Episode 53: Universality and IQ - Part 1

• 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

[00:00:10]  Bruce: Welcome to The Theory of Anything podcast. Hey guys, how's it going?

[00:00:15]  Peter: Hi Bruce, nice to see you.

[00:00:16]  Bruce: Hey Peter.

[00:00:17]  Unidentified speaker: Hear you.

[00:00:18]  Unidentified speaker: Hear you.

[00:00:19]  Unidentified speaker: Hi guys.

[00:00:20]  Bruce: Hey Cameo. You know, the next episode that's gonna come out is one that Peter is in. But it's not the first one we recorded with him. The first episodes we recorded with Peter were actually on the topic we're talking about today, universality and IQ. We recorded three very long episodes, and then we didn't really like them that much. Peter pointed out we made some mistakes, and then we couldn't find one of the episodes. Like, the recording just vanished. So we decided we're going to re-record, and so that was actually Peter's introduction to the show. So they're gonna hear Peter on the show, and they're gonna not have any idea where he came from or who he is, or, you know, because the first ones that had to come out with Peter, there's no introduction to him, because...

[00:01:07]  Peter: I almost think the 45-minute tangent we went on about, are we weird? Could have been like its own mini podcast or something.

[00:01:14]  Unidentified speaker: I don't know.

[00:01:15]  Bruce: You know, I may go, I still have that, I think. So I may go, like, pull that one.

[00:01:19]  Unidentified speaker: Just putting that out there.

[00:01:21]  Bruce: Peter, why don't you give us an introduction, since we didn't give you a good introduction in the last episodes, people are gonna have heard.

[00:01:26]  Peter: Okay. I connected with you, Bruce, through the Facebook group that I created called the Many Worlds of David Deutsch. I would encourage anyone out there listening to this to join that group if they're on Facebook. For years, I was participating in similar kinds of groups to about Sam Harris and other kinds of guys in that world. And I just got, it wasn't actually through the Sam Harris podcast initially, but through checking out David Deutsch's, first his TED Talks are what really pulled me in. And then his books, of course, and just everything else out there, I just got really, really turned on to the whole worldview. I mean, I've just never, never heard anything that made as much sense to me, him and Popper, of course, too. And so that's how we started connecting there. And here I am on your podcast. And I think we've had some wonderful conversations here.

[00:02:29]  Peter: And I'm excited to talk about IQ.

[00:02:33]  Bruce: I think Peter's been a great addition to the show, by the way.

[00:02:38]  Unidentified speaker: I agree.

[00:02:40]  Bruce: Even though we threw away more than half the episodes you've recorded.

[00:02:42]  Peter: Oh, it's okay.

[00:02:44]  Unidentified speaker: All right.

[00:02:44]  Peter: May I just throw out one other thing there to introduce myself that I wanted to get out of the way that my professional experience, I'm not a scientist, not a philosopher. I think I'm on my 17th year of teaching, hard to believe for me, but as a special ed education teacher. So a lot of my kind of professional experience, life experience, I guess just indicates to me it's really hard for humans to change their minds, change their brains, I guess, is really what I mean. In the past, I have been really hooked into an evolutionary psychology view of humans and what we are. I think that getting into David Deutsch has really made me question a lot of these assumptions and gave me just a new way of understanding humans. To be fair, one I sympathized with to a good deal even before reading David Deutsch, I've always considered myself, not always, but at least as an adult, been pretty tuned into optimism.

[00:04:02]  Peter: And I think that's why his ideas about knowledge and reality and humans and universal explainers and all this, it just was like, yes, this just feels right to me. However, I do have a lot of still, I guess, trepidation about this one area. It really does seem to me that we're kind of born with the brains we're born with and it's really hard to change. It's no matter what theory you have. So this is an area where I think maybe we are somewhat in alignment on in that we have a lot of questions.

[00:04:44]  Unidentified speaker: Yes.

[00:04:44]  Bruce: So Peter asked me some questions on Facebook about this and I said, hey, why don't you come join our show? We're going to do a podcast episode or three on this. That was what got him on the show and then he did such a great job. We just started inviting him back every single time. That's how Peter ended up on the show and that's why he suddenly emerges as a new player in the podcast out of the blue.

[00:05:12]  Cameo: I hate it when a TV show drops an entirely new character and it gives you no backstory.

[00:05:17]  Unidentified speaker: Yes.

[00:05:18]  Unidentified speaker: That comes from. [Unclear/overlapping speech.]

[00:05:21]  Bruce: Let me also just mention something else. There are two groups that associate with the Twitter critical rationalists, which I call the Twit Rats and the Facebook critical rationalists, which I call the Fbook Rats. There's actually a lot of difference between the two groups. Peter is one of the main people in the Facebook group. As he mentioned, he started one of the main groups, the many worlds of David Deutsch. He's a strong voice there. Peter's group is heavily associated with the various Karl Popper groups that are on Facebook, which are made up of what the Twit Rats call the Old Guard. It's the students of Karl Popper, people who actually knew Karl Popper. People who are distinguished in their field in philosophy, and in Popperian philosophy, so Bruce Caithness, and who are some of the other ones? Luke, I don't even know how to pronounce his last name, but some people who actually did know Karl Popper personally, Margaret.

[00:06:25]  Bruce: I don't know if she knew Karl Popper personally, but she's someone who's published in that field and things like that. And it seems to be more oriented towards critical rationalists who are aware of David Deutsch, they did not come to critical rationalism through David Deutsch, they're usually older. Whereas the Twit Rats are the ones who tended to be came to critical rationalism through David Deutsch, know about critical rationalism as part of, their interest in critical rationalism came from the interest in David Deutsch's theories. They're almost always younger. I'm like one of the single oldest ones on Twitter. You can just tell the difference when you're talking to the two groups. It's not good or bad, it's just there are distinct cultural differences between the two groups. Dwarkesh Patel, who I don't think I've seen amongst the F-book rats.

[00:07:19]  Bruce: I wouldn't know if I'd call him a twit rat because he doesn't really fully agree with David Deutsch, but he certainly has associated with them a lot and interact with them and he's got his own podcast, The Lunar Society podcast. He's interviewed David Deutsch as well as lots of other people. It's a very popular podcast, [from what I understand].

[00:07:35]  Peter: Wonderful podcast, I think.

[00:07:37]  Unidentified speaker: Yeah.

[00:07:38]  Peter: He's interviewed a lot of interesting people. I mean, Charles Murray and all kinds.

[00:07:44]  Bruce: He wrote an article called Contra David Deutsch on AI with the subtitle, The Universal Explainer's Hypothesis Incorrect because it cannot explain why some animals are smart and some humans are stupid. He then writes this whole article. I'm going to give you first a summary of what I felt was the most salient points from his article. I'm going to try to summarize it as something much shorter than what he actually wrote. Of course, Bruce's rules of summary applies here. The summary is never the same as the actual thing. You should go read the actual article. And I definitely think it's worth reading. In this episode today, I'm going to summarize his point of view, and then I'm going to summarize the opposing viewpoint of the twit rats. I'm not going to express my opinions this time. I'm really mostly curious what Peter and Cameo think. And then in future episodes, I'll be willing to express my opinions a bit more strongly.

[00:08:39]  Bruce: But I found the discussion quite interesting for a variety of reasons on both sides. So let me first try to summarize both the sides of the argument.

[00:08:50]  Unidentified speaker: Okay.

