Episode 124: Popper’s Evolutionary Theory of 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
[00:00:00] Blue: This week on the Theory of Anything podcast, Bruce continues his exploration of evolutionary epistemology, or the idea that knowledge creation in human minds is analogous to natural selection. Specifically, Bruce discusses how Karl Popper’s critical rationalism is and is not related to Donald Campbell’s concept of evolutionary epistemology, which it seems Popper mostly endorsed. But is Popper’s conjecture and refutation really the same as Campbell’s blind variation and selective retention? I enjoyed listening to Bruce in this somewhat shorter episode and I hope someone else out there does too.
[00:00:49] Red: Welcome to the Theory of Anything podcast. Hey, Peter. Hello, Bruce. How are you doing today? Doing well. Today we’re going to talk about Karl Popper’s evolutionary epistemology, or that’s really not the right term for it. It’s going to be Karl Popper’s theory of evolutionary theory of knowledge was what he prefers to call it.
[00:01:10] Blue: But those are the same things. It sounds kind of similar, but okay.
[00:01:16] Red: So in episode 114, we talked about Donald Campbell’s evolutionary epistemology, and we found that most of Campbell’s epistemology actually came from Popper first by Campbell’s own admission. Moreover, Popper said he strongly agreed with Campbell. I read the first part of the first couple paragraphs of Popper’s response to Campbell, where he gives a glowing response to Campbell. So I thought we would do an episode where we look at what Karl Popper actually said about evolutionary, what Campbell later called evolutionary epistemology. A lot of it written before Campbell and then some of it written after Campbell. I’ll try to make somewhat of a distinction. Now, Campbell wrote about it over the course of 20 years, so there’s no clear period of before and after Campbell. But Campbell does a lot of quoting of Popper, so we know all that must have come before. And then there’s a bunch that Popper does near the end of his life that would have been after Campbell, and so we’re going to quote some of that too.
[00:02:16] Blue: Now,
[00:02:16] Red: recall that Campbell coined the term evolutionary epistemology, but that Popper did not like it. Popper preferred the term evolutionary theory of knowledge. I’ll actually give you the quote here when he talks about that. So responding to Campbell’s term evolutionary epistemology, Popper later says in a talk entitled Towards a Theory of Evolutionary Epistemology, so he titled his talk Towards a Theory of Evolutionary Epistemology, just to be clear. In that talk, Popper says, I now feel that evolutionary epistemology, that’s in quotes, it’s a term, sounds pretentious, especially since there exists a less pretentious equivalent. So please let me change the title of my talk to its equivalent, and let me call this inaugural address Towards an Evolutionary Theory of Knowledge, that’s all from All Life is Problem Solving, page 57.
[00:03:14] Blue: And did you say this is the 70s we’re talking about? No,
[00:03:18] Red: no, All Life is Problem Solving. I guess I should have looked up. That was like a later book. I think that was his last book.
[00:03:24] Blue: Yeah, yeah, yeah, that’s right, but I’m in the 90s basically.
[00:03:28] Red: And then on page 49 of the same book, All Life’s Problem Solving, he says, this is a different talk, he says, what others have called my evolutionary epistemology, I have pointed out before that I did not apply that term to my theory of knowledge. It was others who described my theory of knowledge as evolutionary. So for this reason, throughout this podcast, I’m going to call this Popper’s evolutionary theory of knowledge in respect to what he wanted to call it. But let me be very blunt here, it’s exactly the same as saying Popper’s evolutionary epistemology. And honestly, the modern term is evolutionary epistemology. And so I don’t know why he thought it was pretentious. I have no idea why he thought it was. One thing I would note is that he uses the word towards such a theory, which I think clearly does mean that he doesn’t feel we have a full fledged theory yet. Finally, we’re going to discuss what Popper disagreed with Campbell over, which as we’ll see, wasn’t very much.
[00:04:28] Blue: May I interject? Yes. At least according to chat GPT, which sounds pretty accurate here, that there was a foundation built on this idea in conjectures and refutations and the logic of scientific discovery. So in the 30s through 60s, but this says that the explicit evolutionary epistemology was not used by anyone until including Popper until 1970s, basically. That would be Campbell’s paper where he coins the term, the one we covered in episode 114. And it sounds like towards a theory of evolutionary epistemology is it says around 74 to 76.
[00:05:19] Red: Okay, so that would be soon after or maybe that isn’t after, that might be right at the same time.
[00:05:26] Blue: Okay, interesting. Okay, yeah, here it says 74.
[00:05:30] Red: I mean, I guess it would have to come after when Campbell coined the term. He can’t have used the term before it was coined or he would have coined it. So okay, yeah. Okay, so a quick review of Campbell’s version of evolutionary epistemology. So this is from page 56 of evolutionary epistemology, rationality and sociology of knowledge, as we covered in episode 114. So quote, a blind variation in selective retention process is fundamental to all inductive achievement to all genuine increases in knowledge, to all increases in fit of system to environment. So this will become important later when we’re attempting to fit this first point, we’re attempting to fit evolutionary Campbell’s evolutionary epistemology into modern machine learning. Since no one that I know of would deny machine learning is a legitimate inductive achievement of generalization from observations and an example of fit of system to environment. So machine learning offers us a potential counter example or corroboration to Campbell’s evolutionary epistemology that Campbell could and Popper could not have known about. Number two, in such a process, there are three essentials. A, mechanisms for introducing variation, B, consistent selection processes and C, mechanisms for preserving and or propagating selected variations. Number three, this is still all quoting Campbell. The many processes which shortcut a more full, full blind variation in selective retention process are in themselves inductive achievements containing wisdom about the environment achieved originally by blind variation and selective retention. This is Campbell’s idea of a hierarchy of blind variation selective retention processes. Number four, in addition, such shortcut processes contain in their own operation a blind variation and selective retention process at some level substituting for overt locomotor exploration or the life and death winnowing of organic evolution. Okay, that’s the summary.
[00:07:40] Red: That’s Campbell’s own summary of his evolutionary epistemology. Keep that in mind now as we talk about Popper’s early views that influenced Campbell, which is what I’m going to now quote Popper on. Let’s begin with Popper’s view on epistemology itself, which he says is specifically about the problem of, quote, growth of knowledge. This is Popper in Logic of Scientific Discovery in the introduction, so page XIX. The central problem of epistemology has always been and still is the problem of the growth of knowledge. The growth of knowledge can be studied best by studying the growth of scientific knowledge for the most important way in which common sense knowledge grows is precisely by turning into scientific knowledge. Moreover, it seems clear that the growth of scientific knowledge is the most important and interesting case of the growth of knowledge. Popper adds that it is easier to study scientific knowledge rather than common sense knowledge, quote. But all past philosophers discovered that scientific knowledge can be more easily studied than common sense knowledge. That’s from page XXV. So let’s be clear. To Popper, epistemology is specifically about understanding how knowledge grows and his study into scientific knowledge, his scientific epistemology, which is what he’s most famous for, was simply in his mind an attempt to study that because it was the most clear cut and interesting case of the growth of knowledge. But he was actually interested in the concept of the growth of knowledge in general. Now, why am I emphasizing this?
