Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
2 | David Wallace — The Emergent Multiverse
428 Plays1 year ago
We live in a branching universe. If it can happen, it does happen.
These are the almost incredible claims of the Many Worlds Interpretation of quantum mechanics. Yet today’s guest, David Wallace, makes a case that this is the most grounded way of reading our best theory of nature.
While at first sight quantum mechanics seems to say that things (famously, cats!) can occupy impossible states, David argues that a careful reading shows we can take seriously “superpositions” (these apparently weird states) not only at the microscopic level but all the way up to the scale of the universe.
This way of thinking about quantum mechanics was first proposed in 1957 by Hugh Everett, David has made important contributions — particularly in the “preferred basis” or “counting problem” which asks how many worlds are there; and also in understanding how a deterministic theory of the world appears indeterministic — probabilistic — to agents.
David has PhDs in both physics and philosophy from the University of Oxford and currently holds the Mellon Chair in Philosophy of Science at the University of Pittsburgh.
- References and discussion on the Multiverses website
- David Wallace’s research page — https://sites.pitt.edu/~dmw121/
- The Emergent Multiverse — the most comprehensive book so far on the Many Worlds intepretation
- Very Short Introduction to The Philosophy of Physics — does what it says on the tin (very well!)
Recommended
Transcript
Introduction to Quantum Mechanics
00:00:00
Speaker
Quantum Mechanics is our best theory of nature. It's incredibly predictably successful, and yet it's hard to understand what it's telling us about the world, what claims it's making about reality.
Introducing David Wallace and His Work
00:00:10
Speaker
Our guest today is David Wallace. David is a professor at the University of Pittsburgh, and he's one of the leading advocates of the Many Worlds interpretation of Quantum Mechanics. He's written a wonderful book, The Immersion Multiverse, which is perhaps THE book on this interpretation.
00:00:25
Speaker
and it's gained glowing praise from people like Sean Carroll and David Deutsch themselves very active proponents of this way of thinking about quantum mechanics.
Many Worlds Interpretation Explained
00:00:35
Speaker
The many worlds interpretation was first proposed in 1957 by Hugh Everett and is also known for that reason as the Everett interpretation and if it's correct it has almost incredible consequences.
00:00:50
Speaker
would mean that we live in a branching multiverse where anything that physically can happen does happen. But David will argue that it has very unassuming origins. It just reads off the mathematical formalism of quantum mechanics without adding or taking away anything. And it doesn't require that we contort ourselves philosophically either, but rather it agrees with a naturalist view of things, or a realist view of science, where
00:01:20
Speaker
science is making claims about things in the world. I was very privileged to have David as my tutor when I studied physics and philosophy at Oxford over a decade ago, and I'm very privileged to be able to have had this conversation with him. I hope you enjoyed as much as I enjoyed having it.
Podcast Overview with James Robinson
00:01:41
Speaker
I'm James Robinson. I'm one of the founders of the technology company Open Signal, but in another world, I'm not. This is multiverses.
Exploring the Quantum Measurement Problem
00:02:05
Speaker
David Wallace, thank you for joining me and welcome to multiverses. Right in the very first sentence of the introduction of your book, The Emergent Multiverse, you make
00:02:20
Speaker
what for some will be quite a bold statement, that there is no quantum measurement problem. For others, that will be, so what? Perhaps you can start by explaining why one might think and many do think that there is an issue with measurement in quantum mechanics. Good. Okay, so quantum mechanics, which I should say is, you know, our best theories are very small, but because big things are made out of small things, it's effectively our best theory of more or less everything in physics.
00:02:50
Speaker
So quantum mechanics seems to say something very weird about what material objects do. And the way it does that, it has a thing that physicists call a superposition principle, which basically says if a given thing can have one property, and it can also have a different property, you can have both properties at the same time. So if I've got a particle that could be here,
00:03:17
Speaker
it can be here and there at the same time, or if the particle could be going this way and could be going that way, it still has a state where it could be going this way and that way at the same time. And that's weird in itself, but that's not the quantum measurement problem. The microscopic world is allowed to be weird. What we regard as weird is a
00:03:38
Speaker
a consequence of our sort of evolutionary and cultural history, and that didn't involve engaging with subatomic particles much. So we could live with that weirdness if it kept itself in the middle of microscopic. But if you take the equations of quantum mechanics literally, what they say is that when you try to measure microscopic systems, that indefinitelyness, that two things at the same timeness gets scaled up to the macroscopic human scale.
00:04:07
Speaker
So if I've got a particle that's here and there at the same time, and I measure where it is, then my measurement device goes into a measure here state, and I measure their state at the same time, according to the equations of quantum mechanics. And to use the famous, if slightly ethical, dubious illustration of this due to Schrodinger, if the way I register where the particle is, is if the particle is here, I kill my cat. And if it's there, I leave my cat alive.
00:04:34
Speaker
beneath the particles here and there at the same time after the experiment, my cat is alive and dead at the same time. And that's weird. And that's not the kind of weirdness that we think we can brush out of the carpet by waving our hands at the culturally evolutionary origins of what we find weird. That's something that just seems to be a flat contradiction to what we observe in the world. And this is a really successful theory. Really successful theory is not in the business of making routine predictions that flatly contradict the world.
Debating Changes in Physics vs. Philosophy
00:05:01
Speaker
And so what at least in textbook discussions gets done in quantum mechanics is we do a kind of ad hoc fiddle to the theory when we apply it to the context of measurement or more generally to the context of scaling microscopic stuff up to the human scale. And what the fix is, is when we have a state according to the maths in the theory that says the system's doing this thing and that thing at the same time,
00:05:32
Speaker
we reinterpret it and say, no, what we really mean is it's either doing this thing or it's doing that thing and we don't know what, and there's some probability of it doing this thing and there's some probability of it doing that thing. So instead of saying that the cat is alive and dead at the same time, we say, well, the cat is one of alive or dead, but the theory doesn't predict which, it just gives us some probabilities as to which occurs. So what that amounts to is a sort of fundamental
00:05:58
Speaker
shift in how you think about the theory that happens when you apply it to the context of measuring things. I should say, incidentally, you might ask yourself, well, couldn't we have that sort of ignorance reading of this indefinitelyness, even at the level of particles that come, couldn't we say, well, when we said that the subatomic particle was here and there at the same time, couldn't we just have meant it might be here and it might be there and we don't know which? And the short answer is a phenomenon called interference, which kind of allows the future of particles to be
00:06:28
Speaker
affected by all the various places it might have been earlier. The phenomenon of interference makes it really, really hard. Most philosophers would say impossible to take that sort of, it's all about probability, it's all about ignorance, reading of quantum mechanics, and take it all the way down to the microscopic. So there's this fundamental change in how we think about the theory between the microscopic and the macroscopic. And that's the quantum measurement problem. And the reason it's normally thought of as a problem is
00:06:56
Speaker
despite our species' delusions of grandeur, we normally think humans are made of atoms, they're just more physical systems. The way humans behave should be governed by the laws of physics that apply to the bits humans are made of. So it kind of doesn't make coherent sense for there to be new laws for how things happen on human scales that aren't a consequence of the playing out of
00:07:24
Speaker
the laws that govern the constituents of humans. And just generally that kind of, you might call it a sort of soft reductionist way of thinking about the world is really ubiquitous in science and in general really, really successful. And its failure in quantum mechanics would be, to put it mildly, something of a shock. Right. It's a pretty major issue. It's scratching. Yeah. And another way of putting it is measurement is just a hopelessly vague word.
00:07:54
Speaker
I mean, if we had a law of physics that said particles behave in such and such a way, except when the aggregate mass of systems exceeds 2.07996 times 10 to the minus 11 kilograms or something. That's a sharp law. That's the kind of thing that physicists can go with. But things obey the laws of quantum mechanics until a human looks at them. That's a weird thing to put in a physics textbook.
