Episode 99: Critical Rationalism and Solipsism

• Links to this episode: Spotify / Apple Podcasts

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

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

Transcript

[00:00:00]  Peter: Hello out there. This time on the Theory of Anything podcast, Bruce takes a deep dive into solipsism, specifically in the form of the Brain in a Vat thought experiment, Nick Bostrom's Simulation Hypothesis and related ideas. Does the Church Turing Deutsch thesis suggest we could live in a simulation? What does critical rationalism say about these theories? Do these theories make any sense at all, even as something fun to talk about? Along the way, Bruce once again moves me incrementally closer to getting my mind around Frank Tipler's Omega Point Cosmology. I had a great time during this conversation, and I truly hope someone else out there likes it too.

[00:00:56]  Unidentified speaker: Welcome to The Theory of Anything Podcast.

[00:00:58]  Unidentified speaker: Hey, Peter.

[00:00:59]  Unidentified speaker: Hello, Bruce.

[00:00:59]  Peter: How are you today?

[00:01:01]  Unidentified speaker: Good.

[00:01:02]  Bruce: Today, we're going to talk about David Deutsch destroys solipsism.

[00:01:09]  Peter: Yes, sounds good to me.

[00:01:11]  Bruce: So I thought it was interesting that in Fabric of Reality, David Deutsch has two arguments that are in different chapters. And one is actually about solipsism directly. And the other one is more like the simulation hypothesis, although I don't think he ever actually calls it that. And there's enough similarity between these. I'm trying to like pull the threads together, and I feel like something interesting comes out of it. And I think it really strongly relates to... Well, let me put it this way. I think it's a beautiful example of how to apply critical rationalism to metaphysical or philosophical theories. And I think it's much better than the way Karl Popper addresses the same topic.

[00:01:58]  Peter: Now, could we clarify something first? When you say the Simulation Hypothesis, are you talking about the Church Turing-Deutsch thesis, which obviously he endorses? Or are you talking about Nick Bostrom's Simulation Hypothesis?

[00:02:13]  Bruce: Nick Bostrom, yeah. I guess I use the same term for both.

[00:02:17]  Peter: I feel like I've heard you do that. I just wanted to make sure. Because I think I'm sure that most people, when they hear the Simulation Hypothesis, they're thinking about what Elon Musk has endorsed and Nick Bostrom believes, and which actually came from, I believe, The Matrix movie came out before his paper on that. So I don't know if The Matrix is appropriate [wording uncertain] to credit them, but yeah.

[00:02:46]  Unidentified speaker: Yeah.

[00:02:48]  Unidentified speaker: All right.

[00:02:48]  Unidentified speaker: Well, let's talk.

[00:02:50]  Bruce: I actually think that David Deutsch has lots of really super interesting things to say. Like I said, I feel like it illustrates what I've been explaining in past episodes about how to apply critical rationalism properly to philosophical theories. So just as a review, I've come to the conclusion that enough of these episodes are interconnected that I'm going to have to sometimes do quick recaps of past episodes. So here's the quick recap. I talked about the crit-rat loophole. So in a previous podcast, I talked about how Popper in Conjecture and Refutation, Chapter 8, accidentally created a loophole in his epistemology. And I called this the crit-rat loophole, or sometimes I've called it the crit-rat backdoor. All you need to do is declare your favorite theories to be philosophical and, seemingly, if you interpret Popper this way, Popper says you get to ignore all potential refutations and, instead, you get to have a discussion about much iffier criteria, such as, and here's the actual quote from Popper from Chapter 8, Does it solve the problem? This is questions that he believes that you can apply to metaphysical or philosophical theories.

[00:03:59]  Unidentified speaker: Does it solve the problem?

[00:04:01]  Bruce: Does it solve it better than other theories? Has it perhaps merely shifted the problem?

[00:04:05]  Unidentified speaker: Is the solution simple?

[00:04:06]  Unidentified speaker: Is it fruitful?

[00:04:08]  Bruce: Does it perhaps contradict other philosophical theories needed for solving other problems? That's from page 269 in Conjecture and Refutation. Now, I've argued in past episodes that the crit-rat loophole allows bad explanations to masquerade as good philosophical explanations by using the following foolproof trick. First, declare your theory to be philosophical. If someone points out that your theory can be refuted via testing, ad hoc save the theory, for it is always possible to ad hoc save any empirical theory by introducing a new wholly untestable auxiliary theory. If someone points out to you that you just violated Popper's no-ad hoc rule, respond that you already said your theory was only a philosophical theory. Thus, the no-ad hoc rule doesn't apply to your theory. Then point out that your theory is fruitful and simple and how it solves some arbitrarily selected problem. At a minimum, whatever theory you are against can itself be declared a moral problem, because everything can be declared a moral problem, and your theory can be fruitful by solving this moral problem. Literally, any theory can be declared a best theory using this simple formula, and yet you can argue that it's consistent with exactly what Karl Popper said in Conjecture Refutation, Chapter 8. Now, of course, Popper never intended to encourage people to declare their theories philosophical, just so they could get easier standards of criticism. So from Conjecture Refutation, page 48, Popper says, irrefutability is not a virtue of a theory, as people often think, but a vice.

[00:05:53]  Bruce: Popper's demarcation criteria was meant to say, look, don't intentionally make your theories philosophical if you can avoid making them philosophical. So the crit-rat loophole is, I'm arguing, an anti-critical-rationalist viewpoint, or in other words, it's an epistemological mistake. Now, I discussed an alternative approach to how to deal with metaphysical theories using critical rationalism in episode 93, Philosophical Theories vs. Bad Explanations, which I felt was actually one of my most important episodes that had been building up for a while in my own mind, how do I resolve this problem? And I felt like I came up with a pretty good way to do it. Here's a quote from Conjecture and Refutations, page 81, Karl Popper. One can show that the method of science and the method of that [restart wording uncertain] and the history of science becomes understandable in its details if we assume that the aim of science is to get explanatory theories, which are as little ad hoc as possible, emphasis mine. A good theory is not ad hoc while a bad theory is. Now, Popper is specific. What he means by not ad hoc is that it is that it is both independently testable, i.e. the theory does not solely explain the one problem you're working on and that it has actually passed some corroborating tests. I think in the past I've summarized that as simply independently testable, and that's probably technically not correct. I'm implicitly assuming independently testable meant that it had been independently tested. And hopefully people can understand why I would just call that independently testable.

[00:07:28]  Bruce: But technically speaking, a theory could be independently testable, but you've made your test so science fiction-y that there's no realistic chance that it's going to get tested. And in which case, the theory could still be ad hoc according to Popper, okay? So we're kind of looking for actual corroborating tests being passed, which by the way, I pointed out, means that Popper's epistemology does have a concept of positive evidence and a useful role for it, okay? It just isn't what people thought it was. So if you always follow Popper's no ad hoc rule, you must always increase the overall testability of your theories collectively to solve problems because every save of your theory requires that the new auxiliary theory that you're introducing to save your theory to itself have independently testable consequences. I argued that actually the no ad hoc rule that Popper has or as I call it Popper's Ratchet applies just as well to metaphysical theories as to scientific theories. So Popper's Ratchet or the no ad hoc rule is actually the correct epistemological approach to metaphysical and philosophical theories. Essentially, I argued that even metaphysical theories could have testable or I prefer the term checkable maybe because Popper's already using the word testable to mean something else, can have checkable implications. For example, they can be logical implications that we can go check instead. I used Popper's Refutation of Induction in a past episode, that's also in Conjecture and Refutation Chapter 8, as my key example.

[00:09:11]  Bruce: Induction is supposed to be a metaphysical or philosophical theory, just like Popper's epistemology is supposed to be a metaphysical or philosophical theory. Yet Popper wasn't content to merely level iffy kind of criticisms of the kind he raises in Chapter 8. Instead, he argued, he didn't argue, well, my theory solves this problem better than induction. I mean, he does argue that, but he goes so much more deep than that, right? In fact, he, I would dare say that Popper in Chapter 8 of Conjecture Refutation goes for the jugular on induction. And he shows that induction is not logically, is not logically possible at all, right? I mean, like it's, it's an outright, he's showing it can't be done. He's not saying this solves this problem better. He's saying this is impossible. So, and it's a beautiful refutation of induction that he puts together in C&R Chapter 8, okay? So Popper basically treated induction as a universal theory that essentially said something like this. It's impossible, or he treated his refutation of induction, I should say, as essentially a universal theory that essentially said something like this. It's impossible to come up with an algorithm that can start with no starting assumptions at all, or no theories at all, and then produce a generalized theory based only on observations. So he showed that that was logically impossible, so he essentially is creating a universal theory, right?

[00:10:47]  Bruce: Now, by formulating induction in such a specific way, he turned this metaphysical or philosophical theory into something that's really pretty easy to test in principle, if I can be allowed to use the word test to apply to a logical test in this case instead of an experimental test. And what I mean by that is, is that you can refute induction the way he's formulating it by producing an actual algorithm that can produce a generalized theory without needing any sort of starting assumptions or theories. OK, so it should be easy. If Popper's got this wrong, if he's made some sort of mistake in his proof, you can easily tell you should be able to find a counter example to it, and we should be able to easily check that it is a counter example to Popper's argument, and therefore we can show that his argument doesn't work. OK, because his argument is so specific.

[00:11:40]  Peter: I just wanted to interject that I just re-listen to the Unended Quest Intellectual Autobiography, and there's a great chapter on this too, and he is very, very clear that induction died, I killed it, it doesn't make sense.