[00:08:51]  Bruce: So Dwarkesh Patel, here's probably what I see as the key quote from him, that summarizes his view the best. IQ scores are heavily correlated with job performance, school performance, income, lack of crime, and even health. IQ is clearly measuring some underlying trait that determines your effectiveness across a broad range of cognitive demanding tasks. In many cases, even years down the line. You can call this trait whatever you want, but I call it intelligence. I think that summarizes his view well, and I think that's a pretty powerful argument that he's making. Now, throughout his article, he makes a number of what I would consider pretty bad arguments. One of the things that we do when we're arguing with people is we cherry pick. If you don't like what somebody's saying, you pick out the bad arguments and you show they're wrong and you ignore the good arguments.

[00:09:44]  Bruce: I don't want to do that. First, let me just acknowledge, he does make a number of bad arguments and I'm going to, for the most part, just ignore them and I'm going to concentrate on his good arguments because he makes a number of really good arguments. Here's an example of a bad argument. He says, about 21-24 percent of Americans fall in level 1 of the National Adult Literacy Survey, meaning they can't even perform tasks identifying the expiry date of a driver's license or totaling a bank deposit slip. Now, the problem with an argument like this, I mean, this could be a good argument for all I know. Maybe this is because some people are low IQ. But he's not offering any reason for me to differentiate that from the view that these are just low-educated people. Nobody doubts that if you're not literate, then you can't read, no matter what your IQ is.

[00:10:37]  Bruce: In fact, there's not even any real doubt that education plays some role in the development of mind and that you will score higher on an IQ test if you received an education at a younger age versus if you just didn't. So whatever we mean by education, that might be unschooling for all I know, but just educated in some meaningful way where you've received knowledge as you were growing up. So because of that, an argument like this makes no real attempt to distinguish between theories and therefore isn't useful. Even if he's right, that that is caused by low IQ.

[00:11:14]  Peter: But I might have to push back on that a little bit. I mean, it sounds to me like basically what he is trying to get across with that fact is that there are a huge number of humans in this world, probably more than most people would imagine, who are unable to perform basic intellectual tasks. I mean, these tasks don't even necessarily have to do anything with school, necessarily. I mean, they really struggle with basic things that other people just never had to learn, really. I mean, just because it just comes so naturally to their brains that probably a lot of people would never even consider that there are so many people out there who are unable to do.

[00:12:09]  Bruce: You make a good point.

[00:12:10]  Peter: It seems to me the IQ bell curve, I guess, kind of captures it pretty well when you consider it.

[00:12:17]  Bruce: So I was concentrating on the fact that an individual case like this is inductive. It doesn't really show us anything. What you're pointing out is that the really salient point is the 21 to 24 percent.

[00:12:29]  Unidentified speaker: Yeah.

[00:12:30]  Bruce: Okay, I'll buy that. That does improve the argument if I understand it in that way. What I said, though, is still true. He does make a number of arguments that I feel like don't really differentiate between theories. I want to concentrate on the arguments that I felt were actually good arguments. By the way, Peter sent me this just before the show. So people at home aren’t going to see this, but this is a picture that's kind of typical of how people look at the prevailing view of intelligence. So it's this view of people at the top, the small number of exceptionally gifted people like Albert Einstein who made our technical civilization possible. That's IQ of 160 or above. And then kind of below that in the hierarchy is highly gifted people like Jeff Bezos and Elon Musk who invent our infrastructure and create jobs, IQ of 145 or above. Then below that is moderately gifted people like software engineers and computer scientists who implement the ideas of highly and extremely gifted people, IQ of 130 and above.

[00:13:30]  Bruce: And it kind of just goes down until you get down to, at the bottom level, the least able of the population who survived due to social welfare, IQ of below 85, below or equal to 85. And this kind of goes with what you were just saying, that there is this kind of bell curve that exists. This isn't in doubt, right? We know there's this bell curve of IQ that exists, whatever IQ is. But this picture goes further. It's trying to use it to explain things. It's trying to say, yeah, so you have to be high IQ to be someone like Einstein who creates, you know, a brand new physics theory. And then you have to be somewhat high IQ to be a captain of industry that is going to create an Amazon, you know, and then moderately high IQ to be someone who implements that and kind of goes down and it's trying to explain how society works in part based on IQ. And this is certainly typical of how the prevailing theory of intelligence looks at things.

[00:14:34]  Bruce: I should probably note, there's probably no such thing as the prevailing theory of intelligence. There's probably more like many competing theories, but they group together around ideas like this. So saying the prevailing theory of intelligence is probably a meaningful statement, even while it's not literally true.

[00:14:50]  Peter: And I think it's possible to sort of basically accept that this kind of worldview has some validity without seeing IQ as being like the end of the line way to classify human intelligence. I mean, I think probably most people recognize that there's different aspects to this.

[00:15:10]  Bruce: So even people who believe strongly in IQ, don't really believe in anything but roughly measures intelligence. [Unclear wording.] Yeah. So IQ is this measure we're familiar with that captures an idea that different people have different levels of intelligence. Peter, you also made a comment when you were sending me this about the idea of rank ordering. So IQ kind of rank orders people, right? It says, you know, I've got an IQ of 130, so that's better than an IQ of 100 or something along those lines, right? There is actually an alternative opinion on that even amongst the prevailing view, which is that there's multiple types of intelligence. And we'll get into this later when we talk about this in a bit more depth. And people who believe in IQ don't necessarily deny that. They definitely believe that, you know, there is such a thing as someone who's really good at math, but not very good at English grammar or something, or language, or art, or something along those lines.

[00:16:15]  Bruce: And they aren't necessarily claiming. So I think the prevailing theory is probably more like a partial rank order than a true rank order. But IQ itself is a rank order, but then it's only perceived as this super rough, best we can do right now, attempt to measure intelligence.

[00:16:33]  Peter: I think they may also want to point out that if someone is higher in, say, math, it tends to correlate with being higher in [unclear] ability as well.

[00:16:45]  Bruce: And that's called G factor.

[00:16:46]  Peter: Yeah, the G factor.

[00:16:47]  Unidentified speaker: Yes.

[00:16:48]  Bruce: So in fact, this is something that if G factor hadn't been discovered, I doubt we would have IQ today. When they first started to invent the idea of an IQ test, early in the 20th century, I guess, they had all sorts of tests, the tests of different things, and there was no real basis for why should we accept this test versus that test, until they started to notice that people that scored high on one test would also score high on another test, even though they were completely different ideas of intelligence. This led to the theory of G factor, which kind of grows up at the same time as IQ. IQ is perceived today as an attempt to measure G factor. And G factor means general intelligence, which is this idea that some people actually are, as Patel put it, better at a broad range of cognitive tasks than other people. IQ would be a measure of G factor, but G factor is this attempt to at once accept there are different kinds of intelligence, but for some reason they correlate.

[00:17:49]  Bruce: Now, here are some key arguments from Patel. While we can't account for the entire genetic contribution to IQ through known causal mechanisms, we have identified enough to know that genes can cause disparities in intelligence. If culture is the basis for all differences in intelligence, as Deutsch believes, between people, Deutsch must explain why twins separated at birth and adopted by different families with different cultures end up having highly correlated levels of intelligence, or why identical twins have a higher correlation of IQ than fraternal twins. Such studies show that intelligence is between 50 to 80 percent heritable, meaning most of the variance in intelligence between people is explained by genes.

[00:18:31]  Peter: I simply can't understand why people completely dismiss that as an uninteresting conclusion.

[00:18:36]  Unidentified speaker: I agree.

[00:18:37]  Unidentified speaker: I agree.

[00:18:38]  Unidentified speaker: Okay.

[00:18:41]  Bruce: Also, this wasn't in the article, but in his interview with David Deutsch where he actually challenges David Deutsch on this fact, he points out, for example, hardware correlations, the fact that size of brain 0.3 correlates with IQ. We haven't given the Deutschian viewpoint yet, but if the Deutschian viewpoint is correct, as Patel at least understands it, we'll discuss whether he understands it correctly or not, why does genes make such a big difference and why does things like hardware make a big difference? By the way, for those who are interested, episodes 47 and 48 of this show talk about how genes can influence us. When I get around to expressing my own view, which won't be in this episode, I'm going to use an approach very similar to how I did in episodes 47 and 48.

