AMA | October 2025

0:00:00.2 Sean Carroll: Hello everyone. Welcome to the October 2025 Ask Me Anything edition of the Mindscape Podcast. I'm your host, Sean Carroll. Usually, one of the things that slows me down when I'm doing the AMAs, the Ask Me Anything episodes, is I got to think of something to say here in the intro. I suspect nobody cares about the intro. I suspect many of you skip it. But I want to start with something interesting. I'll be honest, right now, this semester, there's been enough going on. I have nothing interesting to say right now. Here's the most interesting thing I have to say, the Philadelphia sports scene has been a disaster. You know, I live in Baltimore, but we're next to Philadelphia, and I grew up in Philadelphia, so all of my teams are still Philadelphia teams. Both the Eagles and the Phillies had very, very high hopes. And also the Penn State football team, which is not Philadelphia, but Pennsylvania, really high hopes completely dashed in different ways. The Eagles won the super bowl last year, so they can't really complain, but this year, they're not nearly doing as well. The Phillies zoomed into the playoffs and then completely crashed and burned.

0:01:01.5 SC: It's gotten so bad that my beloved Philadelphia 76ers are like the hope for Philadelphia sports fandom. That's not a good position to be in because the Sixers have not been making people happy when the chips are down for the last many years in a row, let's put it that way. Anyway, it made me think of, like, why we put so much of our... Why I do anyway, why sports fans do, put so much of our emotional energies into these things that are absolutely beyond our control, in part because it's a distraction from the terribleness of the world around us. But I would like my distraction to be a little bit more rewarding. I really, really need the Philadelphia 76ers to be successful in the NBA this year to preserve my sanity as the world collapses around us. That's it. That's the best thing I have thought of to say to begin the podcast, for what it's worth. The other thing is, various scheduling issues have come up. Usually I try to put the Ask Me Anythings at the first Monday of every month. For various reasons, that's not happening. But there will be one AMA in October, in November, in December, as usual.

0:02:09.3 SC: So don't worry about that. It might not be the first Monday, but they're going to happen. For those of you who are new here, these Ask Me Anything episodes are sponsored by supporters of the podcast on Patreon. You could be a supporter of the podcast on Patreon if you go to patreon.com/seanmcarroll, I can promise as someone who has been a Patreon supporter and is a Patreon supporter of other people's efforts in various directions, it's way easier to do than you would think. It's simple to sign up. You just decide who to support. It's cheap, it's much less than you spend on coffee or pizza or whatever for a typical American, and it's rewarding. You get hours of entertainment, you get no ads on the podcasts, and you get to ask these AMA questions as well as other little tiny benefits that accrue to being a Patreon supporter. So many, many thanks as always to the existent Patreon supporters. You keep me going, you keep me entertained and enlivened doing these AMAs. So let's go.

[music]

0:03:27.8 SC: Tim Falzone says, can you expand on how you understand David Lewis' modal realism and possible worlds in the context of Everettian many-worlds? They're obviously describing very different things, but is there something to explore in their connection? How should we think about counterfactuals in terms of a branching universe?

0:03:45.1 SC: So for those of you who aren't familiar with either one, the Everettian many-worlds, which you probably are familiar with if you're listening to this podcast, is the idea that in quantum mechanics, if you just take super duper literally, what the Schrödinger equation predicts predicts, and you think that the quantum wave function is a physical real thing, not just a way of calculating probabilities, but the real stuff that exists, and you think that everything is quantum. So even big classical things like observers and video cameras are truly quantum mechanical at heart. Then the reason why in quantum mechanics when you observe, say, the spin of a particle, you could get multiple outcomes of your measurement, but in fact you only get one is because the wave function of the universe branches into different worlds and in each world there is some experimental outcome, but those are the different outcomes. So every outcome becomes true, but in a different world. At least it becomes true if the probability in the usual way of thinking about it was not zero.

0:04:47.4 SC: Okay? So it's not true that in Everettian many-worlds everything happens. Number one, only things that are predicted by the Schrödinger equation happen. And number two, sometimes those either have zero probability or can't even be imagined to happen because they would violate the underlying dynamics of the theory. Like, for example, the classic example is electric charge is always conserved in the Schrödinger equation, you never see a proton decaying into an electron or something like that. Okay? So there are things that don't happen. There are worlds that don't exist, even in the Everettian multiverse. In contrast, David Lewis was a fan of what he's called modal realism. So David Lewis is an extremely distinguished philosopher. He was trying to make sense of what is known as modal logic. I think that's his motivation. Anyway, this is not like super duper my area, even though I know something about it. So Modal logic is perfectly straightforward and important. It's logic where you add the idea of a possible truth and a necessary truth, right? And so you have not just if P then Q, but you have possibly P and necessarily Q and things like that. So you try to invent a logical structure that includes all of these operations.

0:06:03.6 SC: And Lewis is trying to sort of make sense of that, trying to say, what does it mean that something is possibly true? Like aren't things either true or not true? And Lewis says no. We can contemplate different possible worlds, and in one world the dark matter is an axion, and in the other world the dark matter is a weakly interacting massive particle. Okay? In fact, we can contemplate all possible worlds and there's no restriction here about the laws of physics or any laws of physics at all. He really means all possible worlds. And Lewis apparently thought that we should consider all these possible worlds as actually real, that there's nothing special about our world except that we were in it. All these other worlds have equal modal or ontological status in some sense. So these two different ideas, the Everettian multiverse, many-worlds, and Lewis' modal realism are different. Clearly very different. Right? The mechanism for making these different worlds is completely different. In Lewis, the mechanism is you imagine the existence of every possible world. In Everett, the mechanism is you obey the Schrödinger equation. Okay? Those are very, very different. But so I think that...

0:07:18.3 SC: So the quick answer is no, these are completely different things, but there are some similarities between them. One of the things that Lewis was emphasizing is that you can sort of think of all your uncertainties about the world you live in as indexical or self locating uncertainties. Every world exists, right? And so when you say I don't know what the dark matter is, that's just a question of which world am I in. And so I can know everything about all the worlds, but just not know which one I'm in. That's indexical or self locating uncertainty. So he thought about all these scientific questions, all the Bayesian probabilities we would attach or credences to different things about the universe as not knowing which possible world we're in.

0:08:04.8 SC: Likewise, as Chip Sebens and I have argued, you can think of probability in many-worlds of Everettian quantum mechanics as self locating uncertainty. I measure the spin of an electron. It's either spin up or spin down. But there's a period where I don't know what the answer is even though the universe has branched. So I have self locating uncertainty about where I am. People like Alastair Wilson, the philosopher, has developed the connection a little bit more deeply. So I think that the answer is there are absolutely close connections and analogies between David Lewis' modal realism and Everett's many-worlds. They're not exactly the same thing by any stretch of the imagination, but we reason with them in similar ways and they have sort of similar implications in various senses.

0:08:55.8 SC: Sean Bentley says, my ninth grader has an interest in STEM and physics and asked me recently what does a theoretical physicist do and how does it compare to an experimental physicist or engineer? I provided him the best 30,000 foot overview I could. But curious if you could share your perspective.

0:09:12.1 SC: I'm very tempted to say that a theoretical physicist just spend their days answering emails and teaching classes, but maybe that's just like my little tiny horizon right now. Of course, what is being asked is the research side of things, the fun side of things, of being a theoretical physicist. It's quite different actually than being an experimental physicist or an engineer. I'm not the one to actually say what engineers do very well. I know that there are engineers who put a lot of thought into things before bringing their hands to actually create something. So they're sort of something like theoretical engineers as well. But I'm really not an expert or even at all knowledgeable on how that division of labor occurs. In physics, you know, it used to be not that long ago, 150 years ago, there wasn't really a difference between theoretical physicists and experimental physicists.

0:09:58.6 SC: They were just physicists and they did their work. But as the domain of physics has become larger and the specific knowledge that you need to be a cutting edge, top flight, either theoretical physicist or experimental physicist, those have become specialized and now there's a difference between them and it's quite a different thing. Theoretical physicists like myself, it's a pretty good job if you're into physics, right? You get to think about the universe. Again, we're putting aside all the nonsense of being a Professor and stuff like that. And just focusing on the fun part of doing research. You ask yourself big questions about the universe. It's almost always true that these big questions come from talking to other people or reading other people's papers. Sometimes you just have a pure, completely unprecedented idea, but that's usually not how it goes. Like, if you think about some of the papers...

0:10:56.9 SC: I did a, I guess an end of year episode a while ago, hits and misses, right, where I talked about some of my favorite papers that I've written, some of which became popular and some not. So you can actually hear in that episode how some of these papers came about. And I say papers because writing papers is what physicists do, theoretical physicists or experimentalists, you do research, but then the point of the research is to get a publishable result and write it as a paper. And it's like sometimes it's just a what if question. Like, what if there was a violation of Lorentz invariance that involved the electromagnetic field? What if inflation happened at different rates in different directions in space in the early universe? What if we modified gravity and attempt to do away with both dark matter and dark energy? These are all sort of what if questions that might lead a theoretical physicist to write a paper. Sometimes someone else writes a paper very, very often, and that gets you thinking. And then you start either criticizing that paper or building upon it in a certain way and trying to do better. So you're driven by these questions that are, you know, in the real world of real theoretical physicists, the questions are usually very specific.

0:12:09.5 SC: One way you can tell a crackpot is if they say they're a physicist, but the only papers they've ever written are on theories of everything. Real physicists don't do that. Okay? Real physicists, maybe some times, very, very rarely, you would come up with a theory of everything. But nobody, not Ed Witten, not Isaac Newton, not Albert Einstein, not Stephen Hawking, not Stephen Weinberg, writes most of their papers about proposing a theory of everything. There's usually steps along the way that you got to do. Sometimes you have a brilliant idea about doing a calculation that really helps people out. That's wonderful. Other times you're working with a model. So in other words, some things you can do is to take physics as we know it and think about either how to calculate or how to explain it in a better way than ever before, or to work out some of its consequences in a more full way. Like Richard Feynman mostly did that. Richard Feynman actually didn't propose a lot of new theories.

0:13:05.6 SC: He thought very carefully about how to understand existing theories. Whereas other people will propose models. You propose a new set of particles or fields or laws of physics or other ways that physics can work. And you see what that does, does it explain what we already know? Does it predict something new? All sorts of things. So Alan Guth suggesting inflation is sort of like suggesting something new. Steven Weinberg saying how electricity and magnetism can be unified with the weak interaction, that's building a model. The problem with model building is it's a very standard thing to do. It's just very hard to predict the right model, right, to just guess that. And so that's a little bit more difficult. And once you have these ideas, you work them out, right? You sit down, you pencil and paper it, you write down equations, you think about, okay, if this were true, so what if we did violate Lorentz invariance with electromagnetism, what would that mean? What would the equations look like? What would the predictions look like? What are the data that I could compare them to, all those things you figure that out, you write the paper, that's what you do.

0:14:05.0 SC: Usually, but not always in collaboration with other people, as opposed to experimental physicists, who of course, have a very different day-to-day. For one thing, they have equipment, they have instruments that they have to either build or maintain or use or something like that. There are still absolutely plenty of physicists who have their own labs. As I would say the majority of experimental physicists have a lab in their physics department. They build equipment and they do experiments. Some experimental physicists, or for that matter, observational astronomers, work at large facilities like Large Hadron Collider in Geneva, or a telescope, a satellite, something like that, you don't know.

0:14:47.4 SC: Usually, even for something like the Large Hadron Collider, the experiments and the analyses are spread out among many, many universities around the world. And individual universities will build a little bit of the equipment that eventually goes into the LHC, even something as large as that. So building equipment is important. Talking to the theorists is important. Like what can we do an experiment to test, what would be the interesting experiments to do? And then there's enormous amount of work in just like data analysis, data collecting, data storage, data transfer, all sorts of ways to take the output of your experiment and to turn it into a scientific result. So the best time in physics is of course, when the experimentalists and the theorists are in close contact moving things back and forth. But at any one moment, they're both doing very interesting things.

0:15:37.2 SC: James Allen says, it's a commonly held belief in classical logic that you can't prove a negative. Strictly speaking, there's no way we can prove that by consuming exactly the right mixture of pizza toppings, Sean Carroll won't turn into a cat and turn back at midnight. But surely we know enough about the natural world to know that is not the case. Are you comfortable calling that sort of knowledge proven, or do we need a new verb to go with this kind of knowledge that we are basically certain of?

0:16:01.7 SC: So I would never call anything in science proven. Science doesn't work that way. This is... I'm giving... I'm talking about this now, but I have talked about this many, many times. Science does not prove things. That's number one. And number two, the statement that it's a commonly held belief in classical logic that you can't prove a negative is certainly not a commonly held belief among people who know something about classical logic. You can always prove negatives. Proving negatives is very easy. Let's say you have a proposition X and you would like to prove not X. Just define Y equals not X and then prove Y. There's literally no difference in classical logic between proving a positive and proving a negative. The statement that you can't prove a negative, number one is not right. But also number two is mixing up classical logic or logical thought where you actually have axioms and prove consequences of the axioms. And science, which is a completely different thing. Science does not work by deduction, by starting with axioms and then proving things, science makes hypotheses and then tests the hypotheses and uses something like Bayesian updating to improve or diminish your credence in those hypotheses.

0:17:13.4 SC: So for the statement that, the proposition that consuming exactly the right amount of pizza will turn someone into a cat, that's a proposition. You can have a credence in it. The credence should never be either one or zero, which it presumably would be if you could prove it. Because maybe tomorrow someone exactly has that amount of pizza and does turn into a cat. You don't know. Right? Science has to be open to changing its mind tomorrow when new data comes in. Of course, it is the case that some propositions, even though maybe you could get enough data to eventually believe them, seems so unlikely to us. Our credences in them are so low that we treat them as more or less established and we move on. That's fine. I don't think that we need a new word for it. I do think that it would be nice if people better understood the difference between proving things in mathematics or logic and being confident in things, in science or more empirically based things. But it's not the biggest thing we have to worry about right now. Certainly introducing a new word is not going to... I think in the real world, that's actually not going to take off.

0:18:21.2 SC: Frank Russler says, when we measure an electron's spin, the Everett theory says that the observer splits into two universes. One universe observed the electron spinning clockwise and the other witnessed the electron spinning counterclockwise. My question is, where are these universes compared to each other? How would an entire universe be created by a simple measurement?

0:18:42.9 SC: Again, this is something I have talked about before, but sometimes it's good to update people to remind them of where we are. There's no such thing as where the universes are in Everettian quantum mechanics. You need to abandon, if you really try to understand quantum mechanics intuitively, you need to abandon a lot of your views of how the world works at a sort of informal classical, folk physics kind of way. And one of the things you need to abandon is the idea that there is space in which everything has a location in space. Okay? That's just not how things work. In Everettian quantum mechanics in particular, the motto would be, space is in the universes, in the worlds. It's not that the worlds are in space. The worlds are in Hilbert space in some sense. But Hilbert space is just an abstract mathematical language that we use to describe the worlds. It's better. It's perfectly okay just to say the worlds don't have locations. The kinds of things that have locations are objects in space. And Everettian worlds are not those. The multiple worlds all exist simultaneously. They can all be traced back in time to a previous undifferentiated world. And that's how the equations work. And that's what you have to accept if you're going to believe this point of view on quantum mechanics.

0:20:03.3 SC: How would an entire universe be created by a single measurement? That's what the equations tell you happens. I'm not sure what to say. It's not like you're copying the universe over and over again. I did write a book about this. Sometimes I want to say... I try to bite my tongue, but sometimes I want to say, if you're really interested in these things, and you know I wrote a book about it, you could read the book and you would definitely find the answer. So the idea in quantum mechanics is that the universe branches, and each branch is thinner than the one that came before. The branches have the same overall thickness.

0:20:39.6 SC: And this thickness is a very precisely quantifiable number. If you have a specific wave function that is doing the branching. And so you get more and more universes, but they are in a very real sense, thinner and thinner. You can think of the thickness or thinness as just the amplitude squared for that branch, which is basically the probability that you will find yourself in that branch. So, of course, if I am with probability one in a certain branch now, and I know that I'm going to do an experiment, and with probability one half, I'll be in branch A, and with probability one half, I'll be in branch B, that means that the two branches A and B have half of the thickness that my current branch does. So the overall thickness is conserved. You don't need to do any work to create a new universe. You're just differentiating or splitting the universe you already have.

0:21:31.2 SC: Brandon Wheeler says, how do you feel Chicago is handling the actions of the President? So for those of you who have not been following or listening, 500 years from now, we're in a situation where the President, Donald Trump and his administration are sending troops of various sorts, a sort of motley collection, not yet the army or the Navy, but the National Guard and various Homeland Security and immigration enforcement personnel into American cities, even though the cities don't want them and they're misbehaving in all sorts of ways, let's just put it that way. There are lots of lawless behavior. There's a long list you can look up online. I'm not going to recite it here. Of blatantly illegal activities on the part of these people who are being sent into the cities. And I think it's pretty clear to most people, including plenty of Republicans who have spoken out about this, that the point of sending these troops into American cities is not to squelch violence and insurrection or anything like that, but to foment it, to act in such a way that people react in such a way that they can say, oh, look, you're being violent. Okay?

0:22:42.4 SC: It's not that it's happening, it's that they're trying to make it happen. So what do you do if you're a locality like Chicago? Chicago is one of the cities that is a target of this behavior. And it's a very difficult issue. I mean, I am not super duper detailed informed about what specifically Chicago is doing. So I don't have reliable opinions about it. I think that they're kind of trying to do the best they can because you can't just wall off your city and not let in the federal forces. Okay? That would be a bad precedent for all sorts of reasons. Local law enforcement and officials have a history and a tradition and norms of working with federal law enforcement. I mean, basically, you'd be declaring a civil war if you tried to use your local forces to prevent federal forces from coming into your city. But at the same time, of course, you don't want to just let them run roughshod over your citizenry. You want to protect the people in your city. So Chicago is trying to do that. They've done what they can to sort of make it uncomfortable and difficult for the federal forces to raise the havoc that they're trying to raise.

0:24:00.0 SC: Are they doing in the best possible way? I'm really not sure. I don't know. It seemed competent and sensible, as far as I can tell. JB Pritzker, the governor of Illinois, where Chicago is located, has of course been very, very vocal about saying that state governments more broadly should unify and stand up against this federal intrusion into their places that are under local control. But we're going to have to see. This is all very, very unprecedented circumstances. It is hard to know what to do, hard to know what the right action is to take. There are visceral, emotional responses. You say, like, you got to stand up, you got to stop this. But someone has to be the adults in the room. Someone has to recognize that hopefully we're still a functioning democracy 10 years from now. And that's the goal that we need to work toward. And sometimes that requires a delicate hand rather than just giving in to your instincts of the moment.

0:25:02.1 SC: Josh Dobbins says this may be a dumb question, but when we say the universe is 14 billion years old, or talk about what the universe looked like at one year or one billion years, I get confused because years seems so relative a term for time. If the early universe is rapidly expanding, didn't it experience some kind of time dilation from a different part of itself? So the age in years becomes a kind of term that doesn't apply because different parts are experiencing time along different velocities.

0:25:29.1 SC: So, no, actually, this is actually pretty... Well, this does make perfect sense, talking about the age of the universe. You do have to know what's going on here. And actually, there's a lot going on here. So one thing is, time doesn't go along at different velocities. Time always goes at one second per second as I like to say. I know that when you get certain casual introductions to the theory of relativity, you can be told that time moves for different observers or in different gravitational fields at a different rate. That's just confusing you. This is why you got confused, because some physicist has explained this to you badly. Okay? There is a true statement to make that between two events in spacetime, where an event in spacetime is specified by both the location in space and a moment in time. So, like this particular place in Baltimore, Maryland, on October 7, 2025, et cetera.

0:26:23.6 SC: And another event which might be the same exact location, but at a different time, you can take different trajectories between these two different events, and those trajectories can experience different amounts of time. It's exactly like, as I often like to say, two different points in space can be connected by two different curves with two different distances along the curve. That's really the insight of relativity that time is not a universal thing. Time depends on the trajectory you take through the universe. But it's not that the rate of time is changing in any way. The rate of time is always one second per second. It's that you have taken a different trajectory through the universe. Maybe your trajectory is very curved, or maybe your trajectory just moves to a place where spacetime itself is very curved, like near a black hole or something like that. Now, in the early universe for cosmology, of course, you could imagine different observers moving at different speeds and therefore, starting at the same point, one goes off near the speed of light and then comes back, just like a twin paradox kind of motion. And they would experience different amounts of time.