[00:11:59]  Unidentified speaker: Right.

[00:11:59]  Peter: Yeah, he's very explicit on that.

[00:12:02]  Bruce: So, and then in the same episode, I also showed that the machine learning community, using Tom Mitchell's textbook, tried to come up with such an algorithm, as Popper is saying is impossible to come up with, and they found they could not, and finally concluded it was impossible. So you can think of this as a corroborating example to what Popper is saying. Notice how everything in how Popper treats induction is identical to the way you would treat any scientific theory. It's basically just testability, increased content of the theory, logical content in this case instead of empirical content. But there just is no difference between his refutation of induction and a scientific refutation, okay? And we don't even need to go to the iffier kind of criticisms that he makes because we can go for the jugular on this stuff, okay? Now this version of induction that Popper refutes is not a scientific slash empirical theory, but it is testable or if you prefer checkable in exactly the same sense that Popper's Ratchet and the no ad hoc rule calls for. Now why is this so? Because it's so specific that we know exactly what a counter example to it would look like. Further, I showed that inductivists try to avoid refutation by redefining induction to merely mean you can in some unspecified sense generalize. Now when they do this, they violate the no ad hoc rule or they violate Popper's Ratchet, okay? This is how you know that they are making an epistemological mistake and engaging in a bad explanation.

[00:13:42]  Bruce: The definition of induction at this point, if you by induction you simply mean you can in some unspecified sense generalize by using observations, if that's all you mean, it's just it's too vague, so it's impossible to know what a counter example to it would even look like. It is thus irrefutable in the logical sense rather than the empirical sense. And I called this vague manning your theories. You just make your pet theory so vague, it says nothing specific about anything anymore, and thus can't be tested or checked even logically. In fact, that definition of induction, if this kind of vaguer version of it, is so vague that even Popper's own epistemology now qualifies as induction. Because Popper's epistemology is a means by which you in some sense use observations and you generalize. Sure, it's actually based on conjecturing theories first, you know, and there's way more to it than what induction said, but it fits that really super vague definition of induction.

[00:14:45]  Unidentified speaker: Okay.

[00:14:46]  Bruce: So we might as well go with Popper's epistemology instead, at least until it has a problem that requires us to invoke some other theory.

[00:14:54]  Unidentified speaker: Okay.

[00:14:55]  Bruce: But even if that day comes, there would be no point in invoking this vague version of induction, as it has no explanatory consequences. So it's just a kind of a useless theory. So now let me point out that what I just said, it does fly in the face of Popper's claims that metaphysical theories by definition can't be refuted. In fact, they can be refuted. They just can't be refuted by experiments.

[00:15:21]  Unidentified speaker: Okay.

[00:15:21]  Bruce: And I think this is the right way to read Popper on this. But a good explanation that is metaphysical or philosophical should be able to be logically refuted by offering what a contradiction to it would look like. Now, I'm giving one specific example. And yes, I know it's more complicated than that. It might be a contradiction to another theory. And Popper mentions that as one of his possible criteria. So when I say contradiction, it wouldn't necessarily be a counter example, but it would be a contradiction of some sort, would be an appropriate criticism of philosophical theory. So now in episode 92, Popper on Philosophical Theories, I offered a series of criticisms of Popper's treatment of metaphysical and philosophical theories. In 93, I pointed out that Popper, as I just repeated, didn't hold to his own approach to metaphysical theories when dealing with induction. And instead he decided to just outright refute it logically.

[00:16:19]  Unidentified speaker: Okay.

[00:16:20]  Bruce: Somewhere in there, I think it was episode 92, I mentioned in passing that Popper's treatment of solipsism, what Popper calls idealism, in chapter 8, was inferior to Deutsch's treatment of the same subject, and that I wanted to do a podcast on just that topic. That's what this today's podcast is, is making good on that off-the-cuff comment that I made.

[00:16:45]  Peter: Well, I'm excited, and I've got some questions on this that are burning a hole in my brain right now, but I'll wait for an appropriate time to ask them.

[00:16:54]  Unidentified speaker: Okay.

[00:16:55]  Bruce: Let me just say, I think Deutsch's treatment of solipsism demonstrates the correct Popper's ratchet, or if you prefer, application of the no ad hoc rule, style approach to metaphysical theories using critical rationalism. I think it's so perfect. It's like, this is it. This is what I had in mind. It's not a coincidence this is what I had in mind. I figured this out by reading David Deutsch's books. I learned all this by reading either Popper or David Deutsch. So let's talk about solipsism. So solipsism, it's the theory that, if you're not familiar with the terms, if you've read David Deutsch's books, you know what solipsism is, okay? But it's the theory that actually the whole world is just your dream, okay? So solipsism is today debated in philosophy school like it's a serious explanation that needs to be addressed. And the brain in the vat thought experiment is often invoked. In fact, the two are often kind of treated as if they're one and the same, okay? Now suppose you were actually a brain in the vat where the nerve impulses were being given to you to make you think you were seeing the real world. And as you thought about what you wanted your body to do, the computer that's tracking these impulses updated what you saw and what you felt so that you felt like you were moving your real body, okay? Given that your actual brain is just receiving your actual brain in real life is just receiving electrical impulses through nerves and doesn't directly interact with the real world in any way.

[00:18:34]  Bruce: And in fact, as David Deutsch puts this, it creates a sort of virtual reality representation of the real world. If you received the very same electrical impulses when you're actually a brain in the vat, it would be indistinguishable in principle from what you see and experience in real life, okay? So Deutsch in Fabric of Reality acknowledges this as a possibility and refers to this as virtual reality. In fact, he has a picture of something exactly equivalent to the brain in the vat, experiment on page 112 of Fabric of Reality, where he shows this computer connecting into somebody's brain, the person's been put unconscious, and so they can only see what the computer is feeding into the brain signals into the brain, and it's basically the brain in the vat experiment. Only Deutsch is using it to explain how virtual reality might work in principle, okay?

[00:19:25]  Unidentified speaker: So, he's not...

[00:19:27]  Bruce: Deutsch isn't like saying the brain in the vat experiment is some sort of impossibility. On the contrary, he's saying it's entirely a possibility, okay? So, there's a modernized version of solipsism that is popular today, slightly different, but similar, particularly among techie Silicon Valley types that's called the Simulation Hypothesis. As you pointed out, this is... Is Nick Bostrom either created it or is at least the one who popularized it? Now, I've seen actual serious, you know, quote unquote serious scientific papers written trying to determine if we actually live in a simulation or not. So there's like real scientists out there who take this at least seriously enough that they're publishing about the topic, okay? And there's a popular argument that gets used by some very smart Silicon Valley types, Nick Bostrom included, that goes something like this. Someday we're going to create simulations so realistic that we'll model reality nearly perfectly. The people living in that simulation will eventually outnumber those of us in reality here. Those simulated people will eventually make their own simulation. Well, sometimes they don't include this part, but sometimes they do. Those simulated people will eventually make their own simulations. Eventually most people will actually live in simulations. Therefore, it's rational to assume that we currently live in a simulation. Now, the simulation hypothesis is admittedly a bit different than solipsism. So, technically solipsism is that there's only one mind and everything else is just that mind's dream.

[00:21:05]  Bruce: By comparison, the simulation hypothesis makes your mind as unreal as every other mind. So, you aren't in a privileged position in the simulation hypothesis. However, other than that may be one completely inconsequential change. Both of these theories have much in common. And both are bad explanations for exactly the same reasons. So, for the purposes of this podcast, even though I know solipsism and the simulation hypothesis aren't quite the same theory, I'm going to treat them as more or less interchangeable for our purposes, okay?

[00:21:40]  Peter: Can I say a few things about that?

[00:21:42]  Unidentified speaker: Yeah.

[00:21:45]  Peter: Well, okay, so, I know that probably what you're going to say very justifiably is that the simulation hypothesis, as Nick Bostrom formulated it, violates every rule of critical rationalism that exists. It's unfalsifiable. It's not, you know, it's not an empirical theory. It's a, I get it. It's a bad argument. But at the same time, like on another level, I cannot completely dismiss it. It makes, you know, David Deutsch, okay, Church Turing, tell me if you disagree with this analysis. Okay, so the Church Turing-Deutsch thesis indicates that reality is computational, I guess. Not a computation, but it's computational, which I guess is another way of saying that it follows laws of physics, and that it's algorithmic. I have like 10 questions about that right there, but let's put those aside. So reality is not a computation, but it's a computational. I guess that no one really knows what reality is. I think that probably Deutsch would agree with that. But like, is it really such a bad logical leap to suggest that if reality is computational, that we're not living in some kind of artificially created computational reality, that our descendants created? I mean, I'm not saying that it's like true, but necessarily, or there's even really good reasons, empirical reasons for believing that other than, I mean, it's not like a terrible logical leap, is it?

[00:23:54]  Bruce: So, it's interesting that you raised that. I'm actually really glad you did. You raised it a bit early, because I actually raised that question myself, and then in future slides.

[00:24:03]  Peter: Okay, I know, I know.

[00:24:05]  Unidentified speaker: I'm jumping ahead.

[00:24:07]  Peter: I had a feeling I was doing that, but I couldn't help it.

[00:24:09]  Bruce: So, I'm gonna hold that question for a moment. Let me just capture what you're saying, though, because I feel like you're saying something that's meaningful, that needs to be brought out. I think I could say it in a couple of ways. One is that it really is impossible to refute the Simulation Hypothesis.

