Are We Living In A Simulation? - Sabine Hossenfelder

1. 0:00 – 2:53

   Intro

   1. SHSabine Hossenfelder0:00

      We know this, climate scientists know this. Uh, weather people know it. It's a problem. And so people are talking about the simulation hypothesis kind of would totally ignore the problem that we have with putting our reality on a computer. (wind blows)

   2. CWChris Williamson0:14

      You've got a quote at the start of your new book that says, "It is far better to grasp the universe as it really is than to persist in delusion, however satisfying and reassuring." It's from Carl Sagan. What's that mean to you?

   3. SHSabine Hossenfelder0:27

      I, when I came across this, I, uh, I just thought this captures exactly what I'm trying to i- express with the book. Um, so i- it's very tempting to fall for some pleasant explanation, uh, and try not to look at the evidence. But I think in the end, it's better to actually look at the evidence, at least for me. Uh, so, so that's been my conclusion. Because otherwise, you always have this feeling that you're lying to yourself.

   4. CWChris Williamson0:55

      Mm. Do you feel like your Twitter account is mostly you dashing the dreams of people who have, uh, non-evidence-based ideas about what might be happening in the world or in the world of physics?

   5. SHSabine Hossenfelder1:08

      Well, except for the animal pictures (laughs) basically. Um, but y- yeah. Well, um, you know, partly the reason that I was writing this book is that I felt, um, what I do on social media is a little bit too destructive, a little bit depressing almost. Um, if there are some headline that claims, "Well, soon we'll be able to send information faster than the speed of light with a quantum internet," or, "Physicists have, uh, created negative mass," or that kind of stuff, then I'm the one who has to say, "Well, actually, no, you can't do this and it doesn't work. And no, we have not made contact with parallel universes," that kind of stuff. And I, I do think it's important, but it's a very one-sided picture of physics. It, it raises the impression that physics just tells you what you can't do (laughs) , what, what, what isn't possible. And I think physics has another side where it opens new possibilities. It tells you what you can do. Um, it, it brings up, um, new ideas that you might not have thought about otherwise.

   6. CWChris Williamson2:13

      Mm. So rather than just playing defense all the time, there's the opportunity for you to put forward some ideas?

   7. SHSabine Hossenfelder2:19

      Well, not, not my own ideas. I'm talking about, uh, stuff like the multiverse or, um, I mean, o- other ideas that have already, I would say, an almost established place, uh, in the public mind is ideas like that time slows down near black holes or wormholes or maybe teleportation. Uh, those are all ideas that came out of physics and I think it's, it's really inspirational and I understand that people like it and they want to hear more about it and, and so I thought, "Well, I'll write a book about all those big ideas." (laughs)

2. 2:53 – 11:15

   The Problem with the Simulation Hypothesis

   1. SHSabine Hossenfelder2:53

   2. CWChris Williamson2:53

      Speaking of ideas that have captured the public consciousness, the simulation hypothesis is something that's been thrown around an awful lot. What's your problem with the simulation hypothesis?

   3. SHSabine Hossenfelder3:03

      So the simulation hypothesis is the idea that our reality is just a computer program basically, and, uh, once you buy into this, uh, there needs to be some kind of programmer who's written the code and, um, w- God knows what this programmer is doing. Maybe he is God. Um, and so I think that as, um, you know, a talking point over a glass of wine, uh, that's well and fine, or maybe philosophers want to go on about it, that's also fine. But if you claim that it's actually based on science, that's when I get a problem because it's a pretty big claim about what it takes to reproduce our observations, and, uh, it, it's a claim that says, well, you can reproduce everything that we observe, um, with, um, an algorithm put on a computer, and I want to see the algorithm. I'm like, "Show it to me," because basically, um, you just claimed that you have a theory of everything and that falls into the terror of, uh, the foundations of physics, which is where I work. And so I think philosophers who, um, write about it and make statements like, "Oh, well, um, you don't have to think about, um, the computational capacity of the computer because if there's some corner of the universe where no one's looking at the moment, you don't have to compute it," I'm like, "Yeah, well, (laughs) you know, show me how you want to write this into the code." So I'm, I'm just not buying it. I think the- those people are, um, they're trying to get away too cheaply.

   4. CWChris Williamson4:38

      Because it's a problem that's put forward as a philosophy problem, right? It's like a- an idea playing with interesting ways that the universe perhaps could exist. But I think your argument is that you need to drag that across into physics. It's not really a question of philosophy so much as it is a question of physics.

   5. SHSabine Hossenfelder4:56

      Well, it was put forward by a philosopher, and it has certainly been discussed by philosophers. Like, uh, I think David, uh, what's his name, Chalmers has just written a book about it, um, which I haven't really read. I've looked at it, but I (laughs) I will not, um, pretend that I actually read the whole thing. Uh, and it, it's, it's fine to discuss, you know, on this abstract basis, maybe one day there'll be someone who'll be able to write a code on some kind of computer, you know, on that level. But when you... when you're making a claim that it is actually possible for what we currently know about the laws of nature and what we know about computers, then I think that's pushing it too far. And I think that some people have made claims about it, like, like Elon Musk, for example, and, um, uh, Neil deGrasse Tyson is, uh, one of them who- who's made statements about it. I, I think they, they overemphasize how much of it is actually based on science, and, and that's the point where I get a problem.

   6. CWChris Williamson5:50

      Which bit is not based in science? The fact that we don't have the computing power or the algorithms to be able to do this?

   7. SHSabine Hossenfelder5:57

      Well, we don't have the algorithms. Uh, we don't even know how to do it. Like, I mean, there, there are really basic problems with, um, how do you describe a chaotic system? Uh, thi- this is something, uh, just to stick with this example, that climate physicists, uh, have to cope with, uh, in reality and it... and it's really difficult, because, uh, as you probably know, um, you have something like the Navier–Stokes equation, it's scale-invariant. So strictly speaking, you can't just ch- chop it off at the finite grid. Um, you have to-

   8. CWChris Williamson6:27

      What is that? Go, go, go into that.

   9. SHSabine Hossenfelder6:29

      (laughs) Well, um, what, the Navier–Stokes equation or finite grid?