0:27:32.4 SC: But that has nothing to do with the expansion of the universe. That's just because the twin moved off near the speed of light. When you talk about the age of the universe, you're taking advantage of the fact that in cosmology, as we know it, the Universe is pretty uniform in a particular reference frame. Right? There's a reference frame you can be in where you're mostly at rest with respect to all of the matter around you, or at least the average velocity of the matter around you. In the universe today, for example, we can look at the cosmic microwave background, which looks pretty uniform to us, but there is a rest frame for the microwave background. If you're moving with respect to that rest frame, it's blue shifted in one direction, red shifted in the other direction. And that's a very noticeable, easily measured effect.

0:28:16.1 SC: So when you talk about the ages of the universe or something like that. You are implicitly imagining measuring that via a clock that is more or less at rest with respect to the matter and radiation in the universe. And when you do that, the notion of what the age is, becomes perfectly well defined. The relative velocities of different things in the universe are rarely more than something like 0.1% the speed of light. They're very, very slow compared to the speed of light. So relativistic effects just don't matter. Of course, spacetime is highly curved in the early universe, but it's not curved in a way that affects clocks in any noticeable way, because it's curved in the same way from place to place because of the uniformity of the universe. So as long as your clocks are all basically moving not very quickly compared to the matter around them, they will all read the same amount of time. That's the time that we attach to the age of the universe and such like quantities.

0:29:15.7 SC: Ben Lloyd says, in light of the recent news about Charlie Kirk and the surprising amount of people actually celebrating it, I was wondering about when political violence is and isn't justified. So, obviously political violence is almost always unjustified, but I wouldn't say literally all political violence is unjustified. I mean, would anyone really be upset if Hitler was assassinated during World War II? So my question is, in your opinion, where exactly are the lines we draw for when it is and isn't justified? I'm thinking it has a lot to do with if you're directly carrying out inhumane acts and if you live in a democracy which allows other pathways to change.

0:29:51.8 SC: Well, I think it's a difficult question. I don't think that there are exact lines that you can draw for when political violence is and isn't justified. I think it is very, very rarely justified. And one's first move, one's default, should not be to be violent. Okay? Or should be to not be violent, I suppose I should say. Charlie Kirk getting assassinated was certainly not justified. That was terrible. No one should celebrate it. I don't think many people are actually celebrating it. I think that there's a lot of people, whenever this kind of thing happens, who want to imagine that their opponents do bad things, so they look for opponents politically doing bad things, and then they amplify those voices. So it seems that a lot of bad opinions are out there. I didn't come across a lot of people celebrating the news about Charlie Kirk. What I came across in my own readings is a lot of people saying, it's terrible that Charlie Kirk got shot and killed.

0:30:46.3 SC: I didn't like him at all, and I don't need to pretend to like him just because he got shot and killed. I think that's perfectly fair. But as far as the actual question, when is political violence justified? One reason why things like the Charlie Kirk shooting were not justified is because there's clearly no benefit to doing it. You know, first off, your first instinct should always be not to be politically violent at all. If you want to be politically violent or you're tempted or you're thinking, is it called for right now? Try to be consequentialist about it. Like are you really doing any good whatsoever? In some ways, all revolutionary wars are political violence, right? I mean, in some even broader sense, all wars are political violence. But a revolutionary war is a war within a country to overturn the government, whatever it may be. The American Revolution counts as political violence, right? And we can argue that it was for a good cause, but there was also some consensus, some agreement, not 100% consensus by any stretch of the imagination, but there was a broad movement to change the political system, and it was thought that the only way to effectively do that was to have a revolutionary war, right?

0:31:56.0 SC: That's the kind of place where, in my mind, revolution... Political violence is justified. Would it have even mattered had you killed Hitler? Like, this is a fun thing, again, at the emotional level to imagine, would it have changed things? Do you think that the people under Hitler are like more amenable to being peaceful or something like that? I do think there was a moment when that was true. Historically, it is claimed, from what I understand, that there was a moment at the end of World War II and Hitler's generals were basically ready to end the war, and Hitler himself was not. And maybe if you had that information and it was trustworthy, then doing an assassination attempt on Hitler would have been justified for that reason. And you could have saved some thousands or millions of lives that were lost at the very end of World War II. But those events, those moments are pretty darn rare. You need to understand the effectiveness of your action, what the consequences will actually be. So I think you really have to think about, are you doing this just out of frustration that you want to lash out in some way, or is there like a very broad belief that you can actually make the world a better place by doing this? And again, I think that almost never would you make the world a better place. But there can be times when you would be, and then it would be justified in a sort of political philosophy kind of sense.

0:33:22.9 SC: Lee Vermeulen says, suppose we assume the universe is finite, quantum mechanics is fundamentally discrete, and therefore the total number of possible quantum interactions is countable. If we hypothesize that the universe's size is exactly what is required to encompass all such interactions, what would be the steps and unknowns involved in trying to estimate that size?

0:33:45.7 SC: So I actually wrote a paper about this, finite... I forget what it's called, but it was basically finite quantum mechanics, okay? Finite discretized quantum mechanics. But it was not strictly speaking quantum mechanics. Because what I was talking about in the paper was, could you make a truly finite theory in the sense that there's no continua or infinite numbers. You know the real numbers have an infinite number of numbers in them. Even the amount of numbers between zero and one is infinite in size. So quantum mechanics, as it usually is talked about, has infinities in it, because Hilbert space is a vector space. Every vector space has an infinite number of elements in it. So in my paper, Completely Discretized Finite Quantum Mechanics, something like that, what I try to do is say, could you discretize Hilbert space?

0:34:38.1 SC: Could you make the quantum state live on a lattice that was embedded in the Hilbert space vector space, rather than just smoothly move through the vector space itself? And the answer is yes, you can formally do it. It would probably not, although I couldn't say for sure, be compatible with observations as we know them, because time could run. Basically, you get a recurrent universe, you would have the same states happening again and again eventually, because there's only a finite number of states that there could be. And likely you would spend most of your time in thermal equilibrium with Boltzmann brains fluctuating back and forth. I can't say that exactly. I did not prove that that would be the case. But it seems that would be the most likely case. So the actual question you're getting at is how big would it have to be? It depends on details. So there's no simple answer to that. I think it's easier to say how big would it probably be?

0:35:35.4 SC: Rather than what is the minimum size? You're asking for the minimum size, I think that it's better easier to ask how big would it sort of... What's the most natural number for it to be? And that's actually a little bit easier because we think that the universe is accelerating with a cosmological constant, and that gives us a horizon around us with an entropy. We know what that entropy is. It's roughly 10 to the 122. And we know the dimensionality of Hilbert space. When you have an entropy that big, it is e to the 10 to the 122. Or equally well, 10 to the 10 to the 122, because the numbers don't matter when your numbers get that big. So that is the dimensionality of Hilbert space that we believe is enough to describe the universe. Now, as I showed in my paper, the number of steps you would take stepping through the universe, stepping through Hilbert space over time, is actually much, much larger even than that. Okay? But if you're just asking, sort of, when you say how many degrees of freedom are there, we can sort of translate that in our minds to what is the dimensionality of Hilbert space? And the answer is something like 10 to the 10 to the 122.

0:36:47.0 SC: Joey says, what is your take on the Ship of Theseus thought experiment?

0:36:52.2 SC: Yeah, I think it's a famous thought experiment. I contributed a little bit to making it more famous by writing about it in 'The Big Picture'. I think it's a good example of the need to let go of some of your casual intuitions about how the world works. So for those of you who don't know, the Ship of Theseus thought experiment, Theseus, back in ancient Greece has a ship. Of course, like any ship, if it's going to be around for a long time, there'll be little dings and you're going to need repairs. And so you might take out a board of the ship and put in a new board of wood to fix up the ship. And then you do that again and again. And at some point, you've done this enough that literally none of the pieces of the ship at the current time were there in the original ship. It's like a rock band who have changed literally all of their members. Right? Is it the same band?

0:37:42.1 SC: So is the Ship of Theseus the same ship, even though it contains none of the original material that was in the ship? And I think there's sort of an obvious correct answer here, which is, there's no such essence of shipness of Theseusness that adheres to this particular ship. What actually happens is there's stuff in the universe, it sort of groups together in ways which it is convenient for us at the higher emergent level to call an object the Ship of Theseus. And strictly speaking, one second later, it's not the same ship. By the way, the reason why people care about this thought experiment is because you actually want to ask this question about people, not just about ships of Theseus and so forth. But anyway, it's even more obvious in the case of people that from minute to minute, you're not the same person, right? You have heard things, you have seen things at one minute that you hadn't seen the minute before. So that has changed your mind a little bit. It is hugely convenient to attach an identity to your existence over time for all sorts of reasons, because there's a lot of commonality from you at one moment of time to you at another moment of time.

0:38:56.4 SC: Even if it's not perfect sameness, there's still a lot of continuity. I can use knowledge about things that happened to you in the past or things that you did in the past to predict your future behavior. All sorts of things. Likewise with the ship of Theseus. I have no trouble saying it's the same ship, even though we replaced all of the wood, all the pieces, all the boards in the ship. Why? Because if it functions in the same way, if it's been governed, owned and run by Theseus this whole time and used for all sorts of different things, I gain some information, some knowledge, some understanding by attaching the label the ship of Theseus to that thing. But of course, that's necessarily casual and approximate. It is not perfect, okay? So the only way to be perfect is to understand that at every moment there's a different ship, there's a different thing, there's a different collection of atoms, there's a different status of the wave function of the universe, or however microscopic you want to go. So the question is entirely about the convenience to us. Is it useful to you to think of the ship of Theseus as being the same, even though you've changed all the boards or is it not? Entirely up to you. There's no absolute fundamental truth about it.

0:40:12.1 SC: Ilya Levov says, is there any intuition why among all the fundamental physical interactions, only gravity has an associated unitful physical constant, Newton's constant G, that contributes to the natural units alongside C and h-bar?

0:40:29.0 SC: There is actually, it's a little technical, I guess. There's probably a simpler down to earth way of saying it, but it is a thing that comes out when you write down the dynamical equations for these different theories. So all of the forces of nature, such as we know them, well, the usual four forces of nature, right? Electromagnetism, the weak nuclear force, the strong nuclear force, and gravity, they're all gauge theories in certain ways. Which is a fancy way of saying there is a symmetry group associated with them that acts independently at every point in space and time. But there's also a difference between gravity and the other gauge theories. All the other gauge theories refer to what are called internal symmetries. So for example, at every point in space, we have three different kinds of quarks. Right? Well, let's pick a kind of quark, like the up quark.

0:41:22.5 SC: The up quark has three different directions in color space that it can be in. It can be in the red direction, the green direction, or the blue direction. So we talk about three different colors of quarks. Talking about three different colors is not exactly right. It's really a three dimensional vector space. And any individual quark field has a direction in that vector space. And so it can be a linear combination of a certain amount of redness, greenness, and blueness. Okay? And the gauge symmetry in question is the fact that we can change our definition of what is red, what is green, what is blue. We can rotate those axes to our heart's content, but that doesn't constitute a real rotation in space. Right? We're not actually moving the direction in which we're pointing when we change red, green and blue. These are just labels we put on these fields that exist at every point in space. Whereas in general relativity, the gauge theory is a different kind of thing because it is spacetime itself that is being described. What that means is, there's a different set of fields involved. In general relativity, you start with the metric tensor, which tells you the distances along curves in spacetime.

0:42:33.7 SC: From that you make a connection field, and from that you make a curvature. Whereas in other gauge theories, you start with the connection, there's no metric. The connection is the fundamental object and you make the curvature directly from it. And what that means at the technical level, that I'm sure this isn't going to make a lot of sense, but the equations you can write down for the dynamical fields are a little bit different in general relativity than in the other gauge theories. And it is a consequence of the difference in the equations that the leading term, the sort of first thing that shows up in general relativity has a dimensionful coupling constant, Newton's constant.

0:43:14.1 SC: Whereas the first thing, the dumbest thing you can write down in all the other gauge theories is dimensionless. The strong coupling constant, the fine structure constant, et cetera, are all dimensionless numbers. There's still other dimension full coupling constants you can write down in the other theories, but they're thought of as higher order corrections to the sort of leading classical version of the theory. So I wish I had a better reason to give you that was a little bit more intuitive for why the equations have that form. The very, very short version of the answer I just gave is, general relativity has the metric and the other gauge theories don't. That gives you a little bit more freedom. And that freedom ends up giving you a leading order term in the equations of the theory that requires a dimensionful coupling.

0:44:04.0 SC: David Kudavarian says, occasionally I've come across the term density operator, but I've never been able to understand in which context it's useful. Sometimes it's mentioned that it is used to in the context of a system for which we don't have complete information. I can't quite wrap my head around the idea behind it. Could you please shed some light on this subject?

0:44:22.6 SC: I can try. It's a little bit technical for quantum mechanical purposes here, but I think that the idea of what you're trying to do with the density operator is not hard at all. It's analogous to a probability distribution for a classical system. So classically, in statistical mechanics, we'll very often say something like in the air in this room, there are specific states of the positions and velocities of every molecule, but we don't know them. Instead, we can describe what's going on in the room with a probability distribution over what might be going on. And it turns out that's enough to calculate things like temperature and pressure and density and even predict the future evolution of the system to quite a lot of, a very good approximation. So what if you wanted to do the same thing in quantum mechanics?

0:45:11.2 SC: In quantum mechanics, you describe a system, you're taught to describe a system using a wave function or a quantum state, and you can use the wave function to make predictions. The predictions are not 100% predictable. They're a probability distribution. You take the wave function as a complex number assigned to every possible observational outcome, and then the probability of getting that particular outcome is that number squared. That's the Born rule. So anyway, what would you like to do if you wanted to do a statistical mixture, a probability distribution of quantum states? Well, you might think what you want to do is take every possible quantum state and assign it a probability, just like what you do in classical mechanics. You take every possible state of the gas in the room and you assign it a probability. It turns out that in quantum mechanics, you don't need to do that. Indeed, it sort of doesn't even make sense to do that, at least in the...

0:46:04.9 SC: There's different levels we can go into here depending on one's favorite interpretation of quantum mechanics. But let's put all that aside and just use the textbook picture where you have a wave function and you use it to make predictions for observational outcomes. The point is that two different quantum states can still give you some non-zero probability of getting the same measurement outcome. So if you take a probability distribution over all possible quantum states, that turns out to be super redundant in terms of the important question, which is what is the probability of getting a certain measurement outcome? Okay? What you really want is just the statistical mixture of measurement outcomes rather than the statistical mixture of quantum states. Because quantum states are vectors, you can't even tell once you start mixing them in some statistical mixture, which one is which that went into the mixture in the first place. So what you want is a mixture of observational outcomes, of probability, of observational outcomes, and that is the density operator. So a density operator is a way of saying here is the probability for every measurement outcome without saying the system is in a particular quantum state.

0:47:15.0 SC: It could be in a superposition or a statistical mixture, I should say, not a superposition, because that's a technical quantum mechanical term. It could be a statistical probabilistic mixture of different underlying quantum states. Now, so that's the little technical bit about why you want to do something different in quantum mechanics to describe a statistical mixture. But one of the reasons why you might be confused about the idea of the density operator is that it comes up in quantum mechanics quite a bit more frequently and centrally and crucially than in classical mechanics. And that's because of entanglement. If I have two systems that can be entangled with each other, what that means is that the system as a whole, with both subsystems included, might very well have a pure state, as we call it, that is to say, a wave function or a vector in Hilbert space describing the full quantum state of the system. Nothing statistical or imprecise about it. But because the two states, the two systems are entangled, each subsystem cannot be described as a pure state of any sort, it can only be described by a density operator. And you can even quantify this, as John von Neumann figured out how to do, in terms of the entropy. In classical mechanics, if you have a probability distribution, there's a formula that comes from Boltzmann and Gibbs for the entropy of that probability distribution.

0:48:41.0 SC: And likewise in quantum mechanics, there's a probability... There's an entropy formula for a density operator. And also likewise in classical mechanics, if the probability distribution is a delta function, that is to say, if the probability is exactly one, that it's a certain state, and zero that it's everything else, the entropy is zero. The entropy is a measure of how spread out all the possibilities are. The more spread out the possibilities are, the higher the entropy. Same thing goes true for density operators in quantum mechanics, there's a formula to calculate the entropy of a density operator, sometimes called a density matrix, by the way, just because operators are often represented as matrices in quantum mechanics. But anyway, if your state is a pure state, if it can be represented by a single wave function, then you plug it into von Neumann's entropy formula and you get zero.

0:49:35.4 SC: There's zero entropy for that. The entropy even in quantum mechanics, even when you already have a theory that only makes statistical predictions for measurement outcomes, you still have a formula that tells you the difference between being in an exact pure state and being in a probabilistic statistical mixture of states. And when you apply that to two subsystems that are entangled with each other, each subsystem has a non-zero entropy, even though the two subsystems as a whole have zero entropy because they're in a pure state. That state of affairs is 100% purely quantum mechanical. Nothing happens... Nothing like that happens in classical mechanics because there's no analog of entanglement. I have no idea, David, whether that made any sense or not, but I think that's the best I can do without knowing more about what level to pitch the explanation at.

0:50:28.6 SC: Okay, I'm going to group two questions together. Marie Roscue says, I've heard and read a lot of myths about physics on so called scientific podcasts or in books for kids. Like atoms are mostly made of empty space, or the earth is in a constant motion toward the moon, et cetera, I struggle to persuade people that things like this are myths. What is the best way to argue and/or explain things like that to people out there, either on social media or in real life? And Jeff B says, I'm teaching an 8th grade science class for the first time this year. For the most part it's going great. But I have one student who appears to have gotten a science is BS attitude from somewhere. Adults at home, I would assume. I like that Jeff doesn't assume that it's from social media. Good for you Jeff. But it probably is. I've tried my best to explain the scientific method and the power of collecting evidence and data, but he's not buying it. I feel like I have a little flat earther in the making, but I don't see what I can do if he's already made up his mind to distrust what I'm saying. Do you have any advice?

0:51:24.2 SC: So there's a lot going on here. And I want to sort of focus on the particular question of, you know, how do you persuade people who seem to already be completely convinced otherwise? But for Marie's question, the two things that you've picked, the atoms are mostly empty space and Earth is in constant motion toward the moon. These are not cut and dried questions. Okay? I think that a lot of these depend on choices of language and even choices of assumptions about what you mean by how quantum mechanics works, things like that, what relative motion means in general relativity, things like that. So it's less about there being a myth and more about there being a slightly less than perfectly helpful way of describing a true scientific situation. So you can try to persuade them out of it. But calling it a myth might be a step in the wrong direction, if that's what your goal is.

0:52:16.1 SC: The bigger picture thing is, look, I'm not really a believer in human being's ability to use the force of reason to convince other people to believe things they don't want to believe. I'm not even interested in trying to be perfectly honest. I think that most people do not have their minds made up about most issues. So for most people it's perfectly sensible to try to persuade them of something because they haven't made up their minds yet. But there's a state of being human in which you've made up your mind and there's just nothing that's going to change you, right? If that's really what you want to believe and you're not asking questions, you're not sort of in good faith trying to figure out a better way to understand things, then nothing anyone says is going to change your mind.

0:53:03.9 SC: Especially if that what the person says is trying just to use logic. You know, logic is not the tool in these situations. I think whatever put them in the state of believing the BS in the first place probably wasn't logical. And there's this idea that you can't reason someone out of believing something that they weren't reasoned into in the first place. If anything, what you should do is just in some very vague sense, try to convince them that it's cooler and better to be on the side believing the true things than on the side believing the false things. How do you do that? Well, I don't know. I mean, there's not an algorithm here. It's more about being a good example, just not being a hectoring scold and being like a cool person who they want to be like, especially if they're young and impressionable that way. I know that's not entirely helpful advice, but I think it's the best thing that I can say.

0:54:01.3 SC: David Harper says, in the 20th century, there was a big debate in linguistics over Chomsky's notion of a universal grammar underlying all language. At the time, the idea of a machine that could speak language was quite far away. But now LLMs do a convincing job at acting like they know language. Do you think LLMs can help us settle this debate or teach us anything about the human mind itself? Or is the transformer model just a trick that is completely unrelated to the human brain?

0:54:27.6 SC: Well, I think it's in between a little bit. Noam Chomsky himself wrote an Op-ed piece saying that LLMs could not possibly teach us anything about human language because the way that they learn it is completely different than the way we learn it. You know we, in Chomsky's view, have a language instinct and LLMs do not. They're just copying what we do. They're not necessarily going through the same steps or using the same architecture that human minds do. And therefore you can study LLMs all you want, but don't expect to learn anything about how human beings use language. I think that's a lot exactly right. I think that's... I mean, I get what he's saying. I think that there is a point to it. But LLMs are studying what human beings do. Right?

