don't love quantum Darwinism as a phrase because there's not dissent with modification, which to me is the essence of Darwinism. It's just the survival of the fittest part, which is not really the essence of Darwinism, but that's okay. That's the phraseology that is used.

So he's not quite explaining why a particular classical appearing world gets singled out. He's trying to explain why you and I all agree on the classical appearing world that we're in. And that's a very important thing to establish. Johann Yartelius says, so I gather that space is big. But what is it? I'm reading your excellent book, The Biggest Ideas in the Universe, Space Time, and Motion, and I've gotten hung up on the is-ness of space. I understand that on the one hand, it is distance between things, but it is also true that this distance can warp and stretch.

What is it that does the stretching? What properties, if any, can we attribute to space itself? I realize this is a naive question, but it's been bugging me for a while. That's okay. It can bug you. It bugged Isaac Newton, too. And Gottfried Wilhelm Leibniz. I tell this story in that book about the substantivalism versus relationalism debate, which is to say,

is space a substance, or is it just a convenient summary of a bunch of relations? Like, does there actually exist space in between me and something one meter away? way? Or are there just a bunch of a table of numbers saying this object and that object have one meter in between them, with nothing actually doing the existing in between them? And, you know, like many philosophical questions, this one has a slightly ambiguous answer, I think. Of course, maybe we don't even know the once and for all final answer. Maybe that depends on some future theory that we don't yet have. But according to our best modern physics, I would say that we don't know the answer to this question. In classical general relativity, you more or less come down

quite firmly on the substantivalism part of the debate. Space and space time, for that matter, are things. They have properties. They have a metric. You can measure it, right? You might not be able to see it space. You don't see space with your eyes, but you can definitely feel it. You can definitely objectively measure some of its properties, so it just makes sense to think of it as an existing thing. What it is is a manifold with a metric, with a curvature that interacts with the matter and radiation in space time in certain definite ways, given certain definite equations. But of course, we know the general relativity is not the final answer, right? There might be quantum gravity. There might even be a better classical

understanding or something like that. In Hilbert space fundamentalism, in Mad Dog Everettianism, where you just start with a single vector in Hilbert space, and you find space as something that is emergent from that vector in Hilbert space, space is not a thing. It's not a fundamental thing. It's an emergent thing, but it's not something that is built into the most fundamental way of talking about reality. It is something that emerges from the relational properties of different parts of the wave function.

So I think that the short answer is, you know, we don't know what space is once and for all. We don't even know if it's at a more fundamental level, something more substantive or more relational. But hopefully that will become clear. I think this is something, this is a case where the philosophizing isn't especially useful until you understand the physics better,

better than we do right now. Anyway, Pete Newton says, should we restrict what AI is used for before too many people lose their livelihoods? Well, I don't know whether we should or not. I think it's essentially impossible to do that. I don't think that's a feasible thing to do. Like, you're just not going to be able to say,

don't use AI for this thing that human beings have a job doing right now. I don't think that's a practical way of getting things done. I think that it would be like saying, you know, don't let people drive cars because we need horse and buggy carriages or whatever. It's just not going to work. People are going to do it. And, you know, countries that try to do that are going to fall behind. the countries that don't do it.

I think when you have technological transformations that disrupt people's lives, what you need is a really sustained effort to give people soft landings, to give people other things to do, to provide ways to retrain and find a new place in the world. I really have no doubt that there will be plenty of things for people to do, no matter how effective AI becomes. But the journey to getting them there might be very painful. and very tricky, and as a society, that's what we have to worry about.

I think there's plenty of things that AI should be regulated about, but just saying you can't use it to do something that might lose somebody their livelihoods doesn't seem very practical to me. Christoph Radomsky says, you're an experienced podcaster, but I wonder if you've ever had a discussion that went so bad, either due to technical issues or argument with a guest or simply guest turned out not to be a good choice for the podcast, that you decided not to publish it. I certainly have never had an argument with a guest or disagreement with a guest that's so bad that I've never published a podcast.