[00:09:23] Red: If you heard the previous podcast to this one, one of the things I made a very big deal about was when looking at Deutsches, or maybe we say pseudo -Deutsch theory of knowledge, there was this idea that somehow a genetic algorithm does variation selection, does find some sort of adaption that allows the robot to walk, and yet somehow isn’t a growth of knowledge. I’m emphasizing this to really drive home the point that to Popper, that just doesn’t make any sense. It’s the growth of knowledge. So this is at odds with the Deutsch view, where some kinds of variation in selection or trial and error processes find solutions to problems, like the walking robot example, but aren’t to be considered knowledge. To Popper, it was all considered knowledge. So we know Popper would have rejected Deutsches’ two sources hypothesis based on this. So Popper never saw his epistemology as explicitly about science, and in fact, he thought there was only a single epistemology even for fields outside what we’d normally think of as science. So from logic of scientific discovery page XIX, he says, and yet I am quite ready to admit that there is a method which might be described as the one method of philosophy, that’s in quotes, that’s like quotes within the quotes, but it is not characteristic of philosophy alone. It is rather the one method of all rational discussion, and therefore of the natural sciences as well as of philosophy. The method I have in mind is that of stating one’s problems clearly and of examining its various proposed solutions critically. Notice how this ties into our episode on rational fallacies, including not not merely the idea that we criticize our theories, but that we choose to formulate our theories such,
[00:11:23] Red: meaning sharply or explicitly, that they can be easily criticized. This is what Popper was saying all along. Okay, but what about animal learning? So David Deutsches argued that all an animal’s knowledge is in its genes, and by implication that animal learning is not a growth of knowledge. This is a necessary implication of Deutsches’ two sources hypothesis. So of course, that’s why he puts this forward. So Deutsches even goes so far as to claim that animal memes, which he admits really do exist, are somehow actually knowledge in the animal’s genes, even though the term meme is generally understood to tautologically be a kind of culture of knowledge outside of the genome. So famously, Donald Campbell, episode 114, strongly disagreed with Deutsches on this. Campbell took the idea that animal learning was a kind of evolutionary epistemology, rather than induction, from Popper himself. Here is one such Popper quote that Campbell used, conjection refutation page 292, although I shall confine my discussion to the growth of knowledge in science. My remarks are applicable without much change, I believe, to the growth of pre -scientific knowledge also, that is to say, to the general way in which men and even animals acquire new factual knowledge about the world. The method of, and this is the method of learning by trial and error, of learning from our mistakes seems to be fundamentally the same, whether it is practiced by lower or by higher animals, by chimpanzees or by men of science. So notice how Popper is agreeing with Campbell here, not Deutsches. Popper again emphasizes, that this study of science was really to simply understand the broader idea of the growth of knowledge.
[00:13:11] Red: Quote, my interest is not merely in the theory of scientific knowledge, but rather in the theory of knowledge in general. Yet the study of the growth of scientific knowledge is, I believe, the most fruitful way of studying the growth of knowledge in general. For the growth of scientific knowledge may be said to be the growth of ordinary human knowledge writ large. Note how, between these two quotes, there should be no doubt that Popper did see animal learning as a kind of growth of knowledge, typical of this more general epistemology of trial and error learning that Campbell called evolutionary epistemology. Thus we know for sure that Popper would not have agreed with David Deutsches, that all of an animal’s knowledge was in its genes. Now I suspect some Deutsches and crit rats will try to argue with me here, that actually that last quote was meant only that animals learn by dying, so we’re talking about biological evolution. I don’t read it that way, and I think that’s a stretch, but let’s, for the sake of argument, assume that that’s maybe a possibility. Let’s see if we can get a more clear cut quote here. In fact, there are quotes from Popper, we just partially quoted one of them, where he claims that Einstein can learn in his lifetime by criticizing his own theories, whereas Amoeba has to die for the species to learn. So there are quotes you can find of Popper, where he is talking about animals dying, but he doesn’t always talk about it that way. So if you are in doubt that Popper did agree with Campbell that animal learning is a kind of evolutionary epistemology, and that knowledge growth here is a far clear example of Popper stating this.
[00:14:47] Red: So this is now from realism in the aim of science, one of his later books, page 97. Dispositional preparedness, so think about how genes give you expectations and so you have some sort of disposition towards something, that’s what he’s talking about here. Dispositional preparedness for what is to come seems to be the true biological analog of scientific knowledge. In an animal organism, disposition to react in a certain manner to certain kinds of stimuli are partly inborn, in other words partly genetic. My thesis is that so far as these dispositions are acquired in the lifetime of the animal, they are modifications of inborn dispositions which are plastic and which develop and change upon being activated by stimuli and especially also under the influence of failure and success, coupled perhaps with painful and pleasurable feelings he’s talking about classical conditioning and operant conditioning here. In this way the organism develops its inborn dispositional knowledge, the genetic knowledge, and it learns by trial and error. So that was all page 97. Hopefully this puts to rest the idea that Popper would have ever agreed with Deutsch that all of an animal’s knowledge is in genes. That whole position just is inconsistent with Popper’s theory of evolutionary knowledge. Now let’s consider Popper’s view on induction. This is from conjecture refutation page 60. Hume, having cast out the logical theory of induction by reputation, he struck a bargain with common sense, merely allowing the re -entry of induction by repetition, in the guise of a philosophical fact. So Popper here is referring to the fact that Hume, despite his criticism of induction, accepted induction still as real and correct, just unjustified or worse irrational.
[00:16:39] Red: As we discussed in episode 116 when covering Strevins’ The Knowledge Machine, Hume’s attitude was that humans did use induction to create knowledge, it was just irrational to do so. But don’t worry, Hume’s view was, don’t worry about that, it’s just how things are, move on with your life. At least that’s the way Strevins presents him. Popper basically is presenting him in the proposal to turn the tables on this theory of Hume’s. Instead of explaining our propensity to expect regularities as the result of repetition, I propose to explain repetition for us as the result of our propensity to expect regularities and to search for them. Thus, I was led by purely logical consideration to replace the psychological theory of induction by the following view. Here’s the view now. Without waiting passively for repetitions to impress or impose regularities upon us, we actively try to impose regularities upon the world, that is to say we come up with a hypothesis to explain the observations. And now going back to Quoting Popper, we try to discover similarities in it, in the world, and to interpret it in terms of laws invented by us. Without waiting for premises, we jump to conclusions. These may have to be discarded later should observations show that they are wrong. This was a theory of trial and error, of conjecture and refutation. It made it possible to understand why our attempt to force interpretations upon the world were logically prior to the observations of similarities. Is Popper talking about science specifically here? No, he goes on to clarify that he sees science as just one example of this. Quote, since there were logical reasons behind this procedure, I thought that it would apply to the field of science also. Campbell here adds, quote,
[00:18:30] Red: Popper has effectively rejected the model of passive induction even for animal learning and advocates that here to the typical process involves broad generalizations from single specific initial experiences, generalizations which subsequent experience edits, that’s from page 50 of the other book, the evolutionary epistemology book that I mentioned. Again, we need to call out machine learning here. Can Popper’s view that induction by repetition requires prior theories that are tested be squared with actual modern machine learning? If so, then machine learning is actually a kind of evolutionary epistemology. Or does modern machine learning offer us a counter example and thereby refute Popper’s expressed view? Now, you may want to check out my episode 91 with the clickbait title, the Critical Rationalist Case for Induction, for a tentative first answer to this question. But as far back as episode 26, I’ve argued that this is a more difficult problem than it first appears and that there are some interesting potential counter examples that are hard to explain within the way Popper and Campbell see this. That being said, Popper then saw trial and error as the universally rational way to adopt, to adapt to the world. So from page 68 of Conjectural Refutation, Popper says, if we have made this our task, that is to adapt to the world we live in, then there is no more rational procedure than the method of trial and error, of conjecture and refutation, of boldly proposing theories, of trying our best to show that these are erroneous and of accepting them tentatively if our critical efforts are unsuccessful. From continuing the quote, from the point of view here developed, all laws, all theories, remain essentially tentative or conjectural or hypothetical, even when we feel unable to doubt them any longer.
