Episode 80: Knowledge vs. Simul-Knowledge
- Links to this episode: Spotify / Apple Podcasts
- This transcript was generated with AI using PodcastTranscriptor.
- Unofficial AI-generated transcripts. These may contain mistakes. Please check against the actual podcast.
- Speakers are denoted as color names.
Transcript
[00:00:08] Blue: Welcome back to the Thurman Anything podcast. Hey Peter. Hello, Bruce. How are you doing today? Good. All right. Well, this is actually going to be our final episode diving into the theory of knowledge. Deutsch’s constructor theory of knowledge and his beginning infinity theory of knowledge. But you said you had a question you wanted to ask me. So should we start with that? Sure. If that’s okay with you.
[00:00:33] Red: Sure. I thought perhaps it could kind of tie in what we’ve talked about in the last two or three episodes to what we’re exploring today. So the two sources of this hypothesis, this idea that there are two ways that at least we currently know of, of creating knowledge in the universe, Darwinian evolution and human brains, human minds, which, you know, I think it’s, I still think it’s a pretty compelling way to look at, look at it. And it just, I mean, I don’t know. I’m not, obviously I’m not down into the weeds with it as much as you, but it seems it still is something that kind of rings true for me. I’ll say that, but not to say that you haven’t done a very nice job sort of deconstructing it a bit, exploring it a bit, and, you know, really trying to tease out if that’s true or not. But getting back to your walking robot, or I say you’re, it’s, it’s gorgeous, I think, walking robot thought experiment. Although
[00:01:38] Blue: I am greatly abusing it to its maximum potential. So you can call it my walking robot if you want.
[00:01:45] Red: Okay. The, we’ll say the walking robot thought experiment where, as I understand it as, you know, a layman, not a computer guy, is that the, there’s an evolutionary algorithm running on this robot to help the robot through trial and error, I guess, learn how to walk properly. And so your sort of assertion, I guess, is that it seems like this is kind of a way of creating knowledge that would move outside of the two sources hypothesis, which, you know, I’m not going to say, argue it’s not. I think you put it in a really compelling way. But I guess that kind of what still sort of rings true for me, regardless of exactly how we define what knowledge is or isn’t, is that the human that is kind of still bringing the knowledge to the table. And the walking robot could, could more be seen as something like a sophisticated calculator. So, you know, kind of analogous to, you know, when an engineer has a extremely sophisticated calculator, they might not even know, like, all the math involved on what they’re, what they’re doing. And I mean, I guess I assume someone does, but they might not know, but they can still use it to create knowledge. I mean, couldn’t the walking robot thought experiment evolutionary algorithm be seen as just a highly sophisticated calculator? Am I making some sense there? You are. That’s a good question. OK.
[00:03:31] Blue: The answer is not what I’m saying maybe doesn’t come across the most compelling in the sense that it still just seems like there’s just something important missing. And I even kind of in our previous episode admitted that, right? I basically said, yeah, when it comes right down to it, humans are in fact doing something special. And I don’t really deny that in the slightest. I completely accept that human beings are doing something special. I my my feeling is, is that it’s something special about the search algorithm itself, not physically what it produces knowledge itself. And I think that’s a better way to look at it. And yet I can totally see Peter where you’re coming from and everybody is coming from when they kind of look at me showing these problems. And they really kind of want to say, oh, but come on, Bruce. I mean, like, obviously humans make airplanes and rockets like clearly what they’re producing is special in some way compared to the immune system compared to what animals are doing. Clearly, clearly, Bruce, something special about what humans are producing compared to everything else. Okay. And I think that that’s a completely intuitively obvious, intuitively legitimate question to ask. Now, there is one aspect to your question, though, that I want to make a correction on because I know people see me this way. But to some degree, your question, the way you phrased it, Peter, there’s an assumption that I want to say, oh, well, see, there’s all these other sources of knowledge. And I have this other theory of knowledge in minds better than Deutsches, and I’m advocating for my theory of knowledge. And everyone reads me that way, even though I never really say it, right?
[00:05:23] Blue: I just think it kind of causes people to say, oh, you’re advancing this other theory, you’re advancing this other theory. And I really kind of want to say, no, that’s not what I’m doing. You know, my interest is in AGI. I want to understand what it is that’s special about humans. And to understand that, what you have to do is you have to take problems seriously. You have to look at every single niggling little problem and you have to say, okay, I don’t know how to solve that problem. That’s good. That means that’s an area I should be searching. And if you’re going to kind of hand wave away the problems, if you’re going to kind of say, oh, that’s not really a problem because intuitively I can see you’re wrong. Even if I can’t tell you specifically what’s wrong, you’re already off the critical rationalist bus, so to speak. And yes, if this were a debate between two people and you needed to win, that might be an effective thing for you to do. But like, I’m really seriously trying to actually solve this problem. And if I can show that that the criteria that Deutsche says, Deutsche says knowledge is adapted information that keeps itself instantiated causes itself to remain so. And if I can show counter examples to that and I can say, look, there are other things that keep themselves to remain so that are also adapted information, but don’t come from the two sources. That means there’s a problem to be solved. It’s it does not that I’m advocating for anything. I’m I’m advocating for let’s take our problems seriously. That’s about the only thing I am advocating for.
[00:07:00] Blue: OK, so in this episode today, I’m going to do two things. One, I’m going to bend over backwards to be as as much as I can. It try to try to allow for this this intuition that people clearly have that there is something special about human knowledge in particular. But biological knowledge created by by a biological evolution as well. I’m going to try my hardest to bend way over backwards to accommodate that viewpoint because I acknowledge that there’s something to it. OK. And then after I do all that, I’m going to demonstrate why even bending over backwards as far as I can just doesn’t really do the trick further.
[00:07:45] Red: Since everybody thinks I’m advocating for
[00:07:47] Blue: a alternative theory of knowledge. Anyhow, let’s develop an alternative theory that kind of accommodates both viewpoints. So let me take you through it, though. And let me try to explain it to you. So this I made up this episode not knowing you’re going to ask me that question, Peter. But in some ways, this episode is like meant to answer the question you just asked me. So let me just do a quick recap. Recall that Deutsches criteria for knowledge in beginning infinity and fabric of reality consisted of things like it’s stored in a general purpose medium, it causes itself to remain so it’s adaptive information that causes itself to remain so it has causal power and it converges across many worlds that last one is from fabric of reality. And also recall that Hervé tried to take the constructor theory of knowledge and boil it down to three criteria, which he put out of what constitutes knowledge physically, which he put as it is capable of enabling its own preservation. It can be copied from one embodiment to another without changing its properties. It can enable transformations and retain the ability to cause them again. Now, I believe these two lists are actually conceptually identical. So I don’t care which of those two lists you go go with. I’m going to go with Hervé’s three criteria if for no other reason than because A, I helped him come up with it and B, I ran it past a whole bunch of Deutschians on your Facebook page and all of them agreed it was a good summary of the three criteria that you needed to match to be counted as knowledge.
[00:09:21] Blue: I should note that he also had four criteria of what counts as information, but I’m kind of assuming it’s information. So I’m trying to concentrate on the three criteria that make information distinct from knowledge. Okay. Now, let me actually point out that this is actually part of the answer to your question, Peter. You asked, well, why not just what what what makes the walking robot distinct from a calculator? Well, a calculator doesn’t do any variation in selection. Okay. It is an algorithm that produces a result without having any need to try out different solutions to the same problem. Okay. It has a single solution to the problem and it’s done there. That doesn’t mean that your questions invalid though. Like I could stop there and I could say, well, that’s the difference. Right. Is is if the algorithm is doing variation and selection, that means it’s creating knowledge. And if it’s not, then it’s not. And the truth is that was I’m not even sure I believe that, but I can at least tell you that there is no nothing in what a calculator does that could possibly pass Hervé’s three criteria. So I’m kind of starting with the assumption that we’re taking the Deutsch constructor theory of knowledge seriously as a theory, which we’re summarizing as Hervé’s three criteria. And if it doesn’t fall on the side of that, then it’s not creating knowledge and maybe that’s right. Maybe that’s wrong. Theories are just conjectures. Okay. So maybe your question could still be valid. You know, why, why don’t we consider a calculator as creating knowledge, but I’m kind of just upfront assuming that we’re starting with Deutsch’s theory of knowledge at the starting point.
[00:11:02] Blue: Therefore, a calculator gets eliminated by the three criteria, whereas the walking robot does not get eliminated by the three criteria, at least not when stated the way I just stated them. So keep in mind that I’m not advocating for a theory. I’m advocating for the existence of a problem that I want people to take seriously. Okay. So I feel comfortable saying, look, I’m not even going to consider the calculator example because I know it doesn’t match Hervé’s summary of Deutsch’s constructor theory. Whereas the walking robot does match it at least in its and the way Hervé originally wrote it. Okay. Without further clarification or further, you know, attempts to say, well, I actually met this. Okay. Does that make sense so far? Why, why I’m not even considering the calculator example? It’s got nothing to do with I have some theory and I can eliminate the calculator example. It’s got everything to do with that. I’m concentrating on a problem with Deutsch’s theory.
[00:12:01] Red: Oh, yes, it does. I guess one of the things I was just reflecting on is that I, one of the things I like about, I guess just Deutsch and his whole ideas about knowledge is that he does. He doesn’t seem tied into one definition of knowledge. No, he probably doesn’t.
[00:12:22] Blue: He’s refined this several times over time.
[00:12:25] Red: We’ve got information with causal power, useful information, information that solves a problem, information that once instantiated causes itself to remain instantiated. Information that gets copied and then that, you know, that doesn’t even get on into the definitions of an explanation, which is a whole other thing. So, you know, it’s quite open -ended in a way, but it’s still, I don’t know, to me, that kind of indicates that I guess one person, you could look at that uncharitably as it’s all very fuzzy charitably. I think you can look at it in that he’s not just interested in defining words. He’s trying to get to something deeper.
[00:13:11] Blue: Yes.
[00:13:12] Red: And I think that something deeper has got to be that there is something special about human knowledge creation, which I think you pretty much agree with.
[00:13:24] Blue: Yeah.
[00:13:25] Red: Right. I mean, it’s just we’re trying to work out the details. You know, there’s only one force in the universe that’s going to divert the asteroid or, you know, hear death or whatever. I mean, there’s just, it just seems, I mean, I hate to just say it just seems obvious and leave it at that. Well, I think that’s
[00:13:46] Blue: how everybody does it, right? It just seems obvious and they’re done, right? Yeah. And only a crazy person like me says, no, I want you to get more explicit. I want you to tell me exactly what it is, right?
[00:13:59] Red: Yeah.