[00:24:29]  Unidentified speaker: Sure.

[00:24:30]  Bruce: So, if it is, can you really treat it as if it's refuted, if it can't be refuted? In other words, can you ignore it because it's false, just because you couldn't refute it?

[00:24:43]  Peter: I get that on one level, but it's still like, I don't know, it still has some resonance for me. I can't ignore it.

[00:24:49]  Unidentified speaker: Well, no.

[00:24:49]  Bruce: So, you're acting as if I'm saying we should treat it as if it's false because it can't be refuted.

[00:24:55]  Unidentified speaker: Fair enough.

[00:24:55]  Bruce: I'm asking, should we treat it as false just because it can't be refuted, or is that a different category? Is there a category of refuted and then a category of irrefutable, and those are both different kinds of problems?

[00:25:09]  Unidentified speaker: Okay.

[00:25:10]  Bruce: And I'm actually going to argue that, is that there are different kinds of problems that have to be treated in different ways. And then, in fact, one of the implications of this is that you should not rule out the possibility that we live in a simulation. You just don't take it very seriously. Fair enough.

[00:25:31]  Peter: And I know, as Deutsch has pointed out, it gets into this idea that, oh, well, maybe do you live in a simulation within a simulation within a simulation, and then you kind of...

[00:25:42]  Unidentified speaker: Yeah.

[00:25:43]  Peter: It's a good way of illustrating how stupid it is, in a way.

[00:25:45]  Bruce: So, well, hang on to what actually... I really feel like what Deutsch says on this is just so spot on that it really does answer your questions in a fairly compelling way.

[00:25:57]  Unidentified speaker: Okay.

[00:25:57]  Bruce: So, now, you may not have realized... I mean, I know you've read Fabric of Reality probably multiple times, but you may not have realized how Deutsch answers this. And I think part of the problem is, is that he does it in two different chapters and they're far enough apart. I'm now going to pull them together in a way that you've probably not seen before.

[00:26:13]  Unidentified speaker: Okay, good.

[00:26:15]  Bruce: So, now, what Deutsch calls solipsism, Popper calls idealism. Now, here's an actual quote from Popper in Conjecture and Refutation, Chapter 8. Another philosophical theory is idealism. We may perhaps express it here by the following thesis. The empirical world is my ideas or the world is my dream. Notice that that is precisely what Deutsch calls solipsism.

[00:26:39]  Unidentified speaker: Okay.

[00:26:40]  Bruce: So, just keep that in mind. When I'm reading Popper, he's going to call it idealism, but he means exactly what Deutsch means by solipsism.

[00:26:49]  Unidentified speaker: Okay.

[00:26:50]  Bruce: So, this is good because this is a case where Popper and Deutsch both made arguments against the same bad metaphysical explanation. And I'm going to argue that Popper's criticisms are problematic, whereas Deutsch's are intellectually and epistemologically full on, a full on takedown of solipsism that really does show it's just a bad explanation.

[00:27:14]  Unidentified speaker: Okay.

[00:27:15]  Bruce: And in my opinion, Deutsch does that by adhering to critical rationalism much better than Popper does. So, okay, so here we go. Popper criticizes solipsism like this. Popper emphasized that we can criticize a philosophical theory by looking at the problems it attempts to solve and showing some other theory solves it better. So based on that, that was that quote I gave earlier. He now says, page 269, Let me refer to a specific example. The idealism, remember that solipsism, of Berkeley or Hume, which I have replaced by the simplified formula, the world is my dream. So no doubt that's what he's talking about. Popper points out that these idealists weren't very happy about adopting idealism, but they felt forged [word uncertain] to it by a false theory that they happened to hold. Quote, page 269. Now, if we try to understand the problem situation which induced them to propound this theory, we find that they believe that all our knowledge was reducible to sense impressions. Also on page 269, Hume was an idealist only because he failed in his attempt to reduce realism to sense impressions. Also on page 269, it is therefore perfectly reasonable to criticize Hume's idealism by pointing out that his sensualistic theory of knowledge and of learning was in any case inadequate, and that there are less inadequate theories of learning which have no unwanted idealistic consequences.

[00:28:45]  Bruce: So put simply, Popper is arguing that Hume and Berkeley didn't want to be solipsists or rather idealists, but because they believed in the false theory of empiricism, that knowledge comes through sense impressions, they felt they had no choice. But empiricism is itself a false theory. Had they lived in a time where Popper's alternative theory of knowledge, via conjecture or refutation, existed, he believes that Hume and Berkeley would not have adopted idealism as a theory. So Popper offers this exercise of how to go about criticizing philosophical theory when it can't be refuted.

[00:29:22]  Unidentified speaker: Okay.

[00:29:24]  Bruce: Now, let me say that I agree with Popper that idealism and solipsism is a bad explanation. And I believe Popper came to the right conclusion here. And I don't really disagree with his argument. Like, I agree that his epistemology solves this problem better than empiricism. And probably Hume and Berkeley were only sucked into a bad explanation because they couldn't figure out how to get around the false theory of empiricism.

[00:29:56]  Unidentified speaker: Okay.

[00:29:57]  Bruce: But as I pointed out in episode 92, Popper also used this identical argument to explain why determinism was wrong. It's the same argument.

[00:30:10]  Unidentified speaker: Okay.

[00:30:11]  Unidentified speaker: So determinism, so here's the problem.

[00:30:13]  Bruce: Determinism is actually a correct theory, by which I mean it's the best theory with zero comparable competitors. Determinism is such a good theory that according to computational theory, even if we lived in a stochastic universe, we could still easily assume the universe was deterministic, and nothing would change because a deterministic universe has pseudo-random numbers and they're identical purpose-wise to any algorithm that uses stochasticity. So throwing stochasticity in there does nothing, and that's why the Turing machine doesn't require a randomizer to be able to be a universal computer. Okay, so Popper's argument against solipsism, even though he got the right answer, really does strike me as problematic because his whole argument can be just as easily used on a correct theory, was used on a correct theory by Popper, and he got the wrong answer. This explains why even though I agree with Popper here at some level, surely empiricism is wrong. Surely Hume and Berkeley would have changed their minds if they had realized that. I find his overall argument less than impressive. Now I lay that out as one example. This is the Popper using the CNR Chapter 8 approach to criticizing philosophical theories, and I'm trying to show that it's inadequate in some way. It's not wrong. All the questions he comes up with and the criticisms he uses strike me as valid and may have even worked in this particular case, right? But there's something just inadequate about it compared to what we're about to see come out of Deutsch attacking the same theory.

[00:31:57]  Unidentified speaker: Okay.

[00:31:59]  Bruce: So if Popper's approach to metaphysical theories like solipsism, as outlined in CNR 8, is problematic, what's the correct epistemological approach to such a theory? My answer was that there is no substantial difference, and this is my answer back in episode 93, that there is no substantial difference between scientific and metaphysical theories under critical rationalism. Both are cases of applying Popper's ratchet using my slightly modified version of the no-ad hoc rule that allows us to apply it to logical cases instead of empirical cases. I mentioned how philosophers often considered the brain in the vat argument for solipsism as correct and even a knockdown argument. I'm going to argue this is precisely because most philosophers don't really understand good epistemology, and if they did, they'd not feel this way. In fact, the brain in the vat argument is a terrible argument on nearly every level. It's one of those arguments that a layperson immediately laughs at as being a bad explanation, and you have to have special philosophical training to treat it as a serious question. This is an example of how folk epistemology of a layperson is often superior to the epistemology of maybe a trained philosopher. So let me first ask a question, and this is really kind of the beginning now of me explaining Deutsch's argument. I'm going to end up massively quoting Deutsch. In fact, a huge part of this is going to be just reading from Fabric of Reality. But let me kind of lay out some things that I think help set the stage well.

[00:33:32]  Bruce: Is solipsism, or the simulation hypothesis, whichever you prefer, is it actually a metaphysical theory? I want you to really stop and think about that for a second.

[00:33:43]  Unidentified speaker: Okay.

[00:33:44]  Bruce: Now the reason why I asked that, so if I were to ask you that, is solipsism a metaphysical theory, and you didn't know I'm about to try to argue that it isn't, what would you say? What's your intuition say about simulation hypothesis or solipsism?

[00:34:01]  Peter: Well, I will say that, I guess, I don't know if this answers the question, but to me, the brain in the VAT experiment feels more like a thought experiment, kind of an abstract thought experiment that may or may not apply to our reality, whereas the simulation hypothesis seems to me something a little bit more serious, I guess.

[00:34:30]  Bruce: It's definitely taken seriously, right? Yeah, like probably nobody takes solipsism seriously. That's probably true.

[00:34:36]  Peter: Well, yeah, in the brain in the VAT style solipsism.

[00:34:41]  Unidentified speaker: I mean, I don't know.

[00:34:42]  Peter: I mean, maybe you could also argue there might not be functionally much different now that I think of it, between a brain in a VAT and in a simulation hypothesis. So I guess they are kind of very similar, aren't they?

[00:34:56]  Bruce: They're very similar, but they're really kind of the same thing.

[00:35:00]  Unidentified speaker: Yeah.

[00:35:00]  Unidentified speaker: Okay.

[00:35:01]  Peter: Fair enough.

[00:35:01]  Unidentified speaker: Fair enough.

[00:35:02]  Peter: But so is it a metaphysical theory? Well, I mean, it's not an empirical theory.

[00:35:11]  Bruce: So, okay, so I'm going to argue that stating solipsism is not an empirical theory is not a technically correct statement, although I understand what you say, what you mean and what you mean is actually correct.