0:55:12.7 SC: And so the optimistic take would be that by studying what human beings do and by being trained to repeat what human beings do in some way, we might then after the fact be able to open up the LLMs, look at the weights, do some interpretation of what's going on inside, and figure out how that maps onto how human beings might actually do it. I don't actually think it's the best way to do it. I mean, it's kind of indirect. Like we have real human beings we can talk to, we don't need to do LLMs. But the nice thing about the LLMs is you can crack them open and you can poke around. The bad thing about them is that they have billions of parameters inside. So it's not at all obvious where to poke or what it means to poke one place or another. We'll have to see. I'm happy to let this kind of research play out and see if it is actually empirically useful or not, it's worth a try.

0:56:09.3 SC: Christopher Smith says, when I was a kid, I used to sit by myself on the couch and think about the possibility that my life is on an infinite loop forever, including my current contemplation about that possibility. I have no idea where this idea came from as I grew up Catholic, so there wasn't a religious component to it, I don't think. I eventually forgot about it. Years later, in my 20s, I discovered Nietzsche's idea of eternal return, which I think originally goes way further back as an idea, possibly to Aristotle. Suddenly, what I thought was my own strange idea as a 10-year-old, I came to realize was an ancient philosophical concept. And this has led me to wonder on occasion if there's any validity to the reality of eternal return. So my question is, is there any physics at all that could in any way validate the idea that the universe is on an infinite loop and that things play out exactly the same way over and over forever?

0:57:01.1 SC: You know, there's some physics that you can imagine has something like that happening. I will first say that the fact that you thought of it when you were 10 years old and the fact that Nietzsche thought of it and the fact that Aristotle thought of it, is not that much evidence that it's real or true or valid. It might just be evidence that it's not that hard to think of it, right? So I think that you have to be a little bit more careful when you're trying to be a working cosmologist here. And you need to say, you can't just start by asking, is it possible that there is eternal recurrence? You have to say, is there a model based on reliable physics that would include an eternal recurrence series of universes that happen over and over again? The real problems here are the arrow of time, which points in one direction in our universe and therefore does not necessarily suggest any recurrence or cyclic behavior. And also the problem of Boltzmann brains that we've talked about before.

0:58:07.3 SC: As I just taught in my philosophy of cosmology class, Boltzmann in 1895 was the one who first suggested the idea of the multiverse and the anthropic principle. And his idea was basically an infinitely big and infinitely old universe characterized by eternal thermal fluctuations. And he said, look, I admit that most of that universe would be in thermal equilibrium and therefore dead, as he put it, and therefore there would be no human beings living there. But occasionally there would be a fluctuation that would give rise to a fluctuation that just a random arrangement of molecules large enough to go from completely featureless thermal equilibrium all the way to like a galaxy with stars and people in it and the whole bit. And he said, maybe we live in the aftermath of just that kind of fluctuation. So the problem with this, as is well known now, is that for whatever kind of observer you think you are, whether it's you're just an intelligent observer, a conscious observer, or whether you want to sort of restrict yourself to exactly human like observers or whatever it is, in this eternally fluctuating ensemble, most such observers are going to be minimal entropy deviations from thermal equilibrium.

0:59:18.2 SC: They're not going to be what we think of as conventional observers that live in the aftermath of a hot big bang with a very, very, very low entropy, surrounded by hundreds of billions of galaxies and all that. They're going to be lonely and all by themselves, maybe even just isolated brains floating in the otherwise empty space, if you want to include those in your reference class of observers. So people have tried to invent ideas of what are called cyclic universes. Turok and Steinhardt famously had such an idea. Roger Penrose has such an idea. I don't really think any of them work for a combination of reasons involving Boltzmann brains and the arrow of time, but I certainly don't have a proof that they don't work. So if you want to know if it's possible, yep, it is possible. It's worth thinking about.

1:00:07.2 SC: David Maxwell says, is fundamental science being sacrificed on the altar of applied science with strategic geopolitical implications a short term effect of our changing world or a deeper enduring challenge?

1:00:19.5 SC: I think it's always a deep enduring challenge. People want to get a return for their investment, right? People want to spend their money, and not just individual people, but societies also, on something that will benefit them in a tangible and predictable way, right? So if you're saying give me money to do science, society has every right to say what's it worth to me? What am I going to gain from this? And you can try to make the case that, well, you gain knowledge and knowledge is good, okay? But that will not necessarily speak to everyone. And other people are only going to be concerned if it's actually giving me better technology, solving some disease, improving productivity or something like that. So I don't expect it to be a short term effect of our changing world at all. I think this is always going to be true and I think that's actually okay. I think that we scientists who care about that should be good at making our case that the kind of science that we do deserves funding.

1:01:16.2 SC: All right, I'm going to group two questions together. One is from Marcin Chady, who says, can you explain why the early universe, incredibly dense as we're told, didn't immediately collapse into a black hole? And Andrew J says, given the extreme mass density, why didn't conditions just prior to the Big Bang result in one giant black hole?

1:01:34.5 SC: So this is a fact. It's a frequently asked question, but I'll take a stab at it, because you never know. Like, maybe trying to explain in a different way reaches a different audience. I will raise one puzzlement in my mind. Andrew's question says, why didn't conditions just prior to the Big Bang result in one giant black hole? And I truly don't know what that means. There might be conditions prior to the Big Bang, or there might not be. The Big Bang might have been the beginning of everything. If there were conditions just prior to the Big Bang, we certainly don't know what they were like. We might have some ideas in certain models, but I don't think it's... We have enough information even to begin to answer that question. So I'm going to sort of pretend the question was, why didn't conditions just after the Big Bang result in one giant black hole, for the reasons that Marcin says, the early universe was incredibly dense, right? There's a lot of matter there. Why didn't it collapse to a black hole? So the way to think about it, one of the ways to think about it, you know, the motto that I always say is, the early universe is like a white hole, not a black hole.

1:02:34.0 SC: It has a singularity in the past, not in the future. So just to elaborate on that a little bit, if you have an ordinary general relativity, if you have a collection of mass, let's say it's collisionless mass, let's say it's all dark matter, okay? So that there's no heating up when it bumps into each other or dissipation or anything like that. And you point all the mass to concentrate at one point in space, okay? So like a cloud of collisionless particles is just shrinking and moving in toward one point in space, that will create a black hole. And what it will look like is that there wasn't a black hole and now there is a black hole, which seems from the outside to be a horizon which if you go into, you're going to eventually hit a singularity, and you can never come out again. That's what the black hole is. But general relativity is a time reversible theory, okay? So we can take that entire solution to all of spacetime, which says, start with no black hole, just diffuse matter all over the place, let it come together and make a black hole, and we can run it backwards in time.

1:03:41.4 SC: And that is going to give us a perfectly valid solution to general relativity. And what that solution looks like is, you start with a singularity, it spits out matter and energy, and that matter and radiation and energy just diffuses into the surrounding empty space. Okay? So there's no rule at all that when you have a lot of density that you must get a black hole. There is a rule that if you have enough density, you must either get a black hole or a white hole. That is to say, you must either have a singularity in your future or your past. But having a singularity purely in your past is 100% allowed by general relativity. So there is zero reason to think that the early universe should have been a black hole.

1:04:29.8 SC: John Kunjumin says, if there is one skill you wish you had spent more time developing as a student or early career scientist, what would that be?

1:04:39.5 SC: I don't know. There's so many skills, a lot of skills that I wish I had better. I certainly wish I knew Quantum Field theory better. I mean, I learned Quantum Field theory and I used it a little bit. But then I went through a long phase of my career where I didn't use it that much. And so now I'm actually, like, using it more. And I wish I had remembered it better or had time to take a class again. That's not really a skill. When I was an undergraduate, I was a really good programmer. I was a really good coder. I don't think we used the word coder back then. It was a long time ago. But I wrote a lot of our own programs to help reduce data and visualize things on old IBM PCs in the astronomy lab at Villanova University. And I let that skill atrophy very much. I am not any good at just whipping up a little routine to do something that I need to do. And we have graduate students to do that now, so what would be the point in doing that? But I do wish that I was better at that. I mean, I was good at it for a little while. I wish I'd kept it up. That's what I would tell my younger self.

1:05:40.1 SC: Make3333 says, are you 100% sure that conscious qualia can be explained in a physicalist framework? I'm usually 100% physicalist like yourself, but this is the one thing where I'm unconvinced. Not that I can imagine anything else being a good fit, but I'm sure I haven't imagined a lot if most of the good ideas in the space of possible good ideas.

1:06:02.1 SC: No, I'm not 100% sure of anything. I mean, I find it hard to believe that no one... That someone doesn't know that. Like, if you ask me, am I 100% sure of anything at all, I'm going to say no, that's just not the way I think it works. But I don't think I need to be 100% sure. There is an idea that, okay, you're pretty sure about something, whether it's consciousness or whatever, and you're asking yourself, what am I going to do with that level of credence? There's many, many, many things I'm not sure about. And to me, once I become sufficiently high credence in something, I'm going to take that as true and move on until I come across some evidence that importantly changes my mind in some dramatic way. So on the one hand, I'm not 100% sure. On the other hand, I'm not worried about it. I'm going to act like a physicalist for all intents and purposes.

1:06:51.2 SC: Zack McKinney says, what's your current thinking on the relative probabilities that AI will or can can facilitate broad epistemic convergence across American culture by providing a somewhat unified source of information versus driving further epistemic divergence and polarization by sycophantically reinforcing everyone's priors in response to leading prompts.

1:07:12.3 SC: You know I don't have deep thinking about this, Zach. I have not really thought about this problem specifically, although stay tuned for upcoming podcasts. But the evidence, such as we have it certainly favors more of epistemic divergence than convergence. I don't think that there's a lot of evidence that AI or really LLMs, large language models, are somehow bringing people to a common truth. There is... It is funny, of course, that people have tried to tune their LLMs to have the same political opinions of themselves like like Grok on X the Everything App. And it fails. Like it doesn't work very well. Like it quickly clicks over into being way further extreme than you wanted it to be. So maybe deep down in the nature of LLMs, which are trained on all of human knowledge, there might be considered some mean field approximation that is going on, right?

1:08:13.8 SC: Like if the LLMs have read everything that human beings have written, maybe they're good at giving the average human being's thought and they would like to stick to that, if you let them stick to it. But also they're trained to be, like you say, sycophantic. Right? Like the LLMs might think something, but if you ask them questions and you make it very clear to them that you want a certain answer, they're going to give you that answer. Believe me, I get... I think I've told this story. But, soon after ChatGPT came out and people began to be able to use LLMs in their everyday lives, someone on the Mindscape AMA asked me, like, has this led to a great increase in the number of crackpot emails that you get with the theory of everything? And at the time I honestly said, no, it really hasn't. And I hypothesized that maybe the kind of person who is a physics crackpot just really wants to believe it's their own ideas and therefore does not ask LLMs. But I was 100% wrong about that. Man, was I wrong about that. These days it's a complete flood of crackpot emails saying, the LLMs and I have worked out this theory and the LLM is telling me it's a great theory. Will you take a look at it? Right? Both on social media and in direct emails and things like that. So that little bit of evidence is making me think that the training of LLMs to kiss the butts of the people asking them questions is much stronger than the training of LLMs to bring everyone to a epistemic convergence.

1:09:47.8 SC: Matthew Wright says, do you have a favorite recipe?

1:09:51.6 SC: Short answer is no. I actually... I don't want to be that guy who just has a recipe and makes it over and over again, is very proud of it. I would rather be the guy who can cook different kinds of things. Right? I'm really not... Like, I don't... I'm not bragging about my abilities here, but I'm trying to be, I'm trying to stretch my abilities to cook different kinds of things, to learn different kinds of techniques and so forth. Of course, everyone who makes things at all in the kitchen has some recipes that are crowd pleasers and you like to make, probably the biggest crowd pleaser I've ever made... Sorry. I should also say, by the way, in terms of full disclosure here, I don't make up recipes myself. So if you're asking, do I have a favorite recipe that I've created? Oh my God, no.

1:10:37.4 SC: I get recipes from out there in the world, from cookbooks of all sorts, from the Internet and so forth. So there's a YouTube video by Heath Riles, who is a barbecue, a pitmaster guy, and he gave a recipe for smoked ribs on the grill, on the smoker, low and slow barbecuing. And I make them and they're super good. I get a tremendous number of compliments on that one. So thank you, Mr. Riles, for that recipe and I have a couple of other recipes. I guess the other thing that I really enjoy the results of is my canelé recipe. A while ago, Jennifer and I, to celebrate one of her birthdays, went to France and visited Bordeaux, among other places. And in Bordeaux you can't help but get canelés everywhere. They're these little pastries, right? Little things with like a kind of slightly custardy but puffy inside and then a crisp, crunchy outside. And they have smell like some mixture of rum and vanilla. Because Bordeaux back in the day was the places where all the... Was the place in France where all the trade ships came from the West Indies.

1:11:48.7 SC: So they would get these exotic flavors like rum and vanilla. And they used them to make canelés. And so if you're in Bordeaux, you get them breakfast, lunch and dinner as little treats. And I fell in love with them and learned to make them. And it's a non-trivial thing because the authentic canelés require copper tins which you coat the inside of with beeswax to let the canelés fall out after you're making them. And that's a level of commitment that I really don't want to get into. It's like 40 bucks for one little canelé pan. So not pan, individual canelé, which is like three bites, right? So the little tin that makes a single canelé is quite expensive. If you're going to make 12 of them at a time, it's a huge investment. So I did search around and figure out that you can also make them in a single carbon steel, appropriately shaped muffin pan. And that's very much more my style. And it turns out to work very, very well. So I'm proud of that one too.

1:12:47.5 SC: Orbital Magpie says, I just listened to the fine tuning episode. I don't do anything remotely related to physics, so I am having a wildly incorrect thought right now. But does it even make sense to talk about fine tuning at all? Because while listening to the episode, the impression I got is that some physical constants have very specific values that defy our intuitions. But why limit to the constants? Isn't everything fine tuned? Why don't we have a completely different set of laws governing the universe? We don't even know what can possibly be in the cosmic hat we're pulling from. So does it even make sense to talk about fine tuning?

1:13:20.6 SC: I think it does make sense to talk about it. I don't think it's perfectly well defined. But I do think that you're putting your finger on something very interesting. Like why in that conversation we were having did we stick with the same framework for the laws of physics, but allow the constants that pick out a specific theory within that framework to vary? And the idea is we're doing our best. That's honestly the answer here. It's not like we have some principled a priori way of saying here are all the possible sets of laws of physics, let's sort through them to see which ones fit our universe and which ones are fine tuned or whatever. We're not that smart to be able to do that.

1:14:03.0 SC: So when we have laws of physics, it's just much easier to say, let's stick with laws that are kind of like what we know, what we've learned over the course of centuries of hard work of theory and experiment to come to. But it's very easy within that theoretical framework to let the constants change. Okay? So it's just hard to get good laws if you... To get good results if you completely change the laws. People like Stephen Wolfram are trying, but I don't think you're getting very far because again, the laws that we have came about because of hundreds of years of really hard work. And it's hard just to sort of throw all that out and start anew and get something very interesting. And if within the laws as we know it, you can point to certain numerical parameters that seem to be fine tuned, that seem to be number one, not what you would have expected, and number two, perhaps just what they need to be to allow lifer to exist, then it's perfectly legitimate to focus on why is that particular parameter like it is, rather than asking the more grandiose question, why are the laws what they are? Not that you can't ask that question. It's just much, much harder to make progress.

1:15:16.3 SC: Cindy Flake says, I've been curious to know what scientists mean when they say that an unexpected result has fallen out of the equation. How does the math in the form of numbers and symbols get translated into a concept that can be expressed in words. In fact, how do they even know what it means?

1:15:33.5 SC: Ah, this is a great question. I like this question and I think that it's one of these things, I think I personally don't necessarily do a great job of explaining. It's just, it's a little bit too familiar to me. Like I've been doing it for too long, solving equations and getting results. So that little bit of translation I don't think I pay enough attention to. So the classic example here would be Einstein and his theory of relativity. So how does Einstein invent general relativity? Well, he already had special relativity from 1905. So that theory came about from the work of a whole bunch of people, Lorentz and FitzGerald and Poincaré and others, and Einstein sort of more or less put the finishing touches on it.

1:16:13.8 SC: Not quite the finishing touches because Minkowski also helped even after Einstein did. But that gave him a framework in which to work. And he was trying very hard to incorporate gravity into that framework. And it eventually took this huge conceptual leap in order to do that, which was to say that not only is there spacetime, not only are space and time unified into four dimensional spacetime, which was part of the special relativity point of view, but that that spacetime can be curved, it can have a geometry that changes in response to matter and energy. And that's a idea expressed in words. And it was Einstein's self appointed task to turn that idea into equations. And eventually he did. So there's a lot of constraints, a lot of rules to govern your actions. You're not just randomly picking equations out of a hat. You want to fit the data. You knew that Newtonian gravity worked pretty well. You want equations that are well posed in the sense that if you give them initial conditions, they will spit out predictions in the future, in a more or less reliable way. So there's a whole bunch of things you want to happen. Okay?

1:17:19.1 SC: And eventually he is led more or less uniquely to what we now call Einstein's equation for general relativity. And right away he checked, does it give the right predictions for the solar system for things like that, and he also made new predictions. He made new predictions for gravitational lensing and for gravitational redshift and things like that. So it was all very good. And he was... He felt good about it. But now that he has an equation, the equation exists independently of him, right? He can't tell the equation what to do. He was guided to get the equation by a certain process of doing physics. And now he has to ask the equation, okay, what are you predicting? Which is a sort of anthropomorphic way of saying, what is the set of solutions to this equation for the geometry of spacetime? The equation relates the geometry of spacetime to the matter and energy inside it. And it was Karl Schwarzschild who, very soon after Einstein put forward general relativity, just a couple of years after, found the Schwarzschild solution, which is the solution that is true outside a spherically symmetric distribution of matter. Okay? So it's what you would use, for example, to predict the motion of planets around the sun.

1:18:33.4 SC: Now, I said Einstein had already done that. That's because he did approximate solutions, and they worked very well. But Schwarzschild found an exact solution to the equations. It had this weird property that there was a certain distance from the center at which things seemed to break down. Neither Schwarzschild nor Einstein knew what to do with that. And so they never figured it out. They both died long before we figured it out. Eventually we figured out what it was, was the event horizon of a black hole. Okay? So the equation was there because Einstein put it there, and Einstein came up with it. And then the solutions to the equation are whatever they are.

1:19:12.2 SC: Like Einstein doesn't have any say over what the solutions are. This very, very important and famous and familiar solution turns out to be hard to interpret because people were very unfamiliar with the idea of what a black hole would be. And in fact, the real difficulty was separating the physical reality of the geometry of spacetime from the specific features of a coordinate system that you use to describe spacetime. That was the thing that even Einstein never became really super good at. And it was people like David Finkelstein in the 1950s who finally figured out how to do it. And to say, yes, this is really a black hole. So the expressing in words is always a back and forth. You have the equations, you try to solve them, then you try to interpret what are the equations trying to tell me. And that process can take a long time, even for the smartest people that we have.

1:20:03.6 SC: Ellen Lubel says, I think you've said that your best guess as to why there is something rather than nothing is that it's a brute fact that the universe has always existed. And so we can't... And so can't we then assume that the universe will always exist? And if that's true, wouldn't we get to live our lives over and over again for eternity?

1:20:24.8 SC: So, not quite, like you're close, but not quite. My best guess as to why there's something rather than nothing is that it's a brute fact that there is something rather than nothing. That's all I would be willing to say. It's not that there's a brute fact that the universe has always existed. That's something that I do not have a strong opinion about. Again, I think the way to answer those questions is not just to try to think about the question for its own sake, but as a good physicist, you try to make a model that tells you what the laws of physics are in that model and tells you what is allowed to happen in that model.

1:21:00.3 SC: And then you're going to compare that model or theory or hypothesis, or whatever you want to call it, to the data, to experiments. Okay? So we have models in which time existed forever. We also have models in which time started at some point. And we have no way of judging which ones are closer to the truth right now. So that's okay. I think that either one is completely compatible with the idea that there's a brute fact of the existence of the universe. I think that when it comes to the universe existing with an earliest moment, there's a really difficult mental block that people have because the idea of time having a first moment, of time beginning, is completely alien to our everyday experience. So they phrase it as something came from nothing. And that's just wrong, that's just misleading, because that assumes there was something before the universe.