[00:14:00] Blue: Okay. So by the way, I that list that you were kind of just making that sounded a lot like Bart’s list. Is that Bart’s list that you? Yeah, that’s Bart. Yeah, I should have given him credit. Can you actually read Bart’s list for us on the air here?
[00:14:12] Red: Well, I kind of just did. Okay. Well,
[00:14:15] Blue: you said that doesn’t include the ones about explanation. Read though.
[00:14:18] Red: The explanations.
[00:14:20] Blue: Yes.
[00:14:21] Red: Information that explains things seen in terms of things unseen and an account of some phenomenon in reality what it is, how it behaves and why.
[00:14:32] Blue: Yeah.
[00:14:33] Red: So those are from David Deutsch.
[00:14:37] Blue: Yes,
[00:14:38] Red: but through Bart.
[00:14:40] Blue: Okay. Now let me point something out because this is actually important.
[00:14:43] Red: Okay.
[00:14:44] Blue: Several things on that list are equivalent conceptually in my mind and the last two are not.
[00:14:50] Red: Okay.
[00:14:51] Blue: So it’s important that we realize this though. If you think so read the last two again.
[00:14:57] Red: About explanations.
[00:14:59] Blue: Yes.
[00:14:59] Red: Okay. Information that explains things seen in terms of the unseen.
[00:15:05] Blue: Okay. So let’s stop and let’s just look at that one for a second. Okay. Okay. That understanding of knowledge is does not include knowledge created by biological evolution. Can you see why?
[00:15:19] Red: Yes.
[00:15:20] Blue: Yeah. Absolutely. Right. Because it does not create explanations about the scene via the unseen.
[00:15:26] Red: Yeah. That’s that would only apply to the human.
[00:15:29] Blue: Okay. This is the point I’m trying to make though.
[00:15:32] Red: Okay.
[00:15:33] Blue: Is that Deutsch has had many different ways of looking at knowledge and he he understand like even if maybe the fans of David Deutsch don’t understand this. Deutsch understands that when he says creativity is the capacity of humans to make new explanations. He knows that that’s not the only way to understand the word creativity or that he’s not even using the word creativity consistently that way. Okay. And the same thing happens with the word knowledge. Right. He he when I asked him I said what about you know I didn’t use the walking robot example but I said like your own example and he goes you know what like I’ve got a robot that like mows my lawn and it has to like go out and try things and I have no problem calling that knowledge. He said to me.
[00:16:17] Red: Right.
[00:16:18] Blue: So I mean Deutsch gets it that he’s using these words in different ways in different circumstances even if maybe some of his fans don’t get that that’s what he’s doing. Okay.
[00:16:28] Red: So it’s really the David Deutsch stands. That’s what I heard on Twitter. The stands you know that term right. David Twitch stands who are interested in tying down these definitions more than more than he himself is.
[00:16:45] Blue: What is this? What I don’t know what that term means stands.
[00:16:48] Red: Oh, it just it comes from an M &M song and a stand is a fanatical fan of someone.
[00:16:55] Blue: Yeah. Well, so it’s actually might be the opposite though. So if you understand, so I’m looking at the concept not the word and this is where I actually think I differ from every single person I’ve talked to about this. Okay. There’s a concept. So read that. Read that definition of knowledge based on explanations again.
[00:17:16] Red: I’m information that explains things seen in terms of things unseen.
[00:17:22] Blue: That’s a legitimate concept. Okay. And I don’t have a word for that concept. So I may well refer to it as knowledge.
[00:17:29] Red: Okay.
[00:17:30] Blue: As a label.
[00:17:31] Red: Okay.
[00:17:32] Blue: But now let’s let’s talk about her vase list. All right.
[00:17:35] Red: Okay.
[00:17:36] Blue: Three criteria. It is capable of enabling its own preservation. It can be copied from one embodiment to another without changing its properties. It can enable transformations and retain the ability to cause them again. Those three criteria are themselves an interesting concept that I don’t have a word for. So I may refer to them as knowledge. Okay.
[00:17:54] Red: Okay.
[00:17:55] Blue: But it’s not the same concept as the one you just read from Bart. And in fact, what there’s a relationship between the two Barts is a is a subset of her vase. Okay. Anything that counts as knowledge under that one from Bart would also count as knowledge under her vase, but not everything that counts as knowledge under her vase would count as knowledge under Barts. That doesn’t we don’t have to look at this as who’s right. We look at it as you’ve got two different concepts with two different sets of properties. One’s more restrictive than the other. Okay. And they have a relationship and I don’t have a term for it. So I call them both knowledge that doesn’t but that doesn’t mean that they’re the same thing that they’re both interesting concepts in their own right. Okay. This is really what I’m trying to get people to think about is the concepts, not the words and it’s hard because I have to use words to explain the concepts and that causes confusion because then people immediately think, Oh, Bruce, you just try to define knowledge differently. No, I’m trying to get at the concept and I have to use words to do it. So it’s okay that we call both those concepts knowledge. What I want you to understand is that they are not the same concept. They may be related concepts, but they’re not the same concept. Okay. But they’re both interesting concepts in and of themselves and Deutsch gets this. So he reuses the word knowledge to mean several different concepts, each of which is by itself interesting. Okay.
[00:19:23] Blue: But the fans of David Deutsch don’t get this often and instead they become very confused and it’s super hard for me to explain to them. No, look, her vase list of what constitutes knowledge is by itself an interesting concept. Just exactly the way it was originally written. There’s no need to refine it further. Okay. Yes. It might be and I actually ended up agreeing with Henry when Henry brought this up to me. He said, well, those three criteria are fine. But actually what he or Milo had in mind was that they were created by replicators that they exist as replicators, even though that’s not listed as one of the three criteria and she had in mind that they would last for generations. And I wasn’t sure I believed him. So I went and I looked this up in Kiarra’s book and it turns out Kiarra does write as if she’s assuming we’re talking about replicators and we’re talking about keeping themselves instantiated for generations. Okay. Here’s the point I want to try to make though is that doesn’t the fact that that’s true doesn’t make her vase original three criteria uninteresting. It may not be precisely what Kiarra had. It may be a little not not precisely what Kiarra had in mind, but it is by itself still an interesting concept. Nothing changes just because it turns out that Kiarra actually had these two additional implicit criteria that that her vase original three didn’t explicitly list. Okay. We’re interested in the concepts. Okay.
[00:20:55] Blue: The problem we’ve been discussing for the last I don’t five episodes or something is that these criteria that her very originally listed are not aligned with the two sources hypothesis that the soul that the soul’s two sources of knowledge is biological evolution and human creativity. In other words, the three criteria aren’t explicit enough to limit us to the three sources to the two sources. Sorry. Okay. That’s an interesting problem. The fact that when people try to wave me away and say, oh, but Bruce, it’s just obvious. It’s like, no, you’ve totally missed the chance to learn. You’ve totally missed the chance to realize that the three criteria don’t match the two sources hypothesis. Okay. And that this is an interesting problem that you should be interested in. Okay. And you don’t want to wave it away. Yes. You can wave it away and it is obvious that there’s something special about human knowledge. It just doesn’t happen to be the three criteria. Right. I mean, it’s there’s something missing. There’s something more and as I pointed out, you can add in Kiara’s. It’s requires to be held in replicators and it needs to instantiate itself for generations. And that’s still not sufficient to get us to the two sources hypothesis. In fact, we tried all sorts of different additional criteria and none of them get us to the two sources hypothesis. So you’re really left with two possibilities. Okay.
[00:22:20] Blue: Possibility number one is that we’re going we’re barking up the wrong tree that there is no set of criteria that physically separates knowledge in such a way that it only the two sources hypothesis creates it or there is a set of criteria and nobody knows what it is and the goal would then be to figure out what it is. Okay. How is that not an interesting problem? I mean, I feel like the way I’ve laid this out, if you can get past this idea that nobody seems to be able to get past that I’m not advocating for a theory, I’m advocating for an interesting problem. Okay. I want to know what the criteria is that allows me to get to the two sources, which are the two special sources, or I want to admit that there is no such criteria and that we’re going to need to look at how the knowledge was created that what makes the two sources special is not physically what they create, but how they create it. It’s got to be one of the two. I don’t see how it could be anything but one of the two. There’s either the criteria exists and I don’t know what it is or it’s physically impossible to come up with a list of criteria that defines knowledge that gets you back to only the two sources and therefore I must expand the question to a different question. Okay. It’s got to be one of the two. Logically, it has to be one of the two. Okay.
[00:23:39] Red: Well, I kind of like the, I guess, the former, I would say, because I, you know, I still like this, the two sources hypothesis, it’s one thing it just sounds cool to say. Right. There’s two sources of knowledge, human minds and natural selection, but then also it just sort of rings true. It really does that these are two. There’s just something special about these two mechanisms or whatever you call it, you know, there’s something there. I would not want to just abandon that, but I said,
[00:24:18] Blue: well, I think that you’re expressing exactly how people in general feel, right? Yeah. So remember that when Henry, when confronted with this problem, he on the fly added additional implicit criteria to the three criteria that Hervé had come up with, that knowledge needs to be capable of enabling its own preservation through replicators for generations. And likewise, James, when I was arguing with James, he said, but I’ll hold you responsible to say what you find so special about adapted information created by the immune system. It doesn’t seem special to me because the trial error process in the immune system is 100 % deterministic. So the result is two and thus a mechanical consequence of mechanical consequence. I see therefore nothing special or relevant about this kind of adapted information. So in essence, James is also adding additional criteria on the fly. He’s in essence saying adding one of two additional criteria. One might be that it is not knowledge unless it is special by which we mean it was created by a non -mechanical, non -deterministic process or possibly it’s even vaguer than that. Maybe it’s just, it needs to be special. James needs to in his gut be able to see it as special. Maybe that’s the criteria, right? That would be kind of a nasty criteria. But maybe that’s what it is, right? And either way, the issue is is that you are adding criteria. If you recall, Henry didn’t want to allow me to write it down, but he insisted that it existed. Okay, so we got to be able to talk about this. We got if I’m not allowed to write it down. That’s a problem, right?
