Theories of Everything with Curt Jaimungal - The Many Worlds Theory of Quantum Mechanics | David Wallace

Starting point is 00:00:00 It won't take long to tell you Neutral's ingredients. Vodka, soda, natural flavors. So, what should we talk about? No sugar added? Neutral. Refreshingly simple. It's weird. It's alien. Super unintuitive. But the systems physicists study are really simple systems. There's a really wide-spread misconception that physics is complicated.

Starting point is 00:00:42 Professor David Wallace of the University of Pittsburgh is one of the world's top philosophers of physics who targets simplicity, seen through the lens of the Everettian interpretation of quantum mechanics, which is commonly known as the many-world theory. Today we explore this theory in-depth, dispelling common misinterpretations and tackling questions like how do you gain empirical warrant for the Born Rule itself, the probabilities of quantum mechanics, when you're confined to a single branch? We later explore what Wallace calls the greatest mystery in physics, that is, why microscopic physics treats past and future identically, however, our macroscopic reality experiences them as fundamentally different. We then explore other probability problems in many worlds, like if all possibilities occur, then how do probabilities make sense? This is where Wallace's decision-theoretic approach gets explained. Finally, we investigate Wallace's conception of reality itself, how emergent patterns at different scales can be considered equally real despite emerging from something else. This is where emergence gets explained and Wallace shows that the best understanding

Starting point is 00:01:48 comes from not abandoning math for pure philosophy nor from shutting up and calculating, but from their thoughtful integration together. What's the largest misconception in physics that you have to dispel to even other physicists? So there's a bunch of things I could try, but let's try this. There's a really wide mis-conception that physics is complicated. And physics is, in a certain sense, very simple compared to other sciences. It's weird, it's alien, super unintuitive, sometimes really expensive to do experiments in, but the systems physicists study are really simple systems. If you compare how complex the nucleus of an atom or even a crystal is compared to a living thing or the human brain or the American economy. Now those are complicated systems.

Starting point is 00:02:45 I'm scared of complicated systems. I do physics. The physics is nice and simple. But you say to even other scientists, let alone lay people, that the physics isn't complicated and they look at you as if you've gone mad. So yeah, that's probably my biggest example. Yeah, I'm similar in that people ask me why is it that I focus on philosophy or fundamental physics on this channel and why don't I touch on the economy or politics.

Starting point is 00:03:09 And to me, to make a political statement, it requires an extreme amount of knowledge and also assumptions. And same with the economy. And even though physics is mathematically complex, it's simple in that, well, it's simple in what's being studied. So where is this complexity coming from, or this perception of complexity? I think partly it's just the math is difficult.

Starting point is 00:03:32 I think partly because, just precisely because physics is studying quite simple systems, we've penetrated much more deeply into physics than really into any of the other sciences. I mean, our level into any of the other sciences. I mean, our level of understanding of the systems physicists study is incredibly accurate and precise and comparatively complete compared to what any other science has managed. Again, not because physicists are wonderful, though some of them are, but because the systems

Starting point is 00:04:01 they're choosing are these easiest systems to study. If you think about how good are we at doing detailed, quantitative predictions of what humans do, and the answer is basically we're terrible. Certainly we can make reasonable guesses that are better than chance, that's perhaps the best you could say. How accurately can we measure the magnetic moment of the electron? Ten significant figures last time I checked. So the very fact that these are systems where our tools are more tractable than in the complexities

Starting point is 00:04:32 of the life sciences or the social sciences mean we've learned vastly more and that means that to go further from where we are now requires this kind of very sophisticated mathematical technology. You know, physics is ambitious in that sense. It wants to make these detailed, precise statements, what's going on. Plus physics is alien. I mean, all of science is a bit alien. Every science has places where you have to put aside intuitions you kind of learned in the ordinary world, but physics is further from the ordinary

Starting point is 00:05:02 world than most of science. You know, your instincts about living things will get you into trouble sometimes, but your instincts about quantum mechanical systems are almost useless. Okay, so do you think that latter part is what Feynman meant when he said no one understands quantum mechanics? Because you just said that we understand physics more so than other areas of science or social sciences, etc. So what does it mean to understand here then? Well Feynman is talking about the quantum measurement problem and the specific weirdness of quantum mechanics That's a whole other story and that to be fair is not irrelevant to why physics is weird and difficult But I think if you if for the moment you think about something

Starting point is 00:05:44 Separate from the measurement problem though I'm sure we'll loop back to it, do you think outside those specific weirdnesses of quantum mechanics, do you think how well do we understand the physics in the middle of the sum or the physics of why a beaker of helium 4 behaves like a superfluid at very low temperatures or something like that? Those are things that we just understand really well and I think Feynman would have agreed with that. What I mean to say is what does it mean to understand?

Starting point is 00:06:10 Does it just mean that we have a model and it's precise enough to make accurate predictions? Okay, that's a really good question. I mean sometimes to understand is to make intuitive but I don't think that's a great model of explanation and understanding even elsewhere in the sciences. I mean, you can get your head around Darwinian natural selection, but I'm not sure you ever find it intuitive. It sort of relies on ideas about space and time and time scales and things that human intuitions don't connect with well. So I think a better model, well, you're right that to some extent, can you model something?

Starting point is 00:06:46 Can you construct a description of it that's predictively effective? That's part of understanding, but I think it's not, you need to move beyond having just a formal mathematical model. You want to be able to understand what the sort of bits in the mathematics mean. But that's a slightly delicate game because you want to avoid saying what they mean is just their translation into our sort of everyday categories because maybe strange things at the microscopic scale or the galactic scale don't translate very well into our ordinary categories.

Starting point is 00:07:16 So I think understanding in physics is a bit of a bootstrap process. You kind of understand things initially sort of badly and partially in terms of metaphors and analogies with things you already understand. And after a while you get enough practice in using those tools that you understand them in their own right and then you kind of go back and you reinterpret your ordinary understanding in terms of your sort of deeper grasp of what's going on. But it's a subtle question. It's not completely obvious and there are lots of controversies in philosophy of science as to just what explanation and understanding mean.

Starting point is 00:07:53 Hmm. Okay. So we're going to get to ever reading quantum mechanics shortly. Before we do so, I want to know other than the measurement problem, what's a problem in physics that you think about constantly, but yet for years you haven't been able to make progress on? Well, I don't know about no progress, but in terms of things I still find mysterious, I mean, let me give you two that are connected.

Starting point is 00:08:17 One is why do we get interesting sort of large-scale physics? Why isn't everything like subatomic particle physics? Why can we say stable things about fluids and solids and chairs and tables? Why are there special sciences at all? Why are there physics sciences above the subatomic scale at all? And then the other question is related to that. The physics and the science we discover for large-scale systems almost always cares a lot about the direction of time. You and I care a lot about the direction of time. The past and the future seems very different to us. The past and the future seems very different to a biological system or to a hot object

Starting point is 00:08:57 that's cooling down. Microscopic physics doesn't give a damn about the difference between the past and the future. And it's been a puzzle for more than a century to understand, like, what's the secret source that can be added to a microscopic physics that thinks past and future are all the same? To get us a macroscopic physics or just an ordinary world in which past and future are manifestly completely different. Is this different than just having some classical limit?

Starting point is 00:09:29 Yeah, because suppose you take a theory that's not classical and you try to get classical mechanics out as a limit. Well, did you get out the bit of classical mechanics that cares about the difference between the past and the future or the bit that didn't? If you got the bit out that doesn't care about the difference between the past and the future, and you know, something like the physics of the solar system doesn't care very much about the difference between the past and the future, for instance, if you get that out, then you still haven't worked out where the difference between the past

Starting point is 00:09:56 and the future comes from in our kind of everyday life-scale work. If you did get it out, if you recover that bit of classical mechanics that does care about the difference between the past and the future, then again we want to know how did that happen? There's nothing in the depths of quantum mechanics that cares about the difference between the past and the future. So if you manage to get something out in the limit that does care, what made it care? I'm putting this slightly figuratively, but in slightly more formal ways, how did the symmetry break? If the microscopic physics treats past and future as symmetric, some extra ingredient has to come in to break that symmetry and make the past and the future look different in our emergent physics.

Starting point is 00:10:36 And this extra ingredient of a low entropy past, is that not satisfying to you? I think it's part of the story, and you're right to pick on that. I mean at some level as a matter of logic if you break the symmetry you either had to break it by putting some asymmetry into the laws or you had to break it by putting some asymmetry into the boundary conditions and there's very little evidence for asymmetry in the laws so the boundary conditions look like the way to go. But that's the beginning of an answer to the question, not the end of an answer. I mean, one way to think about it is, at some level, we seem to understand quite a lot quantitatively about

Starting point is 00:11:19 how to derive large-scale physics from small-scale physics. So suppose I want to derive large scale physics from small scale physics. So suppose I want to derive the kind of viscous flow equations that describe how a sticky liquid flows and I want to derive that from the microscopic physics of that liquid. We've got a reasonable grip on how to do that, but notice that the sticky fluid dynamics has asymmetry and time in it and the microscopic physics doesn't. And I will guarantee you that if you go to the textbooks that talk about the relation between them, they will not at any point say, and now let's assume the Big Bang was like this. So the way in which we're actually, it might nonetheless be the case that indirectly

Starting point is 00:12:02 things about the Big Bang are what matter to getting out that asymmetry, but it's not obvious what the route to make that work is or how the components fit together. I mean, it's not, this isn't a complete head-scratching mystery about which we know nothing. There's lots of kind of partial answers to how that can be. But the point is in isolation, just saying, well, maybe the very early universe had low entropy is the, it's logically possible that could solve the problem because it does break but the point is in isolation just saying, well, maybe the very early universe had low entropy, it's logically possible that could solve the problem because it does break the symmetry, but that's only the beginning of explaining how it solves the problem.

Starting point is 00:12:35 Do you imagine that boundary conditions will always be contingent in the sense that it could have been otherwise and that we can't derive the boundary conditions from first principles? It's pretty difficult to know even how to think about the question. I mean, as a practical matter in physics, normally when we say the boundary conditions are a problem of contingent, we mean, well, you know, there's lots of ways we could set the problem up. And you could tell there's lots of ways we could set the problem up, because look, over here in the lab we set it up this way, and over here we set it up that way, or maybe in maybe in space like one of the stars was this way

Starting point is 00:13:07 around another one was that way around. So when we say it's like contingent which way stars are spinning we can kind of cash that out in things we can get our hands on. We've only got one universe or at least only one we can get at. So there's a philosophical question about even what we're saying when we say the initial state's contingent. And even though we can say formal things in the language of the philosophy of modality, it's not completely clear to me we understand what we're saying when we break away from the kind of intuitions we build up from sort of small, duplicatable systems. And then on top of that, there's the fact that our current best physics

Starting point is 00:13:46 of the early universe is certainly not the last word in the physics of the early universe. So we don't really have a story to tell about how the physics we don't yet have brings about the quote initial condition, the first stage of the universe we can see. And in the absence of that theory it's hard to even assess the whole is it contingent, is it not idea. I mean, for instance, if you've got something like the theories of inflation people talk about, then actually the universe is much bigger than it looks. And there's lots of kind of Big Bang like things going on all the time. And that would give you if that's true, that'll give you both a metaphysical grasp

Starting point is 00:14:26 of what we mean when we say it's contingent how the Big Bang is, and a kind of detailed scientific model that represents that contingency as some probability distribution set of equations. But that's all speculative. There's some evidence for theories of that kind, but nothing conclusive.

Starting point is 00:14:43 Broadly speaking, physics is like, you have evolution laws, which is like this black box, maybe it's not so black, but there's a box, and then you have inputs, which are the boundary conditions or initial conditions, and then there's some output. Is there an alternative model to think of physics as a whole, other than that, or does every model of physics ultimately boil down to that? Very nearly, I'd say. And some of the exceptions are probably not crucial to this question.

Starting point is 00:15:13 If I've got something like an open system where I want to treat effects from the environment around it as just some sort of extra input, then that sort of dynamical model doesn't apply. But in some ways you can then just think about the extra input as more boundary conditions. That's the general model, otherwise we've used in dynamics. I mean, it's still the model we kind of apply in cosmology. Maybe that's wrong, but it's not as if we've got other good case studies of how to run it. I mean, the other thing I suppose worth saying, the slight qualifier gives you your story is quite often,

Starting point is 00:15:46 we don't normally treat the initial condition in physics models as just a complete black box. That could be anything we feel like. Often we have principles like the system went to equilibrium or maximized entropy or minimized energy or something. Or lots and lots of possible ways the system might have started all converged to the same place. So we may of possible ways the system might have started all converge to the same place so we may as well assume the system starts there.