[00:19:26]  Unidentified speaker: [Unclear words] coerce us.

[00:19:27]  Peter: I love how you put that.

[00:19:29]  Unidentified speaker: Yes.

[00:19:29]  Bruce: Genes can and do influence us, often coercively. There might be a similar argument to be made here. Continuing Patel's summary of his view, animals are stupider than humans, but still capable of a limited form of creativity. If creativity is universal, then why do animals have it in a constrained form? Okay, so this one, to understand this, David Deutsch claims, as we've mentioned in past episodes, that only humans are creative and that animals are not, and that all the animals' knowledge is in the genes. Now, I have already taken a very strong stance in my past episodes that Deutsch is wrong about that, unless you're very narrowly defining creativity and knowledge in really specific sorts of ways, which after I've had a chance to talk with David Deutsch about this, I think that's what he's actually doing. But I don't think most fans of David Deutsch know that, right?

[00:20:18]  Bruce: I think that the very fact that he said animals, all their knowledge is in the genes and that they are not creative and things like that, that's what Patel's pushing back against here. Then he goes on to say, if only universal explainers can address problems which require novel solutions with a large search space, then why have non-universal programs been able to solve problems? This is in ML, which is some of Dwarkesh's background. Protein folding, writing poetry, creating art, beating the best Go players, discovering new drugs, a hundred other really impressive examples, and if you don't need to be a universal explainer in order to solve these problems, then why should we care about universal explainers? The general point I'm making is that you can build programs that are not true AGI's, but which are still extremely powerful. You can say they have no creativity, but if extremely important and previously impossible work can be done by, quote, uncreative programs, then this diminishes the value of your concept of creativity rather than the value of the programs themselves.

[00:21:19]  Bruce: I bolded that one because of all the arguments he makes. I overwhelmingly think this is his best argument. And I think he's spot on with this one. And you should know that by now, if you've been listening to the show, that I have used almost the exact same argument when I've been discussing knowledge creation and AI and things like that. This really is something where I think he's called out a legitimate problem that needs some explanation. Okay, he continues. Deutsch dismisses any example of animal cognition as unthinking imitation or genetic knowledge, which accumulates slower and has minuscule reach in comparison to explanatory knowledge. Deutsch is trying to explain complex animal activities by using Richard Byrne's theory of imitation as behavior parsing. This theory states that animals do not need to understand cause and effect relationships or have intentionality in order to fulfill multi-step procedures for gathering and processing food, building shelter, hunting, etc.

[00:22:16]  Bruce: Stating that a task involves statistical pattern recognition does not explain how a chimp extracts patterns of ordered subtasks just from observing instances of the whole task. How does the chimp even recognize that the action he is observing right now is an instance of a task he's supposed to learn to imitate? That's the end of my quote here, but he goes on to point out that if you could actually write an ML program that did this, you'd be rich, which is an argument that I've used in my podcast before he ever said this. He's right about that. There's a lack of explanation here. For Deutsch to just simply use this example, I don't think Deutsch is wrong if you listen to my episodes on this. Episodes 37 and 38 discuss this very topic. Episodes 39 and 40 go over Byrne's actual theory. I have some criticisms of Deutsch, but I don't think Deutsch is entirely off base here.

[00:23:11]  Bruce: But I don't think Deutsch is explaining everything either. There's an explanation gap and Patel is calling it out.

[00:23:17]  Peter: I guess for myself, I feel quite agnostic about some of these details you've brought up. And I am sort of an animal lover. I love my pet, but animals... Sometimes the Deutsch people will say animals don't have feelings, which I don't agree with. And maybe they can be creative in their own way. But I do think that Deutsch really gets at a fundamental truth, which is so often de-emphasized, to say the least, by scientists, which is that humans really are special in the natural world. There's something really different about our brains and the way we are creative and create knowledge and do incredible abstract things.

[00:24:01]  Bruce: By the way, one of Deutsch's arguments that I completely agree with is there's not really that much fundamental difference between our brains and a lot of animals, particularly the other mammals. So the difference has to be in what we would call software. Physically, they are the same, so why are they so different than us?

[00:24:19]  Unidentified speaker: Yeah.

[00:24:21]  Bruce: Also, by the way, episode 34 of this show, I give an example of how AI can be creative using AlphaGo. He mentions AlphaGo also. Then episodes 24 through 26 are my views on AI and knowledge creation. Patel is hitting along a lot of the very same things that I've talked about previously on the show. First, let's talk about Patel's arguments. We've been doing it as we go along. What's your overall thoughts about his arguments, Cameo and Peter?

[00:24:52]  Cameo: There's a lot I agree with. I mean, that's just my initial reaction is I think the things he's calling out, Deutsch on, are legitimate concerns with Deutsch's arguments.

[00:25:07]  Bruce: Peter, any thoughts you want to add?

[00:25:08]  Peter: Well, my overall, I think that the vast majority of what he says kind of rings true for me, but I guess I'm pulled in some different directions. I guess I fundamentally can't for the life of me see why the something like the universal explainer hypothesis, which makes a lot of sense in a more theoretical level. I mean, you know, even if it's extremely difficult to change our brains and develop new abilities, or we can still do it, at least theoretically. I mean, there's always a theory that could probably get you up to the next IQ point or something. And then the next one after that, maybe it's even harder. But there's a way in this natural world, which I think, as I think of it, it's the principle of optimism applied to the human brain. I mean, if something is possible under the laws of physics, then it's just a matter of having the right knowledge to get you there.

[00:26:12]  Peter: I strongly agree with that point. But at the same time, I kind of, I do feel like we are creatures of the natural world, and our brains are, you know, I look at my own children, and it just seems to me that they're, at some level, their personalities and their temperament and probably their capabilities, to a large degree, are really formed early on. You know, it doesn't mean that it's, that everything is set in stone. It's one of the beautiful things about humans is that we can change so much about the natural world, even our own brains. So I guess that maybe some of the harder edged things that Deutsch and some of his followers says about this, I don't quite ring true for me, but I think there's a lot of, there's kind of a middle path that we can look at that accepts a lot of validity from both sides.

[00:27:11]  Unidentified speaker: Yes.

[00:27:12]  Bruce: So a question that I kind of asked when we first recorded this, and I didn't maybe do it as well as I should have, but is Patel straw manning the Deutschian view, right? It's not immediately obvious that he's getting the Deutschian view right. So interestingly, since that original recording that we never released, Brett Hall just recently came out with a series of tweets where he explains the Deutschian view. So I'm now going to quote Brett Hall giving the other side of the argument. So now instead of me guessing at is Patel straw manning the Deutschian argument, we can actually have the Deutschians for themselves explain their viewpoint. So here is summary of the Deutschian view from Twitter. Forget degrees of intelligence. People all share the capacity to learn and to the same extent. It's what we choose to focus on learning that separates us.

[00:28:10]  Bruce: It's there in choosing what to learn the value judgments being made. Those who choose to focus on learning mathematics, physics, science, coding, medicine, law, language, et cetera, are ranked higher. Those who choose to focus on learning sports, dancing, influencing, trades, crafts, art, pop music, and soap opera storylines, office gossip, and politics are ranked lower. Said to be incapable of learning what they choose not to, unlike the previous group. There's no mystery why the first group, on average, performs better on IQ tests. It, IQ tests, selects the kind of knowledge, what they've chosen to learn, what the higher group has chosen to learn. But the capacity to learn period is common across all people. It's only what people choose to do with that capacity that differs. The desire to rank humans on the thing, making them qualitatively different from all other life is perverse.

