Why I am not an Everettian (contd.)

0:00 ...of coming up with a better theory. Again, I think that both points of the dilemma are very problematic. I think there's a research program of finding the satisfactory dimension of collapse theory. There's research programs of finding the satisfactory hidden variable theory. I don't think we've gotten... either one of the research programs have gotten much beyond the point of toy models. I actually don't have a preferred theory, and I think it's right. Now you say, well, no, then you have to accept standards. You haven't gotten better to replace this. Well, it's certainly not true that... No, I don't say that, but what I do say is that that makes Everard worth looking at. If you look at Everard, then there are a whole lot of things you've got to add on. Well, I agree that Everard is worth looking at, and I think Everard is worth it. I think it's definitely worth asking whether there's that eventual problem too. Why I'm an Everettian, I suppose, comes down to the fact that it's only under Everett that I can make sense of monocality without being a solipsist. Well, I actually think that there's, you know, on the dynamical collapse approach, there's... I hope for a truly relativistic collapse model. In fact, Robert Kamulta, I think, has, again, this is still in the toy model stage, but he's recently, I think, last year or something like that, come up with a relativistic version of CRW, which is problematic in a few ways. What's the concern about non-locality from the first point of view? It makes it look like it's the last theory you're going to have to pick out of your preferred relation of physics, science, and engineering. In that very paper that I quoted, the 5.1 and 6.0, it's there on average to come back over. So it's talking about the CRW theory. Non-relativistic theory for non-relativistic quantum mechanics.

2:30 Nevertheless, what Bell points out in that paper is that it's about as barren and barren as it can be because in a certain limit there's a non-relativistic residual of relativity as time can aid, as Bell calls it, time translation. And the non-relativistic ERW theory has that. The non-relativistic ERW theory does not pick up a preferred relation of physics and mathematics. And, um, there's, you know, yeah, so. Is that all right? Is that all right? Yeah, that's all right. Okay, I'll read, I'll read and think. My question concerns the second to last slide. You don't have to put it back up, but just the expression. Credence of E and QM Everett divided by QM Everett, credence of QM Everett. So clearly that's the right way to parse the conditional credence, I agree with that. I have two questions about it. One, that observation doesn't seem to me to be sufficient to claim that the number is low. Even if, of course, credence of QM Everett is very low, it doesn't follow that that ratio is low. The first question is, why can't one make exactly the same point about Qm, never mind Qm-Everett? It seems pretty anti-seemingly unlikely from the point of view of the first 19th century that you've described. And so, doesn't that fact actually make it clear that we're not considering quite the right credences here, and we should instead talk about the credence of Everett, let's say, given what I now believe about quantum mechanics. So, give me quantum mechanics and its predictive success, and now ask me how much credence I have in Everett. Personally, mine is still pretty low, but a lot higher than it would have been without quantum mechanics. Other people's goes up somewhat higher than that, but given the fact that it's proposed as a solution to various interpretive problems, one might think that actually the denominator of your thing is kind of high.

5:00 Okay, good, good, excellent. I'm glad you asked that question. Um, because I wasn't arguing that that quantity was low. I, I, I, um, I, I, um, gave an argument that... I gave a testing argument that that quantity is high and said why do you think that argument failed. And I said absolutely nothing about what that quantity should be, the s of e comma two or whatever. I have no idea what that quantity should be. But why can't Simon say, given quantum mechanics, it's pretty high for me. He can say that. He can say that. And, but, what, here, um, let me go to the other side. If I'm not being censored, this is, um, the, um, credence that person who put in quantum mathematics would do very well. It would be great if it was a skeptical hypothesis. It would be, it would do very well. And hence the credence that person who put in that evidence, um, Atiyah, Witten, would do very well. That's what makes the conditional probability high, because the probability of the hypothesis conditional on the evidence is proportional to the prior probability of the hypothesis, divided by the prior probability of the evidence. So evidence that comes out true with low prior probability is what reduces the probability, if it has a high entity. Now, someone could say, well, I've got the presences according to which this quantity s of u, q, and y is so-so. I just don't have any arguments.