Starting point is 00:16:09 So there are kind of moves of that kind that you use to justify certain initial state considerations. If I'm doing the physics of stars or something I'm not actually going to construct that just as a kind of throughput system that says like whatever the star starts with then it is about turns up. I'll try to make some substantive claims about what's reasonable as a starting assumption about the star and I'll give physics reasons for that. But again, applying that kind of reasoning in cosmology where we're talking about the whole universe, that's trickier.

Starting point is 00:16:38 Hmm. In the year 2000, you had a paper called the quantization of gravity and introduction. I do my research. Yeah, I read that on grad students. I guess there's probably still stuff in there, but I wouldn't stick. I don't promise to stick with most of what it is. Okay, well, that's great because this will help the question. Firstly, I wanted to know, in it you talked about the difference between canonical and

Starting point is 00:17:00 covariant quantization. You didn't touch on string theory. First, it would be useful for the audience to know what the difference is between canonical and covariant quantization in quantum gravity. And I'm curious why you didn't cover string theory. Okay. So the practical answer to the latter thing is I wrote that thing as a grad student, as a graduate exercise in my physics PhD, and put it on the archive.org at a time when archive was a bit more free-reeling about what you put on it, and even then slightly against my better judgment at the advice of my supervisor.

Starting point is 00:17:41 So I would not have done that. I would not advise a student to do that at this point. So there's a lot in that paper that's fairly half-baked. It was never published. I didn't touch on string theory really just because of reasons of scope. What I was trying to do in the more serious piece of work that I was doing that led to the write-up of it was try to pin down and clarify what the sort of almost like the historical starting points of thinking this way are. So you know, string theory is a cool place you get to. But if you ask like, why did you start thinking those terms in the first place? Part of that comes

Starting point is 00:18:18 from ideas in particle physics, part of it comes from saying, okay, suppose just from first principles, I take classical gravity, and ask, how do you quantize classical gravity? Well, that story doesn't actually lead you to string theory. It might lead you to string theory by a long winding road, but it's not the starting point of that story. But to be truthful, the other reason I didn't know any string theory, like I say, it was a pretty half-baked piece of work. The covariance is canonical though. So, looking at it this way, and this is, I think, a serious insight that's worth having, that's not my insight, is well known. Suppose you, there's various ways in which I think this is an out-of-date way of putting it, but it's still helpful for some purposes.

Starting point is 00:18:59 Sure. Suppose you're in the mid-20th century, electrodynamics, you've got a good quantum theory of electrodynamics, you're kind of making your way to a quantum theory of the strong interaction, say, things are looking good, but you're interested in gravity. So you've got a whole bunch of bits of technology that you might want to apply to classical gravity in the hope of getting a quantum theory of gravity that are the sorts of bits of technology that to apply to classical gravity in the hope of getting a quantum theory of gravity that are the sorts of bits of technology that you apply to other classical theories, most obviously electromagnetism, but also maybe like point particle mechanics that got you

Starting point is 00:19:35 out quantum theories. So the obvious thing you want to try is let's try applying that machinery to the gravity. And one way you might do this, this is the covariant way, is to say, well, ultimately, we've got quantum field theory, it tells us how relativistic fields interact, it gives us Feynman diagrams and path integrals and all this good stuff. General relativity is a field theory and if I'm interested in weak general relativity it looks like it's the field theory of a spin-2 particle or spin-2 field I should say on Minkowski's space time. People who are deeply into general relativity are tearing their hair out when I say that but there's at least a sense in which it's

Starting point is 00:20:20 true. So you might say let's just try applying the machinery of the relativistic particle physics to that sort of spin-to-field and see what comes out. That's covariant quantization. The other thing you might say is quantization starts with theories that are kind of thrown in the classic sort of Hamiltonian dynamical form where I have a space of instantaneous states and an evolution rule the moves are forward. Let's throw general relativity into that format and try using the general technology we have for quantizing theories in that format and see if that gives us a good quantum theory. That's canonical quantum gravity. So there are the precursors of those ideas in, well, Dirac's looking at this stuff in the 1340s, I think, but Bryce DeWitt kind

Starting point is 00:21:07 of writes some very seminal papers on this, I think, in the 50s that very much coin this kind of covariant and canonical language. And to some extent, the path to quantum gravity has followed those two alternatives ever since. I mean, the canonical approach to quantum gravity eventually matures into loop quantum gravity in the hands of people like Smolin and Ashdekar, Revelli. The covariant root kind of in a somewhat more indirect path matures into string theory. And you've got quite different traditions playing this game. The people who are trying to do the canonical quantum gravity approach are generally people whose true love is classical general relativity and the people who are doing the covariant approach are people whose true love is particle physics.

Starting point is 00:21:53 Now, in which sense are people in GR tearing their hair out at the statement that weak gravity gives rise to spin-2 field theory? Oh, um... I want to take a moment to thank today's sponsor, Huel. Specifically, their Black Edition ready to drink. So if you're like me, you juggle interviews or researching or work or editing, whatever else life throws at you, then you've probably had days where you just forget to eat. Or you eat something quickly and then you regret it a couple hours later. That's where Huel has been extremely useful to eat. Or you eat something quickly and then you regret it a couple

Starting point is 00:22:25 hours later. That's where Huel has been extremely useful to myself. It's basically fuel. It's a full nutritionally complete meal in a single bottle. 35 grams of protein, 27 essential vitamins and minerals, and it's low in sugar. I found it especially helpful on recording days so I don't have to think about prepping for food or stepping away to cook. I can just grab something in between conversations and keep going. It's convenient, it's consistent, it doesn't throw off my rhythm. You may know me, I go with the chocolate flavor. It's simple and it doesn't taste artificial. That's extremely important to me. I was skeptical at first, but it's good enough that I keep coming back to it.

Starting point is 00:23:08 Especially after the gym. Hey, by the way, if it's good enough for Idris Elba, it's good enough for me. New customers visit Huel.com slash theories of everything today and use my code theories of everything to get 15% off your first order plus a free gift. That's Huell.com slash theories of everything all one word. Thanks again to Huell for supporting the show become a Huell again by visiting the link in the description. The statement that weak gravity gives rise to spin to field theory.

Starting point is 00:23:43 Right. So the way general relativity works is you have this object, the space-time metric, you normally call it G, which determines the distances between points in space-time, indirectly determines the curvature of space-time, all the stuff that defines space-time. And we know, for instance, that flat spaceetime, Minkowski spacetime, special relativistic spacetime is an example of a spacetime like that. So I can write, as a particular choice of G, call it G zero, I can just choose the Minkowski spacetime metric, the flat metric.

Starting point is 00:24:21 And now I can take, formally speaking, I can take any metric g, at least if I assume space-time sort of topologically looks like, you know, Slat Minkowski space-time or if I just work in just a region of it that's not too complicated. I can just formally say for any metric g, g equals g zero, the Minkowski metric, plus h, everything else. By definition, h is g minus g0. And then I could say, well, you know, h is just another field on flat spacetime. And I could treat it as a field just the same way we treat the electromagnetic field and any, well, any field you like, really.

Starting point is 00:25:00 So what's wrong with that? I mean, in a sense, there's nothing. I think it's an informative way of thinking, but the concern you might have about it. So firstly, you could say that splitting of space time into flat space time plus a field on flat space time completely breaks the kind of geometric understanding of gravity that was very much how Einstein looked at it. And, you know, if you look at Weinberg, who has a book on general relativity that takes this kind of route, Weinstein will basically say, yeah, absolutely, I'm breaking the geometric understanding of gravity. Say what, the geometric

Starting point is 00:25:34 understanding of gravity is overrated. And that's a big cultural disagreement between, you might say, relativists and particle physicists. The other more technical reason to be concerned is that that, well, in particle physics language, that split I just made is wildly gauge dependent. In a slightly bowdlerized version of the general relativity description, I'd say it's wildly coordinate dependent.

Starting point is 00:26:00 So the geometry of the problem is being quite badly messed with by making that kind of division. So that's the kind of reason why that way of talking about things is controversial. And where do you lie in this controversy? I'm a pluralist. I think you can use a theory in whatever way you find convenient.

Starting point is 00:26:25 I think if you get insight out of the structure of the theory by thinking about it in geometric terms, good for you. If you get insight out of the theory by thinking about it in particle physics terms, also good for you. People ought to be able to understand these different roots and go back and forth between them. How has your view on quantum gravity changed across the years? Um, okay, so there's a lot I could say there, but here's the simplest one. The way quantum gravity gets described in a lot of semi-popular literature and in philosophy

Starting point is 00:26:57 adjacent circles is something like this, that basically we've got two theories, quantum theory, which is great in its domain, general relativity, which is great in its domain, general relativity, which is great in its domain. Those theories are inconsistent. We don't know how to combine them into a theory that captures the best features of both. That's the kind of classic way people talk about the project of quantum gravity. And I think it's probably how I talk about it in that ancient paper of mine. I think that's largely wrong. In papers I've been more willing to defend how I've argued why I think it's wrong. The reason I say it's wrong is because, well, there's a couple of things wrong with it. The first thing is it assumes that we don't

Starting point is 00:27:40 already have theories that combine quantum mechanics and gravity. And we do very well tested theories. If you just drop your pen and it falls on the ground and stops, then you need both quantum mechanics and gravity to explain that. Gravity to explain why it started falling in the first place, quantum mechanics to explain why it stopped. If you want to understand the structure of white dwarf star or neutron star, which are both understood reasonably well, you need both gravity and quantum mechanics. Nuclear fusion requires us to think in terms of gravity inside stars, think about gravity and quantum mechanics. So we do actually have lots of examples of gravity and quantum mechanics together. What's really going on in modern terms is that's because we don't have much trouble doing quantum theory of gravity, provided

Starting point is 00:28:21 we keep away from energies that are close to the plank energy, the sort of breakdown energy of general relativity in this set up. And so I think the right way to think about quantum gravity is that we're looking for, in particle physics terms, the ultraviolet completion of the low energy theories of quantum gravity we already have. And I think those theories, while we don't understand them nearly as well as say quantum electrodynamics, are things we understand reasonably well, certainly well enough to calculate with. And so I think the perspective that says what we want to do is just take classical general relativity, take the abstract

Starting point is 00:28:58 ideas of quantum mechanics and put them together is not recognizing how much quantum gravity we already have and already actually use in mostly astrophysical, cosmological contexts. Now in your book, The Emergent Multiverse, you talk about ever-riding quantum mechanics. Is that distinct from the many-worlds interpretation? Basically no. And it's somewhat a style choice to mostly talk the way I talk. When I refer to these things in sort of semi-popular context, I'll normally say the Everett interpretation

Starting point is 00:29:37 is sometimes called the Many Worlds Theory. Okay. The reason I tend to use that stylistic approach, and it goes to the title of the book, is to talk about a many worlds theory tends to give people the impression that somehow the worlds are a thing you've added to quantum mechanics to make a new theory, a theory of many worlds. And that's not how the Everett interpretation works. It is a many worlds theory, but only because the worlds are sort of dynamically emergent from quantum mechanics.

Starting point is 00:30:06 Yes, and if anything you've subtracted from the regular quantum mechanics. So you're trying to simplify your axioms. Yeah, yeah, if you think of collapse as an axiom in quantum mechanics, which in some formulations it will be put that way, then yes, exactly. The Everett interpretation simplifies by removing that axiom. Now I don't argue that axiom is used much less in the practice of quantum mechanics than people think it is, but yeah, it's certainly true that if you go to say Dierdakov on Moinem's books, they have it in as an explicit axiom.

Starting point is 00:30:35 So we're going to get to what other axioms are actually added across the span of this book, the emergent multiverse. But before we do that, I want to know what is the largest or one of the largest misconceptions about Everettian quantum mechanics slash the many worlds theory that you have to dispel even to your colleagues? Yeah. So I think in many ways, it's what I've just said, it's that the Everettian interpretation adds something to the theory. I mean, the way I was put in this is a very conservative theory. It tries to take quantum mechanics exactly as it is, and it tries to understand it exactly the kind of way we've been used to understanding physics theory since Newton, i.e. they describe

Starting point is 00:31:19 physical systems that evolve and do their own thing, and the processes of humans intervening in their systems is just more dynamics and to be handled by the equations that govern those systems. And I think because in a certain sense of course these ontologically extravagant, then what people think when they hear about many worlds is somehow you've decided to add this ridiculous ontology to the theory, whereas the ontology was in there anyway. The mathematical structure of the theory isn't altered by studying it. Does quantum gravity have any implications for the many worlds? Like does it provide any difficulties that weren't there before or solutions even?