[00:29:09]  Bruce: It is itself a misfiring of that very capacity to learn, to define, categorize, and explain. All explanations can contain fundamental misconceptions. IQ can be predictive precisely because the interests [and hence] choices some people make mean the knowledge they gain will help them score higher or lower on tests, asking questions about a particular kind of knowledge. Some kinds of knowledge are more useful for getting higher paid jobs, going into particular careers, and knowing what society values in scare quotes. But none of this means differences in capacity to learn. All knowledge can in principle be learned by any person because all people are defined by a universal capacity to learn, and that means learn anything. But no one will choose to learn anything or everything. They will focus on their very personal quirky interests. Of course, defenders of IQ just define intelligence as performance on IQ tests or some other tests by Fiat.

[00:30:10]  Bruce: Like all such claims, it's entirely circular and non-explanatory. Now, this is the overall point of view that he's expressing. He makes moral claims, you cannot actually hear some moral claims from the previous quotes I just did, but he makes moral claims based on this. He says, Karl Popper wrote, while differing widely in the various little bits we know, in our infinite ignorance, we are all equal. So minds cannot be rank ordered, unless we rank the little bits we know, i.e. we make a value judgment entailing some minds some people have less value. The idea some people have less value is a category error. People provide knowledge of more or less value to you or to solving a particular problem. You pay for someone else, someone's knowledge and time. But people are equally valuable, i.e. priceless. A person cannot be bought and owned. This is, I think, a pretty good summary of the Deutschian view.

[00:31:04]  Bruce: I've never heard it expressed quite so all in one place so straightforwardly, but just based on the bits and pieces I've picked up, this is pretty close to what I would have expected. Let's now discuss, I would like to know, Cameo, let's start with you. I'd like to know your initial reactions to this viewpoint that Brett is expressing.

[00:31:26]  Cameo: Like I totally get what he's saying. But it feels a little bit circular also to me. I don't know if you can talk about intelligence and then say, things like having skills with doing plumbing is the same thing as having skills with higher math functions, like the ability to do any number of really, really complicated math. It's not intelligence if it doesn't have to do with what we can do with our minds. Learning things is not actually the same thing as having intelligence. And I don't know that I believe that all peoples are capable of learning all the same things.

[00:32:24]  Bruce: And Peter, give me your thoughts.

[00:32:25]  Peter: Well, a lot of it rang true for me. I will say that. The end part about value, valuing, I think what you called the moral aspect of the argument, I almost felt like he may have been straw manning the sort of the IQ, the other side a little bit. Charles Murray makes an interesting point about this where he says that, well, his pushback against people saying that he's sort of dehumanizing people with lower IQ, is to say, well, do you think of people with higher IQs as better than you? I think that kind of rings true for me. I certainly don't. If I think of someone with a 160 or 170 or whatever, I don't even know what my IQ is, but people who are super intelligent, I might envy some of their abilities, but I would never think of them as fundamentally better than me. I wouldn't—wouldn't—I would say the same about people with IQs that are lower than me.

[00:33:33]  Peter: I've seen that in my own life experience teaching. Sometimes people will talk about the idea of a splinter skill, whereas someone with a lower IQ might just be amazingly good at just this one thing. I see that all the time. I mean, the idea of putting people in boxes based on their IQ I think can be a bit dehumanizing. But I see what he's saying there. But I think there's a way to accept the validity of the idea that IQ can be heritable without dehumanizing people. But a lot of what he said actually rang true for me too.

[00:34:16]  Bruce: By the way, I didn't capture this as a quote, but first of all, I should probably explain who Brett is, although probably people in the show know. Brett is a podcaster who is probably the largest of the Deutschian podcasters. He gets retweeted by David Deutsch all the time. His podcasts are excellent stuff. I actually highly recommend them. But he is really, in my opinion, very careful to match David Deutsch's view. I don't think he always does. He doesn't understand the material as well as David Deutsch. So sometimes he ends up going off a slightly different direction than David Deutsch does. But in many ways, I do see if Brett says this is the Deutschian view. Even if I've not seen David Deutsch say that, I tend to believe Brett, if that makes any sense.

[00:35:09]  Peter: Yeah, I've gotten the feeling there are probably a lot of back, behind the scenes, communication that have gone on for a year. I mean, maybe you're more tuned into that than I am that assume it's gone on for a year.

[00:35:21]  Bruce: There's at least some discussion for sure that has gone on there. It's not an accident that David Deutsch retweets Brett all the time. In many ways, Brett is the apologist for David Deutsch, and David Deutsch knows it. Some people refer to Darwin's Bulldog.

[00:35:37]  Peter: Yeah, I was about to say that.

[00:35:38]  Bruce: To Brett as Deutsch's Bulldog.

[00:35:40]  Unidentified speaker: Yeah.

[00:35:41]  Bruce: So there is a fair question though, because Brett, I don't mean this in a negative way at all. I really don't. I use the word apologist. Brett is not a professional philosopher. He's a podcaster, he's an influencer, and he does a really, really good job of understanding Deutsch's viewpoints, and he communicates it incredibly well, in some cases better than Deutsch does, and he's done a ton of thought into it. But when you actually talk with Brett, he does not understand the alternate viewpoints very well at all, which is why I think he qualifies more as an apologist. He's not a scientist trying to understand the viewpoints and trying to get at the truth. He's totally in the Deutsch camp, if that makes any sense. Now, the world needs great apologists like Brett. There's nothing wrong with that. There's nothing wrong with that. We really benefit from having someone who just mines out a single viewpoint and tries to defend it to the best of their ability.

[00:36:40]  Bruce: We just did episodes on dogmatism and said even dogmatism is somewhat okay for exactly that reason. For the exact same reason, there's nothing wrong with Brett deciding, I'm in the Deutsch camp, I'm going to defend it to the hilt. But Brett doesn't necessarily have a strong understanding of alternative viewpoints. Because of that, whereas Deutsch does, because of that, there is the possibility that Brett is getting Deutsch's view wrong. In this case, I'm going to argue that he isn't. Here's why. In Deutsch's books, Deutsch talks about this. Not a lot, certainly not to the degree Brett does. But here is what Deutsch actually says, talking about Jared Diamond's books. He says, Diamond says that his main reason for writing guns, germs, and steel was that unless people are convinced that the relative success of Europeans was caused by biogeography, they will forever be tempted by racist explanations.

[00:37:32]  Bruce: Well, not the readers of this book I trust. Then he goes on to say, in reality, the difference between Sparta and Athens had nothing to do with their genes, but their ideas were different. Now, you might say, how am I getting out of this what Brett said? It may not be immediately obvious. But Deutsch is stating here that you cannot explain differences between race in terms of IQ. That's what he's saying. He's saying that if you've read this book and you understand universal explainership, then IQ just doesn't matter between races because we all have the same capacity. You can't even make sense of Deutsch's argument here unless you actually have something like what Brett is saying in mind. I actually think Brett is accurately portraying Deutsch's viewpoint. Now, me and Brett actually had an exchange over this Twitter thread where I kept referring to it as Brett's theory.

[00:38:28]  Bruce: And he kind of got a little upset with me and said, it's not my theory. I agree. It's not Brett's theory. It's Deutsch's theory. I completely agree with Brett that he's not the one who invented it. I'm going to continue to say Brett's theory though. And hopefully we can all be matured enough to understand that what I mean is the theory he was advancing in this thread. And I'm just using a shorthand.

[00:38:49]  Peter: Can you just say that culture matters more than biology, but that biology still has something to do with it?