7:30 They gave a very good argument that the Intersquare law of gravitation and that someone who paid close attention to that argument looked at the evidence that he presented and retained skepticism about that would be really not doing justice. You know, would be remiss in giving useful justice to the evidentiary there, but I do have that to open some of my classes. Am I similarly remiss if I don't attach, given all the evidence I know of the quantum mechanics, if I don't attach high degree to that example? We accept quantum mechanics in a sort of usual textbook presentation, which is sort of, you know, as they talk about in the lab, and if we become dissatisfied with that, look for something better, everything seems better. So why would we just accept every given that we've already accepted quantum mechanics? Well that's fine, as long as the acceptance of that every doesn't undermine the reasons we have for accepting quantum mechanics in the first place. And that's what it seems to me that it does, and if it doesn't, I don't understand why not. Because it does involve reconceptualizing those experiments that we took to be evidence for quantum mechanics in the first place. I'm sorry, I think we must go on to day four. If I both am ostentatiously not late, perhaps you can follow up on that. Okay. As a physicist, why do you believe principles? As a physicist, why do I believe the principle principle? The principle principle is... It's not a fact about the world. It is a normative kind.

10:00 So the question is, justification for the principal principle. My answer isn't going to satisfy you. Can I just defer that for a second? I don't think there is something more basic. Now, if you look at the beginning of probability theory, the sort of questions that people asked were, like, what's a fair bet on this particular chance setup? And that presupposed that if you know the chances, You should weight the results of chance set up by the chance values, expectation values. In fact, one way to approach decision theory is to take that as basic. Introduce those values as chance weighted expectation values and then have conditions on rationality that They will then constrain your creed to satisfy the principal's principles. But essentially we're taking something that's really one of the principal's basics. So, that's not going to satisfy you. There's a lot. I'm going to admit to some of the reasons I have a good bunch now. In my case, if you want to say, you want to say, well, if you need to work, how do you like, what I mean by chance is chance is all there in the picture.

12:30 Now, what I like to mean by chance is, it's essentially, it's functionally defined, it's whatever it is that's like the principle. So then the situation is to say, look, we don't have, we don't have the application of the principle, but we're going to have it for, for example, Ignorance is, or stochastic, or anything like that, and so, and what the stationary program would purport to do would be to get in the direction of actually being able to prove adjustment of the principle. Now, I think the response you're going to want to make to that is, okay, what we do have is something, and let's just come to the challenge of that, is the notion of uncertainty. And one of the things we didn't bring up in the talk, and I don't know why, except that it's anonymous, is the idea of I would certainly be in the correct objective attitude at the branching, because I think the notion that you attack is not strong enough, but it's not one bad sign that anything from the Penrose point of view doesn't have any roots, you don't need exactly that notion of parent measure, and I think you're talking very deeply about explanation of why I think you're looking at a serious aesthetic problem in what you mean, why you're looking at the other side of it. But I think for Simon Lee, for instance, and I think on any sensible epistemology of interpretation, it is actually necessary to justify that sort of uncertainty attitude towards branching. If one justifies it, then I think people want to say, ever a chance is now on all fours with other candidates for a given chance. So for the human, I'm just, I don't know if it's certain, but do you think that we have no hope of an uncertainty attitude towards branching? Okay, good, yes, yes, I actually do think that, and that was probably, and that was in the paper, and that was probably not the right, yeah, that was not the right, well, it probably shouldn't have left that, you know, I, obviously, I, you know, I had to pick and choose what to, what to, to, I'm assuming it cost me way over time anyway, but perhaps that wasn't the right thing to do, but, well, okay. And we're going to have to talk more about that, but I don't care, so the objective is uncertainty as you put it, frankly, so here I, you know, Mike and I have never met, and I'm, and if, I know the given statement I am, it's actually, you know, and as you put it, you know,