I can't imagine doing that, both because I try to pick my guests a little bit carefully. Sometimes I do better than others, but I try. And, you know, I kind of owe it to the guest once they've spent an hour and a half of their life being on the podcast that I should at least publish it. obviously some episodes are more successful than others, but that's definitely going to be what you expect going in. Technical issues do happen. Sometimes the audio quality just isn't good. And of course, it's my job in the moment to be listening to the podcast and judge how good the audio quality is. But sometimes it's on the boundary, right?

It's never been like so, well, actually it has. a couple of times, only like, let's say, three times, maybe two or three times in the however many episodes we've had of Minescape, has the audio quality been so bad that it was unsalvageable. And in all of those cases, I just re-recorded the episode. And the person who I was talking to was nice enough to do that. Sometimes the audio quality is pretty bad, but you can touch it up. You can touch it up either in your DAW, your digital audio workstation. which is just a program you use here on the computer. Or there's even nowadays there's little AI programs that will, you know, lower the background noise and things like that.

They're not perfect. They're, you know, if you, they give you a slider. Do you want to like 100% processing or 50% processing? If you put it at 100%, it removes the noise, but it removes a lot of the signal also. So you have to be a little bit careful about that. So I don't usually use that, but sometimes the quality is not so good and, you know, that's the best you can do. So generally no. Like your job as a podcaster is to try hard to make sure that the quality is pretty good the first time that you do it and I've become pretty good at that.

Brett Sloggs says, when it comes to political and electoral strategy for advancing the cause of liberalism, there's a debate between popularists and their critics. The popularists emphasize avoiding political messaging and campaign positions that are particularly unpopular, according to polling, optimizing around existing voter preferences and focusing on incremental persuasion of voters in the middle. The critics generally assume that voters are more persuadable than that and emphasize leading voters on moral issues, shaping voter preferences, accepting short-term cause for long-term transformational change, and not compromising on certain unpopular positions, such as trans rights. Where do you stand in this debate?

I think I stand pretty strongly on the side of the critics in this debate. So basically what's going on is that some people think, some people are entranced by the median voter theorem. Okay, the median voter theorem says, imagine that all voters, just for simplification purposes, exist on a single one-dimensional continuum of views from like most conservative to most liberal. And whoever gets the most votes is going to win, then the median voter theorem says, oh, and an extra assumption is the voters are going to vote for the person who is most close to their views. to their policy preferences or whatever. Then you can prove the median voter theorem that says that you should try to be as close to possible. You want to be as close to the median of the voter distribution as you can so that you get more voters close to you than voters close to your opponent.

Now, anyone who has done any looking at actual elections knows that most elections are not between two people who are basically right in the middle, moderates, right? That's not actually the empirical result that you would get, which is what you would predict if the median voter theorem were true. And the so-called popularists kind of act as if it's true. It is true as a mathematical theorem, but act as if it matters, if it's relevant to real-world politics. The popularists say, look, you might be liberal yourself, or for that matter, it works just as well the other way. You might be super conservative yourself, but if you want to win an election and get your favorite policies enacted, you better appeal to the voters in the

middle because those are the ones who are going to have a tough choice. If there's one fairly conservative candidate and one fairly liberal one, the far liberal people and the far conservative candidates have no choice but to vote for the one closer to them. It's the ones in the middle that you're looking for. Okay? That's the popularist point of view. It's clearly false empirically, like it's hilariously false. And the reason why it's false is because the assumptions of the median voter theorem just aren't true in real world elections. Among the many ways in which they're not true is that individual people are not accurately summarized by living on a one-dimensional spectrum from liberal to conservative. And for that matter, they don't

vote for the person who is closest to their policy preferences. Voters are not coherent in their policy preferences or in their preferences for candidates. Think about Joe Rogan, right, who, you know, whatever his virtues and flaws is closer to an every person, closer to an average person in politics than many of the political pundits out there. The two candidates for president he ever came out in favor of were Bernie Sanders and Donald Trump. How in the world, in the set of all possible political candidates, presidential candidates, are those your two favorites, one of the most most conservative and one of the most liberal, right? What about the median voter theorem? What matters to people like him is not policy preferences. It's as simple as that. What matters to most

voters is not policy preferences. I know that makes people sad, but it's true. And so I think the truth matters in these situations. Lots of evidence shows that voters are much more impressed by people who they think are sincere in their beliefs, and willing to fight for their beliefs, then people who just have the same beliefs that they do. And so this is just a mistake that especially the Democrats make more than the Republicans do. They come across as not being sincere in their beliefs because they're trying to triangulate their way into the median voter. And it doesn't work. They come across as insincere.