[00:38:57]  Bruce: Here's the problem with saying that. By the way, I agree with what you just said. So I'm not arguing against what you just said. I'm explaining why it doesn't fit well with what Deutsch is saying. If you admit that there is other factors that matter other than universal explainership, then you've reopened the possibility that IQ is a real thing, and that Jared Diamond's concern with the quote, racist explanations is a thing. If IQ really exists and there really is cognitive capacity differences between humans, then you don't have an explanation for why there is no cognitive capacity difference between one race and another. There could be because race is at least somewhat related with genetics. So when Deutsch is trying to put that down, the idea that his explanation of universal explainership undercuts that whole viewpoint as a necessity, it undercuts the very concept of IQ.

[00:39:53]  Bruce: I've actually seen statements from Twit Rats to the effect of the left claims there are no IQ differences between races, but they don't have a good reason for why that's true. They'll argue there are no IQ differences between races, but they don't have an explanation for why there wouldn't be. Why couldn't there be IQ differences between races? Well, from the Twit Rats viewpoint, that explanation is universal explainership undercutting the whole concept of IQ. Yes, of course, a completely valid viewpoint might be culture matters the most. In fact, that's something that I would agree with completely. But if you're going to admit that there are other factors, then the moral argument that Deutsch is making here has become undercut. You have refuted it, in fact, at least to some degree. Then you have to find a different explanation. I think that's why I don't know that currently, at least, this is why Brett's taking such a strong stance.

[00:40:56]  Bruce: You can't disconnect the moral explanations from the scientific explanations in a lot of cases. When you're talking about a cultural belief system like Twitter, Deutschianism, Twitter critical rationalism, they're all tied up together, which is why there's maybe even some resistance. The very fact that I'm willing to make an episode and look at Patel's viewpoint and challenge the Deutschian view, that's going to suddenly be a little scary for people who buy into the moral worldview that comes from this. Now, having said all that, let's be honest, IQ really does have a really, really controversial history. And the Deutschians are hardly the only ones who have really big problems with theories around IQ. And again, let's be honest, left has giant problems with IQ these days, to the point where if you try to, if you're kind of a left-leaning scientist, even talking about IQ and studying it is looked down upon.

[00:41:59]  Bruce: So Dwarkesh Patel actually interviewed someone who was an expert in this. When I say this, I'm not even sure what I mean. He was actually someone who was using, he interviewed Stephen—I don’t know how to pronounce this—H-S-U, and the title was Intelligence, Embryo Selection and the Future of Humanity. So this guy is, by the way, excellent interview and super interesting stuff. The guy he interviewed is an expert in using machine learning to work with genes. I'm not sure I remember enough to give a more detailed explanation to that at this point. I listened to it a while back. One of the things that Stephen mentions in the interview is that they can measure IQ in a 15-minute test that's highly accurate. And right now, when they're doing the genetic studies with machine learning, nobody measures IQ. But we could be collecting all sorts of data on how genes relate to IQ.

[00:42:59]  Bruce: And because of the left, there's so much terror over doing that, that they're intentionally not doing it, even though it would be very cheap to do so. And we could actually be doing all these great studies into how do genes actually relate to IQ. And we're just not even doing them at all because of political reasons. Having said that, I don't think the left's entirely wrong. I mean, IQ is controversial enough in history, and there's enough problems with the theory that, yes, they're probably doing it mostly for political reasons, but they're not completely out of left field either, right? There really are concerns around, similar to how the twit rats have gone about handling, the Deutschians have gone about handling this, that IQ may not be a fair measure. It may just be measuring certain kinds of knowledge, and we're just valuing kinds of knowledge and things like that.

[00:43:53]  Bruce: So this is a place where normally Deutschians are very much at odds with the left. This is one where there's at least some overlap with a leftist viewpoint.

[00:44:03]  Peter: I think it's somewhat similar to culture in that fundamentally, it's even if we accept the fact that culture matters a lot, that it's still irrational to judge an individual based on what culture they come from. I mean, we have to just look, the way to get around all this morally is to just to look at people as individuals. That's the only thing that makes sense.

[00:44:28]  Bruce: So here's what I'm going to argue though, and it's okay if you want to disagree with me, because I don't think my argument is completely certain. But my argument is very simply this. I think this idea that Brett is expressing, the theory that I'm calling Brett's theory, does in fact come from David Deutsch, and that this is David Deutsch's opinion. I don't know that for sure. Like I said, Brett does sometimes, perhaps not intentionally, but departs from David Deutsch's views. But I don't think this is one of them. And so I have a relative level of confidence that this idea does in fact, come from David Deutsch and from his books, and that this is something that he would agree with. Who knows for sure? I'm open to the possibility I'm wrong here. But I'm going to continue to call it Brett's theory, just in case I'm wrong. If nothing else, in case I'm wrong.

[00:45:17]  Bruce: And just because conveniently that means I'm referring to specifically what Brett was expressing on Twitter. So now I interacted with the Twitter critical rationalists over this. And a number of them talked with me about their views, with me challenging their views. Furthermore, let me just say that the ones I talked to are without a doubt the cream of the crop of the Deutschian worldview on Twitter. That is just not true for some. I mean, some of the Deutschians on Twitter are just annoying. Really pretty bad. That was not who I was talking to. I was talking to people who have really shown open-mindedness in discussion. Don't just get angry in discussion. Really do think this through. Are willing to differ with Deutsch on things if they feel like Deutsch is wrong. I feel like I was getting the cream of the crop responses from the Deutschian worldview when I asked them, okay, what if somebody said this?

[00:46:16]  Bruce: What if we tested this? Again, you can decide for yourself how good the responses are. And I'm not even going to express my opinions right now. I'm going to try to just simply list out what I said, what I saw as challenges and explain as best I can, as steelman as I can, what their responses were. So first of all, if IQ, so Brett's theory is that IQ tests, IQ is really just measuring interest in certain kinds of knowledge. Well, now if that's true, if that theory is, if we really take that serious as a theory, then it seems like we could test it. It seems like what we could do is we could test children with IQ tests before they've clearly taking an interest in mathematics, physics, and science versus sports, art, etc. We can see if IQ is stable or not. If this is partially just our perceptions that we value certain kinds of knowledge over others, then testing children before they're clearly in, I'm going to be a mathematician versus I like sports at a young enough age.

[00:47:22]  Bruce: We could test them, then we could test them again after they've now made a choice and we could see if IQ is stable or not. If you know anything about IQ, you know they've done tests like this. I'm not even sure.

[00:47:32]  Peter: It is pretty stable, right?

[00:47:34]  Bruce: From what I understand, it's relatively stable. But we can argue whether the tests are valid or not. But my point was not, I'm trying to put out there this test has been done and you've refuted, I'm trying to say, why would you not accept this as a valid test? It seems like this directly takes your point of view seriously, takes the consequences of your point of view seriously, and comes up with a legitimate test. Shouldn't you accept this as a valid test? That if IQs are stable before interests are chosen and then after, doesn't that refute Brett's viewpoint, Brett's theory? Then I also said, couldn't we test Brett's theory? So he basically said that people are interested in sports, dancing, influencing, trade, crafts, arts, pop music, soap opera, storylines, office gossips and politics, that we see them as lower IQ because we less value that kind of knowledge.

[00:48:29]  Bruce: Again, it seems like you could test this. Let's say you took people who are actors or into art or something like that, and we wanted to see who dominates those fields, even something like office gossip, who dominates the office gossip in a certain office or across offices. We could then find out if the people dominating them are higher in IQ or not. We could test it. We could see if IQ influences those kinds of knowledge or not. If we find that all categories, including the ones Brett is saying, are dominated by high IQ, that seems to me like it refutes his theory. So it seems like his theory is testable. Now again, I'm sure that everyone was aware that they've done tests like this. At least not maybe for office gossip, but yeah, actually, actors tend to have really high IQs. Because to be a really good actor is a cognitively demanding task, and Tom Cruise has a really high IQ, probably higher than mine by far.