15:00 What does it mean to have a future self? Well it means to have very certain relation to my present self. So if I'm uncertain about... so what can happen is the wave function is going to branch and I'll have successors on. At least some of those branches. Of course, there'll be cases where there'll be branches where I don't have any successors, and you're going to have to respond to the pathways to those. When I can want to weigh all the probabilities just on the weights where I have successors, they don't need that as a suggestion to us very much. Yeah, an organization with very mental situations doesn't matter very much. It would matter a lot if they fought the Russian War through us, right? We don't want that always to be good to ask of you. Ok, so imagine I've got successors on all the branches. If to be a future self is just to be appropriately related to my present self, and I'm uncertain about which one of those, you know, which one of those successors I'm going to be, doesn't that just simply mean I'm wondering which one of those successors has that, bears that appropriate relation to my present self? I think that I can come up with an answer to the entire thing and that's my answer to the whole question. Well, I mean, that's what it seems, if that's what it is to have a future self, it's just to bear this physical causal relation to my present self. And it's just simply not going to be the case that I am going to be one of those because all of them are going to bear that. Thank you for your attention.

17:30 I think where that comes from is mixing those levels in a different way, but I don't think there's a perfect way to classify those. I think sometime while you're here, we have to book a room for a day, especially if Simon's in town all the time, and Matthew come over. So, isn't it possible? Well, that's kind of following on from this, actually. So if we're asking the question what evidence we would have for quantum mechanics, the immediate question is not what kind of things would we predict for the future, how would we evaluate it, would we look at our current data. But it's a synchronic problem. And there, you could ask, given that this theory is correct, What probability should we assign to us currently making these observations and having this current record in front of these past observations? And then we'll use this difficulty idea to evaluate the conditional probability. The average can do a lot better than the typicality of the function. Everagian can do a lot better than just say, I should assume I'm on a typical branch. You can actually argue that the Everagian should assign frequencies in those branches proportional to the weights of the branches. Then I guess I don't see why there is this gap in probability of statistics. But the question isn't. What should someone convinced of the Everett interpretation believe? The question is, why should someone who is not already convinced of Kahn-Kent on the Everett interpretation believe it? And so even if it can be argued that it is rational for an Everettian agent, Even if it's rational for that person to assign a very high degree to S and hence believes that the average interpretation is well-reported by the evidence,

20:00 And actually that's quite plausible. I haven't gotten to the point where I know a knock-down, drag-out argument from that. But let's, for the sake of argument itself, suppose it is rational for an Everett agent to believe that an Everett interpretation is well supported by the evidence. Why should I believe the Everett interpretation? Because that's how it seems to be. And I have to think, by the way, if there's an additional reason for accepting some kind of principle, like the five-digit-caliber principle, the five-digit-tenancy of all the things about quantum mechanics. So why do we need the five-digit-caliber principle? You know, that means, you know, that's a certain, assume that your results are typical. Well, you know, sometimes you are going to get acycle results, but they're just going to be, um... There are a number of different ways in which you can measure the probability of an academic lecture. The first way is to measure the probability of an academic lecture. The second way is to measure the probability of an academic lecture. The third way is to measure the probability of an academic lecture. The fourth way is to measure the probability of an academic lecture. The fifth way is to measure the probability of an academic lecture. The sixth way is to measure the probability of an academic lecture. The seventh way is to measure the probability of an academic lecture. The inference that I give you to use two-thirds rather than one-half doesn't depend on assuming the result is typical, you don't have to invoke that kind of result because it's automatic when you conditionalize on the evidence that unless you start out really convinced that it was a fair, that I used one-half, You're credence in the two-thirds is going to go way up. And so we don't need any assumes that the results are typical. Yeah, you see the pre-calcate here just as a short term for the more correct version that I laid out before, which is that this is really just green hair. It's not that kind of example. If you look at cosmology and you will have identical observers with the same evidence in different proportions, then you need some kind of principle to make sense of that.

