AMA | February 2024

0:00:00.2 Sean Carroll: Hello, everyone, welcome to the February 2024 Ask Me Anything edition of The Mindscape Podcast. I'm your host, Sean Carroll. So things have been a little bit hectic here at Mindscape world international headquarters. They're not supposed to be hectic. It's the beginning of the new year. I'm not even teaching this semester. I doubled up on teaching last semester so I could have a breather and work on my next book. But it's been hard, there's a lot of things going on, and as a result, this month's AMA is a wee bit later than usual. It's not a big deal. The questions are eternal questions; it doesn't matter what week they appear on.

0:00:36.1 SC: And it might also turn out to be the case that it's a little bit shorter than the average AMA. It's still pretty darn long by most measures, I gotta say. So, I was thinking, while going through the questions for the AMAs... By the way, the AMAs are sponsored by Patreon supporters. So it's the Patreon supporters who get to ask the questions, and once in the lifetime of a Patreon supporter, you get to ask a priority question that I will do my best to answer in good faith. You, too, if you're listening to this and are not a Patreon supporter, could be a Patreon supporter. Just think how good that would make you feel. You can go to patreon.com/ seanmcarroll, chip in a dollar per podcast. If you think that the hour that you spend listening to Mindscape is worth a dollar, then going to Patreon supporting is a nice thing to do.

0:01:24.6 SC: Anyway, sadly, the number of questions I get, well over 200 questions per month, and nowhere near what I could actually do, even in a two or three or four-hour AMA, so I have to pick and choose. And I was pretty brutal this time, because I knew I wouldn't have a lot time to do it, so I was cutting out a lot of good questions. And as I always say, that breaks my heart to have to do that.

0:01:47.3 SC: In particular, I noticed that this month, there were just a lot of questions about quantum mechanics and Many-Worlds and things like that, and I do try to keep the AMAs varied, right? I don't want five questions in a row about the same topic, even if they're really good questions. So, one possibility, let me just throw it out there, if the community, that is to say, the Patreon Community, the folks who are paying for this, if you all would like it to have a themed AMA. So rather than literally just bouncing back and forth between every topic, which is my preference, that's what I actually do, but if you, since you're the paying customers, would like to have just, "This month's AMA is nothing but quantum mechanics," or "This month's AMA is nothing but cosmology," or whatever it is, I would absolutely consider doing something like that. So I'm throwing this out to the Patreon supporters out there. Think about that, chime in in the comments, and we'll consider that.

0:02:40.4 SC: The AMAs are different than the regular episodes. The regular episodes, I take suggestions, I'm very happy to get suggestions, but basically, my criterion is I wanna do what keeps me interested. Whereas the AMAs are my attempt to give back to thank you, the Patreon supporters, for supporting the podcast. So I do want to fine-tune the AMA strategy to be whatever the Patreon supporters really wanted it to be. So feel free to chime in about that, and we'll try to work it out. It's a process. You have to trust the process, as someone famous one said. And with that, let's go.

[music]

0:03:33.5 SC: Linus Melberg says, "If we are living in a simulation, what if the exact same simulation were executed again, so the exact same thing is simulated twice. Would we be alive twice?"

0:03:46.8 SC: Well, the short answer is no. The best answer is no to this. There would be two people who are alive, who are completely identical to each other. This is exactly like the Many-Worlds interpretation of quantum mechanics where you have multiple copies of people. If these people have not actually interacted with some particular quantum measurement that branch the universe into multiple copies, then they are completely identical with each other, but they're not the same person. The same thing happens if the universe repeats in time, as Nietzsche would have imagined, or it's the same thing that happens if you live in a simulation, you run the simulation more than once.

0:04:24.9 SC: There's no causal connection. There's no memory from one to the other. There's no influence it passes in either direction. There's no communication or interaction of any form. So these are two separate people, even if they are exactly the same people. So, two people are alive that are identical to each other, but it's not one person being alive twice.

0:04:48.6 SC: Tucker Hyatt asks a priority question, priority question, meaning, once per life, you get to ask a question that I'm gonna try my best to answer. "According to what is sometimes called David Hume's problem of induction, we cannot explain the regularity and persistence of a laws of physics. Are you ever or always surprised by, and even grateful for, the apparent continuity of our existence?"

0:05:10.0 SC: I don't know whether or not that's the kind of thing it makes any sense to be surprised by. Which, let me be super duper clear, it might be okay to be surprised by that. There are many, many more ways for the universe to be irregular than for it to be regular. So, it's okay to have a vague intuition that the existence of regularities, which we call the laws of physics, in the universe is something special and unusual. But we certainly don't know. We have literally no right to expect to have any intuition whatsoever about whether or not the universe is, in some sense, more likely to be randomly scattered in its events versus orderly in events. Where would that intuition come from? Where would this measure on the space of possible universes come from? This is always a problem for... Well, on the abstract side, for vague theories of all possible worlds, like philosopher David Lewis, or for that matter, Max Tegmark, would have us believe in, or for scientists in the real world to bring it down to earth and make it much more tangible who want to understand the world by proposing theories that may or may not be right about the world.

0:06:27.7 SC: We certainly, in our everyday scientific theorizing, treat certain possible sets of rules as more likely than others. They're simpler or more beautiful or whatever. What right do we have to do that? We get some right, maybe, from the fact that it has worked for us scientifically for a long time, but it's not like there's some clear, crisp philosophical reason why it had to be that way, some justification for the values that we choose to put on different scientific hypotheses. That's why I try to actually sort of be honest about this. When we're talking about theories where there's multiple competitors to explain a certain phenomenon and we don't yet know which one is right, everyone is always gonna bring in our intuitions our our values, in some way. And we have to be humble about whether or not those values are gonna be really all that determinant when it comes down to actually doing the experiments.

0:07:24.2 SC: So anyway, that's a long-winded way, this sort of sparks a whole bunch of other interesting issues out there, Tucker's question, but I don't see why it should be surprising that the universe is more regular. Or let me put it this way: If there hadn't been any regularities in the universe, I want to say I would not be surprised by that, but I can't even really say that, 'cause I wouldn't be here, right? There's a very real anthropic thing going on where there's no observers around to talk about the universe if there's literally no regularities that we'd recognize as the laws of physics. There's no persistence of information for a moment to moment, so you can't learn and reason about anything. But that might seem a little cheap. I guess what I want to say is, in the space of all possible universes, I would be willing to admit that more of them don't have regularities than have regularities. But maybe there's something I don't know about the space of all possible universes. I think that that's just something that right now is way above all of our pay grades.

0:08:29.3 SC: Aaron Perrin says, "The more I learn from your podcast, the more I wonder if the universe is an optimization similar to a Markov decision process, that is particles are selected from the wave function in a way that optimizes some currently unknown natural policy, e.g., complexity. We see optimization processes in other places in nature; evolution is one that comes to mind. As an amateur, how would I take an intuition like this beyond the crackpot stage? I figure I first need to do a literature review, that I need to figure out how to test it and falsify it. What else?"

0:09:00.9 SC: You know, I love this question. I especially love the way that you asked it, Aaron, because you're sort of on the right track, even though the answer I'm gonna give is, no, you are completely wrong about what it is you have to do next. Let me first state that without knowing any details, of course, the idea that the universe is an optimization is completely plausible, except that maybe it's trivially true. Whatever a system does, if there's any regularity, as we were just talking about, if there's any laws of behavior, laws of dynamics governing the behavior of any system at all, you can always cast those laws as a minimization or maximization problem. That's just a mathematical fact. Now, maybe it's interesting to decide what it is you're optimizing, but the fact that a certain law of physics can be stated as an optimization problem is almost content-free. Okay? Let me just mention that.

0:09:56.7 SC: But more to the substance of your question, what does person who has an idea about the fundamentals of physics, but who is not a professional physicist embedded in the research discussion and so forth, what should they do to sort of raise, elevate the level of their ideas to a more respectable level? The first thing you have to do is not review the literature or figure out how to test it, anything like that; the first thing you have to do is learn to understand and comprehend the physics that we already know. If you haven't done that, no one is gonna listen to you, no matter how good your theory might be. It's like saying, "I've developed a new way of designing a car engine," someone says, "How is it different than the existing ways?", and you go, "Oh, I don't know, I have no idea what the existing ways of doing car engines are, but I'm sure mine is better." People are just not gonna listen to you for that. Even if in the unlikely event that you're right, they're not gonna waste their time, because probably you're wrong because you don't know what the possible worries are. You propose a theory and someone says, "Okay, is the Hamiltonian bounded below? Is a renormalizable, is it local, is there violations of causality?", and you're like, "I don't know. How would I know what any of those words mean?"

0:11:16.4 SC: So what you actually have to do is learn physics as it is basically learned by a typical first-year graduate student in physics. So you need to know the basics of classical mechanics, electricity and magnetism, statistical mechanics, quantum mechanics, probably quantum field theory for the particular thing that you're thinking about. And that can sound like a lot, because you also need to learn all the mathematical techniques that go along with these different subfields of physics. You probably don't need to learn general relativity, maybe not even quantum field theory, depends on exactly how deep you wanna go, but maybe you have to learn those two. It seems like a lot, but guess what? Literally hundreds and thousands of physics students do it all the time. So, if you really think that many, many physicists for generations have missed this great idea, and you've had it, then probably, you're at least as smart as the average of the thousands of physics students who are doing this all the time. And if you're really dedicated to developing this theory and making people take it seriously, then you should be dedicated enough to learn these subjects.

0:12:25.8 SC: I always encourage people to go to Gerard 'T Hooft's web page. 'T Hooft, of course, is a famous physicist, Nobel Prize winner, one of the geniuses of the quantum field theory era in theoretical physics, and he has, on his web page, a guide on how to be a good theoretical physicist. And what it is is an overly detailed list of everything you need to know, plus resources, textbooks, online notes and things like that. So, it's all there. You can find it. In fact, the only problem with 'T Hooft's version is that it's over complete. He's way too ambitious about what he thinks you have to learn. To be fair, he's saying if you really wanna be a great theoretical physicist, this is what you have to do, not if you need to just pass the minimum bar. But the point is you can find resources to learn all of these things, in different places, and that would be the first step you have to take. If you really want to take an idea that you have, which might very well be promising, and make it into a form that is understandable and perhaps interesting to professional theoretical physicists, you have to speak their language, you have to know what problems they're gonna have in the back of their minds. If you want them to respect your idea long enough to pay attention to it, then you have to respect their training well enough to catch up just a little bit.

0:13:50.5 SC: Cubit says, "In your subtle episode about immortality, you mentioned the possibility of baby universes, which come into existence a long time after the universe has reached its maximum entropy state, and which are formed due to random quantum fluctuations. I wonder if that kind of fluctuation is special, or if it occurs all the time in our universe just to a much lower degree. Due to the lost microscopic information, such fluctuations do not seem to be governed by a unitary time evolution. Is there even an idea what kinds of laws we are talking about, since the Schrodinger equation no longer seems to apply?"

0:14:22.2 SC: Well, you're sort of half on the right track, half not on the right track here. Yeah, they happen all the time. In the picture, for example, that was put forward by Jennifer Chen and myself, when we talked about the arrow of time in an eternal universe, we talked about fluctuations in a universe like ours and calculated the rate... I don't think we calculated it correctly, and honestly, I know more now than I do, then we could probably improve that calculation, but who cares? In some sense, the actual number is not crucially important. It's very, very, very, very unlikely that you will actually have a random fluctuation that will create a baby universe. It would look to you, on the outside, if you were standing there in the room and a baby universe was somehow created, it would not be invisible. What would it look like is a whole bunch of matter randomly focusing in on some point in space, making a little microscopic black hole, which then very quickly evaporated. And that's like an explosion; that's like a grenade going off in the room you're in. So you absolutely would notice it, okay? And yeah, it happens all the time. It's just so very unlikely that we don't need to worry about things like that, when we're talking about what to expect in our everyday lives or even in the entire history of the observable universe. It's probably never happened even once.

0:15:42.0 SC: Now, having said that, there's nothing about it that violates the Schrodinger equation, et cetera. It's exactly the usual picture of quantum fluctuations, or let's take quantum tunneling. When you have a nucleus that is unstable to alpha decay, that is to say, it's spitting out a helium nucleus and becoming a lighter nucleus itself, that can be thought as a tunneling process. To the observer, that looks sudden. Random. That's why random numbers seem to come into our predictions about quantum mechanics. But if you are an Everettian, everything is perfectly explicable in terms of the smooth evolution of the Schrodinger equation. There is a superposition of states that look like a single nucleus, plus states that look like two nuclei, one is a little alpha nucleus, and the other is the remnant, and the amplitudes, the relative amplitudes, change, that gradually, the one that is multiplying, the single nucleus, gradually fades away; the one that is multiplying all the possible ways to get two nuclei gradually grows and becomes more important.

0:16:47.6 SC: So, were you to observe the system at any moment, there'd be a certain probability that you would see it to decay and not decay. So it's all perfectly described by the Schrodinger equation up to that measurement event, which then, even an Everettian would say even that is completely well-described by the Schrodinger equation. So all of that is exactly the same story for baby universes. There's a universal wave function that is a superposition of just one universe plus a universe plus a baby universe. And for that matter, since baby universes cost zero energy, zero charge, et cetera, there's also super positions with two baby universes and three and four, an arbitrarily high number. And the real wave function of everything is a superposition of all those things, and the amplitudes in the superposition change with time, and there's apparent branching because of decoherence and all the usual words. So, Schrodinger equation still holds up perfectly well.

0:17:45.4 SC: I'm gonna group two questions together. One is from Ilya [0:17:49.4] ____, who says, "In classical general relativity, nothing special is felt by an observer free-falling into a black hole at the event horizon. But what do they actually see before and after the crossing? Would it be just darkness ahead due to the light falling to escape the black hole? After the crossing, would it also become dark to the left right of the observer?" And then Rob [0:18:09.4] ____ says, "What is your credence for there being a firewall at the event horizon of black holes?"

0:18:15.1 SC: These are clearly two very different questions, but they both involve the ultimate question: What do you see when you fall into a black hole? So, for the first one, it's easier. So, Ilya is very clear that we're talking about classical general relativity, which is not the real world but might be a good approximation to the real world. That's what we don't know. So it's not that mysterious, honestly. When you're falling into a black hole and you're very, very close to it, in front of you, that is to say, in the direction of the black hole, if you're falling face first, it's black. [chuckle] No light is coming out of it, it's a black hole. So don't over think it. Don't think that because you read all the stuff about photon rings or gravitational lensing or whatever, that somehow, you would see anything from the black hole. The black hole's health is just black. You might see a stray photon that just got really, really lucky. But for the most part, you see nothing at all.

0:19:06.1 SC: And likewise, when you look back and you look back at the rest of the universe, you see the rest of the universe. Now, you're gonna see it highly distorted, because as soon as you cross the event horizon, half of the rest of the universe will have sort of been gravitationally lensed out of your view. So, what you're seeing when you go toward the black hole is a black region, a circular black region, which is the black hole, that grows and grows and grows until you cross the event horizon, and it's more than half of what you see. And if you turn around, you see everything else in the universe that can be sort of warped... The light from everything else in the universe is warped by the black hole so that you can see it. So you see the outside world, and it's gonna take up less and less of an angle as you fall closer and closer into the black hole.

0:19:56.3 SC: Whereas what Rob is asking about is the firewall paradox. This is a particular idea that says that maybe classical general relativity is dramatically wrong at the event horizon. In classical general relativity, you don't feel anything when you see yourself crossing the horizon. There's no sign post there. Of course, you see the black hole growing, as we just said, but there's no pain or there's no wall you slam into or anything like that. Now, we talked about firewalls. Did we talk about firewalls? Did we ever have a podcast directly devoted to firewalls? Maybe we didn't, I don't know, but we mentioned them, and certainly in passing in different podcasts.

0:20:35.3 SC: The idea came from [0:20:39.1] ____, and they pointed out that if you wanted to have black holes evaporate, so if you want to get information out of black holes, as many theoretical physicists do, so not only do black holes evaporate, but they do so in a way that conserves information at the deep down quantum level, a lot of that information is contained not just in individual particles, photons, et cetera, coming out of the black hole, but the entanglement between those particles. Quantum information relies on entanglement between different kinds of particles. So, they, said if all of the information gets out, all of the radiation coming out of the black hole is going to have to be somehow entangled with all the other radiation that has ever come out of the black hole. In particular, the radiation that comes out early is gonna have to be entangled with the radiation that comes out late.

0:21:37.3 SC: Now, everyone knew that, that then people would nod along if you said that, "Yes, that makes perfect sense." But they point out something else, that this idea that there's nothing that you see or bump into at the event horizon can be translated if you just look in a very small region of space-time as saying that it looks like empty space at the event horizon. It looks like the vacuum, the Minkowski vacuum, if you have a big enough black hole that you're falling into. And in the Minkowski vacuum, those photons that will grow into Hawking radiation. Remember, the story of Hawking radiation is that there are particles going out, but there's also particles coming in that effectively have negative energy, from the point of view of the outside observer. And those sets of particles also have to be highly entangled with each other. There's a very specific entanglement structure that those ingoing and outgoing particles have to have in order for it to look like empty space, to look like the vacuum. The vacuum has a known entanglement structure, and there's a lot of entanglement there.

0:22:37.1 SC: So those outgoing photons have to be entangled with the ingoing photons, but they also have to be entangled with photons that are emitted at a completely other time. That is not allowed by the rules of quantum mechanics. You can be a little bit entangled with both things, but in these cases, you can show they have to be sort of maximally entangled with both things, and you can't be maximally entangled with both things at once. So, [0:23:04.7] ____ suggested that maybe there was not a vacuum state to the black hole horizon, but rather a wall of fire that would incinerate you when you hit it.

0:23:16.1 SC: Now, they were in part trying to be provocative by saying that, and they succeeded. They did a good job on the marketing, and most physicists don't believe that there really is a firewall there. In fact, when we talk to people like Netta Engelhardt recently, I guess it's not recently anymore, huh, it was a while ago, but Netta was inspired in her work on thinking about how to get information out of black holes by attempts to get rid of the firewall, to allow the radiation that gets out to be maximally entangled with other radiation without destroying the vacuum structure of the radiation inside. So that's a very promising way to get rid of those firewalls.

0:23:54.3 SC: I actually wrote a paper myself with some students and post-docs back in the day saying that in Everettian quantum mechanics, there is another way to get rid of the firewalls, because secretly, you've mixed up two different kinds of statements. One kind of statement is about the wave function of the universe. The wave function of the universe is the one that is supposed to conserve information. Any good Everettian any good knows that when you apparently make a measurement, when you have a spin and you pass it through a magnet and you see it spin up or spin down, the whole wave function of the universe is just solving the Schrodinger equation, but you see an apparent wave function collapse. That means that you see information apparently being lost. There's a difference between what the wave function of the universe does and what individual observers actually see.

0:24:42.9 SC: And we made the point, I think our paper was called Branches of the Wave Function... Branches of the Black Hole Wave Function Need not Contain Firewalls. So we made the point that the existence of a firewall is a statement about what observers see, because it's observers passing through the event horizon that will or will not see a wall of fire there. That's not a statement about the wave function of the universe. If you can take the wave function of the black hole and express it as a sum of things without firewalls, but in a way that all the radiation gets out, then you avoided the firewall paradox. And I think that's right, but of course, basically, we're just pointing out a logical loophole in the original [0:25:25.9] ____ argument. We're not actually saying that nature takes advantage of that loophole. That's a harder thing to do. But anyway, the point I hope I'm getting across is most people don't think the firewalls are there, but we're not 100% sure on what mechanism nature uses to actually get rid of them.