[00:20:36] Red: Before a theory has been refuted, we can never know in what way it may have to be modified. That quote claims trial and error is the most rational, but in fact Popper elsewhere makes it clear he really meant it’s the soul way knowledge can ever grow. Quote, this is quoting Popper now, when discussing how to move from observation statements to a good theory, i.e. how to induce things from observation statements, the answer is quote, by jumping first to any theory and then testing it to find whether it is good or not, i.e. by repeatedly applying the critical method eliminating many bad theories and inventing many new ones,
[00:21:19] Blue: there is no other way,
[00:21:22] Red: end of quote. So Popper is arguing there is only one kind of true kind of quote unquote induction, and that induction in this case meaning just moving from observation statement to general theory, and that is his epistemology of trial and error of conjectural refutation. This is why Popper claims that you can refer to his epistemology of critical discussion as induction if you wish, quote, this is from Myth of the Framework. So I may say now that I do not believe there is such a thing as an inductive method or an inductive procedure, unless indeed you decide to use the name induction for that method of critical discussion and of attempted refutations, which I have discussed here. And then he goes on to say, I never quarrel about words, and I have of course no serious objection if you wish to call the method of critical discussion induction. But if you do, then you should be aware of the fact that it is very different from anything that has ever been called induction in the past. Popper is claiming this is the sole and only way to induce anything, and by extension the sole way of knowledge to grow. Note here that Popper is clearly equating what Campbell called inductive achievement with growth of knowledge. Those are to Popper and to Campbell, one and the same thing. So when Campbell argued in episode 114 that all inductive achievement was actually evolutionary epistemology, Campbell was getting this from Popper. I raised this in preparation for a future podcast in which I will discuss my disagreements with Ella Hopner over how to best interpret Campbell’s theory, but we don’t need to get into that today. A similar quote from Popper to drive the point home here.
[00:23:04] Red: While I read this quote, recall in the previous episode where the genetic algorithm that causes the robot to walk from Deutsche’s beginning of infinity and the immune system both solve real world problems, yet Deutsche claims that they do not create knowledge. So Popper says, quote, this is from Objective Knowledge, page 142, epistemology from the objectivist viewpoint becomes the theory of problem solving, or in other words of the construction, critical discussion, evaluation, and critical testing of competing conjectural theories. Again, I see no way in which you can reconcile this statement with Deutsche’s view, views any more than I see how you can reconcile Deutsche’s views to Campbell’s epistemology. So let’s summarize. We’ve seen that Popper has argued so far. All generalizations from observation may look like induction, but it’s really his epistemology of trial and error. Then he equates this to knowledge growth and claims all knowledge growth comes from this same process of trial and error. And we’ve seen that Popper sees animal learning as one kind of this trial and error epistemology, and that there simply is no other way by which these things can happen, arguing for a single unified epistemology that rejects any sort of induction unless, by induction, you mean his epistemology. I’m hoping you’re seeing at this point that Popper really was the basis for all of Campbell’s theory of evolutionary epistemology. It is probably more accurate to refer to it as Popper’s evolutionary epistemology, rather than Campbell’s. I’m doing no injustice to Campbell when I say this. Here is Campbell saying the same thing himself from page 89 of Evolution of Epistemology. This essay has identified Popper as the modern founder and leading advocate of a natural selection epistemology. The characteristic focus is on the growth of knowledge.
[00:24:58] Red: The problem of knowledge is so defined that the knowing of other animals than man is included. All right, now let’s take a look at Popper’s, he gave a speech and it became a chapter in his books that’s called of clouds and clocks. And it was one of the main basis for Campbell’s theory of evolutionary epistemology. And we extensively quoted from it in episode 114. We were quoting Campbell quoting Popper in this case. Let’s go directly and quote in full what Popper actually said. That speech was actually Popper’s attempt, in my opinion, a failed attempt to refute the philosophy of determinism. Determinism is in fact true, so you can’t refute it. But it contains an excellent selection on Popper’s views that inspired Campbell’s evolutionary epistemology with Campbell quoting it at length in his article. So I’m going to read that portion of the talk in full. So this is from Objective Knowledge, page 242 plus. I went for a number of pages. Popper says, my theory may be described as an attempt to apply to the whole of evolution what we learned when we analyze the evolution from animal language to human language. And it consists of a certain view of evolution as a growing hierarchical system of classical controls. There are plastic controls. There is the hierarchy that Campbell always talks about. And of a certain kind of organism as incorporating in the case of man evolving exosomatically, this growing hierarchical system of plastic controls. The neo -Darwinist theory of evolution is assumed, but it is restated by pointing out that its mutations may be interpreted as more or less accidental trial and error gambits. And natural selection is one way of controlling them by error elimination.
[00:26:52] Red: I shall now state the theory in the form of 12 short theses. Number one, all organisms are constantly day and night engaged in problem solving. Notice that this is all organisms, not just humans. And so are all those evolutionary sequences and so are all those evolutionary sequences of organisms that would be biological evolution. So notice who makes a distinction between animals being problem solvers and evolution as problem solvers. The phyla which begin with the most primitive forms and of which the now living organisms are the latest members know that Popper sees a unification here between animal learning and biological evolution, but he does not see them as one and the same. Number two, these problems are problems in an objective sense. They can be hypothetically reconstructed by hindsight, as it were. I will say more about this later. Objective problems in this sense need not have their conscious counterpart. And where they have their conscious counterpart, the conscious problem need not coincide with the objective problem. Number three, problem solving always proceeds by the method of trial and error, new reactions, new forms, new organs, new modes of behavior, new modes of behavior, animal learning, new hypotheses are tentatively put forward and controlled by error elimination. Again, notice that Popper unifies animal learning and biological evolution. Number four, error elimination may proceed either by the complete elimination of unsuccessful forms, the killing off of the unsuccessful form by natural selection, or by the tentative evolution of controls, which modifiers suppress unsuccessful organs or forms of behavior or hypotheses. Notice that Popper is outright saying it, animal learning is a legitimate kind of knowledge creation via evolutionary epistemology distinct from the biological process, biological evolution I mean.