[00:26:00] Blue: So we got to be able to say, okay, what is the actual criteria that you’re advocating for? Okay. The key commonality here though was adding criteria to the definition of knowledge on the fly to be able to rule out counter examples like the walking robot, the immune system, animal means, et cetera, so that you could get back to the two sources. So normally the implications of a theory aren’t considered core to a theory. Normally we start with a theory and we try to boldly and precisely explain the theory and then we have no choice but to accept its implications. That’s called reach. Okay. But for some reason, the two sources part of the theory is considered the core part of the theory and we fill the need to adapt the criteria on the fly until we’re back to the two sources. This is like, this is the analogy I used in the past podcast. If Euclidean geometry saw its core feature as all triangles always have 181 degrees. So you show Euclideans that actually using their own axioms it always comes out to be 180 degrees. So the Euclideans start to ad hoc vary their axioms until they get back to 181 degrees. Now maybe that’s an unfair comparison and I actually argued that it was in a past podcast, but you can see the issue here is is that the two sources is the key core part of the theory. It’s the part that isn’t allowed to vary and the criteria no matter who’s defined them, no matter how it’s been defined, have always been allowed to vary to get back to the two sources hypothesis.
[00:27:31] Red: So it sounds like what you’re kind of saying is that by having a sort of sentimental attachment to the two sources hypothesis, we’re really just off the rails with papyrinism.
[00:27:46] Blue: Well, no, I’m not saying that.
[00:27:48] Red: Okay. Okay. Okay. I mean,
[00:27:50] Blue: I think one could read me that way and I think everybody has read me that way.
[00:27:54] Red: Okay.
[00:27:55] Blue: And to some degree, it’s kind of true. Like if when I explain, look, I’m not trying to say it is knowledge. I’m trying to understand the criteria better if they then refuse to have the conversation, which James in particular did, right?
[00:28:12] Red: Then
[00:28:13] Blue: yes, I think they’re now at that point completely off the papyrin rails. But I don’t necessarily think that has to be the case, right? I think that they could say to me, Bruce, we’re trying to figure out what the criteria are and we don’t necessarily know what they are,
[00:28:30] Red: right?
[00:28:30] Blue: So we know it must include these three, but it could include other things. Okay. And we don’t know what it is. But the reason why we’re always trying to adjust the criteria to get back to the two sources is because we’re really the theory is about what’s interesting about the two sources. Okay. And we’re trying to figure out what that is. Now, I’ve never had anybody say that to me, but if they had said that to me, that would have been completely consistent with papyrin epistemology. So I guess, in essence, yes, I actually do believe that both Henry and James in particular, when they actually talked with me about this, that both were off the papyrin rails. But I think the reason why wasn’t because they kept varying the criteria, but because they refused to admit that there was a problem to be worked on when, in fact, there was objectively, there was. All right. And it could have been either either of them could have handled it in such a way that it was completely consistent with papyrin epistemology. They just chose not to. Okay. Now, what happens when you do that? Well, the conversation kind of ends and nothing gets resolved, right? We’ve passed an opportunity to see, oh, there’s an interesting problem here. Let’s try to solve it and to fall in love with the problem. Okay, which is what’s happened to me. I’ve fallen in love with the problem because I don’t know how to solve it. All right. And I try to ask people for help and I always get this kind of hand wave and I can clearly see that they jump off the papyrin rails at some point.
[00:29:57] Blue: I can also clearly see they don’t have to like they’re choosing to jump off the papyrin rails. I don’t know why, right? It to me, it seems a little dogmatic, but there’s an attachment that doesn’t need to be there. But it may well be the case that the answer to my question is we just don’t know what the criteria is yet, Bruce. That’s a completely legitimate answer to my question.
[00:30:18] Red: Okay.
[00:30:18] Blue: Does that make sense?
[00:30:20] Red: No, it does. Yeah.
[00:30:21] Blue: Okay. Let me let me defend James and Henry here though because I don’t think that they’re like completely wrong. Like they sent something special about the two sources and they are conjecturing that the specialness is in what is being created rather than how it’s created. And I favored the theory that the specialness is how it’s created through an open -ended search process. Like I totally I could be wrong. Like I don’t know what the general intelligence algorithm is. How do I know for sure that I’m right? Okay. So we’re going to we’re going to talk about that. Okay. What we’re going to do though is we’re going to allow the Henrys and James and the Peters of the world to still have their two sources hypothesis and even that to be what we define as the two sources of knowledge. Okay. Like I’m going to totally give you everything you can ask for. Let me let me go show you how far I’m willing to go and you everything you’re asked for. Okay.
[00:31:17] Red: Okay.
[00:31:18] Blue: So what I’m going to do is we’re going to develop a new theory. This theory is not about the two sources hypothesis and therefore is not about creation of knowledge. Okay. It’s instead a theory about adapted information of a certain kind and that kind includes the walking robot example. Okay. And it includes our stuff that comes from an evolutionary algorithm. It includes the immune system. It includes animal learning and animal means includes machine learning algorithms, at least the ones that use variation and selection. Okay. It includes AI search algorithms. It includes trade secrets. I bought up trade secrets as a counter example. So we’ve got this kind of adapted information that I’m talking about and I’m going to say Peter, you’re right. Those are not knowledge. They clearly are not coming from the two sources and only the two sources create knowledge. I’m going to give that to you 100%. Okay. That it doesn’t make those examples go away. It just means that there’s some interesting kind of adapted information that’s got some similarities or relationship to knowledge, but isn’t itself knowledge. That’s all that really means. And so what we’re going to do is we’re going to develop a theory about those that kind of whatever that is since it’s not knowledge. Okay. And I’m going to just say you’re right. It’s not knowledge and I’m done. Okay. There’s clearly something interesting about these examples because they are all examples of the power of poppers and Campbell’s evolutionary epistemology. So recall that Campbell’s evolutionary epistemology that Popper bought off on was that they’re that all knowledge creation and all inductive achievement. He actually uses that term comes through blind variation and selective retention. Okay.
[00:33:03] Blue: Now, I don’t even I’m not even advocating for Campbell’s theory. Like I actually think Campbell’s theory is going to prove to be wrong. Okay. But I’m taking the theory seriously. All right. And I’m saying, okay, what if it’s true? What are the implications? All right. The walking robot example is an example of evolutionary epistemology of Campbell’s evolutionary epistemology. The immune system is an example of Campbell’s evolutionary epistemology. All of these are all of these are examples of Campbell’s evolutionary epistemology. So they have and recall that Campbell’s evolutionary epistemology was an attempt to generalize critical rationalism to cover everything. That was its purpose. All right. Whether it succeeded or not a different question, but that was what Campbell was attempting to do. So they these examples, the walking robot example, the immune system, animal learning, they are connected to Popper’s epistemology in a deep way. And therefore they are interesting from an epistemological standpoint. So we do need a theory of this kind of adaptive information even if it is not a theory of knowledge. Okay. You’re with me so far. Have you have you bought into my argument that? No, it’s not knowledge, but it is something interesting.
[00:34:14] Red: I feel convinced right now.
[00:34:16] Blue: Okay. Here’s what we’re going to now do. People at home aren’t going to be able to see this. I’m going to share my screen to Peter. Check the show notes and you will find a link to a PDF that you can take a look at where you can actually see what I’m talking about with Peter.
[00:34:32] Red: I’m going to actually write this
[00:34:34] Blue: so it’s easier for you to follow. Okay. What we’re going to do is we’re going to start with the theory of knowledge by which we mean. The Deutsche’s constructor theory of knowledge. And we’re going to start with the three criteria that her vae came up with. Okay, which you know, a lot of people bought off that it was a good summary of Deutsche’s constructor theory of knowledge.
[00:35:00] Red: All right.
[00:35:01] Blue: And then on the other one, we’re going to call this Bruce’s entirely new theory that is not a theory of knowledge, but is instead a theory of adapted information that is not knowledge. Definitely not knowledge, but is related to Campbell’s evolutionary epistemology. Campbell’s and poppers evolution epistemology since popper did endorse it. And so is interesting in its own right. Okay, you’re with me.
[00:35:46] Red: Yes, I am.
[00:35:47] Blue: And right now the two lists are exactly identical. This is just her base three criteria. Okay. Okay. So now what we want to do is we want to modify these two lists so that they are now distinct theories from each other. Okay. Now, I’m fine with adding anything to the theory of knowledge, Deutsche’s constructor theory of knowledge that you want, no matter what, like I do not care. I will, I will literally not criticize anything you want to add to that list. But let’s say James were here today and James said, well, you know, Bruce, it’s not special. I would be fine with saying one of the criteria for theory of knowledge is that James must find it to be special in his gut. Those of you listening at home, you can’t see, but I’m typing this as we speak. That’s why I’m going slow. I’m totally fine with that. Like I’ll accept anything as part of the criteria of knowledge at this point.
[00:36:49] Red: Okay.
[00:36:50] Blue: And if that’s not good enough, I’ll even add and Bruce gets no say whatsoever as to what counts as special. Only James gets to decide that. Like I’m totally fine with that. Okay. And now we have two different theories, right? We’ve got the original three criteria and we’ve got a modified version that gets us back to the two sources. Okay. And we can now see that my theory clearly isn’t about knowledge and this other one is about knowledge and it should be obvious now what separates my theory from theirs. It’s James’s gut feel as to what counts as special. Okay. Now, obviously a lot of people, I mean, I’m being humorous when I say that, but a lot of people would immediately say, okay, yeah, obviously we can’t differentiate two theories based on James’s gut feel as to what’s special. Okay. So, so let’s, but the reason I’m bringing it up is not to be silly, but to show how far I’m willing to go to be accommodating. Here to James’s viewpoint, to anyone’s viewpoint that feels that knowledge should only come from the two sources. Okay. Let’s try to, let’s try to do this in a better way though. Okay. So like, let’s say that let me erase that one and instead we’ll make the four, let’s, let’s take Henry’s argument for a second just to show how far I’m willing to go be accommodating. So one of the things that Henry argued is that he said that criteria number one is it is capable of enabling its own preservation. Okay. Now, I pointed out to him that the walking robot was an example of adaptive information that was capable of enabling its own preservation. And he responded to me.
[00:38:38] Blue: Well, no, you’ve misunderstood that criteria really means that it has to be a replicator and it has to exist for generations. So I said, well, can I write that down? So I attempted, I actually tried to do this with, with Henry. Okay. I said, okay, can I, I pulled up the two lists just like I’m doing with you, Peter. I said, can I just write down it is capable of enabling its own preservation through replicators for generations. And I was going to show him that mine was still this interesting theory and it was related to Campbell’s and Popper’s theory and he immediately said, no, you’re not allowed to do that. That, that is, it’s implicit. It’s just obvious. I, he kind of treated me like I was stupid, right? Like, how can you not see that that criteria meant that, right? And it turned out he was even right, right? I got what I read and here it does seem to have this in mind. So since he doesn’t want me to write it down explicitly, I’m going to put it on mine. I’m going to say is capable of enabling its own. Here’s what I’ll write it as on my list so that his can stay implicit. I’m on my side. I’m going to write it’s capable of enabling its own preservation, regardless of if done via replicators and it doesn’t matter how many generations. Okay. So now this is a difference between my theory and the theory of knowledge, right? Is that I’m, I have in this case, I haven’t, this is a non restrictive criteria. It’s actually opening the door up, right?