00:08:17
Speaker
Indeed. So we have this itch. How do we scratch it? And I sort of, I know that your answer will be, it doesn't really exist. It's sort of a fountain itch, but maybe we start with the attempts to actually, you know, solve the problem as it were. Sure. Okay. So the paradox in the way I just presented it is something you get to if you, apparently at least, if you combine our best physics,
00:08:46
Speaker
with some very widely held attitudes to how we should think about scientific theories. What I mean by our best physics is just quantum mechanics and the equations that seem to govern the subatomic world. What I mean by very standard ideas about what scientific theories are, are things like scientific theories aren't supposed to reach out and make direct claims about human observers. Scientific theories are supposed to be in the business of describing and explaining what the world is like,
00:09:17
Speaker
scientific theories are not just supposed to be sort of brute calculational devices who can't be understood beyond letting us work out how to build transistors and nuclear weapons and what have you. And scientific theories are supposed to give you a description of the world. That's third personal that makes sense, even in the absence of humans, isn't irreducibly from the perspective of a human. So
00:09:42
Speaker
So those are sort of standard positions in the philosophy of science that explicitly or tacitly most scientists hold. So the way I sometimes look at it is, if you've been led to the measurement problem by this combination of the physics and philosophy, then at some point you've got to either change the physics or you change the philosophy. And so, by the way, as you alluded to, I don't ultimately think that's right, but it's a good starting point for analyzing it. So if I change the physics, I mean something like change the equations.
00:10:11
Speaker
So I said that a minute ago, if our laws said something like everything proceeds according to the equations, except when the mass exceeds a certain sharp threshold and then there are different laws, that would be an example. If you fleshed out that proposal, that would be an example of trying to change the laws in a way that made the
00:10:30
Speaker
made the theory avoid this measurement problem that would make the shift in the way macroscopic stuff behaves explainable in a sort of sharp, mathematically clean way. You could also imagine changing the equations by adding new stuff to them. Maybe the stuff the equations describe isn't the stuff that observed physical reality is made of. Maybe we need to add some more stuff to the equations to describe the observed physical reality. Those are change the physics strategies.
00:11:00
Speaker
problem with trying to change the physics is that the physics works really very, very well indeed. Quantum mechanics is, by a country mile, our most precisely and thoroughly tested scientific theory ever. I mean, it's probably up there with the theory of evolution as our candidate for best ever theory, with apologies to Einstein. So
00:11:26
Speaker
philosophers like to change the physics strategy, but physicists are usually pretty dubious about it. And the other strategy we have is to try to change the way we think about scientific theories. You could say, no, quantum mechanics is teaching us a deep lesson about the nature of science. Quantum mechanics is teaching us that we were wrong to suppose our theories could describe an objective world, or we were wrong not to think that humans play a central part in the nature of reality, or we were wrong to suppose that physical theories could be in the business of describing things,
00:11:57
Speaker
that are accessible without being irreducibly from the perspective of the human observer. And if you look back to the infancy of quantum mechanics, people were taking quite seriously the idea and the founding fathers of quantum mechanics, some of the philosophers of the time, taking seriously the idea that quantum mechanics would presage a general change in our scientific attitudes. That hasn't played out. I don't think philosophers should be complacent about how we think about scientific theory, that our worldviews do change from time to time.
00:12:27
Speaker
but quantum mechanics didn't kind of herald a large scale change in the way we think about physical reality and about scientifically.
Impact of Quantum Mechanics on Scientific Thought
00:12:37
Speaker
So I'm going to hesitate and say that quantum mechanics didn't presage a fundamental change in how we think about the activity of science, the rest of science, whether it's the way particle physicists talk or chemists or geneticists still stick to this kind of science describes
00:12:57
Speaker
what's going on, science is objective, science is third party, science treats humans as just more physical stuff. I mean, medicine is a clear example of this, that core paradigms of medical and biological science treat humans as physical systems. And so from any sort of general how science works perspective, the price of changing the philosophy is pretty extreme. And so you find that while physicists often like the idea of changing the philosophy,
00:13:25
Speaker
philosophers are extremely unkeen. Right, so you have the philosophers want to change the physics, the physicists want to change the philosophy. Exactly. And if you're optimistic, you might say, well, this just shows the importance of overmindedness.
00:13:43
Speaker
If you're a little bit more cynical, you might say that maybe each candidate doesn't fully appreciate the scale of the project they're proposing to carry out. And I want to just briefly touch on this. I guess maybe we should give a little bit more detail on how it works in changing the philosophy. But I suppose, I mean, firstly, just to defend the kind of naturalist
00:14:08
Speaker
view of science, which, as you say, is probably very common now. I'd like to say very common in antiquity as well, and probably very common in the popular imagination that when what science tells us is things about the world and things about the stuff in the world, not just the results of, you know,
00:14:32
Speaker
pointers on dials and, you know, lights, you know, illuminating the results of experiments. So probably that, I would guess that'd be the kind of go-to view of most people who
00:14:48
Speaker
are interested in science who become scientists. And maybe it was just an interesting time when quantum mechanics was being formulated, where actually, I guess the philosophers were themselves also thinking along different lines, there was this logical and persistent view of things where there are certain things we can't speak of, we'll just talk, we'll stick on the very solid ground of pointers on dials and so forth. Yes, absolutely. Yeah, I mean, the
00:15:18
Speaker
A lot of the philosophy, particularly in the sort of Anglo-American tradition in the first half of the 20th century, took very seriously exactly those sort of ideas. It supposed partly as a kind of pushback against what was seen as the extremes of metaphysical speculation. People in that philosophical tradition tried to ground things very tightly in the accessible and the observable. And for them, a scientific theory was effectively a device used
00:15:48
Speaker
a fiction, you might say, in order to make claims about the observable. So if you think about, take a modern example, the LIGO gravity wave detector is described by physicists and by most contemporary philosophers as a device that detects the presence of gravity waves, which would have been there even if the detector wasn't there. From the logical empiricist's point of view, gravity wave is a convenient fiction used to describe what the LIGO detector does.
00:16:18
Speaker
And that position is largely defunct in modern philosophy of science. Sometimes I think it's a slightly bad rap. People misremember their history and start thinking what was wrong with it was just a sort of huge failure of nerve that people forgot that their job was to describe the deep structure of reality and they was to help and just try to say observational stuff. And that's unfair. There were motivated reasons why the positivists, the empiricists went in that direction.
00:16:45
Speaker
But why the program really falls apart, I think, is that to make sense of that picture of science, you have to suppose that the notion of what is observable is something basic and primitive in your science. And that's not what observation is. I mean, we have this kind of pretty picture, if you like, that observation is my visual field and the various colors in it or something. But that's a picture we get from science. That's a picture we get from our detailed scientific picture of resonance.
00:17:14
Speaker
If you ask what counts as an observation to somebody in modern physics, then their answer is going to be extremely technical and use a ton of technical terms that are not in the vocabulary of somebody who's not a physicist. And it's very theory-laden to use a jargon term that philosophers use. You can't really understand what the observations are until you have access to the theory. I mean, is LIGO an observation device? Well, only if you've learned a ton of general relativity and a ton about
00:17:42
Speaker
complicated vibrating interferometry stuff, which you can tell no one's nothing about from a gloss of it. And I think that's the real reason that kind of strategy doesn't ultimately succeed. It's not just that we don't want our scientific theories just to be devices to connect observations, it's that we can't make sense of the actual theories we have in that way. I just have to say, if measurement devices were black boxes that had been scattered across from the Harvey desert by the gods,
00:18:11
Speaker
And quantum mechanics had been a science that developed by trying to work out what these black boxes did. Well, then maybe you can make sense of quantum mechanics as a theory, which is entirely about these primitive measurement devices and observations. But measurement devices are not scattered across the Mahatma desert by the gods. Measurement devices, I'm reliably informed, are made in labs by experimental physicists in conformity to physical principles. And if you ask an experimental physicist,
00:18:40
Speaker
how does your interferometer work they won't say oh well it's a dark mystery measurement is primitive we just have to take measurement as a basic idea they will tell you a very detailed story involving laser optics and exactly how they calibrate their mirrors and things like that. Right and I think as you alluded that that works all the way down you can think about you know magnifying glass or even the retina right it's absolutely yeah.
Wallace's Realist Perspective on Quantum Mechanics
00:19:06
Speaker
Observation is something that we understand downstream of our theories of how observation happens, whether that's human observation, or the much more verified things we see in the sort of heights of physics. So coming back to, I guess, where we are now, you have the philosophers who are dissatisfied with how the physicists are trying to enter quantum mechanics and vice versa. And here it's maybe worth letting listeners know that
00:19:36
Speaker
that you're sort of on the fence in that, or maybe not on the fence, but you're in the unusual position of having done both a doctorate in philosophy, a doctorate in physics, in the reverse order, in fact, I think. So you are presumably pretty unhappy with both of those things. Yeah, exactly. I mean, my view is that the philosophers have really good reasons not to want to change philosophy, and the physicists have really good reasons not to want to change physics.