[00:35:29]  Unidentified speaker: Okay.

[00:35:30]  Bruce: But I'm going to try to get you to tease this out a little better because I feel like this is where we need to start.

[00:35:34]  Unidentified speaker: Okay.

[00:35:35]  Bruce: So let's just stick with the Brain in the Vat experiment or thought experiment.

[00:35:39]  Unidentified speaker: Okay.

[00:35:40]  Bruce: Testing that you are a Brain in the Vat seems like it would be trivially easy if it were happening today in real life. For one thing, there is no computer in existence that has enough power to keep up with your thoughts. So if you were in existence today, so if you were a Brain in the Vat, you'd easily work out ways to prove you were. Think via experiments, right? Think about this in terms of, say, me putting your Brain in the Vat and then making your reality be Skyrim.

[00:36:10]  Unidentified speaker: Okay.

[00:36:11]  Bruce: I have VR Skyrim, by the way, which is awesome. The data fidelity of Skyrim compared to the real world is enormously different, right? Almost every aspect of Skyrim shows immediate signs of not being reality. Even just little things like the feeling of a jump, gravity, or the responsiveness of your neurons trying to look around or move your arms or something to how long it actually takes for the video game to catch up.

[00:36:39]  Unidentified speaker: Okay.

[00:36:39]  Bruce: You could easily come up with simple experiments that would empirically show you were a Brain in the Vat.

[00:36:46]  Peter: Well, yeah, but isn't the Brain like a more like steel manned version of the Brain in the Vat that we would be that we're not in a literal like wetware Brain in a Vat, but we're something more like running on a computer, presumably a quantum computer. I mean, don't they say that like just a microgram of quantum computer could simulate all of reality or something like that?

[00:37:16]  Bruce: So let's make this a little more specific then, okay? Because you're raising a valid point, but I want to be explicit about it.

[00:37:23]  Unidentified speaker: Okay.

[00:37:24]  Bruce: That when we talk about the Brain in the Vat experiment, we absolutely do not mean an actual Brain in the Vat today with modern technology. We absolutely don't mean that.

[00:37:34]  Unidentified speaker: Okay.

[00:37:34]  Bruce: Because if we did mean that, it would be trivially easy. It would be an empirical theory, and we would be trivially easy to come with an experiment to falsify it, okay? That say, oh, look, I'm actually in a simulation.

[00:37:46]  Unidentified speaker: Okay.

[00:37:46]  Unidentified speaker: Yeah.

[00:37:48]  Bruce: Now, this, what I just gave, the example I just gave, don't even include things like simple experiments, like say noticing if you get too close to textures and the wall pixelates, or because they're made of pixels, or finding floating trees that were designed wrong, which does happen in Skyrim, or more complex experiments, like deciding to build a microscope and finding that you can't, or finding that atoms in this world might be an entire wall with a texture on it, because there's no smaller components that make it up. For that matter, just the fact that the world can't keep entirely consistent and has to cheat and update while you're fast traveling and things like that, because it's not really a full-blown world that's autonomous. For that matter, the rest of the people in the world are so clearly not creative beings. So in short, coming up with an experiment to prove that you're in a simulation or brain in the vat, at least if it were to happen today, would be an empirical theory and it would be easy to falsify.

[00:38:52]  Unidentified speaker: Okay.

[00:38:52]  Peter: Isn't part of the idea that reality can be simulated, the idea that what's going on with these subatomic particles really is not that complicated and that it could be simulated given enough computing power down to every last subatomic particle that that is potentially could potentially run on a on some kind of computer in the future?

[00:39:23]  Unidentified speaker: So, okay.

[00:39:24]  Bruce: So let's use that now as our argument.

[00:39:28]  Unidentified speaker: Okay.

[00:39:28]  Bruce: So what we've done, I want to point out that what we've done is we've unratcheted the argument a little bit.

[00:39:33]  Unidentified speaker: Okay.

[00:39:34]  Bruce: Well, we started off by saying brain in a vat and we've now clarified. Well, actually, we mean a brain in a vat in some futuristic reality where their computers are much faster and much more capable than ours today.

[00:39:47]  Unidentified speaker: Okay.

[00:39:48]  Bruce: So we've now unratcheted the theory and we're a little more specific about what we actually mean.

[00:39:53]  Unidentified speaker: Okay.

[00:39:53]  Unidentified speaker: I know.

[00:39:53]  Unidentified speaker: I know.

[00:39:54]  Peter: I'm violating the tenets of critical rationalism, but I can't help it.

[00:39:59]  Unidentified speaker: Yeah.

[00:39:59]  Unidentified speaker: So we noticed that we did.

[00:40:00]  Bruce: We did violate Popper's Ratchet by doing that. We've ad hoc saved a specific and testable version of this theory by making the theory a little vaguer and a little less testable.

[00:40:09]  Unidentified speaker: Okay.

[00:40:10]  Bruce: Because now I don't know what I mean by futuristic reality. Like I can imagine almost anything now. There's no way to test that anymore.

[00:40:18]  Unidentified speaker: Okay.

[00:40:19]  Unidentified speaker: Okay.

[00:40:19]  Bruce: But even this unratcheted version of this theory still has testable consequences.

[00:40:25]  Unidentified speaker: Okay.

[00:40:27]  Bruce: So it's difficult to formulate a theory to have no testable consequences unless you make your theory very, very, very vague. So let's actually do this. So you just brought up the idea. Maybe we have a quantum computer. So it's in the future. We got a quantum computer. And so we're able to simulate down to the subatomic level of some area of space.

[00:40:49]  Unidentified speaker: Okay.

[00:40:49]  Bruce: That's not an inconceivable possibility in the future. So especially if we had quantum computers. So let's assume now to allow the un-ratcheting of this theory. And let's now see if it has testable consequences. And I'm going to argue it does.

[00:41:07]  Unidentified speaker: Okay.

[00:41:08]  Bruce: See, here's the problem. No matter how fast your computer is in the future, I can surely still make an experiment to tell me I'm in a simulation of reality. Because unless your computer is larger than the whole universe, it can't simulate the whole universe down to the atom.

[00:41:23]  Unidentified speaker: Okay.

[00:41:24]  Bruce: Now, there might be an exception to that. I'm going to cover that exception, the Omega Point computer, in a second. But at the moment, we're not taking the Omega Point computer into consideration.

[00:41:33]  Unidentified speaker: Okay.

[00:41:34]  Bruce: So you are guaranteed to have to cut corners even in the future, in a way that is inconsistent with reality. Now, perhaps, let's say that you have a big enough and fast enough computer that you're going to simulate the entire planet Earth down to the subatomic particle.

[00:41:55]  Unidentified speaker: Okay.

[00:41:56]  Bruce: In fact, you really kind of need to do that, right? Like if you aren't simulating the Earth down the subatomic particle, then the way I'm going to prove I'm inside of a virtual reality is I'm going to build a microscope and I'm going to discover that there are no subatomic particles. So we almost have to assume that this is a computer that can do subatomic particles down for the entire Earth at the same time. Even then, I can still perform an experiment to show that I'm inside of a virtual reality. I'll go fly to the moon and do the experiment. Now at this point, you can just simply keep unratcheting your theory. You can say, well, it's big enough to encompass the moon. Okay, I got to go to Jupiter. It's big enough to encompass Jupiter and you just keep growing it. It's big enough to encompass the whole galaxy down to the atom. It's like, well, I'll leave the galaxy or someone will at some point and then the jig is going to be up. It'll prove that I was actually in a simulation. At some point, you have to realize that it will be possible to come up with an empirical experiment to show that this is not that we actually live in a simulation unless you're literally emulating the entire universe on a computer, okay, which would require a computer larger than the whole universe. So now, let's reconsider Nick Bostrom's argument from before, okay. So someday we're going to create simulations so realistic that will model reality nearly perfectly. The people living in that simulation will outnumber those in reality. I won't even worry about the simulations within the simulations.

[00:43:31]  Bruce: Let's just stop right there, okay. So it's rational to, so goes the argument, so it's rational to assume we currently live in a simulation, okay. We just made a very easy proof that this just isn't the case, okay. At least not to the level that we've unratcheted so far where we're assuming futuristic. Really at this point, you can assume almost any futuristic technology short of an Omega Point computer, right? And we should be able to tell that we live inside of a simulation. So Pedro Domingos is a famous ML guy, responded to this argument by saying in a tweet, if there are multiple levels of simulations, each has exponentially fewer resources than the previous ones. So we're most likely to live in the bottom one, reality. I would have called that the top one. But anyhow, I love this argument from Pedro because it forces you to take a position about the nature of reality.

[00:44:30]  Unidentified speaker: Okay.

[00:44:31]  Peter: So he's saying we live in base reality.

[00:44:33]  Unidentified speaker: Yes.

[00:44:34]  Bruce: He's taking Bostrom's argument, and he's pointing out that it misses the fact that there's exponentially fewer resources, that the computer to emulate the whole universe must be exponentially larger than the whole universe.

[00:44:48]  Unidentified speaker: Okay.