[00:40:00] Blue: I’m saying on the one side, theory of knowledge means it must be capable of enabling its own preservation and we’ll all secretly know that that means by replicators for generations. We won’t write it down because Henry was concerned about me writing it down and instead I’ll explicitly stand on my side. It’s capable of enabling its own preservation, regardless if done via replicators and it doesn’t matter how many generations. Well, now you can immediately see that a lot of these examples like trade secrets and things like that get captured by my side. Okay. So now trade secrets are part of whatever this theory is. We don’t have a name for it yet. And over here, not trade secrets because they are not replicators. I’m fine with that. Okay. Like that’s what I’m trying to do. I’m trying to figure out what is it that is the commonality between all these different interesting things that people want to tell me aren’t knowledge. Okay. So and hopefully this shows just how accommodating I’m willing to be here. I think let’s let’s do this. Okay. I am willing to do this reset list so that they’re exactly the same again. I’m going to allow this and also other criteria. This is on the knowledge side. So this is the three criteria. Try to get back to the two sources and also other criteria X where X is not yet defined that are needed to get back to the two sources being the only two sources of knowledge. Okay. I’m totally okay with accepting this. Like I just said that, right? Like if this is really what we’re doing and we’re admitting that we don’t know all the criteria, but the two sources is special.
[00:41:42] Blue: And that’s what we really care about. Then you should have no objection to me doing this. You should allow me to add criteria X where X is currently undefined and now this is only the two sources and over on my side, it includes non two sources examples such as the walking robot. Okay. And I admit these are two different theories now. Like I’m totally okay with admitting that these are two different theories. Okay. Now here’s what I will insist upon though. Okay. Now I’m going to insist upon a few things as a critical rationalist, I feel like I must insist upon the following things. However, I want to emphasize just how ridiculously reasonable the following requests are. Okay. So number one, I insist that we do give a name to this to my theory. Okay. Like I mentioned when James, I said, look, I tried to make this list for James. I said, look, I’m saying it’s a different theory. I’m saying it’s not knowledge and James would immediately say, no, it’s not knowledge. I’m saying, yeah, I know, I’m agreeing with you. I’m agreeing with you, James. It’s not knowledge. Okay. And I’d say, can we give it a name though? Like I get tired of calling it bruises entirely new theory that is not a theory of knowledge, but is instead a theory of adaptive information that is not knowledge, definitely not knowledge because it’s related to Campbell’s Popper’s evolutionary epistemology. So I’m interested in it. Right. Like it was just too much of a mouthful. So I said, allow me to call it something. Right. Whatever it is, it’s not knowledge, but allow me a name and he refused to allow me a name. Okay.
[00:43:12] Blue: Now, I think the reason why he refused is might be my my own fault. I tried to call it quasi -knowledge and that was going to say, unknowledge or something. Yeah. Okay. And well, okay. So unknowledge is better and the problem with me trying to call it quasi -knowledge is that that’s a very positive sounding term, right? And James clearly had a very negative view of it, right? Okay. And so I think it was it was more or less the same as me saying, yeah, we all know it’s secretly I’m right. This is knowledge and he was wasn’t able to wrap his mind around the fact that I was actually trying to develop a real issue here to solve, right? A real problem. So I had after my conversation with James, I came across the fact that David Deutsch already gave a name to this kind of adapted information. And so I’m going and I don’t think he did it intentionally, but he accidentally did. And so let me get back to that though. I’m going to give it a name here in a second. So but I am going to insist we give it a name. All right. I’m also going to insist that we let people decide for themselves if this new theory is special or interesting. Now don’t get me wrong here. I’m still okay with us making one of the criteria that James must find it special. If that’s what James wants me to do. I’m totally fine with that. I’m not saying that we can’t use is James got feel of specialness as one of the criteria for what counts as knowledge.
[00:44:34] Blue: I’m just simply saying that once we’ve used that as one of the criteria, I get to for myself decide if the walking robot example is special or interesting. It just has placed no role in whether we call it knowledge or not. Okay. And then I’m going to insist that every implicit requirement of Deutsche’s theory either has to be written down or it has to be an explicit counter requirement for the alternative theory. So when Henry tells me, no, you’re not allowed to say it’s by replicators and it’s for generations. I must be allowed to on my side right. I don’t care if it’s replicators and it doesn’t have to be for generations other than that. I make no insistence on how we write it down. Okay. And then the last thing I’m going to insist upon is that we use the substitution method. So we’re going to treat these as just labels. So if the left side here, the theory of knowledge is knowledge and my side is not knowledge or whatever we’re going to end up calling it unknowledge or something like that, that they are just ultimately treated as labels that what we really care about is the underlying concept. So what I have in mind here is when you now say all an animal’s knowledge is in its genes, I’m allowed to take the criteria, whatever it is you came up with, I’m allowed to substitute the word out the word knowledge and place the criteria in its place and drop the word knowledge altogether. Substitution method is very important for getting away from essentialism. Okay. And I get to restate your theory without using the word knowledge or without using the word unknowledge or anything like that. Okay.
[00:46:13] Blue: As long as you agree to that criteria. Now again, the criteria can be vague. It can be James’s gut feel. It can be something that’s undefined criteria X. Like I don’t care what it is, but I’m always allowed to take that final set of criteria, no matter how it’s worded and I’m allowed to rip out the word knowledge and put the criteria in its place and your theory has to hold water on its own without referencing the halo of philosophical baggage that comes with the word knowledge. If you’re not a willing to agree with that, then I’m wasting my time. Okay. So I’m going to insist upon the substitution method. I’m being completely reasonable here. Like I’m really bending over backwards to be as ridiculously accommodating as I possibly can other than those four things that I said, I must insist upon. I will basically accommodate anything. The key thing here is, is that I’m giving the defenders of the two sources theory everything they can ask for period end of story. No exceptions. And I’m not disallowing anything they want no matter how absurd or ridiculous. I might personally feel it is and I’m accepting that their theory is the correct theory of knowledge and mine is some other theory that isn’t specifically about knowledge anymore. Okay. That’s how much I’m giving them. I honestly feel like it’s impossible for me to be any more accommodating to their view than I am being here. Okay. I really want to emphasize that. Now, what are we going to call my theory though? All right. That’s the question. Now quasi knowledge got James’s hackles up. I can understand why.
[00:47:41] Blue: In fact, I’m going to admit it was bad for me to try to call it quasi knowledge. I like your unknowledge better because it’s a negative term and I don’t mind if we use it yet to get a term we can call it stupid not like knowledge. If we if we want to we can call it Fred if we want to like I don’t care what we call it. I just need a label for it. Okay.
[00:47:57] Red: I kind of like the quality. You know, it’s sort of you’re talking about something that’s could be seen as knowledge adjacent. Yeah. Quasi knowledge kind of works for me. I don’t know.
[00:48:08] Blue: Okay.
[00:48:09] Red: Well,
[00:48:09] Blue: we could call it quasi knowledge, but I actually like the term that Deutsch came up with.
[00:48:14] Red: Okay.
[00:48:14] Blue: No, I’m going to since Deutsch came up with it. I’m going to use his. Here’s where I’m getting it from. It’s it’s his interview with Eli. Right. He accidentally named it. He says so what happened was is Eli asked if deriving ideas from premises creates knowledge. Okay. In in his interview, go go look up the interview and you’ll see that he does talk about this. Okay. And Deutsch insists that it does not create knowledge. Like if you come with a bunch of axioms and then you like work out what the implications are and you come up with some surprising result that nobody knew from the axioms. Does that count as knowledge creation and Deutsch says that it is not, but then he says this. He says it is a very good simulacrum of knowledge. Okay. Here’s the actual quote. It’s a good example because I also think that isn’t knowledge either, but that looks much more like a simulacrum of knowledge than the previous examples that Eli had raised, which included the immune system. I’m getting a lot of these examples from Deutsch himself. Okay. So I actually really like this idea. Let’s call this simulacrum knowledge or Stymol knowledge for short. Okay. And the reason why the reason why I like this is because I’m like quasi knowledge, which is a positive term. This is clearly now a negative term. We’re saying something negative about it. It’s just a simulacrum of knowledge. It is not knowledge. You’re clearly it’s we’re not even making like you said quasi knowledge implies knowledge adjacent. This doesn’t even imply that, right? It’s a negative term meant to show, you know, oh, no, it’s just a simulacrum of knowledge. Not really knowledge. It’s just a simulacrum of knowledge. Okay.
[00:49:55] Blue: So we’re going to call this the theory of Stymol knowledge, which is basically the original three criteria of Deutsch’s constructor theory of knowledge without the two sources hypothesis and without any other implicit or explicit being added. Okay. So it includes the walking robot example. It includes animal learning animal memes. It includes trade secrets. It includes absolutely everything that we weren’t so sure was knowledge or that didn’t come from the two sources. That’s all now part of my theory of simulacrum knowledge. Now here’s some questions I’ve got for you now that we’ve kind of worked this out. Okay. So we’ve got the we got the two thing now what I’m going to go for. I don’t care what the other criteria are that we use or if I have to clarify mine to make them more expansive. I’m okay either way, but for the sake of arguments, since I can’t show every single possibility at the same time, we’re going to go with the criteria acts approach because it’s the most accommodating. So the theory of knowledge, which is constructor theory of knowledge now consists of these four criteria. It is capable of enabling its own preservation. This is the left column. It’s capable of enabling its own preservation. We secretly know that means with replicators for generations, but we don’t write that down. It is can be copied from one embodiment to another without changing its properties. It can enable transformations retain the ability to cause them again. And the fourth criteria is criteria X. We don’t know yet what it is, but there’s some other set of criteria that will ultimately allow us to get down to see that there’s only the two sources are the only two sources of knowledge. Okay.
[00:51:28] Blue: This is the most accommodating way. I think I can write this for the moment. Okay. And then the other way I have it is either just the original three criteria. So the right hand column is the theory of simul knowledge. It’s the original three criteria that Hervé used as a summary of constructor theory. And right now I just have it written exactly the way Hervé originally wrote it, but I’m totally okay with writing as the top one doesn’t require replicators. Doesn’t need to be for generations. Whatever I need to do to accommodate everyone’s viewpoint, assume it’s now written there. Okay. So we have these two theories in front of us and I want to ask you some questions about these two theories. You can see them on the screen. You understand what I’m saying, right, Peter? Yeah,
[00:52:12] Red: I think so.