0:25:45.8 SC: Jim asks, "My question regards the exploration to unify gravity with other forces at a quantum level, find gravitons, et cetera. Why is it not satisfactory to simply consider gravity as a separate aspect of our reality instead?"

0:26:01.8 SC: Because they speak radically different languages, honestly. That's like saying, "Why can't we just add together the integers and the colors?" [chuckle] There are different kinds of things; they don't add together nicely. General relativity posits that the basic thing is space-time, four-dimensional or higher dimensional with some dimensions hidden, something like that. Quantum mechanics doesn't say that. Quantum mechanics says, the basic thing is a vector in Hilbert space. I know that when you were first taught quantum mechanics and you first come across the wave function of an electron, it looks like it lives in space-time because it's psi of x, or x and t, where x is position and t is time. So, at every point in space-time, the wave function of the electron has a value. But as soon as you have two electrons, that's no longer true. The wave function of two electrons is not a function of space-time; it's a function of all the possible places where two electrons could be, the configuration space of a two-electron state. And when you have quantum field theory, the wave function is a very, very different-looking thing.

0:27:07.9 SC: So, quantum mechanical wave functions don't live in space-time. General relativity says that space-time is where things live. They are, in principle, incompatible with each other. Now, I know people have tried, okay, people have tried to somehow make it work, but trying is gonna do dramatic violence to either general relativity or quantum mechanics in some way that is completely unclear to me. To be said another way, here's one way to try. You can just say that general relativity is the more correct thing, that there really is a space-time that really is there, and that the curvature part of Einstein's equation is just correct, it's just what Einstein said. It's the energy and momentum part that is fundamentally quantum mechanical. And for example, you could say that what you really mean by the energy is the expectation value of the energy, the average energy in some quantum wave function.

0:28:05.4 SC: The problem is, it's very easy to imagine situations where the quantum wave function is a superposition of two things very far apart from each other, not two things almost on top of each other. That's a very non-classical looking source, and it's completely unclear what to do. Naively, if you just take the expectation value, that would mean that if you had a superposition of a gravitational object here and a gravitational object there, you would feel the force of gravity to be the sum of both of them, and you'd be pulled in between them. Nobody expects that to actually happen. You'll be either pulled toward one or toward the other, because basically, you're collapsing the wave function of the universe by measuring them. But that only makes sense if the wave function of the universe includes the gravitational field. It doesn't make sense if the gravitational field is purely classical, so most people think that that's kind of a non-starter.

0:28:57.2 SC: Robert Greneze says, "We hear a lot about the simulation theory. My most common reaction is, if it is true, so what? It changes nothing about our existence if it is true or if it isn't. Can we assume that the simulators have built a system that cannot be hacked so we can never know or find the source? My question is, if it were true that even quantum physics would be made up. What does it then mean for Many-Worlds?"

0:29:22.4 SC: So I think you have some interesting issues here, Robert, but they didn't quite congeal into a question. Yes, if we live in a simulation, then all the laws of physics are made up. That is true, because the simulators made them up. Just like when you program Space Invaders, the programmers made up the rules of Space Invaders. That would be exactly the same for our universe. What does that mean for Many-Worlds? I don't know. If the rules that the programmers programmed are Many-Worlds, then it would mean that Many-Worlds is still the right way to describe the universe in which we live. If the universe programmed some other version of quantum mechanics in then it would mean something else.

0:30:00.4 SC: This is always what happens when you start talking about the simulation theory. It is so ill-defined what the rules are that basically anything goes, and you're at the end of the day, left without any special expectations that differ in any way, I think, from ordinary physics where we actually live in the base reality. So, let me put it this way. I can imagine a version of simulation theory where there are literally no departures from our expectations between single universe, we live in the base reality view, and the simulation view. That's possible. And if that's true, then who cares? I don't really see what the difference is. Or I can see people making an argument that I would expect something different. If simulation argument was true, then if we live in the base reality, and if that's true, all the things that I expect would be different aren't different.

0:30:54.3 SC: So in other words, the universe we see, if I'm trying to be honest about what I would expect it to look like if we lived in a simulation, is very different than what it would be like in the simulation view. So, to the extent that we are allowed to be good Bayesians and construct likelihood functions for what the universe should look like if the simulation argument were true, it doesn't look like that to me. So I don't spend a lot of time thinking about it.

0:31:20.6 SC: Tim Converse says, "I'm reading From Eternity to Here," good for you, Tim, that is a wonderful thing to spend your time doing, "and I'm doing my best to understand the arrow of time as a consequence of increasing entropy plus the past hypothesis. If we were to construct a sealed and insulated box of gas and let it come to equilibrium, would it be reasonable to say that our normally perceived arrow time does not apply within that box? That conclusion would surprise me, as I would have expected the arrow time to apply throughout our observable universe rather than being dependent on what's going on in particular local regions."

0:31:53.2 SC: Well, that depends, that depends on the precise meaning of what you mean by the arrow time applying within the box. If you are just imagining the box and not letting it be observed or interact with the rest of the universe in any way, then I think the right thing to say is, there is no arrow of time in that box. You have reached equilibrium in that box. There's no place for entropy to go, and indeed, if you were able to look at it without disturbing it from outside, you would see the same thing from moment to moment. There's no arrow of time in that box. If you think about possible interventions, if you think about possible ways that you can interact with the box, then the fact that you are not in thermal equilibrium and you still have an arrow of time comes into play, and now there's an arrow of time. If you poke a hole in the box and the gas goes out, that's because the wider universe is not in thermal equilibrium.

0:32:48.0 SC: So, I think that you might wanna say, "Well, what if I were living in the box, and everything was in thermal equilibrium?", but that's not a logical statement, because you would not be living in a box that was in thermal equilibrium, 'cause people can't live in thermal equilibrium conditions. So, it makes perfect sense to say there's no actual measurable arrow of time confined into the box, but the box is part of a wider environment where the arrow of time is very clear and important.

0:33:18.9 SC: Fernando del Quito says, "In your recent episode, AI Thinks Different, you emphasize that LLMs very likely do not model the world, providing examples as evidence. You argue that their apparent world modeling is merely a by-product of their proficiency and language processing enabled by a vast training corpus filled with diverse human input, and any semblance of modeling the world is just a mirage. However, after a year of extensive usage, I have developed a strong sense that these systems do, in fact, model the world in some way, by excelling in language and compressing [0:33:48.2] ____ this vast knowledge, LLMs must be modeling reality in some way. The language sphere intersects with the mental, which, in turn, intersects with the physical and mathematical spheres. Don't you think that your definition of modeling the world might be too stringent?"

0:34:03.2 SC: No, I do not think my definition of modeling the world is too stringent. As I said very, very clearly in the episode, it is absolutely the case that large language models are able to talk as if they are human agents understanding the world. That's because they're trained on things said by human agents who understand the world. It's not that surprising that they're able to sound like that. The question is, how do they sound like that? What is the method that they use to sound like that? Is it in fact to think like a human being, which means to have a model of the world, which means to have categories in your mind, like frying pan and ice cream and so forth, and these categories have different features and they interact in certain ways. And it's a model, that's what a model is, it's a little toy representation in some way. Do they do that? Or are they just learning that, given certain sentences, certain tokens appearing in a sequence, there's a certain probability for other tokens to appear next.

0:35:10.4 SC: So that's a question, and that question is not answered by looking at large language model outputs and saying, "Wow, they sound very human," or "Wow, those are very impressive," or "Wow, that sounds like knowledge," because that would be true under either one of the two hypotheses. You have to specifically look at questions that are designed to illustrate whether or not there's actually a model working underneath, or instead, the system is just sort of putting together most likely word combinations without ever understanding things. That's why I chose those particular examples that I did.

0:35:48.8 SC: So, again, I try to be clear. I don't know. On the one hand, it is absolutely possible that, in practice, despite the fact that they were not trained to do so, or let's say they were not programmed to do so, it is absolutely possible that large language models have implicitly developed a model of the world. That is secretly the way that they are able to sound so human. It's also possible that it's just a mirage, that they have figured out how to sound human without having that model. But that is the question. It's not whether they can sound human, 'cause we admit that they can sound human; it's how they do it. So the point is not just that they sometimes make mistakes. Of course. People make mistakes, large language models make mistakes, that's not a big deal. The question is what kind of mistakes they make.

0:36:44.6 SC: So there was some philosophers on Twitter the other day were joking, because they were asking ChatGPT or GPT4, supposedly the good one, "Who was the most famous philosopher whose name begins with the letter M?" And GPT4 says, "Aristotle." [chuckle] Which is interesting, because if you had asked GPT4, "Could you please summarize the view of the nature of motion in Aristotle's physics?", it would do a great job. And you go, "Wow, it's so smart," right? Because it's been trained on not only Aristotle, but other people talking about Aristotle and so forth. So it sounds smart, but then you ask it, "Who's a famous philosopher whose name begins with the letter M?", and it says Aristotle. Which is not true, by the way. Not the correct answer.

0:37:32.1 SC: The point is not that it makes a mistake. The point is that it makes a kind of mistake that would be nearly impossible to make if that ChatGPT, GPT4, rather, actually had some model of what it meant to be a word beginning with the letter M. There's the kind of mistakes that you make if you have a model of the world, but you're just not analyzing it very well or whatever. And there's other questions that if you have a model of the world, you know what these words mean, in some sense, you would never make those mistakes. And this is an example. So, yeah, I don't know what to say, because all you've said, Fernando, is that you've played with LLMs a lot and you feel differently. That's not really evidence. I don't know what to do with that. But I think that you can read the papers. There's been papers in the technical literature arguing back and forth about this issue. And as a Patreon supporter, you know that in the reflection video, which I do for Patreon supporters, I asked GPT4 what it thought about these issues, and it was very clear. It was like, "No. We do not have a model of the world. [chuckle] We are just large language models looking at correlations between words." That doesn't mean it's true, because it makes mistakes. But it should make us think.

0:38:46.0 SC: RFD says, "Let's say God exists. If we were to ask what was God's purpose, that seems like an easy question to answer. But having literally done everything an infinite number of times, is it even possible for God to answer the question, 'What is my purpose now?' And if so, what do you think he would say?"

0:39:06.4 SC: You know, I'll just be honest. I think all questions like this are meaningless. Not that they're bad questions or dumb questions; they just make some assumptions that don't fit together. I don't think that God is a coherent concept. My reason for being an atheist is not that I think if God existed, the world be a nicer place. I just don't think that the idea of God makes sense. God is supposed to be omniscient and omnipotent and omnibenevolent, at least according to some conceptions of God. Of course, there's many different conceptions of God, and they're mutually incompatible, that makes it harder to have a conversation about this. But these ideas of omniscience and omnipotence, et cetera, aren't well-defined. Sorry.

0:39:52.0 SC: If you say that God is perfect, let's home in on that idea, God is perfect, which most traditional Western religious people would agree with, God is perfect, right? What does that mean? I understand what it means to be a perfect sphere. I can geometrically define what it means to be a sphere. I can say that certain shapes are approximately sphere, but not exactly. A cube is nowhere close to a fear, and I can imagine what a perfect sphere would be like. But I can also imagine a perfect cube. I cannot look at a perfect square, a perfect cube, and a perfect sphere and say, "Which is more perfect?" There's no single measure of perfection. That's just not a thing. [chuckle] To say that God just is perfect, not perfect at any particular task, but just perfect, that doesn't mean anything.

0:40:45.5 SC: So, when you say, if we were to ask what was God's purpose, that seems like an easy question to answer. No, I don't think that is easy at all. Purposes are things that are attached to we, finite, imperfect beings. We have goals. We have things that we would like to bring into existence that don't exist now. That's where our purposes come from. For God, that would just make no sense, as far as I can tell. I don't even know how to think about that kind of thing. So, sorry for not being more helpful, but my immediate response to questions like this is to just un-ask it, to say that's just not a well-defined question, this whole God idea was a bad idea from the start.

0:41:31.1 SC: Anonymous says, "My question relates to the morals of Many-Worlds. In one of your books, you clarify that the most moral thing to do isn't to just stay in your basement measuring quantum spins to duplicate the universe. This is because each world has a weight. However, you also clarify that we don't notice the universe branching because we split, too, keeping the ratio of our personal weight to our overall branches weight the same. These two things seem, to me, to be inconsistent. If the weight of our branch of the universe is irrelevant to our personal experience, but the utility of our entire branch should discounted by our branch's weigh, shouldn't the absolute number of these ratios, I.e., the number of branches, be what matters from moral judgments and not the weight?"

0:42:14.2 SC: No, I don't think so, but honestly, I have not followed the argument that you're trying to make there. So I might not be giving you a successful explanation here. But yes, when you do, when you're in your basement, here's the thought experiment for those of you who do not read Something Deeply Hidden, you convince yourself that you're a utilitarian. You wanna maximize the utility of the world. And you also think that the current utility of the world is positive. Maybe we don't know exactly the number, but there's more good than bad in the world, it's better that the world exists than not exist. And somehow, you've convinced yourself that in Many-Worlds, when you branch the world into two copies, the way to calculate utility is just to add the utility of the separate worlds. No weighting by the fitness of the world or the thickness, the amplitude squared of the worlds.

0:43:07.3 SC: So then you say, "Okay. The way that I can be the best possible moral agent is just to make a lot of quantum measurements and branch the world as often as I can, 'cause I'm making more and more universes and would make people more and more happy. There's more and more utility in the world." Now, of course, because you're not an idiot, you know that it doesn't actually make any individual person happier, 'cause they don't even know. You're in your basement making quantum measurements. It absolutely has zero effect on any other person in the universe. But you have somehow done some mental gymnastics to convince yourself that you are morally pure by doing this.

0:43:45.2 SC: So, the argument in the book is, how do you reconcile this, because you should count utility, not just by adding up the number of universes times their individual utilities, but by weighting, by the branch. So if you have two branches that were 1 over square root of 2 spin up plus 1 over square of 2 spin down, you would weight them by 1 over square root of 2 squared each, that is to say 1/2 plus 1/2. So if you have equal utility and then you branch them in two universes with 1/2 times one utility and 1/2 times the same utility, you end up with exactly the same utility you had before. And it's completely invisible to the people in that universe. So to me, that's completely consistent. I'm not quite sure what to say. From the inside the universe perspective, the weight is irrelevant, nothing's changing, and the number of universes is irrelevant, nothing is changing. From the God's eye-view perspective, the weight is super relevant and the weights are changing as you branch the universe. So, you just have to be consistent in which perspective you're choosing.

0:44:48.3 SC: Ken Wolf says, "My understanding of cosmic strings is that they are a theoretical artifact from the early universe that have as yet not been detected. If they do exist as predicted, is there much we can say about their attributes? Will they generate gravitational field, electromagnetic radiation or any other phenomena?"

0:45:08.1 SC: Yeah, absolutely. And they are quite theoretical, I should say, so, when you say, if they do exist as predicted, whether or not cosmic strings are predicted depends on unknown features of the laws of physics. We don't know whether the fundamental laws of physics predict the existence of cosmic strings or not. They're hypotheses that we're thinking about. Back when I was in graduate school, they were much more popular to think about thatn they are now because they were a competitor to inflation as seeds of the perturbations that grow into large-scale structure in the universe. That is because, to answer another question of Ken's, yes, they have a gravitational field. Indeed, everything has a gravitational field, so it's not surprising that a cosmic string would have one. And if you imagine these cosmic strings initially kind of randomly scattered throughout the universe and they move around and they push matter around, they could serve as seeds for galaxy formation, structure formation on large scales, and so forth.

0:46:05.4 SC: The nice thing is, we think, or we thought, that we knew enough to make predictions about what the patterns, the statistics, of large-scale structure would be if it were generated by cosmic strings or related relics of the early universe. And we also think that we can do the same thing if the initial perturbations come from primordial times, such as inflation. And they're very different. And we've taken the data now, in particular, the cosmic microwave background data, the anisotropies of the CMB were completely clear here. The anisotropies we see in the CMB are 100% compatible with primordial fluctuations, like we would get from inflation. That doesn't mean it was inflation, but inflation or something else that worked in the very, very early universe. They are completely incompatible with the kinds of fluctuations you would get from constantly regenerated perturbations, such as you would get from cosmic strings.

0:47:06.2 SC: That doesn't mean the cosmic strings are ruled out; it means that cosmic strings are not the origin of large-scale structure. Since we don't know whether they exist, there are plenty of free parameters you can play with. So one free parameter is basically, in fact, the single most important free parameter is the tension of the string, basically, the energy per unit length that the string has. If that's just large enough, it's interesting that if you had grand unification, so if you unified electromagnetism, the weak force and the strong force at the energy scale that we think it should be unified at, and that grand unified theory gave you cosmic strings, which is not necessary, but it was absolutely allowed, then the amplitude, the size, of density privations in the universe, is roughly what you would expect from the cosmic strings. That's why people were very interested in the idea.

0:47:57.5 SC: But that idea didn't work. So instead, you can imagine lighter cosmic strings. That is to say, cosmic strings with a lower tension, with less energy per length of the string. And that would just not leave an important impact on large-scale structure in the universe, and therefore, it's not yet ruled out. It also makes them harder to find, it makes them a little bit less interesting. But they could, as you also ask, interact with electromagnetism. In fact, Ed Witten and others pointed out that cosmic strings can be superconducting, which leads to very interesting electromagnetic effects. They might affect magnetic fields in the early universe. So they might still play a role, but there's no evidence that they exist, and there's no sort of need for them, as far as anyone knows right now.

0:48:45.4 SC: Matthew Wright says, "How seriously do you think the idea of superdeterminism should be taken when it comes to interpreting quantum mechanics? It sounds like there are some physics who believe it to be a serious possibility for how the world works, and others who consider it a completely implausible and even unscientific idea."

0:49:03.8 SC: I don't know if it's implausible or unscientific; I just think it's uninteresting, to me, personally. And that's for two reasons. One is because of superdeterminism itself. It just seems to be bending over backwards to do weird things. So for those of you who don't know, John Bell, when he proved His theorems about predictions of quantum mechanics, proved that there could be these non-local correlations between measurement outcomes that you could get in quantum mechanics that will be very hard to reproduce in any non-quantum mechanical theory. And in particular, the kinds of correlations seem to be non-local. So it seems to be exactly, as Einstein worried about, that you measure something here, and then there's spooky action at a distance, and that seems to affect the measurement outcomes far away. So different people who don't like this spooky non-locality will try to wriggle out of this and superdeterminism is a possible way to do it.

0:50:00.9 SC: And the way to do it is, everyone agrees that the Schrodinger equation, by itself, is deterministic. There's no randomness there. And if you knew the wave function of the universe at initial times, you could just evolve it. And we, both theorists and experimenters, as physicists, are part of the wave function of the universe. So, Bell made an assumption that we physicists can make any measurement we want to. In principle, we could measure anything we want. We could rotate our measurement devices, our Stern-Gerlach experiments, measure spin along any axis, et cetera, et cetera, and he used that to prove his theorem. So superdeterminism says, what if the wave function of the universe is set up to create certain physicists doing certain measurements, but not other physicists doing other measurements?