[00:44:50]  Bruce: And if it's not, then you're not simulating the whole universe. So based on this argument, he's saying he's drawing the exact opposite conclusion that the majority of people will be living in base reality, and therefore we should be assuming we're in base reality. Now, I posted Pedro's counter-argument, and I had numerous crit rats that in my opinion, probably should have known better, immediately respond with, okay, but maybe real reality has 10 to the 10 to the 10 to the 10 times the computational power of what we see in the universe. So, aha, look how far we had to un-ratchet the theory to ad hoc save it. Now we're starting to literally make up theories about supernatural realities vastly different than the one we actually live in, okay? So we're starting to see the real problem here. If a solipsist were talking about any imaginable simulation, the vast majority of them would get detected almost immediately. Really what solipsists have in mind or simulation hypothesis people have in mind is that their brain with their brain in the vat experiment is some outer reality profoundly different than the one we see, that it has the real computational power necessary to have an entirely internally consistent reality yet somehow is subpar [word uncertain] compared to real reality down to the atom across the entire universe. They surely do not mean a brain in the vat in any straightforward sense that we might realistically imagine without invoking a supernatural world very different than our current reality.

[00:46:27]  Bruce: Now understanding that this is what a solipsist really had in mind is key to understanding why they are making an epistemological mistake. Now for the record, the crit rats involved who argued this with me, they were not arguing for solipsism obviously, but they were arguing you couldn't refute it. Okay, so now let's dig into that a little bit. To see why this is the case, let's first lay out the correct epistemological view summarized as, prefer theories that are more constrained. From CNR, Popper says on page 48, some theories are more testable, more exposed refutation than others. They take, as it were, greater risks. Also on page 48, every good scientific theory is a prohibition. It forbids certain things to happen. The more a theory forbids, the better it is. So Popper argues that we should prefer theories, prefer such theories that make prohibitions and constrain more over their unrisky counterparts. Or if you prefer, we could word it like this. Don't bother with theories that can explain any possible outcome because they explain nothing. So this is one that Deutsch is often quoted as saying some form of this, a theory that explains everything explains nothing.

[00:47:43]  Unidentified speaker: Okay, listen, this is what he means.

[00:47:46]  Bruce: This is exactly equivalent to the easy to vary versus hard to vary distinction that Deutsch uses. And it's also exactly equivalent to the ad hoc explanation versus bold explanation distinction that Popper uses, okay? We're increasing risk through increasing empirical content. Okay, so that is to say the no ad hoc rule or what I call Popper's Ratchet or what Deutsch calls the hard-to-vary criteria all right off the bat pronounce realism to be the superior theory to solipsism.

[00:48:20]  Unidentified speaker: Why?

[00:48:21]  Bruce: Because realism constrains more and because it doesn't explain any outcome imaginable. Now, Deutsch points this out with a quote, an extended quote I'm going to have to read from Fabric of Reality. He says James Boswell relates in his life of life of Johnson, how he and Dr. Johnson were discussing Bishop Berkeley's solipsistic theory of the non-existence of the material world. Boswell remarked that although no one believed the theory, no one could refute it either. Dr. Johnson kicked a large rock and said as his foot rebounded, I refute it thus. That's page 86 of Fabric of Reality. Now if the world were a dream, there is no particular reason why your foot should go through the rock versus the rock kicking back. Either outcome would fit the dream theory. For that matter, so does kicking it and having it turn into a pig. Or really anything else, okay? So realism has an advantage as an explanation over solipsism in that realism constrains the possible outcomes. If the rock doesn't kick back, you've refuted realism.

[00:49:26]  Unidentified speaker: Okay, but wait!

[00:49:27]  Bruce: Isn't realism a metaphysical slash philosophical theory and thus irrefutable? Does this mean, does the fact that we expect the rock to kick back at you, does that mean that realism is actually a scientific or empirical theory? I mean, isn't kicking the rock an empirical experiment that might have refuted realism if your foot had gone through it? So I want you to chew on that for a while. I'm not going to answer that, but let's see what you think about that. Just kind of think about that in the back of your mind for a moment. Okay, so but what is solipsism? Solipsism, is solipsism really refuted by Dr. Johnson kicking a rock? And this is why I said the crit rats that responded to Pedro Domingo's remark to Nick Bostrom's Simulation Hypothesis, that they should have known better. Okay, and we're going to get into this. Does, when Dr. Johnson kicks the rock and says, I refute it thus, has he refuted solipsism? Okay, what does it mean to refute false solipsism? So surely solipsism is a philosophical or metaphysical theory, to which Popper's ratchet or the no-ad hoc rule doesn't apply, right? Solipsism is something like the quintessential example of a metaphysical theory, isn't it? So is Deutsch and Dr. Johnson wrong to claim kicking a rock refutes solipsism? If we can literally refute solipsism by kicking a rock, this may put in jeopardy our assumption that metaphysical theories are irrefutable by experiment. So it bears looking at this a bit more closely, making sense of what Deutsch really means here. Okay, here is how Deutsch explains this.

[00:51:10]  Bruce: Dr. Johnson's point was that Berkeley's denial of the rock's existence is incompatible with finding an explanation of the rebound that he himself felt. Solipsism cannot accommodate any explanation of what [quotation wording uncertain] that experiment or any experiment should have one outcome rather than another. Notice that's the same argument I just made.

[00:51:31]  Unidentified speaker: That's from page 86, by the way.

[00:51:33]  Bruce: So this is identical to what I just said. A good explanation constrains more. But that's not the same as a refutation, is it? So, but here is the problem. We just saw that solipsism in general is quite a testable theory using our Skyrim example in most forms. It's really only a certain vague version of solipsism or the Simulation Hypothesis where we hypothesize a supernatural world that has 10 to the 10 to the 10 to the 10 times the computational power of reality. It's that version that's irrefutable. Or put another way, solipsism or the Simulation Hypothesis is every bit as testable as realism, but for the fact that the defenders have chosen to formulate it as an irrefutable theory. IE they chose to ad hoc save their theory from both experiment and, as we'll see in a second, criticism by violating the no ad hoc rule. So yes, of course, if you choose to formulate your theory as irrefutable, it is an irrefutable theory. But the only way to ever do that is to make your theory incredibly vague while adding wholly unexplained parts to your theory. In this case, the crit-rats defending the view were adding a supernatural reality with 10 to the 10 to the 10 to the 10 times the computational power of reality. This was the problem the crit-rats were missing. They were first intentionally formulating the theory to be irrefutable. Otherwise, it would be refutable. And Pedro Domingos' refutation would be correct. If this is a confusing way to think of this, and it probably is, there are, think of it like this.

[00:53:11]  Bruce: There are many versions of solipsism that you are free to conjecture under Popper's ratchet. For example, you could assume you are in a simulation running on current computers, running Skyrim, then come up with an experiment to refute that theory. In fact, all specific versions of solipsism that you can possibly think of today will have testable experiments that will refute those theories. Now, it's tempting to claim I'm wrong and to say, no, I'm being specific when I claim that real reality has 10 to the 10 to the 10 to the 10 times more computation than the universe that we see. But in fact, this is incorrect. The claim real reality has 10 to the 10 to the 10 to the 10 times computation than our universe is, as Deutsch is about to explain, actually quite vague. It's a case of vague manning your theory. Deutsch shows that we could discover we live in a simulation even without the ability to do experiments just through the process of criticism. So now this is from Fabric of Reality, page 137.

[00:54:13]  Unidentified speaker: I'm going to extensively quote this.

[00:54:15]  Bruce: So forgive me. I'm about to quote a whole bunch of what he says. Suppose that someone were to were imprisoned in a small unrepresentative portion of our own reality, for instance, inside a universal virtual reality generator that was programmed with the wrong laws of physics. What would such a prisoner learn about? What could such a prisoner learn about our external reality? At first sight, it seems impossible that they could discover anything about it at all. It may seem that the most they could discover would be the laws of operation, i.e. the program of the computer that operated their prison. But Deutsch points out that this isn't the case. Actually, it would be possible to come up with criticisms that would force you to conjecture an outside world. So now quoting page 137, But this is not so. Again, we must bear in mind that if the prisoners are scientists, they will be seeking explanations as well as predictions. In other words, they will not be content with merely knowing the program that operates the prison. They will want to explain the origins and attributes of the various entities, including themselves, that they observe in the reality they inhabit. But in most virtual reality environments, no such explanation exists, for the rendered objects do not originate there, but have been designed in the external reality. Page 138.

[00:55:35]  Bruce: You would notice that the chessboard, it was a chessboard reality is what he used as his example, is too simple an object to have, for instance, thoughts, and consequently that your own thought process cannot be governed by the laws of chess alone. Similarly, you could tell that during any number of games of chess, the pieces never evolve into self-reproducing configurations. And if life cannot evolve on the chessboard, far less can intelligence evolve. Therefore, you would also infer that your own thought process could not have originated in the universe in which you found yourself. So even if you had lived within the rendered environment all your life, and did not have your own memories of the outside world to account for as well, your knowledge would not be confined to that environment. You would know that even though the universe seemed to have a certain layout and obey certain laws, that there must be a wider universe outside it, obeying different laws of physics. And you could even guess some of the ways in which these wider laws would have to differ from the chessboard laws. Now, page 138 to 139. But reasoning in this way is not the only way that the inmate in our hypothetical virtual reality prison could gain knowledge of an outside world. Any of their evolving explanations of their narrow world could at the drop of a hat reach out into an outside reality. For instance, the very rules of chess contain what a thoughtful player may realize is fossil evidence for those rules having had an evolutionary history.