[00:28:55] Red: Number five, the single organism telescopes into one body as it were, the controls developed during the evolution of its phyla, the biological evolution, just as it partially recapitulates in its ontogenetic development into phylogenetic evolution. Number six, the single organism is a kind of spearhead of the evolutionary sequence of organisms to which it belongs, its phyla. It is itself a tentative solution, problem solving into a new environment, niches choosing an environment and modifying it. It is thus related to its phyla almost exactly as the actions behaviors of the individual organism are related to this organism. Like, I don’t know how you can get more clear that Popper is saying that not all of animals’ knowledge is in its genes. The individual organism and its behavior are both trials, which may be eliminated by error elimination. Number seven, using P for problem and TS for tentative solution, a lot of people have quoted Popper saying this, this is a great quote, and EE for error elimination, we can describe the fundamental evolutionary sequence of events as follows. Problems lead to tentative solutions, which lead to error elimination, which lead to a problem. But this sequence is not a cycle. The second problem is in general different from the first. It is the result of a new situation which has arisen in part because of the tentative solution which has been tried out and the error elimination which controls them. In order to indicate this, the above schema should be rewritten. Problem one, tentative solution one, error elimination, problem two, eight, but even in this form an important element is still missing, the multiplicity of tentative solutions, the multiplicity of trials.
[00:30:35] Red: Thus our final schema becomes something like this, problem one, tentative solution one, error elimination, problem two, and then a number of different tentative solutions to try out. Number nine, in this form our schema can be compared with that of neo -darwinism. According to neo -darwinism, there is in main one problem, the problem of survival. There is, as in our system, a multiplicity of tentative solutions, the variations or mutations. But there is only one way of error elimination, the killing of the organism, and partly for this reason, the fact that problem one and problem two will differ essentially is overlooked, or else is fundamental importance is not sufficiently clearly realized. Number ten, in our system, not all problems are survival problems. Okay, we just said that all the problems of biological evolution are survival problems. Now he’s saying that in his system that’s not the case. There are many very specific problems and sub -problems, even though the earliest problem may have been sheer survival problems. For example, an early problem P one may be reproduction. Its solution may lead to a new problem, P two, the problem of getting rid of of or spreading the offspring. The children which threaten to suffocate, not only the parent organism but each other, it is perhaps of interest to note that the problem of avoiding suffocation by one’s offspring may be one of those problems which was solved by the evolution of multicellular organisms instead of getting rid of one’s offspring. One establishes a common economy with various new methods of living together.
[00:32:11] Red: Number 11, the theory here proposed distinguishes between problem one and problem two and shows that the problems or the problem situation which the organism is trying to deal with are often new and arise themselves as products of evolution. The theory therefore gives implicitly a rational account of what has been called by the somewhat dubious name creative evolution or emergent evolution. Number 12, our schema allows for the development of error eliminating controls, warning examples, warning organisms like the eye, organs like the eye, feedback mechanisms that is controls which can eliminate errors without killing the organism. There you go again. Animal learning is a form of knowledge creation to popper and it makes it possible ultimately for our hypotheses to die in our stead. Notice that he’s talking about animals here, not humans. Each organism can be regarded as a hierarchical system of plastic controls, of a system of clouds controlled by clouds. The controlled system makes trial and error movements which are partly suppressed and partly restrained by the controlling system. We have already met an example of this in the relation between the lower and higher functions of language. The lower ones continue to exist and play their part but they are constrained and controlled by the higher ones. Another characteristic example of this if I am standing quietly without making any movement that according to the physiologist, my muscles are constantly at work contracting and relaxing in an almost random fashion. See temporary solutions n to m of dc8 of the preceding section but controlled without my being aware of it by error elimination so that every little deviation of my posture is almost at once corrected.
[00:33:59] Red: Notice that this is an example of non -explanatory knowledge creation that he popper sees something as simple as your body subconsciously moving your muscles to keep you standing upright as a form of his evolutionary epistemology. Nothing at all to do with explanatory knowledge. Okay so I am kept standing quietly or more or less the same by more or less the same method by which an automatic pilot keeps an aircraft steadily on course. Notice that popper places the subconscious biological processes into the very same evolutionary epistemology as he does everything else but he also considers an AI which in this case is the the airplane autopilot as part of the same evolutionary epistemological framework and and as thus relevant to his overall epistemology. Campbell got this all from popper right like like it’s all there. This example also this is back to quoting popper this example also illustrates the theses one of the preceding section that each organism is all of the time engaged in problem solving by trial and error animal learning that it reacts to new and old problems by more or less chance -like or cloud -like trials which are eliminated if unsuccessful. If unsuccessful they increase the probability of the survival mutation which simulate the solution so reached and tend to and tend to make the solution hereditary by incorporating it into the spatial structure or form of the new organism. Here he’s talking about the Baldwin effect. Thank you he’s referring to the Baldwin effect here which we’ve talked about in past podcasts. Ah
[00:35:39] Blue: yeah.
[00:35:40] Red: Okay so let’s then so that was all prior to that’s the end of that of what popper said in that notice that that Campbell really got all of it from popper like it’s all there okay now popper then does work in evolution epistemology or in his theory of of evolutionary theory of knowledge after Campbell so let’s actually take a look at that so after Campbell coined the term evolution epistemology spent 20 years developing it popper also tried to develop the same theory you especially see this in the book all life is problem solving his final book i’m going to now quote a few important passages from that book that further clarify and elucidate what we’re calling popper’s evolutionary epistemology or his evolutionary theory of knowledge popper’s main emphasis was different than ours today he spent a huge amount of time in his effort developing his evolutionary theory of knowledge to refute empiricism which was as popper defined that term the theory that knowledge comes from our senses so from page 63 of all life’s problem solving popper says philosophers and even scientists often assume that all our knowledge stems from our senses the sense data which our senses deliver to us they believe that the question how do you know is in every case equivalent to the question what are the observations that entitle you to your assertion but to do this so that’s the end of the quote from popper but to do this popper had to make a distinction between an animal’s genetic knowledge found in its genes and the knowledge created via its ability to learn so quote page 63 but seen from a biological point of view this kind of approach is a colossal mistake for our senses tell us senses for our senses to tell us anything we must have prior knowledge here the prior knowledge is the genetic knowledge i’m going to pick out a few choice selections from popper where he makes this point thus argues popper quote i assume that it is only genetically a priority and not valid a priority not a prior not a prior necessary
