Episode 27: Chiara Marletto and Constructor Theory

• Links to this episode: Spotify / Apple Podcasts

Unofficial, reviewed transcript. It may still contain mistakes; check the podcast when wording matters.

Speakers are labeled by first name where identified. Unidentified-speaker labels may refer to different people. Bracketed question marks indicate uncertain wording.

Transcript

[00:00:07]  Bruce: Welcome to The Theory of Anything podcast. We are very excited. We have Chiara Marletto with us today, and we have an interview planned with her. Plus we have everybody else. We've got Cameo, Tracy and Sadia here. Everybody say hello.

[00:00:21]  Sadia: Hi, everyone.

[00:00:22]  Unidentified speaker: Hello.

[00:00:22]  Chiara: Hi, everyone.

[00:00:24]  Unidentified speaker: Nice to be here. All right.

[00:00:26]  Bruce: I think I'm gonna get just straight into it so we can utilize time here. But Chiara, how did you get involved with David Deutsch's Constructor Theory?

[00:00:36]  Chiara: So I was a PhD student in Oxford at the time. So it was 2011. And at the time I was looking for understanding this idea that von Neumann had about constructors that are these generalized programmable machines that can emulate features of living systems. And the way I landed on that is that I was interested in quantum biology. So I kind of my initial goal in PhD was to try to figure out how to apply some of these ideas from quantum information to the realm of biology and specifically understanding some of these features that self-reproducing entities like bacteria have, and whether they can use quantum effects to speed up their functionalities, et cetera. And somehow, as I was starting, I was sort of digging into this work that von Neumann produced. This was something that he did in the 50s. I just stumbled upon this nice talk that David gave at the Clarendon Laboratory, which is the part of the physics department, about constructor theory.

[00:01:53]  Chiara: And I thought, hey, there's someone who's actually working on an extension of the ideas that von Neumann had. And then I realized that actually this was much more powerful than von Neumann's original ideas. And so I got hooked and I ended up basically working full time on that during my PhD, and then later on also developing it on my own in various directions. So that's how it happened.

[00:02:20]  Bruce: Do bacteria use quantum effects for information? Like, do you know?

[00:02:24]  Chiara: Oh, yeah, I think... Well, I mean, there's this trivial sense in which they do. So of course, the bacteria just work according to the laws of quantum theory. So in some sense, all physical systems in this universe obey the laws of quantum theory. But then there is this other deeper sense that has fascinated both theoretical physicists and biologists in equal measure in the sense of whether natural selection has come up with methods to exploit quantum effects a bit like quantum computers do to speed up some functionalities. And there are some theories about, for example, photosynthetic bacteria. So these are bacteria that live on the bottom of the sea, so they receive very few photons. So they need to really maximize the efficiency of conversion of the photon energy into chemical energy for their own things. And there are theories that kind of suggest that they might be actually using quantum speed up like effects.

[00:03:31]  Chiara: And these theories are at the moment are not fully confirmed. So I think in a way they are conjectures. We haven't had, you know, it's hard to make experiments because it's hard to test the system. They're very messy, but I think this is like what motivated quantum biology, which is like an existing field.

[00:03:50]  Bruce: OK, that makes sense. I guess I had heard that there was some theorization around photosynthesis and quantum effects.

[00:03:57]  Chiara: Yes.

[00:03:58]  Bruce: Now, in your book, you say, wouldn't it be wonderful if it were possible to take inspiration from these principles which relate to counterfactuals and imagine an entirely different way to formulate the laws of physics, one that takes counterfactuals as primitives and the laws of motion and initial conditions as derivative. Now, this is something that caught my attention, although this is something that David Deutsch in his first paper says something somewhat similar. And I've always been a little confused by it, so I wanted to ask about this. The issue that I don't think I understand is why would it be, it seems like constructor theory would be a very valuable theory just exactly the way it is, even if it didn't turn out that the laws of motion initial conditions were derivative of it. What is motivating the desire to see it as underpinning all of physics, rather than being something new that gets added into physics and expands the way we think of physics?

[00:04:55]  Chiara: I really like this question because it touches on a tension, which is actually at the heart of this program in a way. The tension is exactly what you, the thing that you just mentioned, that on the one hand, you couldn't [wording uncertain] think of constructor theory of the program of adding principles that are based on counterfactuals to the existing laws of physics as a supplement of existing physics in a way that it gives us more tools to describe the universe and that's already quite cool if it works in that way. So it's like inventing a new methodology that allows you to go deeper. And in that sense, that doesn't need to, as you said, doesn't need to represent anything more fundamental than, say, dynamical laws themselves. It could just be something that you add on the side, which can help understanding things better. But then, and I think this is actually the vision that David had when he wrote his sort of initial seminal paper where he kind of outlined the whole philosophy of the approach.

[00:06:13]  Chiara: The hope is that not only will this new approach add methodologies that are helpful to understand things better, but it could also provide a deeper foundation for the existing laws. And that is necessary, that's considered as necessary, at least to a different degree by both myself and David, because the dynamical law plus initial condition framework isn't in itself very satisfactory as an explanation for lots of phenomena that exist in the universe and that should be actually included into physics, but currently are not or are only imperfectly included because the dynamical laws are not quite sufficient. And so, as always in physics, it seems natural that when you find an explanatory tool that can dissolve or solve problems within an existing framework, that framework becomes somehow derivative or secondary with respect to this new set of laws and tools that you have discovered.

[00:07:26]  Chiara: And so, that's the second and more ambitious way of looking at constructor theory. We don't know if it will succeed this kind of vision, but I think if it's not constructor theory, it should be something else that provides a deeper explanatory framework for the dynamical laws that we currently know. And I think if it's not constructor theory, it should be something else, but there should be some additional stuff that we discover as being deeper and more fundamental than these laws that we are currently using.

[00:07:56]  Bruce: Okay, that makes sense. In the conclusion of David's original paper, he says the principles of constructor theory that I propose may be false. And then he goes on to say, but if the idea is false, then something else will have to remedy. And then he lists out several of the things that in your book, you show that constructor theory can remedy. So I had often wondered about what he meant there. And I guess it's what you're saying that it may be false, that it underpins all of physics, but it clearly has something true to it. It's already solved substantial problems in physics. Just from reading your book, you go through several examples of that. Is that correct?

[00:08:37]  Chiara: Yes, this is correct. I think we have to be careful here because there are lots of ideas that this question somehow brings into focus. The first point to make is that constructor theory is made of a set of principles. These principles are, on the one hand, supposed to be new laws. They add to the existing laws, as we know them. They have continuity in terms of logic with some principles that exist in thermodynamics, such as the second law or the conservation of energy and so on, in the sense that they are not formulated as dynamical equations, but they are formulated in these more general terms with constraints about counterfactuals, what's possible and what's impossible. But they are new laws. And in that sense, they, as new laws, may be false. So for example, it could be that the principle of interoperability of information, which is one of these new principles that we proposed recently, which basically kind of captures the idea that information can be copied from one system to another with no restrictions and so on.

[00:09:56]  Chiara: So this principle could be actually false in the form we proposed it. But the point is that the approach in itself, the fact that it's kind of using counterfactuals as additional things that are necessary to explain features of the dynamical laws, et cetera, et cetera, seems to be a robust new way of formulating laws, which is here to stay. And in a way, it was already there to stay because it exists already, as I said, in thermodynamics. It was used largely also in the theory of quantum computation and in information theory in general. So in a sense, what we are doing is a continuation of this sort of tradition just on a much broader scale. So I think that's what David meant with that comment.

[00:10:40]  Unidentified speaker: Okay.