[00:52:14] Blue: Okay. Here’s my first question for you. Okay. Is simul knowledge special?
[00:52:20] Red: Special in what sense?
[00:52:22] Blue: Well, that’s a good question. Okay. Now recall that one of the criteria we considered was that there was nothing special about simul knowledge. That was James’s argument to me because it was just came through a mechanical process. Oh, I see. Okay. So in fact, let’s let’s add that as one of the criteria. Let’s go back to must be special in that it doesn’t come from a mechanical automated process of persuasion. So now for those at home, we can’t see this. I have now added that as a fifth criteria or a sub point to the fourth criteria X that is required for the theory of knowledge. So is simul knowledge special under this way of thinking of the word special? Well, yeah, I would say so. Okay. No, let me be clear. Fifth criteria. Look at the fifth criteria. Oh, theory of knowledge must be special in that it doesn’t come from mechanical automated process of perspiration. So special is just a word that now means doesn’t come from a mechanical automated process of perspiration. Is simul knowledge special? By the way, just in case this isn’t obvious. This is doesn’t. So it does not put not is simul knowledge special.
[00:53:45] Red: It’s okay. So the fifth criteria in the theory of knowledge, Deutsches theory of knowledge must be special in that it does not come from a mechanical automated process of perspiration.
[00:54:00] Blue: Yes.
[00:54:01] Red: And in your the simul theory of knowledge, this is not a requirement.
[00:54:06] Blue: This is not a requirement.
[00:54:07] Red: Okay.
[00:54:08] Blue: Is simul knowledge special? Or let’s put it this way. Does does is the simul knowledge. Come through a mechanical process automated by perspiration since that’s what special here means. Notice the substitution method there. Okay.
[00:54:25] Red: Let me let me put this in a different way.
[00:54:28] Blue: Must be special. However, must be special. And simul knowledge has no such requirement. So the theory of knowledge, it must be special. Some knowledge has no such requirement. Is simul knowledge special?
[00:54:45] Red: I don’t see why not. I mean.
[00:54:47] Blue: Okay, explain. This is actually a this is actually a logical question. This is like I’m asking you a question about logic.
[00:54:55] Red: Okay. So the theory of knowledge, Deutsches theory of knowledge must knowledge must be special. And then in the simul theory of knowledge, there’s no such requirement that it’s special. But it could be special. It could be special or it could be not. There you go.
[00:55:14] Blue: Okay. This is the thing that I think people have generally missed. And it’s the point that really needs to come out. Okay. Let me. Okay. Let me draw. You can see what I draw. If you look at the link that’s in the show notes. So it there is a temptation. Okay. So there’s a temptation. So what James was actually doing is he was trying to argue to me that this can’t simul knowledge can’t be interesting because it’s not special. That was really his argument to me. Okay. And he basically agreed with him kind of sort of that there’s something special about the two sources. The issue here is, is that because the theory of simul knowledge is less restrictive than the theory of knowledge that other than that, they’re the same theory. Basically. It must must must be the case. That simul knowledge is the larger circle and that knowledge is the smaller circle and that knowledge is contained within simul knowledge. Check the image in the show notes to be able to see what I’m talking about. Okay. There’s no way around that as long as you are trying to use new criteria to get back to the two sources. You are automatically making the theory of simul knowledge include the special stuff. There’s no way around that logically speaking or to put this in another way. The theory of no theory of simul knowledge will always using this approach also include knowledge always. There’s no way around that logically speaking. Okay. Or the theory of simul knowledge is the larger set and theory of knowledge is this is a subset or to put another way, all knowledge is also simul knowledge, but not all simul knowledge is knowledge.
[00:57:02] Red: I think that actually putting it like that makes it makes a good distinction, I think. I mean, I guess what I’m kind of hung up on is as we’ve talked about before this idea of open -endedness. I mean, we’re looking at knowledge versus simul knowledge. Simul knowledge is not going to build a Dyson sphere and conquer death.
[00:57:28] Blue: Well, it will because simul knowledge includes knowledge.
[00:57:32] Red: Okay.
[00:57:33] Blue: See, this is the thing that people are missing. The moment we talk about it as simul knowledge, it immediately gets forgotten that that knowledge is the subset of simul knowledge that the way we’re going about this, we’re trying to add criteria or clarify criteria or whatever that all you’re actually doing is you’re turning simul knowledge into the superset. So simul knowledge, now what you really mean and you didn’t say it this, but this is typical of how people talk, what you really meant was simul knowledge that isn’t knowledge isn’t that’s what I’m Dyson sphere. Yeah. Okay. But simul knowledge absolutely will build a Dyson sphere. Okay. Because it is just a superset that includes knowledge.
[00:58:16] Red: Fair enough.
[00:58:17] Blue: Okay. This is the point I really have been trying to make to people and I can never seem to get them to understand it is that if you want to say the theory of simul knowledge is uninteresting, that could never be true because the theory of simul knowledge always includes the theory of knowledge. Okay.
[00:58:34] Red: In fact, in some sense, it must always be more interesting than the theory of knowledge because anything interesting about the theory of knowledge is also interesting about the theory of simul knowledge.
[00:58:43] Blue: Plus, if there is anything interesting about simul knowledge that isn’t knowledge, that would be interesting also and therefore the theory of simul knowledge will always be the more interesting theory. It’s impossible for it not to be the more it could at you could argue that it’s no more interesting than the theory of knowledge because all the interesting parts are contained in the theory of knowledge, but you can never argue it’s less interesting than the theory of knowledge. Does that make sense?
[00:59:10] Red: Yeah. I mean, it seems to me it’s less interesting in the sense that it’s less mysterious. I mean, something like the walking robot. We pretty much we know how that works.
[00:59:20] Blue: Yes.
[00:59:21] Red: Right. We don’t know how human creativity works. We don’t really know. Well, I don’t know if you’ll agree with this or not, but there’s the same kind of mysteries seem to be contained in Darwinian evolution as well. So, what was the word you say? Interesting.
[00:59:45] Blue: Yeah.
[00:59:47] Red: Well, I think you might, I’m not going to try it completely. It depends on how you are thinking of that word, I guess.
[00:59:56] Blue: But so maybe let’s look at it like this then. When we talk about the theory of simul knowledge, we’re talking about what we currently understand about knowledge and we talk about the theory of knowledge. We’re talking about what we currently don’t understand about knowledge.
[01:00:10] Red: That’s a good way to put it.
[01:00:12] Blue: Okay. In fact, I don’t think that’s just a good way to put it. I think that is a far more accurate straightforward way of saying what we’ve been saying all this time and that everything else around it has actually been a red herring. Okay.
[01:00:24] Red: Okay.
[01:00:25] Blue: Let me now make some interesting points though. Pulling back up my little picture of my superset, which is the simul theory of simul knowledge and the subset, which is the theory of knowledge. Again, if you want to follow along, take a look at the link in the show notes and you can pull up the PDF. They do have a relationship and we actually understand much about the relationship. Okay. The relationship would be whatever they share in terms of those criteria, which is basically her base original three criteria in this case. Okay. So we know that both are created via evolutionary customology of variation and selection or trial and error. We know that both are adapted information. Both keep themselves instantiated by being useful compared to their competitors. Both allow transformations in the world that have causal power. Can you see how this is something interesting that the fact is, is that when we talk about the latest theory of knowledge, sometimes we seem to really just be talking about the theory of simul knowledge. We’re talking about the stuff we currently know and and that it has deep ties to evolutionary epistemology. Therefore, poppers epistemology. So now let’s be very clear. Jeans and human ideas are simul knowledge, but they’re also knowledge. There’s no way around this. Okay. Every single argument I’ve had with defenders of Dwight’s theory of knowledge have always tried to modify the criteria to declare the walking robot example out of bounds and they’ve never once noticed despite me trying to explain it to them. They’ve never once understood that what they’re really doing is they’re creating two theories and genes and human ideas are both knowledge and simul knowledge and that they fall under both theories.
[01:02:12] Blue: They’ve kind of dug their own grave here. Okay. So now we can ask questions like now we can ask really interesting questions and we can get clear answers now that we understand the theory of simul knowledge. So for example, are all animals simul knowledge in its genes? Okay. We’ve debated over and over ceaselessly is all an animal’s knowledge in its genes. Okay. And Dwight has insisted all an animal’s knowledge is in its genes. Okay. I’m now asking a totally different question. I’m asking are all an animal’s simul knowledge in its genes? The answer is no. Okay. Because we know animals learn animal learning creates simul knowledge.
[01:02:53] Red: Okay.
[01:02:54] Blue: We can also ask this question. Are all animals knowledge in its genes? Yes. However, we need to understand what that yes actually means using the substitution method that I insisted upon. What we really mean now is this all of an animal’s simul knowledge was created by replicators for generations that that was created by replicators for generations and that James find special and non mechanical and exists only in their gene exists only in their genes. While the vast majority of their simul knowledge, the kind that was created by the animals ability to learn and that James finds mechanical and non special was created in the lifetime of the animal via its own ability to learn. Sorry. That was not the greatest sentence. It was horrible, but the point I’m trying to make here is let’s replace that with criteria X and it might be a little more clear. So when we say all an animal’s knowledge is in its genes now that we understand that animal learning does create simul knowledge. It just doesn’t just doesn’t create knowledge.
[01:03:52] Red: Okay.
[01:03:53] Blue: Once you realize that what we’re really saying when we say all an animal’s knowledge in its genes, we mean this all of an animal’s simul knowledge that was created by replicators for generations and that matches the here to for unknown criteria X exists only in their genes while the vast majority of their simul knowledge, the kind that was created by animals ability to learn that does not match the unknown criteria X was created in the lifetime of the animal via its ability to learn and was not contained in its genes. Okay. Once you use the substitution method like this, I think basically the statement all an animal’s knowledge in its genes while the answer is still technically yes is now a completely silly statement and it truly ever always was. Okay. Of course, what they’re really trying to get at is that genes are special in some way because they come through this open -ended biological process, but it is absolutely not the case that every every ounce of simul knowledge that an animal has comes from its genes, the vast majority of it comes through animal learning. Okay. You realize what a misleading statement the original statement was when we say all an animal’s knowledge is in his genes. It’s actually a simple tautology and hopelessly ridiculously misleading the fact that they don’t create any knowledge. What I mean here is the fact that animal learning does not create any knowledge or any of the other methods that we’ve talked about such as genetic programming, the fact that these other methods don’t create any knowledge is basically a misleading statement to the point at this point because they do create simul knowledge. Okay. And that is really just a super set of knowledge.