Starting point is 00:32:03 Yeah, I mean probably not, but it's a bit difficult to tell till we've got sort of fully in place theory. I mean, the many worlds approach is like a recipe for understanding quantum theories. You give me a quantum theory and I'll tell you how to understand it in many worlds terms. So insofar as quantum gravity is just one more quantum theory, I would expect the ideas to go across to it. I mean, again, because we don't ultimately know what that theory is, that's speculative. But certainly, for the fragmentary bits of quantum gravity we have so far, I don't think the Everett interpretation has any particular difficulties with them.

Starting point is 00:32:43 What would you say are your specific ontological commitments? So for instance, Sean Carroll is committed as far as I understand to the physical existence of the universal wave function. Yeah. I mean, I think I've talked to Sean about this. I think this is, I push back against that way of putting it, although in the sense which I'm similar to you. I'd say I'm committed to the physical existence of that which is described by the quantum wave function.

Starting point is 00:33:14 But that's not... It's going to be easier to illustrate that if I think of an analogy. So think about classical mechanics. Yeah. I do classical mechanics in a configuration space framework, let's say, so I've got a single point in the configuration space represents all the states of the particles. And I can write down dynamic equations for how that point evolves. I can also put a measure over the phase space of the classical system, I can evolve that measure forward

Starting point is 00:33:47 under Louisville's equations. And that's also a set of dynamical equations. And at a very abstract level, both the classical mechanical equations for the phase space point and the Louisville equations for the phase space distribution are the same kind of thing. You take a mathematical object, you plug it in a differential equation, you get out a mathematical object later. Our interpretation of those mathematical objects, classical mechanics, is different though. For the phase space point, we think about that point as representing actual physical features of the specific system we're studying. Like for instance, if the phase space point says particle 7 has

Starting point is 00:34:25 velocity 15 in the x direction then we think it does. That's saying a thing about the physical system. Different phase space points correspond to different ways the world could be. We don't think about the probability distribution in phase space the same way. We think about that as more, well there's lots of ways to think about it but you might think that it codifies our ignorance about what the true properties of the system are or maybe it abstracts over a large collection of similar systems and says like some of them have these properties and some of them have these properties but in any case that phase space distribution is not representing

Starting point is 00:35:00 physical properties of a single system and so what I want to say is the quantum state is like the phase space point, it's not like the phase space distribution. It's always fundamentally like the phase space point rather than the distribution. It represents physical features of the system being studied. Different quantum states represent different physical features the system might have. But I'm reluctant to go from that and say the phase space point, the quantum state is itself physically real. For the same reason I'm reluctant to say that the phase space point is physically real. The phase space point is a convenient mathematical representation. What it's a mathematical entity, what's physically real are the various physical systems that

Starting point is 00:35:54 the phase space is representing and the particular properties that the physical systems have that the particular phase space point is representing. So if I'm doing quantum field theory for instance, what's real are the points of space-time and the quantum fields on that space-time which have lots of very complicated non-classical properties. Which properties do they have at any given moment in time? Well, the quantum state tells you but I think it's an error to think, to say that the quantum state itself exists or is real or is physical in that story. It's representing physical properties. In some ways that's just a philosopher's piece of pedantry, but if you're not careful,

Starting point is 00:36:37 you go from the quantum state is real to there is a higher, the reality is really this high dimensional space and reality is this kind of waving complex field on it or something and that's that doesn't make any more sense than starting with classical mechanics and saying well the world is three n dimensional and there's only one thing in it and it moves through this complicated path. Well another name for this channel other than theories of everything could have been philosophical pedantry. So I understand we're quibbling over semantics, but what is the difference here between the physicality of the wave function, your version, and Sean's

Starting point is 00:37:09 position? Like, where's the beef? So, I mean, part of it's a sort of philosophical issue, but the substantive beef is that Sean doesn't think that to understand what's going on in quantum mechanics, I need to bring in some understanding of what those properties are that the quantum state is representing. So if I try to do the ontology of a quantum field theory, for instance, that ontology is going to rely on rich spatial temporal ideas about the symmetry structure of space-time, the locality of the interactions, a whole bunch of stuff of that kind, which the dynamics encodes and then the quantum state tells me what those features are and how they change over time. And those features are like weirdly quantum

Starting point is 00:37:54 features, they're super positions of all sorts of stuff of course, but in any case. Sean wants to say, no look, all you've got is the Hilbert space and the Hamiltonian. And all that stuff about space-time geometry and locality, all of that is stuff you've got to extract from a careful analysis of the dynamics of the theory. So somehow the detailed spectrum of the Hamiltonian for sure is supposed to ultimately encode all of that locality data and none of that is to be taken as the fundamental description of physical reality. It's just a sort of a convenient overlay. That's the big difference. My take on this is that might be right, but it's a research project. It's not an interpretive

Starting point is 00:38:39 claim about the physics we currently have. It's not currently the case that we can understand quantum field theory that way. And I don't think our understanding of the Everett interpretation needs to rest on the possibility of doing that. If we can do it, great. Are there any challenges with the Everettian interpretation and quantum field theory in particular? I don't think QFT brings up any particular problems for Everett that don't occur elsewhere.

Starting point is 00:39:09 So both QFT and Everett have interpretational conceptual problems. Everett has questions about what's the nature of the branching structure, how does probability work? And in QFT, there were questions about how do we think about normalization, infinities, all this kind of stuff. I don't think you acquire a further novel problem when you put those ideas together. Okay, I want to linger on this word physical. Are people physical? Well, I guess, was I using the word physical? I shouldn't be a bit

Starting point is 00:39:45 careful about it. I mean, people exist. Some aspects of people are usefully studied by physics. And I'm enough of a reductionist to say that a sufficiently precise physics description of a person would give you an accurate probability prediction about what their physical state will be in the future and that other facts about them like, you know, which political party they support or something ultimately supervene on those physics facts. So that in that sense, people are physical. And does something physical have to emerge necessarily from something else physical? Uh, well, again, I don't think I know how to say what physical means a priori outside the concrete context of the physics we have.

Starting point is 00:40:34 I mean, uh, you know, you could imagine a world with weird pluralistic dynamics. And some of them would be some as you'd call physics and some you'd call some other dynamics or science or something and then maybe emergence would be this weird complicated framework. I mean it's clearly not true that something biological has to emerge from something else biological. It's certainly not true that something that's usefully studied by the theories of electrical conductivity has to emerge by something else from something else that's used to study by the methods of electrical conductivity. So the claim that everything physical is emerging from something else physical is basically relying on the kind of dynamical priority of physics, I guess. I think the evidence

Starting point is 00:41:21 for the memable priority of physics is pretty good. I don't think we live in that kind of disconnected patchwork, pluralistic world, but it's not conceptually impossible that we could live in a world like that. And the evidence is compelling, but not, I think, totally unchallengeable. I mean, that's a broader question about emergence, isn't it? I mean, I think you hear a lot said about the autonomy of different levels of science, but we, and then people will talk about the extent to which some of our biology is autonomous from physics and things, but, and there's obviously some degree to which that's true, certainly it's true methodologically, but it's also true that there are definitely context in which we're confident that physics will give the right answer for a prediction,

Starting point is 00:42:04 even for a complicated biological system. If I take some exquisitely complicated animal and then I drop it in a volcano, we all know what's going to happen to it. We don't need to consider biology to answer that it will just become asinirated, even more so if I drop it in the middle of a nuclear explosion or something. and we understand why as well we're clear stories what's going on we and the story says the thing is made up of an extremely complicated tangle of atoms and molecules and the methods of physics are not very reliable in telling you what that complicated tangle will do because as we were saying earlier they're optimized towards quite simple

Starting point is 00:42:43 systems but we still think that ultimately the laws that govern the system are the laws of physics. And one of the tells that we think that is if you put the system in a case where the complexity goes under control, like you drop it in a fire, why can I predict what it'll do in a volcano or a nuclear explosion or the sun? Well, because the energy levels now are high enough that those exquisitely complicated interactions that make a living thing what it is are just not scale relevant compared to the energy scales of the volcano or the explosion. All of that, I think, is sort of part of what we mean by saying ultimately we're pretty confident the underlying dynamics of the system is physics dynamics. And what's the difference between your interpretation of many worlds and Simon Saunders?

Starting point is 00:43:33 Not very much, I think. I mean, Simon was my PhD supervisor and we worked together on this a lot in Oxford in the 2000s. There were differences of style and emphasis. I mean, I focus much more on the sort of dynamical process of emergence and the kind of language of worlds. Simon, certainly in some of Simon's earlier work, was interested in somewhat more metaphysical questions about what the analogies were between Everett branches and different instances of time in special relativity and whether somehow the ontological strangeness of the Everett interpretation was ameliorated if you thought that other worlds existed in something like the sense that other times existed. The

Starting point is 00:44:25 way I've tended to think about probability in Everett has been following the sort of decision theoretic strategies that David Deutsch developed and Simon, certainly in his recent work, has been interested in taking rather different ways of approaching probability, more to do with sort of relative frequency ideas and trying to apply those ideas in Everett. But these are relatively subtle distinctions and it's not as if for the most part I think Simon's take on these things is wrong or vice versa. It's more an emphasis issue. So is it a difference as to how you assign the weights on to the different branches?

Starting point is 00:45:04 So is it a difference as to how you assign the weights onto the different branches? No, it's an it's not it's not a difference that's transparent at the level of the mathematics That's one of the reasons I don't think it's ultimately substantive. So the weights on the branch is just given by the Born Rule. That's I think that's a good that's if one abuser that I mean if they're not you've got an empirical problem but the the question is, what's the conceptual justification of interpreting the branch weights as probabilities? And you can pull that in either direction. I mean, the quasi-dismissive answer, which I have at least some sympathy for, is you can say, look, given the kind of assumptions about decahedrons and emergent classicality, which are true dynamically of

Starting point is 00:45:48 these systems, then the branch weights have the right formal property to be probabilities. What else in physics did we ever require of some piece of the mathematical structure that it did more than have the right formal properties in order for us to stipulate that that's how we're interpreting it. So you could say, look, once you've established that the mathematical structure of quantum mechanics in the immersion regime is that of a stochastic quasi-classical theory, what more do you want? And the other take in the other directions is more go something like well look probability by its very conceptual nature is something that applies to alternative possibilities or where only one thing can happen and as a more detailed level probability seems to emerge ultimately because of some fundamental

Starting point is 00:46:42 indeterminism or because of some relevant ignorance of initial conditions. And there's no indeterminism in Everett and there's no fundamental level and there's no relevant ignorance of initial conditions. So goes this objection, we just don't understand how that could possibly be probability in a theory. Right. That's that. That's how I put the case in the other direction. So different people have quite different kind of starting points here. And as a small personal story on this, a lot of my work in the early mid 2000s was about these decision theory approaches to quantum

Starting point is 00:47:18 mechanics, to Everettian probability. And I got invited at various points to give talks to kind of quantum information groups, quantum foundation groups, mostly to talk to physicists on these ideas. And I found it was quite difficult to get physicists to be, physicists even who are sympathetic to the many worlds theory, to be concerned in the first place that there was any problem of understanding why mod squared amplitude was probability. And over a period of time practicing giving talks of that kind to that audience, I got better at trying to persuade them in the first half of the talk that there was a problem, only to spend the second half of the talk persuading them that after all there wasn't

Starting point is 00:47:59 a problem. And eventually it dawned on me there was a quicker way to get to the same solution. So I started to sort of think it was perhaps less useful use of my time to confuse and then unconfuse physicists if they weren't very worried about the probability problem in the first place. But philosophers are very, very worried about the probability problem. So certainly internal to philosophy, this is very hotly contested. Okay, well, let's linger on this word decision here.

Starting point is 00:48:25 So decision theoretic approaches. What is it? So what are they just explain it more simply and then give an example. And then also what led you like, what was the decision that led you to decision theoretic approaches? Was it you attended a talk from someone else? You read a book. How did you land on that approach?

Starting point is 00:48:43 Just a moment. Don't go anywhere. Hey, I see you inching away. Don't go anywhere. Hey, I see you inching away. Don't be like the economy. Instead, read The Economist. I thought all The Economist was was something that CEOs read to stay up to date on world trends. And that's true, but that's not only true. What I found more than useful for myself personally is their coverage of math, physics, philosophy, and AI. Especially how something is perceived by other countries and how it may impact markets. For instance,

Starting point is 00:49:11 the Economist had an interview with some of the people behind DeepSeek the week DeepSeek was launched. No one else had that. Another example is the Economist has this fantastic article on the recent dark energy data which surpasses even scientific Americans' coverage, in my opinion. They also have the chart of everything. It's like the chart version of this channel. It's something which is a pleasure to scroll through and learn from. Links to all of these will be in the description, of course. Now the Economist's commitment to rigorous journalism means that you get a clear picture

Starting point is 00:49:40 of the world's most significant developments. I am personally interested in the more scientific ones like this one on extending life via mitochondrial transplants which creates actually a new field of medicine. Something that would make Michael Levin proud. The Economist also covers culture, finance and economics, business, international affairs, Britain, Europe, the Middle East, Africa, China, Asia, the Americas, and of course, the USA. Whether it's the latest in scientific innovation or the shifting landscape of global politics,

Starting point is 00:50:12 The Economist provides comprehensive coverage and it goes far beyond just headlines. Look, if you're passionate about expanding your knowledge and gaining a new understanding, a deeper one, of the forces that shape our world, then I highly recommend subscribing to The Economist. I subscribe to them and it's an investment into my, into your intellectual growth. It's one that you won't regret. As a listener of this podcast, you'll get a special 20% off discount. Now you can enjoy The Economist and all it has to offer for less. Head over to their website www.economist.com slash toe, T-O-E, to get started. Thanks for tuning in and now let's get back to the exploration of the mysteries of our universe.