[00:37:39] Red: not um apodactic apodactic apodactic i forget how to pronounce that meaning absolute certain knowledge is what that term means to put this in another way we learn only through trial and error there we go again our trials however are always our hypotheses they stem from us not from the external world all we learn from the external world is that some of our efforts are mistaken that’s page 47 so popper here argues that there is no way to learn other than via his generalized epistemology that we but not he is calling evolutionary epistemology continuing quoting popper quote my starting point is a very simple proposition indeed an almost trivial one the proposition that animals can know something that they can have knowledge if we stopped here dutch and crit rats would undoubtedly respond yes but that knowledge is purely genetic but popper continues on the quote for example a dog may know his master returns home on working days at about 6 p.m notice that this is clearly animal learning we’re now talking about you and that a dog does not have a genetic knowledge that his master is going to return home on weekdays at 6 p.m okay that is not what popper’s talking about he is talking about animal learning as actual knowledge that fits his epistemology and then he continues and the behavior of the dog may give many indications clear to his friends that he the dog expects the return of his master at that time ah so popper is now talking about genetic not talking about genetic knowledge here his thesis is actually that animals grow their knowledge via learning continuing the quote page 58 now I shall show that in spite of this triviality the proposition that animals can know something completely revolutionizes the theory of knowledge as it is still widely taught quote continuing same page now if you are interested in the theory of evolution you will find that part of it is the important theory of homology and that the dogs nose and my nose are homologous my um attribution of knowledge to the dog is therefore an anthropomorphism but it is not a metaphor rather it implies the hypothesis that some organ of the dog in this case presumably the brain has a function that corresponds not only in some vague sense to the biological function of human knowledge but is homologous with it all page 58 page continuing on page 59 now the existence of animal knowledge is not a mere metaphor but a serious evolutionary hypothesis
[00:40:19] Red: so popper is arguing that animal learning is knowledge growth how does this help his argument against empiricism to understand this he develops the idea that there are at least two kinds of knowledge relevant here first there is a priori knowledge in the genes that is of a general kind second there is knowledge created in the lifetime of the animal that is it’s learning that is specific to the animal and to its specific circumstance quote the distinction between adaption to or unconscious knowledge of lawlike and long -term environmental conditions such as gravity and the cycle of changing seasons on the one side that’s one kind of knowledge and adaptions to or knowledge of environmental short -term changes in events on the other side for example the the dog expecting his master to come home is of the greatest interest while short -term events occur in the life of the individual organisms gonna read that again is there’s just no doubt what poppers talking about here guys while short -term events occur in the life of the individual organisms that that kind of knowledge occurs in the life of the individual organism it is not part of the germ line it’s not in the dna he specifically says it’s in the brain the long -term and lawlike environmental conditions are such that adaption to them must have been at work throughout the evolution of countless generations notice the distinction there short -term knowledge that’s animal learning created through animal learning long -term knowledge that’s the biological genetic evolution that’s in the dna continuing the quote and if we look more closely at short -term adaptions at the knowledge of and the response to environmental short -term events then we see that the ability of the individual organism to respond appropriately to short -term events such as a particular pole of the wind or in the animal kingdom the appearance of a foe is also a long -term adaption and also the work of adaption going on through countless generations because it’s dependent upon that genetic knowledge so popper argues for two kinds or forms of knowledge long -term genetic knowledge and short -term knowledge via animal learning the reason poppers emphasizing this distinction between an animal’s genetic knowledge and their knowledge created via learning is because knowledge created through the senses is the second kind it’s part of animal learning poppers point is that you can’t have learning knowledge even through the senses without first having a priority genetic knowledge that one’s built on top of the other it’s the hierarchy that cambell always talks about quote i hope i have been able to give you some idea of the importance of the distinction between long -term and short -term adaption long -term and short -term knowledge and the fact that it must always proceed short -term or observational knowledge and of the impossibility that long -term knowledge can be obtained from short -term knowledge alone that’s page 64 note here how this fits nicely with cambell’s hierarchy of line variation and selective retention algorithms knowledge comes knowledge comes from knowledge and again we see there is no way to reconcile poppers evolutionary theory of knowledge with deutch’s two sources hypothesis or with his claim that all of an animal’s knowledge is in its genes quote page 51 these forms of knowledge or these adaptations especially among animals may be described as expectations the dog expects its master at half past five animal learning not biological knowledge it becomes restless and one can see it is prepared for its master to come home at half past five these are forms of knowledge says popper and these forms of knowledge are in are in every respect expectations that’s all page 51 still popper explains how animal learning is a type of adaption not unlike biological evolution this is from page 47 from primitive forms of life onward from the earliest cells adaption is an invention on the part of living creatures living creatures adapt and themselves improve their adaptions okay notice again the two kinds of adaption genetic and learning popper here refers here referring to adaptions in animal behavior due to learning as opposed to genetic knowledge built into genes he goes on to therefore argue quote page 49 life must be adapted to the future conditions of the environment in this sense general knowledge that is knowledge in the genes comes earlier than momentary knowledge than special knowledge i.e knowledge created via animal learning right from the start we must be in this way equipped with general knowledge i.e in the genes and with the knowledge we usually call knowledge of the laws of nature with with the knowledge we usually call knowledge laws of nature page 49 page 50 to 51 he says my position in evolution in epistemology goes further than that of all other epistemologists whether evolutionary quantian or non -evolutionary general adaptions proceed momentary adaptions they exist first they are a priority
[00:45:35] Red: uppers point in context is that even knowledge gained via the senses
[00:45:41] Red: requires pre -existing more general genetic knowledge first thus the senses are not a pure source of knowledge as was then assumed but require pre -existing theories that could be overridden or refuted making the more general genetic knowledge itself a conjecture that gets overridden by animal learning it’s doubtful popper would have ever even conceived that future critical rationalist would doubt that animal learning was a kind of growth of knowledge and that quote all the animals all knowledge was in its genes but he’s trying to argue that what he’s trying to argue is that animal learning is explicitly a growth of knowledge consistent with his evolutionary theory of knowledge via trial and error learning notice from these next quotes that is specifically what he is getting at quote page 72 all organisms are problem finders and problem solvers i would argue then that all living things are active meaning active learners they are always casting about in every direction like Beatles we fill our way towards things with every means at our disposal page 53 we are always know nothings and always trying to find our way with our hands in our feet or our ears in our eyes and with any sense organ organs we have which we use actively to make sure of the reality around us my theory of knowledge is thus quite revolutionary it overturns everything my predecessors have said up to now we humans are active we are constantly testing things out constantly working with the method of trial and error page 53 and continuing from there and that is the only method of knowledge growth we have we also the only method we can assume the earliest animals or plants to have had they move hither and thither primitive animals make trial movements and somehow try to optimize something they seek and they find primitive animals are already looking for better surroundings for better worlds and they are active in their quest for a better world like how do you possibly reconcile what this is being said with the idea that all of an animal’s knowledge is in its genes it’s just impossible that is not how popper saw it notice that so far popper is explicitly talking about animal learning being a growth of knowledge consistent with his epistemology and analogous to how humans learn but he then wants to show that as his epistemology predicts this is only possible due to genetic knowledge making it pop making it possible since all knowledge growth requires pre -existing theories that you then have to refute so quote he says in this quest the they the animals must as I said already have somehow adapted meaning genetically they must already have some general knowledge then comes mutations and further adaption via animal learning that is the empirical method of trial and error quote quoting popper to sum up from a biological point of view animals and human knowledge consists consist in often unconscious expectations or potential expectations page 53 but I would argue that our knowledge and this is this is interesting it’s probably wrong but this is interesting that he says this I would argue that our knowledge meaning human knowledge is 99 % or let us say 99.9 % biologically innate