[00:57:11]  Bruce: There are exceptional moves such as castling and capturing en passant, I don't know what that means, which increase the complexity of the rules but improve the game. In explaining that complexity, one justifiably concludes that the rules of chess were not always as they are now. In the Popperian scheme of things, explanations always lead to new problems, which in turn require further explanations. If the prisoner fails after a while to improve upon their existing explanations, they may of course give up, perhaps falsely concluding that there are no explanations are available. Now, Deutsch argues that this would be very different. Oh, sorry, let me read the next quote here first, and I'll explain this. But if they do not give up, they will be thinking about those aspects of their environment that seem inadequately explained. Thus, if the high technology jailer wanted to be confident that their rendered environment would forever fool their prisoner into thinking that there is no outside world, they would have their work cut out for them. Okay, so what he's trying to say is that even if you could not perform an experiment, and you're inside this other reality, I just showed how you probably could perform an empirical experiment in most cases. But even if you were in a world where that was not possible, you could merely use criticism, and you could still eventually come to the truth, that you live inside of a simulation and that there's an outside world.

[00:58:49]  Bruce: And the only way that wouldn't be true to get around that, you would have to have a world that was internally entirely consistent. And as he puts it, the jailers would have their work cut out for him.

[00:59:01]  Unidentified speaker: Let me continue the quote.

[00:59:02]  Bruce: The longer they want the illusion to last, the more ingenious the program would have to be. It is not enough that the inmates be prevented from observing the outside. The rendered environment would also have to be such that no explanations of anything inside would ever require one to postulate an outside. The environment, in other words, would have to be self-contained as regards explanations. And then he adds, but I doubt that any part of reality, short of the whole thing, has this property.

[00:59:36]  Unidentified speaker: That's page 139.

[00:59:38]  Bruce: Or put another way, even if outside reality really was 10 to the 10 to the 10 to the 10 times the computational power of our reality. Unless our creators were God, perfect in making this, there would still be subtle mistakes that would allow us to detect we live in a simulation. This is what I mean by the 10 to the 10 to the 10 theory being vague. I'm asking, what I'm really asking people who make this argument is to lay out exactly how they intend to specifically build the simulation, such that it is entirely, internally consistent. And I don't think they can do this. In fact, I know they can't do this. So solipsists who think they are making a good argument aren't merely assuming some unexplained supernatural world. That's bad enough with enough computational power to simulate reality, but they're also assuming perfection by the designer. But without needing to specify exactly how this was achieved. In other words, the vast majority of solipsistic worlds can be refuted by experiment, and even those that can't would normally have mistakes that forced you to eventually conjecture the truth that you live, that there is an outside world. In fact, this is the real philosophical mistake people that take solipsism seriously are making. They aren't merely arguing that we might live in a simulation of some sort. They are arguing that we just happen to live in a situation so complete and so internally consistent that there is no need to correctly conjecture that there is an outside world. Now, this is exactly what I would call vague manning your theories.

[01:01:16]  Bruce: If you attempted to make your explanation even slightly specific, you would find your theory currently impossible and easily refuted. You'd have to have, for example, you would have to have an understanding of physics far beyond anything we currently know. So just to prove that's the case, imagine if a Newtonian wrote the simulation that we're supposedly living in. So obviously, it would lack QM's deeper explanations, and it would have holes in it because Newtonian physics literally doesn't explain everything. So we would eventually know we live in a simulation. In fact, that would be just as true if it were written by someone who knew quantum mechanics, because presumably, you have to, at a minimum, know quantum gravity to make the simulation internally consistent. So there's really no one alive today that would know how to make an internally consistent simulation. Okay, so a true solipsist, if they actually existed, which presumably they don't, other than people who are the simulation hypothesis [wording uncertain], must necessarily be very vague as to what our simulation is like, other than to say, well, I don't know, but it will always happen to match whatever we find out quantum gravity and all future physics theories, that's what it will be. So it is not really solipsism or simulation theory that is irrefutable. It's a vague version of it that is irrefutable. This is why kicking a rock does in fact refute all forms of solipsism except the vague manned version of it. So why is realism better even if it's wrong?

[01:02:57]  Bruce: So let me back up a little bit now and get to the question you actually asked. So it's still not logically impossible that we live in a simulation. Okay, even with every argument I just made, you could argue, well, it just so happens it's far enough in the future and there's enough computing power that they did make a perfect simulation. Or maybe it's so perfect, we'll eventually figure out we're in a simulation but that's still years off. You could make that argument still, okay? So you can't really rule out that you live in a simulation. And I would say yes, that's technically true. But there is a consequence for taking this view. So let's talk about the consequences for taking this view. So let's consider the idea of the omega point, okay, or an omega point like computer. So Deutsch develops this idea in Fabric of Reality, especially in Chapter 14 in The Ends of the Universe. And then he walks back the omega point theory in, at least partially, in Beginning of Infinity, when the omega point theory proved to be problematic. But even Deutsch admits that an omega point computer controlled by an omega point society could maybe simulate reality down to the atom. In fact, they could simulate the entire multiverse down to the atom. So here's him on page 348 of Fabric of Reality. From my own perspective, the simplest point of entry to the omega point theory is the Turing Principle. A universal virtual reality generator is physically possible. Such a machine is able to render any physically possible environment.

[01:04:31]  Bruce: In particular, this is page 357 now, in particular, Tipler points out that a sufficiently advanced technology will be able to resurrect the dead, or it could do this in several different ways, of which the following is perhaps the simplest. Once one has enough computing power, and remember that eventually any desired amount would be available in an Omega Point computer, one can run a virtual reality rendering of the entire universe, indeed the entire multiverse, starting at the Big Bang, with any desired degree of accuracy, page 357. So I was going to say, can't the Omega Point offer a way in which we can imagine we are living in a simulation after all?

[01:05:12]  Unidentified speaker: Okay.

[01:05:12]  Unidentified speaker: Okay.

[01:05:13]  Peter: So all this time we've talked about this, I still don't completely understand this, honestly. So when I first heard about the Omega Point computer, I just kind of pictured a quantum computer, honestly. But is the difference that an Omega Point computer requires eternal inflation?

[01:05:37]  Unidentified speaker: No, the opposite.

[01:05:38]  Unidentified speaker: It requires a big crunch.

[01:05:39]  Unidentified speaker: Yeah, a big crunch.

[01:05:40]  Unidentified speaker: No, no.

[01:05:40]  Unidentified speaker: Okay.

[01:05:41]  Unidentified speaker: Okay.

[01:05:41]  Peter: So it's under the assumption that eternal inflation does not exist.

[01:05:47]  Unidentified speaker: I've got that much.

[01:05:49]  Unidentified speaker: Then what is it?

[01:05:50]  Unidentified speaker: What is the?

[01:05:51]  Unidentified speaker: Okay.

[01:05:52]  Peter: Is there any way to just easily explain that?

[01:05:54]  Unidentified speaker: Yes.

[01:05:55]  Bruce: So an Omega Point deals with every problem I just raised.

[01:05:59]  Unidentified speaker: Okay.

[01:05:59]  Bruce: And that's actually what makes it so amazing. So this idea. So if you're a Silicon Valley type, if you're Nick Bostrom, the simple truth is, is you don't believe in the Omega Point.

[01:06:10]  Unidentified speaker: You think it's silly.

[01:06:11]  Bruce: You've maybe never even heard of it. Okay. And even if we were to take the Omega Point seriously here, it is a refuted theory at this point, right? Like, see, it doesn't really, because of the inflation issue.

[01:06:24]  Unidentified speaker: Yeah.

[01:06:24]  Bruce: So you can't really invoke it as a way to save Nick Bostrom's theory.

[01:06:29]  Peter: Because most cosmologists believe in eternal inflation now.

[01:06:33]  Unidentified speaker: Right.

[01:06:34]  Unidentified speaker: Right.

[01:06:35]  Peter: It requires the big crunch.

[01:06:37]  Unidentified speaker: Right.

[01:06:37]  Unidentified speaker: Okay.

[01:06:38]  Unidentified speaker: I'm there then.

[01:06:39]  Unidentified speaker: Okay.

[01:06:39]  Bruce: But what we're going to do here is we're going to ignore that fact.

[01:06:44]  Unidentified speaker: Okay.

[01:06:44]  Unidentified speaker: Okay.

[01:06:45]  Peter: Why does it require the big crunch? I guess that's what I really want to know [ending uncertain].

[01:06:48]  Unidentified speaker: Okay.

[01:06:48]  Bruce: So think about my argument that to have a computer large enough to simulate the entire universe down to the subatomic particle, you would have to have a computer exponentially larger than the universe.

[01:07:00]  Peter: Oh, so it's because the universe is more finite under the big crunch cosmology.

[01:07:07]  Bruce: It's actually better than that. Okay. As you do the big crunch, the amount of computational resources goes up. Even as the size of the universe goes down. And by the way, that means the computer is getting faster because the computer is moving at light speed.

[01:07:24]  Peter: And why does it go up because of the speed of light?

[01:07:28]  Unidentified speaker: Yes.

[01:07:28]  Bruce: So there's a number of reasons why. It's because the number of oscillations grows to infinity over time. So you're using the oscillations rather than trying to use subatomic particles. I'm not sure I could give you the exact mathematics. I mean, like I went through them when I read Tipler's book. And he has worked out a way where there is actually an infinite amount of computation just before the universe pops out of existence.

[01:07:56]  Unidentified speaker: Okay.

[01:07:57]  Bruce: And because of that, there is infinite computational resources. Now, the reason that exists, though, is because of the specialness of the omega point cosmology. And you don't get to just invoke infinite resources unless you can explain it. Because normally, you can't explain it. And this is exactly the problem, okay? Is that Nick Bostrom, if he believed in the omega point, and if the omega point was, in fact, a best theory, then Nick Bostrom's argument might kind of make some sense.