Starting point is 00:50:55 Again, that's economist.com slash toe. How did you land on that approach? Yeah, sure. Okay. So in terms of what it is, so at an abstract level, decision theory is just, it goes back to people like Ron Neumann, the kind of theoretical analysis of decision making under uncertainty. So things like how we understand the principle of maximizing expected utility is an example of that content if you disconnect from quantum mechanics.

Starting point is 00:51:24 Why does that have anything to do from quantum mechanics. Why does that happen in the quantum mechanics? Well, go back to this problem of probability. If you've got this situation where everyone agrees that Modscoed amplitude has the formal property through probability, but people are still worried that it can't be probability because intuitively it's not doing the kind of, it's not in the context we naturally think about probability. Well, one way to make progress there and resolve the on-pass is to go operational and say, well, never mind whether it has the deep metaphysical nature of probability, does it behave in the ways probability does? Does it plug into our theories in the ways probability does?

Starting point is 00:51:57 So then you ask yourself, well, how does probability plug into our theories? And at least one very substantial aspect of probability plugging into our theories is decision theoretic. It's in the fact that when we say something in the future as high probability it translates to the fact that we're prepared to, well in game theory type terms to bet on it with high odds. So if you're to say that you're 99% confident that, oh I don't know, the sun stupidly doesn't know, you're in Death Valley, you're 95% confident it's going to be sunny tomorrow. What you mean by that is something like if somebody offered you a bet on whether it's going to be sunny in Death Valley tomorrow, you take that bet at 21 odds and you definitely take it at 2 to 1 odds, but you wouldn't take

Starting point is 00:52:49 it a million to 1 odds because it's not always sunny in Death Valley. Right. So at least in a certain tradition that goes back to sort of mid-century politicians and mathematicians, then it's kind of constitutive of probability that it plugs into our decision calculus in this kind of way. So if you could establish that the mod squared amplitude needs to plug into our decision calculus in this kind of way, then you'd make a big step towards understanding why the mod squared amplitude plays the role of probability in our science. Perhaps to connect to something a little less stylized and betting on

Starting point is 00:53:32 death value and some bit more connected to scientific experiment, you might imagine you're testing a scientific hypothesis. Let's say your theory predicts a certain value for the half-life of the neutron. How do you test that theory? Well, you get a big stock of neutrons, you measure them and you see how many have decayed in a certain length of time and you say, okay, that's the half-life. And if that number matches your theory, you say, hooray, my theory is confirmed, Nobel Prize, please. And if it doesn't match, you

Starting point is 00:54:04 say, damn, you throw the theory out. But of course there's a bit of risk going on there because each of those radioactive decays is probabilistic. If I say the half-life of the neutron is, what is it, 800 seconds, I can't remember, it doesn't matter. If I say the half-life of the neutron is 800 seconds, I don't actually mean that if I take a sample of neutrons, exactly half of them will decayed in 800 seconds. I mean that each individual neutron has a 0.5 probability of decaying in 800 seconds. If I've got a thousand neutrons, there's a one in two to the thousand chance that none of them will decayed after 800 seconds. One in two to a thousand is not

Starting point is 00:54:45 a large number, but it's not zero. So any time you're confirming or falsifying a probabilistic theory, you're always kind of sticking your neck out. You're saying something like, well, I think it's extremely likely that the theory is true because I think conditional on theory being true, it's extremely likely I'd see this and I did. Conditional on theory being false, it's extremely unlikely I'd see this, but I did. And so my decision, if you like, is to publish the paper, claim the result is true, defend it at conferences, build technology that relies on it. In a sense, there's a decision theoretic component in just scientific theory testing. If you could show that it would be decision theoretically sensible to accept quantum theory

Starting point is 00:55:40 if you've got a large amount of data that confirm quantum mechanics as predictions and reject it if you didn't, even if quantum mechanics should be understood in many worlds terms, that would be a very large step towards establishing that the probability-like things in Everettian quantum mechanics really are probabilities. So that's the answer to the why care about care about it in terms of the slightly more biographical question so David Deutsch did some work on this in the Well, I guess it wasn't published till 99 It was insert you're so being circulated informally in the mid 90s where what he basically does is he takes some

Starting point is 00:56:19 They simplified piece of decision theory and he says if you assume the axioms of this decision theory but without assuming anything about probability and if you assume the axioms of quantum mechanics but without assuming anything about probability and you put them together, you derive the fact that the decision theoretic agent bets according to the Mod's Green Antichodes. So effectively you derive the Born Rule from these non-probabilistic assumptions of decision theory and quantum mechanics. And that paper didn't get a lot of attention initially. There was a refutation or response to it written by a bunch of people in the quantum information space, people like Howard Barnum. I think Chris Fuchs was one of the authors, Sengelstein was on it, a bunch of people in that kind of approach wrote a quite interesting response paper to it. I thought they'd missed

Starting point is 00:57:11 the point which was that tacitly Deutsch was assuming the Everett interpretation, which you know if you knew anything about Deutsch you knew it was true, but he doesn't say so explicitly in the paper. So I did a bunch of work that was initially about sort of exegesis of Deutsch and trying to, you know, philosophically clarify and tidy up and reconstruct the arguments. And then I got interested in how we could go beyond those arguments, how we could get stronger versions of Deutsch's proof. So I got into getting interested in some of the general foundations of classical decision theory. And I sort of did what Deutsch did, which is take these assumptions coming from classical decision theory plus the did what Deutsch did, which is take these assumptions coming from classical decision theory plus the structure of

Starting point is 00:57:47 quantum mechanics and put them together into a proof of the Born rule. But I was interested in doing it from somewhat more non-committal and arguably less controversial assumptions about decision theory. So this is the Deutsch-Wallis theorem that you're referring to? Yeah. Yeah, I mean, that word kind of covers a constellation of results. I mean, I normally use it to refer to

Starting point is 00:58:13 the kind of result I proved in my book and in early work led to that. Deutsch wasn't directly involved in that work, but it's very much inspired by his earlier work on it. Speaking of your book, in case people have just skipped forward, the book is called The Emergent Multiverse, and the link is on screen and in the description, and I highly recommend it. So does the proof of the David Wallace, sorry, the Deutsch-Wallis theorem, does it assume non-contextual assignments in decoherent subalgebras? I don't think so, but I'm not 100% sure I get what you mean there. So, oh, you're saying

Starting point is 00:58:54 does it assume non-contextuality in the sort of Gleason's theorem sense of it? No, although the way, non-contextuality is a consequence of some of the things it does assume. So to some extent, it's giving you a justification inside the Everett context of the non-contextuality assumptions. Um, so the, uh, I don't know how technically one is an answer to this and I actually haven't thought about it for a while, so I may confuse myself. A lot of what's going on in my versions of these proofs relies on the physicalization

Starting point is 00:59:29 of the processes to which there is grinding probabilities. So if I am, if you have a framework where measurement is primitive, then of course it's a primitive matter as to whether you want a contextual and non-contextual probability assignment across measurements. So if you just want to say, look, I am assigning measurements to algebras of, you know, Boolean algebras of commuting projectors or whatever, and to every physical measurement I do in the lab, that corresponds a projection-valued measure or positive-operative-valued measure, what's the rule for the correspondence? Don't ask me. It's primitive. God told me. If that's your starting point, then of course non-contextuality can't be proved. It has to be assumed. But in the Everettian framework, of course, all of these measurements are physical processes.

Starting point is 01:00:18 What's actually going on is some complicated set of unitary interactions between the system being measured and the stuff measuring it. And one and the same physical process might be considered to be many formal quantum measurements. And one and the same formal quantum measurement might correspond to many different physical processes. So suppose for instance I'm measuring the spin of an electron say, or that's the spin of a silver atom in a Stern-Gerlach framework. So if you're just being axiomatic about it you'll say fine this is a spin measurement and my spin measurement is represented by these projectors onto the spin up subspace and onto the spin down subspace. And there'll be a separate thing you might do which is position measurement, which is going to be represented by some you know,

Starting point is 01:01:11 if any, I'm going to cash it out, some kind of collection of projectors onto coarse grained regions of position space. And all of that will just be primitive, which will just stipulate that that's what these things are. If you've got a dynamical story about measurement, if you ask how do I actually measure spin, for instance, well, I say the way to do it is I've got an inimmoginous magnetic field and I separate the beams so that some of the particles, the ones that spin up, in classical terms, the ones that spin up go towards the top of the apparatus and the ones that spin down go towards the bottom of the apparatus and then I measure where they are by slamming them into a screen or something. So was that a measurement of spin or was it a measure of position? Well it's the same physical process whether I decide to regard it as a measurement

Starting point is 01:01:53 of spin or a measurement of position is just a convention of my part. And so there's a necessary connection between the self-same physics being described by those two different choices of projectors that put some constraints on how you want to assign probabilities across them. So the audience may be confused because they hear decisions which they think of as tied to agents, which they think of as tied to people slash observers. And we were trying to come up with the physics that was mind independent or realist. So let's go back to the early part of the universe before there were no observers,

Starting point is 01:02:35 before you could even have a Dutch book argument because you have no concept of cost. Like a quark glue on plasma doesn't have a concept of cost, seemingly. Sounds right. Okay, so help them understand how quantum mechanics applies when we're thinking about something that is decision theoretic, which is predicated on agents, which seems to be predicated on minds and people and so on. Good. Yeah, no, it's a good question. So what can we say objectively about the world according to Everettian quantum mechanics?

Starting point is 01:03:04 It contains a whole bunch of stuff with weird quantum properties, but because of the way the dynamical interactions between the stuff happens, then on a coarse-grained description, the system evolved, the collective degrees of freedom of the stuff evolve so as to have the same formal structure as a stochastic dynamical process and in terms of what's the physical goings on that's being represented it's basically a whole bunch of parallel goings on. There's no interference between the various different ways for instance structure forms in the early universe, I have a superposition of all the way structure form in the early universe, I have a measure over those ways given by the Mod's grand amplitudes, and those different ways don't interfere with each other. So that measure compounds over time in the same formal

Starting point is 01:04:00 way a probability does. So all of that's observer independent, all of that stuff you can say about the early universe long before the humans. Does that measure that I said has the formal properties of probability, is it probability? Well, to some extent, that's a semantic definitional question. But one way you can preciseify it is to is to say well does it play the same? Operational roles as probability does well some of those operational roles are just things about dynamics I mean, so for instance does it does does it does it compound over time the way probability does yes it does again that's totally objective fact totally independent humans the Mathematical one of the things probability does is obey certain synchronic diachronic axioms, debase the Kolmogorov axioms in instant in time, it compounds over

Starting point is 01:04:53 time in accordance with various updating rules, Mod's Grammitude does all that stuff. If you're happy that doing all that stuff exhausts the nature of probability, hooray, you're happy with Maywell's theory. If you're not happy, then you want to know, okay, what else has been left out? And arguably what's been left out is the probability itself has to have a certain conceptual connection with the scientific method. That's a defensible, I think that's probably right, in any case, it's def is defensible that statement is independent quantum mechanics. It's the statement that there has to be a conceptual connection between according to the theory X happens with extremely high probability and an optimal use of the scientific method is such that X should be accepted as supported by the evidence. So that's the place in which the connection happens. If you think that

Starting point is 01:05:55 it's part of the conceptual nature of probability that probability statements have to interact in the right way with methodological statements about scientific experimentation, then you're going to have to say something about what would happen in experimental contexts and what would be an appropriate method there. And this is a little more controversial. I want to claim that ultimately if you want to analyse what optimal scientific method is, that's a decision theoretic question. And at that point you can't avoid some consideration of the scientists. They don't need to be like rich, fully-fleshed humans with deep desires and wants and needs. They can be extremely minimal algorithmic devices

Starting point is 01:06:45 set up there to collect scientific data. But we can still ask what's the correct strategy for those systems to adopt. Does that help? Yeah. Now, where I'm confused is that earlier we talked about the many worlds as having a lesser set of axioms, at least it's put forward as such, as one of the advantages to this interpretation. And then it seems like to derive the Born Rule, there is still the introduction of rationality axioms, like ordering or diachronic consistency, or branching state indifference.