0:50:56.7 SC: So in other words, we don't have free choice about what measurements we're gonna make, because we're just part of the wave function of the universe. And what if that initial determination, if I can put it that way, of the wave function of the universe were just the right one to make it look like Bell's theorem was made at the predictions that it has for ordinary quantum mechanics? So, then you can have a local theory that gives you all the predictions that Bell's theory has. So, I'm sure that superdeterminism fans and advocates will not completely agree with my description of the theory, because I'm sure there are bells and whistles in choices that I don't know about. But that just sounds weird to me. That just sounds like a very, very, very bizarre way of getting out of what is staring you in the face in quantum mechanics. So, there's nothing about superdeterminism when it is explained to me that makes me go, "Oh, yes, I need to learn more about that."

0:52:00.9 SC: The other thing is that, and I've said this before, I am not that interested in the foundations of quantum mechanics, because I think I know what the foundations of quantum mechanics are: I think that it's Many-Worlds. My credence that Many-Worlds is on the right track is not 100%, but it's sufficiently high, and my life span is officially short, that I'm not gonna devote my time to worrying about other possible models. I'm just not interested in superdeterminism. So it might be right, and if it's right, ultimately, the people who thought it was right all along will absolutely have the right to laugh at me and make fun of me for not paying attention to it when I should have. But you gotta roll your dice and take your chances, and I think that it's just much more productive to put my own efforts into thinking about Many-Worlds and its implication and how it fits in with other things we know about physics.

0:52:57.2 SC: Everydayhuman@icloud.com says, "Short question: Are you coming back to Mastodon anytime soon?"

0:53:04.2 SC: No. Short answer. [chuckle] So, this is good timing for this question, actually, because Mastodon is a social media site, which is one of the many competitors to Twitter that has sprung up, and I did go on Mastodon originally, but it's just clunky. I just don't like using it. It's very, very annoying, let's just put it that way. It's kind of using Linux, if you know about that. There are people who love it and it's good for them, and that's great. If you like Mastodon, go nuts. And don't try to convince me that I should like it, 'cause I just don't like it that much.

0:53:39.3 SC: I am on Bluesky, and that's why the timing for this question is very good, because Bluesky just opened up to a wide... Anyone can join now. It used to be they were working out the kinks for various things. It's still not completely settled, Bluesky; it is basically a Twitter clone. It looks like Twitter, it works like Twitter. It's a little bit better, in some ways. You have a little bit more control over moderation and what you see in the algorithm and things like that, but there's different sort of multimedia embedded, higher-level tech questions that are not quite as put into in Bluesky as they are in Twitter. But the vibe is way better, and it's super easy to use. I think that, over and over again, people who try to invent a Twitter clone, whether it was Threads or Spoutable or Post or what have you, they try to make improvements. And what they didn't realize is the improvements were what they wanted, but not what other people wanted.

0:54:40.1 SC: And so, Bluesky is just like Twitter, but better, because there's less promulgation of racism and trolling and bad faith and things like that. So, I like it better. The people who I enjoy following are more on Bluesky now than they are elsewhere. I felt guilty about inviting everyone to come to Bluesky back when it was closed, because it felt like it was behind a walled garden or something. But now it's open. They've opened it up; everyone can join. That's where I'm going to be active going forward. I don't promise to be super active; I'm finding it hard to find any time to be active in social media at all, but right now, I only use Twitter to like say, "Hey, I have a new podcast out," or "Hey, I have a new book out." I'm just using it for promotional purposes. If I'm saying something amusing or interesting, it will be on Bluesky instead.

0:55:30.0 SC: Kevin O'Toole says, "Priority question: Bayesian inference is often claimed to be fully dependent on choice of priors. However, it's less discussed that as the evidence piles up, differing credences converge, diminishing the significance of the prior choice. For instance, two scientists with different beliefs about the odds of heads in a mysterious set of coins would increasingly agree as they start flipping. With the world of evidence and information, it seems plausible that even an absurdly broad set of priors could be exposed to enough evidence to increasingly converge until it puts enormous credence on specific scientific conclusions, like perception is mostly reliable and electrons exist. Obviously, this process would not be computationally feasible, but if it were, do you think the credences this holistic Bayesian process would arrive at would match with what is generally accepted in science?"

0:56:22.3 SC: Well, there's a whole bunch going on here. Let's try to pull it apart. When you say an absurdly broad set of priors, let's focus in, let's say you have a particular proposition, and you say, "My credence in this proposition is very, very, very, very small." And then someone else says, "Well, I'm gonna do some experiments, I'm gonna collect some data." What happens if I keep collecting data for which the likelihood of that data under this weird proposition is high, and the likelihood of that data under other propositions is low? Then my absurdly low credence in this proposition will indeed get larger over time. And you're completely in a position where you're just saying, "I have a small number, and I multiply it by a big number." Is the answer big or small? That's a completely ill-defined question, unless you say really exactly how small and how big. It's absolutely the case that for any fixed small number, there is another number big enough that, if I multiply the small number by the big number, the answer is still big. That's true. So, no matter how small your initial credences are, there is an amount of evidence you can collect in their favor that will eventually bring those credences up. But it might be wildly and practical to do that. It might take longer in the age of the universe or something like that. So this is very much a thought experiment kind of thing; it's not a, in any sense, practical kind of thing.

0:57:53.3 SC: The other thing is, it's not at all obvious what all of your credences are supposed to be in this scenario. For science, we have to say not just, "I have one proposition and I'm gonna calculate its credence," but to be a good Bayesian, you have to say, "Well, what are the likelihood functions for all of the other possible propositions that are not compatible with the one you're thinking about?" That's a little ill-defined here, and I worry about these things where you start with an ill-defined set of credences and try to draw strong conclusions from them. Certainly the underlying spirit is correct. It is actually among Bayesians, I don't know among people who spend their time doing more profitable things, but among professional Bayesians, it is very well-known that the priors ultimately go away if you collect enough data, that different people starting with different priors will converge on the same ultimate conclusions. That's a well-known fact. But to make that specific to all of science, if you just did all of possible things, is a little bit ill-defined for me to actually reach any particular conclusion. So the specific question you ask is, "Do you think the credences this holistic Bayesian process would arrive at would match with what is generally accepted in science?"

0:59:11.6 SC: Well, the words would awry that, are a little bit ill-defined. After what period of time? After a finite period of time? If you tell me what finite period of time, I can always pick a credence so small that that would not lead to the right scientific conclusions. So, I don't think that question is completely well-defined there, but if what you are trying to get at is, are the currently accepted scientific principles of the world robust enough that very different sets of reasons starting from very different initial assumptions would more or less converge on them? I think, partly yes, partly no. I think it depends on exactly what your starting point was, what your evidence was, and so forth. In our universe, we invented the second law of thermodynamics before we invented general relativity. And so there was a period of time where we knew about the second law, but we didn't know about general relativity. So if you ask that question then, they would think that Newtonian mechanics was right about space and time, but they would agree with the second law.

1:00:20.2 SC: Whereas I can easily imagine an alternative history where general relativity was invented before statistical mechanics and the second law of thermodynamics. And so then if you ask that question to those people, they would think of general relativity as part of the generally accepted scientific principles. So I think that even in the real world, even if you just take history and imagine small alterations of it, it's easy to pick moments when the scientific consensus was different, and very plausible histories would lead you to that. So I can't promise you that very, very different starting points would lead to the same scientific consensus at any one moment in time. I do think that if you had very... So, I guess let me back up and say this correctly.

1:01:07.9 SC: I also think that the thought experiment is a little bit not quite matching what happens in the real world, because I think that something that is underemphasized, in my view, is the fact that, sure, different people are welcome to different priors. But in fact, people often pick priors that are pretty compatible with each other. Even if they're different by small amounts, they're not wildly, wildly, wildly different. So, that's why science works, not because people have wildly, wildly different priors, but they just do an infinite amount of data-collecting and eventually converge. It's because in fact, reasonable human beings kind of are similar to each other and kind of have similar or at least compatible priors. And it might take more evidence and more evidence to convince some people, but the vast majority of people converge in a simple way. So I think that the actual empirical success of science is both due to the effectiveness of gathering evidence and using Bayes's theorem, but also because people are not completely incompatible with each other when it comes to their actual priors.

1:02:16.5 SC: Eric Fast says, "I enjoyed your solo podcast on artificial intelligence, and I agree with most of your points. But I don't think I agree with the claim that AI doesn't have values. One way we might define value is in the economic sense of revealed preferences." And then Eric goes on, says much more, you can read that on the Patreon page, et cetera.

1:02:37.0 SC: I don't think that's what a value is. Yeah, I don't think that counts. It's not sufficient. That's part of what having a value is; a value is a revealed preference. But if I have a rock and I roll it down the hill and it reaches the bottom of the hill and comes to rest after bouncing around a little bit, the rock has revealed its preference to be at the bottom of the valley. That is no sense of value. [chuckle] That is just not what we need, right? I mean, it's a revealed preference, in the sense that you rolled the rock and that's where it went. But it's not what we mean when we think about a human being having a preference for some set of outcomes, because the rock is dumb. The rock just rolls down. If you put a barrier up in front of the rock, it would stop rolling and it would do something different.

1:03:20.1 SC: Whereas a human being, if you put a barrier, if they were trying to walk down a hill and you put a barrier up, would try to get around the barrier, if they actually had a preference for being at the bottom of the hill. It's subtle, and I don't claim everything makes perfect sense or even has been worked out, but when we talk about human beings and their values, we don't just mean what they do, we don't just mean their revealed preferences; we mean a whole set of counterfactual statements. We mean not only have they done this, but if conditions had been different, they would have been doing that. And the way to understand all of these things, that they both did do and would have done, can be bundled up in a simple idea that they have some values. So they actually have some goals or some preferences, however you wanna put it, in a way that rocks falling down hills do not have. And it's much easier to describe the rock with local laws of physics rather than with some teleology, the rock wants to be at the bottom of the hill. That's a difference between human beings and rocks.

1:04:24.5 SC: Where does AI fit in to this? I would love to know. I think that the AI's way of thinking is sufficiently different, and its way of being constructed is sufficiently wildly different than human beings, that values and revealed preferences, and preferences, not just revealed ones, are not the best way of thinking about what the AI does. And you could convince me I was wrong. The way to convince me I'm wrong is to say, "Here is a set of behaviors of AIs that is conveniently summarized by this future goal-directed behavior rather than by some local sort of dynamical behavior, like a rock pulling down a hill." I don't think you're gonna do that, and I could be wrong, but I don't think you're gonna do that because AIs don't have biology. They don't have an arrow of time. They don't have dissipation built into them. They don't rely on free energy being converted into entropy. They're just very different things than human beings are. Again, and with all the footnotes usually that this is the current generation, they could easily be, 'cause they're just physical systems, after all, this is just a statement about the current state of the art. I do really think it's different.

1:05:37.2 SC: Thenpy Stein says, "On the episode with Adam Frank about detecting life outside our planet, you made a very interesting comment that I've never heard before about Dyson spheres, but you went through it very quickly, and I didn't fully understand your point. Were you saying that Dyson spheres wouldn't be detectable? Could you tell us more about why seeing them would be so difficult?"

1:05:57.0 SC: Well, I wouldn't take too seriously, what I was trying to say, 'cause it was just off the top of my head, without sitting down and doing the calculations. But my point is that, as we were just saying about what makes human being special, one of the things that makes you being special as living organisms is that they feed off of free energy. That is to say, free energy in the technical sense of a kind of energy that is low-entropy energy, a kind of energy that can be used to do work. And that is characterized by being out of thermal equilibrium. The sun is a hot spot in a cold sky. That's what makes life possible. It's very far from thermal equilibrium. If the whole sky were one temperature, life would be impossible. No matter what that temperature was, it if was high temperature, low temperature, if there was an equilibrium distribution all over the sky, there'd be no engine driving life.

1:06:52.8 SC: So, to me, when we start talking about Dyson spheres, it sounds to me too simplistic to say, "Well, I surround a star, and then I come to its temperature and I radiate at that temperature." If the temperature you're radiating at is much higher temperature than that of the background radiation, then you are still, cosmic microwave background radiation, you are still a hot spot in a cold sky. There's still people who could live outside of your Dyson sphere and take the fuel that is being given out by your Dyson sphere, because it's giving off a lot of free energy. Maybe not as much as the original star would have, but still, it's giving it off. And so, if you're really into squeezing all of the free energy you can out of your radiation, then you're gonna keep squeezing it until the heat radiation you're giving off, I think, is that of the temperature of the cosmic microwave background, and you sort of converted it into as a high-entropy thing as you can convert it into.

1:07:54.7 SC: There's probably loopholes there or a fine print that I haven't thought of, but that would be my impression. And if that's true, it still might be possible to detect them, because maybe they're concentrated. Maybe the temperature of the Dyson sphere is the same as the microwave background, but it's just brighter than the background so it looked like a bright spot on the CMB map or something like that. But I think that's different than what they're actually looking for, so I don't actually understand the argument completely.

1:08:27.1 SC: CJ says, "Suppose for a moment that your proposed quintessence field from Episode 127 on the screwy universe turns out to be real. If the quintessence field is creating positive vacuum energy, could there still be a negative cosmological constant that is overpowered by the quintessence field where in that scenario is the net effect considered the cosmological constant?"

1:08:48.4 SC: Yeah, so this is a good question. The answer is very easy, but we don't know what the right answer is. The possible answers are very easy, but we don't know what nature actually does. So the point here is, in Episode 127, solo episode, I talked about a particular form of quintessence, which has an effect on radiation in the universe causing its polarization to rotate in an effect known as cosmic by refringence. And that's an interesting thing to look at, and there's even very slight hints in the data that maybe it is there, which is very exciting. And one way of talking about this is in terms of a scalar field. Just like for inflation in the early universe, you could have dark energy that is due to a scalar field, rolling down a potential. Rather than just the cosmological constant, which is the easiest and probably correct answer for what the dark energy is, the cosmological constant is just a constant; it's not rolling, there's no dynamics, there's no new degrees of freedom or anything. So it's either a cosmological constant, or it's a rolling scalar field, or it's something else that might be more exotic.

1:09:53.0 SC: The thing about the rolling scalar field is, it's rolling down a potential, so you literally imagine a ball rolling down a hill, and it has a height above ground, in some sense. There is the value of the potential, which would be zero dark energy. And for whatever reason, right now, the scalar field is at a higher value of dark energy, and that's what's making the universe accelerate. But balls roll down hills, and so typically, the scalar field is rolling down the potential. And so the amount, the effective amount of dark energy in the future, will be a lower number. And so the question is, will it be a number that is exactly zero? Or can it be a lower number that is still positive? Or could it even become a negative number? And the answer is absolutely all of these are possible. We know so little about the constraints on things like that. So, when people, including myself, say that the acceleration of the universe most likely means that it will accelerate forever, that is assuming that the right theory is the cosmological constant. And I think that that's the most likely way to bat right now, but it's far from certain.

1:11:03.3 SC: So it's also very possible that you have a slowly rolling scalar field, quintessence field, and it will roll down so far that the total amount of energy density empty space will effectively be a negative number. And if that happens, the universe will re-collapse, and there'll be a finite lifetime. We'll be something closer to anti-De Sitter space than what we have right now, which is closer to De Sitter space. So all of these are absolutely on the table. My money is on the good old cosmological constant, but my interest is absolutely quintessence.

1:11:39.0 SC: Michael Masami says, "Is general relativity emergent? If and when we have a quantum description of gravity, GR will not lose its power and usefulness to describe the world, but is it at the most fundamental level?"

1:11:52.1 SC: I think it might very well be. We don't know is the short, honest answer here, 'cause we haven't figured out quantum gravity completely. In string theory, straightforwardly interpreted, in good old-fashioned string theory, let me put it that way, general relativity is not emergent. It's a fundamental thing. There are spin to vibrations of the string that look like graviton, and that is, gravity. But string theory has also given birth to the holographic principle. And in holography, you can very well say that there is a different theory in one lower dimension that doesn't have gravity, but gravity is emergent from quantum entanglement in that different theory with one higher dimension. That will be holography.

1:12:42.5 SC: My collaborators and I have pursued a vision where general relativity is emergent in a different way from entanglement in our good old three-plus-one-dimensional space-time. It's only gonna be valid in the weak field limit, that's why it's actually compatible with holography, but it's certainly very much of the spirit that the fundamental stuff of the world is not space and time; it's the wave function of the universe evolving with time. And general relativity emerges from that. So, I actually think that not only is it possible, but it's probably the right way to go, to think of general relativity as emergent. But we just don't know yet. We'll have to think about that.

1:13:19.6 SC: Nathan says, "I am a philosophical naturalist, and I am terrified of death. You have indicated that you are at peace with mortality. Did this attitude come easily to you, or did you have to struggle to obtain it?"

1:13:31.9 SC: Well, I don't know exactly what I indicated, so, let me try to be clear about this. I am resigned to death, I am reconciled to it. My credence that I will die and cease to exist and have no afterlife, no life after death, is very, very, very high. Sufficiently high that even though it's not 100%, it might as well be. I'm not gonna spend a lot of time worrying about the possibility of life after death. That doesn't mean I'm happy with it. That doesn't mean that it doesn't make me anxious or sad, et cetera. I would like to live a nice long life. As I said in the episode on immortality, the solo episode I did recently, the people who were at the conference I was at at Santa Fe largely voted that they would not want to live 10,000 years. And I was absolutely in favor of living 10,000 years myself.

1:14:23.5 SC: So, I think that when I come close to death, I want to think that I will accept it, and I will not go into denial or anything like that. But it is the end. That's dramatic. That's something that is absolutely okay to take seriously, to struggle with, in some way. There's no use in struggling, there's no benefit that you get it from fooling yourself or from imagining that there was something else going on in the universe. We have to take reality at its face value. And so I like to think that when it comes, I will do that with equanimity. But I can't tell you that for sure, 'cause I haven't faced it. I haven't been in that situation. I don't wanna die, I wanna live a long time, so we'll see what happens. But I know that it's not gonna be forever, even if it does turn out to be quite a while.

1:15:16.0 SC: And let me also say, one of the reasons why it's hard is because there is nothing that it is like to be dead. There's nothing that's like to not exist. When we have our imagination, we can imagine the world after we're gone, and maybe these are some people who are sad, maybe there are some people who remember us fondly and we have a legacy, maybe not, maybe we just disappear, whatever, and our current state of thinking and coming to emotional grips with reality is affected by that imagination, affected by that mental time travel, as we talked about with Adam Bully a while back. And that's legit. That's okay, that's fine. That is as it should be. A lot of human life involves conceptualizing and imagining the future. And the fact of death means that that comes to an end. We can conceptualize it, but we won't actually be there to think about it.

1:16:13.8 SC: So, different people react differently to that. Different people feel very, very strongly about their legacy and their memory and things like that. My own personal take is that I'm trying to maximize what goes on in the world while I'm here: My happiness, the happiness of others, the value I get out of the world. That's what I can do. After I'm gone, if people wanna remember me fondly, that's great; if they don't remember me at all or remember me badly, that's out of my hands, 'cause I will be gone. I really think that putting the emphasis on what I can actually change by my actions is the more healthy way to think and to go. But I'm not claiming there's any easy thing. There's no simple nostrum here. This is one or the fundamental issue of being a mortal human being. Struggling with it is the most natural thing in the world.

1:17:08.2 SC: James Allen says, "What is the physical mechanism that gives rise to half-life in radioactive decay? If I wait a certain period and half my sample is gone, if I wait the same amount of time, why isn't all of it gone?"