[00:10:41]  Chiara: On the other hand, the other stuff that you mentioned, the new ideas that we proposed, some of these new principles actually have consequences, and these consequences appear to be directly testable. And therefore, as I said, it could be that they will be falsified at some point, they could be tweaked. But the hope is that this new approach with counterfactual will stay. And that's the kind of vision behind Constructor Theory.

[00:11:06]  Sadia: Actually, let me ask one thing here. I was actually wondering that, would you say that determinism is basically assumed for Constructor Theory? But by that, I mean like, you know how in the Many Worlds interpretation, it's a pretty much deterministic view where it doesn't really leave any room for saying that there are certain possibilities that just never actualize, but may just remain. And kind of, you know, that other approach might leave room for indeterminism. Would you say that the determinism is kind of assumed as a priori for constructor theory?

[00:11:47]  Chiara: I think not necessarily in the sense that determinism is, it's a nice, it's a very nice question. I think the, so constructor theory doesn't have dynamics as a kind of primitive thing, right? So determinism is often, at least most, you know, most formulations of what deterministic means require there to be some dynamical law that takes states into other states. And then one way of thinking about determinism is that, let's say there's no stochasticity involved in the laws, a bit like, as you said, in the laws of quantum theory, where stochastic evolution is not really there. I mean, this kind of Schrodinger equation, which is deterministic, and then you add on top some probabilistic stuff with the Born rule and so on. But I think the, so in a way, Constructor theory in itself doesn't require determinism to be, you know, principle of nature or anything like that.

[00:12:46]  Chiara: However, when we, when you go particularly to things like the Constructor theory of information, for example, where we lay out these principles that are needed for the counterfactuals that are required for information to be physically possible, then it's implicit in some of these principles that it must be possible, one of the tasks that, so one of the kinds of tasks that must be possible for information to be instantiated are one-to-one tasks. So things that don't destroy or create information out of nothing, as it were.

[00:13:22]  Sadia: Wouldn't that, I'm sorry to interject that, but wouldn't that imply that determinism?

[00:13:28]  Chiara: Yes, so it would imply if you were to make a model, so if you, so the concept of a task doesn't necessarily map onto dynamical law directly, but once you, you know, let's say you want to make a model for what it means for a constructor to perform a task under a certain dynamical law, then for these principles to be true, then it seems like these dynamical laws should be deterministic in the sense that you were envisaging. So you could say, it could be that there is a different constructor theory where these principles are not true, in which deterministic dynamics isn't required. So I'm not ruling that out. But for the moment, at least, if you take these principles as correct in capturing what are the regularities needed for information in the universe, et cetera, et cetera, then yes. So when you want to transfer these principles into constraints on dynamical laws, this would suggest that the dynamical laws have to be deterministic.

[00:14:27]  Sadia: So for example, let's say that our universe, actually there was some sort of an evolutionary principle whereby the universe is actually, the laws are changing or maybe even new laws are being created. So that means that the constructor theory should be able to handle because as long as laws are, you know, we can identify something as lawful, that could, okay.

[00:14:50]  Chiara: Yes, yes, exactly. And again, these laws are not the primitive things. So somehow they should be a consequence of these more fundamental principles.

[00:14:58]  Sadia: In a weird way, it's almost like laws or something that is, it turns out to be a useful way for us to talk about things, you know.

[00:15:04]  Chiara: Yeah.

[00:15:05]  Sadia: Okay, yeah.

[00:15:07]  Bruce: Karl Popper in his epistemology, particularly in his book, The Logic of Scientific Discovery, he formulates his epistemology in terms of universal statements. So for those who aren't familiar with logic, a universal statement would be for all. So for all swans, color is white. That's a false universal statement because we know that black swans exist. Therefore, we can refute or falsify that universal statement. But universal statements, what really makes them useful is that they forbid things so that you are able to refute them. I couldn't help but feel like there was an analogy there to constructor theory. Is that accidental, coincidental, or is constructor theory an attempt to inject Karl Popper's epistemology into physics?

[00:15:52]  Chiara: That is a nice point. I think, well, in this sense, I think constructor theory is not at heart any more Popperian than other bits of physics. It's just that they are not explicit about it in some sense. So let me explain. So all laws that [are] traditionally regarded as fundamental, so they are primitive or somehow they kind of they underlie everything there is in the universe. You know, you can take different definitions for fundamental, but broadly speaking, I mean something like that. So all these laws have the feature that they forbid things in a kind of universal fashion. So and then, sorry, on top of this, I think to falsify them, you have to go along the kind of process that Popper envisaged. So in this sense, I think, you know, the way in which you would falsify a constructor-theoretic law isn't different from the way you would falsify like, I don't know, Newton's laws or something like that.

[00:17:02]  Chiara: And likewise, Newton's laws, as well as the constructor-theoretic laws we proposed, intended to be universal. So they should, you know, apply to everything there is in the universe. And that's why we actually were able to falsify Newton's laws, because at some point we realized that that wasn't the case and it turns out, you know, that they're actually false. But yeah, I think that's kind of one bit of the answer to your question. The other aspect is that I think because of the way both, well, David in the first place, but also in some sense myself and other people working on this, because of the way we are thinking about science, it becomes more explicit, perhaps, that there is a Popperian influence in the way we are posing the problems that we want to solve and so on. And in this sense, maybe in other branches of physics, this is not as obvious because people are not perhaps so interested in these epistemological aspects of the theories that they are investigating.

[00:18:05]  Chiara: But if they thought about what they are doing, they would actually kind of discover themselves as Popperian after all. So I think I don't think Constructor Theory is any more Popperian in its foundations than the rest of science or physics specifically. There is one particular point of contact with Popper's epistemology, which is the theory of knowledge. This is something that I think David has been very keen on for a long time. Knowledge intended as a particular kind of information that can be then defined in a sort of physical manner and puts this idea of knowledge without a knowing subject that somehow is dear to Popper directly into fundamental physics. So I think that's one point where say Constructor Theory gets more into the Popperian epistemology camp than other physical theories. But this is in this specific regard, not as far as the general foundations, which I think are Popperian, broadly speaking, for physics as a whole.

[00:19:08]  Bruce: Okay, that makes sense. Now, on the other hand, I noticed that Karl Popper formulates his epistemology using something that, at least to me, seems very analogous to what you call the prevailing conception. For example, here's a quote from Popper in The Logic of Scientific Discovery. He says, For a universal statement without initial conditions, no basic statement, which would be a testable statement, can be formulated. And he's very specific about the need for and the importance of initial conditions. And really his epistemology formulates things in something that is basically initial conditions plus universal statements, which would be laws, which seems to me to be somewhat analogous to the prevailing conception.

[00:19:51]  Chiara: Yeah, this is a very nice question. And I think the so I guess that there is this aspect that perhaps is relevant to mention that. So in physics in general, not just in constructor theory, when you have a law that doesn't mention initial conditions and dynamical laws, for example, the conservation of energy, the way you test it is ultimately to, you know, to find a prediction about a specific situation which does involve setting up an experiment under some initial conditions and seeing what happens. So I think the logic of testing a theory does involve, as Popper said in that quote, in one way or another, the fact that you do check the kind of predictions of your specific laws in a particular control condition where you can set up an experiment of sorts. So that's one way in which you can, say, falsify a particular law. And this is particularly true of principles.

[00:21:03]  Chiara: Some people think that principles are not testable, precisely because they are not formulated in terms of dynamical laws and initial conditions. But that's untrue in the sense that you can then, given a principle, you can check a model that's compatible with it and then you can specialize that model in a particular initial condition plus dynamical law type of situation and check the predictions of the model. Now, the important thing is that what we are trying to suggest in constructor theory is that the description or an explanation that only uses statements that resort to initial conditions and dynamical laws, that this explanation isn't satisfactory, is not sufficient to describe the whole of physical reality. So the subtle point is that in order to explain some of these predictions that use dynamical laws and initial conditions, you have to resort to these other things that are not formulated in this way, but can be tested through predictions about initial conditions and dynamical laws.