Conservatism in the Many Worlds Interpretation
00:20:03
Speaker
So we can't change anything. Are we completely stuck or where do we go from here? So as, as of course you realize, there's another tradition which denies the way I set up the premise. And this kind of comes back to that provocative claim, intentionally provocative claim in the beginning of my book that there isn't really a measurable problem after all. And it goes back to the comment I kind of slipped past you right at the beginning that of course the cat can't be alive and dead at the same time.
00:20:34
Speaker
And so, why do we say that? Well, the obvious answer is like, it's intuitively obvious that cats can't be alive and dead at the same time. And that's the intuition they don't tell us much in science. Well, much better answer is it doesn't look as if the cats are alive and dead at the same time. But does it? I mean, if you ask yourself, what would it be like to see a cat that's alive and dead at the same time? And well, I think we have an intuition of that.
00:21:00
Speaker
it would be like seeing a sort of blurry double image of a cat with a life and the cat was dead. Somehow you'd be simultaneously seeing two cats and it would look as if you're seeing doubles, if you'd had like way too much to drink or sort of muzzly waking up from sleep or something. That's what the intuition says it's like to see a cat with a life dead at the same time. But again, intuition isn't a great way to tell what it's like to look for things. And again, looking at things as a physical process governed by physical laws.
00:21:29
Speaker
So we better ask our physics what it would be like to see a cat was alive and dead at the same time. So seeing a cat was alive, that's a physical process. So I've been in America too long. That's a physical process. My brain goes into some kind of live cat registering state. We don't know the details. I will say things like, oh, a live cat. I will report the cat is alive. If I keep a diary, I might not have
00:21:58
Speaker
extremely boring person, I might just make it over. Likewise, if the cat's dead, there's a well understood, at least an outline process by which I registered the death of the cat, like it's a quarter of my brain somewhere. I comment on the cat being dead, I wrinkle on my nose or whatever. So what does quantum mechanics say happens if I observe a cat was alive and dead at the same time? Well, that same process and magnifying we saw earlier plays through.
00:22:26
Speaker
It doesn't, the quantum mechanics absolutely doesn't say I go into some weird cognitive state, which corresponds to seeing double. Quantum mechanics says I go into the ordinary seeing a live cat state and the ordinary seeing a dead cat state at the same time. And that process ratifies. If you ask me whether the cat's alive or dead, you don't hear me babble simultaneously in two different ways. You at the same time, according to the equations,
00:22:54
Speaker
go into a perfectly ordinary state of hearing me tell you the cat's alive and another perfectly ordinary state of hearing me tell you the cat's dead at the same time. And if anyone listens to the podcast, they'll hear, they will at the same time be in a state of hearing themselves be told the cat's alive and hearing the cat's dead at the same time. And if they then tweet about the podcast,
00:23:21
Speaker
their Twitter feed will be in a reporting the cat's alive state and reporting the cat's dead state at the same time. And all of this is correlated. So the world at the same time is in a state where the cat dies, I report the cat's dead, you hear me report that, you pass it on to your listeners, your listeners, tweet about it. And the cat's dead. Whatever the other ones is, there's a live cat
00:23:51
Speaker
way things are and a dead cat way things are and they're happening at the same time and each of those ordinary ways things are is perfectly ordinary but they're simultaneously present and humans are physical systems humans are processing systems a situation where we've got these two parallel goings on at the same time is a situation in effect where we've got two two chunks of physical reality a cat being a live chunk a cat being dead chunk
00:24:21
Speaker
And those chunks don't interact with each other. Each one goes along for all the world as if it was the only goings-on. And I've given a situation where physical reality is simultaneously doing two things, each of which looks like an ordinary set of goings-on, an ordinary world, and the two don't interact with each other. What you have is a many-worlds theory, many-universes theory, a theory where the act of measuring the quantum particle, killing the cat, and so on,
00:24:50
Speaker
causes physical reality to bifurcate, to come into a part where the cat's alive and a part where the cat's dead. And the really crucial thing to get across here, anyone listening, if you just take one thing away from this, take this, is this isn't a new idea that's being glued into the theory. It's not to change the physics move in the way I was describing earlier. We're not saying there is a measurement problem, we will solve the measurement problem.
00:25:20
Speaker
by introducing this idea of parallel universes. It's that when we think deeply about the quantum theory we have, we realize that it was a many-universes theory all along. And that's what I had in mind when I said we can avoid either changing the physics or changing the philosophy. You get the many-worlds theory by holding on to the equations of quantum mechanics, not modified at all, holding on to that picture of
00:25:49
Speaker
a scientific theory as something that gives an objective third-party observation independent description of what what's actually going on in the universe and then just taking those together and thinking very carefully about what the theory is actually describing what it's actually representing about physical systems. So bizarrely that way of thinking about quantum mechanics in one sense is extremely conservative doesn't change
00:26:18
Speaker
doesn't change the philosophy, let us make sense of the theory as we find it in the way we used to think of theories. But of course, in another sense, it's radically not conservative, because it really is saying that every time we see a quantum measurement, then we split into multiple copies, which is clearly has a science fictional sound to it. Yeah, I definitely want to get into the that aspect of it. But I think it's, as you say, this is probably the crux of
00:26:47
Speaker
your book and the main world's interpretation, how it arises very simply, as you say, from just the quantum mechanics. You have a nice quote from DeWitt, I think, where it says, quantum mechanics supplies, you know, the mathematical formalism of quantum mechanics supplies its own interpretation. And we're not adding in, you know, there's no
00:27:15
Speaker
Oh, by the way, apart from the equations, we have to have this concept of measurement. And when you do a measurement, something else happens, which is completely different. And we're not, neither are we saying, oh, there's an extra, there's some extra terms we add to the equations, which explain why as systems grow bigger, you know,
00:27:35
Speaker
similar to your suggestion that there was just some physical limit in size at which suddenly things go from being in super positions to not. And if people are interested in this, there are serious attempts to
00:27:53
Speaker
make, you know, less crude versions of that work. But they do have, as you say, lots of challenges because the physics is just so well established. And, you know, touching anything is, it's very difficult. And it's worth saying that,
00:28:12
Speaker
what I would call toy models that demonstrate the possibility of these strategies in some places, and their advocates would say that toy models are an unfair description. There are models of change to physics that seem to work inside a certain quite narrow descriptive regime. They work quite well if you ignore relativity and if you ignore light and if you ignore electromagnetic effects to a large extent. The challenge is expanding beyond that.
00:28:40
Speaker
to the much richer regime of particle physics, of radiation, of photons, and all of that good stuff. Yeah, indeed. And one way I like to think about these things emerging is, we can't do this visually, but if one imagines an equation, and you've got a particle which is in a superposition, literally you can represent that as
00:29:06
Speaker
a term which says it is in one place. I don't know, it's going through slit A if we talk about a double slit, which is very famous. It's just an experiment where light passes through two slits and single particles. There will be a term saying it's going through slit A. If it goes through slit A, it hits a detector and the poor cat meets its demise. Through slit A, the cat's given a treat and everything's good.
00:29:37
Speaker
So if you kind of, you start with a state which just looks, you know, particle going through a slit A, particle going through slit B. And both of those things.
00:29:50
Speaker
according to the physics are happening at the same time. As we said earlier, it's not just a representation of ignorance and actually there's experiments that you can do that really very strongly show this. These things aren't manipulated according to the pure calculus of probability. They can interfere with each other, those two states and we can
00:30:16
Speaker
we can actually manipulate them in ways that may feel very apparent. And that's where the kind of, if people have heard of the wave particle dualities, it's just saying, well, these particles are behaving like waves in as much as they can interfere with one another or interfere with themselves, indeed. There are different histories or paths that they take into fear with each other. And in terms of things I was saying earlier, that the idea of a particle being a wave is really a
00:30:47
Speaker
a way of saying that the particle is in many, many different places at the same time. The way the description of the particle is a useful description to use. When the particle isn't really in one place or another place, it's at the same time in many, many different places.