1:17:22.0 SC: Well, that's a good question. It's a classic kind of question, but that's okay, we can examine those here. This is gonna depend, sadly, the exact right words to say, will depend on what you think about quantum mechanics. Here is the right way to think about it. Think about every individual nucleus in a radioactive sample. Imagine that your sample is all radioactive nuclei of exactly the same kind. And imagine that each nucleus is initially not decayed, so you start with some sample, all not decayed nuclei. Then what quantum mechanic says is that every nucleus obeys the Schrodinger equation. And as we briefly alluded to before, that Schrodinger equation will say that each nucleus, the wave function of each nucleus, we don't even need to worry about them being entangled with each other or whatever. Each individual wave function will evolve into a superposition of decayed and not decayed.

1:18:15.2 SC: So when you look at it, when you make an observation of that particular nucleus, you will have a probability of seeing it either decayed or not decayed. And the thing is, those probabilities are completely independent for every single nucleus. It's not like there's a half-life of one minute and all of the atoms remain undecayed, and then boom, at one minute, they all are decayed. It's like at 10 seconds, 20 seconds, 30 seconds, there is a certain probability that's gonna be less than a half, if we're talking about the half-life, that each individual nucleus will have decayed. But some of them will; you have a lot of nuclei in there. So if you have a very large number of nuclei, and there's a probability that each one of them will decay, probably some of them will. And those probabilities per nucleus are completely independent of each other. So at every moment, you could ask the question, what is the probability that only one nucleus has decayed? Only two nuclei have decayed? Et cetera, all the way up.

1:19:18.2 SC: And what you find is that the probability that most of the nuclei have decayed is very small when you just start out, when you start out very close to completely undecayed. But it grows. It's always a probability, so there will always be appointing to where half of the nuclei have decayed and have not, and then after you go past that point, there will still be a probability that each remaining nucleus will decay or will not decay. And if you have enough of those nuclei, the time when it takes half of them to decay will actually be pretty well-specified. Each individual nucleus is hard to predict exactly when it will decay, but when you have large numbers of them, the large numbers give you a pretty sharp view of what time that is. Just like if you flip a 50/50 coin many, many times, it gets very, very close to 50% after you flip it millions of times, even though when you flip it four times, you can get a lot of fluctuations. I hope that helps just a little bit.

1:20:22.4 SC: Kyle Cabasares says, "When you started creating multimedia content such as podcasts and videos, were you doing all the filming and editing yourself initially, or did you have an editor help from the get-go? Trying to figure out how to balance my own content creation interest with my professional responsibilities."

1:20:38.9 SC: Well, the podcast and the videos are two very different things. But the answer is the same, which is that, no, I did not have any filming or editing help myself, it was all me doing the work there. I did have help getting things set up in terms of equipment to buy and software to use and things like that. I was certainly happy to appeal to people who know better than I do about those things, and they were very helpful. But actually doing it, I always wanted to be me. So for the podcast, that's not so hard. I think if you listen to early episodes of Mindscape, the audio quality was not as good because I was still learning on the fly. Sometimes it just all miraculously came together; other times, it was a little sketchy. It's more uniformly respectable now. But it's not that hard. You spend some time learning it and you do it. I would encourage anyone to just figure out the basics of recording and a little bit of massaging the audio in a digital audio workstation of some sort.

1:21:40.9 SC: I use Audacity, but there's also a very good online tool now from Adobe. You have to pay for it to get the real version now, but it's an AI audio clean-upper. [chuckle] It improves the audio quality of spoken word audio, if you recorded it in an echoey room or with a cheap microphone or something like that. It's not perfect, but if you have what would otherwise be unusable bad audio, I can save it sometimes. For video, I'm not still a video person, I don't do a lot of video, but the little videos that I did for The Biggest Ideas in the Universe, that was a pandemic project, that was literally the time when it was not really going to be possible to get a lot of help. I bought a green screen the day before everything shut down and I would not have been able to buy the green screen anymore.

1:22:34.7 SC: So, yeah, that, I also taught myself. That's why the lighting is very bad in some of those original videos. And again, it gets better. I learned, and that's also just part of the fun for me. I'm not sponsored by any big magazine or media corporation or anything like that, so I'm not doing it in a completely mercenary way. I wanna learn the skill, I wanna figure out what's going on. And that part has been fun. But whether it's right for you, it's hard to tell. Everyone is different, and I have no objections to people who are part of a bigger machine where they have professionals doing these things, and that's completely okay, of course.

1:23:13.8 SC: Kyle Steven says, "Unlike many other podcasts, you never have had the same person on the podcast multiple times. Do you foresee ever having former guests return in the future?"

1:23:24.6 SC: I go back and forth about this. I think the answer is no, I do not foresee that. Or let me put it this way: I sometimes fantasize that when the podcast is coming to an end, when I declare a series finale for Mindscape, then maybe I'll revisit some of my favorite guests or even panels with multiple favorite guests. Who knows? I could do anything. No rules when you're near the end there. But what I told myself at the beginning was, there are a lot of interesting people to talk to. And I feel this way about doing research as a scientist. When you're a scientist and you first learn to do research as a graduate student, it's all hard, 'cause you don't know any of it. [1:24:08.9] ____ learn to write a paper, to get a publishable result, it's very, very hard. You do a lot of work when you were a graduate student to get up to that level.

1:24:16.3 SC: But then once you're at that level, it is easy to either basically do the same kind of thing or to wring small changes on what you have been doing. And that's an easy rut to fall into. And sometimes maybe in my life, I've been guilty of falling into that rut, but I don't want to, so I wanna nudge myself out of it. That's why I'm thinking about... One of the reasons why I'm thinking about things like complexity and philosophy these days. Likewise, for the podcast, I absolutely know that there are people who I've had on the podcast already who would love to come on again and who will be great. And I feel bad 'cause sometimes they ask if they can come on again, and I had to say, "Well, no, we have this policy." But the point is, I don't wanna fall into the rut. There's so many people out there who are super-duper interesting, some of whom you've heard of, and I just haven't gotten around to yet; some of whom you've never heard of, and I only recently discovered and I'm trying to bring to you. And that kind of freshness really keeps me interested.

1:25:14.0 SC: It's also much more work. [chuckle] There's a certain amount of work that goes into having people on the podcast I'm not familiar with already. And if I've had them on before, then by definition, I'm at least a little bit familiar with them. So, am I gonna get lazier with time and start having people again? Probably not, but I'm not gonna make any super-duper strong promises either way.

1:25:37.1 SC: Richard Cashtan says, "Many people believe that the earth had an advanced civilization before ours, so long ago that all hint of them was very deep, where we will never find it. If the earth had a previous civilization like that, wouldn't they have had used up all of the oil and natural gas so that we wouldn't be finding all the huge pools our modern civilization has used?"

1:25:56.4 SC: Well, I don't think that many people believe that. Some people maybe believe that. It's certainly not a scientific consensus. I think it's quite unlikely. It was interesting that we talked about this a little bit with Adam Frank when he was on the podcast, and he said signs, techno signatures of previously existing civilizations would be harder to find than you might think. Things do get wiped out. Your point about natural resources being used up is a perfectly good one, but it's also a little presumptuous. It's presumptuous that they would use the same kinds of technology that we would. Maybe they would. Maybe there's an argument that it's just an obvious thing to do to dig up oil from the ground and use that. But maybe not. I really, honestly can't say for sure. Also maybe plate tectonics, et cetera, has brought different pockets of fossil fuels up to accessible parts of the Earth's surface than were there before. I don't know that, either.

1:26:57.2 SC: So I think it's very unlikely there ever was an advanced civilization on Earth. You make a good point that one... So as a Bayesian, let's put it that way, if you have a proposition that there was an advanced civilization on Earth and you have forgotten temporarily whether or not there is a fossil fuel on the Earth now, the likelihood that the fossil fuels were all used up, if there was an advanced civilization, is larger than the likelihood if there was not an advanced civilization. So the fact that we have considerable resources of fossil fuels should decrease your credence that there ever was an advanced civilization. I just don't know by how much.

1:27:36.9 SC: David Rabinowitz says, "In Something Deeply Hidden, you object to Max Tegmark's quantum immortality thought experiment by arguing it was the wrong way to evaluate the costs of dying. But this unintentionally sidesteps might be an equally interesting part of the question. Do you believe the quantum immortality experiment would work? That is, do you believe someone playing quantum coin flip roulette indefinitely would, in fact, find themselves miraculously surviving, and does every conscious entity who ever lived and who had a possible survival trajectory under the laws of physics believe they are still alive and/or immortal in the [1:28:09.1] ____ that selected survivalist branches of the wave function they occupy?"

1:28:11.6 SC: I mean, yes, but in a completely trivial sense. Every entity that has things that you would describe as beliefs also has the property that they're alive. There are no dead people who have beliefs. So everyone who is alive believes that they are in the part of the wave function of the universe where they're still alive. That's true, whether Many-Worlds is true, whether it's truly stochastic. Any alive creature thinks thinks that they're still alive. That's all you're ultimately saying there. So I absolutely believe that Tegmark's set up, where if you had many worlds and you do some quantum coin flip, and one of you dies, one of you survives, leaves you with a large number of branches where only some of them or a very tiny number of them, maybe only one, has an alive person on it. And that alive person in that ensemble is the only one you can ask, "How are you feeling about it?" I just don't think that before you do that experiment, you should be sanguine about the fact that most of the branches are not gonna have you on it anymore.

1:29:19.2 SC: Tyler Haley says, "I have a question about dark energy. Should we think of dark energy as having substance like dark matter does? The energy part of dark energy implies there's some sort of mass or momentum associated with it, by Einstein's relation e squared equals m squared plus p squared." Tyler is indicating that he can't do the superscripts in the question.

1:29:43.5 SC: No. It does not necessarily imply that at all. That equation, the energy is the square root of mass squared plus momentum squared, is meant to apply to some things and not others. The things it's meant to apply to are objects. Point-like objects, honestly. If you have something that has internal structure, then there's other kinds of energy that it can have. But objects that can be idealized as localized at some region of space, those are the ones that have things we call "mass" and "momentum." Why? Because those are things that you can push with your finger, and the resistance to being pushed is what we mean by the mass. If the dark energy is the cosmological constant, that is a property of space-time itself. It is the energy density per cubic centimeter of space-time itself. It's not something you can push. It's not something that can resist being accelerated. It's not something that has momentum. So it has energy, but it doesn't have any of those other things, because it's not the kind of thing that mass and momentum are associated with.

1:30:50.0 SC: Now, we don't know whether the dark energy is a substance like dark matter, or whether it's something different. I've written papers on both sides, to be honest, but if it's the simplest thing, like the cosmological constant, then it is not a substance really, and there's certainly a priori argument that it has to be anything like that.

1:31:11.2 SC: Only Normal Person says, "Reading on The Origin of Time by Mindscape guest Thomas Hertog, I was surprised that part of the book was interested in rehabilitating Wheeler's participatory universe into a real theory. Is that concept something other theoretical physicists are interested in legitimizing, or would the vast majority consider to be unsalvageable?"

1:31:31.4 SC: I think the vast majority would just consider it to be too vague to matter that much. But deep issues of quantum mechanics and Cosmology are places where things that other physicists have the luxury of ignoring suddenly become relevant. I'm not a big fan of Wheeler's rhetorical flourishes when it comes to quantum mechanics and the observer and things like that. I think it sort of obscure things more than it clarifies. But here is a case. Thomas was trying to say something very specific, that the way we should think about quantum cosmology is to start with ourselves and build outwards, rather than starting with the Big Bang and letting it vibrate forwards. It's compatible. These are just two ways of thinking about the same underlying physics, but they're trying to make the case, he and Stephen Hawking, that certain cosmological puzzles make more sense, if we start with the fact of who we are and our observed universe around us, and then ask how that fits in with the wider cosmos, rather than pretending to be God and saying, "Here's the whole cosmos, where are we in it?"

1:32:45.3 SC: I don't know whether that's the right attitude to have or not. I don't know whether it's just a family resemblance to what Wheeler was trying to say or whether that's really what he had mind all along, but it's all ongoing research. I'm happy to consider these slightly fanciful ways of thinking.

1:33:01.9 SC: Bryan says, "Here is what I believe I understand from you: The expansion of the universe is accelerating. Here is my related question: What is the acceleration rate? Is it constant, or is it different in different places? Do we even know it?"

1:33:17.6 SC: So, parts of this are easier to answer than others. It mostly is constant. I'll explain why the "mostly" comes from in just a second. As far as we know, it is not different in different places, and we do know what it is, if it exists at all. So I say, "if it exists at all," here's what I mean. Usually, when we talk about something accelerating, what do you mean, accelerate? You mean that the velocity is changing with time. So, if you use the phrase "the acceleration of the universe," that you might naturally think there's something called the velocity of the universe, and it's increasing with time. But the universe doesn't have a velocity. A velocity is distance divided by time. What is the distance we're talking about here?

1:34:09.4 SC: The universe has a scale factor, which gives you the relative distance between galaxies as a function of time. So, something that makes perfect sense is to say, the scale factor of the universe today is two times what it was a million years ago, or whatever. That makes perfect sense. To ask what the scale factor is doesn't make sense. That's a choice of units, that's kind of arbitrary. So, the universe isn't really something to which we should attach the idea of an acceleration in the conventional sense. Why do we do so anyway? Because that factor, the scale factor, that tells you the relative distance between galaxies, is a function of time, and we can plot it, and then we can take its first derivative, and we can take its second derivative, and its third derivative, that is to say, the rate at which the scale factor is changing, the rate of which its rate of change is changing, and so forth.

1:35:08.2 SC: And what we mean when we say that the universe is accelerating really is just that the second derivative is positive, that the rate of change of the universe, the rate of change of the scale factor of the universe with respect to time, is increasing. So if you could plot the scale factor as a function of time, it is not only going up, but it is going up in the sort of accelerating-looking way. What I've hidden from you in that is that if you say, "Okay. If it's going up so it has a derivative, what is the derivative? What is the rate of change of the scale factor?" But I just told you that's a meaningless question, 'cause there's no such thing as what the scale factor is. It's only relative to other things that it makes any sense. If you look into the details in a cosmology book, the Hubble parameter or the Hubble constant, but it's really not a constant so we call it the parameter, is the number that physically is real and does actually characterize the rate of expansion of the universe.

1:36:08.7 SC: But it's not the derivative of the scale factor. It's the derivative of the scale factor divided by the scale itself. So if you multiply the scale factor by 2, you get a 2 in the numerator and denominator and they cancel out. So the Hubble constant, a dot over a, derivative of the scale factor divided by scale factor, that's a real physical thing, and we can actually measure that. So, now, I hear you saying, "Okay, good. What it must mean to say the universe is accelerating is that the Hubble constant is increasing." No. It doesn't mean that. Precisely because what the university accelerating means is that a dot is increasing, but the Hubble constant is a dot over a, scale factor derivative over scale factor.

1:36:53.9 SC: So, if both a dot and a are increasing at the same rate, then the Hubble constant is going to be constant. And indeed that is where we're going. That's why I said the acceleration rate is mostly constant. As we empty out matter and radiation from the universe and are left with nothing but cosmological constant, we asymptote to a condition where the Hubble constant does really become a constant. But that doesn't really mean a constant velocity. A Hubble constant being constant means a dot over a is constant, time derivative of a divided by a is constant. So time derivative of a is proportional to a, and I can solve that equation. It's an exponential. It's e to the t. In fact, it's e to the h, not t, where h is a certain fiducial value of the Hubble constant.

1:37:41.4 SC: So, Hubble constant means exponential expansion, which indeed is a curve that looks like it's accelerating. So all of this is just very confusing because we choose to use vocabulary that resembles velocities and accelerations to describe the expansion of the universe, where that vocabulary really isn't legitimate at all. Sorry about that. I like to tell you that's just what the conventions are.

1:38:11.4 SC: Okay, I'm gonna group two questions together. Ryan Sage says, "I was hoping you might clarify the debate about the recent Nobel Prize in Physics regarding local realism. The anti-physicalists, not Philip Goff surprisingly, but Hoffman, Kastrup, et cetera, appear to be jumping all over this as case closed, it's over, but even the recipients of the prize appear to still hold to realism or think that their submission is not evidence against it in a general sense."

1:38:38.3 SC: And then Jace Forbs says, "In your book, Something Deeply Hidden, you talk about locality in such interesting ways. And I'm curious if you have, since writing the book, have anything to expand on regarding your ideas of locality and how to think about what it is." There's a typo there, but that's okay, "What it is."

1:38:57.7 SC: So both of these questions have to do with locality in quantum mechanics, and in particular, the idea of local realism. It's a little weird. The fact that people like Don Hoffman, Bernardo Kastrup, etcetera, are celebrating the Nobel Prize as case closed for local realism is just bizarre. IT kinda give the game away that they're not really serious because this is just... The Nobel Prize is great, the discoverer of these bell equality violations that are predicted by John Bell are just predictions of the Schrodinger equation. They're quantum mechanics as it was written down in the 1920s. There's nothing new that has been done to quantum mechanics; it's only testing certain predictions of quantum mechanics and showing that they're right.

1:39:45.3 SC: So, what we learn from these experiments, which were absolutely tour de forces of experimental physics is that quantum mechanics is right. That's good. I'm very glad to hear that. But if I already was a realist about quantum mechanics and thought it was right, then nothing in these experiments changed my mind. I'm a realist about the wave function in particular... About the quantum state, I should say, the state vector. Other people are realist about other things, but all of these people have already priced in the Bell inequalities. Whatever your particular favorite version of quantum mechanics is, it better be compatible with Bell's theorem, or otherwise, you would have given it up long before these experimental results came in. So I don't think that there's anything new that happened with these Nobel Prize... I mean, it's a very worthy Nobel Prize, but we already thought that these results were gonna be there. So if you've somehow changed your mind about anything because of these results, then your mind was not correctly settled in the first place.

1:40:48.8 SC: Now, as Jace gets to, the question of locality is still super interesting in quantum mechanics. My take on it is different and idiosyncratic, so I try to be clear about my beliefs versus the conventional wisdom here. My belief about locality is not the conventional wisdom. The conventional wisdom in quantum foundations goes something like this: If you have the Schrodinger equation, or if you have what you might call a unitary evolution of the quantum state according to the Schrodinger equation, that evolution follows local laws of physics, the Hamiltonian, the thing that makes the wave function go, the thing that encodes what the wave function is doing, is local in a very specific mathematical sense.

1:41:30.7 SC: But then you have measurements. Quantum measurements famously do not seem to be described by the Schrodinger equation. An Everettian and says, secretly they are, but you have to do extra work. But anyway, they don't seem to be, at the most naive point of view. So you can have locality of the dynamics according to the Schrodinger equation, as you do in quantum field theory, but then non-local apparent correlations in measurement outcomes, and that's what the Bell inequalities point to. So, it's interesting, because people who work on quantum foundations care a lot about the measurements and bang on about the non-locality in quantum mechanics. People who do particle physics and quantum field theory mostly care about the Hamiltonian and the Schrodinger equation, and they are very proud of the fact that physics is perfectly local. And they're both right, 'cause they're talking about two different things.

1:42:21.8 SC: My attitude, which is very, very different, says, "Look, if the primary object of your theory is a vector in Hilbert space, your question should not be, 'How in the world can we get non-local measurement correlations even though space and time have these properties?' Your question should be, 'Why does physics ever look local at all?'" [chuckle] The fundamental starting point of a vector in Hilbert space has no idea of space in it, much less a locality. So the interesting question is not like, "Oh, my goodness, how can you possibly do what Bell's inequality says you should do in these measurements?" That's easy, that's obvious, that follows directly from the reality of the wave vector. The hard part is, why does the Hamiltonian have the features that physics looks local when you're not making measurements?