[00:48:54] Red: note here that poppers arguing that both humans and animals are 99.9 % genetic knowledge but what is the rest of our knowledge then quoting popper the rest is a modification a revolutionary overturning of some previous knowledge meaning genetic knowledge just as that knowledge was itself once a revolutionary overturning of something that went before but in the end all knowledge meaning genetic or learning goes back to innate knowledge and to its modification page 54 thus in poppers disprove of empiricism or the idea that knowledge comes directly from the senses
[00:49:31] Red: that that’s his argument in a nutshell so instead popper shows that even learning via our senses is a form of his evolutionary epistemology just like Campbell made the same argument so page 115 now let’s now talk about popper’s reaction to Campbell’s evolutionary epistemology so I’m going to quote I already quoted this first part but I’m going to give some extensive quoting from poppers response to Campbell including where he says he disagrees with Campbell or where he might have a disagreement with Campbell so starting the first part of quote page 115 of evolutionary epistemology okay here’s popper now responding to Campbell professor Campbell’s remarkable contribution chapter 2 above shows the greatest agreement with my epistemology and what he cannot know an astonishing anticipation of some things which I had not yet published when he wrote his paper in addition it is a treatise of prodigious historical learning there is scarcely anything in the whole of modern epistemology compare with it certainly not in my own work his historical references are all highly relevant and they are a real treasure house and they often surprised me greatly for me the most striking thing about Campbell’s essay is the almost complete agreement down to even in minute detail between Campbell’s views and my own I shall try to develop one or two of these points a little further and she’ll then try to in the very rare and comparatively minor points turn to the very rare and comparatively minor points where there may be some difference of opinion so let’s actually go through what he talks about so thanks to Deutsches two sources hypothesis the idea that there are only two sources of knowledge biological evolution in genes and human ideas there are numerous crit rats today that strongly believe with disagree with Campbell’s more ubiquitous view of knowledge growth the two most most important disagreements are that Deutsche crit rats don’t believe AI machine learning the immune system or animal knowledge even animal memes should count as knowledge creation as we’ve seen Campbell’s actually got his ideas from popper and popper doubled down on this view after Campbell published his paper and how might crit rats explain or in my opinion explain away popper’s strong endorsement of Campbell’s theories now I would anticipate that they’d claim that popper agreed with some aspect of Campbell’s theory but disagreed with him on others and I suspect crit rats will be tempted to try to drive a semi truck through that small loop hole because popper did say minor disagreements unfortunately popper seems to have anticipated this problem and already closed the loophole popper specifically tells us what he disagrees with Campbell over which we’re going to cover in detail there are a number of points where popper feels Campbell doesn’t say as much as he’d like the popper claims they aren’t disagreeing so for example page 116 Campbell does not say very much about my theory of world three and it is here that I would like to make some remarks in amplification of his popper that immediately argues that animal learning is a form of knowledge growth for for that animal quote a point which follows almost directly from his critical realism is that I have described what is what I have described as the knowledge situation of animals and of men so far as knowledge is not genetically built into them oh animals and men have knowledge not genetically built into them okay animals and men can only gain knowledge if they have to drive if they have a driver instinct for exploration for finding out more about the world whoa popper just said it outright the knowledge situation for both animals and humans is that all their knowledge is that all their knowledge is not genetically built into them and instead they are given a driver instinct towards curiosity and exploration to gain additional knowledge not found in their genes popper contrasts this with the more with the more than popular idea the knowledge is gained passively through the senses quote rather than passive recipients of information impressed upon from the outside landmark ism and inductivism page 116 thus it isn’t by the way I just can’t help but say this notice that this idea that the animal is just simply learning through its senses and there’s no knowledge being gained through trial and error that he just sees that as landmark ism or inductivism okay so in essence the crit rat view is a form of landmark ism inductivism from popper’s viewpoint thus it is an almost immediate consequence of critical realism together with evolutionary an evolutionary approach to regard ourselves and our knowledge as continuous with the animals and with animal knowledge that’s page 116 the main task of the theory of human knowledge is to understand it as a as continuous with animal knowledge and to understand also its discontinuity if any from animal knowledge so we have some overlap between how an animal learns and how a human learns and some things that are different about us page 117 that is to say human learning and animal learning might may share some epistemological similarities rather than dismiss animal learning is not creating knowledge but we should also expect human learning to be different in some way or we couldn’t explain why humans lack while human learning lacks the limits of animal learning going back to quoting popper in in all this there is I believe complete agreement between Campbell and myself page 117
[00:55:14] Red: popper next praises Campbell’s idea of blindness and points out that that he popper had previously made somewhat vague similar arguments quote I am thinking of what Campbell puts calls the blindness of the trial of trials and in a trial and error method I have sometimes compared the human situation to the quest of the quest for new knowledge with the proverbial proverbial situation of a blind man who searches in a dark room for a black hat which is perhaps not there this is not saying much but it indicates that the searcher is at least acts as if he has a problem I have often added that the trial movements of the searcher will not be completely random there are various reasons for this both positive and negative the positive ones are in the main that the searcher has a problem to solve and that this means he has some knowledge however fuzzy previously acquired by essentially the same trial and error method this knowledge serves as a guide and eliminates complete randomness page 117 so complete there’s complete agreement here between popper and Campbell on the idea of blind variations and also the idea of a ubiquitous hierarchy of by of variation selection algorithms in nature that build on each other as opposed to being only two quote pop popper quote now Campbell also explains why he does not call them random blind variations random and calls them blind an excellent term and insists on the fact that so far as they are trials in the trial and error movement that is so far as they are forays into the unknown they are blind page 117 earmark that last quote for later when I talk about my disagreement with Ella Hopner over how to interpret Campbell the key thing to note here is that to that Campbell’s blindness is defined in terms of being nothing more than a foray into the unknown Ella had a disagreement with me over that where I wanted to define blindness that way she wanted to define it some other way popper continues well to the to the degree that past knowledge enters their blindness is only relative it begins where the past knowledge ends okay but where do popper and Campbell disagree because so far it’s all agreement so page 119 popper says I come now to two points where I may possibly slightly deviate from Campbell the first of the two disagreements is underwhelming in my opinion popper claims that Campbell quoted him popper and then said quote this insight is the earliest and most frequently noted aspect of an evolutionary epistemology popper then says however I may perhaps say that in the long passage quoted by Campbell from me there are I hope several insights and that the insight denoted by Campbell as this insight is not the one I wish to bring out as important that is their first disagreement he’s not clam claiming Campbell was wrong per se only that the passage quoted him quoted had multiple insights and Campbell found an insight that wasn’t the main one popper intended that’s not much of a disagreement right now here’s the second disagreement that he raises quote I come now to my last comment it is I think an important one and it is related to the difference between man and animal and especially between human rationality and human science and animal knowledge this is all page 120 Campbell speaks very interestingly of the language of bees and he also mentions my stress on criticism but he nowhere seems to allude to my view that human descriptive language differs from all animal language in being also argumentative and that it is human argument argumentative language which makes criticism possible and with it science and that’s it that’s their biggest disagreement he then continuing a quote from popper there is a world of difference between holding a belief or expecting something and using human language to say so the difference is that only if spoken out and thus objectivized does a belief become criticized before it is formulated in language I may be one with my belief the belief is part of my acting part of my behavior if formulated it may be criticized and found to be erroneous in which case I may be able to discard it notice this is an example contrary to what we’ve talked about in past podcasts where popper is talking about humans having beliefs and doesn’t seem to have a problem with it okay in this context