00:31:02
Speaker
Yeah. So we have this particle passing through the two slits. And it's in quite a simple state initially. We can think about that particle on its own. But as soon as it starts to engage with the world and it hits the one detector, which is unfortunately releasing some poison gas or something, and the other detector, which we prefer, which is going to supply the cat with some treat,
00:31:31
Speaker
it's entering a much more complicated state. And as that state gets ever more complicated, as you say, as in one part of this equation someone appears and sees what's happened to a cat, something else happens in the other term. And you literally have a plus sign between these two states. But those two states are very much separated by that plus sign.
00:32:01
Speaker
there is a point at which it is possible, as we said, to interfere them back to each other. But by the time the world has become very complicated, if you like, or by the time the effects have propagated outwards, it becomes very, very difficult for those, in practice, impossible for those those paths to
00:32:28
Speaker
I guess touch each other, so have any influence on each other. Yes, exactly. How do we, what's the best way of understanding why that is? Why these interference experiments, effects become harder and harder. Exactly. We might be able to do that at the small scale, but then something happens, which means we don't see anything from the other terms of the equation.
00:32:55
Speaker
Good. So imagine you've got an electron that's in two places at once, then to do an interference experiment that really tests the claims in two places at once, all you need to control is the electron. You need to control the electron pretty well, but it's only one particle. But now suppose, you know, electrons are charged. Let's suppose it's gone, it's gone near another electron that happens to have slipped into your apparatus. Well now, you know, the electrons repel each other like charges of hell.
00:33:21
Speaker
So let's say the electron was here and there at the same time, and let's say that the second electron is close to wherever here is. So now if the first electron is here, the second electron is deflected. If the first electron is there, it doesn't get deflected. So now the second electron is deflected and not deflected at the same time. But the combined state of the electrons is something like first electron here, second electron deflected, plus first electron there, second electron not deflected.
00:33:51
Speaker
So now I've got a quantum superposition of both particles at once, and they're correlated, which is what physicists call entangled. It's not that each particle is separately doing two things at the same time, it's that the two particles together are doing three things at the same time. So now to demonstrate the interference, you need fine control of both electrons. And that's okay too, we can do multi-electron states, but you can now start to see how this can ramify. I mean, if I imagine there's not just one
00:34:20
Speaker
electron that was close to the original electron where there's like a million of them. Now this thing happens with all one million electrons. And now to do the interference experiment, I need fine control of a million particles. And that starts very rapidly getting computationally and experimentally impossible. And so what's, what suppresses these experiments that show interference, what hides interference from us.
00:34:50
Speaker
isn't per se how big a system is. It's how many moving parts the system has that are all independently caught up in the entanglement process. And once we have more than a relatively small number of those particles, the interference experiments become inaccessible. So to give an example of how you can kind of bypass this sometimes,
00:35:13
Speaker
People have done the two-slit experiment you described with electrons and photons, but they've done it with much bigger things. They've done it with buckminsterfullerene molecules, for instance, these kind of carbon-60 buckyball things that have several hundred subatomic constituents. I've lost track of if they've done these experiments yet. I think it's certainly been explored doing these experiments with frozen bacteria. It may even have been done by now. So you put a bacterium in a superposition where the bacterium is here and there at the same time.
00:35:42
Speaker
But the reason these things are doable is that even though the system is quite heavy, it's only really got one relevant moving part, something like where is it in space? It's not as if we've arranged things so that if the buckminsterfullerene particle molecule went along the left path, it gets heated up. And if it goes along the right path, it doesn't get heated up. If something like that happened, then all the internal moving parts of the
00:36:11
Speaker
molecule would get entangled with the position of the molecule. And then to do the interference experiment, you'd have to control all of those moving parts and bring them all into alignment, which would be probably realistically impossible even for the buttons, the fullerene molecule, let alone for a bacterium. And so it's that kind of effect. Physicists call these decoherence, the
00:36:35
Speaker
the redundant recording through entanglement of where the system is in many, many, many bits of the environment the system's in. And once that happens, it basically becomes unrealistic to do interference experiments. And it also becomes unrealistic to stop it keeping happening. So by the time I've got a million electrons in a superposition of here and there, then the next million electrons alone want to get in the act.
00:37:03
Speaker
and then the next million electrons after that want to get in the act. And so once your superposition gets above a certain scale it just becomes uncontrollable and it spreads out on the speed of light and effectively within a fairly short period of time a respectable chunk of the whole world will have recorded whether the electron went one way or another and so now you need to control the whole world at a microscopic level in order to
00:37:32
Speaker
reverse the decoherence and show the interference effect and that becomes basically impossible. I mean, if you've heard, if people have heard of these, these ideas that, you know, heat is just molecular motion. And so in principle, one can imagine taking a warm cup of water that was previously ice and water and has melted and somehow arranging it for it all to go backwards so that the ice spontaneously reforms and the water spontaneously heats up. That kind of thing is theoretically impossible.
00:38:02
Speaker
But in practice can never happen. And it's very similar for the sort of interference at the large scale. It's theoretically possible that you could reverse all the digaherons and all of the environmental entanglement and get back to the pure superposition. But in practice, you're not again. Yeah, I think that's a really useful analogy. And as you say, it's actually a very similar, you know, the mechanism is quite similar in a sense.
00:38:31
Speaker
So yeah, you know, in principle, yes, branches that have significantly diverged could sort of re-merge, but that's just, you know, the vanishingly small probability of happening, so it's almost not worth considering, but usually, you know, on the very small scale, there is a sense in which that is happening all the time, at least within labs, but
00:38:57
Speaker
in many other effects. Yes, exactly. I just want to clarify something that may have been obvious, but in case people didn't get it, I guess when one thinks of
00:39:12
Speaker
The way which we can demonstrate interference or the way we can kind of do these experiments. The reason why it involves having quite precise over control over states is because you want to do things like don't crudely speaking cancel out peaks and troughs and we're talking about
00:39:29
Speaker
noise cancellation and that's the classic example of interference cancelling which many people will probably be using all the time without necessarily knowing about it. You might have noise cancelling headphones where they have a little microphone in them and they are measuring the
00:39:49
Speaker
the sound outside your headphones and creating a sort of inverse version of that. So all the loud noises, all the loud frequencies are kind of cancelled out by an out of phase, but similar frequencies. So you can kind of imagine two waves coming along, touching each other and just becoming a single line. And that's,
00:40:17
Speaker
Well, it is the same mechanism with interference in quantum mechanics. But within quantum mechanics you have, your waves aren't kind of neatly localized in one place, one thing, they're entangled in one place. It's the entanglement that makes the problem become exponentially, literally exponentially more difficult for quantum interference than for ordinary sound. So for ordinary sound, if I've got one source of sound, then I have
00:40:46
Speaker
one wave in three-dimensional space stand, or if I've got 10 sorts of sound, I've got 10 waves in three-dimensional space, and those waves can interfere with each other. If I've got one quantum particle, I can represent it as one wave in three-dimensional space. If I've got 10 quantum particles, I can't represent that as 10 waves in three-dimensional space. I have to represent it as one wave in 30-dimensional space.
00:41:10
Speaker
Right. Because of the interference. I need to keep track of all the various places that all the particles could be. I need three numbers to say where one particle is. And so if I think about my wave as tracking all the various, the way which it could be in lots of places at the same time, then my wave is in three dimensions. But I need 30 numbers to say where 10 particles are. X and Y and Z coordinates for all 10 particles.
00:41:39
Speaker
So now I've got a wave in 30-dimensional space. And while doing the math to cancel waves in 10 waves in 3-dimensional space is a non-trivial challenge, but doing the math to cancel a wave in 30-dimensional space is another thing again. And it just gets worse and worse. I mean, if I have an entangled superposition of all the particles in my body, there's somewhere around 10 to the power of 25 particles in my body.