1:43:15.2 SC: I have some ideas about that, but the short answer is nobody knows. Right now, it's just a fact. It's not incompatible. You can just say that's a fact. The Hamiltonian is local in some very well-defined sense; measurements are not. That has no relationship whatsoever to whether you are a realist.

1:43:33.5 SC: Henry Goldstein says, "What physical mechanisms lead to complexity emerging out of the whole universe? The requirement of energy flow by non-equilibrium thermodynamics seems critical, but how and why do molecules then emerge from atoms, cells from molecules, and ultimately life?"

1:43:48.9 SC: Well, yeah, I don't know. If I knew that, [chuckle] I'd be rich, or at least I'd have a couple more publications than I do right now. I'm thinking about exactly this question. I think it's a fascinating question. I wrote a paper with former Mindscape guest, Scott Aronson, and others about this really, very preliminary, tiny step towards answering this question. But the short answer is we don't know, so I'm trying to figure out what are the ingredients in the laws of physics that make it possible that complex structures come into existence. Right now, it's completely compatible with everything we know about the laws of physics, but we can tell you which aspects of the laws of physics would still allow for complexity to emerge if they weren't there. What are the necessary ingredients that go into this? We're still thinking about that. We don't even agree on the definition of complexity, so, this might take some time.

1:44:41.1 SC: Claudio says, "The standard story about the non-locality of entanglement says that two particles become entangled in the lab, and one can be moved as far away from the other is one wants and remain entangled. Is the opposite also true? If distance doesn't count, are particles capable of becoming entangled with particles that are far away and not in their immediate surroundings?"

1:45:00.6 SC: This is a great question. I should have put it next to the other ones about locality and entanglement, but it is a great question. The short answer is no, the opposite is not also true. Because that becoming entangled is a physical interaction. So, if I have two photons and I move them close to each other, they interact and an become entangled with each other. Or even two billiard balls, if I ignore the fact that the classical limit is pretty good and describe the billiard balls using quantum mechanics when they bounce off each other, they would become entangled with each other. But that's the local part of physics. That's not a measurement. The word "measurement" never appeared there. That's just evolution of the wave function of all of these particles. And so, for the particles to interact and therefore become entangled, they need to be close to each other. They need to be able to interact. Or slightly far away, but some field is bouncing back and forth between them, which essentially is the same as saying they're close to each other.

1:46:00.2 SC: Whereas measurements can happen no matter where the particles are. So the point is you become entangled when you're nearby, then you move far away, or not, that's up to you, and then when you can make a measurement, you're gonna get some correlations because you were entangled.

1:46:18.3 SC: Roland Weber says, "Could you please give us an update on Puck, the outdoor cat you were considering to adopt?"

1:46:24.0 SC: Yes, I'm glad you asked this, Roland. I'm always happy to talk about Puck or the other cats. So for those of you who are not around, back last fall, I think September or October, Jennifer and I noticed a stray kitten out on our backyard, not much of a kitten, maybe six months old, something like that. And being the cat lovers that we are, we instantly felt it was our responsibility to take care of this little cat. And we named him or her Puck, we didn't know whether it was him or her. Puck seems like a fairly androgynous name and also follows the Shakespearean team of Ariel and Caliban, even though it's from a different play. And actually, if you're fans of The Sandman comics, Puck, the Shakespearean character appears in there, and there's a slight resemblance between the Sandman Puck and this little kitten Puck, so we named them Puck. We think now that it's probably a he, so I will refer to Puck as he, for the simple reason that Puck has not gotten pregnant. And by this time, several months later, most female cats would have gotten pregnant by now.

1:47:32.6 SC: But anyway, we thought that it was our responsibility to take care of Puck, so we started leaving out food and water and shelter and things like that. And Puck is a smart kitten. Puck figured out very quickly that we were the food providers and began to hang out. But this was September or October, and a six months old cat roughly speaking meant that Puck had never experienced winter, in our best estimation. And so we got very worried the Puck was gonna either be hurt or just be uncomfortable in the cold outside. There was really no option to bringing Puck inside with us, because Ariel and Caliban are not sociable creatures. Even a few years ago when we had little foster kittens who were the most adorable, lovable balls of fluff, Ariel and Caliban had no truck with these other kittens. They did not want other animals to be in their house. So the idea of bringing in a feral into their house was not on the table.

1:48:34.5 SC: So what we did was we do now live in a big grownup East Coast house, so we now have a basement that we've never had before in our lives. So we started coaxing Puck to come into the basement by leaving his food closer to the basement, and then taking some time and being in the basement, keeping the door open, and putting the food inside. And then we actually... [chuckle] This sounds quite ridiculous to people who are not cat people, but we had a cat door put into our basement door so a cat can come in, and we don't need to be there; we can just leave the food there, and the cat can warm himself up. And indeed, Puck is smart enough to instantly have figured out the cat door. I was very worried that Puck would not figure out how to work the cat door, how to just squeeze his way in, but he figured that out right away.

1:49:21.6 SC: So now, we're in a very happy equilibrium where we leave food out for Puck in the basement, he comes by, I think roughly a few times a day. He doesn't hang out. He's not interested in sleeping in the basement or hanging... Even in the coldest days that were certainly below freezing, he seemed to spend most of his time outdoors. He didn't love the snow, that's true, but he certainly... Let's put it this way. He was very skinny when we first met him. He has chonked up quite a bit now because we fed him a lot of food. For all we know, we're not even the only people on the block who are feeding him food. The good news is our block is sufficiently large and sufficiently grassed over and treed over that there's no reason for him to ever cross the street. So, at some point, we're gonna have to trap him and get him shots and get him neutered and things like that, but we haven't done that yet. We just wanna keep him alive during the winter. And honestly, right now, Puck's leading his best life. He's having a grand old time. He loves running around outside.

1:50:22.1 SC: It used to be like he would... He's very cautious, I think that's why he's lived this long. So he would stick very close by to the fence or to the house or whatever. And nowadays he just saunters right out on the lawn or on the driveway, just like he owns the place. So, he has a big winter coat in addition to his extra chonkiness, so, he's kept warm against the elements, he gets to run around, play with the birdies, et cetera. I think that Puck is having a great time.

1:50:47.8 SC: Dan Inch says, "Would the Many-Worlds interpretation of quantum mechanics be affected much if we knew the answer to quantum gravity? Also is Ariel getting a nice shower every morning?"

1:50:58.1 SC: So this is another cat question as well as quantum mechanics question. Those are the best questions that we can get. That's when you're allowed to have two questions in one AMA salvo there. So, for Ariel, since we're in the cat theme, again, people who've listened to a long time know that Ariel, our female cat, liked to get showers in the morning, which meant that in addition... The human would go to the shower, Ariel didn't like that, that was too much water. But Ariel would like it when you would just make a little drip of a few drops at a time coming down from the shower faucet, and she would sit under there, let it drip off of her, groom herself, drink the water, and the whole bit.

1:51:38.1 SC: That lasted through when we moved from LA to Baltimore. For a while, we were staying in an apartment while our house was not yet free for us to occupy, and she discovered the shower in the new place, et cetera. But for some reason mysterious to me, she is less interested now that we're in this house. She does not like showers. She did occasionally walk into the shower and we did put on the drips, it just wasn't quite to her liking. She's a very particular cat, so we've given up on that. Now she drinks from the sink, so we have to turn on the sink and she'll drink from that, but she's not actually getting, what? I don't know, cats, I cannot predict their behavior very well. It's almost as if they have free will or something.

1:52:20.1 SC: About quantum gravity, I do not think that the right way to think about it is to say Many-Worlds would be affected if we knew quantum gravity. I think it's the other way around. I think that, I suspect, and this is just a vague hope, this is not a very strong feeling, but I suspect that we will get insights into quantum gravity from taking Many-Worlds seriously, because Many-Worlds, in philosophy jargon, has a certain kind of ontology. It has a certain kind of notion of what is fundamentally real that provides a very good starting point for re-conceptualizing what we mean by space-time and quantum gravity and things like that. That's something I've talked about. I think that there's a solo podcast back there, if you look far enough on how space-time maybe emerges from quantum gravity, from quantum mechanics, rather. And that particular approach really is at least inspired by, if not completely dependent on, Many-worlds as an interpretation of quantum mechanics.

1:53:14.5 SC: For the simple reason that every other interpretation that I know of takes space or space-time or something equivalent to that as a primitive object. It puts it that into the theory, and then it turns out that's difficult to quantize that kind of theory, whereas Many-Worlds doesn't take space as a fundamental objects. It takes the wave function as a fundamental object, and space is supposed to emerge from that. If it does, if you Google "how the universe emerges from the wave function," you'll find talks online by me where I go into some of the technical details about that, if you're interested.

1:53:53.9 SC: [1:53:53.9] ____ says, "What do you think of the medicalization of neurodiversion traits? Research suggest that up to 15% to 20% of the US population is neurodivergent. That fraction, is it still reasonable to say that they are divergent as opposed to the norm being significantly misrepresentative of the population?"

1:54:12.4 SC: This is a very good question. I have kind of philosophical perspectives on it, but what I don't have is an educated medical or psychiatric perspective on it, so, don't take my view too seriously here. What I think, thinking about it philosophically, is that these categories of neurotypical, neurodivergent, et cetera, clearly are socially constructed. They're clearly invented by human beings. Go back to the Sally Haslanger episode where we talked about the social construction of reality. These are categories we invent. That doesn't mean they're not real. You can be real and socially constructed, but we have invented these categories to help us understand and account for and explain the world that we see.

1:54:53.2 SC: And as we learn more and gather more data and are just more careful, we can adjust the meaning of those terms to fit the data better. And so, I absolutely think there's a temptation to make the mistake of coming up with a label, attaching it to something, and then mistaking that label for reality, rather than wondering how well the label actually fits on the reality. So I think what happens with things like neurodiversions, and for that matter, different sexual behaviors or different social behaviors or a million different kinds of human-scale things, there is a norm, and that norm might be literally the middle, but it might also just be what the dominant group likes and we tend to sort of measure everything else with respect to that norm. And then other things look divergent.

1:55:43.8 SC: And then it might take a lot of work, a lot of work mentally and scientifically and philosophically to say, "Actually, no, we shouldn't treat that as the norm and other things as diversion; we should just look at the different things that exist and talk about them for their own sake rather than labeling them as typical or divergent." It's not quite on the specific example of neurodiversions, but we talked about this with Joseph Henrich, because remember, he's the psychologist who studies weird populations, Western-educated, industrialized, rich, democratic. And an enormous amount of psychology research has been done on these people. And not only are these weird people not average, not only are they not the majority, they're not even typical. They're not even in the middle of most distributions. They're on the extreme of some distributions. So, you get a very wrong opinion about human psychology by taking them as the norm, which professional psychology has done for many years.

1:56:43.8 SC: So I think that as a general principle, even though I'm not an expert on neurotypicality, neurodivergence, and the differences thereof, I think as a general principle, we are too quick to label some things as typical, some things as diversion, and we should be very careful about doing that.

1:57:00.6 SC: Jay Peski says, "Is complex system theory something you hope to truly dig into? I was curious about what are the main aspects that draw you to it. Been loving the guests and seeing how the work is applied to so many different fields' topics is exciting."

1:57:14.7 SC: So, yeah, I'm trying my best to dig into it in a more serious way. It's a little bit different than being a student. When you're a student, there's a couple of things that are going on. One is you have an advisor, so you have someone to sort of guide you and pick the research programs that you're working on. And also, you kind of have time to learn new things, 'cause that's your job, to learn new things when you're a student. Whereas as a professor, as an old person, let's put it that way, there's a lot of things going on. You have to be an advisor to students and to post-docs and make things happen and write books and do podcasts and whatever, and as well as doing the research that is ongoing and trying to learn new things. So it's harder to focus, honestly. But yes, I'm trying to, bit by bit, learn the basics, I'm hoping to teach not this year, not this upcoming year, '24-'25, but the next year 2526, hoping to teach a course in complex systems at Johns Hopkins, and hopefully by then, I will learn it. If not, I will have to learn to in the process of teaching the course. And I have lots of ideas about questions I wanna ask within that field, so we'll see if I can make any progress.

1:58:21.1 SC: Tyler Whitmer says, "You've helped me get to a place of understanding fundamental physics and the philosophy of physics way better than your average person, but obviously nowhere near the level necessary to be a working physicist or philosopher of science. Do you think there are societal and/or scientific benefits to having more of the population at that intermediate level of understanding beyond just the fun of learning for the individual?"

1:58:46.2 SC: I think that there are, yes, but I am kind of all about the fun of learning for the individual, honestly. That's my primary motivation. I think that... Look, there are things we'd like to do. I'd like to have this podcast. I like to do lots of things, so, having the podcast is sometimes a time-suck, but all else being equal, I enjoy doing this podcast. I could pretend that I was doing this podcast as a gift to the world, as a public service. I'm spreading the word and educating people, et cetera, et cetera, and so forth, but honestly, I wouldn't do it if I hated it, if I actually actively disliked it. And likewise, for more generally publicizing science and things like that. I do think that getting more people to a better level of appreciation and understanding of physics and philosophy would make the world a better place.

1:59:42.3 SC: But my evidence for that is not super-duper strong. I can't honestly say that that's perfectly clear. What is perfectly clear is that I enjoy doing it, I get reward from doing it, and when individual people come to understand things better, in part because of resources that I have given them, that makes me feel good. So I'm self-centered enough, selfish enough, that that is what keeps me doing this kind of thing.

2:00:08.2 SC: Leo Beji says, "You may be familiar with the story of FDC Willard, the cat who co-authored a physics paper with his owner, the physicist Jack Hetherington. Hetherington had apparently typed the whole paper using "we" instead of "I" and needed a second author. After the publication of the paper, FDC Willard was invited to join the faculty at Michigan State University full-time. Now that FDC Willard has broken the glass ceiling for cats in physics, who do you think would be more amenable to co-authoring a paper with you, Ariel or Caliban?"

2:00:37.3 SC: I like all the cat questions this time around. So I think, as a footnote, as a friendly amendment to this question, I don't think it's true. The story about the cat being co-author. It's because in physics writing, in physics technical professional publications, there is ambiguity about single-author papers. Some of them are written in the first person singular, some of them are still written the first person plural. And the plural is very common, even among papers that only have one author. But some journals are different and say, "No, if you have only one author, it should be first person singular." And so yes, apparently, this physicist did write a paper in the first person plural, even though there was one author. The journal complained, so he just added his cat as a co-author kind of as a joke.

2:01:30.8 SC: I don't think that the cat was ever offered a job to join the faculty at Michigan State University. That is completely unlikely. No one will be offered a job by people they never met. So, you need to have more than one publication. You need to actually do interviews and things like that to get a job offer as a faculty member. It was apparently said that Willard was invited to give a talk. That, I believe, because sometimes you will say, "Oh, this paper was very good. We've already heard from Hetherington, so let's hear from their co-author." That makes perfect sense to me, but not being offered a job.

2:02:06.8 SC: As far as Ariel and Caliban are concerned, it's absolutely Ariel who would be a co-author on a physics paper. And as I said before, Caliban lives in the moment. He's present-oriented. I don't think that his individual imaginative horizon stretches more than a few seconds into the future. Is Caliban happy? Then he's happy. Is he hungry? Then he's hungry. That's all Caliban thoughts stretch to. Whereas Ariel definitely contemplates different possible future. She's like, "What would happen if I jumped up here?" And the usual cat behavior where they ask you to open the door and you open the door, and then they sit there going, "Hmm, should I walk through this door or not?" So I think that even though it's not quite the same, Ariel is closer to having the imaginative capacities necessary to actually think about writing physics papers.

2:02:54.4 SC: Paul Haas says, "In response to a question from Brian Keating in October 2022, you once sent a message advising your younger self to be more proactive in shaping your education to pursue your true interests early. Do you worry that if your younger self had taken the advice for a more formal, well-planned educational path, you might have lost in uniqueness whatever you gained in pure educational quality?"

2:03:19.1 SC: So, maybe to clarify, I don't know if I read that as clearly as I could have, Brian Keating as a cosmologist who lives in San Diego, who also has a podcast, and I was on his podcast and he asked me what advice I would give my younger self. I did not actually send the message to my younger self. We don't know how to do that. We don't know how to travel in time, et cetera, so, that was a thought experiment. And the advice was not to have a more formal, well-planned educational path, but in fact, almost the opposite. Maybe because I did not grow up in an academic environment, didn't have role models who were academics or anything like that, but also absolutely in part because of my own personality, the way that I pursued education, once I figured out I wanted to be a physicist, was pretty conventional.

2:04:08.2 SC: I was pretty like, "Okay, what am I supposed to do next? I'm gonna do that." I did not take as much time as I might have to step back and contemplate alternatives. And I don't mean I'll turn to like go to medical school or go to Wall Street or whatever, but even within theoretical physics, like, what research to do? What topics to focus on? I took the opportunities that were right in front of me and easy to take, because Puck becoming a little overweight because he doesn't know as a feral cat whether the food will continue to appear, as someone who didn't grow up with a lot of academics, I didn't know whether research opportunities would continue to appear, so I took the ones that were immediately available to me. And what I would like to have done is to have really thought deeply about what are the most interesting possible topics to work on that I could, in principle, make contributions to. I think that's always a good thing to think. I was just late to figuring it out. And I think I'm doing it now, but I don't think I was doing it when I was 25 years old to the extent that I could have.

2:05:10.5 SC: So, no, I don't think I would have lost in uniqueness whatever I gained in pure educational quality. I think the opposite: I would have gained more uniqueness by taking a step back and thinking more about all the possible ways of doing good research, rather than just doing what I thought I could do quickly and efficiently and productively.

2:05:32.6 SC: Robert [2:05:33.3] ____ says a following thought experiment. "Satan comes to Earth," the typical thought experiment, "and gives you two options: Either you flip a fair coin, and if it falls heads, Earth is destroyed, and if it falls tails, Earth gets unlimited energy, diseases are cured, everyone's happy, and so on. However, there's an alternative: You can measure the spin of an electron... " Sorry, let me state that with the correct emphasis. "You can measure the spin of an electron instead of tossing the coin with the same outcomes: Spin is up, Earth is destroyed; spin is down, Earth gets unlimited energy, diseases are cured, and so on. With 50/50 probabilities for the outcome. You are forced to do one of these two, either the coin or the spin measurement. What would you do?"

2:06:18.4 SC: I love this thought experiment. It's a great thought experiment, because it's putting legitimate pressure on a perspective that I have advocated in the past, which is that in Everettian quantum mechanics, where there really are two alternatives that become equally real, you should treat it, you should deal with it, you should act exactly as if you were going to see actual stochastic events with real probabilities, not just you are going to apparently see probabilities, 'cause you don't know which of the branches you're going to end up on. So, if I am true to my previously stated beliefs, I would say that both of those alternatives are the same. The alternative of destroying the Earth with a 50% chance, and having 100% chance of destroying the Earth on one branch of the wave function and zero percent chance on the other should be treated equally. And this is a good... This is what Dan Dennett would call an intuition pump, because this is saying, "Okay, I have made the apparent differences between these two scenarios as vivid as possible by literally having it not just be some minor reward that I calculate the expectation value of, but literally destroying all life on Earth." Are you really gonna have the courage of your convictions and stick by it?