   10. CWChris Williamson6:32

       Yeah, I'm not familiar. I'm not familiar with either of those.

   11. SHSabine Hossenfelder6:34

       Oh. Oh, the, the Navie- the Navier–Stokes equation is the equation that describes how, um, fluids and gases behave. Uh, for example, the atmosphere, but also oceans water, uh, and that kind of stuff. Um, so, um, it's like the central equation that you have to solve on a computer if you want to describe the weather or the climate, uh, in the long run. And, uh, this equation has the property that it's scale-invariant, which, uh, basically means it, um, it, it draws on all different scales, like the long ones and the very short ones. Uh, but if you want to calculate what happens for the globe, you can't actually, on a computer, you can't really calculate it on all scales. So, um, what you have to do is you put a grid on it, um, and, um, the resolution of the grid depends on, uh, whether you want to do a calculation on a country basis or on, on, on the entire globe and for what time span. But typically, for climate models, it's right now, I think something on, on the order of, uh, several tenths of kilometers and maybe for weather models, they can go down to one or a few kilometers. So I'm, I'm not a climate scientist-

   12. CWChris Williamson7:46

       Mm-hmm.

   13. SHSabine Hossenfelder7:46

       ... so (laughs) I hope this is roughly right. It should be, kind of, in, in that order of magnitude. Uh, but th- that's clearly not scale-invariant, um, so you, you modify the equations, uh, and that has consequences. So there, there are... there are just situations which those models will not correctly predict and we know this, climate scientists know this, uh, weather people know it, it's a problem. And so people are talking about the simulation hypothesis, kind of, we totally ignore the problem that we have with putting our reality on a computer.

   14. CWChris Williamson8:18

       And that's only at the scale of a world or perhaps even a country with the grid size or the pixel, kind of, the pixel equivalent of one square kilometer perhaps. So trying to make something that's Planck length across an entire universe is signifi- Okay, I understand. So if we can't simulate a universe, can we create one?

   15. SHSabine Hossenfelder8:37

       (laughs) Well, that's, that's an interesting question. Um, I, I made a video about this and I was trying to argue that, um, if everything we know about, um, the early stage of our universe is correct, that's a pretty big if, but let's just assume (laughs) it is correct, then the answer is yes, quite possibly, one day we'll be able to make a universe. And (laughs) a lot of people thought I was joking because they know me as someone who, who usually says, "No, you can't do this. (laughs) This is all rubbish. Don't believe it." And this was another reason why I wanted to write a book, because I felt that I've maneuvered myself into a corner where I, I'm taking this very, this very negative destructive point of view.

   16. CWChris Williamson9:22

       Mm-hmm. How is it that it could be possible for us to create a universe but not simulate one?

   17. SHSabine Hossenfelder9:28

       Well, um, (laughs) well, that's-

   18. CWChris Williamson9:30

       Or more possible, should I say?

   19. SHSabine Hossenfelder9:31

       ... a very good point. Uh, yeah. Well, it, it's, it's... it depends on the question what you mean by, um, a simulation, uh, and what people normally talk about, they talk about a computer that has been programmed by someone, and that something which you would not be able to do with this universe. Um, you would just create the conditions under which it comes into existence, uh, and then what, uh, at least for, for we currently know about how those things work, it would create a small bubble basically that would pinch off from our universe, so it goes away. You have... You, you, you don't control it. Um, you don't program it. Um, it probably inherits most of the conditions from our universe and, um, you can never get into contact with it again. (laughs)

   20. CWChris Williamson10:15

       Would you ever actually be able to know if you'd created one in that case?

   21. SHSabine Hossenfelder10:20

       Uh, well, that's a good question. Um, y- you have to think about some observers, observables that would be, uh, related to it. Um, people have thought about some. Um, unfortunately, it would look pretty much like a tiny black hole, so it might be very difficult to tell apart. Uh, but, um, I should probably add, like, this is not something we're going to do in the next 1,000 or maybe 10,000, uh, years. Um, it, it would require to put, um, a macroscopic size of material, so estimates say something like 10 k- kilograms into a particular quantum state where th- uh, that kind of stuff, uh, could happen, yet. Um, so it's theoretically possible. There's nothing in principle that stands in the way of one day doing it, but it's certainly not something that's, that's going to happen in our lifetime.

   22. CWChris Williamson11:11

       Okay. No new universes just yet. So what about free will? That's something

3. 11:15 – 18:13

   How Physics Impacts Free Will

   1. CWChris Williamson11:15

      else that's been popularized, people like Dan Dennett and Sam Harris over the last few years. I actually had a friend who, uh, I, spiraled into a three-week depression by sending him a video of Sam Harris explaining free will on Joe Rogan, and, uh, after that, he came out on the, the other side actually being quite thankful for it. But for those three weeks, I'm sorry, Luke. Um, what about-

   2. SHSabine Hossenfelder11:34

      (laughs) .

   3. CWChris Williamson11:34

      ... physics' relationship to free will, what, what have you come to believe about that?

   4. SHSabine Hossenfelder11:39

      Yeah. I've, I've, I've been there (laughs) but it took me much longer than three weeks. Um, so it, it's one of the reasons why, uh, my book has kind of a warning, uh, in the beginning where I'm like, "Some of the things that we're going to talk about, uh, are not easy." Uh, and I think that it's one of those points. Uh, but I also think, uh, that, um, you know, everyone who knows something about physics, um...... will stumble over this problem sooner or later, so let's just talk about it and then you can get through it and come out on the other side. And, and I hope that my book facilitates that. Um, and so, y- yeah, I'm, uh, basically, I'm, I'm, uh, I'm with, uh, Sam, uh, I, I like his book. Uh, I mean, to begin with, it's a very short book, (laughs) so I can really recommend it. You'll get through it very quickly. I like the book except for the one paragraph that he has about physics, (laughs) where I'm like, "Um, I don't think... This, this wasn't quite right." It was something about the Planck scale, which, uh, the book would have been better without that paragraph. Uh, but yeah, I mean, so, so there's this basic p- problem that, um, for all we know about the, the fundamental laws of nature, they are a combination of, uh, determinism, where the future is predicted by what happened in the past, and then every once in a while, there are some, uh, random event that comes from quantum mechanics, but you can't influence it, um, because nothing influences it. It's entirely random. And, um, now, whether you think that this rules out free will or doesn't depends on what you mean by free will. So, this is why there is so much debate about it. So, the way that I've tried to approach the problem in my, in my book, uh, is by just first explaining what we know about, um, the laws of nature, which is what I just said, this combination of determinism with the, uh, occasional random quantum jump, and then you can ask, "Well, does this mean that free will does not exist? W- what do you make out of this?" Now personally, my conclusion from this is just, "Well, if free will doesn't exist, let's get over with it, um, and, you know, think about ourselves in a different way." Um, but of course, you can try to come up with different definitions for free will, and this is something, uh, that David Alberts, uh, has done, for example, and other people have put forward slightly different, um, definitions. And this is all fine with me, so long as we know that we're talking about different things.