[01:08:32]  Unidentified speaker: Okay.

[01:08:34]  Unidentified speaker: There's still a problem.

[01:08:35]  Bruce: And I'm going to take you through what the problem is. Okay. And it's a big problem. It's like a huge problem. But it would kind of make sense in an omega point cosmology to say, well, you know, in the future, you got this giant group of people and they want to find out what the past was like. And so they simulate the Big Bang and the entire multiverse. And then they decide, you know what?

[01:09:00]  Unidentified speaker: It's humane.

[01:09:00]  Bruce: Let's save all these people. So they wait for them to die. They play out their lives.

[01:09:04]  Unidentified speaker: They yank them into heaven.

[01:09:05]  Bruce: And you know, it's, this is kind of what Tipler's theory was, okay? And Deutsch had a number of criticisms of the theory, even in Fabric of Reality. And then he had stronger criticisms in Beginning of Infinity, and even pronounced that inflation was a problem for it, and it wasn't an acceptable theory anymore. But let's just, for the moment, let's go ahead and ignore all that. And let's invoke the omega point to save the simulation, Nick Bostrom's Simulation Hypothesis, even though Nick Bostrom would never even know about the omega point, and if he did, he wouldn't invoke it.

[01:09:42]  Peter: But the omega point you see is a more like empirical testable kind of...

[01:09:47]  Bruce: Well, I mean, it got refuted, so yes, it was empirically testable.

[01:09:50]  Unidentified speaker: So...

[01:09:50]  Peter: Theory, okay.

[01:09:51]  Unidentified speaker: Yeah.

[01:09:53]  Peter: So I find it interesting is just sort of a way that from an optimism sort of perspective, that there are, even if it's not specifically correct, there are theories out there that go around that counter cosmological heat death, right?

[01:10:17]  Unidentified speaker: Yeah.

[01:10:17]  Peter: I mean, we've talked about this in this context before. It's still kind of like cool that there are optimistic scenarios out there, even if it's not like 100% correct.

[01:10:29]  Unidentified speaker: Right.

[01:10:29]  Unidentified speaker: Yeah.

[01:10:30]  Unidentified speaker: All right.

[01:10:31]  Bruce: Here is Deutsch's excellent response to this kind of argument. And I'm now going back to page 86, where we're talking about solipsism, and I'm going to use it to respond to Nick Bostrom's simulation hypothesis, now unratcheted and saved by the Omega Point.

[01:10:49]  Unidentified speaker: Okay.

[01:10:50]  Bruce: So, Deutsch says, and my argument here is, yes, it's possible we're in a simulation if you're assuming an Omega Point like cosmology. That is possible. There is a consequence for taking that stance. This is what Deutsch is going to work out now. He says, solipsism cannot accommodate any explanation of why that experiment or any experiment should have one outcome rather than another. To explain the effect that the rock had on him, Dr. Johnson was forced to take a position on the nature of rocks. Just like Pedro Domingos' tweet forced you to take a position on the nature of reality and to suddenly pony up and admit, well, I'm actually saying that the outer reality has 10 to the 10 to the 10 more computational resources than the universe, right? Deutsch again, were they part of an autonomous external reality or were they figments of his imagination? In the latter case, he would have to conclude that his imagination, in quotes, was itself a vast, complex, autonomous universe.

[01:11:58]  Unidentified speaker: Okay, page 86.

[01:12:02]  Bruce: Referring to a theory that the planets were moved by angels, Deutsch continues and he says, and the Inquisition would have to take a position on the source of the underlying regularity in the motion of planets, a regularity that is explicable only by reference to the heliocentric theory. For all these, remember, the Inquisition believed that angels were moving planets, right? But they happened to be moving them according to the heliocentric theory. For all these people, taking their own position seriously as an explanation of the world would lead them directly to realism and Galilean rationality, page 86. Okay, so in both of these cases, the supposedly correct theory, so we're assuming now that the Simulation Hypothesis is correct, okay, still has to invoke the explanatory power of the supposedly incorrect theory to make sense of it. So, Deutsch continues, page 87. But in fact, the rebounding of kicking the rock depended on what the rock did, such as being in a certain place, which was in turn related to other effects that the rock had, such as being seen or affecting other people who kicked it. Dr. Johnson perceived these effects to be autonomous, independent of himself, and quite complicated. Therefore, the realist explanation of why the rock produces the rebounding sensation involves a complicated story about something autonomous. But so does the solipsist's explanation. In fact, any explanation that accounts for the foot-rebounding phenomenon is necessarily a complicated story about something autonomous. It must in effect be the story of the rock.

[01:13:51]  Bruce: The solipsist would call it a dream rock, but apart from that claim, the solipsist's story and the realists could share the same script. In other words, even if this version of solipsism or the Simulation Hypothesis, which is impervious to experiment and criticism, was correct, we'd still need to think of the rock as being something effectively real. Let's put this answer in perspective, and we'll see that it applies just as well to the Simulation Hypothesis, as it does to solipsism. What Deutsch is pointing out is that even if reality is just a dream, the simple truth is that so long as the dream happens to be consistent with realism, then realism is the correct theory and is a better explanation about reality. So even if solipsism is entirely correct, realism is still also the correct explanation and is in fact a better explanation, since even the solipsistic explanation must invoke it. Hopefully you can see how this applies to the Simulation Hypothesis. Even if we do live in an entirely internally consistent simulation running inside an omega point computer, run by godlike omega point beings, that is still exactly like saying realism, from our point of view, is a correct explanation. In fact, one of your starting assumptions is now essentially, we don't need the Simulation Hypothesis. I mean, that is, when you say it's internally consistent, you're saying we don't need the Simulation Hypothesis. We might as well just treat this as reality.

[01:15:26]  Unidentified speaker: Why?

[01:15:27]  Bruce: Because we really only needed solipsism as a theory or the Simulation Hypothesis as a theory or as an explanation, if we were not in a perfect simulation or perfect dream of reality.

[01:15:41]  Unidentified speaker: Okay.

[01:15:41]  Bruce: Now, Peter, do you find this answer maybe at once correct in every way, but still unsatisfying?

[01:15:51]  Peter: I find it very satisfying as a critical rationalist, as a guy that might like listen to Nick Bostrom talk to Joe Rogan for three hours on the podcast.

[01:16:03]  Unidentified speaker: I don't know.

[01:16:05]  Peter: I think I still think it's a pretty intriguing idea.

[01:16:11]  Bruce: So let me point something out here. You now know what it feels like to be a theist.

[01:16:17]  Unidentified speaker: Okay.

[01:16:19]  Bruce: You know exactly what it feels like to be a theist.

[01:16:21]  Unidentified speaker: Oh yeah.

[01:16:22]  Unidentified speaker: Yeah.

[01:16:23]  Peter: I've heard that connection between the religion and simulation hypothesis before.

[01:16:30]  Bruce: It's not a strange connection. Simulation hypothesis is a belief in a certain kind of God. It is a religious belief. Not kind of, not sort of.

[01:16:40]  Unidentified speaker: It is.

[01:16:41]  Unidentified speaker: Fair enough.

[01:16:43]  Bruce: And honestly, theists do find such arguments unsatisfying, and so do you.

[01:16:49]  Unidentified speaker: Okay.

[01:16:50]  Bruce: Even though you consent to them, even though you say, well, as a critical rationalist, I can see you're right, that we don't really need to invoke God. We don't really need to invoke a simulation. We don't need to really invoke a dream. Even though you can see that's exactly correct, there's just a part of you that just finds it unsatisfying.

[01:17:09]  Unidentified speaker: Yeah, that's fair.

[01:17:11]  Unidentified speaker: Okay.

[01:17:12]  Bruce: But from a critical rationalist standpoint, we aren't really rejecting the simulation hypothesis. And I think this is where maybe I'm different than most critical rationalists. Clearly I am different than most critical rationalists, crit rats in a lot of different ways, right? And I've kind of argued that the crit rat community has gotten away from Popper on some of these things. And this is one that I'm going to argue they have gotten away from Popper on.

[01:17:36]  Unidentified speaker: Okay.

[01:17:37]  Bruce: If you were to look at the Bayesian community, they're going to say, oh, I think that it's probably true or it's probably not true, and they're going to take a stance. As a critical rationalist, I feel no real need to take the stance, the Simulation Hypothesis is false.

[01:17:54]  Unidentified speaker: Okay.

[01:17:56]  Bruce: What I do instead is I throw it into the bucket that I call the morass. It's a giant bundle of an infinity of explanations that are indistinguishable from each other because they have no actual implications that can be checked or tested. The morass is a great place to find interesting conjectures. But you have to take the explanation out of the morass by making it checkable or testable before it can be distinguished from every other explanation inside the morass. There is really no point in even taking a stance if we actually, secretly live in a perfectly self-consistent simulation run by gods, precisely because a perfectly self-consistent simulation is actually fully explained by realism, by definition, even if by the logic of this, even by the logic of solipsism itself. There's really no point in even taking a stance if we really actually, secretly live in a perfectly self-consistent simulation, precisely because a perfectly self-consistent simulation is actually fully explained by realism, even by the logic of this form of solipsism, which is a preferable theory, realism is a preferable theory, because it constrains more and therefore explains more. Indeed, it is hard to see how you could ever get the perfectly consistent self-consistent simulation hypothesis out of the morass and become a good explanation, because it's intentionally framed to never come out of the morass. Okay, that's, they're doing that intentionally, all right? Perhaps by, perhaps you could, in a case like this, have something like the Second Coming.