[00:49:27]  Bruce: So people who dominate even the fields that Brett is talking about, tend to be higher IQ. Not necessarily always, you could argue that that's not true for sports. But I suspect that it would depend on the sport and depend on the position. But you would probably find that good quarterbacks have much higher IQs than not as good quarterbacks or something along those lines. Pretty sure they have done studies in this, by the way. A fullback or something, maybe not so much. Do you think Michael Jordan has a low IQ? I doubt he does.

[00:50:01]  Unidentified speaker: I don't know what his IQ is.

[00:50:02]  Peter: The decisions he's making on the court and how quickly you're making those decisions. I mean, I think there's a lot going on in his mind, for sure.

[00:50:12]  Bruce: Then I also suggested, couldn't we measure neutral factors that have nothing to do with specific field? For example, and I just threw these out there. Off the top of my head randomly, not. Fluid reasoning, knowledge, quantitative reasoning, visual spatial processing, or working memory. I threw those out there. What if we just measure those? It's got nothing to do with specific fields or what we value or we don't value. What if we found that these correlated? Now, of course, I'm sure everybody knows that what I did is I just looked up G-factor, looked up what they measure, and listed them. Of course, we already talked about G-factor does correlate. That's what led to the continuing theory of IQ, was that they kept finding that when you try to measure different types of capacities that really could have been unrelated, that they turn out to highly correlate.

[00:51:01]  Bruce: If you found that these factors do not correlate, let's be honest, IQ would be dead as a theory. It would be gone. That is an example of how IQ as a theory, the prevailing theory, there is a difference in cognitive capacity between humans, could have been refuted and wasn't. It was corroborated by an actual empirical test. Shouldn't the opposite, therefore, also be seen as refuting Brett's theory? I raised each of these issues with them. Here was their response. Let me do my best to try to explain their viewpoint. The Deutschians denied these tests would matter. They doubled down on a few arguments previously mentioned. For example, they said that it just shows they did well on a particular test. It doesn't mean that it actually correlates to their capacity to understand explanations. It just shows that they are not interested in the subject being tested.

[00:51:55]  Bruce: A key argument that was made multiple times, by the way, was that since the prevailing view offers no detailed explanation as to what cognitive capacity actually is, and instead is just about a correlation, that that means it has no explanatory power. Therefore, in their minds, it does not count as a good explanation and it's dead, as far as they're concerned, leaving theirs as the sole surviving explanation. And then they would ask, how would any of these tests account for the fact that maybe some people just don't care about the test and don't try very hard? Because they're not that interested in it. And perhaps this is the main difference in IQ is level of interest in the test and how much effort to put into it.

[00:52:39]  Peter: I mean, it kind of makes sense. And I can't say they're completely wrong there, but to someone who's like an IQ researcher, I mean, I think they would probably argue that they've, oh, they've accounted for this and they have, you know, 50 different ways to get around that. And I don't know.

[00:52:57]  Bruce: I'm trying to not express my view too much, but basically attempt to say, you know, this shows that IQ tests don't matter. That seemed like a testable thing, right? That you could come up with ways to try to test that. Of course, an IQ researcher would try to take something like that into consideration. Whether you agree with how they did that or not would be a different question. So then I gave them examples of refutations. I said, okay, let's talk about actual real life problems with your theory. So I said, my neighbor that lives down the street from me, he has Down syndrome as well as other fairly serious issues. I mentioned in the previous recording, I mentioned him quite a bit. He is severely mentally challenged to the point where he can't take care of himself on his own well. Not atypical of someone that has Down syndrome and autism.

[00:53:45]  Bruce: Yet he can talk, he can engage in conversation, he can understand what you're talking about, as long as you keep it simple. He has interests, he has things that he likes a lot, and that he doesn't like as much. Furthermore, Down syndrome is known to be a genetic disorder caused by when abnormal cell division results in extra full or partial copies of chromosome 21. This is from the Mayo Clinic. We even have an explanation for why, how genes cause Down syndrome. And this somehow Down syndrome people are lower IQ, always. We know there's this really strong link between genetics and have a genetic explanation and low IQ in this case. Doesn't that refute Brett's theory? That was what I was asking. And then I said, we know taking drugs or being, I don't think I actually brought this one up, but this one I should have brought up. No, I did bring this one up.

[00:54:43]  Bruce: We know that taking drugs or being drunk or high reduces one's cognitive ability. How is it, how is this fact, the fact that I can go take painkillers or alcohol or something, and my cognitive ability reduces, how is that even a meaningful statement in the Deutschian view if the only capacity that really matters is whether you're a universal explainer or not. What's going on from the Deutsch, in the Deutschian viewpoint, when I get drunk and my cognitive ability, and we say, quote, my cognitive ability reduces, if it's not actually my cognitive ability that is reduced, what is it that is being reduced? What's your explanation for this? Then I pointed out, one of the main things they argue is that it's all just about interest and time. A baby doesn't have strong cognitive ability, but they just haven't had the time to develop it yet, no matter what you happen to be interested in.

[00:55:38]  Bruce: We've mentioned that one several times. One of the things I pointed out because they kept bringing that up was cognitive decline due to aging. How can you explain cognitive decline due to aging in terms of time and interest? Because now we're talking about someone who's had more time and is clearly still interested, and yet for some reason we can measure a cognitive decline in a person as they age. What is it we're measuring? How does the Deutschian worldview explain that? Furthermore, these are people who used to be younger and used to have higher intelligence, and they can tell you, self-report. Yeah, my intelligence is that my cognitive ability is lower now due to the fact that I am older, even though I've spent more time on the material, even though I'm still interested in it.

[00:56:24]  Peter: It seems to me that all this is getting at a view of the human mind which might be a little bit different than the Deutschian view, but it's one I also sympathize, which says that there's a way to push back against this, where you say, oh, well, these people with disabilities or cognitive decline, they're in another category and their brains are just broken, but the universal explainers are for everyone else. But then there's another way to think of it that maybe we're all a little bit disabled in different ways, and I see that in my own life. I just have, there's just certain things, like I think I brought up a sense of direction before, or some people just can't hear musical pitches. There might just be a thousand million things in our brains that we just have different. Maybe in some areas, we're really high in and then we have certain things where basically, even if we don't have an identifiable disability, we're probably all disabled about this or that, and that's just how our brains work.

[00:57:36]  Bruce: By the way, Cameo in the previous episodes that we recorded, she brought up the tone-deaf thing, which I thought was an interesting example. The other example that this is the one I didn't bring up that I should have, but I brought up in previous conversations with the Twit Rats, is people can lose their sanity. I gave the example of a BYU professor that I know, who has a paper he published about how he went insane temporarily. He had a virus in the brain, and he actually describes what it was like to be insane and to come back out of it again. And an insane person is clearly still a universal explainer. They can still understand things, but there is something just off with an insane person. In this case, we're not even talking about a lifelong mental illness. We could challenge whether the word mental illness is an appropriate term or not.

[00:58:30]  Bruce: We're talking about someone who very literally got a virus in the brain, lost their sanity temporarily, and then in a few weeks, they finally found a way to deal with the virus and came back to his senses. How would you go about trying to... Clearly, while he's insane, he cannot error correct well at all in certain areas. And what's the explanation for that within the Deutschian worldview? How could you possibly explain something like that in terms of just interest and time? So now I raise these issues with them and here's what they responded with. The Deutschians, so you mentioned something and I was actually expecting them to say this, and I was a little surprised that they didn't. I would have if in our previous recorded episodes that we didn't release, I took the stance that my neighbor didn't count because my neighbor was clearly because it's Down syndrome and we understand exactly what's wrong with the chromosomes, that he is not in fact a universal explainer, and therefore isn't what we need to consider in terms of, he's just a different category because of that, which is basically what you just said, Peter.