   5. CWChris Williamson13:58

      Mm. What is it th- that's happening at the quantum level that's causing this random chance? Because as far as I was aware, it means that had you have gone back and run the same period of time again, that something different could have happened. But I, I thought that given the initial conditions, everything is predetermined from there, but it seems like you're saying there is a genuine roll of the dice here, which could one time c- come up with a one, and the next time, come up with a three of some kind.

   6. SHSabine Hossenfelder14:26

      If you take quantum mechanics seriously, the way that we use it right now, um, that's exactly what happens. Um, of course, you can't actually go back in time and run it again, (laughs) so, so there's a slight problem with that. Uh, but at least the way that the theory works, um, there are occasionally those measurement events, uh, for which you can't predict the outcome. You can only predict the probability of getting a particular outcome. Uh, and, uh, th- those are generally random. They're, they're indeterministic, so quantum mechanics is an indeterministic theory. But of course, um, quantum mechanics may fundamentally not be correct. It may just be an approximation to something else that's going on, and that something else might be deterministic again. Um, and then as you certainly know, there's also the many-worlds interpretation, which tries to do away with, uh, the, the measurement update, uh, entirely, so then you only have the deterministic evolution left. So, it's, it's somewhat controversial. But, um, y- yeah, I mean, I, I think for what the discussion about free will is concerned, it doesn't, it doesn't really matter, you know? You, you either have a deterministic evolution with the occasional random event, or you have a deterministic evolution without the occasional random event, but in both cases, I find it difficult to make sense of free will.

   7. CWChris Williamson15:44

      Yes, because... Am I right in thinking the many worlds is where you branch off each time that something occurs? Is that right?

   8. SHSabine Hossenfelder15:50

      Yeah, basically. So when... In, in the standard interpretation, which is often called the Copenhagen interpretation, um, y- you make a measurement. Um, you collapse the wave function in one, into one definite outcome, but you don't know exactly what, so th- which outcome? So, so that's, that's the indeterminism. Uh, in the many-worlds interpretation, you don't collapse the wave function. You just say, "Well, now, there are, there are two universes, or three, or four, or five, uh, depending on how many possible outcomes, uh, there would have been," um, and each of them happens in its own universe.

   9. CWChris Williamson16:25

      Yeah. So, what that would mean is that each universe is deterministic, because you are following down the path of one. If you continue to go down that, that would mean that that is the only way that that could have come out, and all of the others are also deterministic. Is that right?

   10. SHSabine Hossenfelder16:40

       Um, w- no. Actually, what's deterministic is the entirety of all universes.

   11. CWChris Williamson16:46

       Right, understood.

   12. SHSabine Hossenfelder16:46

       But if you, if you are in one of those universes, it'll still look in- indeterministic, and this is exactly what we see, right? It looks indeterministic.

   13. CWChris Williamson16:54

       Ah, I see. So, is it right... Y- you often hear that if we had known the initial conditions of all of the matter, uh, at the beginning of the Big Bang, we could have accurately predicted everything that was going to occur for the rest of time going forward. Is that not true?

   14. SHSabine Hossenfelder17:12

       W- well, in the many-worlds interpretation, it would be true for the collection of all those parallel worlds. Um, in our universe, um, if you collapse the wave function, it's not true because you get this, uh, you get this random, um, quantum element. And it's not necessarily the case that, um, those quantum, um, events remain small. They can certainly grow into macroscopic differences, and this is what, uh, Schrödinger tried to illustrate with his famous, uh, cat thought experiment. Uh, so he was trying to say, "Look, it could be that, um, you, you have a, a cat that's either dead or alive depending on whether or not an atom decays," and so what it basically does is that it amplifies this, uh, indeterministic outcome from a tiny macroscopic quantum thing, which is this atomic, uh, nucleus, uh, to a big thing, which, which is a cat, but it could be you, or it could be the entire planet, you know, if you, I don't know, you blow up a bomb or something like that.

4. 18:13 – 27:09

   Misunderstanding the Universe’s Origins

   1. CWChris Williamson18:13

      Going back to the beginning of the universe then, what are people getting wrong with the way that they currently look at the Big Bang, the origins of the universe? What are the claims that you've been, uh, beaten over the head on Twitter or elsewhere?

   2. SHSabine Hossenfelder18:29

      (laughs) Yeah. So, so the problem is that, um, we have a fairly well-established theory of how the universe evolves. And yeah, there are some niggly bits with dark matter and dark energy, uh, but, uh, let's leave this aside for the moment. So, we have Einstein's Theory of General Relativity that tells us how the universe as a whole, um, changes in time if we know the matter content, matter and energy content. So, we can use this to extrapolate the present state back in time by using those equations. And, uh, so what happens is that at some point, those equations just break down and, uh, we end up with a state at which the energy density, and also the curvature of the universe, was infinitely large. So, this is what's called the Big Bang. And, uh, now the problem is that, um, m- most physicists, me included, think, um, that this is probably not what actually happened. It just means that, uh, those equations break down and we would have to use a better theory, so that would be a quantum theory of gravity, but we don't have it. So, um, how did the universe begin? Well, we don't know because we don't have this theory. Now of course, a lot of physicists are unhappy with this state of affairs, and, uh, what they try to do is that they kind of modify the equations at an early time, uh, and then they attach a different story to the beginning of the universe. So instead of a Big Bang, you might instead have a big bounce. Like, this is something which is popular in, in certain cycles, so you have a previous universe which collapses and then it starts to expand again, so there's this bounce, uh, in the middle, and those bounces could repeat and then you get a cyclic universe. But it doesn't have to be the case; in some, in some scenarios it's just single bounce. Um, but it could be other things, you know? Um, other people have claimed that we came out of a black hole, and it could've been a higher dimensional black hole so we could have five dimensions, or it could've been some kind of collision between higher dimensional membranes, or something with a gas of strings, or some people say maybe it didn't have any geometry but it was just some kind of network. So, there are all kinds of stories. And the problem I have with those stories is that they make a simple story more complicated, which isn't something that a scientist should do. So, I think the honest answer, uh, we can give as physicists to the question, "How did the universe begin?" is, "We don't know."