00:42:09
Speaker
now you're looking at a wave in three times 10 to the power of 25 dimensional space and now it's obviously way beyond our computational capacity, let alone our manipulative capacity. Right, and that's exactly why we're doing this for fairly heavy things, but we're compressing down the dimensions essentially by just, you know, having the C60, the Buckminster ball, the Viking ball in
00:42:37
Speaker
line with all just one degree of freedom, essentially. Exactly. The center of mass of the buckyball is not quantum entangled with any of the other, in physics jargon, internal degrees of freedom of the buckyball. How rapidly one of the electrons in the buckyball is vibrating is not correlated in any way with whether the buckyball is in the first slit or the second slit. And if it was,
00:43:03
Speaker
and then the interference experiments would become automatically more difficult to do. Okay, great. I want to... I'd like to ask the question that you ask as one of your chapter headings in your book, because I think it's really... The question itself is interesting, but I think the answer is more interesting, which is, so how many worlds do we have? We have this branching going on.
00:43:30
Speaker
Can we count the number of worlds? Is it infinite? Or is that not the right kind of way of thinking? The one sentence answer is that it's an indefinite, slightly arbitrary finite number that's definitely larger than the message anything can think of. Right. Definitely more than 10 to the power of 10 to the power of 30, say.
00:43:50
Speaker
So if it comes up in a pub quiz, that's the one to put down. Yeah. One followed by more zeros than you could fit in the universe is a lower limit. What's going on in slightly more detail is the exact way I define worlds is a little bit fuzzy and a little bit pragmatic and arbitrary. So if you think about the CAT experiment, the way I describe that is that this
00:44:16
Speaker
There's two worlds. There's a world in which the cat is alive and a world in which the cat's dead. But it would be better to say there's two lots of worlds. There are live cat worlds and dead cat worlds. And if you zoom in a bit on the live cat world, there's a whole bunch of different ways a cat can be alive. And a whole bunch of other random processes go on that mean that there'll be slight differences between different versions of the living cat. And if you, I mean, I think that perhaps is a relatively vivid example of this. So in Schrodinger's original experiment,
00:44:45
Speaker
he didn't just do a one off measurement, what sort of one on done and kill the cat, you get the wrong answer. He puts the cat in a box. And there's a poison gas file in the box. And there's a Geiger counter in the box. And there's a radioactive source. And when the Geiger counter detects radioactive emission, it knocks out the poison gas and kills the cat. So and then the idea is you wait until the until you until on average, there's a 50% chance of radioactive atom having emitted
00:45:15
Speaker
So now I've got a branch where the cat lives, but I've also got lots of branches where the cat dies because they correspond to it dying at different times. Now, how many branches? Well, we can kind of chunk them, coarse-grain them as we want. We could decide to say, well, let's consider the collection of branches where the cat died between 1,000 seconds and 1,001 seconds after the experiment.
00:45:44
Speaker
And, but there was nothing magic about .001 seconds. We could, we could, we could gradate that more and more finally, and then we'd get more and more worlds. Eventually you'll reach a stage where you're trying to define the world on a timescale so tight that quantum interference effects are coming in between the world where the cat dies at a thousand seconds and the world where the cat dies at a thousand seconds plus 10 to the minus 20 seconds or something.
00:46:14
Speaker
Once you get down to that level, it doesn't really make sense to talk about worlds anymore, because what we mean by worlds are chunks of reality that don't interfere with the other chunks of reality. And at that grain, they have started interfering with each other. So trying to define worlds as more fine grained than that scale isn't a sensible thing to do. It doesn't give you worlds and this sort of macroscopic does its own thing, doesn't interfere with other things way.
00:46:41
Speaker
But there wasn't a magic moment where that was right. It's not as if at 10 to the minus 20 seconds you had interference, but at 1.00001 times 10 to the minus 20 seconds you didn't have interference. You have a kind of blurry, fuzzy space where the world description becomes successively less useful. It goes from being really useful to kind of useful to really noisy, but sort of a bit useful to useless. And it's a bit arbitrary exactly where you want to put that line. So in that sense, the number of worlds
00:47:11
Speaker
is a bit arbitrary, but there's a clear lower limit there. I mean, it's definitely the case that the world's separated by the points where the cat die is separated by at least a microsecond, won't interfere with each other. So you've got at least that many, that that final resolution of the worlds. And I should say this kind of fuzziness, although it looks weird in the case of something that seems as fundamental as how many worlds there are, it's something we're quite used to in
00:47:39
Speaker
the special sciences in our general stories about how high level stuff comes out of low level stuff. I mean, to take a silly example, if you look in, if I look at the sky at the moment, it's cloudy in Pittsburgh right now. There's loads of clouds. If you ask me how many clouds? That's not really a well-defined question because the degree to which I want to break up the clouds into separate clouds rather than treats of things that are whispering together as the same cloud is going to have a level of arbitrageousness. But it's not arbitrageous to say there's lots of clouds.
00:48:10
Speaker
Uh, if you've led a terrible life, um, and in your, on your deathbed, you come to realize it, there's lots of things you regret, uh, exactly how many things you regret. Well, you know, um, that's a bit arbitrary, exactly how do you individuate them, but, but lots of them. Um, so it's, so you determinately regret lots of things, but the number of things you regret isn't determined. And there were more scientific examples of this kind of thing, um, uh, where our higher level concepts are a bit fuzzy, but still extremely useful.
00:48:37
Speaker
Yeah, I think that's exactly where's the boundary of my table. My table definitely is about a meter high, but is it one meter plus a nanometer or one meter minus a nanometer? Well about level, it's a bit arbitrary. Yeah, I love the clouds analogy, by the way. It really helps clear it up. Didn't intend upon that.
00:49:03
Speaker
Yeah. And I think you're right. I mean, we could if we wanted to define some definition of how it is or what a world is. And in some ways, that's well, arbitrary itself, unless you have a very specific interest in a particular description of the world. But I guess, yeah, we don't we don't need to write as
00:49:28
Speaker
Exactly. Recognize the world when we see it, when we find it useful to talk about it, which is, you know, when it's branched off sufficiently from other points so that it's not interfere. There's some fuzziness as to where that happens, but in practical times, it just happens so quickly, right? These things are propagating, facts propagating out of speeds of light that we don't need to worry about it. It's probably more of a worry for
00:49:59
Speaker
It's perhaps like an attack vector, I guess, for people who don't like many worlds interpretation. You know, you've replaced one fuzzy term measurement with like another fuzzy idea of worlds. Yes. I feel like you've done quite a good job at defending that in what you've just said, but I don't know if there's any other things you would add on that. Yeah, I mean, I suppose you might put it this way. I mean, a traditional philosophy attempts to solve the
00:50:27
Speaker
the measurement problem have tried to get rid of the fuzziness. And the way I think about the Many Worlds theory, which I just want to be clear, I haven't talked to all of the history of this, but this is not my idea. It's an idea I've played with and developed, but it's many years, many years older. But what I take to be an important point about the Many Worlds theory is it doesn't get rid of the fuzziness and approximation. It puts it in a place where it's legitimate and harmless to have fuzziness and approximation. It's appropriate for our stories about the relation between the large scale and the small scale.
00:50:56
Speaker
to have a bit of fuzziness and approximation. And we're used to that all the way across science. We're used to the idea that concepts like liquid or conductor or animal or gene are extremely useful and productive and totally robust and real and not in any sense a kind of observer's trick of the night and yet are a little bit approximately defined.
00:51:23
Speaker
in terms of the lower level things going on. So why, you know, when exactly is something liquid? We've got some very robust descriptions of what liquids do, but there are going to be borderline cases and fuzzy places where, you know, I mean, is rock a liquid? I mean, on timescales, that's better than glass a liquid. People talk about that sometimes. And the real answer is it's kind of a little defined because it depends quite what you mean by liquid. But is water a liquid? Yes. Is rock a liquid? No.
00:51:53
Speaker
Are you a liquid? No. The concept is robust. It's useful. There's a well-developed subject to fluid dynamics, but for all that, there's a bit of fuzz and give in the concept of liquid. So all I'm really saying is that kind of fuzz and give is fine in understanding quantum mechanics. And if our understanding of the relation between the macro world where the cats are definitely dead or alive and the quantum mechanical description when they're not has a bit of fuzz and give, that's okay.
00:52:19
Speaker
What's not okay is for this sort of primitive concept of measurement to be actually just written in to the fine structure of our theory. And I guess one way of doing it is that the fuzzing give is fine, provided you've got a non-fuzzing give description of the world at the deeper level. And the full quantum mechanical description, the way I want to talk about it, isn't fuzzy at all.