1:21:57.7 SC: And we're going to call that something nothing, and that nothing turned into something. And none of those words make any sense whatsoever. The correct way to say it is in this set of ideas, there was a first moment of time. There was a moment before there were any other moments. When you say it that way, like, there's nothing objectionable about that. Sure. You know, maybe time goes forever. Maybe it had a beginning, maybe it'll have an end. I don't know. But all of those are possible. And none of that says anything about whether the universe will be eternal toward the future. Maybe it will be, maybe it won't. And none of that says anything about whether we're going to live our lives over and over again for eternity. I can very easily imagine scenarios in which the universe is eternal but never repeats. You know, the example I like to give is the integers. The integers zero, one, two, three, four, minus one minus two minus three, minus four, there's an infinite number of them, and no two of them are the same. There's no repetition there. The fact that something goes on for infinity is not a prediction that it repeats itself over and over. It can just be different every single time.

1:23:04.8 SC: David Sotolongo says, following up on your conversation with Cass Sunstein, how do you think a liberal democracy should deal with illiberal or anti-democratic voters or candidates for office? I understand that in Germany, for example, political parties and even private organizations are unlawful if they are aimed at undermining the democratic constitutional order. Is that something you would like to see in the United States? That is, should the courts or some other body be able to bar someone from running for office on the basis they are anti-democratic or otherwise illiberal?

1:23:38.1 SC: I think that here is a case where we would all like to have hard and fast rules, but we're not going to get them. I think some good old human practical wisdom is going to be called for in these questions. In Germany, there was a very good reason to have specific laws against anti-democratic candidates. Namely, that they had a phase they went through with a super totalitarian regime in charge that did really, really terrible things in the not so distant past. And the people who lived through that, many of them are still alive. Right? And so that makes sense. That's a specific situation. It's not universal. You have to actually take into consideration the real facts on the ground. Here in the United States even before the American Revolution, even before, we got our independence from Great Britain, Great Britain was never that tyrannical.

1:24:31.3 SC: I mean, a little bit tyrannical, but not like super tyrannical to the extent that the Nazis were. So there was already a tradition in the United States of having elections and voting for people and things like that. And so it was more forgiving, naturally it made sense to be more forgiving in terms of allowing people to have whatever crazy opinions they have and let the democratic process handle them. The real danger, of course, is that when you say, could you bar someone from running for office on the basis that they are anti-democratic or otherwise illiberal, how do you know that? Who's judging this? Right? Do you have like a ray that you point at someone, a little ray gun that says, oh, this person has anti-democratic thoughts in their heart of hearts? That's just really hard to do. So I think that the strong presumption would be not to do anything like that.

1:25:23.0 SC: But the facts on the ground can change, right? Like if we do get to a point where there is a constant real danger of democracy collapsing, then maybe we do need laws like that, and you can very sensibly argue that maybe we're already at that point. I think that we're at a point my personal view would be right now we're absolutely at a point where democracy is endangered in the United States. But I don't think it's quite systematic. I don't think it's... I think it's a weird anomaly right now. That is my personal thought. I think it's a very serious anomaly. I don't think that we should pretend it's going to go away. We should deal with it. We should do what we can to shore up democratic norms and institutions and values and all of those things. But I don't think we're at the point yet where most Americans are fundamentally anti-democratic by inclination. I could be wrong about that. But I do think that you actually have to think about the situation, not just invent once and for all rules and apply them and expect them to fix all your problems forever.

1:26:28.0 SC: Sandro Stuckey says, I know you discussed this before, sorry, but I just don't get it. How can the inhomogeneities in the cosmic microwave background be caused by quantum fluctuations in the early universe? From a many-worlds point of view, how can there be decoherence in a situation like that? If the universe is the system... Is the universe the system or the environment, or both or is the division somehow not relevant in this case?

1:26:51.3 SC: This is a good question, actually. So just to give people some background here, there's the Copenhagen interpretation of quantum mechanics, and there is the many-worlds interpretation. There's others too, but let's put them aside for the moment. In the Copenhagen interpretation, you have a wave function that tells you the probability when you make an observation that the wave function is going to collapse to certain values. In the many-worlds interpretation, you have a wave function that just obeys the Schrödinger equation forever. But what happens is that individual quantum systems that are in a superposition get entangled with their environment and the universe branches into multiple separate worlds. And in each world, that system that was in a superposition now has a definite value of one way or the other.

1:27:37.1 SC: And in order to make that former thing work, the Copenhagen interpretation, you need an observer. In order to make the latter thing work, the many-worlds interpretation, you need an environment, the rest of the universe for the system to become entangled with. In some sense, neither one of these are obviously available in the early universe, right? So I'll be honest with you, a lot of cosmologists basically cheat when they think about quantum fluctuations in the early universe. They pretend there's an absorption observer there, and they pretend that the wave function collapses and they treat what is really a quantum mechanical state, as a fluctuating classical variable. And honestly, you really shouldn't do that. But they do it anyway. But sometimes it works, and in this case it basically works. If you are a many-worlds person, you can't just say, well, there's a mystical observer doing the collapsing of the wave functions. You have to divide the system into the variables you're keeping track of and the environment. And in the early universe, it's not that you can't do that. It's in fact, there are too many candidates for what the environment is. Very, very small scale fluctuations are things you don't keep track of.

1:28:48.4 SC: Individual photons, right, you don't keep track of. And those can be part of the environment. But also things outside your observable horizon can be part of the environment as well. I actually wrote a paper once with Kim Boddy and Jason Pollock where we asked this question in the context of eternal inflation. But other people have asked it, how does decoherence happen with regular old quantum fluctuations in the early universe? And the answer is, basically the casual don't pay attention kind of approach gets you more or less the right answer. Okay? So one way or the other, the answer that people predict is actually the right one. I mean, in my heart of hearts, I am hurt by the fact, like we were very hopeful that we would get a different answer for eternal inflation by doing it right as opposed to the way everyone does it. But the sloppy approach actually turns out to work pretty well.

1:29:41.8 SC: Kevin O'Toole says, in your interview with Steven Pinker, you discussed Aumann's agreement theorem, saying that rational people with common priors didn't need to share evidence to reach agreement. They just need common knowledge of each other's posterior probabilities. However, the leap from knowledge, some knowing, sorry, knowing someone's posteriors to having common knowledge of someone's posteriors is a big one. And I don't see any practical way of making that leap without sharing the underlying evidence, to be concrete, say I've seen a coin flip heads twice and have a 70% posterior of the coin being weighted toward heads. So Kevin is suggesting that we don't know if it's a fair coin or not. All we're seeing is a couple of outcomes of its being flipped and we're judging the probability that it's a fair coin versus unfairly weighted towards heads. And he says, someone else might have a 60% posterior for three very different reasons. They might have seen only one of the same flips I saw, they might have seen an independent flip, or they might have seen many more heads and tails. In any example I can think of, I don't see an easier way to distinguish these cases than telling each other our exact evidence. So my question is, in the real world, do you think that there are cases where the conversational dances you discussed can shortcut the evidence, determining whose evidence is stronger or less redundant without stating it fully and directly?

1:31:00.2 SC: So I'm not sure that I'm fully equipped to answer this question, and maybe I should have ignored it for that reason. But it's a perfectly good question. I get it. I've had questions like this myself. I think there's two things going on. Like, first, you could imagine that the two people talking to each other for whatever reason have 100% faith that each other are 100% rational creatures. Okay? So in other words, your personal prior that the other person is doing their best and telling the truth, could be 100% or very, very close to 100%. If you don't want to give 100%, 99.9999%. And basically in that case, when they tell you their posterior probability, their updated priors, you should believe they have a good reason to do that, right? Even if they don't give you the evidence, you should believe that they've done their best Bayesian job, and then that fact should lead you to update your priors.

1:31:57.2 SC: It's nothing more complicated than saying, if I am confused about a particular issue and I'm not sure which way it goes, and I meet a smart, informed person and they have a strong opinion one way or the other, even if they don't tell me why, I should probably say like okay, that increases the chances that I believe their point of view than the former confusion that I had. The other thing is, even if you don't have that strong assumption that the two people believe each other are perfectly reliable, et cetera, that's the dance that you're talking about. They can talk about things back and forth, and I'm pretty sure that one of the things they're allowed to talk to is to share their experimental data to explain, oh, well, I saw this, and that's why I believe this. And they go back and forth and once they converge, they come to agreement. I don't think there's any rule that says they can't actually reveal the data they have. So... But again, this is not actually my area. I'm not completely sure. That's just my impression.

1:33:00.5 SC: Joe Bender says, I recently read about a spectroscopy result from the James Webb Space Telescope that may confirm the observation of dark stars in the early universe. Sorry, I should emphasize, dark stars. That's the idea we're discussing here. I don't think that I've heard you discuss dark stars before on the podcast, with the obvious caveat that these are still preliminary results. What credence do you assign to the existence of dark stars? If they do exist, what implications would they have for other questions like the nature of dark matter, the origin of supermassive black holes, et cetera? If they were observed, would that provide new constraints for any of the current hypotheses on the composition of dark matter?

1:33:35.8 SC: So the idea of dark stars, I'm not super familiar with it. I know a little bit about it. Again, I'm not... It's not really something I followed closely. It's, if you think about the very early... Not the very early universe, like seconds after the Big Bang, but in between the time of the cosmic microwave background radiation 380,000 or so years after the Big Bang and the time of the first bright stars lighting up, okay, the birth of the first galaxies, there is a lot going on, but we can't see it directly. It's the so called dark ages of the universe. And since there haven't been any stars yet in the universe, the ordinary matter in the universe is overwhelmingly hydrogen and helium. Okay? It's no carbon or heavier elements or anything like that. And those heavier elements, as far as we can tell, play an important role in helping stars to form in the current universe. And our old universe today, 14 billion years after the Big Bang, is sort of contaminated by the remnants of a whole bunch many generations of exploding supernovae and other stars with solar winds and things like that spitting higher atomic number elements into the interstellar medium that then get recycled into new generations of stars.

1:34:50.1 SC: But the first generation didn't have that benefit. So it's actually a little bit harder to understand how the first generation of stars formed. The dark star idea... The other thing I should say is, for ordinary stars that we know, like the sun and the other stars we see around us in the Milky Way, we don't expect dark matter to play any strong role at all. And the reason why is because dark matter doesn't dissipate. It's dark, it doesn't collect... It doesn't interact with photons. When two dark matter particles come very, very close to each other, they don't bump into each other and emit photons and lose energy like ordinary matter does. So rather than collect in very, very concentrated regions of high density like a star in an otherwise empty space, the dark matter becomes a little more dense in some over dense regions and we have what are called micro-halos, but there's still these micro-halos much bigger than the solar system is. Okay? So it's nothing like a dark matter cloud is actually attached to the sun or anything like that as far as we know. There could be a few dark matter particles, but it's not in any way a dynamically important part of the energy density in the solar system.

1:36:03.1 SC: But in the early times it might have been different because it was hard to make stars. There might have been cases where there was a common concentration of a little bit extra dark matter and a little bit of extraordinary matter and they helped each other to make a star. And that's basically the idea of dark stars where you have both dark matter over density and an ordinary matter over density, generally much bigger than ordinary stars, both in total mass and in total distance. But what can happen is, if you have the right kind of dark matter particle, these dark matter particles can be concentrated in the middle of this star, "star", and enough that they actually annihilate a little bit and they give off radiation and that affects the glow of this dark star, such as it is.

1:36:50.0 SC: And that might be involved in what JWST is seeing. To me it sounds all a little bit far fetched personally, but again, I haven't looked at it carefully. What JWST is showing us about the early universe is surprising in various ways. I know some people want to claim that it's like evidence against the Big Bang or something like that, which is complete nonsense sense. It is evidence that there's interesting things going on in terms of star formation and galaxy formation and maybe even black hole formation, all of which is great and we'll, learning something about that. To me the dark stars idea is a little speculative, but of course if the data says that it's there, then I'm going to absolutely be happy to change my mind.

1:37:32.2 SC: Mike VR says as I've tried to understand how patterns persist in biology and social systems, anti-fragility seems essential. Systems that actually improve through stress. But I don't see anything like that in galaxies or rivers. Do you think this capacity to benefit from stress is unique to living systems? Or would it... Could it appear in physical ones too?

1:37:53.1 SC: I guess I'm not completely sure that the existence of anti-fragility is a objectively verifiable thing. I mean, you can get a sense of something like that. I know that I don't really improve through stress that much. I mean, maybe you can find a new and more effective mode of being if you're stressed in the current situation that you're in and you're being forced to do it, right? You know, a trial by fire kind of thing. But also stress can hurt you and make you bad. I don't just mean human stress, I mean stressing any kind of system at all. So I think that it would generally be a mixed bag whether or not any system would improve through stress.

1:38:33.2 SC: Now, I do think that there's something going on where living systems have capacities to manipulate information and plan and make modifications in themselves, or adapt in the broader sense to changing circumstances in ways that nonliving ones don't. I just think that living systems have that sort of information gathering and utilizing capabilities that exist but are pretty primitive in nonliving systems. It's an example of the complexogenesis I talked about in the podcast a little while ago, the increasing sophistication of ways to take advantage of the non-equilibrium nature of the world in which we're embedded, which gives us access to various kinds of information. So I think that the general rule is that living systems are just much more sophisticated than nonliving ones at that. But I don't know if I would quite put it through that anti-fragility lens.

1:39:31.0 SC: Nick B says, this year's DESI results, D-E-S-I Dark Energy Spectroscopic Instrument, I think, are raising questions that we thought were answered. Is dark energy evolving? Could the universe's rate of expansion be slowing, et cetera. I came to pop physics about a decade ago, and so the majority of what I read and heard presented heat death as the universe's inevitable future. It's disorienting for me to experience such a rug pull, so I can't imagine what it feels like for people who are actually working in this field. If we're heading towards the Big Crunch, even if it won't affect us directly, it can feel like we're losing a shared and familiar future. A universe where there's enough time for protons to decay and black holes to evaporate and now there's not. Billions of years instead of trillions means there's much less time for interesting things to happen. Do you feel grief at having to abandon such a widely embraced feature of modern cosmology? And is it in any way compensated for by the relief that in a Big Crunch future Boltzmann brains will never pop into existence and no longer haunt your AMAs?

1:40:26.9 SC: So just to be clear, I don't think that anyone ever thought that we were certain about the future of the universe. Certainly I never was. Certainly none of my professional physicist friends ever were. Depending on where you get your pop physics, maybe people were a little bit overly simplistic about what the options were. It's completely okay to think that one option is more likely than others and yet also think that there are other options that are possible. That's just the way science goes. I would still right now think that the most likely option is a cosmological constant rather than evolving dark energy. But I could be wrong, and I'm very happy to update my credences if the evidence keeps coming in. You know, when these first bits of evidence for something very dramatic come in, you shouldn't latch onto them too strongly.

1:41:14.3 SC: They can go away. They usually go away, right? Some of them will stick around and you can say, oh, look, you should have paid attention to that all along. But you don't know ahead of time which ones are going to stick around and which ones are going to go away. In terms of the future and the big crunch and De Sitter Space and heat death and things like that, I have zero emotional investment one way or the other. I have a personal down to earth investment in the sense that I spend some of my time doing research that takes as an assumption, one theory or the other. I've done research on dynamical dark energy, on constant dark energy, on all sorts of things, and I don't want that research to be a waste of time, I guess, so I hope that my choice of what to work on is reflected more or less accurately in what the world actually is going forward. But in terms of, you know, billions versus trillions of years, no, I absolutely do not matter. I'm just very curious as to what the universe does. I would like to know one way or the other.

1:42:13.9 SC: Julio Contio says, what is your opinion on storing or freezing one's genetic material so that a new you can be created, trained, and perhaps continue the work the current you is working on? The idea came from the foundation series in the Emperor's Multiple Clones.

1:42:27.3 SC: I think that is a very silly idea. For one thing, your genetic material is no real near all of you. Therefore, cloning yourself using your genetic material doesn't make a new you. It just makes the equivalent of an identical twin, less than identical twin really, because presumably the environment in which the clone grows up is more different than the environment in which you grow up than your twins environment would be. And twins don't do the same thing. They're not exactly working on the same problems, doing the same jobs or whatever. So if someone made a clone from genetic material of a scientific researcher, the chances that that clone would carry on the work of the original seem pretty slim to me. I think it's much better just to think that there's, you know, your graduate students will carry on the work that you're interested in, and that's a much cheaper and more cost effective way of getting it done than saving your genetic material. I think that you have to appreciate that actually both nature and nurture do matter for these kinds of things.

1:43:32.3 SC: Gerard Sage says, lately I've noticed that complex societal systems made of human agent interactions often behave as if they themselves are agentic. A market responds to economic... To the economic environment and alters its behavior to grow, even if most people think it shouldn't, fossil fuels, et cetera. I've been calling this quasi-agency a nod to quasi-particle emergence from particle interactions. And I worry that quasi-agents threaten the utility of democracy. After all, humans do things we disagree with for systems larger than ourselves all the time. Do you have any thoughts, either broadly or from a physics of democracy perspective about quasi-agency and the validity of this concern?

1:44:10.6 SC: Well, I certainly think that collections of human beings can function as emergent actors in the world, whether they're nations or corporations or fandoms or whatever. We always talk about groups of people as if they have common goals and things like that. And then you look at it more closely, you realize, well, they don't really have exactly common goals. But then again, you look at individual humans more closely and you realize that they don't even have a set of coherent goals. So maybe that's not a difference between individual humans and collections of them. What I guess would be the warning I would suggest is that even though collections of human beings can certainly be thought of as agentic in some sense, that doesn't mean they're just like humans. I think that it would be important if you wanted to go down this road to understand what are the similarities and differences between individual human beings and groups, complex social systems made of human agent interactions, they might not respond in the same way to incentives or to warnings or whatever. They might not communicate with each other the same way. They might move on different timescales, they might have different sets of concerns, all these things, so I think it's important to recognize the existence of multiple layers of emergent phenomena, but also important to give them their own due and rather than sort of stuffing them into a box from what we learned about some different level.

1:45:41.9 SC: Gag Halfront says, we haven't had any cat news in a while. How are the kits?

1:45:47.6 SC: The cats are great. Ariel and Caliban are great. Puck, as you know, has a new home in New York City. We get occasional photographs of Puck. He's doing wonderfully, as far as I can tell. He is a absolute charmer. That Puckster has an enormous amount of personality in him. And I'm very glad that we found him a good home. Having said that, Ariel and Caliban are thrilled that he is gone. And even though he was just in a room, in a very large room in our house, so he had his own little domain, when Puck was here, that became part of the house that Ariel and Caliban were not allowed to count as their own anymore. And they knew he was there. You know, they went through their day, but they were not at all happy about it. And the three cats did not get along.

1:46:27.8 SC: So Ariel and Caliban are now much, much happier that they have the house back to themselves. And Caliban in particular, like, he's a pretty robust cat, super healthy, he and Ariel are brother and sister, of course, and they came out of the same litter. And Caliban was the healthy one of the litter. He was always the one who ate the most food and everything. But he's kind of a hot house orchid, you know, when Puck was here, he basically went on a hunger strike a little bit. And he lost considerable weight, like, noticeable amount. Not an unhealthy amount, but he was a little bit of a chubster. And then now he's very, very skinny, so. But he's put on a little bit, so he's coming back. He's bouncing back to normal. And anyway, since I'm actually teaching every day of the week this semester, I'm here a lot and Jennifer's here a lot. There's no traveling going on. As far as the cats are concerned, this is the best thing in the world. So they're doing great.

1:47:19.7 SC: Anthony Rubo says, in a previous AMA, you were asked if you believe emergence arises ultimately from a fundamental or is it... If it is emergence all the way down? You answered you believed the former mostly because you literally cannot imagine what it would be like to have emergence all the way down. You also mentioned that we should leave our minds open to both possibilities. This got me thinking, as you are very comfortable with imagining infinities in other areas, be they time, mathematics, or other physical properties, what is conceptually different for you about this type of infinity?

1:47:52.1 SC: Well, I will emphasize that we don't know what's going on in terms of what it is all the way down. But there is something different that happens when you look at very, very small scales in the universe. Simply because of quantum mechanics. In classical mechanics, which governs our macroscopic world to a pretty good approximation, things can be any size, right? There's no relationship between what a thing does and how big it is intrinsically. You can just imagine scaling the whole universe down to some really tiny size, it wouldn't matter. In quantum mechanics, once you get down to very small scales, things start changing their behavior. Things start having wavelengths that kick in. The Compton wavelength, in particular, is a particular kind of quantity that we have associated with a particle.