   3. CWChris Williamson20:55

      What's-

   4. SHSabine Hossenfelder20:56

      And I know this is unsatisfactory-

   5. CWChris Williamson20:58

      (laughs)

   6. SHSabine Hossenfelder20:58

      ... but I think that that's, uh, that's how it is.

   7. CWChris Williamson21:01

      What is the fundamental problem that physicists are bouncing up against, or coming up against (laughs) , bounce might be the wrong word, with all of these, uh, extravagant stories that they're trying to tell about what might have happened before the Big Bang, or to create the Big Bang? What is the fundamental problem? Is it there's something from nothing?

   8. SHSabine Hossenfelder21:24

      Um, that's actually an even more difficult problem. Uh, no, it's got something to do with the, with the type of theory that we currently use, um, which is this combination of, um, a description of the state of a system at one moment in time. It's normally called an initial state, but somewhat confusingly could also be at the final time. Um, so it's, this is just, um, uh, one specification of the state at one moment in time, and then we have some equation by which we can tell from this one initial state what happens at any other time. Uh, and so the way that it works for the universe is that we take an initial state, uh, in, in the early phase of the universe; it can't be exactly at the Big Bang because as I said, this is a singularity, so it doesn't work, but it could be something after this, and then you apply, um, your evolution law. So, in the simplest case th- that's just Einstein's equations. Uh, and then you can calculate how the universe should look like today and you can compare this, uh, two observations, and if it works you say, "Bingo. Okay, good theory." Uh, and then now, if you look at, uh, at all those other explanations, basically what you do is you, you attach a more complicated story before this. And, um, this is something which you can always do, um, and it's allowed by our theories, uh, but it makes the theory completely ambiguous because there are many different ways that you can do it, and it's some- something that the scientific method, um, actually doesn't allow, um, which is why I think it's, it's a problem that we can't really resolve with the theories that we currently have. It, it's, it's possible that at some point we'll come up with a different type of theory, um, that might be able to overcome this limitation, but at least for now I, I think we're stuck with it.

   9. CWChris Williamson23:13

      It seems to me like there's a lot of problems, or, uh, physic- uh, theoretical physics is kind of bouncing off the limit of a bunch of the current theories that we have. I had Michio Kaku on the show, I was speaking to him, and it seemed like his explanations, there's always caveats here and there. There's always something that's like, "Oh, we don't quite know about this but it's sort of a best guess, and we've got this but we need to add 11 dimensions in in order to make it work," and there's a lot of assumptions. Like, I'm, I, you know, I appreciate I'm bro-science-ing my way through this, right, but it, it seems like the, there has... I- it seems unsubstantial or insubstantial at the moment what we have to describe the universe, and that this is causing people to retrofit stories around what's happening in order to be able to make the theories work.

   10. SHSabine Hossenfelder23:59

       Yes, right. And all of this is entirely unnecessary to actually explain what we observe. Um, so, so that's the problem with it. So you, so you get those, uh, multiple stories, um, that can all be made to fit to the current observations. And then, if you turn, if you turn it around, um, it has the consequence that you can't use observations to tell those stories apart.

   11. CWChris Williamson24:24

       Interesting one. Talking about how the universe began, what about how it's going to end? Do we... Big crunch, big freeze? What's the other one? Big heat? Is there a heat death? Or-

   12. SHSabine Hossenfelder24:36

       Big, big, big grip. Yeah.

   13. CWChris Williamson24:37

       Big grip, that was it.

   14. SHSabine Hossenfelder24:38

       The heat death is also one of them, yeah.

   15. CWChris Williamson24:40

       Okay. Well, how are we gonna end? How's it, how's it all gonna finish?

   16. SHSabine Hossenfelder24:42

       Yeah. I'm, I'm afraid the, the answer is, again, we don't know. Um, in, in this case, it's... I think it's easier to understand. Uh, so if we're trying to say how the universe is going to end, we have to extrapolate the current state of the universe, uh, into the future, possibly over trillions of years. Um, and, uh, the problem is that there could be some physical processes that are just very rare. Uh, and... or they are so small that we haven't been able to measure them.

   17. CWChris Williamson25:12

       What like?

   18. SHSabine Hossenfelder25:12

       Um, well, one example could be that the cosmological constant is not actually constant, but it might very, very slightly been changing. And now over, over trillions of years or something like this, this could be the decisive factor for the fate of the universe, and it's just something that we, we can never rule out. So basically, what happens is that if we extrapolate the current state into the future, the, the uncertainty just blows up and in the end, you can't say anything. So i- it's kind of an interesting mind game, I guess, uh, to try to figure out... Assuming that nothing else happens, like the cosmological constant is actually constant and there are no other processes that we haven't yet heard of, uh, and so on and so forth, you can try to speculate what's going to happen. But I would say, uh, don't take it too seriously.

   19. CWChris Williamson26:01

       If nothing was to change, if there's no, uh, spooky alterations hiding in the future, what do you think would be most likely to expect in that case?