[00:22:14]  Chiara: And these other things are principles about counterfactuals. And it turns out that actually these are deeper than the dynamical laws themselves. It's a bit like when you're thinking of general relativity, it has some dynamical laws and you can set some initial conditions or boundary conditions, whatever. But there's a whole set of principles that come with GR itself, some of which describe the features of space-time and the geometry of space-time and so on and so forth. And these principles are actually in a sense more robust and fundamental than the dynamical laws themselves. But ultimately, the way you test them is to find a particular prediction that they end up making about a specific dynamical situation. But in a way, the dynamical part of this is not the fundamental part. The principles are fundamental.

[00:23:09]  Sadia: Would you say that one of the reasons why this was inspired, the constructor theory was inspired, was that, you know, any type of concepts that you want to [unclear]—become part of a physical theory or a concept around which a physical theory could be made, such as information, maybe there might be a push from a physicist to understand it in terms of like the so far considered fundamental physics, like in, you know, fields in terms of some somehow to relate, relatable to the physics of the fields or particles in fields?

[00:23:44]  Chiara: Yes, I think the so that's exactly the type of view that we are trying to counteract. So I think in some way, because dynamical laws and fields and so on have been so successful so far in various ways, and because most testing goes through them, there is somehow an assumption that everything that's fundamental has to be formulated in terms of those things. And this is a problem because, as you said, things like information are not—the regularities that underlie information can't be expressed in those terms. And if you try to express them in those terms, you end up with an approximate theory or with a theory that's subjective or brings observers into fundamental physics that doesn't, you know, satisfy scientific criteria of various kinds and so on. So I think, in a sense, what we are pointing out is that in order to stick to this, let's just use initial conditions plus dynamical laws because they're very successful.

[00:24:46]  Chiara: We end up with theories that are non-satisfactory or at best approximative for phenomena that are actually very important and they could lead to further important discoveries. And that's why we're advocating, let's try in a different way and take that way seriously, which is this one with counterfactuals.

[00:25:06]  Sadia: Great. And just tying quickly to what Bruce mentioned about Popper, in a way so I can see that these principles could be tested where the initial conditions aren't necessarily as, you know, sometimes people talk about cosmological theories and the initial condition is meant to be like something at a big bang. But we can just always set up an experiment with some initial conditions and then we could test out, you know, a certain model, which in a sense would kind of, you know, would be a way of testing, I guess. Yeah. Okay. Great.

[00:25:39]  Unidentified speaker: Yes, exactly.

[00:25:41]  Bruce: This idea of a prevailing conception, I had never heard of that until reading David's paper and your book. Now, I'm not a physicist, so that's not too surprising. I actually found it a little hard to believe that the prevailing conception was a part of physics, in part because of what you said in your book, the fact that Newton's laws aren't set up as dynamical laws of motions plus initial conditions. So the foundations of physics really was largely rooted in counterfactuals. But talking with Sadia, who is a physicist, she has assured me that that really is the way physicists tend to think, is in terms of this prevailing conception, even though that isn't part of the foundations of physics. I hope you can see why I was a little bit confused there. How did the prevailing conception creep into physics if it really wasn't part of the foundations of physics?

[00:26:34]  Chiara: Yeah, this is an interesting question. Let me speculate a bit about this, because I have been myself trying to somehow understand how we've come to this, as you say. Well, I think one possible answer is that, as I said, the method that uses dynamical laws and initial conditions is just very successful. And it has led to a number of great discoveries. So in physics, Newton's laws are one example, but then later quantum theory and general relativity and also the atomic theory, the fact that everything, lots of phenomena can be reduced to microscopic explanations about particles like atoms and subatomic particles and so on. Moving about according to some trajectories that are set by the dynamical laws. So all of these great steps in understanding reality through physics have been made by largely by appealing to this, you know, this method of explanation through initial conditions and dynamical laws.

[00:27:50]  Chiara: And I think that has slowly made most physicists lean towards the implicit assumption that any new fundamental law should come in the form of initial conditions plus dynamical laws. So that should be like the ultimate type of explanation you want for the universe. And I guess this is not... I think this has been somehow gone unquestioned for decades. And it's been supported, corroborated. This idea has been corroborated by the fact that it is really a very, very powerful way of doing things. It's stunning. It's just amazing, you know, the amount of progress we've made sticking to this kind of logic. However, there is also a different way of doing physics, which is already implicit in thermodynamics. So there are these high level principles that somehow implicitly use counterfactuals. It has gone largely unnoticed that most of the times when there was a moment of, oh, we need to conjecture a new dynamical law, people were actually using the power of principles, such as, say, the second law of thermodynamics or the conservation of energy and so on, to guide them to make guesses for the future laws.

[00:29:15]  Chiara: So somehow this has gone unnoticed but it’s there. And that's why I think in a way what we're doing with constructor theory isn't that different from this other tradition that exists in physics. It's just we are extending it to a much broader set of things. And we are hoping that therefore this prevailing conception, that I prefer to call the traditional conception, because I'm hoping at some point it won't be prevailing anymore, will be balanced by the new approach that also uses other tools. And my view is that we won't make any progress about things like understanding living systems and understanding the way the mind works and so on from a physics point of view, unless we switch to this or a different mode, but at least a different mode of explanation from the dynamical laws initial condition type of logic.

[00:30:13]  Bruce: So might one say that physics has gotten stuck into an inductive rut where they're expecting the future to resemble the past?

[00:30:21]  Sadia: Yes. That's what I was going to say too. It feels like with quantum theory, the success of quantum theory for the longest time under a sort of like instrumentalist type of way might have actually further, you know, where I think physicists start maybe not taking philosophy, giving philosophy its rightful place. Because even in string theory, you've seen like the push has been so much from like mathematics being the horse that's, you know, pulling the carriage and seemed like they just kind of got carried away without, you know, much philosophy being considered. It seems like Einstein actually took the approach where he was very much into philosophy and then he gave principles for his theory. But he actually, interestingly, Barbour has pointed out, Julian Barbour, that he kind of settled more for an operational approach in the end and never really fully addressed some of the question that he actually took over.

[00:31:14]  Sadia: And sorry, I thought I would just mention that maybe too.

[00:31:17]  Chiara: It's a great point. I think it's true that so I guess physics has a huge bias towards empiricism and instrumentalism in various ways. And I think, you know, once this was a good thing because, you know, it was supposed to sort of counteract attitudes such as maybe dogmatic religious takes on, you know, on explaining things in the universe. So this was, you know, a good thing at some point, but somehow this has gone out of hand. And to the point that I think it really trumps progress in many directions in the fundamental sense. So of course you can, there are, you know, countless fields of physics where people are making great progress and they never have to question maybe their, you know, empiricist take on reality because they are not wondering about some of these issues that are more, let's say, the foundations. But when you're really thinking about the foundations of theories, it's very important to understand that, as you were mentioning, that there is much more to an explanatory theory of reality than just what its empirically accessible content.

[00:32:27]  Chiara: And I think this is a thing that people massively struggle with, and this is also why there are so many misconceptions around there about quantum theory itself, because that's the only theory where this dichotomy, this between empirically accessible content and the explanatory content of the theory is most acute. And somehow, you know, people who are really sticking to empiricism end up forgetting about non-empirically accessible content of quantum theory, which is the most important thing in it. And they try to somehow fit the quantum theory as it is in some mold that doesn't belong to it. And that's how we end up thinking about, you know, all sorts of wrong views about, you know, measurements and stuff like that. So, yeah, I think this is empiricism and empiricism and instrumentalism have somehow a huge responsibility in perhaps this fact that physics has got stuck.