1:48:41.4 SC: And it's just the mass of the particle in natural units converted into a wavelength. And the way it goes is that the lighter the particle, the less massive the particle is, the longer its quantum wavelength is. Okay? It's Compton wavelength, I should say. Sorry about that. So in an atom, for example, the Compton wavelength of an electron is much bigger than the Compton wavelength of a proton or a neutron. And the relevance of this is the Compton wavelength is basically the smallest wavelength into which you can squeeze something such that it actually acts like a single particle. If you take a quantum wave of a field corresponding to a particle of a certain mass, and you squeeze it into a box tinier than its Compton wavelength, then it starts not behaving particle like anymore. Basically, particle antiparticle pairs pop out of the vacuum, and you get a superposition of many, many different sets of particles.

1:49:37.1 SC: And that's basically because you're saying that the energy to make a new particle is less than the energy that is already in the wave because you've squeezed it into this tiny little box. So, for example, in a proton or neutron, the masses of the up quark and down quark that make up protons and neutrons are smaller than the mass of the proton and neutron themselves. And that means that the Compton wavelength of an up quark or a down quark is bigger than the proton in which it is combined. Oh, my goodness. How can this happen? This is all stuff I discuss in the 'Quanta and Fields' book. The answer is it's not very particle like, right? The interior of a proton is not really three quarks orbiting each other. It's actually a very complex soup of quarks and antiquarks and gluons all interacting in certain ways. It's a static state. It's not literally moving, but it's a complex superposition of all those possibilities. So because of this effect, because for every field. Well, let me just add the final thing, for a big macroscopic thing like a baseball. Baseball has a Compton wavelength, but it's incredibly tiny.

1:50:47.9 SC: It's much smaller than a proton, so who cares that it has a Compton wavelength. So basically, if quantum mechanics has any fundamental purchase on the lowest layers of reality, there is kind of a smallest distance you can get to, right, without making things really, really heavy. You know, the Planck mass, which is considered super duper heavy by particle physics standards, is about 10 to the minus five grams. So by macroscopic standards, it's almost nothing. It's very, very light by macroscopic standards. But that's huge by particle physics standards, right? It's many, many trillions times the mass of a typical atom. So if you try to imagine a world made out of things that are lighter, less massive than the fundamental particles we know and love, right, with electrons and protons and neutrons or the quarks that make them up, then you can't squeeze them into small areas. There's sort of an uncertainty principle says you can't be low mass and literally tiny in size at the same time. That makes it very difficult to imagine that there are layers beneath the layer that we know and love right now of quantum fields and electrons and the standard model and so forth.

1:52:00.9 SC: There could be layers beneath in the sense that there's a completely different kind of thing. There are some abstract quantum degrees of freedom that give rise to the existence of space and time and things like that. But it wouldn't be the simple sort of Lego model where you just have some more ingredients coming together. There's almost no reason to think that electrons could be made of smaller things. That's very, very difficult to imagine. If they're lighter than electrons, they can't fit. If they're heavier than electrons, then the electrons should be heavier. Those are basically the choices. That's not that physicists aren't clever enough to think of ways to make it happen, but it's not exactly the most promising scenario going forward.

1:52:41.4 SC: Mikhail Sirotenko says, is your default hypothesis about time that it is emergent? If not, why? Based on your solo episode, it seems that it's totally possible to experience time in a timeless wave function. So should we just assume that a model of the universe without fundamental time is a simpler and cleaner explanation?

1:53:00.8 SC: No, we should not assume that. We should admit that we have no idea right now whether time is fundamental or emergent. Those are both perfectly good options, and we shouldn't leap to a conclusion before knowing what the right answer is. I think that one can make arguments right now in the current state of our knowledge why time being emergent is a more sensible thing. One can also make arguments right now why time being fundamental is a more sensible thing. So until one of those sets of arguments wins or we understand how to overcome the other set, we can just say it's okay. We're exploring both possibilities. Again, that's... As I just said a little while ago, that's how science goes. You don't just say, well, you know, I think it's probably going to be this way, so I'm going to assign essentially full credence to it being this way. I think you have to admit, when you just don't know, and keep that in mind and move forward on that basis.

1:53:53.8 SC: Ed Saidstuff says, recently I've become interested in chemistry and the vastness of combinatorial space. The number of potential molecular configurations is thus astronomical, yet only a tiny fraction exist at all, and even fewer participate in biology. Do you think this asymmetry hints at a deeper selection principle in physics or information theory that guides chemical organization toward life? Rather, is it just a feature of the world we live in?

1:54:19.3 SC: So I'm not exactly sure what is the asymmetry you're pointing at here. Yes, it is true that only a tiny fraction of molecular configurations is used in the universe, but that's almost necessary because if you calculate the number of molecular configurations, given that there are molecules, like organic molecules, that can basically be arbitrarily long and arbitrarily complicated. I mean, genomes alone, right, like DNA molecules alone, can be essentially infinite in configurational space. Therefore, it is completely unsurprising that most of the possibilities that we can imagine are not actually actualized in the universe. There's sort of no choice about that. We live in a finite universe, at least in the part of the universe that we can observe is finite and there's a finite amount of matter in it, and most of it is hydrogen anyway, so, yeah, we're not going to come anywhere close to exploring the entire space. I do think that there are implications for that fact, you know, as evolution happens, or as complex systems adapt and evolve more generally, we're not going to do anything like a comprehensive scan of all the possible spaces of molecular configurations or genomes or whatever you want to call them.

1:55:37.1 SC: And what we do is sort of very, very partial, you know, what Darwinian evolution or synthetic methods do is going to be necessarily very, very incomplete in terms of exploring the space of possibilities. I don't think that's anything good or bad. I think it's just a fact. And what we should do is try to understand how it actually does work. In a way, Darwinian evolution is quite clever in exploring that space, having sort of both the feature of selection by adaptation. Right? If something is working well, you tend to reproduce more and spread it through the gene pool, the population, but also randomness, randomness through sexual mixing and randomness through mutation. So you're kind of generally moving toward what is a better fit to the fitness landscape you're in, while also taking some resources and parceling them out towards exploring combinations that might not a priori be better, but you don't know because you haven't tried them yet. Right? That's a pretty clever mechanism. It's not to say it's the most clever mechanism. I think that genetic algorithms, for example, in computer science have been shown to be kind of interesting, but not the best way to generally solve a problem, which is an interesting feature of the world. So who knows, maybe we'll come up with more and more clever ways as we go on.

1:57:02.5 SC: Schleyer says, how do you decide where to publish your papers? This is a thing in scientific publishing or academic publishing more broadly, which, by the way, is less settled in historical terms than you might think. You know, the whole scientific publishing enterprise never was settled into one mode of operation. It's kind of constantly churning and changing and things are very different. When Einstein was writing papers later in life, in the '30s and '40s, it was sort of during the shift where important journals like the Physical Review started refereeing papers. It used to be you would send a paper in and the editor would decide whether it gets published or not. The idea of sending it to an anonymous referee is relatively recent in time. And Einstein didn't like the idea, infamously. But anyway, there's still ongoing revolutions. The existence in physics of what is called the arXiv A-R-X-I-V absolutely revolutionized physics publishing, because now almost everyone puts their papers online for free before you send it to the refereed journal. It is extraordinarily frustrating to me and to other physicists that other scientific disciplines don't reliably do this.

1:58:25.1 SC: And there's a whole bunch of interesting papers that are still behind paywalls for no good reason at all. It just drives us batty. Like why are you doing that? You're doing all this work to be a scientist. You're writing up your work and you're hiding it from people. There's no reason why someone should have to pay a $100 to read your science paper. So I think that things are going to be very different going down the line. And nevertheless, because people haven't caught up to what the physicists are doing, some areas have, but many have not. The existing system says that there are journals that are sort of more prestigious than others. Right? You get more points for publishing in this journal than another one. And again, in my world that just doesn't really exist. Like sure, Physical Review Letters or maybe Science or Nature are a little bit more fancy than other journals, but come on, who cares? We're actually about, is the paper any good? That's actually much more important than what journal it gets published in. I actually have a slight preference for the American Physical Society journals like Physical Review, Physical Review Letters, et cetera.

1:59:31.6 SC: Because on the one hand it's not like some predatory private corporation that is charging exorbitant rates, it's the American Physical Society. On the other hand, it's more or less established. They're not going to go away, 50 years from now, they won't have disappeared. I kind of worry because there's a whole bunch of sort of small fly by night journals which might grow into super prestigious journals down the road, but right now it's some person in their living room with a computer running a journal. And you can publish there just as much much as you could publish anywhere else. And kind of nobody cares. But I would worry that it's not sustainable in the long term. So that's why I like Physical Review to publish things. But honestly, it doesn't matter. Some of my papers, like I just for whatever reason I was busy, things were going on in my life, I never published them. They're on the archive, people can read them and cite them. The science is still there. So who cares whether it gets published or not, where it gets published, any of that stuff. If I'm collaborating with students or postdocs, which I often am, then it matters a lot that we get them published because they want to have a CV and apply for jobs and stuff like that. Happily I am past the point in my life where I have to need... Where I need to worry about that kind of thing.

2:00:43.7 SC: Thomas Henry says, the transformers may have first coined the term atechnogenesis. But in what sense can the current evolutionary phase transition from the biological to the technological be compared to abiogenesis, that is to say, the origin of life? And to what extent is it accurate to think of something like NASA's Perseverance Rover as having evolved?

2:01:05.6 SC: Well, I don't think it's accurate at all to think of the Perseverance Rover as having evolved. I think that you have to treat it differently. It... In some sense, in some sufficiently broad sense, it came about as a result of biological evolution, namely the very kind of cheap sense in that it was constructed by human beings and human beings came about as a result of biological evolution. Just like the pad of paper in front of me evolved. But the thing itself didn't evolve, I evolved. I was a... Or the people who made the paper, I guess, evolved and they made the paper. And I think that that's what the Perseverance Rover is like. I think what happens is that when life organisms, living organisms, reach the point where they can construct elaborate technological artifacts, then that's a new thing, right? That's not biological evolution.

2:01:56.1 SC: That is something that has appeared on the Earth at some time as a consequence of biological evolution. But it's different. And it's different among the many reasons, not to be too obvious about it, but things like the Perseverance Rover do not actually reproduce. They certainly do not have selection pressures on their reproductive success like biological organisms do. They are designed. That's the most important difference. And what I mean by that is evolution doesn't look ahead, right? Biological evolution just mixes up the different genomes, has little mutations here and there and sees what works. That's all it ever does. Biological evolution never says, oh, you know what? The climate is changing. We should evolve in a way that is more adapted to a different environment, right? That's not how it works. It waits until the climate does change and then it adapts to it a little bit. Whereas something technological which is designed by a human being, you can think ahead, you can run scripts in your brain or on your computer that say what are the conditions that this thing is going to have to adapt to?

2:03:07.4 SC: So that gives us a different way of coping with the future than biological evolution is able to do. On the other hand, of course, biological evolution has the enormous advantage that it's been working for billions of years and precisely because it's not designed for any particular purpose. It often ends up, biological organisms often end up being much more robust and general purpose than technological artifacts tend to be. Because when we design something technologically, we design it for a purpose. And sometimes they break and then when they broke, they're just broken, right? Because they... Their purpose is no longer fulfillable. Whereas biological organisms are resilient, they can bounce back because precisely because they're not designed for one specific purpose. And they're very much designed or they're very much adapted to self repair and survival and perseverance in all sorts of environments. So there are pluses and minuses both ways. But of course, the final thought is that the technological aspect of this thing is just beginning, right? You know, we're still at the start of the emergence of technology in this sense, so who knows where it's going to go. I mean, maybe there'll be a lot more adaptation and evolution in the technological side of things once we get good at it, but we're not quite there yet.

2:04:30.0 SC: Al X, asks a priority question. Remember, the priority questions are things that every Patreon supporter gets to ask once in their life. And I will promise to do my best to try to answer it. So Al X asks, as a former international physics olympiad medalist and current physicist, I noticed that the urgency with which me and the other competitors were studying and solving problems was far greater than that of most PhD students and researchers I know, especially those who didn't participate in competitions when they were younger. I believe introducing more of a competitive environment would boost productivity in science similar to what World War II and the Cold War did. What do you think is the best way to increase urgency in science, in physics especially? Is it introducing a major olympiad hackathon where the event will be live streamed and the winners get the grant money everyone applied for?

2:05:19.5 SC: No, I do not think that at all. That sounds not at all like what would increase actual science, scientific productivity. There's different kinds of science going on. Okay. If you want to write a program to do something that is very well defined and you know what it is, then maybe a hackathon or something like that is useful. If you want to build a device where everyone knows what the device is and everyone knows more or less how to build it, and you just want to get it built, then that kind of urgency might be helpful. If you want to think deeply about nature and come up with a creative new idea about how the world works, then that kind of urgency and competition is the worst possible thing that you could imagine. You need time. You need the opposite of urgency.

2:06:04.6 SC: I think that the real problem right now is that there's too much urgency, if anything, people want to write a lot of papers. People write many more papers as scientists now. The rate of paper production now is much greater than it used to be. And part of that is competition. People trying to increase their publication lists and get hired for jobs because the money is very hard to come by. The positions are very hard to come by. The World War II and the Cold War boosted scientific productivity because money was free, money was... Well, it was a lot more available. There were a lot more jobs, people were being hired, there was grant money and things like that. There were all these big projects going on, that will give people the freedom to take time to think and do their jobs and be creative. And I think that's much more important to really important scientific progress than a sense of urgency or competition might be.

2:07:01.0 SC: Brian Mendoza says, while listening to the 'Quanta and Fields' on atoms, that's a chapter in 'Quantum and Fields', I found myself disappointed to hear that anti-gravity devices are likely to be impossible. The anti-hydrogen experiment at CERN recently confirmed that antimatter falls downward. So no joy there either. In the back of everyone's minds is a Star Trek fantasy for humanity. But reality seems to say otherwise. We will not be traveling to the stars. My question is, do you think that this is actually a good thing, because nature seems to prevent civilizations from interfering with one another?

2:07:32.5 SC: So I do agree that we're not going to do some clever physics thing to enable faster than light travel or warp drive or anything like that. I could be wrong, as I've been saying throughout the podcast, but that's my very, very strong credence on that. I see zero reason why this implies that we won't be traveling to the stars. People are very impatient. We can travel to the stars. It will just take time. Okay? We'll have to learn to be patient enough to put up with that. The idea of technologically solving longevity is infinitely easier than the idea of technically solving warp drive. Warp drive literally violates the laws of physics. But the idea of either putting people in storage or just giving them lifespans that stretch for tens of thousands of years, that's not against the laws of physics at all. That's a hard engineering problem that we're nowhere close to solving right now. But then again, we're nowhere close to building starships right now either. So it's not especially urgent. So I think that there's no obstacle to eventually traveling to other stars. It won't be cheap or easy or quick, but it will be possible. Do I think it's a good thing? I don't know. There is the Fermi problem out there. It's not really a paradox. But there's a fact that we're not overwhelmed with alien visitors right now. And whatever the explanation for that is, it might have implications for whether it's a good or bad thing to run around the galaxy visiting other star systems.

2:09:01.7 SC: Rad Antonov says, what exactly is entanglement entropy and how is it calculated?

2:09:09.7 SC: I realize now that I should have talked about this back up, I should have grouped this with the question on what a density operator is. But that's okay. I put it down here. Entanglement entropy is the entropy associated with the subsystem of a quantum mechanical system when that subsystem is entangled with the rest of the universe or with the other subsystem in whatever system you're looking at. So it's a unique thing to quantum mechanics. It doesn't exist in classical mechanics. If you... There is something that is completely analogous to the classical system, which is that if you just have a single system so it's not entangled with anything else, it can either be in a pure state and have zero entropy, or it can be in what's called a mixed state and have non-zero entropy. That's exactly like a classical system. You can either in a classical system know exactly what its microstate is and it has zero entropy, or have a probability distribution over its microstates and it has non-zero entropy. The formula in quantum mechanics for that entropy that a quantum system has is called the von Neumann entropy.

2:10:09.9 SC: It's minus trace rho log rho, where rho is the density operator that we were talking about before. So that's how it's calculated. If that's a meaningless statement to you, minus trace rho log rho, then I don't know what to tell you except to read a book on quantum mechanics. And it's not a very elementary thing, right? You have to go through some effort to get there. There is a wonderful book by Schumacher and Westmoreland. Benjamin Schumacher is the quantum physicist who invented the word qubit. So he's famous for that. But he also wrote a book, and it's a textbook on quantum mechanics. And the book is called... I'm not going to get it exactly right, but it's like Quantum Information Processes... 'Quantum Systems Processes and Information', something like that. The reason why I'm mentioning it is because it takes a very unusual angle on being a quantum mechanics textbook by starting with information and density operators and entropy and things like that. So if you want a shortcut into it and are willing to do the math, then that would be a good place to start. But what it means morally, the von Neumann entropy is how far you are away from being in a pure state.

2:11:21.1 SC: If your system is in an unknown state, then maybe you're better off describing it with a density operator and you can calculate its entropy. The bigger the entropy, the less pure the state is. The entanglement entropy in particular, as we mentioned before when we were talking about density operators, comes about because there's a completely new thing in quantum mechanics where you can have an overall system in a pure state, so you know exactly the state of the overall system that has subsystems A and B, let's say, if you have a classical system that has subsystems A and B, like you have a box of gas with a partition in the middle, and you tell me, I know exactly the state of the combined system of A and B, then you certainly know exactly the state of the subsystems A and the subsystem B, right? In quantum mechanics, that's not true. In quantum mechanics, you can know the entire system exactly. But because there is entanglement between subsystem A and subsystem B, there's no such thing as the actual true individual quantum state of subsystem system A by itself, because it's entangled with the rest of the world.

2:12:27.5 SC: So let's say you just have two spins owned by Alice and Bob, and your overall state is a superposition of Alice's spin is up and Bob's is down, plus Alice's spin is down and Bob's is up. Okay? That is the pure 100% legit quantum mechanical state of the overall system. What is Alice's state all by itself? And you might say, well, it's in a superposition of spin up and spin down, but it's not. It's entangled with Bob's system. So there is no fact of the matter about even what Alice's wave function is, much less what her spin is doing. And the entanglement entropy is just the von Neumann entropy minus trace log rho, rho log rho, applied to the particular situation where the overall system is in a pure state, but some subsystem of it is in a mixed state because of entanglement. So entanglement entropy is just von Neumann entropy applied to a subsystem that is entangled with the rest of the world.

2:13:31.2 SC: Peter Bamber says, what new information would it take to make you change your mind and decide that the many-worlds interpretation of quantum mechanics is not the most probable?

2:13:40.4 SC: Well, the best information would be to rule it out by experiment, which is easy to do do, because the many-worlds interpretation just says quantum systems always obey the Schrödinger equation. So all you have to do is show an experiment where a quantum system that is completely left by itself, not poked at or interfered with or entangled with the rest of the world in any way, violates the Schrödinger equation. And you might say, well that's just being cheeky, that will never happen. But there are experiments going on right now to look for that and there are theories that predict it should happen. So if any of that happens, then there you go, many-worlds is ruled out by experiment. Of course there's also other ways to lower your credence in many-worlds, the best way would be if someone came up with a better theory. [laughter] That's always possible. I don't think it's very probable, but it's always conceivable. You can absolutely change your credence and theories without any new experimental data coming in. If someone comes up with a better theory, or contrary-wise, if someone shows that a really, really good argument, that many-worlds is a bad theory for one reason or another. So I think that there's plenty of different ways in which it would happen. I'm not worried about it happening because I think that many-worlds is probably right.

2:14:53.1 SC: Sergei says, on the limits of knowledge, a realist view is that the universe just is and we humans are just tiny specks of it with very coarse grained maps of the whole. One can imagine that building progressively more accurate maps or laws could run into the limits of size and/or speed, whether hard or exponentially hard. If so, we might end up never being able to learn the true laws of the universe, assuming it is even a meaningful statement in this ontology. Instead, some observations would always remain outside of our ability to model them and appear to us as Knightian uncertainty or even miracles. In other words, the universe as it might have accuracy... The universe as is might have accuracy limits on lossy compressibility and so not contain such compressed images of itself in our heads or our computers, whether now or ever. How likely do you think this might be? And how might be... What might be an observable or possibly experimental indication of this situation?

2:15:49.4 SC: I think this is perfectly plausible in principle in the space of all possible worlds. I think that the remarkable thing about our actual world is the opposite of that. The remarkable thing to me about our actual world is how accessible and intelligible and discoverable the rules of the game actually are. The ultimate laws of physics. The fact that the laws of physics underlying our everyday lives are completely understood by us, very primitive, not very rational human beings, is an amazing testament to the understandability of them. Now we don't understand everything, we don't know what dark matter is, we don't know how quantum gravity works, et cetera. And it's completely conceivable to me that emergence is actually getting in the way here, or what particle physicists would call decoupling.