[01:19:36]  Unidentified speaker: Okay, that would do the trick.

[01:19:37]  Bruce: So it's not maybe impossible, but you can see how until the Second Coming happens, invoking the simulation hypothesis or invoking God or invoking religion, it's gonna stay in the morass almost by definition. And you can see why some people would kind of believe it and take it seriously. But you can also see, and you can see why there's a good argument against it, but why that argument is a little unsatisfying. But our goal isn't so, as a critical rationalist, our goal isn't so much to figure out what is true as to figure out which theory is best. Yes, that's on the assumption that this will tell us something about true about reality, which it does. Even if the simulation hypothesis is correct, realism is still a good theory on its own, right? Realism is not like, if we're talking about a perfect simulation, run by a perfect being, you know, and it has no mistakes, then realism is still true, right? Solipsism or simulation hypothesis are no longer at odds with realism. Realism is still, in some sense, the correct theory. It's just a theory sub to the other ones. Okay, so now, let me just make a comparison here between Popper's argument versus Deutsch's. Oh, before I move on, maybe let me just summarize my thinking on that. I don't know how many critical rationalists would agree with me, how many critical rats would agree with me here, and I don't have handy the, I had a quote from Popper where he actually does say something to this effect, so I feel like this is something I could have sourced from Popper.

[01:21:14]  Bruce: But what I'm really saying is, I don't really feel the need to reject wild theories like this, right? Instead, I simply label them for what they are: untestable. If you were to ask me, do we live in a simulation? I'm going to say, you know, I have no reason to believe that. And that's kind of the right answer, right? It's not an answer, no, we do, or yes, we do, or no, we don't. And as a critical rationalist, I feel like that's actually a pretty good answer. I can, but I want to be sympathetic to where, Peter, you're coming from, where you say, you know what, maybe we do have this infinite growth, an Omega Point computer, like computer. Maybe we are actually inside the simulation. You know, I think if I were to try to criticize that view, here's how I would probably go about it.

[01:22:03]  Unidentified speaker: Okay.

[01:22:05]  Bruce: And this is going to kind of be a little mind-blowing and it might be a little hard to follow. So I apologize. And I probably should have thought this through better to try to figure out a way to make it simpler. Why would the godlike beings inside the Omega Point, why would they make a perfect simulation of the Big Bang? Well, obviously, it would be because they want to gain knowledge of what actually happened inside the multiverse, okay? There may be moral reasons too. Okay, I know Deutsch tried to criticize this, but it may be that they literally want to recreate their history across the multiverse and then save everybody and say, you know what, these people that came before us, they had hard lives, let's pull them into the Omega Point and let's give them a happier life.

[01:22:53]  Peter: Or, you know, it might be something so trivially easy in that these godlike entities that it would be only take like one teenager or something who wants to do the silly thing.

[01:23:08]  Bruce: So we've now come up with an explanation. So and this is kind of more important than it first seems. You need to explain why the people creating the simulation bothered to make it a perfect simulation. OK, because there's no particular reason why they should want to pull the wool over our eyes. If we were really going to in the future in Nick Bostrom's style want to create simulations, we probably wouldn't make them perfect simulations and there would be no reason for us to do so. And when the people inside the simulation figured out that they were living in a simulation, we would think that was cool, right? We'd probably even at that point want to communicate with them and say, hey, you know, let's have conversations with them. So there's no particular reason why we should be living in a perfect simulation, unless there was some sort of moral reason like this or knowledge-based reason that they wanted to find out what happened in the Big Bang. They wanted to save us. That would have to be the reason, right?

[01:24:06]  Unidentified speaker: Or maybe tourism.

[01:24:07]  Peter: Maybe they wanted to experience what people lived like a couple million years ago.

[01:24:16]  Bruce: You might make an argument that that's a bad reason to put people through a lot of suffering just so you could have tourism. They probably are more moral than that because these are godlike beings living in an objectively more moral set of values than us, right? So, you would really need to make an argument that they needed this knowledge and they wanted to save us. It's hard to come up with what a better explanation would be for why they decided to make it a perfect simulation. So, let's say that they did, though. Okay, so let's say they made this perfect simulation. The perfect simulation is really just a simulation of you, Peter, when you lived across the multiverse. You're in this you that I'm talking to you right now. Let's say it's quote unquote the simulation version of Peter, but it's this simulation version of Peter is identical to the one that actually lived in the real world in the past. Okay, so in fact, this is very much like the multiverse. The fact that they are doing a perfect simulation of Peter and that there really was in the past such a Peter that this is a perfect simulation of, you can't really say the one living in the multiverse is living in a simulation, because it's really just the same Peter that lived in the real world. I know that's a little mind-blowing, okay, but it's like trying to differentiate universes that are identical. We just simply consider them to be the same universe, okay, in the multiverse. There's really an infinity of universes that are identical, so we just treat them as a single universe.

[01:25:52]  Peter: That's the fungible concepts.

[01:25:55]  Unidentified speaker: Yes.

[01:25:56]  Bruce: So the same thing happens here, right? We're emulating every Peter that existed across the multiverse, and yes, there's some version of you in the simulation, but that simulation is completely identically the same as the Peter that really lived. So from a certain point of view, we might as well just say, this is the real Peter, and he actually lived in reality, and he isn't in a simulation, and that would be a correct way to look at it. That would be the most correct way to look at it. Yes, it's a little complicated because there's this copy of Peter that lives in the simulation, that we needed to be able to save Peter after Peter dies, and pull him into the Omega point, right? So even in a case like this, where we're using the Omega point to try to save Nick Bostrom's theory, I think it still doesn't really make sense, right? You can get there, but only in this really weird way, where what you're really doing is you're making a simulation of someone who actually lived, and it is therefore that person, the person that actually lived.

[01:26:58]  Peter: In Tipler's Omega point, they simulate every person who has ever lived, or who could ever live.

[01:27:06]  Unidentified speaker: That's correct.

[01:27:07]  Bruce: Across the entire multiverse.

[01:27:09]  Unidentified speaker: That's right.

[01:27:11]  Unidentified speaker: Okay.

[01:27:11]  Bruce: So even with this way of doing it, I don't know that the Omega point can really save the truth of what Bostrom was trying to get at. It may look a little like it, but it's different in so many ways, and it just doesn't have quite the same impact that Nick Bostrom intended. So let's now compare Popper's argument versus Deutsch's, okay? Popper's argument, while ultimately correct, could just as easily be applied to an incorrect argument, and in fact was just as easily applied to an incorrect argument. So that isn't desirable for arguments. Is that true for Deutsch's argument?

[01:27:52]  Unidentified speaker: No.

[01:27:53]  Bruce: Deutsch's argument first points out how easily we could test or criticize something like the Simulation Hypothesis if you were trapped in a simulation, and how it is only untestable under very special but wholly vague and unspecified conditions, where they're entirely internally consistent. Okay, then his view takes the view, the other viewpoint seriously and says, okay, but if you are right, the competing theory realism is still the better theory, because we still don't need your solipsistic theory. So it defeats solipsism on its own turf, okay, by showing that this version of solipsism, the kind that is untestable, the kind that is uncriticizable, is really just a fancy way of saying realism. So it defeats every form of both solipsism and the simulation hypothesis simply by taking it seriously and working out the consequences and implications of the theory itself. Here's from Popper, Myth of the Framework, page 60. By contrast, the correct method of critical discussion starts with the question, what are the consequences of our thesis or our theory? And also on page 60, a kind of rational discussion, critical discussion, does not seek to prove or justify or establish a theory, but tries to test the theory under discussion by finding out whether its logical consequences are all acceptable. In other words, Deutsch attacks solipsism along both of Popper's axes, see episode 83 if you don't know what I mean by that, simultaneously.

[01:29:33]  Bruce: Deutsch first shows its defenders could have made their theory testable, in which case it would have been refuted or at least criticized out of existence. You would discover you were actually in the simulation. And then showed their choice to make it irrefutable meant that their theory had no explanatory implications of its own in the vague man form they've chosen. This is the correct way to apply critical rationalism to philosophical and metaphysical theories. Indeed, it seems to me to blur the lines between the two sides of the demarcation line, since there seems to be little or no difference in how Deutsch goes about applying critical rationalism to a philosophical theory as you would to a scientific theory. So, that would be, sorry, my summary of how Deutsch destroys solipsism.

[01:30:26]  Peter: Okay, well, I've enjoyed listening to you, Bruce, and I hope someone out there does too, and I love the title, and I...

[01:30:36]  Unidentified speaker: Thank you.

[01:30:37]  Unidentified speaker: All right.

[01:30:38]  Unidentified speaker: Thank you.

[01:30:46]  Unidentified speaker: Hello again.

[01:30:47]  Peter: If you've made it this far, please consider giving us a nice rating on whatever platform you use, or even making a financial contribution through the link provided in the show notes. As you probably know, we are a podcast loosely tied together by the Popper Deutsch Theory of Knowledge. We believe David Deutsch's Four Strands tie everything together. So we discuss science, knowledge, computation, politics, art, and especially the search for artificial general intelligence. Also, please consider connecting with Bruce on X at BNielsen01 [spelling uncertain]. Also, please consider joining the Facebook group The Many Worlds of David Deutsch, where Bruce and I first started connecting.

[01:31:30]  Unidentified speaker: Thank you.