[01:05:31] Blue: And it’s in fact the more interesting super set because it encompasses knowledge. So when we say things like our animal memes knowledge, well, okay. Now under these two theories, no, they’re not knowledge, but they are simul knowledge. Okay. They’re there’s something that let me take. Let’s let’s get explicit here.
[01:05:50] Red: So to really simplify this, you could just say animals create animal knowledge and humans create animal knowledge and human knowledge. But that kind of just sounds like a tautology and not a very interesting thing to say.
[01:06:06] Blue: And yes, and it is a tautology. It has always been just a tautology. Okay. The interesting question is what actually separates animal knowledge from human knowledge? That’s actually the interesting question. As long as you’re insisting on calling animal knowledge, not knowledge, you’re blind to what the actual quite interesting question is.
[01:06:26] Red: Okay. That rings true. Okay. Okay.
[01:06:29] Blue: Okay. So I’m going to argue that it is more than a little obvious at this point that what we are calling Deutsches theory of knowledge now that we’ve added whatever criteria acts we need is something more like the quintessential example of a bad explanation. It’s it consists of a very easy to very criteria, whether that’s James’s gut feel or criteria acts currently undefined. And those are pretty much the only things that separate knowledge and simul knowledge at this point, because everything we actually know about what constitutes knowledge is now covered by simul the theory of simul knowledge because of the way I chose to word things. Okay. Um, but but for the moment, let’s ignore that fact and let’s assume that someday we’re going to have that criteria X whatever it is. Okay. And now I’m going to so now I’m literally giving you everything like there’s been making a case against the theory of Deutsches theory of knowledge, but I’m now giving you everything. Okay. Can this explanation be saved? Can is there is there possibly a criteria X that we just don’t know today that’s going to allow us to make a clear distinction between the theory of knowledge as separate from the theory of simul knowledge. Okay. Let’s let’s explore that as a possibility. I believe I can demonstrate why this is impossible and I’ve used this example in past podcasts. Let me reuse it again. The immune system.
[01:07:53] Blue: Let’s imagine a hypothetical where under in one set of the one part of the multiverse, there’s this giant pandemic and the immune system finds an antibody for somebody and then we duplicate that immune system and we go out and we copy that antibody and we save humanity using that antibody created by the immune system. Okay. Let’s say in a different universe that to save humanity nobody’s immune system happens to come up with the right antibody, but a bunch of scientists using inspiration managed to figure out how to put together an antibody and they then give that as a vaccine and they save humanity.
[01:08:41] Red: So that’s a compelling example because it’s basically the exactly the same thing that
[01:08:45] Blue: as part of the thought experiment, we’re going to specify that the final antibody is physically identical,
[01:08:52] Red: but
[01:08:53] Blue: one came through the immune system and the other one came through human inspiration and creativity.
[01:08:58] Red: Okay.
[01:09:00] Blue: Under this example, we are now forced to say that one is that that one, even though they’re physically exactly identical by definition, that one counts as knowledge and everyone counts as simul knowledge, but not knowledge. Okay. Well, that’s clearly a contradiction at least at least unless it’s I guess it’s not quite true that it’s clearly a contradiction. If you’re insisting that the criteria are physical differences, then it’s clearly a contradiction. But if you’re allowing for the difference to be not what’s physically different about them, but how they got created, then it’s not a contradiction. And that’s how come I think I know that the actual difference between two sources and the non two sources examples is not what makes them physically different, but instead how they got created. There’s something special about the two sources in its process, the fact that the process is open ended. It’s not physically what gets created at the end. Now there can be no criteria X that separates them if you’re insisting that criteria X is physical, physical differences. Okay. Now let’s say you disagree with me. Okay. Maybe you think there’s some sort of problem with the example I just used. I think that’s a correct example. I think it’s fatally shows that there’s no criteria X that exists, but maybe I’m wrong. Maybe a lot of times when we do this, we’re making little assumptions that we don’t even realize we’re making assumptions. So there’s always some room for I could be wrong. Okay. Here’s the challenge though. If you don’t agree with me that this example forever closes the door on saving the two sources hypothesis, I issue this challenge to you.
[01:10:39] Blue: You must explain to me how two identical antibodies by definition we’re defining them as identical can be physically different in some way such that one can be identified as knowledge physically and one not with and one not without having to appeal to which process created them. Now it seems to me that this is obviously a completely impossible task. But if you disagree with me, you now have a very explicit challenge that will refute me. Okay. In other words, my theory is 100 % refutable, at least in principle. Just not let me
[01:11:15] Red: let me just explore one thing here.
[01:11:17] Blue: Okay. I mean, could we define the creation of the of the knowledge created by the immune system as part of
[01:11:31] Red: the first source, meaning Darwinian. Knowledge creation.
[01:11:37] Blue: Yes. And that was why that was one of the last part in the previous two podcasts ago or something. I addressed the credit assignment problem.
[01:11:45] Red: Okay.
[01:11:46] Blue: So, yes, if you allow the credit assignment problem, I explained in that podcast why I consider the credit assignment problem to be fatally flawed and why you are not allowed to do that. I guess you’d have to tackle my argument there. Okay. In essence, what I said was, look, all knowledge begets all other knowledge. So if you’re going to allow yourself the credit assignment where you get to just decide who gets credit, even though there was actually multiple variation selection processes that play to roll, that obviously you now have a completely irrefutable theory. I can never refute it, not even in principle. And therefore, I now get to call it a bad explanation and I’m dismissing it on the grounds that it’s a bad explanation. Okay.
[01:12:25] Red: Fair enough.
[01:12:25] Blue: Notice that my challenge here is the exact opposite of that. Okay. You need to show me how there could be a criteria X that physically differentiates the two antibodies, the one in the part of the multiverse created by human creativity and the one that was created by the immune system.
[01:12:42] Red: Okay. When
[01:12:43] Blue: I’m declaring them to be physically the same, you need to either show that there’s some sort of logical flaw in my hypothetical. That would be a fair way to go about this. I don’t believe there is. So I don’t believe you can do that. Or you need to actually come up and show me. You know what? There actually is a physical difference between these and this is how I would define them. Maybe you could like come with some way to define it across the multiverse or I don’t know, like it’s not a complete impossibility. Like, sorry, it is a complete possibility in my opinion, but I can’t prove that, right? Of course, I can’t prove that. That would be justificationism if I could.
[01:13:19] Red: I have one more possible thing that just occurred to me that like a way to differentiate those two kinds of knowledge is, you know, okay. So let’s say we have a communication with an alien species or something like that. And, you know, we can take the human created antibody. Is that what you said? We have the human created antibody. Well, we can transmit that into the alien species across the galaxy, but the immune system is not going to do that. You know, it doesn’t.
[01:13:59] Blue: Okay, that’s certainly true. Okay, no ifs ands or buts about it. Now recall though that in the hypothetical example, I said that somebody’s immune system discovered the antibody and then the humans went on to duplicate it themselves.
[01:14:13] Red: And there’s nothing really
[01:14:16] Blue: stopping us from then saying, well, then I mean, yes, the immune system can’t do, but that’s just because it’s not an open search process like a human is.
[01:14:24] Red: So
[01:14:24] Blue: I actually think that example fits my theory, which is that the actual difference is in the specialness of how it gets created, not in physically what it creates.
[01:14:34] Red: Fair enough. Okay.
[01:14:36] Blue: Um, the thing I wanted to emphasize though is that my theory, the way I just explain this, this hypothetical example is it wants a easy to understand what a canner example would look like and B has no counter examples. This is going to turn out to be something important in a future podcast. So that’s why I’m calling it out now. Okay. It’s this idea of what do we mean by falsifiability or refutability? In fact, I think my theory, what I’m explaining here is completely irrefutable because there are no counter examples. That doesn’t mean it’s non falsifiable. What makes it non falsifiable is so let’s use a comparison of the two. So if you are arguing the credit assignment problem, that’s non falsifiable in that I can’t even conceive what a counter example would look like. Whereas what I’m saying here, it’s easy to conceive what a counter example would look like. They just don’t exist. And this is going to turn out to be one of the next threads I’m going to pull about something that I think is super important about critical rationalism that I think many, many critical rationalists have completely missed. Okay. So I’m highlighting that for a future podcast.
[01:15:47] Red: Okay.
[01:15:48] Blue: Let me now make a, now that I’ve used that immune system example, let me actually now explain, offer you an explanation for why I believe the difference is actually how not what. Okay. So let’s, let’s do another hypothetical scenario. The issue is that the specialist of the two sources is not in what it creates, but how it creates it.
[01:16:13] Red: So
[01:16:14] Blue: here’s an illustration. Suppose you want to save time getting home from work each day and you happen to have two devices available to you to help you figure out a solution to this problem. So device number one is a standard GPS with an A star algorithm that finds the shortest or if you prefer a fastest path home. Okay. So that’s your device number one. Then you also have available to you device number two device number two is an AGI that has human level of creativity plus access to device number one.
[01:16:49] Red: Okay.
[01:16:50] Blue: Okay. Which device are you going to prefer to pose that question to given that it’s, it’s free in this case to go to either one.
[01:16:58] Red: Well, I mean the AGI, right?
[01:17:01] Blue: Of course. Well, why? Yeah. I want you to actually think about why you would prefer to go to the AGI.
[01:17:06] Red: Well, I mean, you could probably handle more complexities that might come up. You’re just trying to get home. Is that what you said? Yeah,
[01:17:13] Blue: you’re just trying to get home.
[01:17:14] Red: You’re just trying. I mean, you know, who knows what could come up with there could be a natural disaster or a a war or something. You know, that’s right. Yeah. Okay.
[01:17:28] Blue: You answer the same as mine. I would much rather prefer device to presumably everybody would answer that.
[01:17:33] Red: Yeah.
[01:17:33] Blue: And
[01:17:34] Red: I just want to talk to an AGI. I mean, it could really have a nice, nice conversation on the way home. Okay.
[01:17:41] Blue: So let me let me say let’s pretend like I wanted to go with the kind of the standard pseudo Deutsch theory of knowledge answer this. I might try to answer. Well, I would prefer device number two because device number one voice number two is creative. So it creates knowledge where the other device doesn’t create knowledge. Okay. Let’s say I tried to answer that way. This is clearly a non -explanation like like you haven’t really explained to me at all why you prefer device number two at this point. Okay. It’s it’s a tautology.