[00:59:41]  Bruce: I was expecting them to say that, and if they had said that, that could be used to explain a number of the ones I just gave.

[00:59:47]  Peter: I've heard people make similar claims.

[00:59:50]  Bruce: So one person seemed a little open to that possibility. I'll mention that in just a second. But for the most part, the Deutschians took the stance that my neighbor was still just a case of time and interest, which surprised me a little. They would ask, how is my neighbor different from a child? Is a child's capacity less than adult just because he had less time to learn? One suggested that this was not any different than if a normal person feels sick and therefore has an ability to create knowledge temporarily reduced. But it doesn't mean they don't have the same or equal capacity. I quickly pointed out on that one that actually that did mean they didn't have the same or equal capacity. And they said, I'll have to think about that. That was the one exception. Everybody else kind of really did take a hard stance. There's no real difference here.

[01:00:43]  Bruce: One discussed a severely mentally challenged family member that could play video games really well, kind of similar to what you were talking about, Peter, better, maybe better than someone with a quote normal intelligence level. And they claim this refuted the prevailing, could be used to theoretically refute the prevailing theory. His argument wasn't really that it refuted the prevailing theory. He was really saying, your arguments that these refute our theory are no better than me claiming that the fact that my family member can play video games really well, I could claim for the same reason that it refused [apparently, refuted] the prevailing theory. Then they would ask, why can't you explain the difference with your neighbor in terms of just being a matter of time and interest? They also reused quite a bit, the prevailing theory is not explanatory, so it doesn't count argument.

[01:01:33]  Bruce: They would say, look, it just doesn't matter because the prevailing theory is just a bad theory because it's not explanatory. It's just a correlation. It doesn't have, it needs to have, one of them said, it needs to have a strong explanation that explains what intelligence is in terms of function of the mind, and it needs to explain how IQ measures that. And since it doesn't do that, it's just a bad explanation and it just doesn't count, and ours is the surviving explanation. And then by comparison, they claim that their explanation, Brett's theory, is an explanation and that it has reach and therefore, it is the best explanation. Now, this is something that we did not discuss. Actually, before I go on, let me ask you guys' thoughts on my pushback to them and their pushback to me.

[01:02:28]  Peter: It seems to me, again, I guess I keep coming back to this, that what they're saying has some validity to it, but I just can't like, going back to the twin studies, this idea that non-genetically related siblings raised together have the IQ does not, and not even just IQ, every single personality characteristic there is, from my understanding at least, little, very little or close to zero correlation between the siblings, whereas siblings or twins raised apart, everything is highly correlated in terms of their abilities and their IQ, and there's every way you can measure the human mind. I can't quite, as much as sympathetic as I am to these universal explainer arguments and basically everything that Brett Hall is saying, I just can't quite get my mind around why that's just as a parent and a teacher, and just someone who just wants to know about human brains, why that's not an interesting fact about our world, and why this idea that this stuff is probably, at least heavily influenced by genes isn't a good explanation.

[01:03:55]  Peter: I guess maybe I just don't quite get it.

[01:04:02]  Cameo: Well, I can't agree that enough time or... Like your neighbor, it's not a time and interest.

[01:04:12]  Bruce: He's in his 30s, right?

[01:04:15]  Cameo: Even if he put all of his focus on changing, evolving past the point that he's at, evolving is not the right word, changing that capacity, I don't believe that's something that he'll ever change. And I don't agree that it can be time and interest. I just, I don't know.

[01:04:40]  Peter: I mean, certainly interest has a lot to do with what our capabilities as well. I mean, I really like the point that these are intertwined kind of phenomena. I mean, if you're not interested in an IQ test or anything, you're probably not going to score well. And maybe you have other interests that you would, that are just as meaningful and great.

[01:05:01]  Cameo: But I gotta say, it's a shoddy explanation personally.

[01:05:08]  Bruce: Brett's is, you mean?

[01:05:10]  Unidentified speaker: Yes.

[01:05:10]  Unidentified speaker: I'm not sure I would go that far, but okay.

[01:05:13]  Cameo: Well, and I don't mean, I like the, I like their focus on, you know, that I understand all the problems with our more traditional view of intelligence and intelligence tests and all of that. But the argument that somebody who has Down syndrome, that it's only about their time and interest, that's a really hard argument to get behind.

[01:05:45]  Bruce: I did not discuss this with them. I mean, everybody has to decide where they're going to spend their time. And this was a long discussion. And so eventually it kind of petered out. This is what I would actually wish I could discuss with them. Maybe I will at some future point. So the argument that the capacity to understand an explanation is, the argument is ultimately that the capacity to understand any explanation is the sole important capacity in intelligence. Unless maybe you want to count time and interest as a capacity, which I think you could make the case that those count as capacities too. That argument is not unlike claiming that the capacity of a universal computer is solely in its instruction set, whether it's universal or not. Now you go back to the episodes we did on computational universality. There are different kinds of computers.

[01:06:33]  Bruce: Some are universal, the Turing complete ones, and some are not. And they can't do certain kinds of algorithms. They're incapable of doing certain kinds of algorithms, things like finite automata or whatever. But imagine using the argument that they are using with universal computers, right? You say the sole form of capacity of a universal computer is an instruction set. So it's just a matter of is it universal or not? And you could use the arguments that they're using. You could try to explain everything in terms of, it's just a matter of the software, right? You could say, yeah, sure, this computer is slower than that one, but it's just because that's not as good an algorithm. You could use identical arguments with what they're using to try to justify a theory that universal computers, their sole capacity is their instruction set. But we know that's wrong.

[01:07:26]  Bruce: Nobody would make that argument because we know that's wrong. We know universal computers actually have three types of capacity. The power of the instruction set, whether it's universal or not, what types of algorithms it can actually perform, and then memory and speed. Why should universal explainers have only one single capacity, if universal computers don't have a single capacity? Now, I actually know what the Deutschians would say for this, because I've seen bits and pieces elsewhere. I've seen what David Deutsch has said about it. Let me try to explain their argument against that argument. Their argument is that potentially universal explainers could have memory and speed capacity differences, because the brain is a universal computer. In theory, universal computers can have memory and speed capacity differences. They would, at least in theory, buy that there could be a speed of memory capacity difference between humans because they could have different brains.

[01:08:24]  Bruce: However, they would argue that due to computational universality, those are the only possible capacity differences. We already know from that theory that there's only three capacities possible and that these three capacities are the only ones that could possibly apply to a human as well between differences between brains. Then they would make the following argument. Say, yes, potentially, you could have a difference in speed or memory for one human compared to another. But in practice, we shouldn't expect to see that because human brains are not like CPUs. There's no Moore's law of human brains going on, where we end up with every two years, double the number of transistors. Human brains should all be roughly equivalent in speed and memory. And therefore, while there could be some speed and memory capacity differences between humans, they should be mostly just negligible and really wouldn't ultimately make a difference.

[01:09:22]  Peter: It seems they oftentimes will bring up, well, you hand someone an iPhone or something like that, then that becomes practically an extension of their brain.

[01:09:31]  Unidentified speaker: I know I feel that way sometimes about my phone.

[01:09:34]  Bruce: Brett makes that argument as part of the thread. Somebody else whose name I can't pronounce because it's a string of letters, challenged Brett on this thread and said, well, what about working memory? Working memory is completely neutral to what area of interest you have, but it would benefit from working memory. And we can measure that different people have different amounts of working memory. So why wouldn't that make a difference where some humans have higher IQs, maybe for multiple factors, but one of them is working memory. We know humans have differences in amount of working memory. So why wouldn't that then translate to IQ differences, where we can actually measure cognitive differences between people? And Brett's response was, it wouldn't matter because once you get to the point where a universal explainer and you can write things down, working memory is not that important anymore.