[00:33:23]  Sadia: And I don't know if you felt this way, but now that I've kind of become, after I left grad school and my PhD afterwards, I really dive quite a bit into philosophy. And I've literally, I mean, you're not going to believe this, but right now I'm thinking about going over all the main branches of physics, because I know that I'm not going to look at the things the same way. Why aren't we taught physics with that? You know, it's kind of, I mean, don't, I don't know if you had that experience or maybe it was just where I took physics, that the way physics is taught is very much like the, you know, like the professors hardly ever talk about like the philosophical implications or, it's very much as if you would teach an engineer or something, you know, which would be fine, but not for...

[00:34:10]  Chiara: Yes, I think often I had largely the same experience and, you know, other than maybe for a few fortunate cases of lecturers who were not like that, but I think most, I mean, the tone of the textbooks and, you know, the tone of most lectures is actually based on, as you say, emphasizing, well, only the predictive content of the theory, rather than say, you know, the principles. Yeah, exactly. The principles about physical reality and so on. And I think that's also why I guess physics isn't that popular, even like in, say, secondary school settings and so on, because at least the way I was presented, physics was more like, you know, like a set of, it's like a cooking book or something, a set of recipes to solve problems that appear to be completely parochial, like, you know, you've got to predict what happens to an apple when it falls from a tree.

[00:35:12]  Chiara: And that doesn't sound like mind boggling and interesting, whereas, let's say, at the time philosophy was much more interesting to me because it sounded like, you know, they were asking the right questions about the universe and stuff like that. And then later on, I realized that actually physics has the chance of answering some of these questions much in a deeper way than, say, you know, philosophy or things like religion and so on, of course. And it was quite late in my study, you know, physics, that I realized that physics was much more fundamental than the way I was presented it.

[00:35:47]  Sadia: Oh, I had exactly the same experience. And, you know, I used to feel kind of like the odd one out because of the sort of problems that I used to get interested in. And, you know, it felt like I was just some sort of a dreamer or something. Whereas I really wanted to know what, you know, like when I took my, you know, the quantum theory course, I really wanted to know what the implications were for the nature of reality and stuff. But I remember our professor saying, no, it's all about how we set up the system and think about these things as a filter. And then when you get out, it's all about predictions. And, you know, I'm like, OK. It was kind of a bit of a turn off.

[00:36:24]  Chiara: Yes, and exactly it's not, as you say. I mean, the most important part of the theories is really in its explanatory set of tools. And most of them are not about predictions. They are really about, when you talk about the quantum state, there are all sorts of things that some of them are counterfactuals, some others are even informally expressed.

[00:36:46]  Unidentified speaker: But they're all important for the explanation to make sense.

[00:36:48]  Sadia: So this kind of brings me to another interesting question that I wanted to ask you. I wonder whether you've thought about, because I know in your book you mentioned about the problem of time, and I know that you are familiar with Julian Barbour's work too. I was wondering what are your thoughts on the nature of time? Do you think that time might turn out to play a fundamental role in physics? Because it appears that so far when you look into the theories, and I know that quite a few more physicists are becoming convinced that time is pretty much going to turn out to be an emergent concept, and Julian Barbour has done work quite a bit on general relativity and stuff, where he's shown how time turns out to be emergent. But I'm just curious, what are your thoughts since you've been thinking about these things in a deep way?

[00:37:37]  Chiara: So in constructor theory, time isn't primitive, so it's not one of those fundamental concepts, because when you state, for example, that a task or a transformation is possible or impossible, that doesn't refer to, because it doesn't refer to dynamical law directly, it doesn't refer to time in the traditional way that, say, standard dynamical laws do, because standard dynamical laws have time built in as a parameter that exists and sort of labels different instances of the dynamical evolution. And then there's this improvement that we got with general relativity, where space and time ended up being handled on the equal footing, and there was a very important step forward and so on. Now, the fact that time isn't explicitly mentioned when we say a task is possible and so on, doesn't mean that that isn't important to incorporate in the constructor-theoretic picture of physical reality.

[00:38:50]  Chiara: And so this is something that we haven't worked out yet, because at the moment the kind of statements we've been proving and so on are at the same level as what you do in, say, the quantum theory of computation, for example, where you're thinking of composing possible tasks with one another and you're thinking that whatever is performing these tasks will have a clock in it and it will stop its evolution at some point. And the fact that the task is possible exactly refers to all of these things. It's just that because the tasks are primitives and the possible and impossible statements are the primitive statements, we never have to model what the fact that the task is possible means within, say, a dynamical framework.

[00:39:39]  Sadia: Would you say that it really wouldn't matter? Let's say that if time turns out to be in some sense more fundamental, like really fundamental and global rather than emergent, as in clocks being emergent, I think based on what you said earlier when we talked about determinism, it doesn't seem like it's going to really make much of a difference to constructor theory.

[00:40:01]  Chiara: So it won't make a difference in the same way that it does to the dynamical laws, right? So when people say time isn't fundamental and they offer explanations for how to incorporate time in a non-fundamental way within the dynamical law, you have to work quite a lot because obviously dynamical law has time built in, and then you have to make an awful lot of work in order to go from timeless statements into a dynamical law type statement. So I think this is what, for example, Julian Barbour has been doing for, say, classical mechanics, and there are similar proposals for quantum theory, etc. And so it won't be as hard, it won't be making much of a difference to constructor theory compared to what happens in the dynamical law type of approach when you're trying to say that time isn't fundamental. But it will be important for constructor theory to have a set of principles that underlie the notion of clocks.

[00:41:12]  Chiara: So clocks are physical systems with certain counterfactual properties. And what David and I would like to have, and this is something we're working on at the moment, is a theory that likewise does the same for clocks as we did for information. So we want a theory that tells us what are the principles that have to be satisfied by nature for clocks to be possible. And it's an interesting question that you are now posing and something that we haven't decided yet in the sense whether this will say that time is fundamental or not according to these criteria that are set by these other proposals that you mentioned. So in a way, it will be interesting to see whether these principles rule out or in some of the existing proposals to explain time within standard traditional physics approaches. And the answer is very much open, so we don't know at the moment.

[00:42:10]  Sadia: Interesting. Actually, that's one of the things that I was really kind of interested in looking at Julian Barbour's work and then I looked at David Deutsch's work. And then when Constructor Theory came out, I was so excited because I sort of I knew exactly what was the inspiration, but I still kind of lacked like what to do with it. But Julian Barbour kind of talks about these time capsules or, you know, and I kind of wondered what would be the relation with like some sort of a theory of information.

[00:42:39]  Chiara: Yeah, we expect that there is a relation and I think these principles for clocks will be more—will [be], somehow, necessary principles for, even the principles for information to be satisfied. And so this will, you know, this will be contained in this theory that we are trying to construct at the moment for clocks. But I think it's interesting to see whether these principles will have some consequences in, say, ruling out some of the existing proposals for incorporating time within fundamental physics in a way that goes beyond the traditional approach.

[00:43:20]  Sadia: So would you say that, let's say, and now I'm going to, because based on some of the thoughts that I've had, would non-locality, let's say that non-locality turns out to be some sort of more fundamental to locality, where locality and space are emergent. I mean, I'm thinking that constructor theory is going to turn out then to be maybe like an approximate theory, but what are your thoughts? I mean, this is kind of what I have been thinking, but.