2:07:35.4 SC: After thinking about it, I think, yes, I would indeed have the courage of my convictions and stick by it. I think the intuition here is pushing you in the following way: You think to yourself, in the quantum mechanics case, one Earth will survive on one branch of the wave function, whereas one will surely end, and that somehow seems better than a single Earth having a 50/50 chance of surviving or not. Okay. And I get that intuition, but let's modify the thought experiment by just a little bit. Let's imagine that you flip the coin... Sorry, you measure the spin, and then time passes. Let's say 10 years pass after you do the spin measurement, and Satan doesn't tell anyone what the spin measurement was, but Satan knows what the spin measurement outcome was, and then on the... What is it? The spin-up branch, Satan destroys the Earth 10 years later. So what it seems like to the people in that branch is that they've just been going around on their business, and now the Earth is destroyed. And that seems bad. Now, and the fact that there's another branch of the wave function where they're getting unlimited energy seems like cold comfort to them on this branch where you literally destroyed the Earth.

2:08:54.3 SC: So, I think that this is a case where our intuitions are being pumped but are just still not very good. [chuckle] It's not a super relevant case to real-world questions, but I do appreciate that we should look at extreme... If we think we have some philosophical principles, we should put them to the test in extreme circumstances. And I think that if I try to come up with the right thing to do in this situation, I honestly cannot come up with a better procedure than treating the two branches of the way function with 50/50 weights in exactly the same way that I treat the truly stochastic thing for the Earth. The truly stochastic thing leads you to think, "Well, there's a real chance that the whole Earth gets destroyed, that's super bad." But there's also a real chance that no one gets destroyed, and in the other procedure when it's quantum mechanics and both worlds are real, there's a 100% chance that the whole Earth gets destroyed on one branch of the wave function. So I think that our intuitions are not serving us well in this particular case. But again, I admit, this is one of those things that I'm willing to change my mind about under more pressure, if the pressure really pushes me in some direction.

2:10:16.8 SC: Oh yeah, the other thing I want to say about this was, that's my particular view, but there are absolutely legitimate alternatives. Go back to the podcast we did with Lara Buchak where we talked about risk and rationality, and Lara makes the case that it is entirely rational to factor risks into your... To be risk-averse, but be rational at the same time. In other words, the rational thing to do is not just to calculate the expected utility and maximize it; it's perfectly okay to bias yourself in favor of less risky outcomes. That does not conflict with the basic principles of rationality. And as I said live in real-time on that podcast, thinking that way might, I think, legitimately lead you to believe that you should act differently in Many-Worlds than in a stochastic world.

2:11:14.8 SC: So, I think maybe the combination of those two things, believing in Many-worlds and believing in some sort of risk-weighted version of rationality, might actually give you different outcomes than a truly stochastic world. I'm very open to that possibility, though I don't really know for sure. I haven't really had a chance to think about it myself. Okay, very good question.

2:11:36.3 SC: Chris says, I was wondering if you're many years been looking at life to lens of physics and your vast knowledge about it affects how you listen to music. Do you feel you have more thoughts than the average person about things like waves or the neuroscience of how we interpret sounds by listening to music?"

2:11:50.8 SC: The short answer is no, because I'm just not that kind of physicist. Waves, sound, things like that, these are very, very important for physics, but it's not centrally important to my particular kind of physics. I do think that I have different questions about music, not specifically because I'm a physicist, but because I'm a scientist. And it's not about the science of how we perceive sound or anything like that; what I'm most interested in is music theory. Music theory is very well-developed. We have theories about why certain rhythms work well, certain chord progressions work well, why resolution is very important, why certain scales sound good and how they evoke different moods. I haven't seen very convincing discussions of music theory in the sense of deriving it from fundamental physics. I would like to know, not just that major keys sound kind of cheerful and minor keys sound kind of sad; I wanna know why that's true in terms of the mathematics of what notes are generally found in the chords of major scales and minor scales.

2:13:00.8 SC: I don't know. There's probably people who've done that, but I haven't found that literature, so I'm curious about that, it's not quite because I'm a physicist, but because I'm a scientist, I do have those questions that I would like to know the answer to.

2:13:14.1 SC: David Summers says, "I know you were a consultant for the physics on Avengers: Endgame, but there were no speedsters in that movie. If you were approached to give a plausible-sounding physics explanation to an origin story of The Flash or any other speedster, how would you go about it? Is there anything that could be done with relativity that wouldn't have Einstein turning in his grave?"

2:13:33.8 SC: I think it's very hard. I think that The Flash, et cetera, pose a real challenge to people trying to make physical sense of things, more so than most super heroes, more so than flying. Flying, I can kind of make sense of. But the problem with The Flash is, how do you go fast? You run. How do you run? You move your legs, you push them against the ground. So really going fast is just a matter of having stronger legs. There shouldn't be any logical distinction between being able to go fast and being strong at the end of the day? But if I take my own advice to physics consultants on movies, you have to imagine that this world is different than the real world. The world with Flash in it is different. So, in the Flash's world, moving fast is not just a matter of having stronger leg muscles; it's a different kind of thing. What kind of thing is it? I don't know, I have not really thought about that. You would have to pay me to think about that. It's not something that is an easy answer.

2:14:38.1 SC: Jim Kakalios, by the way, is a friend of mine, a physicist, who wrote a whole book on the physics of superheroes. I'm sure that he's written about The Flash. I would pick up that book and check it out there.

2:14:49.8 SC: P. Walder says, "Fine-tuning arguments are presented prominently in Philip Goff's latest book. In the past, you've acknowledge the fine-tuning is the best available argument for God, but at the same time, you've indicated that it's a very bad argument. Can you explain what, in your view, makes it a bad argument?"

2:15:04.3 SC: Yeah, very quickly, there's a few things that make it a bad argument. The primary one is one that we already mentioned earlier in the podcast, just that the idea of God is not well-defined. The fine-tuning argument only makes sense if you can say that you're being a good Bayesian, there's a likelihood function that the world looks the way it does if God exists and the world looks a way it does if God doesn't exist. I think that that likelihood function, the world looks a certain way if God exists, is just completely ill-defined. I have no idea what that likelihood function is. We know what the world does look like, and everyone, ex post facto, says, "Oh, yes. This is exactly how God would have wanted it." I never see a careful derivation that is convincing from first principles to say that God would have tuned the universe in a certain way. But that's not specifically an argument about fine-tuning; it's just working about the very notion of God, forget even about the existence of God.

2:16:00.4 SC: I think when it comes to the fine-tuning argument in particular, there's a bunch of things going on. One is we don't know the extent to which the universe is finely-tuned for the existence of life. We sometimes pretend that we do. We say, "Ah, if the laws of physics were very different, I can't imagine how life could exist." But we just don't really know. We don't have a measure on the space of the laws of physics, we don't know the conditions that would lead to the existence of something we recognize as life, and so on and so forth. So, at the level of a philosophically careful, rigorous argument, this is not that. Fine-tuning just doesn't rise to that standard.

2:16:38.5 SC: I think, more importantly, if anything, [chuckle] the apparent fine-tuning, if you put aside that issue and say, "But I think it's fine-tuned, I think that most sets of laws of physics would not allow for the existence of life, the ones in our universe do," if anything, that's an argument against the existence of God, not for the existence of God. And the reason why is because the very common problem with discussions of God is that people say God can do anything, God is omnipotent, omnipowerful, and then don't actually take that seriously. Because the idea fine-tuning is, if let's say, we say, if the neutron were a little bit lighter than the proton, the neutrons would be the ground state matter, protons would decay away, we would have no atoms and therefore no life.

2:17:33.6 SC: Well, if you believe in God, you could have life anyway, even if the universe were nothing but neutrons, because God can do anything. The only idea that would prohibit life from existing in a universe with nothing but neutrons is naturalism. If you believe that life is nothing more or less than some configuration of physical stuff, then maybe you have a chance of making an argument, that without fine-tuned constants, we would not be able to have life. But if God exists, we can have life no matter what matter is doing. We can have individual neutrons be alive. God could do that. God can attach your soul to a neutron. No problem. The only time you need finely-tuned physics is if God doesn't exist. So I think that the argument that God must exist because physics is finely-tuned is precisely backwards.

2:18:34.3 SC: Not to mention, of course, we have plenty of plausible, even though not necessarily true, physics mechanisms that can account for the fine-tuning. Whether it's the cosmological multiverse or cosmological natural selection or something, we don't know enough about the origin of the fundamental laws of physics to say we cannot explain those apparent fine-tunings through good old physics mechanisms. So if I were religious, I would not rely on that particular argument very strongly.

2:19:02.5 SC: Yuha, I'm sorry, I'm not gonna get your name pronounced correctly, [2:19:07.1] ____, I think, [2:19:07.6] ____ say, "In a Doctor Who special last year, the Tardis," or I suppose I should just say Tardis, "ended up in a part of the universe that light and manner had not yet reached. The physical rules and constant seemed to be similar to ours. The lifeforms and space ships kept their form and functionality. I've never heard anybody speculating about space without photons and matter. Would current reasonably serious theories allow a space without photons and matter to have come to exist?"

2:19:35.0 SC: Yeah. Absolutely, and in fact, it's a little bit tricky, because we do think that empty space has fields in it, and those fields have quantum states. So, in that sense, there's always something, even in empty space. But we can imagine those quantum states being in what we call the vacuum state, the lowest energy state, the state that is the most analogous to what we would classically think of is just empty space. And indeed, that's not only plausible or conceivable; it's gonna happen. If you wait long enough, the universe is gonna approach the vacuum state that is corresponding to whatever value of the vacuum energy that we have. So, yeah, you don't need stuff in the universe for space and time to exist. You can just have space and time.

2:20:23.2 SC: Subhendu Harsh says, "Elon Musk recently confirmed that Neuralink has been implanted in the first human being. Today, we make advancements in tech at all costs and by any means, and for the most part, that has served us well, but is there a threshold after which we should begin to question this unbounded progress? Is there a need to limit or at least democratize advancement in tech?"

2:20:44.0 SC: Well, just again, a footnote here, there's no confirmation in the Neuralink has done this there's a tweet that says Neuralink has done this. The tweet was not peer-viewed. We would like our standards to be a little bit higher for scientific accomplishments than that. And the other footnote, of course, is that Neuralink is way behind. There are other companies that are much more advanced in the field of brain/computer interfaces that have done a lot both, invasive and non-invasive, to interface brains with computers. So this is an ongoing field that is much, much bigger than that just one company.

2:21:21.8 SC: Having said that, I completely sympathize with the thrust of your question, that we should be very, very careful. We should absolutely think about the consequences here. It's gonna happen. I don't think that that's even a useful question to address, because like you say, when we have new technologies, we use them, that's gonna happen. But we can use them in responsible ways or less responsible ways, and I do think that the rate of technological progress is pretty fast these days, whether it's brain/computer interfaces or genetic engineering, synthetic biology, AI, what have you. It's very hard to place rules in a sort of legal sense, put them in place in ways that allow for innovation and yet keep us safe, because consequences are often unintended. It's just very hard to see what we should allow without hurting people. But that doesn't mean we don't have the responsibility to try. I think we really should try to put rules in place, safeguards tha prevent people from being hurt, while still doing the scientific research.

2:22:30.9 SC: Anonymous asks a priority question, so we're gonna have to believe that Anonymous is true to their word, and never asks another priority question, because anonymous, you know, who knows? But Anonymous says, "What tips do you have for your first-year Physics student aiming to become a theoretical physicist?"

2:22:48.2 SC: Well, I think that... People ask me this, but my tips are pretty anodyne, they're pretty conventional. I don't think I have any super out of-the box tips. Take all the physics courses you can. Take as many math courses you can, but they are subservient to the physics courses. Physics is more important than math, if you want to be a physicist. The general thrust... And by the way, take other courses as well, take Humanities and Social Science courses. Those are important for becoming a well-rounded human being, which even theoretical physicists should strive to be. So, take your academics seriously. It's sad, but not only learn, but try to get a good grade point average as well, because guess what? Someday you're gonna be applying to grad schools. They will look at your grade point average. It's not the most important thing, but it is something they will look at. If you wanna be a successful physicist, you wanna go to grad school. You wanna go to a good one. You wanna eventually get a job. You kinda have to play the game to some level, you have to balance playing the game of academic advancement versus your own personal interests. Try to find that sweet spot where you can do both.

2:23:57.2 SC: But anyway, yeah, learn all the physics you can. Don't necessarily wait. Again, because of my personality, et cetera, I waited sometimes. I didn't think about learning this certain topic like quantum field theory, whatever, until I could take the course. There's no reason to do that. There's no reason to necessarily wait until a course is offered. You can just learn it. You can take online courses, you can read lecture notes, you could buy the books, go through them. Do the homework yourself, whether it's relativity or statistical mechanics or quantum mechanics or quantifier or whatever, learn more physics than your courses are trying to teach you.

2:24:38.5 SC: I would not put a huge emphasis on doing theoretical physics research very early. I think that research is good to do, but it's way easier as a first or second-year undergraduate to do experimental or computer-based research than it is to do real theoretical physics research. And that's okay. Just do that. You get to know what research is like. And the last thing is, go out there in situations where you're not comfortable. Go to the talks that you don't understand, if there are talks of your university. Feel what it's like to be part of that community, because that's where you're aiming to be part of. You won't feel like part of the community, there'll be people asking questions and you're like, "I have no idea what is being discussed here," but it will eventually, the socialization, will happen, and you will become a part of that, and it takes a long time and you'll be impatient, but it eventually will take place.

2:25:34.1 SC: Flame Poof asks, "The latest result from the Dark Energy Survey seems to be suggesting that a flat lambda-CDM model where the equation, the state of dark energy is slightly higher than -1, currently measured as w minus 0.8 and, is time-varying. Could this mean that the current ideas about cosmic expansion could be wrong, or that a more complex model is required?"

2:25:55.7 SC: It could, [chuckle] but I'm pretty sure that there are error bars on these measurements. Snd to the best of my current knowledge, the error bars are perfectly compatible with w equals -1. For those of you who are not slinging the lingo, w equals -1 the cosmological constant. If w is -0.8 or something like that, there is a slowly decaying dark energy. So the question is not what is the best fit value; the question is, is it compatible with the cosmological constant, at let's say, a 3 sigma level or something like that. As long as it's not more than three sigma away from the cosmological constant, I'm not gonna worry too much about the current model.

2:26:35.8 SC: Siddharta says, "In the block universe, the present does not have any privileged status over the past and future. How does it explain why I, a configuration of particles, is experiencing now writing this question and not next Monday reading your answers or last Monday experiencing no AMA?"

2:26:52.5 SC: Well, I've been asked questions like this, and I've tried to give answers. I find that, empirically, my answers are unconvincing to the people who ask the questions. So I must be missing something going on in the minds of people who are asking these questions, 'cause to me, the answer is perfectly obvious, namely, in the block universe, when you talk about I, me, myself, what are you talking about? There is a me at the moment of February 7th when I'm making this recording, 6:55 PM. There's a different me, there's a different collection of particles doing slightly different things one minute later. And there's an infinite number of these collections at different moments of time, each one of them at a single moment of time. And each one of them thinks it's now, from their perspective. I see nothing puzzling about this. Of course, each different person at that moment thinks it's that moment. What else would they think? [chuckle] There's no metaphysical mystery there to me.

2:27:52.0 SC: So, I honestly just don't know what to say about this question. If that is not clear, then there's something going on in people's minds who don't think it's clear that I have not yet quite grasped. And that's my fault, not yours. Sorry about that.

2:28:09.6 SC: Harrison Brown says, "Does e equals m c squared imply that a fully-charged battery is heavier than a spent one?"

2:28:17.1 SC: You know, look, in principle, yes. E equals m c squared says that the mass of a... Well, let's put it this way. The total energy content of a single object at rest is proportional to what we call its mass, and the proportionality constant is the speed of light squared. So that's what mass is in relativity. It's the energy content of a single isolated object at rest. So a fully-charged battery has a wee bit more energy than a discharged battery, and therefore, its mass is a little bit more. But of course, the real world is way messier than that. There is a process by which the battery gets discharged, their chemicals, their molecules and electrons moving around inside the battery. A whole bunch of other things can happen, maybe it absorbs some moisture or it heats up or something like that, and the amount of mass that is equivalent to the energy in a charged battery is so incredibly tiny that I'm sure it is a very, very sub-dominant effect compared to other things that would matter in real-world batteries. So don't expect your load to be any lighter, if you have a backpack full of discharged batteries, than one of full batteries.

2:29:34.7 SC: Philip Dobson says, "I have a question about the quantum arrow of time. Does the Schrodinger equation have time asymmetry built in, or does a simple system evolve and branch in both time directions? Is there a special quantum past hypothesis?"

2:29:47.4 SC: I'm a little bit worried about the second part of your second sentence there: Does a simple system evolve and branch in both time directions? The Schrodinger equation does not have an arrow time built in, exactly as Newton's laws do not have an arrow of time built in. It's Laplace's demon all over again. From any one quantum state, you can evolve it either forward or backward in time equally well. No arrow of time in the Schrodinger equation. But there's an arrow time in the world, because exactly like you are suggesting, there's a past hypothesis. The quantum state of the early universe was in a special kind of state where there were very few branches of the wave function, and as the universe evolves and things become entangled, the number of branches goes up. But that doesn't mean that a simple system evolves in branches in both time directions, precisely because there really is a special quantum past hypothesis that prevents that from happening.

2:30:43.8 SC: Alright, it's getting late, and I think my voice is beginning to go, so I do have several questions left, but I'm gonna have to try to give them slightly more compressed answers if we're gonna make through this. Master Work Tool says, "During the pandemic, our D & D game moved online. The players complained that the dice rolls were unfair because it used a pseudo-random number generator. So I recently wrote a dice roller that pulls entropy from a public feed out of a quantum computer, so people are making everyday scale decisions based on quantum flux. We've joked since the 2016 election that we live in the crappy parallel timeline. If I make an agreement with myself to turn this tool on until the next election when an anti-Democratic candidate wins election and turn it off when the opposite happens, could that, in principle, be used to increase the thickness of realities where democracy is more stable?'

2:31:34.9 SC: No. [chuckle] It could not. Again, I'm sorry, I'm gonna have to keep the answer is short here, but when you make quantum measurements, you're only taking the branch that you're already in and subdividing it. The total weight to the branch there already in does not grow any smaller just because you make measurements. It gets subdivided amongst more and more branches, but the total weight remains constant over time in that branch. So as long as you are not the one either aiding or undermining democracy, your random number generator has no effect on that.

2:32:13.0 SC: Hans Nuten says, "Do you consider Everettian quantum mechanics as a complete theory? I've seen you tackle both unreasonable and reasonable objections to Many-Worlds, but would you consider these objections as a need of a new theoretical framework, or are they mostly about flushing out the philosophy and consequences of our ontology of the world, and explain them in a clear way?"

2:32:30.3 SC: 100% the latter. I think that the Everettian quantum mechanics itself is complete. It is the statement that the world is entirely represented by vector in Hilbert space evolving according to the Schrodinger equation, for some Hamiltonian. That's the whole theory. There's plenty of work to be done. All of the work is taking that theory and explaining how it can account for the world we see. That is non-trivial work; it involves both physics and philosophy and maybe other ideas, but the theory is there. It's our job to understand what the theory is trying to tell us.

2:33:08.1 SC: Tomas Freeman says, "How do you navigate having friends or family who are otherwise rational or reasonable adults, but believe in astrology, human design, or other New Age woo pseudoscience?"