2:16:39.6 SC: If there are new phenomena at very, very high energies that only influence our low energy world through the sort of bundled up renormalized effects on low energy processes, then it might be very, very difficult, if not impossible to discover what those high energy processes are. Or maybe there is some underlying stratum of reality that is either purely quantum mechanical or even pre-quantum mechanical that gives rise to the world we see in terms of quantum fields in spacetime, but it gives a rise to that in such an effective way that we will never actually have empirical data pointing us toward the right underlying theory. These are all very possible. Okay? But I don't think that's... Either one of those is what Sergei is pointing at, the idea of just uncertainty and inability to model some observations, so that they look like miracles, that's absolutely possible. But I see either zero evidence of it or zero evidence that we should even be worried about that coming true. So I don't think it's very likely. But of course there's always some chance.

2:17:47.0 SC: Tim Giannitsos says, during your solo episode 320 on complexity and the universe, you mentioned that the concept of force is not fundamental in modern Quantum Field theory. The familiar forces like electromagnetism are associated with gauge bosons. You then outlined a broader definition of force that includes the Higgs boson and the Pauli exclusion principle. What definition are you using to include these other concepts as forces? Do they still have units of kilograms meter per second squared? I don't know if you meant your definition to be informal or whether there is some comprehensive new list of things we should call forces. ChatGPT includes the Casimir effect and entropy as forces.

2:18:24.8 SC: Yeah, I think this is one of those places where you shouldn't trust ChatGPT too much. It's going to pull in a whole bunch of opinions from the Internet and give them all to you without a lot of filtering there on which ones are relevant to the question or not. But no, I mean, really forces. I mean, really, yeah, you know, the right units, the right expression in Newton's laws of motion and so forth. The point is that the concept of force was formalized in Newtonian physics, right? In Newton's Principia Mathematica, F equals ma. Force is mass times acceleration. And then you had various expressions for what kind of force you're talking about, gravitational force, friction, or what have you. And there were hydrostatic forces eventually developed things like that, forces from pressure and so forth. And you could go very far. When you talk about the structure of the Earth inside the crust, you're basically using Newtonian physics and notions of forces and things like that. Now, you probably know that even in the context of classical mechanics, you don't need that notion of force at all. You can talk about the Hamiltonian version of classical mechanics, which just talks about energy, and that's all it talks about. You have positions and velocities or positions in momenta more generally.

2:19:41.4 SC: And then you calculate the energy in terms of that, and you do all of physics based on that. There's no appearance of the notion of force. Likewise, in Lagrangian mechanics, you calculate an action which is again defined in terms of energies, and you minimize that action over a path. And again, there's no appearance of the word force. So force is not supposed to be something absolutely fundamental and necessary in physics. It's a particular way of describing the classical regime, and it's a way that applies when you're using Newtonian kinds of concepts. In that context, the familiar gauge bosons give rise to forces, but also so do degeneracy, pressure and the Higgs boson and other things like that. The Casimir effect absolutely does count. Entropy doesn't count as a force, but there is a kind of force called an entropic force that does count, that arises in certain very specific circumstances. But none of that is fundamental because none of it's quantum mechanical. Okay? Quantum mechanics just uses a different language. Quantum mechanics is closer to Hamiltonian mechanics or Lagrangian mechanics.

2:20:49.1 SC: You can formulate it in different ways, but the word force never appears in the Schrödinger equation, or in the path integral that Feynman would write down, or in Heisenberg's algebra of observables or anything like that. The word force just never appears. So all of this discussion of forces in the context of particle physics is only supposed to be informal. If you want to be formal, then you have to include these other things like the Higgs boson and the exclusion principle. The phrase that there are four forces of nature, which I'm happy to use in the context where it's appropriate, is just a way of saying there are four sets of gauge bosons that give rise to interactions of the form, that in the case of electromagnetism and gravity, give rise to classical long range forces as well. There's a lot of mathematical similarity between what's going on in the weak nuclear force and the strong nuclear force with electromagnetism and gravity, even though they don't directly give rise to long range forces that we see macroscopically.

2:21:51.7 SC: Michael says, if physics reached a point where all fundamental laws were known and quantum physics was unified with the macroscopic world, and all were experimentally verified, what kind of questions do you think would remain not just in philosophy, but within physics itself?

2:22:06.6 SC: Almost all the questions... [laughter] I mean, the search for a unified picture of fundamental laws in the quantum realm that get matched on successfully to the macroscopic world, is very important. I like doing it, some people like doing it. But it's a tiny fraction of all science or even all of physics. If you go into a physics department, there's a lot of people doing, let's say, atomic physics. And guess what? We know the rules underlying atomic physics, right? We know how atoms are constructed, we know the Schrödinger equation, we know even Quantum Field theory is occasionally necessary. But there's still a lot of questions we don't know the answer to. Because once you get beyond a single hydrogen atom, once you get either to more complicated atoms or to collections of atoms, there's a lot of interesting things that are hard to calculate, hard to predict, not to mention chemistry, condensed matter physics, biology, all of those things. There's a lot of people doing biophysics, there's a lot of people doing plasma physics or astrophysics or gravitational physics. You might know general relativity, that doesn't mean you know what the gravitational wave signature is going to be when two black holes collide with each other, right? So the quest to find the underlying laws is important, but it's a tiny fraction of everything that science does. So most of physics won't even notice if we were able to do that.

2:23:29.2 SC: Henry Jacobs says, are people who want to fight for democracy in an in-person prisoners dilemma, where our only hope is to resist, all resist at once, coordination is hard, do we have any hope?

2:23:39.8 SC: It's not a prisoner's dilemma, I don't think. Remember, the prisoner's dilemma is the weird situation where any individual person is clearly better off by defecting. Of course, if everyone else defects, you're also clearly better off by defecting. It's a kind of a game theory coordination problem, absolutely. Coordination is hard, like you say. And, with Cass Sunstein, we talked about this a little bit, the coordination problem, and with Steven Pinker also. If there's a whole bunch of people who want to resist a totalitarian dictator, but they know that if they're the only one who resists actively, they're going to be tremendously punished, then clearly you have a coordination problem. Yes, we're not under a completely totalitarian dictatorship as yet, as evidenced by the fact that we can actually go out and protest and say bad things about the regime and things like that.

2:24:34.1 SC: They're trying to get rid of that. They're trying to eliminate it. They're trying to spread the idea that anyone protesting the current regime is ipso facto a violent terrorist and therefore should be cracked down upon. So that is clearly a move in the direction of totalitarianism. But at the moment, we can still do it. And as I'm recording this, I think next week, October 18th, there's going to be a big demonstration in Washington DC. And I hope that people go and I hope that it's effective. There are battles that are not just picking this piece of legislation or this court decision or whatever. There are battles of public opinion and there's battles of elite opinion, letting people in positions of power, whether it's in government or in the media or whatever, know that there is a majority of people who don't like what's going on right now is crucially important. So, yes, coordination is hard, but it's not impossible. And we can still do it. We do absolutely have hope.

2:25:36.7 SC: Theo Lind says, consider a software that records sounds below the threshold of 80 Hertz. Is this analogous to effective field theory? Sound is a pressure that passes through a medium, but it is... Is it entirely discretized or do higher frequencies imprint data upon what is recorded below the threshold of 80 Hertz?

2:25:58.7 SC: It's close, but it's a tricky question here. Yeah. You can use techniques from effective field theory in what is basically classical acoustic physics. Right? People use effective field theory all the time, not just in Quantum Field theory. Effective field theory is very useful for analyzing the question we just mentioned, which is predicting the gravitational waves from inspiraling black holes. Okay? Whenever you have fields, there'll be a regime where effective field theory is useful, but it's not exactly the same. You have to also be careful a little bit. The thing about Quantum Field theory is that when you have a cutoff and you look at energies and modes and wavelengths below the cutoff, so you look at the infrared regime of long wavelengths and low energies, it's not that the ultraviolet regime, the higher energy things are simply being ignored, they have effects because quantum mechanics has virtual particles in a way that classical physics does not. And those virtual high energy effects can absolutely have an impact on low energy physics. The point of effective field theory is that all of that impact can be summarized in changes in the parameters of the low energy theory.

2:27:13.0 SC: The fine structure constant, the mass of the electron, all the various numbers you would use to define your low energy theory are affected by the high energy stuff. So they're there. It's not like we're just ignoring them, it's that we don't need to know what they're doing in order to understand the low energy physics. We can measure the renormalized or the effective or the low energy parameters in the low energy regime all by itself. That's something that is very clean and clear in quantum effective field theory. In classical physics, you'd have to like be a little bit more careful in figuring out exactly what happens. You don't have virtual particles. There's some circumstances in which you kind of have something similar to that. You know, I've written papers on what is called the effective field theory of large scale structure in cosmology. And there you have Feynman diagrams, you have renormalization, you even have virtual particles and the whole thing. But it's kind of a trick because there's not really anything probabilistic going on, in the fundamental theory, it's just that you don't know the details of the initial conditions. So you sum over different possibilities and that gives rise to what looks like an effective field theory description with Feynman diagrams and the whole bit. So it's a subtle thing, it's sort of complicated. But there is something very definitely going on in Quantum Field theory that doesn't reappear in the classical world.

2:28:35.8 SC: Ken Wolf says, you've mentioned a number of times that you do not think that ChatGPT and the current crop of LLMs are in any way sentient or conscious, and I agree. Is there anything an artificial Intelligence could do, which you could deem persuasive evidence that it has subjective conscious experience? Would a certain level of intelligent behavior suffice, or would there have to be something more?

2:28:57.1 SC: I think there would have to be something more. I think that the real lesson of the Turing test, for example, Alan Turing, obviously brilliant mathematician and computer scientist, devised this idea that what matters to thinking is how you act, how you would in particular respond to prompts that asked you something. And if the machine gave you responses to the prompts that were indistinguishable from those that a human agent would give, then we should say that that is thinking, by all intents and purposes. I think that what we should have learned by now is that he was wrong about that. I mean, maybe it was true for thinking, which actually is what sort of what Turing cared about more than consciousness. But it certainly doesn't apply to consciousness.

2:29:47.0 SC: Or I shouldn't say certainly, it does not seem to apply to consciousness. It's easier to fake being a conscious agent than we might have thought. I think that LLMs and their performance is incredibly impressive in many, many ways and has moved much more quickly than I might have thought. But in order to say that it's truly conscious, I think that what we've learned is you have to go beyond the input output paradigm and actually look at what is happening inside the LLM. I think that something like this case was made by Anil Seth, as I think I mentioned in a previous AMA. Anil Seth was a previous Mindscape guest and he's written an important paper about the differences between what is called computational functionalism, the idea that all that matters is the computation, the thing does, and a more biologically oriented view of consciousness. And my own view is not exactly aligned with Anil's because I think that he puts a big emphasis on literally biology and a lack of complete substrate independence.

2:30:52.3 SC: I think that substrate independence is possible if you truly get process matching. So, like, not only the input and output are the same, but the process by which you reach that input and output is precisely isomorphic, or at least pretty close to isomorphic. So in other words, I do think that even though we don't know what's going on completely in conscious experience, it is more than just a black box of input and output. It has something to do with what is actually happening in the brain. It probably has something to do with microscopic processes that might be metabolic or entropy generating. It might have something to do with counterfactuals, how would you act in different circumstances, things like that. So I don't know the full answers to these things, but I'm pretty sure that simply answering questions in the right way is not going to give me evidence that something is truly conscious or intelligent.

2:31:51.3 SC: PJ Wenzel says, concerning your child universe ideas from papers with Jennifer Chen circa 2004, I wonder if you would provide a descriptive verbal picture of the physical process, but also list the assumptions and limitations or simplifications that went into that model. The question is, what is your assessment of the chances that our real universe could self replicate in this way?

2:32:13.5 SC: Sure. Jennifer Chen and I wrote a paper on a multiverse cosmology that helps explain the arrow of time. And a crucial role was played by baby universe creation in what would be otherwise empty space, De Sitter Space. You know, the space that we think that our universe might be headed toward if there's a cosmological constant that is providing the acceleration. And the idea is that, maybe if there's a true cosmological constant, the future of our universe is nothing but cold, empty De Sitter Space and nothing ever happens. That's absolutely a possibility. There's also a possibility that quantum fluctuations give rise to baby universes. And by baby universes in this case, we mean universes that are truly disconnected from our spacetime. So what you would see in the background De Sitter universe, which was cold and empty, would be a thermal fluctuation that would sort of be like a lot of photons come together to make a little mini black hole which then evaporates away right away, but from inside, you've actually bubbled off a new little blob of spacetime that can then possibly inflate, undergo the process of inflation, which we called spontaneous inflation, and then create a universe all like ourselves.

2:33:28.0 SC: And it would be... By inflation it could become arbitrarily big, even though it started arbitrarily small, and then this would be completely separate. It's not embedded in our universe. You just get multiple universes that all exist in the wave function of the universe, but exist not within each other spatially or even next to or touching each other spatially. So all of that process is entirely not understand. Right? It's something that Alan Guth and Eddie Farhi and others back in the '80s wrote papers about, and Fischer, Morgan and Polchinski also wrote about it, and other people wrote papers about it. Can you actually have this sort of quantum tunneling into a baby universe, they were mostly concerned with circumstances under which it was triggered. So, like, you made it happen. You set up a situation which looked like it would lead to this. And they found lots of mathematical subtleties. So none of them could really reach a firm conclusion that it did or did not happen. What Jennifer Chen and I needed was a spontaneous quantum fluctuation into that kind of thing, which is therefore even more hard to know whether it's real or not.

2:34:35.3 SC: Anything dealing with the combination of quantum mechanics and gravity at a profound level, we have to be a little uncertain about that, that's okay. In the modern world, now that we have AdS/CFT and things like that, and we understand holography and other aspects of quantum gravity a little bit better. There's been a reluctance to believe in the existence of baby universes, especially because in AdS/CFT, you seem to have a complete description of what's going on in the quantum, in the conformal field theory on the boundary. And it seems to map to anti De Sitter Space without any baby universe. Of course, it's not clear that that's what happens. And even more importantly, we don't live in anti De Sitter Space.

2:35:19.0 SC: We're approaching De Sitter Space and things might be very different. There have in fact been a couple of papers very recently saying, okay, maybe you do get baby universes after all, in either anti De Sitter Space or in De Sitter Space. So I don't really know what the chances are that our real universe could self replicate in this way. I do... I am pretty confident saying I have not heard any other scenario which explains the arrow of time in a convincing way. There are scenarios which have an arrow of time, but basically because you put it in by hand, right? Not because it sort of robustly arises in the way that it does in our model. So for that reason alone, I think that it's worth taking this possibility seriously. And who knows, hopefully we'll understand it better before too long.

2:36:08.2 SC: Lina Musiara says, I think I understand the compatibilist argument for free will. However, aren't you a little bit bothered by the fact that deep down your decisions are already determined in spite of your not being able to know them in advance?

2:36:20.9 SC: No, [laughter] I'm just not bothered at all. I don't know why you folks are bothered by all these deep metaphysical, cosmological, physical things going on here. What's going to happen trillions of years in the future, whether or not things are determined at some microphysical level. Why would that bother me at all? Like, I got other things to worry about, like the world is full of issues, and that is just really not one of them. And look, even if somehow you... It turns out that quantum mechanics is not many-worlds, but it's some truly stochastic theory, right? Imagine that there truly is unpredictability about what's going to happen. There's still even in that world, even like a Copenhagen like world, where you have true stochastic elements and the fundamental laws of nature. What that means is you can't predict what will happen in the future, but there's something that will happen in the future, right? You just don't know it. To me, that's exactly the same as there is some fact of the matter about the arrangements of my atoms, et cetera, that will be revealed in my behavior in the future. But I just don't know it. There's a fact about the future that I don't know. I truly don't see the difference there. Certainly not at a sort of level that would bother me. And one way or the other, even if it bothered me, if it were true, I would learn to live with it.

2:37:41.1 SC: Alex Dubrow says, when writing a trade book, do you work on a schedule or wait for inspiration and do you find yourself rereading and revising a lot during the process?

2:37:50.3 SC: Well, in some sense you work on a schedule in the sense that you take, typically have a contract with a deadline. I've been very, very bad with volume three of 'The Biggest Ideas'. And I feel very guilty. And this is not... I don't want to be this person who hands in the book late, but it's been a lot going on this year that has slowed down my book writing. So generally that puts a timetable on you. But I don't have anything very formal about writing at certain times of day, writing certain chapters on certain weeks or anything like that. My writing is very, very uneven. Some weekends, thousands and thousands of words are produced. Some weeks, nothing is produced. You just don't know. You try and you see what happens, as far as I'm concerned.

2:38:36.9 SC: I do less revising than average, mostly because I try very hard to get it right the first time. Some people will write in the sense of just like scribbling everything that they can think of down on a piece of paper and then just revise and revise and revise and cut and paste and edit and things like that. But the way that my brain works needs to be more systematic. Like, I can't imagine writing chapter seven before I've written chapter six. And many people do. Many book authors do that. So I will stare unmovingly at the open document for a long time before actually starting to type anything and really try to figure out what it is I want to write. And as a result of that, most of what I write is more or less the final product. I will certainly edit it and revise. But it's not major kinds of stuff. The most major kinds of stuff that I have is I will write stuff and go, you know what? It's just not working. I'm just going to erase it. But it's not very often that I write something and then dramatically revise it to a new version of the same thing. That doesn't happen very much.

2:39:42.3 SC: Tucker Hyatt says our best laws of physics seem to be describing what we observe in the universe moment after moment and eon after eon. Do you think we will ever regard such continuity, such lawfulness, as anything other than a marvelous brute fact of reality?

2:39:57.3 SC: Maybe. [laughter] I feel like I'm not giving as definite answers as people want here. It could be a brute fact. It absolutely could be, or it could be as I'm actually very interested in thinking about, that the apparent orderliness of the universe sort of arises out of a different kind of order that we would almost be tempted to call chaos or disorder. I mean, one person's order is another person's disorder. If I flip a coin a billion times, and half the time it comes up heads, and half the time it comes up tails, in one sense, that's maximally disorderly. I have 50% chance of either one happening either time. But the statistics you would get out of those billion flips are very, very predictable, right?

2:40:39.3 SC: Roughly 50/50 heads and tails. And you could even go to higher order moments of the distribution and things like that. So is that chaos? Is that order? Where does the continuity come from? These are sort of questions that are more subtle than you might think they might be. So I'm open to all the different possibilities. My stance on brute facts is that we have to be open to them. It's not that I know what they are. It's not that I can point to something and say, that's a brute fact. Whenever you see something in the behavior of the physical universe that is a little bit interesting or intriguing or puzzling or mysterious, by all means, try to find a better explanation for it. I'm all in favor of that. But you have to be open to the possibility that no such explanation exists. You don't get to decide ahead of time, oh, this has an explanation, I'm gonna go find it.

2:41:32.0 SC: Miwash Vaijor says, your recent conversation with Petter Törnberg reminded me of the idea of strange attractors in the parameter spaces of human societies and institutions. To me, it always comes across as something profoundly revolutionary but very intangible and extremely hard to quantify. Now that we can model ensembles of agents in increasingly accurate social settings, do you think this is something we'll ever be able to map out or do you see any deeper reasons that will remain out of our reach for decades?

2:42:00.4 SC: Well, I think it's in between those two things. I think that it's not just strange attractors by the way, the phrase strange attractors was popularized in the literature on chaos and chaos theory in the sense that you can have the general idea of an attractor. The idea of an attractor is that you have some dynamical system and a wide variety of initial conditions move toward the same kind of final conditions. Okay? A ball rolling on a hill with friction has an attractor where it goes down to the bottom of the hill and it just sits there. A strange attractor is one where it's not just one point where the system sits, but some sort of fractal basin of attraction or something like that. So most attractors are not strange, but they do exist. And I think the more interesting thing is the existence of attractor mechanisms at all. This is why I'm trying to work on the physics of democracy these days.

2:42:53.1 SC: Because I do think that there are ideas from statistical physics, statistical mechanics that apply 100% perfectly to, or at least not reliably. Even reliably is too strong but validly to collections of human beings. In both cases, in the case of like box of gas and a collection of human beings, you have individual particles, individual constituents that have their own ways of behaving. But then there is sort of average behaviors that come about because of the mixture and the interactions between all the individual things. And I absolutely think that there can be times when you see, oh yes, I see why this is happening. Not because of this person's individual choice or that person's individual choice, but because of the overall collective dynamics of the whole.