[00:43:49]  Chiara: So, so far, the constructor theory we formulated is based on a very kind of strong principle of locality. So in a way, any theory that has non-locality built in will be ruled out, or at least it shouldn't be a fundamental description of reality, could be an emergent one or something of that sort. Now, this is really a camp where, it's really a kind of intuition. So there are no other reasons than the fact that we have a number of good explanations that are based on locality. So the fact that we have a number of good explanations for physical reality that are local is mainly the reason why we expect locality to be a kind of good principle to set out and look for better explanations. Of course, it could be that we are wrong and it would be interesting to find out. And I think, as I said, this doesn't necessarily mean the Constructor Theory is in itself wrong.

[00:44:58]  Chiara: It just would mean that one of the principles isn't correct. However, I have a very strong feeling that, given that quantum theory and general relativity and special relativity as well are all local, it sounds natural and nicely rooted in whatever we understand about the universe to expect that [the next theory?] will also be local. And however, I mean, I think it's nice that there are some groups that are looking in different directions and I think non-locality is a thing that has been sometimes misunderstood because there is this idea that entanglement in quantum theory leads to non-locality and that is wrong, I think it's a misconception that has been created by a jargon that people created at some point when Bell's theorem was proposed. Yeah, but I think so other than the specific use of non-locality, which is not quite right, it's certainly interesting to investigate what happens if you drop locality in a theory and see what other things go away with it.

[00:46:08]  Sadia: And it's not even like we're dropping. I mean, I guess what I'm wondering, because I've been thinking about Lee Smolin's work, the recent work where he takes time as global and then locality is supposed to be like an emergent and so our clocks in that theory would be emergent in a theory like that. But I think that would be interesting, wouldn't you agree? Kind of like just sort of a thing to try out to see, you know.

[00:46:30]  Chiara: Yes, I think it's so I, as I said, I have a strong intuition that that isn't the way to go. But I think it's interesting that some people are trying to see what that implies. And it could be that this leads to something interesting, too. Note that the theory of information, I mean, all of the good features of information and the fact that we even can set up tests and the theories are testable, does require locality to be satisfied, at least on some kind of scale. So if you, as you say, you know, you might want to have an explanation which is underlying everything, which is non-local, but then you would want that to be compatible with the fact that locality is satisfied at a certain level of...

[00:47:17]  Bruce: Was it quantum mechanics that introduced, got people starting to take non-locality seriously? By the way, I agree that quantum mechanics does not actually imply non-locality. I think it's very much a local theory. We just did a podcast on that with Sam Kuypers. But is that where that came from? Like, it seems strange in a way that physics would even toy with non-locality as a layman on the outside. It sounds so supernatural.

[00:47:45]  Chiara: Yeah, I think that the intuition of physicists has always been, at least the majority of physicists that put forward various theories, was that they were looking for theories that were local. Newton was greatly bothered by the fact that his mechanics had some action at distance built in. That was something that he really didn't like. The theory of fields, which was proposed with first Maxwell's equations with light and so on, and then later on quantum field theory, have somehow managed to provide a relatively good explanation for a local theory that explains how disturbances are propagated in this local way that allows us to consider the effects on a physical system that is decoupled from the rest of the universe as being confined to that system only unless the system gets then coupled to other systems and then thereby propagating its signals, you know, elsewhere in the universe.

[00:48:58]  Chiara: But I think you're right that somehow, as probably Sam Kuypers mentioned, the quantum theory was initially misunderstood and there were lots of comments even from the founding fathers to the fact that phenomena like entanglement and superposition could lead to a violation of some of these principles of locality formulated in various ways. This isn't true. Quantum field theory is basically based on the fact that there's no action at a distance. But somehow this hasn't quite percolated from the practitioners of that field into the general areas of physics where people maybe are not so familiar with quantum field theory and they're still doing some quantum theory which isn't quite relativistic and so on. And from there to the public. And so I think the idea that quantum theory is non-local is quite widespread and despite being wrong, it's still kind of rooted into the way in which people perceive quantum theory.

[00:50:10]  Chiara: And one of the reasons why quantum theory unfortunately is greatly misunderstood.

[00:50:14]  Sadia: And actually, you know, when we do quantum field theory, locality is actually the starting assumption, right? It's not like something that's pretty much we start by assuming that. OK.

[00:50:25]  Chiara: Yeah. Because the equations are local in that way. Yeah, that's right. Yes.

[00:50:30]  Bruce: Is some of the resistance to giving up on non-locality because that would kind of force us into a many worlds interpretation and many people are bothered by that interpretation?

[00:50:41]  Chiara: Oh, yeah, very much so. I think the so I think that there has been this phenomenon, which I can't quite explain, where quantum theory and general relativity get proposed around about the same time, in the sense that it's kind of around the same period. And they're both mind-boggling theories, and both of them are really interesting and deep and have got lots of counterintuitive implications. And the quantum multiverse is one of them for quantum theory. But while GR, general relativity, is largely considered as very hard to understand, but still fine. So no one, very few people... I mean, there is a discussion about what it means and so on, but somehow no one has created a case about it. Quantum theory, on the other hand, is actually, we are stuck on discussing some things, such as, is it non-local? Do we have a problem with measurement? How do we fit classical reality within quantum universe?

[00:51:53]  Chiara: Blah, blah. And all of these questions are actually solved, answered very nicely within this many worlds or Everettian take or relative states take on quantum theory, which has been perfected since it was proposed by Everett, by many people. But somehow the fact that, as you say, you have to take this step and consider things like different branches of the wave function as coexisting, the fact that the ultimate descriptors of physical reality are not real numbers, but they are matrices, operators, Q numbers, which is basically the central tenet for locality in quantum field theory. These things haven't been swallowed. They haven't been swallowed as much as, you know, the fact that, say, space-time is represented by a four-dimensional manifold and its geometry is very counterintuitive, doesn't correspond to what we expect, you know, in what we see, what we experience.

[00:53:06]  Chiara: Obviously, you know, I perceive my reality around me as just a 3D thing and time is flowing, and it doesn't seem to match the idea that Einstein suggests within GR. But somehow I'm not using this as an argument against thinking that space-time is fundamental. Whereas in quantum theory, even practitioners sometimes get fooled and they say, oh, but I experience a classical reality, how can this be compatible with what quantum theory says about the quantum multiverse? And I still don't know why we haven't moved past this stage, given that there is a perfectly fine explanation for how the quantum multiverse works.

[00:53:46]  Sadia: Just to be fair, and this is something I haven't had the chance to dive into, but I'm merely saying it as I've heard from Lee Smolin, that when you look at the quantum gravity research so far, I remember him saying that the research in quantum gravity has kind of indicated there are some conflicts there that seem to indicate that non-locality might need to be taken seriously at the fundamental level. But since I haven't really explored it, I can't really say much other than just quoting him. I don't know if you know anything about that or...

[00:54:22]  Chiara: Well, I think that the best approximation we have to quantum theories of gravity are local in the sense that quantum field theory is. Now, it could be that... I mean, if I understand correctly, I think Lee Smolin has a view that doesn't quite... That clashes with this idea of the quantum multiverse, and I think he's not happy with it, if I understand correctly what he's thinking. So in some sense, maybe that's why he then goes on to saying, well, given that I am not happy with taking on board that the descriptors of fundamental... Sorry, that the fundamental descriptors of physical reality are Q numbers and not real numbers, then I have to pay a price for that. And one of the prices you have to pay is the fact that you have to... Somehow think that dynamical laws are non-local. By the way, this was true. I mean, this was realized by this, by Bohm.

[00:55:24]  Chiara: De Broglie and Bohm have this theory called... What we call now pilot wave theory or De Broglie-Bohm theory, which is like a theory which is empirically more or less equivalent to quantum theory. But its dynamical laws are non-local and they use additional descriptors that are real valued, which obey these non-local equations. So if you're interested in, I think, possibly Lee Smolin is interested in that and he might...