2:33:19.3 SC: Easy. I let them believe those things. I'd let people believe whatever they want. If they wanna force it on me, then I would begin to wonder why they are my friends, but you have to decide whether or not someone is open-minded about things. Like if they have some New Age belief but they're saying, "Here's my New Age belief, why don't you share it? I have reasons for it. What are your reasons?", and they would like to have a respectful dialogue, then have the respectful dialogue. If they simply won't listen to you or are uninterested in listening to you, then don't talk to them about it. It's actually pretty easy. You shouldn't be annoyed about what other people believe. I suspect that there is no person in the world who, if you knew all of their beliefs, would not be really annoying to you, in some way or another. I think that even your current self would be annoying to yourself 10 years from now, or 10 years ago, if you talked about each other's beliefs, because I beliefs change over time. So, it's not a matter of whether beliefs are backed up by science or whatever, it's a matter of like, are these the kinds of beliefs that people are open to talking about and maybe changing their minds about, or are they just part of who they are, and we have to decide what that means for how we think about them.

2:34:34.0 SC: A user named XLWRP090 says, "Can Bayesian reasoning account for confidence? There's a difference between describing 50% to a hypothesis due to a lack of evidence versus it being due to having lots of contradictory evidence. New evidence should have less impact in the latter case. Would a formula which factors in confidence explicitly be more useful?"

2:34:56.7 SC: I don't think there is much of a difference between those two things. If you have lots of between ascribing 50% credence to a hypothesis because you know nothing about the evidence and having a lot of contradictory evidence. If you have contradictory evidence, by construction, some of your evidence isn't very good, 'cause it contradicts the other evidence. Maybe none of your evidence is very good, but by the very idea of having lots of contradictory evidence, you should be placing very low value on all of that evidence. So, I don't think the new evidence necessarily have less impact in the latter case, unless you have reason to believe that the new evidence is just as bad as the evidence you already have, in which case, sure, it absolutely should have less impact.

2:35:42.1 SC: But that should be, if you're a good Bayesian, built into your likelihood function, built into the answer to the question, under this theory, how likely is it that we get this evidence? If you're honest about constructing your likelihood functions, you always know that there's always a probability of getting evidence that seems incompatible with your theory, just because your evidence is bad or weak, or your experiment was not designed well. That should be built into your likelihood functions. You don't need to change the basic principles of basin reasoning.

2:36:14.4 SC: Price Mitchell says, "Why do you not take the AI fast takeoff scenario seriously, I.e., rapid exponential self-improvement? I don't give it much credence myself, but I don't think I've heard any critics give a reason for doubting it beyond saying it's ridiculous or fantastical."

2:36:29.9 SC: Well, I don't think it is ridiculous or fantastical; I've just seen very little convincing evidence to take it seriously. Consistent with my stance that current large language model types of AI are, on the one hand, very impressive, but on the other hand, not modeling the world, they're gaining their successes by piggy-backing off of the successes of human speech and human writing. There is a natural limitation to that. They're gonna run out of human speech and writing to train off of. And that is exactly the kind of model which is not going to improve by self-improvement, because its objective function is not something out there in the world that can be objectively verified.

2:37:19.1 SC: If you want to build a computer to play chess or to play Go or whatever, self-improvement is absolutely the way to go. Don't let the computer be trained on human chess games; make it just play against itself many, many times, because it's clear when you're winning. You know what it means to win in a game of chess or not win a game of chess. But if the game you're playing is, "Say true things about the world," then the model itself is not able to evaluate whether it's succeeding or not, like it is able to evaluate that in chess. And empirically, once these LLMs start being trained on LLM outputs, they get worse rather than better, because they're losing touch with the human texts that actually gave them the plausibility in the first place. So, I think that exponential self-improvement lasting for a long time, I just don't see the argument for it. Maybe it's there, I'm absolutely open to the possibility, but I see no strong argument to expect it to actually happen.

2:38:25.2 SC: Walter E. Miller says, "What are your thoughts about the gigantic galactic structures discovered in the universe, especially the big ring announced at the recent AAS meeting? Is the cosmological principle broken and it's our understanding of baryonic acoustic oscillations incorrect?"

2:38:38.8 SC: Probably not, and probably not, are the short answers. I don't know a lot of the details about the specific observation because I'm not that interested in it. And the reason that I'm interested in it is 'cause it's probably not a big deal. It could be a big deal. There's absolutely some non-zero credence that this is a big deal, but the basic picture of the Big Bang model and our basic understanding of large-scale structure and acoustic oscillations and things like that is really well-established. And some individual announcement of some observation at some meeting is just never gonna be good enough to overthrow that. Something like the discovery of the acceleration of the universe in 1998, that was a single result that made people change minds very quickly.

2:39:26.6 SC: But notice two things. Number one, there were two teams that did it, so they could check each other's work. And number two, there is instantly an obvious explanation, namely, they have discovered the cosmological constant. If you just have some wacky observation that seems not to fit with the standard model and no good explanation for it, there's no reason to take that seriously until someone else finds very strong independent confirmation of it. And I'm not impatient. I can wait. I suspect it will go away.

2:39:57.1 SC: Paul Torex says, "Tim [2:40:00.8] ____, citing John Bell, said on Robinson's podcast that when a wave function decoheres, any [2:40:04.9] ____ element will get as small as you like, leaving the separate worlds of Many-Worlds. But also, no matter how long you wait, there will be some [2:40:11.2] ____ element that is as big as you don't like." Translation for everyone else out there, what Tim is just saying is that the different worlds and Many-Worlds are not exactly perpendicular to each other. They're very, very close to perpendicular, but not quite. And so Paul's question is, "Is that true? And does it mean that when we decohere a quantum superposition, all we can say is that probably, the measurement outcome will look like a collapse?"

2:40:34.9 SC: It is true, and yes, we can say nothing more than that probable statement, but you have to attach numbers to this to have any impact on how you think about things. When I... What should I say? When I take a pot of spaghetti with all the spaghetti sauce and I'm trying to take it in to the dining room table and I trip and it all falls on the floor, there's a possibility that if I pick up the pot again, all of the spaghetti and all the sauce will just jump back into the pot. So all I can say is that probably, I will have messed up the floor, because I don't know for absolute certainty. But guess what? [chuckle] The probability of that happening is so small that I don't need to take it seriously. And it's just as bad, if not much worse, in quantum mechanics. The amount of decoherence is in enormously big, and it's completely implausible that in any realistic situation, you will get anything other than what looks like ordinary quantum mechanics.

2:41:43.1 SC: Chris B says, "Thought experiment: If you can scale yourself up to proportion whereby a planet is roughly the size of an atom, could you conceive of discovering quantum mechanics or relativity, or will the universe seem quite different?"

2:41:54.7 SC: Well, do I have the book for you, Chris. Volume II of The Biggest Ideas in the Universe, called Quanta and Fields, is coming out in May. And if you don't wanna wait for it, you can actually just look at the video I did on scale for the video series, and I explained that there's no such thing as scaling yourself up. So I'm not sure exactly what you mean by this. I presume you mean making yourself of the scale where quantum mechanical effects are obvious around you. That just doesn't quite fly, because in quantum mechanics, there's a relationship between your mass and your wavelength. The Compton wavelength of different particles depends on their mass. So, if you want to make something small, if you want to make Ant-Man small, you have a choice: Either you make his individual atoms less massive, 'cause you wanna make a Ant-Man not only smaller in length and height; you wanna make him lighter. You don't want a 200-pound Ant-Man being the size of an ant, that would lead to problems. So you need to make the individual particles that he's made of lighter.

2:43:05.5 SC: But guess what? What means their Compton wavelengths get all spread out. And that means he's not the size of an ant anymore or an atom certainly. He's all spread out all over the place. Quantum mechanics means you can't just scale things and keep their physical properties the same. So, I don't know how, maybe I shouldn't say it's impossible, but I don't know how to have a human-like thing exist in a regime where quantum mechanical phenomena are absolutely in your face and necessary, rather than existing in the classical limit. I haven't put too much thought into it, but I suspect that it's just not conceivable.

2:43:45.0 SC: Zay says, "In Descartes' Fifth Meditation, he uses the proposition that something is made better by existing to support the existence of God. Disregarding any God arguments, do you believe that something is made better by existence or exists more perfectly as a concept?"

2:43:58.6 SC: So as you might guess from an earlier answer, no, I do not think that something is made better by existence or exists more perfectly as a concept, because I don't think that the notions of better or more perfectly make any sense in these contexts. I don't know what number of betterness to assign to objects so that they have a slightly less betterness when they don't exist, [chuckle] than when they do. I think that I can be better at running a marathon, 'cause that's something quantifiable, better than another person, or typically worse than another person, 'cause I'd be terrible at running a marathon. But that's a notion with respect to which I can judge betterness or worseness. There's a quantitative number that I can compare. When it just comes to existence, something is made better by existing, you've just taken a concept like betterness that makes perfect sense in a certain quantitative realm, and you've used it in a realm where it makes no sense. So, I would not agree with that kind of reasoning.

2:45:01.5 SC: Astro-Nobel says, "According to quantum theory, the mass of the Higgs boson should be very high, of the order of the Planck mass. Nevertheless, physicists were hopeful to find the hegemon within the reach of the Large Hadron Collider, and so they did. Why did they expect what they didn't expect?"

2:45:17.6 SC: That's a very good question. I'm sure that typical ways of talking about this can be confusing. The expectations are relative to different sets of knowledge, is the short answer. If we knew nothing about the world other than there's something called the standard model of particle physics, there's something called the Higgs boson, there's something called the Planck mass, et cetera, yes, indeed, you would expect the mass of the Higgs boson to be of order of the Planck mass. That would be what particle physicists would call natural. I do think that there's some more work philosophically to be done there, but okay, that is what they mean by that.

2:45:50.2 SC: But we do know more than that. And we knew more than that long before we turned on the Large Hadron Collider. We knew, since the time of Enrico Fermi, that there is a particular mass scale characteristic of the weak interactions. And what I mean by Enrico Fermi is Fermi was the first one to have a theory of the weak interactions back in the 1930. And so he is the one who calculated the weak interaction energy scale. And then when the Higgs boson came on the scene, the sensible place to put the mass of the Higgs was not at the Planck scale anymore. Given that data input, conditionalized on the actual strength of the weak interactions, the Higgs boson mass should be near the weak interaction scale. And indeed, you can be a little bit more sophisticated than that and use other pieces of data and so forth. So we had pretty much pinned down what the Higgs boson mass had to be in the real world, but that was based on other data. That is separate from the argument that in the absence of any data, you would have expected the Higgs to be up near the Planck scale.

2:46:55.5 SC: Josh Hartley says, "Imagine that you were subpoenaed to appear in court to testify at a murder trial. The defense attorney asked you if you thought that humans had free will. There's plenty of evidence that you do not think so. You immediately realize that the defense is trying to use your scientific credentials, expertise in previous stance on free will, to prove that his or her client was not responsible for the horrific crime that had been committed. How do you respond to the defense attorney's questioning?"

2:47:18.9 SC: I respond by, I guess... I'm not sure if there is a legal phrase or not, but in my high school debate days, we would have said, "hypothesis contrary to fact." [chuckle] In this case, the hypothesis is that I do not think there is free will. That's not true. I do think that there is free will. I've said many times that I think that there is free will. I'm a compatibilist about free will. I think that the people who are anti-free will just aren't taking seriously what is the right way to think about it. I do believe that human beings obey the laws of physics, but I also believe that human beings are not Laplace's demon. So the fact that we obey laws of physics does not suffice to help us predict what's going to happen. The right way to think about human beings is as agents making decisions, and they can get praised or blamed for the decisions that they make, in my view.

2:48:13.3 SC: Valor Up says, "At what point is to become useless to talk about a theory if the theory is untestable? I struggle with this thought in the Many-Worlds interpretation of quantum mechanics."

2:48:22.2 SC: So I don't know why you would struggle in that particular case, 'cause Many-Worlds is super-duper testable. Many-Worlds is just a statement that the universe is described by a wave function, and it obeys the Schrodinger equation. That is 100% compatible with all the experimental data we have so far, and it would be instantly falsified if we get evidence that says that either there is something in addition to the wave function of the universe, or that the wave function does not always obey the Schrodinger equation. Both of these are very plausible experiments to do. If Many-Worlds had literally no impact on how we account for the world, then it would be uninteresting as a scientific theory. But as I said before, I think it might actually be very impactful when it comes to inventing better theories of emergent space-time in quantum gravity, for example. So I would say both. Make sure you understand the testability or otherwise of your theory, and also think more about other implications the theory might have for the rest of science.

2:49:22.1 SC: Dave Stern says, "Do you still have your electric car?"

2:49:25.6 SC: Yes. [chuckle] I'm only answering this question because it feels like there's a subtext here, like the expectation is that I don't, or something like that? Yeah. We still have our electric car, a little BMW i3. It's only... I don't know, it's a 2017, so I'm hoping to have it for another 10 years. It's a perfect little car for driving around the city. It's usually the car that we drive around. I love that little car.

2:49:46.6 SC: Bill Maggie says, "If Newton had known that the speed of light was invariant for all observers, do you think it's likely he could have developed special relativity or even general relativity?"

2:49:57.2 SC: Yeah, I have no idea. General relativity, no, because he didn't know differential geometry. He didn't even know non-Euclidean geometry. Special relativity, maybe, but it would have been a huge leap. Special relativity in the real world was developed by many people, not just Einstein, from Maxwell through Lorentz and Fitzgerald and Poincare and other people up to Einstein and Minkowski, who really sort of figured it out once and for all. So Newton would have had to do the work of many geniuses. But maybe he could have. He was the greatest physicist ever to do physics, so, maybe he could have done it. I'm just not sure. He would have had the necessary ingredients, I think. If he had really known that the speed of light was invariant and the principle of relativity, he probably could talk yourself into special relativity, in principle.

2:50:50.6 SC: Jeffrey Seagal says, "As a philosopher, do you find the looseness of definition of words a strength, or a liability? I find it frustrating when philosophers end up arguing about the definitions of terms and how philosophical arguments may seem to differ based on someone's interpretations of words."

2:51:06.3 SC: Well, the joke that I wrote in... I think it was in From Eternity to Here is that physicists are always frustrated by philosophers, 'cause they're always arguing over the definitions of words. Philosophers are frustrated by physicists, 'cause they're always using words without knowing what the definitions are. And I think that both of those are mistakes. It's a mistake for philosophers to imagine there is something called the right definition of a word. We're allowed to invent definitions of words. Some might be more useful; some might be less useful, but there's not out there in the world a right definition. It's there's a choice we need to make, so the right discussion to have is how should we define words, and that's a perfectly legitimate discussion to have. It's very frustrating for people, whether they're physicists or anyone else, to try to make an argument based on a definition, based on a word, and not be willing to define what the word means. I think that should be just as frustrating as spending all of your time arguing about the definitions.

2:52:12.2 SC: Dave Grundgeiger says, "It seems like we're not so certain exactly how we've come to coarse-grain things the particular way we do other than and after the fact, it works. If we were to try to write software from first principles to find useful core screenings and fundamental physics data, I don't think we'd know how to start. Are you aware of anyone who works specifically on trying to find a fundamental theory of coarse-graining?"

2:52:33.1 SC: Yes. I know people who are doing exactly that. I don't know of any great progress along those lines. There's a little bit, computer scientists do things like this all the time, which is called the latent space. If you have some variables of some very complicated system, many, many, many variables. If you analyze the data of what all those variables imply in the right way, you can find that only a subset of those variables actually matters. And that subset might not line up with the variables themselves. Maybe you have X, Y, Z, et cetera, and what matters is x squared plus y squared, and z doesn't matter, and whatever is perpendicular to x squared or y squared doesn't matter.

2:53:17.9 SC: So, that's kind of a coarse-graining. You're asking yourself, "What are the macro variables?' You're asking yourself, "Can we automate the process of finding the emergent structure?" There's no obstacle to doing that in principle, but in practice, it's much easier just look at the macroscopic thing and figure out what variables seem to be relevant.

2:53:40.4 SC: Anonymous says, Anonymous asks the question, "I get the sense you don't think there's more than a Pascal's mugging chance, a less than 10% chance, that we'll soon build superintelligent AIs which are existentially bad. If we break that down into the probability of x and y, is the probability of x times the probability of y given x, is it mostly that there's just a tiny chance we will build in superintelligent AIs any time soon, or that, given we'll build superintelligent AIs soon, there's just a tiny chance it would be existentially bad?"

2:54:13.0 SC: I'm not gonna quite be going along with the format of your question here. I think that the question is ill-posed. I would ask the same questions about the word "superintelligent" applied to AIs as I would about "perfect" or "better" applied to objects that exist, or to God. [chuckle] What do you mean by that? To just say the phrase, "superintelligent AIs" presumes there's some unitary measurable thing called intelligence. And there just isn't. There's many different capacities that people can have that's kind of correlate with each other, maybe a little bit, but not very well, and that we informally relate to intelligence, and therefore there's this strongly anthropocentric idea that we can do the same thing for AIs. But all of the evidence says that we can't. Some people would take being good at chess as a sign of intelligence, but we know there are computer programs that are better at chess than any human being and have no other intellectual capacities at all. So, I don't know what superintelligent is supposed to mean.

2:55:32.8 SC: Furthermore, as we said before, I don't think it especially matters what superintelligent means, if the thing is not an agent with interests and values, like human beings are. Not to say there's not possible dangers there. I just think that... It drives me a little batty that people can't stop thinking about AIs in the same way they think about people. They are different kinds of things. That doesn't mean they're less impressive, it doesn't mean they're not dangerous; it just means we gotta take seriously what they are, rather than putting them into the boxes that we've constructed based on our experience with human beings.

2:56:13.4 SC: David Maxwell says, "What do you think of the Bletchley Declaration and recent approaches by governments to tackle AI from a policy, regulatory, social, and economic perspective? Do you have a preferred position between the unsurprisingly more heavy-handed regulatory approach of the EU versus that of the US?"

2:56:31.1 SC: Well, I think there's a bunch of things. It is a very complicated policy question that is going on. I don't have anywhere near the expertise to give you the right answer here; I can give you vague impressions. One of the reasons... Again, I'm trying to keep the answer short, one of the reasons why I could have said more about the previous question is I think that the best way to mitigate huge existential risk kind of worries about AIs is just to concentrate on mitigating the much more realistic short-term worries, and then we'll learn more about what the actual worries are and how they play out long before we get to existential risk kind of questions. So I think that we should take a kind of incremental approach, but it should be very fast. The pace of the technological change is very fast, much faster than the time scales typically adopted by governments or regulatory agencies.

2:57:28.2 SC: I do have some sympathy for the idea that overzealous regulation can stifle innovation. But I'm not an idiot, and I know that companies who are building AIs are trying to make money, and therefore have a vested interest in not being regulated. And I would not take their word for it on the right amount of regulation that there should be. Regulatory capture is a real phenomenon, and we have to take it seriously. So, I don't know how much regulation there should be. There should be some, it should come quickly, and once it comes, it should remain nimble. Unlike banning psychedelic, banning any scientific research on psychedelic drugs and just keeping that ban in place for a long time, good regulation has to be nimble enough to keep changing in response to the changing technology. This is not something the government and bureaucracies are especially good at, so I'm not especially optimistic that whatever regulation we do get is going to be the best possible kind.

2:58:35.7 SC: John Stout says, " [2:58:36.9] ____ once said that at a fundamental level of particle physics, entropy and thermodynamics are inapplicable. Can you address this? If I find the interview then in a replied comment to this question, I will quote his exact words for clarity."