[01:18:13] Red: So
[01:18:14] Blue: I’m going to offer a better explanation. Device number one by design must start with the assumption that you are driving a car or some other means and that the only way to save time is to search for the sort shortest path. But device number two being a human level creativity can try thinking outside that box. It may find that there is a huge shortcut if you’re willing to take a bicycle. It may discover a free helicopter ride home that you didn’t know about by searching the internet and finding that out. It may research and find that if you ride a motorcycle, you can legally go faster by zipping between cars and therefore this path is actually faster even though the GPS says otherwise. It may discover that a transporter was just invented and recommend that you use that, right? Okay. Now, of course, these are all kind of silly examples. Some of them are more silly than others. The motorcycle one’s probably maybe even a decent one. Okay. The transporter one is clearly silly. The important point here is that when we say that device two is creative and device one isn’t what we really mean is that device two is doing a far more comprehensive search of best solutions across a more open -ended amount of possibilities that in fact, it’s an open -ended amount of possibility. It is not defined in advance what you’re allowed to consider as a possibility. And it can also adapt to new ideas on the fly. Now, this is practically a restatement of what open -endedness means. Go back to the open -endedness episode that we did.
[01:19:44] Blue: On the other hand, most of those other options are probably unrealistic except possibly a motorcycle and most likely most likely device two and device one are going to end up giving you the same piece of knowledge at the end. That is the shortest route home. It is beyond bizarre to insist that even though both devices gave you the same piece of knowledge that just because one of them did a narrow search for knowledge and one of them did an open -ended search for that knowledge that somehow it mystically transforms the result into knowledge in one case and not in the other case. This is my real argument against the existing Deutsche’s theory of knowledge and why I believe he is in fact barking up the wrong tree. He’s trying to figure out what physically is different about the two sources of knowledge when in fact there is nothing physically different about them. What makes them special is that they are open -ended searches.
[01:20:39] Red: But it sounds to me like you’ve also said you see some validity in the two sources hypothesis. The two sources hypothesis is valid exactly this much that those are the only two open sort open open -ended search processes that we know of and that we don’t currently know what they’re doing that makes them open -ended. OK, OK.
[01:21:00] Blue: That is exactly how far I’m going to go with the two sources of knowledge. You can decide for yourself. Is that like 99 percent of the way? I think it kind of is right.
[01:21:08] Red: Yeah, yeah.
[01:21:09] Blue: OK, but like I can see why someone would say wasn’t that exactly the same as saying the two sources hypothesis is wrong? Well, yeah, it is. In fact, no matter how much I’m going along with it, I’m clearly also saying it’s wrong. This is typical of critical rationalism. We like false theories because that’s usually the only theories we have. We’re trying to figure out which one has the most most verisimilitude, right? So yes, that that is exactly. So I do think the two sources hypothesis needs to be no longer be thought of as. There are only two sources of knowledge. I think that is literally just false.
[01:21:43] Red: So what you might agree with if we said there are two sources of mysterious and open -ended. Knowledge creation of knowledge.
[01:21:55] Blue: And then there’s a bunch of
[01:21:55] Red: you would agree with that. But then there’s all. Yeah, all kinds of other ways we can create knowledge, but it’s like they’re really not as interesting. That’s right. You would agree with that.
[01:22:05] Blue: I would totally agree with that. OK.
[01:22:08] Red: OK, well, I guess I agree with that too. Yeah.
[01:22:12] Blue: And I believe this is the correct way to understand the two sources, how the two sources are special because they represent an open -ended search. I mean, the problem of open -endedness, not because they produce something physically different than other processes.
[01:22:24] Red: OK,
[01:22:25] Blue: OK. So I agree.
[01:22:28] Red: I agree. You’ve convinced me.
[01:22:29] Blue: OK. Now, given the above argument, I have a proposal. Here’s my proposal. Let’s stop stupidly calling the alternative theory, the theory of simul knowledge and instead call it what it really is, the theory of knowledge. And instead, let’s refer to the constructor, the two sources hypothesis as the problem of open -endedness. And let’s be done because I think that’s actually the more correct way to refer to these problems. Here’s also if I’m right about that, let’s leave a little the door open. I feel like the argument I just made is completely fatal. Like I don’t think there is any way whatsoever to show that I’m wrong about this because I think I’m right. But of course, I don’t know that for sure. So I’m going to leave the door a little bit open. Let me say this, that I’ve laid out what you would need to do to show me that I’m wrong. You would need to actually come up with an actual set of criteria that really does get us back to the two sources and that doesn’t doesn’t get refuted by the two hypotheticals that I just offered the immune system example in the device one and device two example. Okay. And I don’t think that’s possible, but it could be wrong, but that’s what you would need to do. You would need to explain to me what’s physically different about those two scenarios. Now, as it turns out, the original criteria. So when we talk about the Deutsche’s constructor theory of knowledge, if you drop the two sources hypothesis, if you assume that that’s actually just a poor way of looking at the problem of open -endedness and ignore it and instead only look at the original criteria.
[01:24:01] Blue: It’s adapted information that keeps itself instantiated. It turns out I’m going to argue that that is now identical to Popper and Campbell’s evolutionary epistemology. And again, I don’t know that for sure, but here is what you would do to prove me wrong. You would try to find an example of adapted information that causes itself to remain so that did not come through a process of blind variation and selective retention. Any even a single counter example would show that these two are not equivalent. I can’t think of any. So I think they’re equivalent. I think Campbell’s evolutionary epistemology or Popper’s evolutionary epistemology, since they’re more or less the same thing, is exactly identical to Deutsche’s constructor theory of knowledge, but that the difference is, is that Campbell and Popper thought in terms of how knowledge gets created through a process of variation and selection. And Deutsche was thinking in terms of what’s physically different about knowledge. And I think that is the true value of the constructor theory of knowledge is that it took evolutionary epistemology and it defined it in such a way that it defined it in terms of what it physically was. But I think as long as you hold on to the two sources hypothesis, the constructor theory of knowledge is badly broken and that the only way to save it is to drop the two sources hypothesis and admit that Campbell and Popper are correct, that there is a giant hierarchy of knowledge creation out there and it’s not just the two sources. It’s just some of them are open -ended and some of them aren’t. Okay. This is why I say I don’t really feel like I’m negative on Deutsche’s constructor theory of knowledge.
[01:25:49] Blue: In fact, I think Deutsche’s constructor theory of knowledge absent the two sources hypothesis, which I believe to be fatally wrong. I believe that it is exactly the right thing to do that that when Campbell and Popper tried to define knowledge in terms of how it got created that that was a mistake. It was a good first attempt. Like it’s not a mistake like it’s wrong. It’s just like we don’t really want to define knowledge in terms of how it got got created. We really want to define knowledge in terms of what it physically is and that Deutsche had exactly the right approach to that, but that he accidentally attached a false theory, the two sources hypothesis, which was really just a misunderstanding of the problem of open -ended is and that that has been a weight on the theory ever since and that we really you’ll know the constructor theory of knowledge is back on track once we finally dropped the two sources hypothesis. That’s my prediction. Okay. Now that I’ve shared my opinion on the subject, let me go ahead and try to steel man the other point of view as much as I can here. So first of all, how would you go about refuting what I’m suggesting? Maybe let’s start with the example of creating an antibody. So I’ve been talking as if in the two scenarios is the exact same antibody. It would really have to be the same recipe, but let’s say that the scientists that came up with a creative solution that they created the recipe in DNA that doesn’t seem that unreasonable to me. So we have the exact same recipe.
[01:27:14] Blue: You would probably have to figure out how to define knowledge such that there’s some other connection to something else that is different between these two scenarios. I don’t know what that is. That’s why I’m kind of doubtful that this exists, but I don’t think it’s completely impossible. So that would be where you’d probably need to go with this as far as the two devices that is to say having a GPS versus having an AGI try to find the quickest way for you to get to work or whatever. I don’t really know what you would do to try to refute that example that that’s a tough one, but maybe there is something you can come up with across the multiverse that there’s some sort of physical difference between these two, even if at the end of the day they just produce the exact same route. And while I’ve suggested that we should just drop this whole idea of a theory of simul knowledge versus a theory of knowledge and we should just say there’s one theory of knowledge. Again, that’s my own suggestion. If you’re going to insist to me that there are only two sources of knowledge just by definition, biological evolution and human ideas, then that’s fine. I will around you refer to my theory of how I understand constructor theory absent the two sources hypothesis. I will call it the theory of simul knowledge and I’m completely fine with that because I really don’t care about the word knowledge or any word at all. It’s the concept that I’m interested in. It does seem to me that people are unnaturally attached emotionally to the word knowledge in a way that really isn’t healthy and I don’t want to replicate that problem.
[01:28:47] Blue: So I’m completely fine with anyone who wants to discuss this with me. I will call my theory the theory of simul knowledge, simulacrum of knowledge and you can call yours the theory of knowledge and it does not make the slightest bit of difference to me. Let’s now take a look at some of the problems that we talked about in the past. I believe that this new way of thinking of the theory of knowledge or if you want to prefer to call this theory of simul knowledge, I’m fine with that too. Let’s talk about the credit assignment problem. This expanded view of knowledge also bypasses the credit assignment problem. We sort of just don’t care how much credit to assign to different levels of the hierarchy. Clearly all knowledge is in some sense created in some very large hierarchy of variation selection algorithms. All of them deserve credit. But if I ask, does the robots evolutionary algorithm create knowledge? I know what I’m really asking is, is the algorithm playing approximate role in directly creating the adaptive information that it outputs and the answer is ever sounding yes. The credit assignment argument that gets used can now be seen for what it is. It’s like asking, who is your one ancestor? And if the person answers, my dad is John and you wanted him to answer, my grandpa is Larry. You can just insist they answered wrong because the question was ambiguous. The credit assignment problem is intentionally ambiguous as a means of trying to get back to the two sources as a dogma. That’s what’s really going on with the credit assignment argument.
[01:30:14] Blue: Furthermore, let’s say I’m wrong having said that let’s assume I’m wrong and that someday someone does offer me a criteria acts that objectively separates simul knowledge from knowledge using only physical properties. I’m fine with realizing that both of these theories are interesting in their own way and they have an important relationship that knowledge is a subset of simul knowledge. This makes the theory of simul knowledge the more encompassing theory and the one that aligns with poppers evolutionary epistemology. But now let’s ask, what if I’m right? Okay, so if I’m wrong, that’s not a bad thing. Like all someday go, oh, you’re right. There’s actually a difference between knowledge and simul knowledge. That’s that’s something I wanted to know me me pointing out that we didn’t know what the difference was. That was a legitimate problem that we needed to be working on. Okay, so if I’m wrong, sorry, if I’m I’m wrong, that still means I was making progress towards the real a real problem, which is trying to figure out what that criteria X was. Okay, but what if I’m right? Then there is a fatal flaw in Deutch’s current constructive theory of knowledge, namely the two sources hypothesis and it will need to be air corrected at some point. I hope you can see why that would be an important thing to know and that therefore these arguments should be considered. Okay, need to be taken seriously. This also solves the problem of animal memes. Deutch wanted to declare animal means not knowledge. Here’s the quote page 409 from beginning of infinity. This is how eight memes can be replicated without the impossible step of literally copying knowledge from another eight, another eight.