   20. SHSabine Hossenfelder26:10

       Um, I think if I, if I remember correctly, um, at least in the current standard model of cosmology, i- it would be the heat death. So it's because the cosmological constant, uh, speeds up, um, the expansion of the universe. So th- um, galaxies become more and more distant to each other, and then the stars gradually die, and they collapse to black holes. Everything will, everything will be dark, and then the black holes evaporate, and then you have this leftover gas of elementary particles and, and that's pretty much it. But, I mean, if you have another theory, like, uh, Roger Penrose, for example, has this idea of a cyclic universe, then the whole thing eventually f- transitions into a new kind of big bang. Uh, so I know this is making this a little bit vague, but, I mean, he has some mathematics to show for... And I think it's not entirely crazy. I, I actually have quite some sympathy for it, but I'm, I'm not really sold on it, I guess (laughs) .

5. 27:09 – 32:00

   Is There Something Better than Mathematics?

   1. CWChris Williamson27:09

      Would you say, speaking of mathematics and Roger Penrose's work, is, is mathematics like the ultimate language or the basis of reality? Is that what everything is built on?

   2. SHSabine Hossenfelder27:19

      (laughs) Uh, yeah, that's a very interesting question. Believe it or not, I've actually thought about this for (laughs) , for quite some bit. Uh, well, it's certainly the case that currently it's the best thing that we have. And I think it'll continue to be the best thing, uh, for quite some time. We haven't fully exploited its, its potential, especially when it comes to a chaotic or complex systems that we just, uh, talked about. We've barely begun to understand how the mathematics works. Um, but how, uh, how are we to tell if that's the best thing ultimately? Like, we, we've barely just begun to understand nature and to try to formulate our, uh, hypotheses about it, uh, in, in forms of mathematics. Maybe in 100,000 years somebody will come up with something better than mathematics.

   3. CWChris Williamson28:07

      What else, uh, what else could there be? I don't understand what could be better than mathematics.

   4. SHSabine Hossenfelder28:12

      (laughs) Well, yeah, I guess that's the problem, right? (laughs) May- maybe we're just not smart enough, um. So, um, I, I certainly hope that, um, uh, given time, uh, there will be a more, a more sophisticated, uh, species on this planet. And, uh, God knows what they would come up with. But one, one thing that I've played with, um, is that it, it is in, in principle possible to do science without using mathematics as an intermediary, and we actually do this when we do computer simulations to some extent. So the way that we currently do computer simulations is that we, um, we take the mathematics that, um, we have extracted from, from observations and then we formulate it in terms of an algorithm, and we put that on a computer. So we use mathematics as this middleman, um, to get the simulation. But strictly speaking, you don't actually need this. You could try to find another system that, um, mimics the thing you want to describe directly. And actually, um, this is being done, uh, in quantum simulations. So there... well, for one thing, there are certain types of quantum computers, uh, that are based on this idea. Uh, but more generally, you can, you can simulate certain... the properties of certain fundamental particles. For example, the Higgs is one example, uh, that people have looked at, but also things like, uh, Majorana par- particles and so on. You can simulate them in condensed metal systems. And, um, now, a- again, this is using mathematics as an intermediary, yes, but strictly speaking, you don't, you don't need it. You could just say, "Well, I take the simulation in and by itself, and I just try to figure out what it tells me about this other system," so, so you can map reality directly to reality. And, uh, I think that that's kind of a different way to do science that we haven't explored further enough, and maybe that's, that's a way that we will be able to go beyond mathematics at some point. But this is, like, really, really far out there (laughs) wildest speculation.

   5. CWChris Williamson30:28

      Is it true that other alien civilizations would have had to have discovered mathematics as well? I, uh, I often hear about mathematics being a universal language of the universe and stuff like that. Is, is there any legs to that?

   6. SHSabine Hossenfelder30:41

      It seems very plausible to me, but, uh, how are we to know? Right? We, uh, we haven't spoken to any alien species. Um, so, uh, I mean, it's certainly the, it's certainly the case that mathematics is kind of a universal description of certain regularities. Um, and, and so it, it seems plausible to me that this probably would have been the case. But, um, I guess we'll have to wait until we meet those aliens to find out.

   7. CWChris Williamson31:06

      I suppose if you were to think about any alien sat on its little green planet, and it was to look up at the sky at night, and it would say, "There is a point of light in the sky. There is another point of light in the sky. There is another point of light in the sky." Like, I guess inferred from that is, look, th- that is the fundamental basis of counting, right? There's, you've got some mathematics in front of you, and I guess everything else from there all the way up to E equals MC squared is just more of-

   8. SHSabine Hossenfelder31:34

      (laughs)

   9. CWChris Williamson31:34

      ... counting, counting stuff and manipulating it.

   10. SHSabine Hossenfelder31:37

       Yeah. That, that's how the argument, uh, normally goes. But, but again, I mean, this is kind of very strongly based on, on our experience of the world. Um, so how are we to tell that not aliens, uh, would see reality completely differently?

   11. CWChris Williamson31:53

       Mm. (clicks tongue) You mentioned earlier on about the cosmological constant, which is part of this or

6. 32:00 – 37:45

   The Fine-Tuned Theory

   1. CWChris Williamson32:00

      i- is used to justify the fine-tuned theory of the universe, the fact that you have this unbelievable sort of knife edge that a bunch of different, uh, characteristics of this universe, gravity; strong, weak nuclear force; cosmological constant. Am I right in thinking that if any one of these was even ever so slightly different that basically we probably wouldn't exist, and the likelihood of it occurring seems super, super low, and therefore people say, "Look, this is obvious that the universe has been fine-tuned for life." But that's also super contested. Can you go into the fine-tuning theory for me?

   2. SHSabine Hossenfelder32:34

      Yeah. So this is an argument which has been made for a long time, and it works pretty much the way that you just summarized it. You, you take those constants that we have, uh, in the fundamental laws of nature. Um, I think there are 26 of them. It depends on how you count, you know. A lot of those constants are just zero for all we know, like the mass of the photon. And then you can, you can argue over does it count as a constant if it's zero, um, so, you know, it's, it's split in half. Uh, but, um, yeah, so, so you can ask stuff like if the cosmological constant was a little bit larger, uh, or if it was a little bit smaller, what, what would happen? And, or you can ask about, uh, the fine-structure constant, um, the-

   3. CWChris Williamson33:16

      What's that?