2:43:44.6 SC: I mean, a classic example is the two party system in the United States. Not every country has a two party system. The United States does and always has, roughly speaking. Why is that? It's not because of the orneriness of Americans or anything like that. It's because we have a political system enshrined in the Constitution of the United States that favors a two party system system for all sorts of reasons, most clearly in the direct election of the President rather than through a parliamentary kind of system. So you can absolutely try to explain these large scale structure things by looking at statistical arguments based on many individual persons doing the best that they can. And I think that we don't know a lot about that. I think it's an active research area. It's going to be fun to see where it goes.

2:44:32.4 SC: Philip Ricius says, you learned about the Schelling model in your recent interview with Petter Törnberg. I was surprised that you both seemed unaware of this and the Ising model respectively. Where else have two fields of science independently developed similar tools for very different problems and could learn from one another?

2:44:52.5 SC: We might have given the wrong impression there. I mean, I did not learn about the Schelling model from that interview. I teach the Schelling model in my classes. In fact, it occurs to me now I have written a section on it in 'The Big Picture'. So I've known about it for a long time. And I think that Petter knew about the Ising model too. What we didn't know was whether anyone had sort of mapped them on to each other in any way, which I think... And in fact we got a Bluesky notification from a fellow physicist who pointed out someone who came close to doing exactly that. But you can't do it exactly because they're not the same model. For those of you who are not listening to that one, the Ising model is a set of spins on a lattice with the property that every point in the lattice has a spin on it and they talk to their nearest neighbors and then you give it a temperature or magnetic field or whatever and you talk about its properties. The Schelling model is a little bit different. It has a lattice and it has two kinds of population like red and blue squares or something like that.

2:45:50.0 SC: And they do interact with their neighbors in some way. But there's also empty squares which does not exist in the Ising model. And rather than red and blue flipping back and forth between red and blue, the individual squares move like checkers on a checkerboard. In fact, I think that Schelling actually used a checkerboard when he was fooling around with it when he first invented it. So it's not the same. They're different models. They're very similar in some ways. They might have similar properties, but they're not exactly the same. And different models are going to be different for... Sorry, going to be useful for different purposes. The Ising model represents a very real physical possibility, and the Schelling model tries to represent a very real social possibility.

2:46:31.5 SC: Most physicists would not have heard of the Schelling model because it's not something that shows up in their laboratories. I'm unusual in that way, in that I absolutely have heard about it quite a while ago. In terms of the actual question, where else have two fields of science independently developed similar tools? Oh, all the time. [laughter] There's a lot of science out there. Nobody can know all of science. Nobody knows anywhere close to all of science. It's very, very common for scientists in one field to need a technique and figure it out themselves and then realize or be told, oh, yes, in this other area of science, we already developed that. I mean, famously, Werner Heisenberg, trying to invent quantum mechanics, had a way of talking about two observables not commuting with each other, and he had to be explained by Max Born, you reinvented matrices and it eventually be called... Became matrix mechanics. But yeah, that happens all the time. It's not rare at all.

2:47:29.0 SC: Eric says, I'm curious how you got into the world of writing popular books. Did you have a clear ambition to write for popular audiences? Did you cold call publishers, or did you have social connections that made the path easier to follow?

2:47:40.4 SC: I certainly didn't have social connections that made the path easier to follow. I did have a tiny bit of a public, I don't know, reputation is even too strong because this is before like the Internet or anything. Not before the Internet, but before people became well known on the Internet. But, I would give popular talks and I had a blog and things like that. Right? So I had a public profile, I guess is the word to say. I did want to write for popular audiences, yes. But it wasn't like a high ambition for me. I did put in a proposal and it was rejected. And I went, yeah, okay, that's fine. I have other things to do. And I moved on with my life. It was eventually my agent who contacted me rather than the other way around. So if you're really interested in writing popular books, if that's the subtext here, you certainly don't cold call publishers.

2:48:31.1 SC: That's just not the way it works. You can cold call agents. In the trade book world, agents are the mediators between publishers and authors almost always and certainly every good time. Like, you should have an agent. You shouldn't try to do it yourself. For one thing, agents will listen to you. Agents are... They're... Part of their job is to receive pitches from people they've never heard before and evaluate them and try to figure out whether it would make a good book or not. For another thing, publishers will listen to agents. Publishers will actually listen to a good agent who has a reliable track record of saying, yes, I think this book might be good. And for a third thing, when you're in the process of writing the book and negotiating with your publisher, the agent is your advocate and on your side. So it's the agent that is really, really important there. Social connections are not that helpful, except maybe in finding an agent. And even there, a good agent is actually going to look at the proposals that come in one way or another.

2:49:30.6 SC: Pete Faulkner says, during multiple AMAs, I've always been impressed that you appear to have a wonderful memory for the names and topics of all your former Mindscape guests. Given that there have been well over 300 episodes of the podcast, do you have a trick for remembering them all so well, or do you actually have to go back and check details sometimes?

2:49:47.1 SC: I certainly don't check details for the AMAs. I do very, very little work to prepare for the AMAs, but I actually have the opposite impression of what Pete says. I think I'm very bad at remembering previous podcast episodes, et cetera. I do spend a day of my life, roughly speaking, for each podcast episode, so hopefully there'll be some memory of what happened. And some of them are very interesting and made a big impression on me, and so I keep coming back to them over and over again. But there's a lot. Like you say, there's 300, more than 300 out there. And some of them, I leaf through the archive and go, oh yeah, I remember that one. I hadn't thought about that one in a long time time. So no, certainly no memory tricks to share. Sorry about that.

2:50:26.9 SC: Michael Honey says, how do you plan to wind down your career when the time comes? Are you tending to become a gray beard emeritus professor or do you have other retirement ambitions?

2:50:36.3 SC: Man, that's a harsh, harsh question there, Michael, that I'm ready to wind down my career someday, even though it's not now. The very idea that it would be coming is a little bit depressing. But you know that the true, and this is absolutely true, and honest answer is, I don't know. I could very easily see that my career almost doesn't change when I technically retire. Maybe I don't teach as much or whatever, but I'm still writing books, writing papers, doing podcasts, more or less exactly the same thing.

2:51:06.8 SC: I could also absolutely see completely changing, like just changing gears into something very, very different. I'm absolutely open to that possibility. One of the downsides of an academic career is that your life is more or less similar when you're 50 years old than when you're 30 years old, right? You're a professor, you're doing research. Or when you're 60, or maybe when you're 70 or 80. And some part of me kind of enjoys the romantic idea of just doing something completely, utterly different if I have the option to do that. I'm not saying I will, though. I don't think I'm going to just kickback, but maybe I will. I don't know. I'm not going to rule that out. I'm not going to make any promises that I won't just kick back. I've been working hard. Maybe I deserve a little kicking back. Who knows?

2:51:50.2 SC: Ian Carey says, when talking about hypothetical higher dimensions in discussions of String theory, et cetera, I've heard you and other guests talk about the extra dimensions as too small to be detected. But when we talk about the three spatial dimensions, size doesn't seem to enter into it. Are these extra dimensions fundamentally different from the spatial ones in how they work or is this just a case where the confusion comes from using everyday words like small to describe complex theoretical ideas?

2:52:16.5 SC: Actually, neither one. The extra dimensions aren't fundamentally different, and the word small just means small. I think the confusion comes in because the macroscopic dimensions, the three dimensions of space that we know and love, are sufficiently big that size doesn't enter into it, but they have a size. The size might be infinity. That's very possible. The universe might stretch forever in all the directions. It might not be infinity. It might be that the three spatial dimensions we live in have the form of a torus or a three sphere, or some weird kind of topological shape where there is a size to them. It's just that the size is bigger than our observable universe, so we don't notice. So it's the three big dimensions that we don't talk about the size, about... But we could. And in the smaller dimensions we have to because there's a reason we don't see them, and it's probably because they're small.

2:53:08.0 SC: Darren Vigliotti says, you mentioned last month that you were teaching classes again and how exhausting that can be. As a medical middle school principal and former high school science teacher, believe me, I can empathize. So one month later, how are you holding up? I hope things are well and you're enjoying the teaching.

2:53:23.0 SC: So I am barely holding up, Darren. It's a lot of work and you know, I care about it, so I try to do a good job. So I put a lot of effort. Well, let's put it this way, I put a lot of effort into the actual teaching, into the lecturing, thinking about what to cover, et cetera. I'm less good at assignments, at problem sets and exams and things like that, like that just, I get no joy out of doing that and therefore it's harder to get motivated. And I try, I want to do well, I care about what the students feel, I want to be fair to them, obviously, and it's more important for their lives, how they do on the tests than it is for my life. So I owe it to them to try to do a good job. But I'm not as good at that part of it. Happily, I'm teaching two very, very fun classes right now in undergraduate quantum mechanics and philosophy of cosmology. So that keeps me going. The fact that Quantum Mechanics is at 9:00 AM on Monday and Wednesday and Friday is not good. But I got, like I said before, snookered into that one, they had me accept teaching the course before I knew when it was being taught. So we'll see if they trick me into doing that again.

2:54:32.1 SC: Karage Yu says, between all of your scientific contributions, which one do you want to be remembered for the most?

2:54:40.0 SC: So I'll give two answers to this, both of which are complete cop outs. So I do this sometimes. I appreciate the questions, but I don't always answer them in the way that they want to be answered. The first is, I hope it's one I haven't made yet, I hope it's one I haven't actually published yet. And I actually think there's a chance this is true. The things I'm working on right now in the foundations of quantum mechanics, emergence of space and time, complexogenesis, things like that, I think I have a chance for these to be of much more lasting importance than anything I've done so far. And I think that's how it should be, if you want to sort of be excited by your future work. But we'll see about that. The other even more honest answer, and you might not believe me or not, but I truly don't care how I want to be remembered. Once I'm gone, how people remember me is like zero interest to me. I would feel bad if I were gone because I still have friends and family who are here, but once they're gone, I'm not in it for long term being remembered or being talked about, that's just... I'm motivated by various things and that's just not one of them.

2:55:49.2 SC: Schreiber Bike asks a priority question. Why hasn't there been an attempt to say how many universes there are at several ages of the universe based on the Everettian interpretation? Is that not an interesting question in which the assumptions and uncertainties could be discussed? I'd find it interesting to know based on that list of assumptions and uncertainties, how many universes there were when our universe was one second old, one-year-old, one billion years old, and today?

2:56:14.6 SC: Yeah, there's been almost no attempt to do that because it's not a well defined question in Everettian quantum mechanics, there are details that depend on things like is Hilbert space finite dimensional or infinite dimensional? And we don't know. So like, how do you even talk about it if you don't know that very basic question, is the space infinite or finite? Okay? If it's infinite, then there literally is no answer to the question how many universes are there. There's a continuum of universes. Like David Wallace, former Mindscape guest, likes to say, asking how many universes there are is like asking how many experiences you had yesterday. It's a category error. It's not something that you can count. If Hilbert space is finite dimensional, then you at least have a chance of asking how many universes there are.

2:57:05.9 SC: But still it depends on somewhat arbitrary choices about what's the environment, what's the system, things like that. As I like to say sometimes, the thing about the many-worlds interpretation of quantum mechanics is it's not about the worlds, not about how many universes there are. If you care about the interpretation because you care about understanding quantum mechanics, the question how many-worlds there are is way, way, way down there on the list of interesting questions. It's not one of the ones that we need to understand to talk about what you predict on the basis of this theory, how it helps you understand the emergence of spacetime or anything like that. What you do care about is when you do a certain experiment, what are the fraction of worlds that get a certain outcome, and the fraction of worlds gets another outcome, that's where the probability comes from. But that's just the fraction of worlds, not the number. The number is truly an uninteresting concept in Everettian quantum mechanics.

2:58:04.6 SC: Michael Bright says, when people speak of atheism and God, it often feels like these two views are split between those who believe there's a bearded man in the sky dictating events on Earth versus those who think there's a scientific explanation for why reality exists the way it does. I do not believe that is a bearded Zeus looking man in the sky dictating events, but I do believe in the scientific method for understanding how physical reality works. But I sometimes find atheism to be... To almost feel religious in its certainty. I find that I prefer doubt. Why are so many brilliant physicists self described atheists? Is it that they find the idea of a pagan like figure dictating events silly? Or is it something fundamental in physics that suggests there is no creator that set into motion the laws in which our physical universe exists?

2:58:46.9 SC: So I want to erase from this question any mention of certainty or doubt. No one is certain about these things. No one should be certain. If they are certain, they're just making a mistake. As I said before, sometimes your credence is high enough that you move on with your life and you act as if something is known. But that's not meant to imply that you are metaphysically 100% certain about it. I am completely willing to change my mind about the existence of God if God just does something to provide evidence that he exists. I'm totally happy to do that. The reason why physicists and cosmologists et cetera tend to be atheists is that they look at the world and they see no evidence for the existence of any supernatural behavior whatsoever. In fact, it's by far the opposite. As we've already mentioned, it is amazing how well a physics science based description of the world works at a very fundamental level. I mean in some sense much better than you might have expected has any right to work. Religion had dominance for thousands of years.

2:59:53.0 SC: It had a chance to make some interesting predictions for what the large scale structure of the universe looks like or what the fundamental workings of the universe looks like. The Bible could have said, human beings evolved from other species. It could have said there are billions of stars and galaxies out there. It doesn't say any of those things. There's nothing in any religious tradition that you go, oh yeah, they seem to know something about the world that we wouldn't have known otherwise until we discovered it scientifically. Whereas in science we're constantly making predictions and they come true and it's great. So there's just no need, like following the footsteps of Laplace when he was cheekily asked by Napoleon why there's no mention of God in his treatise on celestial mechanics. He says, I have no need of that hypothesis. And I think that most scientists are more or less sympathetic to that point of view.

3:00:44.8 SC: Nikola Ivanov says, why do so many natural systems, earthquakes, neuronal avalanches, ecosystems, et cetera, operate near critical points?

3:00:54.3 SC: Well, that's a really good question, and I'm not sure what the answer is, but I've thought about it, and we talked about a little bit with Nigel Goldenfeld. I don't know if you listened to that podcast. He's one of the world's leading condensed matter physicists who's worked on criticality a lot. But let me point out like a little bit of a caveat here, and especially I'll explain for the people who don't necessarily talk this language. When you have a phase transition, so when you have ice boiling to water, or let's say you have the Ising model. We were just talking about the Ising model. Right. The Ising model can be defined as, let's say you have a bunch of spins on a lattice and they are interacting with their neighbors. And let's say they're interacting in a way that says they want to line up with each other. Right? Like, they're happiest if each spin is pointing in the same direction as its neighbor. Okay? But there's also a temperature, so there's thermal fluctuations. So even though the neighboring spins like to line up, maybe they don't because they're jiggling around a little bit because of the temperature. And at very, very low temperatures, as you might guess, you have a situation where everything is lined up. Okay? And if there's no external magnetic field or anything, there is even a symmetry breaking because everything could be lined up, or everything could be lined up down.

3:02:09.7 SC: As long as it's the same, that's all that really matters. And that's the situation at low temperature and then at high temperature, no one cares that the spins want to be lined up. The temperature is so high that they're just jiggling randomly back and forth. These are two different phases of the Ising model, the low temperature aligned orderly phase, as it's called, the high temperature disorderly phase. And somewhere in between, there is what is called a critical temperature, a critical point at which it's right balanced in between everything lining up and everything randomly jiggling. So at that critical point, you can do math, you can do physics, you can calculate various statistics, and you find that what happens is there appear little regions, not the whole thing, but little subsets of the lattice where all the spins are lined up one way or another. And you can actually count, like, how many small regions where everything is lined up are there, how many big regions, there's different ways to do this math. And what you find is power law behavior. So the number of regions that are all lined up is some wavelength, some size of the region, to some negative power. Okay?

3:03:23.4 SC: Then you can calculate this exponent, the negative power that is in question here. And different systems, the Ising model or the Ising model in different numbers of dimensions and so forth, will have different exponents you can calculate. And it's full employment for condensed matter physicists. So the critical point is exactly that point, if you're below the critical point, so if you're in the regime where things want to be orderly, you're almost all orderly with a couple of little tiny, almost negligible fluctuations back and forth. If you're above the critical point where there's a very big temperature, almost everything is fluctuating with a very rare, occasional smattering of orderly regions. If you're at the critical point, it's a mixture of both. And so what does that have to do with earthquakes, neuronal avalanches, ecosystems, et cetera, in the Ising model, or in water boiling or anything like that, all these phase transitions seem to have these power law behavior at the critical points where you're doing a phase transition. And what does it have to do with earthquakes or something like that? You know, maybe nothing. In the Ising model or in any condensed matter physics system which is undergoing a phase transition, it seems like you have to tune to the specific temperature or the specific conditions in the parameter space that will keep you at the phase transition.

3:04:43.0 SC: Most values of temperature are not at the phase transition and therefore not at the critical point. So it seems very, very special. So why in the world do you see this power law behavior, which is the actual thing you see in all these other situations like earthquakes, sand piles, firings of neurons in the brain, and so forth? And I think, and talking to Nigel and other people, I think maybe the answer is going to be there's not a universal answer. There's different reasons why different kinds of systems have this power law behavior. For example, there is something in network theory called preferential attachment. If I have a bunch of nodes and I'm actually adding connections between the two nodes, so I'm making a graph by drawing a bunch of circles which represent nodes. And I'm sort of semi-randomly connecting one node to another, one, it depends on the process that I use. So one process is for any two nodes, there's just a random probability I will draw a line between them. But preferential attachment says there's a little bit more probability that if a node already has a lot of lines attached to it, it will get more lines, right?

3:05:54.1 SC: So the rich get richer kind of phenomenon. And you might say, well, that's unfair, that would never happen. Think of web pages, right? Web pages link to each other and when you start having more links to you, you get more links to you because more people find out about you. And it's straightforward to show that under that kind of dynamical system, you get power law of behavior in the number of links to and from different web pages. Zipf's law in linguistics, the sizes of cities. There's a whole bunch of things that obey power law behavior and has nothing to do with a phase transition, right? It's just a different mechanism that also gives you a power law. Now, in some cases, there might be an even different explanation, like an evolutionary advantage.

3:06:37.6 SC: If in the brain there's evidence, and people debate the specifics here, so maybe don't take it too seriously, but there's some evidence, that if you look at the firing of neurons or the timing of different events in the brain, they also obey power laws and they call that critical behavior. But it's really just power law behavior. And maybe there's a purpose to that, maybe that helps the brain function. You don't want the whole brain to be doing exactly the same thing at the same time. Like a low temperature Ising model, you don't want it to be completely random either, like a high temperature Ising model, you want to have it have some structure, so there's some predictability and reliability. But you also want some flexibility as well. Critical points or power law behaviors, which are ubiquitous in complex systems, seem to be a way to get that balance of what you're looking for, which is one way you get that kind of behavior. I don't think you should look for a unified theory of all of them.

3:07:33.8 SC: Brian says, is there any technical difference between the universe expanding versus a static universe where everything is getting smaller?

3:07:40.3 SC: Well, yes and no. If you're just doing general relativity, so you're just thinking about spacetime and what it's doing, you are welcome to use coordinates in which the universe is static, but everything is getting smaller. The reason why nobody does that is that there's more to life than general relativity. There is atomic physics and Quantum Field theory and all those things. It would be enormously inconvenient to think of the hydrogen atom as shrinking as a function of time. You could do it, but it would require an incredible coincidence or conspiracy between all the constants of nature to let it happen. Remember we were talking about Compton wavelengths and how the Compton wavelength is proportional to one over the mass of the particle.

3:08:19.4 SC: So you have formulas for this that involve h-bar, Planck's constant, and also the speed of light and things like that. If you wanted to work in a description where the size of the universe is the same but everything is getting smaller, you basically have to imagine that every constant of nature is changing and every constant of nature is changing in exactly the way that is necessary to keep the usual relationships between them all the same. Right? Why do we always see the same wavelength of light emitted from a certain transition in a hydrogen atom? If hydrogen atoms are shrinking and you are shrinking and the wavelength of light is shrinking, you're allowed to think that way, but it's no gain in either understanding or convenience whatsoever.

3:09:03.0 SC: Jonathan Sirocco says, the big ontological change of String theory is that particles are 1D objects rather than 0D particles. But from Quantum Field theory we already know that the electron is not actually a point particle and is instead represented by a three dimensional wave function. I'm sure the electron is also represented by a wavefunction in all spatial dimensions in String field theory. When the String theory electron is described by such a wavefunction, is the 1D ontology still relevant? It feels like the underlying shape might be obscured. Or would this be the level of description where standard QFT is the emergent description of String theory and that both wave functions describe the same object?