[00:55:58]  Sadia: No, he actually kind of, he talks about it in his book, and then he kind of pretty quickly ditches that idea to showing that, you know...

[00:56:05]  Chiara: Yeah, because that theory is actually problematic in various ways.

[00:56:09]  Sadia: Yes, so he recognizes that. I think the motivation is not to get too sidetracked, but I think seems like his motivation is somewhere coming from certain suspicions from research in quantum gravity.

[00:56:20]  Chiara: There may well be, there are lots of wild different models that have been proposed. The problem with these models is that we are not sure how to link them with somehow predictions that can be then tested. So in a sense, though they might be appealing in some sense, it would be nice to have a way of first explanatorily connecting them with what we know at the scale that we can test.

[00:56:48]  Sadia: Oh, he has actually come up with some really interesting tests, but not to get away from it again, but...

[00:56:53]  Chiara: Yeah, I don't think it's [unclear]...

[00:56:55]  Unidentified speaker: I don't want to get too excited. [Unclear overlapping speech.]

[00:56:57]  Chiara: I think it's, yeah, I'm kind of very skeptical about that.

[00:57:02]  Bruce: Well, his test, the one that you explained to me, Sadia, it is an existential statement. So it actually violates Popper's epistemology because it's not a universal statement. So it's not actually an empirical test yet in its current formulation.

[00:57:16]  Sadia: Yeah, I guess we can just discuss that some other time.

[00:57:19]  Chiara: I think it's interesting. This brings in another element to the discussion in the sense that quantum gravity is... So there are... What happened in that domain is that since we realized that quantum theory and general relativity are fundamentally different as far as the information theoretic structure is concerned in the sense that quantum theory supports universal quantum computing models and GR is a classical theory, so it doesn't support quantum effects or quantum computation, if you like. Well, what happened is that lots of people tried to then come up with very smart proposals to put them together. And we got some kind of working model, which is like a quantum field theory model, which we know is incorrect, because it doesn't quite capture the full of the full GR, but at least can provide some predictions in some domains, but we know it's not going to be the right theory, because it kind of slaughters the whole of GR and makes it not true.

[00:58:18]  Chiara: And at the same time, we have other models that are mathematically satisfactory, but somehow they don't quite... It's very hard to actually make them predictive about some real testable situations.

[00:58:34]  Sadia: So wouldn't it be fair to say that some other options, like right now, as of now, there is an open question and we don't know what might work out at the end.

[00:58:44]  Chiara: Yes.

[00:58:45]  Sadia: Even though, yes, locality has beautifully worked so far.

[00:58:50]  Chiara: What I want to say is that somehow it's nice to have additional principles that can help you and guide you to make guesses in this context, because most of these proposals are now guided by how do we tweak the mathematics of this or that theory or one or the other theory in order to merge them together, whereas somehow, in addition to this mathematical intuition, one should have a very powerful physical intuition, the same way that, say, that Einstein was guided to conjecture his own theories. And somehow we are hoping that some of these theories with constructor theory can be used to that end.

[00:59:29]  Bruce: By the way, just as a side note, Chiara mentioned Q numbers versus real numbers. On the Theory of Anything podcast YouTube channel, Sam Kuypers has a lecture that he gives on what the difference between those two is, and I won't go into it, but for anyone who's curious about that, he explains the difference between those two.

[00:59:54]  Sadia: One of the things in constructor theory, it seems like there's a mention of universal constructors, but I've been kind of wondering what's the relationship between, say, universal explainers and universal constructor, and how do you think the constructor theory will accommodate universal explainers? What would it take, maybe? Or do you think right now the theory has what it takes to accommodate?

[01:00:20]  Bruce: Or shed light on?

[01:00:22]  Sadia: Actually, before I go on, there is another thing that we wanted to ask you which might be tied to it, so I'll just kind of read that out as well. This was something to do with free will that you had mentioned in the book, that unpredictability of action or free will, you said, is therefore another counterfactual, that dynamical law approach does not seem to be able to accommodate, which I kind of understand now that we've kind of talked a little bit about what you mean, but maybe if you could say something about universal explainers.

[01:00:52]  Chiara: Yes, so the first thing to say is that the, so this is in, you know, the context is this theory of knowledge that David Deutsch came up with to, say, connect, I would say perhaps Popperian epistemology with physics. And the, so the seeds of this are back in his Fabric of Reality book, and then they were expanded on in the beginning of infinity. And also in the seminal paper on Constructor Theory, where he kind of mentions this idea of knowledge as being a kind of information that has this counterfactual property of being resilient or robust. And so, in other words, it's got the ability to stay instantiated in, or embodied in physical systems. And when I say information, again, someone might have a kind of fluffy concept in mind, informally expressed. But I think it's nice to point out that in Constructor Theory, you can express what information is in terms of these counterfactual properties that physical systems must have.

[01:02:10]  Chiara: So we somehow brought that concept in physics. And therefore, saying that knowledge is a special kind of information is another way of saying that knowledge is also a physical thing, just like energies and so on. Now, the question is, what are the laws that knowledge obeys? And here is where, Sadia, you mentioned this idea of the universal explainers. Universal explainers are entities that can formulate universal explanations and therefore can conjecture new knowledge. And the question of how does new knowledge come into the world is out of no knowledge, is actually a major question that was somewhat addressed by Darwin's theory of evolution, because the natural selection is one way in which knowledge, for example, the knowledge in genes, comes about and gets tentatively constructed through evolution from very simple elementary beginnings.

[01:03:13]  Chiara: But of course, there is the other process that brings knowledge about, which is the process of thinking that happens in our brains and possibly happens in other entities in the universe, brains or similar things pertaining to aliens, other civilizations that might exist in the universe. And we don't know how that process really works. We know it at the descriptive level in the epistemological sense, so we have this idea of conjectures and refutations and so on. But if we wanted to have physical laws about knowledge and the way it grows once it's created and the way, you know, why is it resilient in some contexts and it isn't in others? What creates and destroys knowledge and so on? We don't have laws about that. So the question of free will is very strongly linked with the question of understanding how new knowledge is created, because whenever someone makes a choice, there is some knowledge being created about, you know, the explanatory reasons why that choice is made versus another.

[01:04:20]  Chiara: And so the whole of this problem of free will and explaining, say, knowledge creation more broadly within physics cannot be phrased even within the traditional conception, because the traditional conception is just about the universe going along a specific trajectory which has a specific initial condition. And once you are on that trajectory, there's no knowledge or there are no choices being really made because it's just one thing after another. So in a sense, the fact that I appear to have the choice now to, after we've spoken, go and read a book or go and have a walk, that isn't quite true. It's just a sort of appearance of my experience, but actually ultimately I'm just executing whatever it is fixed by the initial conditions of the universe. Now, in this context, I think one way of getting out of this problem, which is the thing you mentioned, is that you could still say that there is a certain unpredictability as far as my decision is concerned because the only way in which, say, you could come up with a prediction about what I will do and therefore show that I'm actually predetermined is to have a full simulation of me as a brain, let's say, a copy, a clone of myself, that will then come up with the decision and then this happens before I make this decision, and then you will know ahead of me what choice I will execute.

[01:05:55]  Chiara: But suppose that there were a principle of physics that said that it's actually impossible, so another counterfactual, for you to come up with this simulation with certain limited resources that we have available. Well, that would be one way in which some of these statements that have been suggested in order to explain how the emergence of free will is actually compatible with determinism in physics could be quantitatively explained within physics. So that's one way in which I was conjecturing in the book, along the lines of what David conjectured and others, that Constructor Theory could solve this problem [unclear linking phrase; possibly “or at least”] provide the tools to address the problem once and for all. Yeah, sorry, go ahead.