   4. SHSabine Hossenfelder33:17

      Um, the... Fine-structure constant, that's the... it sets the strength of the electromagnetic interaction, alpha. Um, so, so what would happen if we made this a little bit, uh, large or small or the gravitational, uh, constant and so on and so forth? And in, in, in a lot of those cases, the answer is something would go badly wrong. Um, stars wouldn't be able to shine, or, um, the, uh, galaxies, uh, would never be able to form. Everything would collapse to black holes, or the en- the entire universe might collapse, uh, immediately, or it would be impossible to form any kind of complex molecules that we think are, are necessary for life. Uh, and, uh, people have gone through a lot of examples, uh, of that type. And, um, some have taken this as an opportunity to say, "Wow, that looks really unlikely that something like this would have happened by chance. There must have been a creator." Um, uh, and then there are, on the other side, there are physicists who say, "Well, this seems very unlikely to have been the case, therefore there must be a multiverse." Uh, so they're kind of, uh, two sides of the same, uh, coin, so, so to say, um, though they're both, uh, e- exactly opposite, uh, to some extent. But they come out of the same idea, which is that it seems like there's something in need of an explanation, uh, which is why are the constants of nature exactly those. Now, the problem with this entire argument is that we have no way of quantifying the probability of this happening. So if you tell me, "Well, it seems really unlikely," I would ask you, "Well, um, how do you know?" I mean, it's not like you can collect a sample and ask, "How often does it happen?" Because we have only this one set of constants of nature, and we, we have no way of telling how likely or unlikely it would have been. Um, so typically, these arguments work with, um, some kind of statement about what is or isn't a small change in the constants of nature. But this you can also question. Like, how small is small? And, uh, why is this too small, and why not this other thing? And so I think this is all ill-defined, and I'm, I'm totally unconvinced by any of those arguments. There's also the curious fact, um, that, uh, in, in the past couple of years, you, you wouldn't believe it, but every once in a while, there's actually progress on these matters. Um, some physicists have come up with possible combinations of the constants of nature that are very different from the ones that we have in our universe but that still seem to allow for complex chemistry, uh, to happen. So, physicists don't really talk about life, but it, it seems quite plausible that the, these other combinations of the constants of nature would also allow for, um, complex chemistry and possibly even life. So, um, you know, I think this, this argument is really just wrong, and people should stop talking about it.

   5. CWChris Williamson36:08

      (laughs) Well, it's cool to think that there's another combination out there that would allow complex, uh, life to evolve because that... I mean, that would do away with the fine-tune theory overall, right, that you have another version of this that could work. And I, I really like the idea that even though if we were to look... I think it's the cosmological constant that's a, a very small number, or there's a couple of others, like the weak nuclear force is an unbelievably small number compared with some other numbers that are in the bunch of constants.

   6. ... but when you talk about how much you want to change that by, as you said, it- who is to define what would be a large or a small change? Who is to define whether that's a large or a small number? And then when you talk about something being fine-tuned by design, you're talking about something being moved by small amounts. But if you're the person that's already creating before that, a decision on what big and small is, then it's just you making, I guess, like, a- a value judgment or a pre-design to judge- uh, judgment on whether or not you think it should change.

   7. SHSabine Hossenfelder37:08

      Yeah. Uh, that's one way to look at it. Um, normally, um, physicists draw the arbitrarily, say, um, i- if- if a change is of size one, then that's large, uh, and everything that's below, uh... Uh, that- that already becomes very subjective, but then you can say, is- is it smaller than one over 100 or is it smaller than one over 1,000? Uh, that'd be small, but of course you could just have started with, um, as you say, one over 1,000, uh, or something, and then measure everything relative to that. Um, and- and so th- this is exactly where, uh, this argument that physicists put forward becomes c- circular.

7. 37:45 – 45:10

   Boltzmann Brains

   1. SHSabine Hossenfelder37:45

   2. CWChris Williamson37:45

      Mm. What's your view on Boltzmann brains? I learned about these last year, that there's potentially-

   3. SHSabine Hossenfelder37:50

      (laughs) .

   4. CWChris Williamson37:50

      ... other versions of us floating out there in the ether, and that me and you and this entire conversation might just be the brief flickering of some brain a million, trillion, billion miles away from here, uh, and I might just be imagining it all.

   5. SHSabine Hossenfelder38:05

      (laughs) Yeah. So, um, indeed. It might not have been a- a whole version of you, but maybe just your brain which, uh, has the illusion that it's talking to me. Um, yeah. So, that- that's one thing which, um, comes out of statistical mechanics if you take it too seriously. Um, so basically the idea is that if the universe goes on forever, which, um, as we already discussed previously, uh, is likely to happen if you just extrapolate the- the current state into the future, then all kinds of combinations of fundamental particles that are in the universe should happen at some point, provided the laws of nature fulfill a certain property, which is called ergodic, which I'll come back to in a- in a second. And then it can happen that just coincidentally, you know, the, um, elementary particles combine to form some particular molecule. Uh, and it takes a really long time. You know, we're- we're talking trillions and trillions, uh, of years. Um, and then you wait a little bit longer and they combine to form something like, that's like a cell. And if you wait long enough, it would form a brain. And then, in fact, if you just wait long enough, it'll form any possible brain, uh, brains that will think all possible thoughts. Um, but the downside of this is, um, that the larger the thing, um, the shorter it'll persist, because it's surrounded by all this randomness, um, which- which comes from the interaction of the other molecules, which have not spontaneously assembled to, uh, a large thing. So, um, the shorter the thing lives, the more likely it is. And this is what gives rise to this idea of the Boltzmann brains. Mm-hmm.

   6. CWChris Williamson39:53

      So, somewhere at the end of the universe, in the very, very far future, there are all those brains spontaneously assembling its- themselves, just thinking one thought, "Oh, hello, I'm here," or, "Hello, I'm talking to Sabine," and then they- they fall apart again. And I see you laughing, and yeah, it- it seems a little bit ridiculous. Um, but the thing is that if you- if you take this seriously, uh, what we know about, uh, the fundamental laws of nature and about statistical mechanics, then you kind of have to bite into this sour apple. Um, I- I think that's a German idiom, but I- I hope it- I hope it translates. Um, and so a lot of people think this is just silly. Um, but I think it actually tells us something about the laws of nature. Um, because if you want to prevent this from happening, then we can conclude that the laws of nature can't be ergodic. Um, and that's a statement about the properties of those fundamental laws. There are actually some indications, I should add, that, um, gravity or quite possibly the strong nuclear force might not be ergodic. And so I think that this is an intere- interesting factual. Um, so it's not quite as silly as it sounds, as at- (laughs) first. Why would forces change within the same universe just because they're far away?