Starting point is 01:07:20 I believe there's state supervenience you talk about in your book as well. So do you see these as extra ingredients that come along with the minimalism? Is it no longer minimalism? Like how do you view this? Yeah, so I want to claim these are constitutive assumptions about what agency is and what rationality is and therefore indirectly they're constitutive assumptions about what science is. And those assumptions are neutral in themselves, they're neutral to the many-wells theory. So it's true that if you want to establish that probability in the Everett interpretation, or the multiple amplitude in the Everett interpretation plays the role the probability plays in scientific

Starting point is 01:08:06 inference, then you're going to have to say something in the set-up of the problem about what you take scientific inference to be, and not just something about what the dynamics of quantum mechanics are. But that's not specific to Everett. I mean, if you wanted to establish that some formal measure in any theory played the role the probability plays in the scientific method, you need to say something about scientific inference as well as something about that theory. If you don't feel any need to connect something that has the formal properties of probability to scientific inference, then congratulations, you don't need to say anything about scientific inference and ever by itself will be fine for you. But if you do think it's part of your job to connect

Starting point is 01:08:47 scientific inference to multiple-amplitude, then logically you're going to have to say something about scientific inference. So if you look at what I do say about scientific inference, there's sort of two classes of things I want to say. One class of them is things that I take a constitutive in what it is to be a rational agent and in particular a scientist. And those assumptions are pretty minimal. They're basically just that one can consistently attribute to a physical system over time a pattern of preferences and intentions and that that pattern is consistently ascribed. As I want to say, look, if you've got some complicated biochemical system, but that claim, those claims are not true of it, even in idealization,

Starting point is 01:09:31 it's not meaningful to call it an agent. This isn't like some kind of intuition as what agent should be. It's constitutive of what it would be to be a system that enacts a strategy over time, if you like. And then there are modeling assumptions about how physical systems of that kind could be realized in a quantum world So for instance, I assume that any physical strategy performable by a Human well, it doesn't human a rational agent agent, is going to have to be continuous in Hilbert's space measure, in the Hilbert space topology rather. And the reason for that is just going to be that it, and I'm not exactly sure how you prove it, but I want to claim it's dynamically obvious, that it will not be possible to build any structure that itself follows the unitary dynamics and instantiates

Starting point is 01:10:27 the things it does in unitary records and so on and stores it in unitarily controlled memory data. You won't be able to do any of that stuff unless you're doing it in a way that's continuous with respect to the Hilbert's based topology because the Schrodinger equation is continuous with respect to Hilbert's based topology. So it's assumptions of that kind. I mean, there are perhaps slightly more protestable claims there, but the point is that the kind of assumptions I do that kind of connect decision theory to quantum mechanics are supposed to be modelling idealisations of that kind. They're not supposed to be separate axioms of the kind that one could coherently consider their falsity. They're supposed to be, again, some mixture of sort of constitutive definition and realistic claims about what kind of physical systems

Starting point is 01:11:12 could exist in idealization. Okay. Speaking of coherently, what makes a world a world? That it dynamically can be modeled by a set of equations autonomous to that world, which won't be messed around with by interference with other such things. Now is that a continuum? Like initially they can interfere somewhat and then eventually do they ever interfere zero or is it just vanishingly zero? Yeah, it's going to be vanishingly small.

Starting point is 01:11:50 These things are emissive degrees. That's a very general feature of emergence in the sciences. So what is it to be a fluid is characteristically defined in fluid dynamics as a system which is not resistant to say shapes doesn't work at all to keep its shape. And this is a heavily heuristic way of putting it. I should remember fluid dynamics better and do it better. But basically that's the idea of fluid. Solids keep keep their shape fluids deck. Okay, um, but of course everything keeps its shape a little bit there always time scales and energy scales on which there's a little bit of willingness to To keep your shape and likewise solids don't completely keep their shape perfectly if I take a block of granite

Starting point is 01:12:39 I leave it sitting on the surface of some Dead planet for a billion years and I come back it's going to change a little bit. So what we really know it's not completely and categorically true that we have a totally sharp line here between fluids and solids. Yeah, but come on in practice. While there are genuinely things where it's not quite obvious whether you want to say they're fluids or solids, there's tons of stuff that blatantly are fluids on any reasonable way of making it precise and tons of things that are solids on any reasonable way of making it precise.

Starting point is 01:13:15 You might similarly say like, does the Earth have an atmosphere? Yeah, sure it does. Does the Moon have an atmosphere? No. But there's a little bit of gas on the surface of the moon and there's no completely sharp point when you leave the Earth's atmosphere and enter interplanetary space. The odd particle makes its way just through Brownian fluctuations from the Earth's atmosphere to the moon. So you can't draw a completely sharp mathematically rigorous line.

Starting point is 01:13:43 You say the Earth has an atmosphere, the moon doesn't, the Earth's atmosphere is separate from the Moon's, the Earth's atmosphere stops at a certain distance from the Earth. There's no totally magic place to do that. You'll have to draw some slightly arbitrary lines to say where it goes. But nonetheless, there's a very clear, even if not perfectly precise, extremely precise sense in which the Earth's atmosphere does not extend out as far as the Moon. Or that water is a fluid and granite isn't. And in the same sense, there's no completely precise cutting off of interference here. If worlds are defined by decohering in the formal sense and not having interference of that kind and having autonomous dynamics of their own, that's never going to be perfect.

Starting point is 01:14:29 That's always going to be an epsilon level correction. And it's always going to be arbitrary where you put the line below which it counts, but you're still going to have a very robust division. Does that mean that fundamentally speaking, not only is there a universal wave function, so a single wave function, but there is technically just one world? Well, fundamentally, yes, there's just one world in the same sense that fundamentally there are no planets, fundamentally you don't exist, fundamentally there's nothing except

Starting point is 01:14:59 complicated excitations of the quantum vacuum. So there's a sense of the word fundamentally where you could use that. And for some purposes, that's quite useful. But it can be misleading. So yes, fundamentally, there are no worlds, but fundamentally, there's almost nothing. But what would you say to those who are saying that fundamental physics is attempting to capture what's occurring fundamentally? Well, no, I don't think that's entirely right. I mean, there's a reason the book's called is attempting to capture what's occurring fundamentally? Well, no, I don't think that's entirely right. I mean, there's a reason the book's called The Emergent Multiverse.

Starting point is 01:15:30 I think physics and science generally is trying to capture what's happening. Some of what's happening is what's happening fundamentally, but there's more to life than what's happening fundamentally. Quantum mechanics, for instance, not quantum field theory or quantum gravity, quantum mechanics, the kind of thing you field theory or quantum gravity, quantum mechanics, the kind of thing you study as an undergraduate does not try to capture what's happening fundamentally

Starting point is 01:15:49 because it's studying non-relativistic electrons and atomic nuclei treated as point particles. Those aren't fundamental. So sure, non-relativistic quantum mechanics is not fundamental physics. Guess what? Quantum electrodynamics is also not fundamental physics. It's an effective field theory that's applicable at energies below the electro-week symmetry breaking scale. So are there fundamentally any electrons? No. Is QED a fundamental theory? No, nothing studies this fundamental. Nothing the standard of the studies is fundamental, really. I mean, it's an effective field theory again that's descriptive below Planck scales. We don't actually have any empirically confirmed fundamental theories. Fundamentality is an aspiration, but we're selling physics short.

Starting point is 01:16:32 We think the only thing physics is telling us about is fundamental stuff. So what is fundamental physics then? Well, the way people often use fundamental physics is much broader than that. And I think it's semi sociological what they mean by it, but some in this some sense of fundamental, where fundamental applies to, say, atomic physics, maybe even applies to like classical Hamiltonian mechanics, the context in which you might say the classical electromagnetism is fundamental physics, but climate science isn't fundamental physics. I think that's kind of a relative division.

Starting point is 01:17:09 Physics studies the world on lots of energy scales. Some of the systems it studies have relatively few moving parts and can be studied in a relatively complete way. Some of them have an extremely large number of moving parts and have to be studied by the techniques of statistical mechanics and other sort of approximative methods. Generally speaking, we tend to use fundamental to refer more to the first sort of physics than the second. But in the strict sense that philosophers tend to talk about, they tend to mean fundamental physics. I actually don't like this usage, but to go with it, they tend to use fundamental physics to mean something like that branch of physics which is concerned with, you know, the most, even this is dodgy for activity, the most fundamental distinctions

Starting point is 01:17:52 of nature, nature exactly without approximation on all energy scales. And in that sense of fundamental physics, no, there is no theory of fundamental physics that has any experimental support. String theory genuinely is aspirationally a fundamental theory in that sense. String theory aspires to be a theory of everything in the classic sense, not in the sense that it will predict all the interesting goings on. No one thinks string theory will tell us he'll win the next election, but in the sense that it does do that kind of lowest grade discussion. That's

Starting point is 01:18:28 a glorious goal. I'm a big fan of string theory. By all means, let's pursue what we can get there. And there are reasons to think that the path towards more fundamental physics has been a path towards shorter and shorter length scales and that ultimately quantum gravity tells us that that search will end, that there won't be any length scales and that ultimately quantum gravity tells us that that search will end, that there won't be any length scales below a certain scale. And so I don't think it's quixotic or absurd to look for a fundamental theory in that sense. But we misunderstand physics if we think that what we love the world for physics is knowing about the fundamental. Okay, so let's disregard the fundamental and focus on the practical somewhat, even though

Starting point is 01:19:07 this is a loose way of speaking. Sure. So, personal identity. The person who's listening, when they are splitting, quote unquote, in this many worlds interpretation, what is exactly splitting? Is there more copies of them? Like is their identity splitting? Are there more copies of them? Like is their identity splitting? Are there divergent successors?

Starting point is 01:19:27 How are they supposed to think that look, if the many worlds interpretation is correct, what does that imply for them? Hi everyone, hope you're enjoying today's episode. If you're hungry for deeper dives into physics, AI, consciousness, philosophy, along with my personal reflections, you'll find it all on my sub stack. Subscribers get first access to new episodes, new posts as well, behind-the-scenes insights, and the chance to be a part of a thriving community of like-minded pilgrimers. By joining, you'll directly be supporting my work and helping keep these conversations at the cutting edge. So

Starting point is 01:20:01 click the link on screen here, hit subscribe and let's keep pushing the boundaries of knowledge together. Thank you and enjoy the show. Just so you know, if you're listening, it's C-U-R-T-J-A-I-M-U-N-G-A-L dot org. Kurtjaimungle dot org. What does that imply for them? Yeah, okay. So there are two different ways to talk about this. One way to talk would be to say, I'm going to measure a just a sterling gullock experiment and maybe I'll see spin up, maybe I'll see spin down. So I'm going to have some, in the future, there are going to be some Davids who see spin up,

Starting point is 01:20:35 there are some Davids who see spin down. Now, depending what you think the reference of David is, you could say, well, Right now there's one David and that David will split into lots of David's So I right now will have lots of different future experiences or you could say What I David am should be thought of as a four-dimensional entity And so there's actually lots of David's even now Right now they're all identical and they're going to become different in the future.