[01:06:44]  Bruce: Well, go ahead. I want to finish what you were saying, and then I have a question about it.

[01:06:48]  Chiara: No, I just wanted to finish with this question about universal explainers that Sadia mentioned. This connects to the universal explainer's problem because... so constructor theory is supposed to be a generalization of the quantum theory of computation to cover not just universal quantum computers or Turing machines, but also those objects that von Neumann called constructors, which are programmable machines that can be programmed to perform other tasks than computations, like building a house or constructing a computer, afresh, or doing some other cooking, something like preparing coffee or something like that. And then there is the universal constructor, which is like the most general [unclear] machines that has the ability to perform all transformations that are physically permitted. But constructors don't have this free will thing, so they can be programmed and they will execute whatever problem you put in them.

[01:07:53]  Chiara: But a thinking brain has the ability of somehow be programmed to do something and then do something else. And that's, I think, where Sadia's question was going. Can these entities also be explained within constructor theory? And the answer is we don't know, but we are hoping that they will be explained. And counterfactual thinking is important for this. It's the key to crack this problem.

[01:08:19]  Sadia: So when you said that you think it will be explained, does that have something to do with the type of law or the way this comes about? What would it be like? I guess, Bruce, you had a question because I kind of want to...

[01:08:31]  Bruce: Yeah, let me ask my question. So I was actually surprised by what you said in your book about that, Chiara. You mentioned something called the deterministic nightmare. And the reason why I was surprised is because I had understood that problem as not existing at the level of physics, but being an emergent property of computational theory. So I had a thought experiment that I actually invented for a totally different purpose. But the thought experiment was suppose you have a computer that's running one or more AGI's, and they have some sort of environment that they're living in that's sealed off from our world. Virtual world, it may not be very much like ours. You can imagine the computer being either a classical computer or a quantum computer. Wouldn't they then be living, I mean, like they aren't interacting with our physics. They're completely at the emergent level, but to the best of our knowledge, according to computational universality, we should be able to make a virtual world with one or more AGI's in it that are knowledge creating entities and that are people.

[01:09:35]  Bruce: Wouldn't they then be in this deterministic nightmare, even though what you call the deterministic nightmare, even though they're not part of our physics?

[01:09:46]  Chiara: I think they would be. But the question is, is there still a sense in which even if you assume this deterministic nightmare scenario, is there still a sense in which you can say that all of these AGI's, each of them has some inherent unpredictability as far as their own thoughts and choices?

[01:10:10]  Bruce: I see. Let me just take that one step further then. I would assume, like Stephen Wolfram talks about this, there is no way to predict the vast majority of deterministic computations. Presumably, they aren't predictable the first time. But what if you're running two of them and one's on a slower computer? The second one would be predictable under this thought experiment, wouldn't it?

[01:10:32]  Chiara: Yes, it would be. But the question is, predictable means something. So here is the kind of problem that currently, the sense in which predictable or computable is used somewhat always refer to this realm of the Turing machine. So the Turing machine is a highly idealized thing that is already very complex and is supposed to exist somehow by fiat. And then you can refer lots of statements about computability or possibility of certain things to that. So like if you [unclear phrase about a machine] and blah, blah, you don't worry about the resources that you need to run it. Would you be able to say, I don't know, predict that a certain program will halt or not or something of that sort? Now, these problems are meaningful within the current complexity theory because that's all rooted in the Turing machine model. But to give a physical answer, a physics answer to the question, can I, can, you know, is it possible to predict the thoughts of the kind of brain that's thinking?

[01:11:51]  Chiara: So a physics question to that answer has to take into account other things. Particularly, it has to take into account resources with which this task has to be accomplished. And it's not valid to give an answer to that question, assuming that you've got a Turing machine available. And so that's where constructor theory is relevant, because it goes outside of this complexity theory type of scenario where you're only referring everything to an existing Turing machine. And it's saying, no, Turing machines are actually emergent objects that presuppose, I mean, they're already assuming lots of things to be available. Let's not assume that. And let's ask the question from a fundamental physics point of view. And there we don't have the answer in the sense that in principle, yes, it's possible, but that depends on the constraints that I'm putting on you.

[01:12:44]  Chiara: So for example, if I'm asking you to do this in a limited amount of time with limited resources, you may not be able to do it. And so all of the statements currently are very informal and they are very qualitative and a bit imprecise. But the hope is that with constructor theory, you can actually have a complexity theory, which is more rooted into physics than the current complexity theory. And this will also answer this question about what does it mean to emulate the thinking of a human being or a brain?

[01:13:15]  Bruce: If that were true, wouldn't that violate universal computationality?

[01:13:20]  Sadia: Or the interoperability? I mean, I was actually kind of thinking along the same lines, too, that is the worry that what if, just to give an hypothetical, what if there's some new sort of a law at play? Like there's some element to it where interoperability is not possible, like matter and whatever this new law is operating on kind of communicate, but there are only two [wording uncertain] certain channels, but there may be other things. In other words, there may not be some sort of a unified theory that could connect the two.

[01:13:54]  Chiara: Well, that's one reason why, okay, that's not the physics we think we have, but it could be that we are in the physics of that kind, in which case lots of things to do with universal computer computation may not be true or might have to be revised. So that's one, that's for sure. I mean, that's one scenario that's possible. It's not, I don't think it's plausible. It could be the case.

[01:14:18]  Sadia: Like a [unclear] form of dualism, right, here that we're...

[01:14:22]  Chiara: Yes, although then, you know, because we are thinking that there is always some way of explaining, let's say, if there is a dichotomy of that sort, there should be a law that explains why that is, you know, that there are two kinds of information that are not interchangeable. A bit like, you know, when in GR, we say that mass and energy are equivalent, and mass or energy couples with other things in the same way. There are not different kinds of masses or energies. Likewise, information is similar in that sense. And if you had different kinds of information, then you would really have to have a good explanation for why that is. So it wouldn't be really like a dualism. It would be more like, oh, I have a really new groundbreaking explanation, and this would probably lead to all sorts of interesting phenomena that I can't even conceive at the moment.

[01:15:19]  Sadia: I guess in a way we wouldn't even be able to access that, so it almost would be kind of like invoking supernatural in a way, right?

[01:15:27]  Chiara: Yes. And in response to Bruce's question, I think the universality of computation is based on the model. So it's kind of assuming a certain model of the Turing machine, and the Turing machine is running with side effects. And no one cares when you talk about computation, no one actually ever somehow models those. And that's part of the reasons why complexity theory isn't currently satisfactory, because it's satisfactory for some purposes, but it doesn't quite do the job of addressing all tasks and that's also why the theory of computation, despite the universality of computation, is not the whole of physics. That's one of the key insights, I guess, that David has in his Constructor Theory paper, is that that was one of the things that hooked me at the start of my interest in this topic, is that it's not enough to just study computers in order to understand the universe, because there are things that in programs are not distinguished.

[01:16:39]  Chiara: For example, models of the universe with different values of the fine structure constant. Those programs are all the same as far as the theory of computation is concerned, but as far as physics, they are very different. And Constructor Theory is needed to extend the quantum theory of computation to the whole of physics, precisely because of this reason.

[01:17:00]  Bruce: Actually, can I ask about the dust question? Because I was actually curious about that, because it relates to the interoperability. In the book, Chiara, you mentioned this thing called dust. You made it up. It's not necessarily a real thing, but it's this thing called dust that only interacts with our universe through gravitation or something along those lines, and how that would then violate the theory of interoperability, and there would be no way to build a universal computer in such a universe. Did I understand that correctly? First of all, I want to make sure I understood that correctly, but is that what you were trying to say in the book? Then that was kind of inspired by dark matter, right? Because that's one of the theories of dark matter, is that it's just like what you call dust.