[01:10:30]  Bruce: And that was his response. So based on all this, I then pointed out after having them argue with me for quite a while, I said, it looks to me like you're saying, I've just offered what seemed to me to be several very testable consequences of your theory. And you're claiming that they're not valid tests. Whether I agree with you or not, that doesn't matter. You're making the claim that they're not valid tests. Are you making the claim there are no valid tests of this theory, and thus your theory is metaphysical? Now, I got an interesting response on that. Brett actually became a little offended and put up in a fun way though, and put up a picture of Kathy Newman saying, so you're saying, okay, which was really strange because he then went on to admit that his theory makes the same predictions as the prevailing theory just explains them differently, that there are no testable differences between the two.

[01:11:25]  Bruce: So it seems like I was actually correctly summarizing his view, which is very different than what Kathy Newman did. And then he never did respond after that. One of the other ones I was talking to, he took exception to the term metaphysical. He said, I would call it a philosophical theory, not a metaphysical theory, which was also interesting because Popper, his demarcation is supposed to be between empirical sciences and metaphysical theories. That's what Popper calls it. That's the Popperian term. If your theory has no testable consequences, Popper calls it metaphysical. That's the critical rationalist term for it. He didn't like the term metaphysical, but he said, I would call it philosophical. He says, if you by metaphysical mean, it has no testable consequences, which again, that is the Popperian definition of metaphysical, then yes, I would say this is a metaphysical theory.

[01:12:15]  Bruce: There seemed to be agreement, even with Brett who took some exception to calling it metaphysical, that this is in fact a completely non-empirical theory. They did not push back on that once it was fully explained. They said, yes, there is no way to test this theory, but this theory is a good explanation. It has explanatory power, and that's why it's a good theory. That's why it's a better theory, and the other theory has got no explanatory power. It's refuted. It's not a good theory. This is the good theory. I think this is, as best I can, a steel man version of their argument and how they responded to what seemed to me to be some fairly legitimate pushbacks that you would expect them to get. I've tried to capture what their responses were. At this point, I think probably we can wrap up this episode. Let me just give you guys a chance to make any final thoughts or comments on this.

[01:13:13]  Peter: We didn't talk about my favorite part of the article, which was the universal bicycle.

[01:13:19]  Bruce: I never did justice to that, did I?

[01:13:21]  Peter: I think about that when I bike to work every day. I don't think I'm saving the earth or anything, but just for health reasons and I just like it. I think about the universal bicycle aspect all the time.

[01:13:34]  Bruce: Let's talk about that one. Because I don't want to forget it like I did the last time. Let's talk about it right now.

[01:13:41]  Peter: Basically, the idea, if I could just explain it, and then I'd be curious on your thoughts about it, it's a criticism of the universal explainer hypothesis. The idea is that, well, how is that any different than saying you have a universal bicycle where if you take that bicycle and you change this part, you put this part, put a jet engine on it, suddenly you have a spacecraft capable of going the speed of light. But it really is totally theoretical. You still just on a practical level, you have a bicycle and you're not going to go very fast and you're not going to go very far. Is that fair?

[01:14:23]  Bruce: Yes, it is. I think that's exactly the criticism he's getting at is that even if you accept that universal explainership in some sense is true, it doesn't change the fact that the prevailing theory could still be true. That was one of his better criticisms. I probably should have captured that one, particularly since I knew you liked that one. Thank you for bringing that one out. I'm not going to share my opinions. Not like my opinion is a secret. I've made it fairly clear as I'm going along. But I actually would like to make the next episode or at least a future episode be where I actually go through each of the potential reputations [apparently, refutations] that I offer to them, and take a look at their push backs, and talk through how I feel about their theory and how I feel about Patel's theory. One of the things that we didn't mention, and this can be in the next episode, is that Patel actually tries to offer a more specific alternative theory.

[01:15:27]  Bruce: Basically, he offers a combination of hardware plus the scaling hypothesis. I don't have time to explain exactly what that means. But he believes that the differences in intelligence can be explained by the fact that we actually have hardware differences and that the brains are not equivalent. On top of that, that leads to a problem, because if the differences are in hardware, why are animals not on the human bell curve? He explains that through something called the scaling hypothesis, which is that as you're starting to scale up the number of neurons, the giant leaps suddenly take place at certain levels. Then he likens that to the difference between GPT-2 and GPT-3, or these Midjourney or these Stable Diffusion, these new machine learning algorithms that are coming out, that do these amazing things like being able to type a scene, a description, and it can draw a picture of it.

[01:16:26]  Bruce: This wasn't possible at lower levels of neurons, even though we had the architecture, even though we had the right concepts, you just needed a bigger computer and more data to be able to process, to be able to actually make these cool new algorithms. Even though we technically had the same neural network architectures previous to this point, even though we technically could have done these, it just wasn't actually feasibly possible until we had a big enough computer with enough compute and with enough data and enough ability to actually run that data through the system until these giant leaps came out. His theory is based around hardware plus the scaling hypothesis, basically. I should note that that's different than the prevailing theory. The prevailing theory is actually indifferent why IQ exists. This is where the Deutschians are partially right.

[01:17:24]  Bruce: There is a lack of explanation in the prevailing theory at certain points. They don't really necessarily say what IQ is. They don't necessarily try to explain it, or even how it happens, or why it happens, or how it's related to brain function. They're simply trying to measure that there is a cognitive difference. So Patel is trying to fill in that gap with his own explanation, if that makes any sense. So maybe next time on the next episode, we can talk through what I think is actually wrong with Patel's explanation. I think Patel's explanation is really actually quite problematic, but I also think that the Deutsche one is. And we can kind of talk through the pros and cons of each theory. One of the things that I had suggested to the Deutschians I was talking to, is I said I actually think both theories are wrong, but that both have some verisimilitude.

[01:18:17]  Bruce: That's not exactly uncommon, and that fits perfectly with Popper's epistemology. Is that two false theories may both capture some sort of different truth from the other theory. But what is that? And like I can see the Deutschians, one in particular who kept bringing this up, just saying you can get great [unclear wording], Bruce, explain it, right? And he's right, that's a hard thing to do. It's hard to explain because we know so little about intelligence at this point. It's very hard to explain at this point. So there is a lack of explanation in the prevailing theory, and they're right about that. We can probably stop here. This would be a good stopping point.

[01:18:54]  Peter: Thank you, Bruce. That was wonderful. I really enjoyed that.

[01:19:00]  Bruce: The Theory of Anything podcast could use your help. We have a small but loyal audience, and we'd like to get the word out about the podcast to others so others can enjoy it as well. To the best of our knowledge, we're the only podcast that covers all four strands of David Deutsch's philosophy as well as other interesting subjects. If you're enjoying this podcast, please give us a 5-star rating on Apple Podcasts. This can usually be done right inside your podcast player, or you can Google The Theory of Anything podcast Apple or something like that. Some players have their own rating system, and giving us a 5-star rating on any rating system would be helpful. If you enjoy a particular episode, please consider tweeting about us or linking to us on Facebook or other social media to help get the word out. If you are interested in financially supporting the podcast, we have two ways to do that.

[01:19:46]  Bruce: The first is via our podcast host site Anchor. Just go to anchor.fm slash four dash strands, F-O-U-R dash S-T-R-A-N-D-S. There's a support button available that allows you to do reoccurring donations. If you want to make a one-time donation, go to our blog, which is fourstrands.org. There is a donation button there that uses PayPal. Thank you.