   7. SHSabine Hossenfelder41:11

      Oh, um, no. The- the forces don't change. So- so the argument is that this, um, um, this statement that all possible combinations of fundamental particles happen at some time, this is only correct for, um, laws of nature that have certain properties. So, um... And- and this is this, uh, ergodicity. Uh, this is basically what it means. So, if the theory's, uh, ergodic, then- then all these possible things will happen, uh, at some point. But this is not the case for all possible interactions that you can think of. In particular, if you have some interactions that are very strongly bound, which the strong nuclear forces, it makes it highly implausible. You know, when you- you- you take, uh, you put some particles together and they get stuck, uh, like in a bound state, like the strong nuclear force, uh, binds together quarks in- in a proton, uh, or something like this, um, w- why should everything that can happen, um, also actually happen at some point? Why can't they just... why can't they just remain stuck together forever? Uh, and then so th- this kind of interaction, uh, runs you into problem like, problems like this. And, uh, to- to be fair, uh, Boltzmann, when he was thinking about this, didn't know anything about the strong nuclear force, so probably it just wasn't a problem that- that didn't occur to him.

   8. CWChris Williamson42:26

      Does that mean that it's wrong to say that me and you would be having this conversation further away? Is that the same, the exact same model that people are using that if the universe outside of the observable universe is infinite, then it means that me and you are having this conversation a million different times in a million different ways?

   9. SHSabine Hossenfelder42:41

      ... well, the way that I would put it is that it's either that or the laws of nature are not ergodic and then we learn something from it.

   10. CWChris Williamson42:48

       Mm. Interesting. What do you mean, you talk about knowledge in humans being predictable, what do you mean there?

   11. SHSabine Hossenfelder42:55

       Um, actually David Deutsch talks about this. So, I, I went and interviewed, um, David Deutsch which was v- very interesting. And he has this argument, uh, that knowledge, uh, can't be predictable because if you could, uh, predict the knowledge, uh, from what we already know then it wouldn't be new knowledge, uh, which makes sense. And I think it's also something, um, that we see happening in, in physics but also in other disciplines of science is that any kind of new theory that w- we develop, um, it, it requires this intuitive leap. It requires something new, you can't just strictly speaking deduce it from what came previously. So, y- what you see a lot is that, uh, in hindsight, um, physicists often make up some kind of story for how they came to this conclusion that was deduced from what they had previously heard and so on and so forth. But it seems to me there's always, like, an, an element of magic (laughs) in between somewhere, you know, where, where there's this insight w- uh, which comes in. And I think this is, this is what, what David, uh, is getting at. I should say though that this is, like, something at the, at the very high emergent level. Um, you could, y- yeah, on, on the level that we are talking about it. Um, y- you could make an argument that fundamentally if you're talking about quarks and gluons and th- the laws that they behave, um, by, it might well have been deterministic, uh, it's just that, uh, for practical purposes, we, we would never have been able to predict it.

   12. CWChris Williamson44:31

       Do you think... Does that mean that knowledge is discovered or created?

   13. SHSabine Hossenfelder44:38

       Well, I guess both, right? Um, so sometimes, some types of knowledge, uh, you discover, um, others you create. I, I, I don't... I know people discuss this a lot when it comes to mathematics, but I have, I've always been kind of a little bit ambiguous about it. I, I don't see that it has to be one or the other. It could well be, it could well be both.

   14. CWChris Williamson45:04

       Got you. When it comes to consciousness as well, obviously we've spoken about free will, spoken about simulation,

8. 45:10 – 53:37

   Can We Compute Consciousness?

   1. CWChris Williamson45:10

      is it possible for us to compute consciousness?

   2. SHSabine Hossenfelder45:15

      Um, that's a very good question. Um, as you probably know, Roger Penrose, um, takes the point of view that it's not possible. Um, he has an argument that it's based on, uh, Gödel's theorem, and I have an interview with him, uh, in my book about it. I'm, I'm not really convinced by this argument, uh, but then, you know, he's a Nobel Prize winner and I'm not, so maybe you should listen to him (laughs) and not to me. And this is why I have the interviews in my book so you don't just hear what I think. Um, so, um, he thinks there's an, an uncomputable element, uh, to consciousness and, um, so the, the laws of nature that we currently use, they are computable, and, uh, I think this is one of the reasons why he thinks, um, that there's something about quantum mechanics, uh, which we don't yet fully understand, and so he has put forward some, uh, alternative, uh, theories to quantum mechanics where, um, he thinks this element of consciousness, uh, comes in. Um, so I think it's interesting. Um, I, I'm not convinced that it, it's quite right, but, uh, then maybe there's something worthwhile, um, to it. So, if he's right, then it would mean that, um, artificial intelligence at least in the, in the way that we can currently formulate it in terms of algorithms run on computers would never be really conscious.

   3. CWChris Williamson46:31

      Yeah, well I think that's... Maybe about 10 years ago... Or when, when did Superintelligence come out? 2014? Nick Bostrom's book, something like that? Uh, when that came out and I read that, I was adamant that the artificial general intelligence apocalypse was going to be with us within 10 or 15 years, uh, and it seems like the AI safety community, the discussion around AGI in general has kind of switched a little bit in that time. It's not like I've got my finger on the pulse of the coolest stuff in the development of it, but just from my sort of perspective, and it seems now like well-defined problems are things that artificial intelligence is getting incredibly good at, and poorly defined problems basically haven't made much progress at all as far as I can see. And it's the poorly defined problems that are precisely where the general from AGI would come from, and I think that it seems like some of the, uh, what did they call him? Bostrodamus. Um, some of the Bostrodamus concerns, the pre-apocalyptic people, it, it seems like that's changed. It seems like the narrative around this sort of discussion has changed precisely because of the fact that we're having more and more difficulty in computing this stuff.