So I think it's like this is not even a tough decision. I mean, maybe there are times when you want to moderate your own policy preferences in order to get elected. Like, that's absolutely possible, okay? But the idea that mostly you should take polls of what people think and then say that those are what your preferences are, that's just obviously false. It's just obviously empirically wrong. But I think it's not exactly how Brett is characterizing the critic's view here. He says the critics generally assume that voters are more persuadable than that and emphasize that politician should lead voters on moral issues shaping voter preferences.

I think that as part of it, right, that you shouldn't just listen to what the voters say and then say, oh, yes, I agree with that. You should try to make the voters appreciate that there is a better way of doing things and take a leadership position. But I don't even think that's the main thing. I think that's true. But I think that just having the courage of your convictions and just saying, you know, I'm going to get things done and I'm going to like take a stand here is something that people really respect, almost independent of what your policy positions are. I think that you really, if you really want to win elections, rather than being a so-called realist about policy positions, you should be a realist about the fact that most voters don't vote on the basis of policy positions. It's a little bit more complicated than that.

Okay, Thomas Prunty says, is it possible to fully describe the curvature in general relativity as extradictive? a intrinsic curvature in some higher dimensional flat space time. If so, would we need additional dimensions of space in time or just space? So it's possible. So the issue here is general relativity describes four-dimensional space time, and it's curved. And when we first, back in the 1800s, when we first started thinking about curved geometries in general, it's certainly easiest to imagine a curved geometry as being embedded, being a subset of a larger, higher-dimensional, flat geometry. In fact, it's really interesting to me that the first systematic sort of axioms for non-Euclidean geometry

were for negatively curved geometry by people like Loboshefsky and Boliath, et cetera. and that's weird because negatively curved geometry is like harder to visualize than positively curved geometry. A positively curved two-dimensional surface is just a sphere. It's easy to visualize a two-dimensional sphere in three-dimensional space. A two-dimensional hyperboloid, which would be negatively curved. People say like it's a saddle shape or it's a pringle or whatever potato chip, but you can't actually visualize the full infinitely big, uniformly curved negative

curved surface. That's not something you can visualize because it can't be embedded in three-dimensional Euclidean space. So you might have thought that because the two-dimensional sphere with positive curvature is easier to visualize than the two-dimensional hyperbolic plane, mathematicians would have started doing non-Euclidean geometry with positive curvature first, but they didn't. I think the reason why is because you can embed the two-dimensional sphere in three-dimensional Euclidean space, but you can't embed the negatively curved surface because people probably thought that you didn't need something called positively curved geometry.

You could always just start with good old Euclidean geometry in one more dimension and then think about curved surfaces embedded inside. This is what is called the extrinsic curvature of the surface. It's extrinsically curved because it gets its curvature from how it's embedded in a bigger space. That's what a sphere does. in good old Euclidean space.

When you go to the negatively curved surface, you can't do that, right? You can describe a negatively curved surface of constant symmetry, constant curvature, but not as something embedded in three-dimensional Euclidean space. So you had to sort of invent new rules for it, and that's what they actually did. So that distinction between intrinsic curvature, the curvature that a space or a manifold has of and by itself, without being embedded in some higher dimensional space had to be invented in the 1800s. And this is what Riemann really sort of put a lot of,

gave a lot of extra technological boost to. Remondian geometry was a very general way of talking about intrinsic curvature. So you didn't need to be uniform curvature, like a sphere or hyperboloid. You could have any amount of curvature anywhere changing all around. That's what Riemann taught us how to do. They needed to develop a lot more technology afterward, but people did that.

And then when general activity comes along and says, okay, we have a four-dimensional space time, and it's curved. So what kind of curvature is it? Well, in Einstein's way of talking about it, it's intrinsic curvature. There is no larger space in which the universe is embedded, okay? Now, Thomas's question is, is it possible to describe the curvature of a four-dimensional space time in terms of extrinsic curvature in some larger flat space time, yes. The answer is yes.