[00:59:35] Red: at least I do not sense any difference in opinion in what Campbell says concerning truth in an instrumentalism but there seems to be a slight difference in emphasis in in my stress on the idea of truth on the argumentative function of the human language and on criticism in brief in my predilection for world three it is very likely there is no difference here Campbell’s beautiful essay covers a great many things he may have been reluctant to say more and that’s it popper really doesn’t offer even a single disagreement with Campbell on anything this should close the loophole crit crit rats are likely to want to use to claim Campbell’s evolutionary epistemology isn’t really popper’s theory of evolutionary knowledge okay that is the that’s the end go ahead beater okay let me just let me just
[01:00:23] Blue: react to this a little bit you’ve put a lot out there I appreciate that about you and said some things here and elsewhere on this that I think I agree with and also some things I’m still wrestling with so it seems so let me just make sure I have this this right so it seems like both popper and the the popper Campbell view and Deutsch, Deutsch’s two sources view would fundamentally agree that conjecture and refutation can be thought of as a meta meta theory of knowledge would you say that’s true you
[01:01:06] Red: know I think I think for Campbell and popper the answers are really obvious yes and I can see a case for for dutch’s but I want you to think about this for a second okay the the problem is is that if you were to ask dutch the way you just worded it he’d probably say yes and he’d say yes that’s what I’m yeah okay I think okay but stop and think about his his belief that animal learning doesn’t create knowledge okay that is a trial and error process okay that he’s saying doesn’t create knowledge okay or think about animal memes animal memes is literally cultural knowledge that an animal has that is adaptive behavior that then survives because it helps the animals adapt okay
[01:01:50] Blue: okay no that makes sense
[01:01:53] Red: that is not that that has to either be qualified either you have to either under dutch’s theory call it not knowledge and then you don’t ask what it is or you have to call it somehow that animal meme is knowledge in its genes even though that’s that’s a like a blatant contradiction okay yeah
[01:02:11] Blue: and
[01:02:11] Red: when you realize that dutch is saying that then there is a problem with trying to say that dutch is saying that the overall meta theory of knowledge is this theory of trial and error he’s clearly carving out a an exception case that he doesn’t really talk about he doesn’t really get into that he doesn’t call knowledge but it behaves exactly like knowledge and that he doesn’t believe has any it’s unclear if he believes it has any relationship to the meta theory of knowledge or not right that it’s related to um and this is why I’ve objected to it so strongly right like I I
[01:02:47] Red: understand exactly why the two sources biological evolution and human ideas are so much more impressive examples of knowledge growth than animal learning or any of the others that we’ve talked about and I’ve never denied that like like I’ve made a point of saying look the what’s going on here is they’re confusing two things they’re confusing how cool open -ended knowledge creation processes are compared to not open -ended knowledge creation processes so they’re trying to declare those the only two sources because those are the two impressive sources okay and the ones that are the center of everything else like every other source is going to connect back into biological evolution in some way right because there’s a hierarchy you can always trace it back to biological evolution in some way okay and in doing that they are creating like think about think about how this is I’m going to get into this in future podcasts but just as a short preview let’s say I’m arguing with Ella Hopner or Dennis Hackathol about or Vaden for that matter about Campbell’s and Popper’s evolutionary theory of epistemology okay what they’re going to do all three of them we’re going to see is they’re going to carve out something that’s not part of Popper’s epistemology and say well machine learning it’s doing statistical inference machine learning it’s doing blah and they’re they’re going to Ella didn’t Ella actually called it induction by the way but all three of them are going to carve out something that’s an exception okay Popper did not do that Campbell did not do that okay modern crit rats do do that okay now what is this thing this other thing that exists that animals learn that adapt their behavior that allow their behavior well it’s not Popper’s epistemology under this theory okay it’s not evolutionary epistemology it’s it’s what well Ella as we’re going to see she’s going to call it induction she’s going to say well it’s a kind of induction Popper was was not quite right there is a kind of induction that exists well other crit rats aren’t going to go as far as Ella they’re going to say no no no it’s not induction induction’s wrong but it’s not Popper’s either it’s it’s something else and they’re never going to really say what it is okay I call this crypto inductivism it’s exactly the same of inductivism but it’s it’s just not under that name right and this is one of the main differences but because of the two sources hypothesis because it’s a compelling intuition pump that the crit rat community has deeply bought into okay they are forced into a sort of belief in induction that um even though they don’t necessarily call it that but they kind of have to go back to something that’s just almost the same as induction and you know what’s an animal doing it it observes things and then it changes its behavior and it generalizes its behavior it realizes that uh based on one observation that the food made it feel sick it knows not to eat that food again so based on an observation it generalizes okay well if that’s not being done through Popper’s epistemology then the only thing left is induction right like and that’s exactly what Campbell and Popper are trying to argue they’re trying to argue no that’s not induction either okay it’s actually Popper’s epistemology now is that correct you know what I don’t know like like that’s exactly like we have to first understand what Popper was trying to say which is there’s only one theory it’s evolutionary epistemology or what today we would call evolutionary epistemology since he didn’t like that term and that’s it you got one epistemology it’s the way everything works AI is going to work that way machine learning is going to work that way animals are going to work that way they’re they’re absolutely going for the one universal epistemology approach period of story modern crit rats don’t believe that they believe in something else and it’s a little unclear what they what they think this other thing is okay and what I’m trying to emphasize here is not that one side’s right and one side’s wrong that’s a different thing but that there is a strong disagreement here between Popper and the crit rat community today
[01:06:58] Blue: I guess that how I’m thinking of it what distills it for me is that Popper’s ideas don’t seem quite maybe it’s a matter of emphasis but it doesn’t seem quite realistic about humans how special we are I you know as Deutsch says there’s more differences between individual humans than there are species of animals which I kind of agree with
[01:07:33] Red: yeah
[01:07:34] Blue: but the other side of the coin is that I guess Deutsch’s ideas don’t seem to do animals learning justice for many of the reasons you’ve said I mean like this whole idea that the extreme thing is the animals don’t have feelings
[01:07:54] Red: yes
[01:07:54] Blue: which doesn’t quite it’s just it doesn’t ring true for me and and most people I think I mean if you said animals don’t have human feelings I don’t have feelings that are based on explanations yeah but there’s something going on that seems a lot like feelings and seems a lot like learning and creating something like knowledge not open -ended knowledge but it just doesn’t I think you’ve gotten at some some of the issues with this this how how severely this theory is stated I guess
[01:08:34] Red: yes so I of course what I’m point what I’m trying to suggest is that there’s a way to actually have it all right and I don’t entirely know what it is like I’m still working it out myself I really
[01:08:47] Blue: bring in the open -ended thing
[01:08:49] Red: yeah so I’m bringing the open to try to say okay let’s admit that certain kinds of knowledge creation are more impressive right so yes you’ve got kind of this narrow sort of knowledge creation that maybe an animal does and it’s just not as impressive and let’s admit that let’s let’s admit that Deutsch is onto something there okay and that Popper kind of I wouldn’t say he missed that because you can definitely find places where he treats humans is very special but we just barely read some quotes about how humans are special because of their argumentative function like that’s an example of he this of course what he means is that we’re using explanations there right so Deutsch is getting this that aspect from Popper but Popper didn’t go as far as Deutsch does and Popper should have gone further he should have more strongly stressed the difference between animal learning and human learning