   4. SHSabine Hossenfelder47:46

      Yeah, it's very interesting. Um, they're also increasingly running into resource limits, um, is my understanding. Like, those, uh, running those-

   5. CWChris Williamson47:56

      Like computing power?

   6. SHSabine Hossenfelder47:57

      Yeah, computing power, energy, um, and, and then there's the problem that, you know, all the stuff with the, uh, what, what do you call? The hyper-parameters that have to be found out somehow, and, uh, oh, those, um, neural networks, uh, that are typically used for artificial intelligence, um, they, um, have what's called hyper-parameters that have to be chosen, um, so the network can properly learn. And that's a little bit of black magic, like, where do those hyper-parameters come from? And there, there's a lot of discussion, at least that's my understanding, and l- like you, I don't really work on this, I'm just following it, uh, from the outside, um, that, um, because of this black magic, it's kind of irreproducible what, what, what they achieve because you don't know how did they come up with those parameters? How often did they try it out? Did they fail? Um, and so there's now, um, a, a move in, in the community where they're, where they're trying to make this more transparent so that it's, it's easier to figure out what was actually done.

   7. CWChris Williamson49:00

      Mm. I, I, part of me feels grateful, I think, that giving a, uh, self-reprogramming super intelligent AGI a little bit more time to come around while we could perhaps get some more wisdom to come along with the technology. That seems like a pretty good idea to me. Uh, e- even if it's not by choice, it's simply by programming language restriction and our own, uh, lack of ability to write whatever it is that we need to write. Part of that feels like a, probably a good thing, I think, overall. I think that technology outstrips wisdom and the fact that we're hitting some real roadblocks is, is interesting. Is it, am I right in thinking that Moore's Law, the computing power doubles every two years on average, that that's actually beginning to slow as well?

   8. SHSabine Hossenfelder49:51

      I, I've, I've heard this too, but then I've heard other people objecting to it. Uh, and so I, I'm really not a technology person, so I'm, I'm not entirely sure who to trust. Uh, but yeah, so, so the death of Moore's Law, um, has been proclaimed a few times already. (laughs) So at this point I'm not really sure who to trust. There's also this entire problem that with the world economy after the pandemic and all the supply chain issues, it might not, you know, it might be an anomaly in the data. So-

   9. CWChris Williamson50:17

      Ah.

   10. SHSabine Hossenfelder50:18

       ... it's maybe, maybe a little bit unfair.

   11. CWChris Williamson50:20

       Yes. Of course, that makes sense. How come, how come nobody gets any younger? Why aren't, why aren't any of us getting any younger?

   12. SHSabine Hossenfelder50:27

       Yeah, so that's an interesting question, um, in that we've, we believe we figured out part of the answer. Um, so in, in physics, this goes under the problem of the arrow of time. Why does time seem to look different in one direction, um, than in the other? And, um, we believe we understand part of it. Uh, part of it is just that, um, certain chains of events, uh, are very likely to happen whereas others are unlikely. So, things are likely to break. Um, disorder is likely to increase, um, but it's very unlikely that things spontaneously unbreak, and it's very unlikely for order spontaneously to decrease, which is just another way of saying that entropy normally increases. Um, but this of course brings up the question like, if entropy constantly increases, why was it small in the past to begin with? And, um, the way that we currently deal with it is that we just say, "Well, the universe was born with a small entropy and we have no idea why, and that makes everything work." And, and that's true, but it, it, it... We have no idea why this was the case. And it, it's, again, it's a question about the initial state, which we already discussed when we were talking about the beginning of the universe. So, it's one of those, um, issues where we, we currently don't even know how to answer the question. It's one of the, it's a question that Penrose is actually trying to answer with his, um, cyclic universe.

   13. CWChris Williamson51:54

       Mm. W- And how do you envision, from a physics perspective, how do you envision time? Is everything happening at once?

   14. SHSabine Hossenfelder52:02

       (laughs) Well, um, at, uh, at least in our current theories, time is a dimension, um, so of course if you map out space and time, then it's not that everything is happening at once. Um, it's just that, that the whole thing together is one mathematical construct that just sits there. It's not that there's one particular moment which we call the present moment that is special in any regard. I mean, it's special in our perception, but then you could say in our perception, uh, each moment is special at some moment in time. So, in that sense they're all equal, if that makes sense (laughs) .

   15. CWChris Williamson52:40

       Would this be kind of the same as saying because I'm stood in this spot right now, this is the spot which is special to me, but if I was stood in a slightly different spot, that would be in terms of the three dimensions, right, of space and time, is it sort of an equally arbitrary choice of now and then now and then now again?

   16. SHSabine Hossenfelder52:59

       Yeah, it's very similar except that in space you can go back and forth, whereas in time you can't.

   17. CWChris Williamson53:04

       Why- Why is that the case? That, that there is an arrow of time, that it has directionality at all?

   18. SHSabine Hossenfelder53:10

       (laughs) Well, well, that's what I just said. We have only, um, half of the answer, right? So, if you're asking like, "Why can't you go back in time?" It's because you were already there, and, um, so if you were allowed to, um, to do it a second time, then that could lead to all kinds of, uh, causal products. Uh, this is the, the usual problem with, with time travel, I'm afraid, so that, and it's not going to happen (laughs) .

   19. CWChris Williamson53:35

       Uh, well, if you say so. Look, Sabine

9. 53:37 – 54:17

   Where to Find Sabine

   1. CWChris Williamson53:37

      Hossenfelder, ladies and gentlemen. If people want to keep up to date with the stuff that you're doing, you've got your YouTube channel which is blowing up now, and your Twitter and your new book. Where should they go?

   2. SHSabine Hossenfelder53:46

      Um, well, just Google my name because it's not a very common name and you'll find more out about me than you ever wanted to know.

   3. CWChris Williamson53:54

      Sabine, I appreciate you. Thank you.

   4. SHSabine Hossenfelder53:57

      (laughs) Thank you.

   5. CWChris Williamson53:58

      What's happening, people? Thank you very much for tuning in. If you enjoyed that episode, then press here for a selection of the best clips from the podcast over the last few weeks. And don't forget to subscribe. Peace.

Episode duration: 54:17