There are theorems that say, well, I don't know the theorems for space times. For just spatial manifolds, there's like the Whitney embedding theorem and things like that that say that if you have some geometry that is arbitrarily curved, given a sufficiently large flat space, you can always embed it. And I'm pretty sure that the same thing is true for space times. I'm thinking to myself now. Maybe that's not true because maybe there's global considerations when you have close time like curves or something like that. So maybe I need to actually back that up.

But let's say the answer is yes. In general, I think you would need more dimensions of both space and time. But the whole point of general relativity is you don't need that, right? That would serve no purpose whatsoever except maybe like making you feel better because you could visualize it or something like. like that. More importantly, if you take seriously the sort of spirit of general relativity, according to which when you have gravitating sources of energy or mass or whatever, they affect all of the geometry of space time, you would imagine an experimental prediction from that kind of picture in which a gravitating source would have its gravitational field

leak out into the extra dimensions, the dimensions in which you're embedded. Something very much like this actually is supposed to happen when you have large extra dimensions of space. You could have large extra dimensions, but gravity leaks out into them. Maybe that explains why gravity is weak, right? But those large extra dimensions are still less than a tenth of a millimeter across. They're not that big. If you had infinitely large extra dimensions, you would expect that to be very, very noticeable experimentally. Gravity wouldn't look like an inverse square law.

It would look like something more complicated. So my impression is probably you can do it, although now that I'm, you know, in the process of giving the answer, I realize I don't know technically the results in space times rather than space. So I shouldn't be too confident about that. But I think you can do it. And there's no good reason to do it and a little bit of a reason not to do it. Okay, the last question for today's AMA, for this month's AMA, is actually not a question. I'll read it out loud. It's from the nine-tailed fox.

Not a question. I just wanted to say thank you for doing what you do. I have all your books and wait with anticipation each week for your podcast. I'm autistic and have a rather difficult time with handling my emotions when I'm feeling overwhelmed and unable to focus you are my go-to. The podcast helps drag me out of myself and view things in a bigger picture, not to mention expanding my knowledge of the universe. I never thought I would join Patreon, but after about three years of listening to your work so hard, at always trying to teach everyone no matter what their level of education, I just had time to find a way to support you more. Thanks again for the work you do and for your continuation of the podcast after so many years. So again, there's no question here. And usually I will say that I like getting comments like this in the Patreon comment sections, but I don't choose them as questions. I don't read them out loud. I thought I would read this particular one for two reasons.

Number one, to give a shout out and support to Nine-Tale Fox, you know, a difficult time handling your motions, being autistic. These are things you got to struggle with. And I support you in that struggle. It is not going to be easy sometimes. And other times it will be easy and it'll go well for you. So I hope things are going well for you. But the other is to just send out my appreciation for the kind words here. And, you know, I don't usually read comments like this, just the compliments.

but I do read them and I care about them. I don't read them out loud because you don't want to hear other people complimenting me and I don't want to sound like I'm just complimenting myself. But I do appreciate it. And in fact, furthermore, I want to say probably this is a general thing. Like you will often hear writers or creators of different artistic things as well as scientific things say it's nice to sometimes just get a random email or comment or whatever saying, you know, hey, I really like your stuff.

It's really good. You know, thanks for doing it. If someone is very busy, they might not reply to you or anything. Don't do it because you expect a reply. But, you know, it can be a nice way to support the people, especially the not famous ones, the people who are not doing that well, the struggling creators or young scientists or whatever, who are trying to bring things to a wider audience, they can really be lifted up by the occasional compliment, kind word recognition that what they're doing is.

is actually reaching people in a nice way. So I would like to send my thanks to Nine-Tailed Fox and to everyone else who ever says those nice words. They are part of what keeps me going here at Minescape. And I encourage you to say similar nice words to other people who might be doing good things that you enjoy on the internet or elsewhere. So that's the end of this month's AMA. Thanks, as usual, everyone, for supporting the podcast, for sending in your page. Patreon dollars and just for participating in the discussion in whatever way you choose to do that. I'll talk to you next time.