00:52:40
Speaker
It's a description of a quantum mechanical world that is very alien to our intuitions, but which is perfectly sharply describable, which evolves in its own sharply describable way. It's only if we want to relate that to macroscopic concepts to, you know, special science notions like chairs and tails and cats and dogs that we need to have a story that has a sort of fuzzy aspect to it.
00:53:09
Speaker
And a lot of the problem with, say, the idea that quantum mechanics has a different law of dynamics for measurement is that one sort of doesn't respect that separation. One has fuzz in the very formulation of the microscopic theory. Right. Yeah. I think the title of your book is very well, I'm sure you thought very carefully about this in putting emergence right at the beginning of the title, the emergent multiverse.
00:53:39
Speaker
And it is that emergence, which is doing the work in transporting us from a very precise description of the micro physics. And then out of that, you do get fuzzy structures arising, but one would be crazy to dispute that or try to say that that's not a legitimate thing to happen because, you know, as you say, this is a theory of, you know, you talk about tigers in the book, right?
00:54:07
Speaker
No one denies that at the basis, tigers are obeying the laws of physics, and there is a micro-physics of tigers. But it's not a very useful thing to talk about when you're in the jungle. You're not going to survive too long thinking about that. And tigers themselves are a pretty fuzzy concept. Their cells are changing all the time. There's interactions in the external world, exactly the limits of one.
00:54:37
Speaker
be very hard to define, but you'll certainly know one when you see one. Right, exactly, yes. The tiger-level description is objective and robust even if it's not completely precise. So yeah, so I think, I mean, I feel emergence is doing the work for us there and separating, creating a bridge, I guess, between the fuzz and the precision. Exactly, yes. It's an emergent multiverse, it's not a fundamental multiverse, and you're exactly right, that's the main reasons in the title.
00:55:07
Speaker
The secondary reason is that emergence is sexy, so it sells copies. Very good. That's a good segue because in terms of sexiness, we've just had, you know, everything everywhere all at once. And I can't help wondering if they, there is like a slight allusion to Everett, the originator of this, just in the EV repetitions or the ever repetitions there.
00:55:34
Speaker
Probably coincidence, but who knows? I never even thought of that. That's a lovely thing. I do. Yeah, I'm intrigued about your thoughts on the way that the many worlds interpretation is viewed outside of these physics and communities and doesn't always seem to have particularly accurate representations. Yeah. Yeah.
00:56:04
Speaker
What are your thoughts on that film? I don't know if you've seen it or if, or any other works that you think. Yeah. I haven't actually go around to seeing that particular film, but it's becoming increasingly professionally embarrassing not to have done. So I need to do about this. I've seen plenty of the other bits and pieces. I think the story is going to be complicated. I mean, I think there's, there is genuine flow of these ideas, um, from popular science, from science to popular science to popular culture. And I think, um,
00:56:34
Speaker
It's genuinely true that one of the reasons you see many world's ideas talked about more in popular spaces is they've kind of, well, in the physical science terms, they've come inside the opposite window. They've come inside the range of views that scientists and then science journalists taking the cues from scientists have started taking seriously and seriously enough that one doesn't just laugh the conversation down if it comes up. So go back even maybe 20, 25 years then
00:57:04
Speaker
there would be something seen quite widespread as unserious or inappropriate about talking about parallel universes, many worlds inside physics. I mean, that's a little longer than that, but certainly by the 80s sounds true. And these days, I wouldn't say the many worlds theories, the majority view in quantum mechanics among physicists, but it's very widely known and very widely supported, or at least taken seriously. It's certainly not a fringe.
00:57:35
Speaker
position in physics anymore. And so that legitimates these things being the kind of thing one can write a potentially slightly breathless New York Times or BBC Popular Science article about. And it's the kind of thing you can be asked to give a popular talk about. And then that can be the ideas go for creative professionals who are looking for inspiration. So one
00:57:59
Speaker
particular example I do know is the US show Devs, which was all about quantum information and computing in many worlds. And I do in that particular case, I know because I talked to the showrunner that they'd watched a bunch of sort of semi popular talks on many worlds and Everett by me and other people as part of the prep work for that show. So that's, that's a direct effect to what you see it turning up. And that said,
00:58:24
Speaker
I think there's another slightly more deflationary reason why many worlds are all over the place at the moment, which is just that the Marvel movies have been super popular. And the multiverse idea has been all over comic culture for a very long time. And not really a comic person, but way back into the 20th century. And while some of those ideas perhaps are also inspired by science fiction that's in turn inspired by
00:58:52
Speaker
by physics, I think it's also true that big, complicated science fiction franchises start growing multiverses very naturally just because they start contradicting themselves in various places and then trying to reconcile that. So the comics seem to invent and reinvent the idea of the multiverse as much as a method of managing their continuity snafus and then trying to turn a bug into a feature, as anything else. So that's the boring deflation, or not actually boring, but from a physicist's point of view.
00:59:20
Speaker
the boring deflationary reason why, partially why some of these things are out there. It's sort of downstream of Marvel and Marvel's downstream of the comic continuity culture. But I'd like to think that at least some of it is the scientific contribution. Well, very interesting about this. And that's, yeah, clearly the idea is spreading. I wanted to ask about one other particular work, which I don't know if you've read, which is
00:59:48
Speaker
anxiety is the dizziness of freedom by Ted Chang. No, I've read a bunch. I don't think so anyway. I've read a number of Chang stories, but I don't even call that one by name. If you have read it, it's sort of right in your wheelhouse, I guess. Although, as with all these things, the representation of physics has to kind of bend the rules a little bit. So obviously, you can't just say there's many worlds and
01:00:15
Speaker
Unfortunately, we can't get to them or get any information between them. So I don't want to give away too much of a plot, but as you know, I think 10 characters. I can probably maybe I'm going to try and get across some of the ideas without spoiling it. So it's a in the story, it imagines that characters
01:00:44
Speaker
You can buy a machine, I think it's called a Prism, which I think you press a button or something and there's a quantum event and a branching happens. So you're standing there with your Prism and you have a counterpart who calls Paracels who then goes off with their Prism. And you're able to send limited amount of information between the two worlds after that fact.
01:01:14
Speaker
What I think is quite interesting about the story is it explores some of the, even though the interaction between these characters is very limited, just the knowledge that they're sharing about themselves really changes their understanding of their own lives. So for example, there's cases, there's a case where one character in the past has taken a very rash action,
01:01:44
Speaker
And in trying to understand themselves, they're thinking about, well, what's that thing I did fundamentally part of my character? Is this something caught on me, or is it kind of an aberration and just something that doesn't define who I am? Yes. And so it's a great relevance to these folks that they can talk to one another.