3:09:38.4 SC: Yeah, you're putting your finger on something interesting here. Oh my God, I sound like ChatGPT. Sorry about that. But nevertheless, it is interesting. In Quantum Field theory, you could think of Quantum Field theory as something that grew out of starting with a particle theory and then changing the particles for fields, then quantizing the fields. Or you could think of it as just starting with the fields in the first place like with E and M. In the String theory case, we certainly historically did the equivalent of starting with a particle theory. We started with the individual strings and we quantized that. And a lot of people think that what you really should do is what is called String Field theory, where you have have the equivalent of a field operator that can create or destroy strings and things like that.

3:10:24.6 SC: As far as I can tell, as someone who is not a super expert in this, people knew that was interesting and tried it and it didn't get anywhere. It didn't really seem to solve any problems or give you any insight. So people still use the old fashioned individual particle way of talking about String theory. Part of what you're asking the question is, what is the connection between the way of talking about strings and the way of talking about ordinary particle physics. And you're exactly right. The strings generally have a size. They're one dimensional loops or line segments. But that size is so incredibly tiny that there is an emergent Quantum Field theory description where the quantum fields have their excitations that look like particles. Okay?

3:11:10.1 SC: And we think that, okay, if you're a String theorist, you think that they really are strings, but by any imaginable way that we actually had to measure them, they're going to look like particles to us. So one thing, it is true that when people say the electron is a point particle, that's very misleading because the electron is the excitation of a field. But that's okay because there absolutely is a regime where there's an appropriate approximation where there's just one electron around. The field nature of it is not that important. And you can talk about it like a particle and then in that description, it has no size, it has a Compton wavelength for its wave function. But the thing it's describing is not something that has any size. In String theory, you know, there's a lot we don't know about String theory still. So I think that it's still possible that someday String Field theory will turn out to be an even better way of doing it. But right now we do fine with just thinking of strings as little things, quantizing them and mapping it on to conventional Quantum Field theory in standard ways.

3:12:15.0 SC: Chris Kaltwasser says, we accept that a human developed species, such as a dog or a cow, possesses a basic right to exist free from cruelty. This right is not diminished by the fact that it's a product of our selective breeding. Given that we are now creating increasingly complex, intelligent and autonomous machines, is it not an inevitable ethical conclusion that they will be compelled to grant them... We will be compelled to grant them rights. Where ethically do we draw the line?

3:12:42.6 SC: Yeah, well, like before, we just were talking a little while ago about how would we know if an AI is conscious. I think that something like that is going to be necessary before we think that we should grant AIs rights. If the AI... If you construct something that does what an LLM does by having a giant lookup table, like a really, really big lookup table, like, similar to Searle's Chinese Room experiment, where you ask it and there's a question that you can ask, and for any question you ask, there's a ready answer, and the lookup table is so big that the thing you're talking to essentially is indistinguishable from a conscious creature. No one wants to grant the lookup table rights. Right? The mechanism by which you are constructing those answers actually does matter. And I think that the mechanisms by which LLMs work are very, very, very different than the mechanisms by which conscious creatures work. That's not to say that we can't imagine AIs that do work more like conscious creatures and should be granted rights, I'm absolutely open to that. I don't think we're any anywhere close to that right now, but it's absolutely a possibility. And I would hope that various combinations of philosophers and computer scientists and people think very hard about this question and try to give good answers to it. I think that right now there's been a lot of talk about it. Much of the talk has not been very high quality, but the urgency is going to grow greater over time.

3:14:11.3 SC: Matt Haberland says, I want to be the utility monster for Halloween. What should my costume look like? If you've ever heard of utility monster, Laplace's demon, or similar and unintentionally visualized and incarnate creature, what did it look like to you?

3:14:29.1 SC: So I'm answering this question because I think it's a good question, but I don't know the answer. We'll crowdsource it. Let's see if anyone has any suggestions for how to dress like the utility monster. The utility monster is a thought experiment that is meant to illustrate difficulties with being a utilitarian as part of your moral philosophy. The utility monster... Remember utilitarianism says we should maximize pleasure or happiness or utility, depending on whatever variety of utilitarianism you subscribe to. So the thought experiment is imagine an alien or a monster, or some creature that is just really good at experiencing pleasure, like way better than you are, or all the human beings in the world world, just so that the maximal way to have pleasure or utility in the world, is not to make human beings happy, but to keep the utility monster happy. The utility monster's happiness just counts for much more than all the happiness of human beings in the world. So, a lot of... I've seen sketches of utility monsters and they kind of look like monsters, but I think that's wrong. I think the utility monster is really good at feeling pleasure and happiness. So the utility monster should look happy. That's the single most important thing I could say. What is the most joyous and pleasure filled kind of costume you could think to wear? That's what the utility monster could look like. I don't know what that means in practice, but I'm happy to let people think about it.

3:15:54.0 SC: Ryan Nichols says, I understand that black holes are not quantum excitations of a field with a Lagrangian. But given that the are completely described by mass, charge and spin, much like a fundamental particle, is there any reasonable or sensible way to think of a black hole as a kind of a fundamental particle, perhaps as some quanta of geometry?

3:16:14.4 SC: Yeah, there are ways to do that. I mean, you can absolutely write down theories where there is a black hole field that creates and destroys black holes. And you might say, well, wait a minute, I shouldn't do that because a black hole is a composite object that I make out of other fields, isn't it? But the distinction there is not so clear. You know, like we absolutely do Quantum Field theory with protons and neutrons and pions and other things. In fact, we do Quantum Field theory with sound waves in fluids or solids and things like that. So the thing about Quantum Field theory is you start with some fundamental set of fields, but you might be able to do a transformation into a different looking set of fields. So to say the black hole is composite in one description is not to say it's not fundamental in another description. And indeed people have tried to do that. Again, it's one of those things where, okay, you can do it. Is there a point to it? Why are you doing this? Are you getting any benefit from it? That I'm not so sure about.

3:17:16.5 SC: Brandon Lewis says, I have a new cat and he's super vocal with quite a repertoire of different sounds from short to sentence length. I don't understand most of it, though the cat seems to think he's being perfectly clear. Do your cats talk to you? If so, to what degree do you feel you understand each other? How much information do you think cat speech carries?

3:17:38.6 SC: So I think that it's an interesting question, you know, like the communication, let's put it that way, from non-human animals. It's absolutely clear that cats and dogs and other creatures like that are communicating with us. Anyone who has met any one of these creatures knows that they communicate. Talking is a... Is not necessarily the right way to think about it. They don't necessarily have grammar or words or anything like that, but they have a repertoire, like you say, that is absolutely true. And my cats Ariel and Caliban talk to us all the time. But, usually, like, look, they're not that smart, I got to say. Like, I love my cats a lot, but geniuses they are not. I think that a lot... [laughter]

3:18:21.9 SC: The impression I get, which might be completely false, but the impression I get from my cats talking to us is that my cats think that we're not that smart. And they try to explain things like, isn't it time for us to get treats now? And we know perfectly well they want the treats, but we're just not ready to give it to them. They've had too many treats already. But I think that from the cat's point of view, we're just too dumb to understand that the treats should be coming now or the lap time should be coming now or the scritches or whatever. So, yeah, they talk. I think different cats have a very different range of how talkative they are. Some cats are pretty quiet. Some cats are pretty noisy. Some cats talk with their claws. Some cats purr a lot. Yeah, that's part of the fun. The wide variety of different ways they have of letting us know how short we are falling in their estimation.

3:19:13.0 SC: Alexander Knuckle says, do your derivations of Born's rule and the papers with Charles Sebens use the epistemic separability principle still pretty much reflect your current attitude toward the problem or have things changed in the meantime? If so, I'd be curious to know which new promising directions you see in this field.

3:19:31.9 SC: I'll be honest with you. On the one hand, I mostly think that that paper was on the right track. I don't really see that there's new promising directions in the field. I don't think there need to be new promising directions in the field. I don't think that we need to bang our heads against the same old problems over and over again if they have been more or less solved. But if anything, the change in my attitude since writing the paper is, I'm more relaxed about the whole thing right now. So for those of you who don't know, this is an issue in Everettian or many-worlds quantum mechanics, in many versions of quantum mechanics, you say, if I'm measuring a spin up or spin down, there's some fundamentally stochastic element of what's happening. And there's... All I can say is there's a probability that there will be spin up or spin down. In many-worlds, you can say with certainty there's going to be one world where it's spin up and one world where it's spin down. So where does the probability come from? The usual way of doing it is there's an amplitude associated with those two possibilities and you square that amplitude to get the probability. That's the Born rule. But the amplitude in Everett just sort of goes along for the ride.

3:20:35.0 SC: It's sitting outside the entire universe. So Chip Sebens and I wrote a paper inspired by other papers. Lev Vaidman was really the first to highlight the importance of self locating uncertainty. There were papers by David Deutsch and David Wallace and others deriving the Born rule from decision theory. Everett himself as well as other people used just statistics and frequencies to derive the Born rule. And in our paper, Chip, in my paper, we said, we're not saying by writing this version of deriving the Born rule that other versions are wrong. There can be many right ways to get the same answer and they're all equally true, but they just use different routes to get there. And I think I'm even more in favor of emphasizing that aspect of things now. Look, the thing is, I think that the nice thing about the self locating uncertainty approach is that it's almost clearly there. Like you can't really deny if you believe that... If you understand and accept the basic notion that it could be true that the universe is described physically by a wave function that always obeys the Schrödinger equation. I think the existence of the self locating uncertainty is almost unavoidable.

3:21:53.7 SC: People try to avoid it, but I think it's really unconvincing frankly. And so you're in a situation epistemically where there's a fact of the matter. I'm on the spin up branch of the universe or the spin down branch and I don't know which one is true. And to get through life you need to have a way of assigning credences to those two possibilities. And the only objection is, well, I don't wanna. I think if you know that you have to assign credences to being on the spin up branch or the spin down branch and you ask how should I do that? It's almost blazingly obviously clear that the right way to do it is to use the Born rule. There's no one who actually thinks there's a better way of doing that within the context of many-worlds. At best, there are people who say, I don't want to do it at all and you can't force me to do it. And all I have to say to them is, okay, I can't force you. I'm not going to try. Let's get on with our lives, okay? For me, I'm perfectly happy with that particular understanding of what is going on. And we have other problems to think about now.

3:23:01.0 SC: Ophir Averbuch says, I recently listened to an older episode where you talked to David Kaiser about sources for funding for science, particularly for large physics projects. The discussion focused on which funding regimes are most beneficial for the scientific enterprise. However, as a society we should arguably also consider the scientific value against the alternative cost. The money could be going to hospitals, education, et cetera. Do you have any thoughts about how this calculus should work? Specifically, how do we evaluate progress in scientific understanding against other social goods, especially in cases where science is not expected to improve technology?

3:23:39.2 SC: Well, I think that under a well functioning society, which ours is not right now, but when things are going well, the answer to this question is just democracy. I don't think there's an algorithm. I don't think you say, okay, this fraction of your wealth goes to this thing and this fraction goes to another thing. I think that people in the citizenry of the country in question, vote representatives to represent them. Those representatives rely in various ways on expertise and also on advocacy. So they listen to people say, well, we would like the money for this purpose, we would like the money for that purpose. And they do their best to decide how to allocate the money, knowing that a certain kind of Biotech research might actually be very technologically useful and save lives, even if the people who voted for them have no idea what that Biotech research is, because that's the job of a representative democracy, that we offload the responsibility to know those things to people who are much more experts than we are.

3:24:43.0 SC: So again, in the ideal situation where everything is working, your representatives are not supposed to be knowledgeable about everything, but they're supposed to be responsible and they're supposed to care about your welfare. And therefore they rely on the real experts to make their cases for spending different amounts of money. And I think that's how science should work also. And I think in science, in areas like cosmology or quantum mechanics or whatever, quantum foundations, let's say, areas that don't have direct technological or economic or medicinal applications, we can still make the case. The case is very easy to make. This is a fundamental human drive to understand our universe and it's worth spending some money on it. Not as much as it's worth spending on health care. But guess what? We don't spend anywhere near what we're spending on health care on understanding fundamental physics. So I think that's a system. It's an imperfect system, but it's the best system we're going to get. It's broken down in the last few months, but hopefully it improves going forward.

3:25:47.8 SC: Kaushik Mitra says in Poetic Naturalism you emphasize that multiple ways of talking about the same underlying reality can all be valid depending on context. How do we distinguish between valid poetic expressions of naturalism and those that drift into pseudoscience or quantum mysticism? For example, saying we are all one with the universe can be grounded in physics. We're all part of a single quantum wave function. Yet it often sounds like mystical hand waving. Where do you draw the line between a poetic but legitimate framing and an unscientific one?

3:26:18.1 SC: I think for this question, which is a very, very good question, I would point toward pragmatism in the sort of technical philosophical sense of the pragmatic school of philosophy that is famously like the first American born school of philosophy, William James, CS Peirce, John Dewey, people like that. And somewhat unfortunately, the motto of pragmatism became the cash value of ideas, which is to say, like when talking about ideas in metaphysics or epistemology or whatever, okay, what good is this to me, this particular idea? Is your system that you're proposing for talking about the world actually getting me anywhere? Does it help me understand something or build something or do something, or predict something, or is it just window dressing? I think that works perfectly well with exactly these questions. To say we are one with the universe, what does it mean? What does it get me? To say that there is only one wave function that describes the whole universe, that gets me something. It's very clear what that gets me. I can calculate things, there's entanglement, I can make predictions, the whole bit. To say we are all one with the universe, if that's what you mean, you're fine. I'm 100% part of it. If that means somehow that you're going to leap from that to think that I can change reality by thinking about it without touching it, then I would be more skeptical. So I think that's not very different about poetic naturalism than any other way of thinking.

3:27:48.8 SC: Paul Hess says, the CMB seems to be a good rest frame that we can measure our motion against and maybe even label the CMB to be the rest frame of the universe. Does the way we see it now imply that at the start of inflation, everything was basically at rest with respect to everything else before inflation started, in order to create the CMB rest frame we observe now? Or could it have been, for instance, a hectic universe full of motion that then also inflated?

3:28:14.9 SC: Well, I do think that the whole question of what happened before inflation is a fraught one, an important one, but one that we haven't really settled or not even appreciated as much as maybe we should. But the general idea would be that it could be relatively hectic. I mean, there are results. There's a famous paper by Vachaspati and Trodden that showed that in order for inflation to start, you need a region of space to be relatively smooth. If it's wildly varying in that region, then those wild variations carry energy and the inflaton will never be the most important thing. The inflaton will never dominate the energy density of the universe. The inflaton field needs to be dominant in some region of space in order for inflation to get started in the first place. But you could have a region where that's true.

3:29:05.8 SC: You could have a region where the inflaton dominates, surrounded by regions that are wildly fluctuating. That's perfectly okay because the inflating region will eventually take over and win. Typically, even if the region that starts to inflate is relatively smooth, it won't be perfectly smooth. It won't be exactly the same from point to point. If you could do that, you wouldn't need inflation in the first place. But the action of inflation is to smooth everything out. So the small ripples that might have been there ahead of time are basically dissipated away. There does need to be... This was a paper by... I'm not exactly sure who it was, so I'm not going to try to get all the authors right, but Bob Wald was one of them, I think maybe Tony Rothman was another one, maybe Bill Unruh.

3:29:53.0 SC: I really shouldn't try to guess because I'm not sure. But there's a paper back in the '80s which tried to think about this question, where does the rest frame come from? Because in the approximation where the inflaton field is perfectly constant, right, where it's not varying at all, and you have a De Sitter expansion, there is no rest frame. There's no preferred rest frame there. And typically people used that approximation, but it's just an approximation. So what these people showed is that it's the slight deviation from that approximation that really does matter. When inflation starts, the inflaton field is moving. It's not completely constant. And therefore, if it's moving, there is a single rest frame in which its value is the same, is constant from place to place in that rest frame. And that rest frame survives to give us the cosmic microwave background rest frame today.

3:30:46.0 SC: Friedrich says, during all those years of AMAs, has there ever been a question that occupied your mind much longer than the duration of your answer in the podcast, that either left you thinking in the time between selection and the podcast or at some later point? Alternatively, is there a question that you still remember?

3:31:03.0 SC: I think it's in between that. Mostly, to be super honest, I answer the question and move on with my life. There's a lot of questions and I can't dwell on all of them for a long time, but some of them will stick around. I think that this is a feature just of me, not of other people. But I'm really bad at remembering why I started thinking of something or where I got an idea from or anything like that. So people always want to know, like what was the thing that inspired you to think this? Or where did you start? And I have to disappoint them. I have to say, I don't know. I don't remember that. I'm sure something happened, okay? So I'm sure that there were AMA questions that I thought about or made me think about something for a long time, or took an existing train of thought and nudged it in a crucial direction, or something like that. I just can't remember those things. I don't have... That's not a capacity that I'm very good at. I'm just... My memory is not very good in general. Or maybe I could be more kind to myself. I only remember certain things, let's put it that way. Jennifer and other friends of mine are constantly amazed at how bad I am at remembering the plots of books I've read or movies or TV shows we've seen. Like, I can watch TV shows over and over again and still be surprised at what happens, even though I've already seen them. Just no room in my brain for that stuff. Sorry about that.

3:32:24.1 SC: Richard Cashdan says, what do you think about Kip Thorne in his book 'The Science of Interstellar'? He says something like that all matter in a black hole disappears and leaves behind only a distortion in spacetime. Is that literally true? If that is literally true, then how is information preserved? And how would it be possible for that information to ever be recovered?

3:32:44.3 SC: Well, Kip Thorne is a very famous, very accomplished, celebrated physicist and a friend of mine and a former Mindscape guest. You can listen to his episode on the show. Nobel Prize winner, leader in gravitation theory, co-founder of the LIGO experiment, all this stuff. So I would listen to what he says. But in this particular context, in 'The Science of Interstellar', he's interested in macroscopic black holes, big black holes, right? Stellar mass or million solar mass black holes or something like that. And the thing about such things is you can ignore the fact that they radiate a la... Stephen Hawking, right? The temperature of an astronomical sized black hole is much lower than the temperature of the Cosmic Microwave Background. It's going to be completely unobservable. So he's talking about black holes. Classically, he's not including the fact that black holes eventually evaporate and maybe the information that went into them is conserved in some way. Maybe it is not. We don't know yet. Many of us think that it is, but we're not sure the mechanism behind that. So for the context in which he is talking, a big classical, permanent astrophysical black holes, what he says is completely true.

3:33:57.7 SC: The final question of this month's AMA comes from Peter Kane, who says, when I studied Astrophysics at University in the UK in the early 2000s, many lectures involved professors writing line after line of equations on the board. While talking through them, I found I could either copy the notes but not listen, or listen but end up with little to refer back to later. I've always wondered, was this style of teaching a flaw in me as a student or in the way the lectures were delivered? How does your own approach to lecturing compare?

3:34:26.4 SC: Yeah, I think if you went to school these days, you might have the opposite complaint, that a lot of professors just use PowerPoint and it's much worse than writing on the blackboard. Of course you can show pictures which are fun and cute and it depends on what you're trying to teach. Maybe that's the right way to do it. I am absolutely one who still writes at the blackboard during all my lectures. I do think two things, number one, you can have a style of working at the blackboard which is completely incomprehensible. You just write too fast and talk too fast and don't give the students time to digest. It's absolutely important that you occasionally pause, you spend some time looking at the students, hopefully more time looking at the students than at the blackboard. And you let them think a little bit, maybe even let them ask questions or you ask some rhetorical questions to them. I think all of that is important. Also, I had some very formative courses that I took as a student when I was in grad school from people like Sidney Coleman and Nick Warner, where they would have these very high density lectures with a lot of stuff going on and therefore they would xerox their own personal handwritten lecture notes and hand them out. In Sydney's case, it was not his lecture notes, he would never let anyone see those.

3:35:42.0 SC: But he taught the same course for years and years and prior generations of students took notes and then handed them out to everybody. And in Nick Warner's case, he just xeroxed his own. And so I try to do something like that. It depends on the details, it depends on the course. If it's a small seminar course where most of the focus is on the discussion in class, then I'm not going to hand out lecture notes. But if it's a lecture course like the two I'm teaching this year, I write all of my lecture notes on the iPad using notability and I make a PDF file and put them online for the students to read. So they always have that to refer back to. So different techniques are going to work differently for different people, for different students, for different lecturers. Many lecturers don't care about how good their lectures are. They do them and they get on with their lives. And I think that's a flaw in the system. I think we could do better. I do care a little bit. I try. Whether it succeeds or not, I think it's going to depend on details, depends on the subject matter, the student, many, many other things. But I too try via multiple different channels to convey the information and to put it in the minds of the students.

3:36:49.5 SC: With that, we've reached the end. Thanks as always for supporting the Mindscape podcast. Hope you had a good time. Talk to you next time.

[music]