2:58:49.2 SC: Yeah. No, I think this is standard stuff. This goes back to Boltzmann. If I have Avogadro's number 10 to the 23 or whatever, particles, certain individual features of the particles don't matter. They average out. It's just like flipping the coin many, many, many times. The idea of entropy and temperature and pressure and things like that begin to apply when you have a sufficiently large number of particles that you can average over individual fluctuations. Once you get down to having one or five or 10 particles, that makes no sense. If I take the air in the room around me and I chunk it up into little cubes, and those cubes are one micrometer in size, every cube, even though it's quite small, we'll have an enormous number of atoms and molecules in it. But if I only have, in the room, five molecules, then I probably have zero atoms in my little cube, and I can't do that averaging. So, there's nothing especially wild about this; it's just a feature that things like fluid dynamics, statistical mechanics, et cetera, are emergent in the limit of large numbers of particles.

3:00:01.8 SC: Brandon Lewis says, "I'm not sure I wanna keep working in the software industry. In fact, I stepped away from it back in 2020, after a little over 10 years. After taking care of some personal affairs, I've been wondering what to do next. I'm thinking about going back to school to obtain a higher degree, perhaps in chemistry or material science. Do you have any advice?"

3:00:18.9 SC: You know, probably not. I mean, I certainly should not be the one giving advice about a chemistry or material science degree, let's put it that way, because chemistry and material science can both, in principle, lead to fulfilling careers outside of academia. My experience has all been inside of academia, and I really don't have... I need to be able to say some things about being outside of academia, 'cause I have students who go there and I try my best, but I'm not the world's expert in these things, I'm not the ones to ask. I think probably because of timing, Brandon, you asked this question before we released the most recent podcast with [3:00:58.6] ____ talking about the downside of STEM careers. So, listen to that podcast, keep any of that in mind, but otherwise, chemistry and material science are fields where you're as employable as you are in other areas. If those are the things you're passionate about, then I would absolutely be in favor of pursuing them. Don't pursue them if you don't like them, but just think they're gonna lead to good jobs. That's not the reason to do it. They could lead to good jobs, but you better like doing them before you get those jobs, if you're gonna be happy later on.

3:01:36.6 SC: Pete Faulkner says, "For some years, "I've had a layman's interest in quantum physics and cosmology and read fairly broadly across those topics. Then about five years ago, I discovered podcasts, in no small part due to Mindscape and an online course providers such as [3:01:50.4] ____. This changed my life profoundly. Now I enjoy plugging in to listen and have found that the best time for me to do that was to start going for long walks. Online learning has changed me, not just from intellectual standpoint, but by contributing to a far healthier lifestyle. Has any seemingly unrelated to innovation ever triggered such a profound positive change in your own life?"

3:02:12.2 SC: Whoo. No. [chuckle] You know, I'm almost jealous of you. I love this idea of long walks, listening to not just podcasts, but audio books or lectures or whatever, that sounds enormously rewarding and fun. It is not my lifestyle right now, partly because my day job involves exactly that kind of thing, so my leisure time is not gonna be that kind of thing. I'm much more likely to read a novel in my leisure time than I am a science book. But maybe that's not true; maybe just count the reading science books as work, but it's also a pleasure at the same time. It's a fuzzy boundary between them. But specifically, the question about, is there an innovation that triggered a profound positive change in my own life, yeah, no, not that I can really think of. I mean, does the internet count? I would think that, for me, I'm maybe in a minority where the existence of the internet has greatly improved my quality of life.

3:03:17.4 SC: It's a weird thing. I think that we, as a society, have not yet grappled with the fact that for many people, it has made their lives worse, not just the internet, but in particular handheld devices, and young people are not a good mix, as it turns out. And I don't know what we should do about that. But for me, I have a podcast, I met my wife over the internet, because we both were writing blogs. My life has been greatly improved by the existence of the internet.

3:03:48.2 SC: Mark Smith says, "How much credence you give to Neil Turok's hypothesis about applying CPT symmetry to the entire universe centered on the singularity?"

3:03:56.7 SC: So there is an idea that Neil Turok and others have pursued, that there's a bounce in the universe. There's a moment in the middle of the history of the universe near what you and I would call the Big Bang, and time sort of flows away from that middle in both directions. Which is similar in spirit to what Jennifer Chan and I proposed with our multiverse and the arrow time, but there's a couple of very important differences. One is that, as Mark alludes to, Neil is trying to address not just the cosmological history of the universe, but this underlying symmetry called CPT, which is basically roughly an advanced version of time reversal invariance applied to particle physics. And the idea is that you can sort of symmetrize the whole universe by considering both sides of this initial bouncy thing. Which is great. I mean, maybe. I don't know. I'm in favor of that. I haven't thought about it very deeply.

3:04:53.3 SC: However, the specific examples that they're looking at require a super specific delicately, finely-tuned condition at the bounce. Which is a very typical thing to need an ordinary general relativity. So, the whole point of my paper with Jenny Chan was that the bounce didn't need to be finely-tuned, that it wasn't really a bounce; it was just an eternal universe where there happened to be one moment that you would call the minimum entropy moment. But there's nothing special about that. It wasn't low entropy in any particular moment; it's just that entropy is an unbounded quantity at any time. It has some value, and there's some moment in the history when the value was the lowest it ever was, and it increases in both directions from then. But you didn't need fine-tuning to make it happen. So we're sort of interested in different questions. They're both interesting hypotheses, but they're aimed at different puzzles.

3:05:53.8 SC: Rory Cochrane says, "Re: Your Christmas immortality address. What is to stop humans or successor AIs manipulating the universe to prevent their own equilibrium De Sitter space from eventuating, thus making immortality theoretically possible?"

3:06:09.6 SC: Just because we have literally no known way to do that, given the laws of physics as we understand them. If we have new laws of physics or discover new laws of physics, maybe that will become possible. But that's not something that we even know what it would take to do, therefore, it's not something that I put on the table as a possibility.

3:06:27.7 SC: Tyler Smucker says, "What are your thoughts on homeschooling? Can I get my kids a valuable education, or will I be harming them by impeding their social development?"

3:06:37.2 S2: I think it's very, very specific to circumstances. You absolutely can get kids who are very, very well-educated through homeschooling. Malcolm McIver, previous Mindscape guest, was homeschooled through his pre-college days. It just depends on who's doing that homeschooling. I suspect that a lot of parents think they can give their kids a complete valuable education, but maybe don't know quite enough to do that effectively, if we're really honest. And of course, there is the worry about social development, if you're not actually interacting with other students. I think certainly in college or university, and almost certainly before then, the role of interacting with your peers in the ultimate development of who you become is at least as important as what you learn in classrooms. So, I don't know. I don't have kids, I shouldn't speak to that. I think that either thing can happen, I think that you should just try to be as responsible and honest with yourself as you can be.

3:07:42.1 SC: Nikola Ivanov says, "What physical mechanism drives the spontaneous symmetry breaking in gauge field theories?"

3:07:50.9 SC: Well, it's just the fact that entropy likes to increase. [chuckle] No one else would say that except for me, but it's ultimately true. Why do I say that? Because what everyone else would say is there is a field, in this case, the Higgs field, for example, that has a potential energy, and the potential energy depends on the value of the field. And let's make our lives very, very simple, let's make it that the field just has a real number for its value, so it's between minus infinity and infinity. And so there's some minimum value of energy that it wants to sit at. And if the actual potential energy has the feature that that minimum value of energy is not when the field equals zero, then in empty space, in the vacuum state, the field will have a non-zero value. And that can spontaneously break a symmetry. So it's just the fact that the field wants to roll down to its lowest energy state.

3:08:49.6 SC: Now, why do I mention entropy? Because, of course, we say things like, "Fields wanna roll down to the lowest energy state," but energy is conserved. So what does that really mean? Of course, what it really means is, entropy is higher when the field is in its lowest energy state, and its energy that it used to have when it was not in his lowest energy state has been converted into photons or other higher entropy excitations of other fields. So, ultimately, it's always 'cause entropy wants to increase.

3:09:20.4 SC: Sidhap says, "Now that you've been here for a year or more, how do you like living in Baltimore? What are its main pluses and minuses?"

3:09:27.0 SC: We really like living in Baltimore. I think the single main plus is the real estate prices. Very honestly. Having come from LA, which is a rather overheated real estate market, we get a lot more house for exactly the same price. We do not have to buy a more expensive house, but we got a much nicer house. And the location is better. We're right next to campus. I walk to work everyday. It's a 10 or 15-minute walk, depending on whether I'm going to my physics office or my philosophy office. And it's a very beautiful neighborhood with stray cats around that we can make friends with. Plus, it's the East Coast. It's a vibrant lifestyle. We're very close to other things, we can easily go to Washington, DC or Philadelphia without too much extra effort. We can go to New York or Boston or whatever. There's a feeling of connection that is easier to get here than in California, for example.

3:10:20.9 SC: But I love California. I love the weather there. Right now, the weather is terrible, of course, it's raining to beat the band, as it sometimes does in the winter time, but overall, the weather in Los Angeles is better than the weather in Baltimore. Especially these days, clearly. In the summertime, it's not the winter you worry about anymore. The summers are becoming unbearable around here, and that's going to be an issue. Air conditioning is important, we've got to upgrade our air conditioning system. And Baltimore is right on the edge of being a big enough city to keep me happy. It is a big enough city to keep me happy. There's plenty of places to go out to eat or go to music or whatever, but there's not an overabundance. If we're in LA or New York or whatever, there's just so many things, I would never be able to get to see them all. In Baltimore, I get the feeling like I can kind of go to all the places I wanna go to. It's right there on the edge. So, yeah, pluses and minuses.

3:11:15.3 SC: And of course, Baltimore is an old East Coast city that has a lot of poverty-stricken areas, and that's just bad. It's bad for the city, it's bad for the people who live there. It sort of makes you sad to see that, and I feel bad about that. I don't know what to do about it. But LA, in its own way, he has an enormous amount of poverty and homelessness, et cetera. So that's not really a difference between the two. There you go.

3:11:39.9 SC: Bob Ritchie says, "In Carlo Rovelli's Helgoland, he's a bit dismissive of Many-Worlds. In order to describe the phenomena that we observe, other mathematical elements are needed beside the wave function psi. The individual variables like x and p that we use to describe the world. The Many-Worlds interpretation does not explain them clearly. Do you care to speak to this?"

3:12:02.6 SC: Yes. Carlo should read my papers. [chuckle] I mean, I think that he's right. And look, I'm good friends with Carlo, he was like, what? My second guest ever on Mindscape. He's just kinda goofy when it comes to quantum mechanics. Most people are. That's okay. I'm friends with them, just like everyone should be friends with their New Age friends. It's absolutely an issue for Many-Worlds that things like x and p, position and momentum, are not special. They're not part of the theory. The theory just has a vector in Hilbert space and a Hamiltonian. X and p need to be emergent. But if you've been listening to Mindscape for even a little while, you know that the fact that something needs to be emergent rather than fundamental isn't gonna bother me. I'm just gonna say, "Good, let's do the work. Let's see how we can get x and p, position and momentum, to emerge out of the wave function."

3:13:00.6 SC: And rather than just saying, "Ah! It's not there. I am lost, I don't know what to do. I will invent new theories of quantum mechanics," you can sit down and do the work. You can ask, "Why and under what circumstances would I ever be driven to invent things like position and momentum out of a theory that was just a wave function evolving in time?" And guess what? You can answer that question. We've made progress toward answering that question. So far, just one paper, Ashmit Singh and I on quantum mereology, but other people have written related papers and we're making progress. It's so much more fun to ask the question and try to answer it and make progress than to just say, "Ah, I don't see how it can be done! I'm gonna just invent elements of my theory that answer the question without doing any work at all."

3:13:50.1 SC: Emmett Francis says, "I find that mathematical physics at its best can involve some surprising and clever tricks and manipulations. Do you have a story about some of your favorite mathematical jiu-jitsu you pulled off in your research?"

3:14:02.4 SC: Well, I guess it depends. I've never been... It's weird. When I was a first year graduate student, I took one of the required courses, was Applied Math. So there's an Applied Math Department at Harvard, and I guess it wasn't a required course for me, 'cause I was an Astronomy major. Physicists were supposed to take it, but it was basically like mathematical physics. Complex variables, linear algebra, things like that, series expansions, special functions, all that kind of messy stuff. And weirdly, I was really good at it. When it came to finding a clever change of variables to do an integral, I was like the best. I was awesome at that. I could really have this intuition about, "Oh, yes, make this e of the cosine data, and this integral will really kind of simplify itself."

3:14:52.7 SC: But [chuckle] having said that, I don't like it. That's not my style. It seems like it's just kind of dry and technical and yeah, solving integrals. And if any of you follow Seamus Blackley on Twitter, he is a well-known guy for having been the pioneer of the Xbox, but he was trained as a physicist, worked at Fermilab. We follow each other on Twitter and things like that. And Seamus is the kind of guy, he just loves doing a good integral. That's where he gets his joy from. And I just don't understand it. I could do it, I haven't done in a long time, but I could do it back in the day. So, the kind of math that I typically use these days is different, and it's the kind of math that... For a long time, I was just using simple math. If all you're doing is cosmology and the Friedman equation, I learned that in grad school; I just apply it over and over again.

3:15:46.1 SC: These days, when I'm doing complexity in quantum mechanics, I need to learn more math. I'm learning new math, but it's sufficiently new to me, 'cause I wouldn't say I'm doing any advanced jiu-jitsu with it. The closest maybe that I ever came... I mean, I've been involved on papers that did amazing mathematical things, but mostly, it was my friends'. If you wanna dig into my old [chuckle] papers, there were some with Wade Taylor and Miguel Ortiz on two-dimensional quantum gravity, where we pulled off some wild mathematical manipulations with generating functions and free variables. And it was a great time. And some papers with Ted Pine on topological defects where we did homotopy theory, and that was actually... I did a lot of that. The homotopy theory was me.

3:16:34.3 SC: But the one that I remember the most was actually my paper with [3:16:37.3] ____ on time machines in 2 plus 1 dimensional gravity. And it was almost an accident. And even here, I don't deserve super good credit for it. But we were trying to understand whether or not in a world that was only two-dimensional space, you could build closed time like curves, Richard Gott at Princeton had written down the exact solution to Einstein's equations that looked like it did have closed time-like curves, and we were trying to ask, "Could you start with the space-time with no closed time-like curves and create them by manipulating the matter and energy in that universe?" And we had developed some pretty nice math-y manipulations with holonomy. I don't even know how to pronounce these words anymore. Yeah, holonomy, like if you take a vector and you parallel transport it around some curvature, it will rotate a little bit, and that's the holonomy of the vector.

3:17:34.1 SC: And we realized that all these holonomy transformations are elements of the Lorentz group in 2 plus 1 dimensions, the group of Lorentz transformations, of boosts and rotations. So that much we knew. And then I was giving a talk on this, and one of the people at the talk was Don Page, well-known cosmologists, theoretical physicist, and he off-handedly said that, "Oh, yes, the Lorentz group in 2 plus 1 dimensions, that has the geometry of anti-De Sitter space." And I didn't even... I had a vague idea of what these words meant, but I was not aware of that fact. So I knew that groups could be thought of as manifold, and therefore, they could have geometries. I did not know that the particular geometry of this particular group was equivalent to that of a well-known space-time. This was long before people were doing ADS CFT. So most people didn't even know anything about ADS at this time, this was like 1991 or something.

3:18:38.6 SC: And so, that tickled me. I loved the fact that this Lorentz group, you had a geometry, it was kind of like a space-time all by itself. It was not the space-time we were concerned with; it was another space-time. And so I went back and I thought about that, and I showed how this... We had what we thought was a proof, or at least an argument, that you could not build a time machine in 2 plus 1 dimensions, and it was this long, complicated algebraic thing, mostly Alan Guth had done it, and even did some numerical simulations to verify that it was alright and everything. But it was just not compelling. It was there, and we were kind of reluctantly going to write it up, and what I realized is that you could actually use this fact that the group of Lorentz transformations in 2 plus 1 dimensions has the geometry of anti-De Sitter space to just draw a picture, a space-time diagram of that group, and draw some curves in that diagram and instantly improve the result. [chuckle]

3:19:45.6 SC: So this goes back... It's funny, 'cause I did say earlier in the AMA, my dream is to some day writing paper with no equations, just pictures. This is the closest I probably will ever come. The paper we ended up writing has lots of equations in it, 'cause we were setting up the whole problem, but the actual important part of the demonstration is just a picture that you can draw once you know that the Lorentz group in 2 plus 1dimensions has the geometry of anti-De Sitter space. Does that count as mathematical jiu-jitsu? I don't know, but I was very pleased with it myself. I'm still sort of coasting on that little victory there.

3:20:19.7 SC: Okay. Two more questions, let's see if we can get through them before I lose my voice. Simon Sanliand says, "If the universe is something best described as the evolution of some Hamiltonian in Hilbert space... " By the way, sorry, it's not the Hamiltonian in Hilbert space; it's a vector evolving in Hilbert space under the guidance of a Hamiltonian. But okay, I get the point. The question continues, "but the arrow of time is an internal property of this universe due to entropy and the second law of thermodynamics. Why are people so set on finding explanations or even causes of this universe by focusing on the Big Bang and that which may have preceded it?"

3:20:57.9 SC: Well, look. We don't know whether the moment that we call the Big Bang or the vicinity, whatever happened 13.8 billion years ago, we don't know whether that was the beginning. Like you imply in the question, that might have been the beginning or there might have been something preceding it. But what we do know is it was kind of special. It looks like a very low-entropy state in a space and states where most states are very, very high-entropy. That's a feature. That's something remarkable. It is literally worth remarking on that fact. And so, whatever the explanation is, it's going to have to account for that fact. That is the most blatant thing that needs to be accounted for. So I don't know, ultimately, what the right explanation is, but focusing on that seems to me to be a very sensible thing. I mean, what do you want us to do? Focus on what the universe was like yesterday? [chuckle] By the laws of physics, the universe yesterday is related to the early universe, but it's just much more complicated and messy yesterday. The thing that seems to be most explicable and also has the power to then explain so much else is the condition of the universe near the Big Bang.

3:22:08.1 SC: Okay. Last question is from The Great Deceiver who says, "I'm curious to know if you give any intentional time to silence, however you might define it, either personally or in your teaching. How important do you think it is as a theoretical physicist? It does seem like a good number of scientific discoveries, at least, have come about through various practices of quiet contemplation."

3:22:31.5 SC: Yeah, I have a mixed response to this one. In principle, I'm very sympathetic to the idea that quiet contemplation can play a huge role in scientific breakthroughs. I do value walking in silence or even just taking a shower in the morning and thinking and not babbling or anything. But sort of actively being quietly contemplating is not something I am any good at at all, like sitting and either meditating or just sitting in a chair, not doing anything. That's just not gonna happen. I need a book or a piece of paper or an iPad or phone or something, whether it's watching TV, reading a novel, playing a game, writing a paper, scribbling, writing some diagrams, doing something. So, in order to do quiet contemplation, I literally need to leave all the devices at home, get out, walk around, and I sometimes do that. But if I'm honest, it's not something that I put a lot of energy into doing it. Who has time for that?

3:23:43.1 SC: When I say that, of course, it's ironic. I should have time, I should make time to do that. I agree. You're making me feel bad. Why are you making me feel bad after I talked to all these hours for this AMA? But look, you gotta make some choices in life about how to spend your time. And I hope that you think the last few hours you have spent in a somewhat rewarding fashion.

3:24:04.9 SC: Thanks for listening. Huge thanks, as always, to all of the Patreon supporters of Mindscape. I really appreciate it. You keep me going. See you next week.

[music]