Starting point is 01:21:08 In which case, it's multiple who's speaking at the moment as lots of David speaking. Now the same physics is underpinning both of those stories. There's no difference between those stories at the microscopic level. There's a difference at the level at which we decide to individuate and talk about large-scale persistent objects like humans. In my own view, this is a termological choice. We can choose to talk in the three-dimensional idiom, in which case I split, or we can choose to talk in the four-dimensional idiom in which case I split or we can choose to talk in the four dimensional idiom in which

Starting point is 01:21:45 case I diverge. I don't think it's factive which of those is true. That is controversial. There are philosophers including philosophers who are sympathetic to the many worlds theory like Al Wilson who will say it's a substantive matter which of these is true and in fact who will normally then defend the diverging story. Which of us is right turns on sort of methodological questions in metaphysics I guess. Is every single splitting that's physically possible, does that outcome occur or are there stronger constraints on the quantum dynamics itself for what constitutes a viable world? That's what you mean by possible. I mean every, well there's two things going on. So firstly,

Starting point is 01:22:32 if you have a sufficiently low amplitude branch, this goes back to what you're saying about is there no interference at all. If you had a sufficiently low amplitude branch, then you could expect that it won't actually have a coherent evolution. The noise from interference with other branches is going to wash it out. That noise is extremely small and drops off extremely quickly, but it isn't zero. So to take an example you often see in seeing physics textbooks if you ask like, you know, you might get you sometimes you might have to calculate what's my amplitude just to tumble through the wall? You know, you can, this is the guy remember doing that as a problem sheet problem as an undergrad

Starting point is 01:23:18 physicist, you know, you do it for a particle and of course it's in a potential and it could be quite a reasonable number and then you find appropriate numbers for a human and an actual piece of concrete and you get out some eutically double exponentially small number. From that point of view, although I'd want to try the modeling, my expectation is that there aren't any branches in which I tunnel through the wall because that amplitude is so small that the branch can't be defined as an emergent entity. It's just washed out by interference effects by reasonable size branches. So in that sense, I would expect that that's one sense in which not everything that in a certain sense is possible happens. The other thing to say is

Starting point is 01:24:01 of course only those things with non-zero amplitude will happen. So if there's something which is a dynamically permissible evolution according to the Schrodinger equation, but the actual quantum state is such that it gets amplitude exactly zero, then it won't happen. However, as a practical matter, it's quite difficult to come up with things like that. So within those constraints, the space of things that happens is extremely permissive. I mean, one way to think about it is that any sequence of outcomes from a series of quantum experiments, however unlucky, unlikely is going to be seen. So if I measure trillions of spin path particles, there will be a branch in which I get spin

Starting point is 01:24:42 up every single time. So if we're only going to allow the positive probabilities, maybe this is a foolish question, but if you have a continuous spectrum then, let's say position, then doesn't anything that occurs in this continuous spectrum have measure zero? No, because individual eigenstates of position aren't a good choice for the decoherent basis. If I consider histories that are defined by fine-grained positions, they're not going to decoherent. If you want to find something that decoheres, that actually shows this not having any interference or negligible interference, you're going to have to course grain a bit. So the kind of plausible basis you might find is you might say, interference, you're going to have to coarse-grain a bit.

Starting point is 01:25:25 So the kind of plausible basis you might find is you might say, okay, I'm going to look at the... firstly, I'm going to look at not the position of every particle in my body, but the kind of coarse-grained averages of positions of particles in my body. So I'll look at my mass distribution averaged over cells a micron across or something. And even then I'll, I'm interested in projectors of small but not zero width. So I'm not, I'm not just doing a projector onto an exact configuration of that coarse grained mass distribution. I'm asking for it up to some farce. Again, not much farce, but a little bit.

Starting point is 01:26:07 That's going to give me a discrete, normally finite set of projectors. And then you're not going to have those kind of worries you're raising. Could you still have discrete but infinite? I mean theoretically if you had a, well I suppose either you could be looking at a spatially infinite system or you could start just clumping them smaller and smaller and smaller together. Whether that's actually going to give you a decoherent branch is going to depend on the dynamics. You need to actually check the equations. Normally speaking if you start looking at things on smaller and smaller scales, eventually you'll reach a point where quantum

Starting point is 01:26:48 interference is going to become unnegligible and then the branching language is going to break down. As for things on a spatial infinite scale, I mean, ask your cosmology. My feeling is generally speaking, we know almost zero about real spatial infinities in physics and essentially every time we find ourselves using a spatial infinity it's a modelling idealisation. So I'm generally suspicious of spatial infinities in that sense. And after all the De Sitter horizon of the expanding universe is finitely far away and anything beyond that horizon is going away from us faster than the local speed of light. So it's, what should I say, practice, light will never get from us to it. So in that sense, we're never going to have a physical situation in which a spatially infinite region is going to matter dynamically. What is strong emergence? And do you believe that the only kind of emergence is of the weak kind?

Starting point is 01:27:50 So this language is slightly difficult to pin down. I mean the short answer is yes, of a higher level physics that's not derivable even in principle even with infinite computing power from the microscopic physics. The classic example of people who believe in the hard problem of consciousness. Yeah, lots of philosophers and scientists who, I think wrongly, think consciousness is a fundamentally inexplicable phenomenon, microphysically, they think consciousness is strongly emergent. If the collapse of the wave function was simply something that had to be understood in irreducibly high level terms, that would be strong emergence. I don't believe in strong

Starting point is 01:28:46 emergence. I mean, belief is a weird thing to say. I think the evidence of strong emergence is very weak. I don't think it's conceptually incoherent, but I think we've got no reason to think it exists. There are intermediate things you might imagine. I had a paper on this from a few years ago. In a classical world, at least, you might imagine that there were kind of regularities at the macro level that were consistent with microscopic physics but were nonetheless not derivable from microscopic physics in the sense that they came about because they were encoded in very delicate correlations in the initial state of the universe, if that were true that would be a form of emergence that in some ways would be more like strong emergence

Starting point is 01:29:32 than weak even though it wouldn't be an incompatibility with the laws of physics. I don't understand that. Can you explain that? Yeah, sure. So imagine I've got a classical world, yeah, and I've got some micro dynamics and then I'm interested in what's happening macroscopically at a coarse grain level. So the normal way we do statistic physics is that we derive from the micro physics plus some kind of statistical assumptions about initial conditions, some sort of macroscopic dynamics. Boltzmann's equation for how gases evolve is a classic example of this. And most of those macro equations that are derived are stochastic. There are lots of ways things

Starting point is 01:30:17 might turn out, but none of them, but many of them are unlikely and you have a probability measure over them. So you get in. For Boltzmann's equation, the gas is extremely likely to spread out to cover the whole of the box, but it might not. Okay, so if you're in that situation, ask yourself how much you can say about the macroscopic dynamics if you only know the microscopic laws of physics, especially like again, we're pretending a classical here for simplicity. Sure. And the answer is almost nothing. I mean, some ways the macroscopic stuff might evolve

Starting point is 01:30:57 is flat out impossible, like it violates energy conservation or something. But most of the things we don't expect to happen are not flat out impossible in that sense. I mean, I'm sitting in a skyscraper at the moment. Is it flat out impossible that in the next 10 seconds it could collapse into a swirling mass of dust that reassembles itself into a massive bust of Donald Trump? No, it's not flat out impossible. The bust would have the same energy as the rest of the building. There's no other flat conservation law being violated. If we pretend everything is classical, there'll be some ludicrous, in microphysical terms, ludicrously implausible dynamics that

Starting point is 01:31:38 gives rise to it. But it's not flat rule out. So turn that around. There's some choice of initial conditions or choice of probability measure for initial conditions such that the skyscraper turning into a bust of Donald Trump is certain. Just take all of the micro-states that are compatible with that happening and evolve that distribution back in time to the beginning of time. Again, continuing to pretend things aren't quantum. Evolve it back to the Big Bang, you'll get a weird, totally indescribable probability distribution over initial states, but it's meaningful. Nothing is flat out contradicted in the microscopic physics by claiming the distribution looks like that

Starting point is 01:32:26 so That's what I mean about the fact that you could have Macroscopic phenomena that are completely inexplicable Microscopically without being flat out Inadvertable with it. It's it's slightly more delicate to play this game in quantum mechanics And it requires a little bit of a broadening of the quantum formalism, but you can basically do it to play this game in quantum mechanics and it requires a little bit of a broadening of the quantum formalism but you can basically do it. So that's a form, there's some ways in which I think that kind of thing is more what the strong emergence people have in mind because the thing about strong emergence in other

Starting point is 01:32:54 senses is it can sometimes look as if it just flat out contradicts the microscopic laws of physics. You know the wave function collapses for instance, then that just flatly contradicts the Schrodinger equation. So there are ways you could imagine having Maxcopic physics that was compatible with microscopic physics but inexplicable from the Maxcopic physics because effectively you coded everything into the most delicate boundary conditions you can imagine. I don't believe in that either. Again, I don't think

Starting point is 01:33:19 it's incoherent things that could happen, but I don't think there's any evidence for it. There's lots of evidence against it. I think we have a lot of evidence that our world is weakly emerging, which is to say lots of interesting novel stuff happens at higher levels. And the kind of methodology we need to study things at higher levels is not extractable from low level methodology, but ultimately higher level regularities and autonomy is now the playing out of phenomena which bubble up from the lower level. What about consciousness? What do you think about it? I mean the immediate thing I think about it is that I'm not a philosopher of mind so don't

Starting point is 01:33:58 assume anything I'm going to say here is terribly original or deep. My basic take on consciousness is I think something like this broadly damned end of line unconsciousness is correct. We need to say we are conscious, but we have lots of inflated metaphysical views of our consciousness which are about taking our intuitions about consciousness too seriously and fundamentally there's no clash between consciousness and a complete microphysics. But I mean, I can say things to defend that claim, but none of them would be original to me. They're just really endorsing positions in philosophy of mind that I find persuasive.

Starting point is 01:34:36 Now, have you collaborated further other than the Wallace-Deutch theorem with David Deutsch? Not formally. I mean, when I was in Oxford, he and I chatted periodically. He's a deeply interesting guy. But no, we don't have any formal collaboration. I'm curious if you have any disagreements with Deutsch on any aspect of physics, but maybe in particular, many worlds.

Starting point is 01:35:04 Not massively. I mean, Deutsch is very committed to a quite specific approach to scientific epistemology, which is very, very much sort of picked up from the kind of way Karl Popper approaches these things. Deutsch is in some way the last Popperian. Most of philosophy of science, I think people would say that Popper had very important insights, but there was also a lot wrong with what he thought and that by and large we don't need to put things in straightforwardly Popperian terms. Deutsch is very Popperian. And that means some of the various ways people think about deriving probability are, to him, are based on wrong philosophy of science. I'm a bit more pluralist about that. I think

Starting point is 01:35:53 scientific methodology is messy and complicated business and different ways of thinking about it get at different aspects of how it works. And so I'm a bit more relaxed than I think David is about how to think about those questions. But those are more questions of scientific methodology and philosophy than they are of first order physics. I don't think there's a lot we disagree with, disagree about at the level of the flat physics. You're proponent of the Danetian real patterns. Can you explain what real patterns are? Yeah, I mean this is a somewhat heuristic way to think about immersion dentology. So ask yourself what you're saying if you say that there are

Starting point is 01:36:38 macroscopic objects that are emergent from macroscopic objects. So like I'm, I'll use a relatively mundane internal to physics example rather than going off towards consciousness. You know, I'm sitting talking to you at the moment on my laptop. Does my laptop exist? Well, it seems to be doesn't it's here. I'm currently looking at it. But there's also a whole bunch of atoms that comprise my laptop. So what am I saying when I say that in addition to the atoms that comprise the laptop, there's a laptop? And there's a bunch of things that philosophers have said about that. One thing people have said is they'll bite the bullet and say, no, there isn't really

Starting point is 01:37:21 a laptop. That's a fancy way of talking, but it's not genuinely true. That's kind of difficult to reconcile with the way we actually use language and apart from anything else, if there isn't really a laptop, because it's not fundamental, then going back to what we were talking earlier, there aren't many atoms either because atoms aren't fundamental. We have no idea what that actually is because we don't have a fundamental theory. So something seems problematic there. You might want to be just very pluralist and say, look, yeah, there's an atom level description of what's going on, a laptop level description of what's going on, but we don't know anything much about how they connect together.

Starting point is 01:37:52 But... A laptop level description. Yeah, but the thing is, of course, you know, Intel built the chip in the laptop and Microsoft put together the bulk of the laptop using physics principles. They didn't kind of just make it up. So actually, we seem to understand a hell of a lot about the relation between the atom level and the laptop level. So that's not very plausible. Then people sometimes say, well, the laptop just is another word for the atoms in it. Philosophers have this term, myriological sum, which is like, it's sort of a bit like saying that the laptop

Starting point is 01:38:24 is the set of all the atoms in it, it's not quite that. In other words, one has this idea of composition as just a primitive metaphysical idea, walls are composed of bricks, laptops are composed of atoms. That's also scientifically problematic because inter-level relations look a lot more rich and pluralistic than composition. So if you think about, say, how you derive fluid dynamics from microphysics, we don't really... The fluid is not in any very simple way just the myriological sum

Starting point is 01:38:57 of the atom, the part of the else. Most of the fluid occupies all of space, whereas atoms are mostly empty. So there's obviously some sense in which it's true that the fluid occupies all of space, whereas atoms are mostly empty. So there's obviously some sense in which it's true that the fluid is mostly empty space, but there's a very important sense in which it's not true that the fluid is mostly empty space. So that notion of mereological composition is a bit problematic as well. What Dennett offers us, and his reasons for bringing it up are slightly different from the thing I take from it, but ultimately it's a matter of my thing, is well, a good general thing to say about macroscopic ontology is macroscopic stuff is patterns and structures in learnable stuff.