22:30 But what I was going to say is that if you grant the idea that the Everettian could regard the evidence as evidence for that view, then the question of which theory would accept Everett as a model would then... I think that would be the kind of considerations that we can determine whether everybody ought to come over to the MREP and come, or whether right now we don't have enough considerations in the table to determine that. And whether or not the world is simple is an empirical question, and it's simple in some ways and not simple in others, so that's not going to be a criterion. People keep coming back to what should an Everettian believe? And think of other scientific theories. What should a Newtonian believe about gravitation? Well, my colleague Bill Harper, for a number of years, has been arguing that Newton has a fairly sophisticated methodology and that he makes a very good case on that methodology, a very good case for the inverse square law. And on Newton's own methodology, given the evidence we now have, So, a Newtonian, given the evidence we now have, should believe that general relativity is better supported than Newtonian gravity, and it shouldn't be the case that The different theories have methodologies attached to them on which that theory comes out well-supported, so there's an Everett who should believe that Everett is well-supported and, you know, a collapse theory should believe that a collapse theory is well-supported, and that's, you know, that's, you know, that's coming dangerously close to the sort of cartoon version of Coon where you, where you, where you...

25:00 You may accept a theory on a blind leap of faith and then with that theory comes a notion of rationality on which you then interpret all the evidence that is in favor of the theory. If that's what we're stuck with then, we're not the election. Now, we would really rather run out of time. There are several people who want to, but I, I, Anthony Eagle and Harvey Brown are both here for some months, so if I may, I, I just invite Simon, Matthew, and Michael Vickson, who are... I'm not in. I, I, I said all I wanted to say, except when I said typicality, I meant the totality of all leftists, but that's... Okay, well, I, I, I will reserve it for dinner talk. Simon, do you want to make a final remark? If I could make a final remark. I suppose if I could just confirm really what I was saying in the talk is that essentially the pain is knowing how to float and yet retting. There is no way to do this incrementally, this is a large jump of whatever, we can't incrementally work on to improve that, it seems to be the same. I understand that sort of very well, and may I just say with David that I think if you, to throw out subject and uncertainty is a big boat, absolutely, but I can't work my way through ever giving out subject and uncertainty, so if that's the linchpin of your deeply felt views and so forth, that's what we ought to be talking about, and that's not what we're talking about in the afternoon. But if we look at what we did talk about, Now, I have trouble, may I say, identifying precisely what was the bad move that was done at St. Bertie, what precisely went wrong, and the nearest I can get to it is that what went wrong, in your view, is that through the deceptivising of probability in the late 1800s, when people no longer have access to evidence of power,

27:30 Okay, and so just on that very specific point, it does seem to me that the right thing to say is that we never have access to chance to learn, as we are in ourselves, what we always have access to, to learn the truth, and only learn the truth, on any kind of account. I don't really understand what was said about it. Okay. All right. Let me just make some final remarks. I'd like to know where I found the code metaphor, and then see what's the credit. And in fact, since Monday when I circulated the paper, I've added... The current version of the paper now ends up with the statement that they ever, ever again ought to be able to regard all this, you know, hot load cancelling and these outcomes as the latter they used to get to the quantum mechanics that they then throw away. And it's not clear to me that they're up for that anymore. You're absolutely right, we don't have direct evidence of cancelling. Our evidence about cancers is true relative frequencies. That's absolutely right. Nevertheless, I don't have to invoke any particular theory about the physical underpinnings of those cancers in order to use experiments yielding relative frequencies as measurements of cancers. To understand on the evidence interpretation how that distills the case. And I don't. And maybe I missed something. Well, I'm sorry to have cut off some local folks. I'm very sorry about that. But can I just, by way of finishing off, say that as many people in this field know, Abner Shimoni is a wonderful mind and human being and believer in stochastic deduction and in first-world probability.

30:00 Tradition has been most beautifully and wonderfully represented today by your talk. Thank you very much. Well, thank you. Can I just say, okay, having, since Gary said that, I can't, it's just one, it's just quoting Everett. However, it's said on many, on many occasions that just like the gas, the session of the Everett Interpretation expands the building space available to us. I don't think we've seen it ever after that. So he's cheating. I know he's cheating. Thank you for watching.