01:02:14
Speaker
Yeah, and I'm interested in your thoughts on whether the many worlds interpretation should help us or change our views on who we are, on morals, on moral actions and so forth. Yeah, discuss. Yeah, I mean, I think maybe indirectly. So I'll give an analogy first. I mean, I think
01:02:41
Speaker
I would expect in the longer term, as and when these ideas are taken to genuinely get wise protraction, I would expect it to have an impact somewhat similar to some of the sort of Renaissance realisations, arguments by people like Bruno that we're not alone in the universe, the stars in the sky and other suns, and there could be people living on this, on some of the worlds around those other suns. And in one sense, that's
01:03:12
Speaker
That doesn't directly affect anything that happens here. I mean, certainly in the 17th century, even now, our ability to learn the detailed facts of what might be happening on these other planets in distant galaxies is close to nil. And doesn't directly affect interest rates and people's sex lives and whatever political campaigns. But there's an indirect way, of course, in which our picture of the world and our sort of philosophy and our conception of ourselves as a civilization
01:03:40
Speaker
changed enormously by those realizations when we have a very different conception of our place in the universe, in some ways a more humbling conception than we had before the Renaissance. And it's telling that the Inquisition only threatened Galileo, they burned Bruno at stake. So I think, I don't think the Many Worlds theory directly affects
01:04:05
Speaker
much of anything about what we should find ethical and how we should live our lives. But I think widespread realisation of the sets of ideas potentially does in the longer term. And I think your example from Chang is a good illustration of this. I think one of the things it does is it gives us a different kind of perspective on risk and probability and happenstance. I mean, so suppose you're in an ordinary situation and you're
01:04:34
Speaker
you've gone to a restaurant, you have very unethically chosen to drive home when you've had way too much to drink. And reflecting on that in the morning, when in fact you got home fine, there was no trouble. You can think, well, I took a lower risk, I got lucky. Because extremely plausibly, because
01:04:59
Speaker
There's boundry plenty of chance events determining what other bits of traffic turn up and maybe even how your reflexes respond in certain situations. There are going to be branches in which your decision to drive drunk had tragic consequences, either for your own people. And arguably, if we were sufficiently wise, we'd recognize that even without the crutch of the Many Worlds theory, and we'd realize that
01:05:25
Speaker
the decision to drive home really drunk was quite badly morally wrong and also extremely unwise and the coincidence that as it happens no harm came of it isn't really salient to assessing the moral and prudential unwisdom of doing it. Maybe if we were sufficiently wise we could realise that anyway but we're often not sufficiently wise and I think the
01:05:54
Speaker
the realization of the fact that other branches are there where other things turn out worse is that helpful in that way. And I think similarly these things may make bad stuff that happens in your life seem a little slightly more philosophic about it. I mean, if, you know, you have a chance diagnosis of some cancer or say something and
01:06:23
Speaker
Is there some comfort in thinking that you just happened to get very unlucky and that there are many people just like you who didn't get the cancer and carry on valuing your projects and loving your loved ones or people who are equivalent to your loved ones? I mean, maybe, I mean, it's not directly going to change how you causally handle the tragedy in your own life, but maybe it gets you really respected on it. I mean, again, again, you see these themes, you know, if it's the philosophy, I mean, there's a long standing, you know,
01:06:52
Speaker
theme almost like I guess of various sorts of sorts of sort of sort of the misconceptions where one's ability to distance oneself from the first person perspective and see things in a broader way is can be helpful. And so in a sense, this is just more of that. But again, maybe it makes it a bit
01:07:10
Speaker
bit sharper and more vivid. So those were my guesses of some of the ways in which these things in the long term potentially influence sort of self-deception enough philosophy. Yeah I think that's the other example that just came to my head was if you take a decision that you believe was right but didn't work out for some kind of contingent reason I can well imagine there's some I mean I think there is some extra comfort in thinking
01:07:34
Speaker
Well maybe i'm just in in that kind of technical pollens like a low branch weight. Part of the multiverse and i you know i'm in that bottom you know one percentile that makes the other 99% are possible. Yeah no i mean the thing makes it slightly difficult of course is that often we don't know what the probabilities are and so the very fact that the bad consequence occurs
01:07:58
Speaker
should increase our evidence that the probabilities were quite bad. That's true. So that does complicate it. But even so, there are places where you can feel something happened that was circumstances beyond your control. You can think, well, that's just how the probabilities happen to come out. Yeah. Interesting. I agree entirely with your assessment that this probably shouldn't change any of our moral decisions or
01:08:28
Speaker
the amount of thought that we put into our actions, but it does, it does maybe throw that into sharper relief. And perhaps, you know, that's a good way to, you know, buy, I'm going to have many sort of counterparts or parasols or something spreading out from who I am. It's almost like a multiplication of consciousness, which is, yeah, quite mind blowing.
01:08:59
Speaker
But and not necessarily directional, as it were, in terms of pointing to doing things differently. But it does add emphasis. Coming back, I guess, to Bruno and so forth, it does make me, you know, it reminds me that one of the one of the kind of initial reactions to people hearing about the world versus, oh, my gosh, that's just like too big.
01:09:30
Speaker
We've been through this before, right? And not with always great consequences. Obviously, and in some way, the analogy holds really well in that, as you say, we're not adding any new stuff, right? We've not changed the physics here. We're not saying, oh, you're going to have faster than light travel in any of these branches. We're just saying there's a lot more of the stuff that we know about already.
01:09:59
Speaker
And that seems very similar to this kind of extension that we saw with Bruno. And, you know, perhaps if we find that the space time that we live in is not bound, but is infinite, and there is sort of a finite matter density smeared throughout it that, you know, we might also see or have more reason to believe that there's just way, way more stuff.
01:10:24
Speaker
than we've previously considered. Yeah, there are 10 to the 11 galaxies in the visible universe. They have kind of all of 10 to the 11 stars in each, most of those stars have planets. I mean, I can, I can quote those numbers. But if you're actually asked to visualize it, my, my mind boggles, it's just dizzying. And now
01:10:45
Speaker
The Everett branches, the numbers are even bigger, but frankly, the numbers are already way too big for a tuition to keep in ordinary astrophysics. Yeah. And I think I want to revisit this just one last time, because I want to really try to deal with all the objections one might have to the Many Worlds interpretation, try to anticipate what people might be thinking. Because this links into this idea of, I guess, falsifiability. And yeah, we can't access these other branches.
01:11:15
Speaker
or these other worlds once they've become worlds. And that's almost the definition of a world with kind of- Yes, exactly. However, I think you put it very succinctly where you say, well, quantum mechanics is the best, the theory which has certainly the most evidence towards it of all theories, perhaps. Quantum mechanics entails
01:11:43
Speaker
many worlds interpretation if we if we don't mess with anything. Therefore, you know, all that evidence going in at the front end is then evidence for the many worlds interpretation. And I do find that quite analogous with, you know, this is just the size of the space time that we believe that we believe in we live in now, where, you know, we can't see
01:12:10
Speaker
we can see those, we can't see all the stuff that we think there is, right? And we can't even in principle, perhaps, see some of it, you know, might make discoveries that just say, okay, it's too large to see, certainly it's too large to see the stuff that is outside our light cone, as it were. Do we believe that that stuff doesn't belong in the theory? I would say that most people don't. Nonetheless, I feel there's a different
01:12:39
Speaker
There's a sense in which people struggle maybe with the learner world's interpretation and think that it fits into a different category somehow. Are they right to do that or is it really a very similar situation to what we see with the space time, let's say? So I don't think they're right, but I certainly think that
01:13:01
Speaker
It's a fault on the right side. I think it's always right to say about scientific theories, which bits of the theories are doing the work? How seriously should we take them? You've done philosophy of science. You know that the past is a graveyard of false theories that people used to think were good theories. And if we simply uncritically believe everything that our current theories say, we're just inevitably going to end up believing some things that get superseded by later theories. So it does make sense to be critical about what are the things we really have evidence for.
01:13:31
Speaker
So what can we say here? I mean, can you test the theory? Well, the theory is unitary quantum mechanics. It's unmodified quantum mechanics. You can, of course, test that. If you violate the secret position principle, then you have falsified quantum mechanics and within the Many Worlds interpretation of quantum mechanics. Can you... It's not just helpful to think about testing theories against other theories. Can you test the Many Worlds theory against the theory that
01:13:58
Speaker
all our theory is, is an interpretive is a device to produce results for calculations, not really, almost by definition, but you can't test, you can't test the theory that fossils are dead dinosaurs against the theory that fossils are a good calculation, the dinosaurs are a good calculation of fiction to let them make predictions about fossils. Those are, those are the same theory with a different philosophical attitude to it. Can you test the Many Worlds theory against any specific proposal for how
01:14:27
Speaker
we might change the physics. If somebody says, well, when things get above certain mass scales, superposition stop happening. Can you test them any way else against that? Yes, you can. And people have made many proposals and parameterized ideas for ways you could modify quantum mechanics to make Schrodinger's cat go away. And many of those theories have been tested and falsified.
01:14:53
Speaker
by evidence, and the evidence has supported many-wells theory. Some of those tests are ongoing. It's always right to look for testable alternatives to our best physics, so we should keep doing those kind of experiments. But we do keep doing them, and they do keep confirming that there doesn't seem to be any evidence for this change of the equations of quantum mechanics. Could there be some change of the equations of quantum mechanics that's so very far from what we can get at empirically that yet somehow suppresses
01:15:23
Speaker
the many-world branching structure, well, yeah, there could be. And maybe there are, it's not difficult to come up with contrived examples of that that we'd never in fact be able to test. But the point is we're now, I think the burden of proof at this point shifts a little. I mean, it would always be a bit suspicious of burden of proof arguments, but we're then saying we've got our theory, our theory has all of this structure. We're considering a modification to the theory for which we have no evidence.