Starting point is 01:39:39 So a pattern is a slightly abstract thing, but nonetheless nonetheless I can say, look, the water is a certain pattern in the behaviour of the atoms. The laptop is a certain pattern in behaviour of different atoms. The economy is a certain pattern in the behaviour of humans and companies. The atom is a certain pattern in the behaviour of quantum chromodynamics. And that's a, there's more to say here, I think it's a little bit into the weeds of metaphysics, but as a starting point, that idea of patterns is a bit more ontologically flexible to make sense of the relation between theories on different levels

Starting point is 01:40:18 without either giving up on what we seem to know about scientific reduction and inter-level relations in terms of physics, or committing ourselves to these very specific and slightly Procrustian ways of reconciling macro and micro, these things like these specific kind of mereological sum or composition relations, if you like, is a patterning, is a much more flexible and open relation between theories than is composed of. So it's one way of seeing where it comes from. Okay, so I have two questions. You can explore both of them if you like, but I'll just lay them out.

Starting point is 01:40:57 So one is, I was asking about real patterns. In the word real patterns is the word real. So people think of what is real, and then you said, does the laptop exist? Your word was exist. So people think of what is real and then you said does the laptop exist? Your word was exist. So then the question is, well, what's the relationship between real, what's real and what exists? Does everything that is real necessarily exist and vice versa?

Starting point is 01:41:15 Do we have an equivalence between those? So that's one question that occurred to me. Then another is, well, what's the precise definition of pattern? So feel free to tackle whichever one you like. Good. I mean, let me do this in reverse order. Dennett doesn't offer a precise definition of pattern. And I think he doesn't on the grounds that he thinks our understanding of these ideas

Starting point is 01:41:39 in practice from just looking at what we do in science is more reliable than trying to develop a detailed metaphysical account of it and respond to counter examples. I think he was concerned probably not without reason that if he tried to do it that way, which is like sort of metaphysics best practice, then he'd spend the rest of his life defending the metaphysical structure of this idea. And he didn't want to do that, he wanted to use that idea in the things he really cared about, which is making sense of the mind-body problem. And somewhat similar for me, I mean, I use real pattern ideas in making sense of Everettian ontology, but I don't really argue for those ideas a priori, I just say, look, we have a whole bunch of examples of how emergence

Starting point is 01:42:26 in fact happens in physics. We can extract some common features of that and we can label that under Dennett's notion of a real pattern. And we can take that set of ideas and then say, well, if you apply the driver at what do we conclude and the answer is we conclude that the same kind of rationales that tell us that tables exist and fluids exist tell us that showing this cat exists and the live cat exists and the dead cat exists and they're separate. So that's what I'm doing with it and again it doesn't turn on having a precise analysis available it turns on having a good enough understanding of some other context that we can apply it in this context. That's not to say that I think it's not an interesting question in metaphysics and philosophy of science what the actual answer is to how to think about these notions.

Starting point is 01:43:16 I have an answer to that question in much more recent work, although it's a bit indirect. So the paper of mine that might be useful here is called Stating Structural Realism. Great. The link will be on screen and in the description for those who are watching slash listening. Okay. Great. The other thing about real versus exist. So yeah, I mean, I think I'd want to be quite, this is quite philosophically traditional,

Starting point is 01:43:42 quite minimal about this. I mean, if I say something exists, it just means I can refer to it. In a sense, there aren't any nonexistent things, tautologically. And in the same sense, I'd want to say there aren't any unreal things. And in that sense, yeah, I'd say real exists as anonymous. That's a delicacy in Dennett's work, frankly, because Dennett does want to draw a distinction between what he calls real patterns and mere patterns. So his example of a mere pattern is like something that's metaphysically definable but not scientifically interesting.

Starting point is 01:44:16 So his example is the lost sock center, which is the point at the center of that sphere, which is the smallest sphere you can draw around every sock he's ever lost. Okay. So, Dennett will say that's metaphysically well-defined. There is such a point. He can tell you something about how it's moved over time, but it's not scientifically useful. It doesn't play any role in scientific explanations, so it's not real.

Starting point is 01:44:38 It sounds so close to pragmatism. There's definitely a pragmatic strand in the way Dennett talks about it. What I would say if I try to develop that is that I think what's really going on is that the pragmatism is happening because we're trying to put into natural language things that are really stated in mathematical terms. So the way I think about it is that there's a, take my fluid dynamics situation, there's a mathematical understanding of a fluid as derivable from the mathematical understanding of the particles in the fluid, and that's not pragmatic at all, that's just a derivation. Our decision to use certain language to describe the fluid has a little bit of pragmatism laced into it. But that's, the pragmatism is how we use our sort of human,

Starting point is 01:45:28 biologically derived language to describe things that weren't ultimately developed to understand that way. The pragmatism isn't in the kind of underlying structure of physical reality. But I haven't argued for that, that's just, I'm just saying that's what I think about it. So would you say ghost fields are real? That's a really interesting example.

Starting point is 01:45:53 So I mean two things to say about that. I mean one is that ghosts are theoretically dispensable. I don't have to use ghosts to study a system. I could quantize a neutral gauge or something and get rid of them. There are lots of reasons why it's often convenient and I should say this is not an area of quantum field theory that I claim to be an expert so I'm staying an inch deep here. But I think quite what role you want to play for those kind of gauge-dependent artifacts is a little bit delicate and I don't quite know what the right thing to say is. I would say that insofar as describe, when I'm trying to talk, use

Starting point is 01:46:36 language to talk about this stuff, insofar as using the language of ghosts is helpful, then I should use it. And at some stage, they're like, it's not obvious there's a residual thing to say about ghosts once you've said a whole bunch of truisms about them. So you know, can ghosts be detected? No. What statistics do ghosts have? Opposite statistics. Are ghosts gauge dependent? Yes. What are the scattering coefficients of ghosts? Blah blah blah. It's not obvious to me there's a residual question once you know the answers to all those questions. How much explanatory power does math have in physics beyond enabling predictions? So I could be more specific if you like.

Starting point is 01:47:25 Yeah, I guess so. I guess what's underlying my question is, what do you make of Tegmark's mathematical universe? I actually think those are separate questions and I'll say why. I think most of physics is done at least partially in mathematics. It's not done wholly on even mostly in natural language. I mean, that's been true since the 17th century. Galileo talks about the book of the world being written in the language of mathematics. Mathematics isn't really a

Starting point is 01:47:56 language at all. It's a representational tool that doesn't map cleanly onto a human language with compositional semantics and things. You can broaden the use of the word language if you want to, to include all representational tools. When people talk about the language of love, then you could say that's a broader notion of language. So there's a broad sense in which sure math is a language, but math isn't a language in the way that English or French or Urdu is a language. And our explanations in physics are not purely given in language. If you think about it, you ask your professor to explain why does the electromagnetic interaction get stronger at short distances but the QCD interaction get weaker at short distances, a crucial important fact about gauge theories.

Starting point is 01:48:47 You can get quite a long way answering that question in various heuristics and linguistic descriptions but eventually if you just push your professor on that question they're going to go to the whiteboard, they're going to write down the equation and they're going to demonstrate in that equation how it works and show how certain coefficients contribute to it. So then that, ultimately, that ability to have at least some part of the full explanation just essentially rely on that equation is central to the way physics works. There's no prospect that we can somehow imagine the equation could be eliminated and just losslessly replaced with an explanation purely in words that uses no mathematics.

Starting point is 01:49:25 And I think what's to be said about that is just that we have representational tools as scientists and as humans. We, for lots of purposes, natural language is a great representational tool. It's not our only representational tool. We use maps, we use diagrams, and in science and in particular in physics, we use a lot of mathematics. We use it to represent the world. We don't just use it as uninterpreted formalism but that way in which it represents the world is not something that can be lostlessly translated into words. So in that sense I think mathematics plays an irreducible role in explanation unless you want to say that physics doesn't explain anything, which I think is totally plausible. And to some extent I think maths

Starting point is 01:50:12 plays roles in explanation in other bits of science, so aspects of economics or biology that use mathematics, but not quite so kind of sweepingly as physics does. Tegmark's trying to do something else, of course. Tegmark is not talking about mathematics as a representational tool. He's talking about the idea that the world itself is mathematics. I don't quite know what that means, to be honest. Uh-huh. But in any case, I don't take it as synonymous with what I'm saying at all,

Starting point is 01:50:43 any more than, you know, lots and lots of philosophers and lots and lots of, you know, people in various walks of life think that the world can be fully described in language. It doesn't follow from that that the world is just words or something. That would be a category error. That would be a confusion. So I want to say ultimately our best deep physics of theories describe the world much more in mathematics than they do in natural language. That doesn't make the world mathematics any more than it would have made the world language if the opposite was true.

Starting point is 01:51:13 Suppose Max just joined the call right now. What would be the question you have to him, like the precise question would it be, Max what do you mean? Would it just be that or something else? Probably I mean the truth is if he did, I would feel bad that I haven't read the book recently. I would need to remind myself of the position. He's a smart guy with interesting ideas. But I don't have a very strong grasp at the moment. But I think my main concern would be something like we need to... Mathematics in science is almost invariably interpreted representing mathematics, mathematics that's supposed to represent the world. That doesn't mean that can be translated into a

Starting point is 01:51:50 verbal description of the world, but nonetheless we understand it as representing. And I'm not really sure where that notion of representation lies in Max's system. So before we end, I'm going to have a question for you about what your advice is to young researchers, people entering the field, could even be researchers in adjacent fields if you like. But before that question, I have a question from Emily Adlam, who you know. I do, yes. She just said she would like to know your thoughts on the relevance limiting thesis and whether you think that self-locating information can generally provide reasonable grounds for

Starting point is 01:52:32 updating beliefs about scientific hypotheses. I don't have a systematic thesis on that and I think it's the kind of thing that one has intuitions about. And this is an extremely deep philosophical disagreement between me and Emily, I'm not going to persuade her, but there's an approach that says something like we should have a top-down idea about how we could possibly build an epistemology and from that basic top-down starting point we can say it could never be that learning something about where we are in the world told us what the world is like. That's a plausible intuition, very plausible intuition. I don't find intuition a plausible,

Starting point is 01:53:08 sensible way to do epistemology. I think our epistemology starts extremely deeply situational in the world we actually live in. And I want to say, I want to consider specific representations of reasoning systems and ask what strategies they could or would adopt and what the constraints of the physical situation they're in are on how those strategies work. And from that context, I think you can establish that in some circumstances, rational scientific systems, rational scientists would indeed update on self-locating information. But I don't think I know even what the conceptual starting point would be for answering that question without an embedding in a particular physical context. And I think the way people tend to answer the question is they kind of have,

Starting point is 01:54:11 they run into thought experiments and intuition pumps and cases which are supposed to have obvious answers. I just don't, I have a lot of confidence in that as a methodology in philosophy. So let's give some background as to what this question is referring to. What is the relevance limiting thesis? And then you could also talk about what self-locating information is. And then please tease out some more the difference between you and Emily because you said this is a deep conceptual difference between you. Yeah, I don't know what the relevance limiting thesis is, sorry.

Starting point is 01:54:43 I probably would know what it is talking about if somebody explained it, but I don't actually I don't literally know what that term refers to. Okay. I believe it's the idea that purely self locating information should not lead to an update on non indexical beliefs like general beliefs about the world. That was what I thought. Okay, that's good. I'm relieved. Yeah, so okay. So why believe that's true? This is an example of a sort of very general epistemological starting point people might have. This is perhaps an indirect way of answering your what's the methodological difference. So here's a way a lot of people

Starting point is 01:55:20 in philosophy and some people outside it think about these kind of epistemological questions. They think something like suppose I knew nothing about what the world is like. What can I conclude about what the strategies are whereby as I start to collect data I could inform myself more about what the world's like and then they start having ideas like, okay, what priors should I have over all the possible ways the world could be? And what a priori things can I say prior to any knowledge about the world about what rational constraints apply to decision making and decision collecting there? So one of the things you might think, for instance, is, well, I want to distinguish between what the world is like and where I am in the world. And you might say something like, well, evidence about where I am in the world can't possibly

Starting point is 01:56:18 influence my, be relevant to what the world itself is like, except insofar as it also provides evidence about what the world is like directly. So for instance, if where I am in the world is somewhere with beaches and cocktails, then obviously the world contains beaches and cocktails and I can rule out worlds that don't. But beyond that kind of thing, if it's just information about where I am in the world, that can't tell me which world I'm in, in the not in the Everett sense, which universe I'm in. So people come up with ideas of that kind. And how do they argue for those? Well, there's a classic form of a philosophy argument where you say, okay, here's the principle, doesn't that principle seem self evident to

Starting point is 01:57:03 you? And if it doesn't seem self evident to you, they say, well, look, here's the principle, doesn't that principle seem self-evident to you? And if it doesn't seem self-evident to you, they say, well look, here's a thought experiment where you can imagine, here's a set of consequences which entail this principle, doesn't that kind of seem self-evident to you? And I fundamentally don't believe in that way of doing philosophy. It's what my colleague, Edouard Mashery, calls the method of cases. You start with some particular example case, which is often pretty alien. You're invited to have certain beliefs, often certain intuitions about the case, and then things are supposed to follow from it. At least in these spaces, I don't believe in that methodology for doing philosophy at all. You can perhaps imagine why it's valid in something like ethics. You might say, well,

Starting point is 01:57:47 look, doesn't your principle entail that it's bad to burn babies? Surely you agree it's bad to burn babies, so you shouldn't do it. Even in ethics, I'm a bit skeptical about that, but there may be in ethics, there's some rationale as to why our intuitions should track what's true. I just don't see that in these kind of contexts. And I think one thing we've learned from hundreds of years, thousands of years of the philosophical tradition, at least in the Western canon, is that attempting to start from no prior information about the system and bootstrap your way up to a full understanding of the situation is a fool's game. I mean, Descartes tried. Yeah.