[01:17:51]  Chiara: I think I wanted to be a little bit more general. I didn't want to refer to specific models because all of them have their own details, and I didn't want to enter that specific discussion, so somehow I wanted to be very broad. But I think, yes, I guess this is one way of showing why understanding the physics of information has important consequences also to rule out or rule in or understand the implications of some of these models that we have for the universe. For instance, the fact that we assume that computers are interchangeable in some way and bits are interchangeable and that there are no two different kinds of bits that can't be copied onto one another also means that we are not allowed to assume that there are sectors of the universe which are kind of severed from ours in a fundamental way. And so, you know, any kind of interactions that we are, if we are postulating these two sectors of the universe, any kind of weak interaction that might happen between these two should be strong enough to be amplifiable to then perform a copy-like operation so that information in our sector, information in the other sector are interchangeable.

[01:19:17]  Chiara: And if they are not, then we have to be prepared to consider the fact that, you know, universal computers may not be possible in the way that we think they are now. So that's one way of illustrating why the physics of information is not just some kind of quirky, non-fundamental, emergent theory that maybe might be of interest for engineering and technology, but it is actually quite deep and it puts lots of restrictions, the same way that, for example, Noether's theorem puts restrictions on the kind of Lagrangian mechanics that we expect describe some of our most fundamental models at present.

[01:20:00]  Bruce: This is probably an out-there question, but that made sense what you just said. Why is that not also true for many worlds quantum physics, where you have worlds that de-cohere from each other so that they can't have interoperability between them anymore? Does that also imply that there's no such thing as a universal computer?

[01:20:20]  Chiara: Oh, no, that is not the case because the point of...

[01:20:32]  Unidentified speaker: The...

[01:20:32]  Chiara: So the different branches of the multiverse are not universes in the same sense that, say, these sectors are. So it's still possible to reconcile the existence of these universes with the interoperability of information and with the fact that in each of these universes information is interoperable.

[01:20:54]  Bruce: I see. That actually makes perfect sense.

[01:20:57]  Sadia: Would you say that then the conservation of information is important? Because, you know, like in the multiverse thing, we can define this thing called the super information. So the whole idea is that they're all, you know, they operate under the same laws, but there is the super information. So if the information was not conserved, like first of all, I know that the way even if it's defined, you know, there's, does that question make any sense as to what I'm saying? Is the conservation of the information an important thing for constructor theory, right?

[01:21:31]  Chiara: Yes, yes. Well, I guess maybe one should be a bit clear about what we mean by conservation of information. But I guess, yeah, there is a sense in which the interoperability of information has some kind of analogy with the idea that the type of information has to be preserved in different interactions. And when you put two systems together, each of which can kind of work as information medium, the composite system should still conserve the capability, you know, this kind of conservation of counterfactual properties, if you like. And so, yeah, I think in a way it's a good analogy to think of the interoperability of information as stating some kind of conservation of information.

[01:22:21]  Sadia: But when we, going back a little bit at the beginning of our discussion, we kind of discussed that what if there was, and I don't know if you used the word information there, but what if the universe was creating some new laws through some sort of an evolutionary process? You know, I mean, there I kind of struggle that would that, like, you know, obviously, to me, it would seem like the universe would be creating new information. Any thoughts on that? Or let's say that if, you know, a second law of conservation turns out to have the final say. I know that in, you know, the way the many-worlds interpretation deals with is that the information kind of branches out. But let's say somehow, you know, the universe could create, maybe we'll restrict the universe creating new laws. I guess, do you think the information would be conserved? I don't know, I have some fuzzy thoughts there.

[01:23:16]  Chiara: Well, that is a nice question. I think, as I said, perhaps we should be a bit more clear about what we mean by conserving information.

[01:23:32]  Sadia: Conservation basically just means that, okay, even if there was a new law and we discovered it, it would become part of our picture. But that doesn't necessarily mean that there was some given information since the Big Bang that we're just uncovering, right?

[01:23:45]  Chiara: Yes, you could think of some... So the fact that information is conserved, one way of thinking about it is that things that are distinguishable... So the degree of distinguishability of things doesn't get changed with dynamical evolution. So if I've got two states of a system like a flag, one is red and the other one is blue, and then I apply some dynamically allowed transformation and I change them into some other states, then I expect the final states of these two flags if we are just applying fundamental dynamical evolutions to still be distinguishable. And a dynamics whereby blue and red get glued together and they turn both to be, you know, the flag gets, both flags get colored by the same color, are dynamics that are possible but with side effects. So there has to be some track of the fact that I just erased the difference between these two states.

[01:24:54]  Chiara: That's what a dynamical law that conserves information does. And I guess you could imagine a law like, you know, quantum theory that has this feature that conserves information in this sense. And then a different law that still does that, but it's just different from quantum theory. So the fact that the universe might be going through these laws as, you know, into, in some of these hypothetical speculative models doesn't necessarily mean that the information gets deleted or erased in a way that [unclear verb] conservation of information in the sense that I said. So yeah, I think it's possible that some of these models are compatible with the principles of information as we know them now.

[01:25:38]  Sadia: All right.

[01:25:38]  Bruce: Just as a side note, we are at an hour and a half, so it might be good to start wrapping this up and let Chiara get back to her life here. And I'll be erasing this part of the show where I just said that.

[01:25:54]  Chiara: No problem.

[01:25:55]  Bruce: Do we want to allow any final question or should I go ahead and wrap things up?

[01:26:01]  Sadia: Fine with me. I could go on forever, but yeah.

[01:26:03]  Unidentified speaker: [Unclear: “I think we should better go.”]

[01:26:05]  Sadia: This was so much fun.

[01:26:06]  Chiara: Yeah, it was really nice. I enjoyed it too. By the way, I mean, I'm happy to continue conversation at some other point or through emails or whatever way.

[01:26:18]  Bruce: That would be awesome. I feel the same way as Sadia. This is great having someone to ask a lot of these questions.

[01:26:26]  Chiara: Thanks, guys. It's really nice. I enjoyed it too. Yeah, it was really nice questions too. Very, very nice.

[01:26:31]  Unidentified speaker: All right.

[01:26:32]  Bruce: Thank you. Chiara, thank you for coming on the show. We really appreciate it. This has been an exciting episode. And I know that I'm just very grateful for you being so willing to share your time and your mind with us like this. So thank you.

[01:26:49]  Chiara: Yeah. Thank you. Thank you for having me. It was real fun.

[01:26:53]  Sadia: Yeah. Thank you very much, Chiara. This was such a treat. Thank you very much.

[01:26:58]  Unidentified speaker: Thank you.

[01:27:02]  Bruce: The Theory of Anything podcast could use your help. We have a small but loyal audience, and we'd like to get the word out about the podcast to others so others can enjoy it as well. To the best of our knowledge, we're the only podcast that covers all four strands of David Deutsch's philosophy, as well as other interesting subjects. If you're enjoying this podcast, please give us a five-star rating on Apple podcasts. This can usually be done right inside your podcast player, or you can Google The Theory of Anything podcast Apple or something like that. Some players have their own rating system, and giving us a five-star rating on any rating system would be helpful. If you enjoy a particular episode, please consider tweeting about us or linking to us on Facebook or other social media to help get the word out. If you are interested in financially supporting the podcast, we have two ways to do that.

[01:27:48]  Bruce: The first is via our podcast host site, Anchor. Just go to anchor.fm slash four dash strands, F-O-U-R dash S-T-R-A-N-D-S. There's a support button available that allows you to do reoccurring donations. If you want to make a one-time donation, go to our blog, which is fourstrands.org. There is a donation button there that uses PayPal. Thank you.