01:15:53
Speaker
that we're led to try making essentially, at least this is the many Wells version of it, that we're led to try to make that change to our theory, largely for reasons of aesthetic preference of having some philosophical discomfort with the fact that our theory without these modifications leads to something we find distasteful. And so we'll make the modifications for that reason. That's not normally a good reason to
01:16:23
Speaker
change scientific theory. And as a strategy, it hasn't generally been conspicuously successful in developing good science. So that's the sense in which I'd say I think taking the many worlds theory seriously thinking those terms using those ideas is good science, and science that you report to accept this government in an appropriate way. Even while can see that you'll never do an experiment that very directly detects another branch in the way that we can directly detect
01:16:54
Speaker
I don't know, planets around Alpha Centauri, I say. But yeah, we could also probably never do a direct text for planets that are around, I don't know, some kind of icon, obviously. Yeah, or even some galaxy a billion parsecs away, where the whole galaxy is just a dot on our visual field, you know, I agree. And look, if somebody wants to say, for that reason, you should be less
01:17:22
Speaker
confident of the existence of the many worlds than around the existence of the planets round out for Centaurae. I think that's what Centaurae has, planets actually, it's a binary. Planets around some reasonably close star. And I think I'm going to say, yeah, that's reasonable. And I guess I probably, I am less confident in the existence of the Everett branches than I am of those planets. I'm very confident there are planets around those distant stars.
01:17:49
Speaker
I'm only moderately confident that the many world theories right ultimately, I'm very confident that it's the best way to read the quantum theory hat. And I think a good reason to think that the properties that leads to many world theories are going to survive in any plausible replacement of quantum mechanics. So I think it's pretty likely that something like the many world theory is true, but wouldn't bet the mortgage on it. Right, right. So it's the best that we have. And it is delivered to us by
01:18:18
Speaker
the best theory we have as well. And the best theory we have is a really, really good theory to be clear. It's not just the best of the bad balance. It's a spectacularly successful theory. Yeah. So I think that's pretty compelling. But yeah, it is interesting how one's intuitions of the fact that this is not something we can reach out and touch in principle and say what we could around those planets does sort of fit into a different fold somehow.
01:18:49
Speaker
Yeah, I think that's fair. I mean, and look, you see, it's not the bits of physics as well. I mean, even before we detected gravity waves directly at the LIGO Observatory, it would have been almost impossible for there not to be gravity waves. There was incredibly good indirect evidence for gravity waves, because by neutron stars orbiting each other, were losing energy, and they were losing energy through radiation. And they were losing energy exactly the rate that was predicted on the assumption they were emitting gravity waves. Very hard to see that, and not think it must be gravity waves.
01:19:19
Speaker
yet there was still something important, even if only psychologically, in actually detecting gravity waves directly. Yeah. Yeah. And I think we do see, I don't know if there were any holdouts on the gravity wave. Probably not. Maybe more with black holes, perhaps. Although again, the event horizon telescope literally photographs them. Now we've kind of closed that off, that debate.
01:19:47
Speaker
It took a while for similar reasons. I guess it's a challenge with the Many Worlds interpretation in there. There is probably nothing more that we can deliver in terms of providing evidence for it other than the evidence for quantum mechanics. Yeah, there are bits of evidence that I think speak to it closest. So suppose that we, as is probably going to be the case sometime in the next 20 or 30 years, suppose we have a working general purpose quantum computer,
01:20:15
Speaker
the way that works is effectively to be able to create more or less arbitrary entangled superpositions and manipulate them in more or less arbitrary ways. It's logically consistent to suppose that we could build a quantum computer that demonstrates the universality of the superposition principle to that degree and yet have some collapse mechanism that kicks in at macroscopic levels, but it starts to become
01:20:42
Speaker
somewhat less plausible, quite, quite deeply relevant aspects of the theory would be tested very deeply by quantum computer. If you want to get a little science fictional, if you build a quantum computer and then ran an AI program on it that was, you know, could pass the Turing test, then you start getting at least somewhere towards being able to control branching and report on branching for not for us, but for beings we can interact with. But now we're
01:21:09
Speaker
Now we're a lot further in the future. And by the time we can build AIs at that level, we'll probably have more serious things to write down. Yeah, if we make it that far. Right. Yeah, I think that's interesting. I mean, I guess so on the one point, you know, building that sort of computer would provide evidence against the competing interpretations, I suppose. And that we wouldn't be saying
01:21:33
Speaker
dynamical, if we're not, supposing we're not seeing this kind of spontaneous collapse or disappearance of superpositions, then OK, yeah. You would constrain the space, it would be a different kind of thing that constrains the space of collapse theories. Well, it's never going to constrain that to zero, but they are genuinely constrainable by experiments and the experiments are getting tighter. And I guess the, you know, the general intelligence AGI point
01:22:03
Speaker
you know, is the suggestion there, then that, yeah, on the other hand, there could actually be a more positive proof of the main goal-sense expectation where we could run some interference. It's hard to imagine how we do this without knowledge of the other world, right? Where we would be able to run an interference experiment and show it as it exists. So I struggled to see how it would, given we'd have knowledge of another branch,
01:22:34
Speaker
more extra it would tell us, we wouldn't be able to extract any new information from it as it were. So here's what you could do in principle. And these are very, obviously, these are very speculative thought experiments. But in principle, I can run a computer program, and I can give it a superposition of inputs. And I can then do some collective
01:23:01
Speaker
action, either person sitting outside the quantum computer can then do some collective action on the quantum computer that effectively involves rolling back the computer program's response to the input and reinterfering the computer program states. And it's fairly easy to construct protocols of that kind, where you would get a different final outcome if the program was genuinely in a superposition than if the program spontaneously collapsed into one another.
01:23:31
Speaker
Yeah, that's, that's, that's fairly straightforward. So then if you could say it, then if you want to send, this is where it becomes a bit transfictional. If you want to say, well, my computer program is sophisticated enough that I'm going to treat it as an intentional agent. Um, I'm going to regard it as well, conscious to use a better word. Um, then I'm going to say, well, the quantum and what I'm now distinguishing as well, was the, was the, was this program genuine in a superposition of inputs or did it actually only experience one input?
01:24:02
Speaker
And now that that genuinely is a testable question. Yeah. Okay, interesting. Yeah. So you would be asking, what did it feel like to be in that state, in a sense? Yeah. And there are questions you could ask would have to pass through a very narrow filter, because you need to get the same answer from the computer program. The computer program is talking to me outside the computer. Yeah. Important that its response doesn't split me.
01:24:32
Speaker
Yeah. It needs to diverge, come back again, give a single answer. Exactly. But you could do something like, did you receive the input? Yes, no. And it sends yes. And you can confirm that it determinedly sends yes. That would be a performable action. Okay. Yeah, interesting. As my input is a superposition of C red and C green, you can ask, did you see a colour? And you should get the answer, yes, whether it's all red or green. And you could, you could throttle the communication channels tightly enough that you
01:25:03
Speaker
that you should get one and you could definitely get that answer so you could do certain rudimentary checks of that kind of the program functioning the way you think it's supposed to be. This is an application of AGI which I hadn't considered before. Actually there was a workshop about this in obviously San Francisco that I went to last November that was actually trying to play with these ideas. It's very speculative. I mean again I think the
01:25:32
Speaker
The point to which we can run an AGI on a quantum computer in any realistic length of time is the point long past the stage at which the world will have been radically transformed in good ways or bad. I don't work in AI, I have no deep insight as to whether the threshold of AGI is 20 years away or 1000 years away.
01:25:56
Speaker
I think the answer is always 20 years. That's definitely the answer you get if you survey a good. It's roughly the same as how long it will take to get fusion. Yeah. It's been stable since the 1950s in both cases. I think we've ended up at the sort of the most speculative point we could get to, which seems like a good place to... Yeah, and I want to emphasize,
01:26:19
Speaker
This is playfully speculating, but the ideas of the core, every interpretation, are non-spective in that sense. It's a quite serious, conservative way of thinking about quantum mechanics.
01:26:35
Speaker
Yeah, I do think that is the take home point and unfortunately everything everywhere, all at once, is not the most accurate representation of the many worlds interpretation. They say all publicity is good publicity. Right.
01:26:54
Speaker
Well, thank you so much, David. This has been a real pleasure. It's been many years since I spoke to you, but I still carry the flame for the other interpretation. This has renewed and added extra fuel and thoughts to that. Thanks for having me on. It's been great.
01:27:30
Speaker
you