Starting point is 01:58:24 Descartes said, suppose I know nothing at all, what can I do? What can I get out of it? Descartes proved the existence of God and derived the world from it. And you know, good for him. But most of us think there were some flaws in the reasoning. And I think the lesson from that is really we start very deeply situating the world knowing lots and lots of actual things about the world. Knowing the lots and lots of actual things about the world, and we have, I think we can then start to ask ourselves, well, in particular scientific contexts, would information of a self-locating kind tell on which theories we should rationally adopt? And then I think you can give arguments that it would. And if that's right, then in the actual physical situation we exist in, it does make sense to operate that way.

Starting point is 01:59:09 But that relies on a very kind of naturalized conception of rationality and agency. And I think for reasons I don't want to put in her mouth, Emily comes at these problems from a very different starting point. Oh gosh, Professor, there's so many questions that occur to me. I can keep talking to you for probably hours more, just so you know. I've prepared for this for quite some time and there's maybe six times more topics slash questions to get to. So I thank you.

Starting point is 01:59:44 I hope we can speak again. Before you go, many people are interested, what is your advice for young researchers, students in math and physics, and also where is your head at these days? Where are you heading? Maybe a vacation is on your mind, beaches and cocktails came relatively quick. Yeah, I have small children that complicate the beaches and the cocktails. Well, to answer those in reverse order, where my head is at the moment, at least, quite research, is probably a lot of those questions about statistical mechanics and emergence and relations between

Starting point is 02:00:18 levels that we were talking about earlier, and some of those questions about quantum gravity and how to think about that. I have fragmentary interests in lots of places and they coalesce in different places, but if there's one primary thing I'm doing at the moment in the medium term, it's statistical mechanics, directions of time, relations between theories at different levels. Advice to researchers. So it's a bit tricky to answer that because I know your audience is predominantly math and physics and while I'm kind of intellectually in a lot of that space I'm not institutionally in that space. So I'm not well placed to give advice of that

Starting point is 02:00:56 kind. So I'll say a little bit but that it'll be rather vague in general. It does also comprise academics in philosophy. Also artists, people in general, people who are interested in fundamental questions. So perhaps let me say something, it won't be an exhaustive answer, but something about how to think about these doing conceptual work in physics. You shouldn't only shut up and calculate, but you do need to do the calculating bit. There's a tendency for people interested in the foundations of physics to read the first couple of chapters of the textbook, read the axioms of the theory, if you like, get deeply confused, head into the, I think there's usually correctly, there's

Starting point is 02:01:53 a deep conceptual problem here, head into the foundations of that problem, read lots of the philosophy literature. Much very good work has come out of that. But I recommend reading the rest of the book too. We don't really understand physical theories unless we know how to calculate in them, how to use them to model things. By all means circle back to the conceptual questions that really drive you, but don't suppose that you can bracket the often quite messy understanding of how the theory works. And another reason why one should think that way, and this is perhaps more advice for philosophers and physicists, is that if you only look at the kind of clean axiomatized beginning of a physics textbook, you don't understand how messy physics is. Physics is much messier and full of approximation schemes

Starting point is 02:02:54 and heuristics and rules of thumb and quite complicated connections between different subfields than it looks from the clean mathematics and it's much more like other sciences in that regard. So I think you want to get an understanding for a physics theory, not just what it looks like in the actions, but how it's used and how it works in practice before you're really in a position to think deep thoughts about it. And if you're in physics, the other advantage of that is that just professionally, foundational work alone is not usually a sensible basis for an academic career if you want to do it on the physics side of the subject.

Starting point is 02:03:37 There's absolutely not to say that you shouldn't have interest there or follow things there, but it is to say that probably you want that to be one aspect of the research you're doing as a junior physicist, even more so mathematician and not the controlling aspect. But I think that's not terrible that one has that into play. There'd be good things to come out of it, but just as a career point of view, the way physics works relies on you at least having some substantial presence research-wise in spaces that are not just foundational. So in other words, shut up and calculate can dissuade people from foundational questions, but the calculations are important. You don't need to shut up. Yeah, the shutting up is optional. The calculating bit does kind of matter. And you might, if

Starting point is 02:04:31 you're learning the subject, you might want to at least kind of go quiet for a bit while doing the calculating. Okay. Don't get put off from trying to understand how a theory is used by saying, but that doesn't make sense. Okay, it doesn't make sense, but it's empirically successful. So in the end, it's your job to explain how it's empirically successful if it doesn't make sense.

Starting point is 02:04:53 And that doesn't mean at all you shouldn't go back to the question about how does it work, how do we make sense of it. But you won't really know what is the thing you have to explain until you understand some substantial amount of how the theory is used. So don't get put off from understanding that material. Speaking of material, you have reading material on your website and I believe it was broken down into different sections like GR and quantum theory or quantum field theory. Yeah, I have a couple of lists for what quantum field theory, quantum theory, space-time theory, and statistical mechanics. Some of them are a little out of date, but I think they've still got the cool things in them.

Starting point is 02:05:29 A while ago I did a back-of-the-envelope calculation, and I don't recall it because it was a while ago, but I calculated how many hours would it take to go through each of these and it was something like 4,000 and it may be 4,000 per topic or it may have been 4,000 in total. And then they had me wondering, well, did you actually go through all of these yourself and then maybe my calculation was off because there's no way, plus the academic administrative work and just walking and thinking and talking and doing other miscellaneous tasks that aren't just sitting down. I think I've read everything on those lists, at least the papers. If I've referenced a book, then I haven't necessarily read the whole book. Some I think I've read everything on those lists, at least the papers.

Starting point is 02:06:05 If I've referenced a book, then I haven't necessarily read the whole book. Some of them I've read quite fast. And I read really quickly, but also, as I'm sure you know, there's a level of reading you might have of paper where you need to go through every detail to understand it in rich depth. And there's an understanding where you basically kind of skip read through to get the general sense of what's going on and understand it so that you kind of got a lock in your memory so you

Starting point is 02:06:28 can go back to it, you need no more. So yeah, I wouldn't, I think I've read everything or nearly everything I put on those lists, but I wouldn't say I've read everything in exhaustive detail. Also, I mean, look, there's a rough guess. Those lists probably have, I don't know, a hundred-ish papers on each. So that's 400 papers. So no, I haven't spent 10 hours on each of those papers, but I'll have spent 30 minutes or an hour on each of those papers. Okay. Now, last question.

Starting point is 02:07:00 Okay. What's a favorite paper of yours that you've read in recent memory? A paper that sticks with you and you've read it in the past year or so. You're like, hmm, that's interesting. You had to edit out the pause. I have to think about that. No, it's good. I've read a lot but it all blurs together. It could even be one that you come back to that you reread like bells. Yeah, that's probably right. So I mean, I've come back to this several times, but let's go with Joe Polchinski's paper on effective field theory, which is not one of his sort of classic research papers. He has a nice sort of, I think these election notes actually

Starting point is 02:07:41 from the 90s. And among among other things he has this lovely place where he's talking about effective field theory. This links to what we were talking about about fundamentality and he explains about effective field theory and the ways in which the theory screens out the microscopic and how it works and things and he has this little Q&A part way through the paper with like a sort of fictitious student and one of the questions is something like, question, isn't effective field theory really disappointing? Doesn't it just, doesn't it tell us it's going to be extremely hard to get at the real true fundamental physics of what's going on? And Polchinski's answer in the Q&A is, nobody ever promised you a rose garden.

Starting point is 02:08:23 Sometimes that's just the way the world works. These analogies like quantum mechanics as to classical mechanics, like statistical mechanics as to thermodynamics, are they more than just analogies? Is there something that's similar? I think that's a bad analogy. I just don't think it's true.

Starting point is 02:08:42 I mean, it's not that I can't see what fragment of something it's getting at, but in both cases I think it's misleading. I mean, firstly, I think the way in which thermodynamics is a, this is going to be a long answer again, thermodynamics is a control theory. It's not really a dynamical theory at all. The relation between thermodynamics and statistical mechanics is thermodynamics tells us something about what kind of transformations can be brought about. Classical mechanics is just a dynamical theory. Again, it's in the path of standard physics paradigm. It tells how systems evolve

Starting point is 02:09:20 over time if left to themselves. Thermodynamics says if systems are left to themselves they just stay where they are because it's a theory of equilibrium. So that's part of the worry. But also what we have in quantum and classical mechanics is a whole bunch of inter-level relations that tell us something about how a theory with more coarse grain degrees of freedom relates to ones coarse-grained degrees of freedom relates to ones with finer grains of degrees of freedom. Some of those relations are a cross between quantum and classical, some of them within quantum, some of them within classical. So yeah, I mean, I think there's not, as a slogan, there's some insight in

Starting point is 02:09:57 that, but I think there's a lot that's confusing about it as well. One more thing I'd say about the analogy is that it's kind of, the point of analogy of that kind is somehow to dispel mystery. So the implication is, well, we all understand how statistical mechanics is related to thermodynamics. So we can use that. That's interesting. We understand how classical mechanics is related to quantum mechanics. But we're really confused about thermodynamics and statistical mechanics are related. And different people say different stuff about them. And lots of stuff in textbooks is wrong and lots of students say statistical mechanics and thermodynamics are the things that confuse the most. So it's not exactly a nice clean stable thing we can

Starting point is 02:10:33 use to make sense of the classical quantum transition. Professor, it's been a blast. Thank you so much for spending over two hours with me. Not at all. It's been a pleasure. I've received several messages, emails, and comments from professors saying that they recommend theories of everything to their students and that's fantastic. If you're a professor or a lecturer and there's a particular standout episode that your students can benefit from, please do share and as always feel free to contact me. New update! Start at a sub stack. Writings on there are currently about language and ill-defined concepts as well as some other mathematical details.

Starting point is 02:11:09 Much more being written there. This is content that isn't anywhere else. It's not on Theories of Everything, it's not on Patreon. Also, full transcripts will be placed there at some point in the future. Several people ask me, hey Kurt, you've spoken to so many people in the fields of theoretical physics, philosophy, and consciousness. What are your thoughts? While I remain impartial in interviews, this substack is a way to peer into my present deliberations on these topics. Also, thank you to our partner, The Economist. Firstly, thank you for watching, thank you for listening. If you haven't subscribed or clicked that like button, now is the time to do so. Why?

Starting point is 02:11:51 Because each subscribe, each like helps YouTube push this content to more people like yourself, plus it helps out Kurt directly, aka me. I also found out last year that external links count plenty toward the algorithm, which means that whenever you share on Twitter, say on Facebook or even on Reddit, etc., it shows YouTube, hey, people are talking about this content outside of YouTube, which in turn greatly aids the distribution on YouTube. Thirdly, you should know this podcast is on iTunes, it's on Spotify, it's on all of the audio platforms. All you have to do is type

Starting point is 02:12:25 in theories of everything and you'll find it. Personally, I gain from rewatching lectures and podcasts. I also read in the comments that hey, toll listeners also gain from replaying. So how about instead you re-listen on those platforms like iTunes, Spotify, Google Podcasts, whichever podcast catcher you use. And finally, if you'd like to support more conversations like this, more content like this, then do consider visiting patreon.com slash KurtJayMungle and donating with whatever you like. There's also PayPal, there's also crypto, there's also just joining on YouTube. Again, keep in mind it's support from the sponsors and you that allow me to work on toe full-time. You also get early access to ad free episodes,

Starting point is 02:13:05 whether it's audio or video, it's audio in the case of Patreon, video in the case of YouTube. For instance, this episode that you're listening to right now was released a few days earlier. Every dollar helps far more than you think. Either way, your viewership is generosity enough. Thank you so much.

Your Ad Here

Theories of Everything with Curt Jaimungal - The Many Worlds Theory of Quantum Mechanics | David Wallace

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.