[01:31:41] Blue: And he had previously explained that it was all statistical, right? Just a statistical correlation. But this is a contradiction to the very idea that memes are knowledge outside the genes, which he admits on page 37 like genes, all memes contain knowledge, often inexplicit of how to cause their own replication. The issue here is that Deutch is conflating in the page 409 quote knowledge and explanatory knowledge as if they’re the same thing when in fact, they aren’t. So on page 407, he actually says this. He says, but recent discoveries have revealed how apes are able to imitate such behavior without ever creating any explanatory knowledge. He pulls this shell game. I don’t think he did it on purpose at all. I think he just didn’t occur to him that explanatory knowledge and knowledge aren’t the same thing that one’s a subset of the other, right? That what animals create with animal means is an explicit knowledge instead of explicit knowledge. And that’s actually what separates human memes or at least most human memes from. I think human memes are often actually in explicit also factors. I read a book about this, that there’s actually overwhelming evidence that for a very long time, most human memes had no explanations behind them at all and that if you ask the people to explain their their knowledge, their ceremonies, they do. Most of them had never given it thought. So I actually don’t think that all human memes are explicit knowledge and do it kind of argues that they are and I don’t think that’s right. But human memes include explicit knowledge and implicit knowledge, whereas animal memes are entirely implicit knowledge.
[01:33:19] Blue: But they are knowledge and this need to try to remove them from being knowledge leads you to a contradiction, namely that memes are by definition knowledge outside the genome, but animals don’t create knowledge. And what you’re really doing is as you’re switching between two uses of the word knowledge, one being simul knowledge and one being the more restricted mysterious open -ended kind of thing that gets created by the two sources. And Deutsche never seems to have picked up on the fact that’s what he’s actually doing is that he’s actually conflating these two things by accepting this expanded understanding of knowledge. The problem goes away. Animal memes are knowledge outside the genome, but they are non -explanatory. I also note here that Bern may actually argue that point. Bern argues that some animals do have a limited kind of explanatory knowledge, so I’m not going to try to resolve that. But what I’m going to say, though, is that the contradiction in beginning infinity goes away. What Deutsche was really trying to say was that animal memes are knowledge outside the genome, but they are non -explanatory knowledge. And if read that way, he’s completely correct, right? At least with the examples that he used.
[01:34:29] Red: Just to be clear, are you saying that Deutsche’s theory of knowledge can be rescued by just inserting the word explanatory?
[01:34:39] Blue: No,
[01:34:40] Red: it can’t. Into it? No, you’re not saying
[01:34:42] Blue: that. It always has to be rescued the opposite way. So if you don’t believe me, try to do it. Try to rescue it by inserting explanatory into it. So let’s say you do that. So you say, okay, to be knowledge, it must be explanatory. Okay, you just eliminated one of the two sources. So it cannot be rescued that way.
[01:34:59] Red: Oh, I see. Yeah. Okay, of course. Okay.
[01:35:02] Blue: Now, of course, you could rescue it in the sense of you could declare the one source hypothesis.
[01:35:08] Red: The one source. Okay.
[01:35:10] Blue: And but then I would still argue with you. I would still say, well, there’s not actually one source because there are other sources of explanatory knowledge. And I would probably cite burn that some animals create explanatory knowledge. I would probably cite explanation based learning that that’s a kind of explanatory knowledge. And what you would quickly find is that you still can’t rescue the hypothesis this way.
[01:35:32] Red: I see. Although, although you’d be a lot closer. Like you might be able to convince me that animals don’t really create explanatory knowledge that there is something physically different about animal explanations versus like, since we don’t know enough about explanations, my case becomes much weaker at that point.
[01:35:50] Blue: But you have for sure eliminated biological evolution, Darwinian evolution.
[01:35:54] Red: Okay. What if we just inserted the word open ended then? Okay. Can I save it? Is it that it could be that simple? Okay.
[01:36:03] Blue: No, you can’t. And here’s why.
[01:36:05] Red: Okay.
[01:36:06] Blue: Because. Deutsch is on the right track with this constructor theory of knowledge that the goal is to not define knowledge in terms of how it got created.
[01:36:15] Red: Yeah.
[01:36:16] Blue: So the moment you insert and was created by an open ended source of knowledge creation, you have now ruined the entire point of the constructor theory of knowledge.
[01:36:26] Red: Okay. I get it now. Yeah. It’s it’s more about open ended. It’ll be more about the process. It is like the actual thing. Okay.
[01:36:34] Blue: That’s right.
[01:36:35] Red: Okay.
[01:36:35] Blue: Okay. So let me now summarize. My point of view now that I’ve kind of taken you through this and etc. The theory of song simul knowledge that I laid out started out. I started out acting like we’re just making a theory about something that isn’t knowledge and I did my best to bend over backwards to make that possible. But then it was just some kind of adapted information that wasn’t knowledge, but to even be able to go about creating such a theory. It forced us to confront the impossibility of dutch’s theory ever being squared with the two sources hypothesis without just turning his theory into a bad explanation. So I personally declare the two sources hypothesis dead and I even explained why it can never work both by giving an example that would be the identical antibodies thought experiment of how to demonstrate the impossibility of the task as well as an explanation as to why the task is impossible. That would be the GPS versus the AGI device thought experiment. Okay. So that is why I believe the two sources hypothesis is dead. I think you could reform the two sources hypothesis as not being a theory of knowledge, but instead as being the same as the problem of open -endedness. And I think you could save that much of it, but no more. Now, if I’m wrong, I’ve shown that the theory of simul knowledge is still a sub super set of dutch’s theory of knowledge and thus equally or more interesting. It would be pointless to discuss any theory of knowledge without also referencing this theory of simul knowledge. Just like it would be pointless to have a theory of of finite state machines while ignoring a theory of Turing machines.
[01:38:17] Blue: Sadly, I know from experience that my extended argument here, although I do think it is it is fatal to the whole two sources of hypothesis. I know it will likely not be accepted. And what we’re really going to turn to now is why I think my argument, even though I think it is lock solid is not likely to be accepted by the online critical nationalist broader broadly because I believe that they have a flaw in their epistemology that they haven’t caught on to yet. And that’s what we’re going to turn to next. And in fact, this whole thing about the theory of knowledge was actually a setup for what I really wanted to talk about, which was what I see as a fatal problem with the epistemology of most online critical rationalists and how to fix it.
[01:39:10] Red: Well, I think you’ve made a very compelling case. I hope that our audience will be willing to stick with the number of episodes required to get there. But I guess I cannot no longer try to sound smart at a party by talking about the two sources hypothesis. So
[01:39:34] Blue: thanks, Bruce. You’ve ruined that, ruined that for me, but yeah. So we’re going to actually now turn to so Peter has challenged me many times to put together a podcast where I explain what I understand critical rationalism to be with the adaptations that I’ve talked about on the podcast. So for example, we did several podcasts about Popper without refutation. So Peter’s challenge is can you like give me on our or less podcast that explains critical rationalism, but without referencing the word refutation, right? Right or not directly referencing it. And so I’m going to have not not going to do that just yet, but I’m kind of building up to that. There’s one more piece of thing I need to work out myself and I need to explain. And then I think I’ll be ready to try to address Peter’s challenge where I’m going to attempt to give critical rationalism in a completely it’s the same as critical rationalism that Popper taught in my opinion, but it’s reworded in such a way that it has it really causes you to think about it in a completely different way. And I think removes several errors that I’ve seen online critical rationalists use that I think causes them to not think rationally very well at times because they’ve misunderstood some aspect of Popper’s epistemology because of the way it was worded or the way it’s sort of been digested. And so I like really one of the main things I’m going to do is I am going to try to actually explain Popper’s epistemology without using the term refutation.
[01:41:10] Blue: And I think that that actually causes you to see Popper’s epistemology in this new light that really does enlighten your thinking and you go, Oh, wow, like people are really misunderstanding certain aspects of critical rationalism because of the use of the word refutation. And so I’m going to take that challenge up, but we’re still not quite there yet.
[01:41:30] Red: Well, I will look forward to that. And we’re wrapping up wrapping this up for the or are you talking about you’re going to do this today? No,
[01:41:38] Blue: no. OK,
[01:41:38] Red: OK,
[01:41:39] Blue: yeah,
[01:41:39] Red: we’re wrapping up for now. OK, well, if you are our audience and you’ve stuck stuck with this.
[01:41:47] Blue: First of all, why did you like? I mean, like this is there’s so many episodes, you know, you have to understand that this podcast is at the end of the day, you know, my search for the truth, right? It’s I’m on live live on air trying to work out the problems that I see and that I find the problems that I find interesting and I’m trying to find solutions for them. And I don’t always succeed, right? I’m much better at finding problems than finding solutions for them, which is, you know, a common problem. But it is ultimately you’re listening to this podcast if you think that these are interesting problems and you find them fascinating. If you don’t, then this is a lot of pain, right? To have to go through.
[01:42:35] Red: Well, you’ve convinced me that there’s their interesting problems and I and I have found your solution and actually very, very compelling. And I hope someone else does. If you do, please, please reach out and let us know and or at least leave a leave a positive review on Spotify or something. Maybe a like. I don’t have five stars right now. Please.
[01:43:02] Blue: Yes. Thank
[01:43:04] Red: you, Bruce. I thank you for all of this and I hope you have a have a wonderful day. All right. Thank
[01:43:10] Blue: you very
[01:43:10] Red: much. Thank you.
[01:43:11] Blue: All right. Bye -bye. The theory of anything podcast could use your help. We have a small but loyal audience and we’d like to get the word out about the podcast to others so others can enjoy it as well. To the best of our knowledge, we’re the only podcast that covers all four strands of David Deutsch’s philosophy as well as other interesting subjects. If you’re enjoying this podcast, please give us a five star rating on Apple podcasts. This can usually be done right inside your podcast player or you can Google the theory of anything podcast Apple or something like that. Some players have their own rating system and giving us a five star rating on any rating system would be helpful. If you enjoy a particular episode, please consider tweeting about us or linking to us on Facebook or other social media to help get the word out. If you are interested in financially supporting the podcast, we have two ways to do that. 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 for strands.org. There is a donation button there that uses PayPal. Thank you.
Links to this episode: Spotify / Apple Podcasts
Generated with AI using PodcastTranscriptor. Unofficial AI-generated transcripts. These may contain mistakes; please verify against the actual podcast.