[01:09:39] Red: I’m trying to find a way to have it all and it’s not an easy thing to do this idea the idea of Popper that there is one epistemology okay that that there is one way in which all truth about the world is known that may not be true like Popper could be wrong and we shouldn’t assume even though it’s it’s very appealing point of view we shouldn’t assume it’s correct okay but assuming it is correct when it’s the only one you know about that is the right point of view okay so it makes if you know that you can learn through trial and error and that’s a way to do it and you don’t know of an alternative way to do it then your first assumption should be trial and error is the universal way in which learning takes place okay
[01:10:29] Blue: okay
[01:10:29] Red: and until somebody can actually say not just it’s done through induction and then we don’t describe what induction is but somebody says here is what induction is as a competing theory to Popper’s theory and here’s a working example of induction that is not trial and error until somebody can actually put up or shut up like that then actually this this idea that there’s a single epistemology that’s exactly the right universal stance you should be taking because otherwise your theory is not falsifiable okay if you simply throw on there there’s one way to learn you know one form of epistemology and that’s Popper’s and then there’s this crypto inductivist thing that that animals do right then that’s animal learning if you’re doing that right like that at this point you since you haven’t defined what it is there’s no way to prove you wrong right it’s a completely ad hoc attachment to the overall epistemology that that there’s just nothing we can do with okay that’s the problem with that approach so
[01:11:32] Blue: take home point we shouldn’t necessarily rule out that something like a xenomorph from aliens could exist that gains knowledge through harvesting DNA or or the or the Borgs from Star Trek who gain knowledge by essentially colonizing other human minds these things these things could happen it’s you know I I’m wearing the multiverse let
[01:12:04] Red: me let me say let me go ahead and give my opinion like okay I’m gonna do an episode where I’m gonna actually give you an algorithm that is an adductive achievement it’s it’s just linear regression right like if you you do a linear regression and you don’t use backprop and instead you use the normal equations you can get an adductive achievement and you can form a model one that reads digits the MNIST data I
[01:12:32] Blue: never thought about that so these are these are species from science fiction obviously yeah based on induction in a sense
[01:12:40] Red: maybe you could argue that they still evolved and that they have a learning algorithm like like since they’ve never really explained how it is but yeah I guess you could argue that they’re inductive okay the problem with induction is that it’s very vague and so it’s it’s never clear what counts as induction and what doesn’t which is why it doesn’t make a good competitor to popper’s theory right
[01:13:01] Blue: well then you also would have to get into some of the stuff gone from Prometheus where yeah indicates that the they were actually the xenomorphs created by the engineers but then the xenomorphs end up I think they were kind of a bio bio weapon but then they end up actually killing the engineers so I don’t know I’m trying to I’m trying to wrap up to a concluding statement Bruce but okay let me
[01:13:28] Red: make my concluding statement
[01:13:29] Blue: okay because
[01:13:30] Red: I can point to an example like linear regression that really does seem to violate Campbell’s evolution and as we’ve seen popper’s evolutionary epistemology there’s something wrong with it there is it’s not clear if it’s a big problem or a small problem like and for a long time I had this problem that I didn’t know how to resolve it this go back to episode 26 and I basically admit I have no idea how to resolve this right like this is long ago now years ago and I’m still working on it and I’ve talked to so many people at this point Ella and Dennis and Vaden who all had very similar answers to me which I found very unsatisfying like truly unsatisfying answers the thing is I can’t really say they’re wrong like it could be that they’re on the right track and popper was wrong like I see no particular reason why that couldn’t be true but I definitely find it very appealing like what popper’s trying to do is he’s trying to make everything be a single epistemology and that’s a super appealing view if nothing else if there were two epistemologies then the correct single epistemology would be the combination of the two like if you if you if poppers was the right epistemology one of the things that Vaden we’re going to see does is he tries to declare this other thing not epistemology okay but like if you’re doing a machine learning model and you have one that overfits or you have one that that works and it actually makes good predictions both using the exact same method okay inductive method okay why does one work and one doesn’t now clearly the answer is that it’s got something to do with the fact that it’s captured something true about the world so machine learning models cannot be dismissed as not related to epistemology at all they are somehow related to epistemology because they connect into they build models that say something true about the world okay no it’s not explanatory knowledge no it’s not the same as the scientific epistemology I don’t deny any of that okay but the idea that you can just declare it not an epistemology is just not true okay based on that what you really are getting is a declaration of two epistemologies out of the critrack community they don’t realize that’s what they’re doing they would not put it in those terms but I think starkly speaking that’s the correct terms under which to understand what the critrack community is arguing so you’ve got these two epistemologies poppers and this other thing that works for machine learning and AI and animal learning and things like that okay when you look at it in that way you really can’t say poppers epistemology is correct because it’s missing this other significant way in which which things can learn about the world and can adapt themselves to their environment okay once you realize that this is really just the same as saying poppers epistemology is wrong now that might be true poppers epistemology might be wrong but we need to get on board with the idea that that’s really what the critrack community is claiming about in here okay they’re claiming poppers epistemology is wrong they’re very shy about that today they want to break it off not call it epistemology not call it poppers epistemology they want to call it something else this is a problem if they were more bold with it and they were just say yeah actually we think popper got this wrong I would totally find that acceptable but I’ve got a problem with just kind of shyly implying it but not saying it okay be be bold about it this is where you’re going now is it true that there that popper is wrong probably not I eventually did discover Deborah Mayo’s theory which offers a way to fold these counter examples back into poppers epistemology through the concept of severe learning uh severe testing and I don’t know if it’s a complete answer I’m not even sure I fully understand her theory yet so I’m hesitant to endorse it too strongly until I’m I’m trying to like understand Bayesian reasoning go back and read her book again see if I can get it this time but let me just say that the idea that actually all these examples machine learning artificial intelligence even the linear regression example with the normal equations that she does offer a possible way to fold it back into all being one epistemology and that’s that’s actually a very appealing thing about her theories okay again is it right I you know I don’t know we’re not here to back a single theory and be memoids we want to get it to whatever the truth is and that means we have to really stop and think about this carefully and have to think about the criticisms very carefully and what the alternative theories that are being offered are okay now if you were to push me and say Bruce what do you think I think probably Debra Mayo is right and probably the crit rack community is wrong I think the crit rack community is on a wrong path and the Debra Mayo has the correct path and that really we should be looking at Debra Mayo’s theory and we should jettison the two sources hypothesis that led to this bad path that we’re currently on but you know what I don’t know that like like it may be that I’m going to change my mind tomorrow and I’m going to say you know what the crit rack community has found a legitimate problem here and we really need to start taking it more seriously we need to maybe we need to look to the Bayseans to help us understand epistemology better and come with a synthesis of Bayesianism with critical rational I don’t anticipate this happening but who knows maybe that’s exactly where we need to go if that’s the truth let’s go there like let’s not just decide we’re memoids defending a single theory and this is really why I’m trying to make such a big emphasis here on what Popper actually said and compare it with what Campbell actually said and trying to show that there is zero way in which you can reconcile it with Deutch’s two sources hypothesis it’s impossible to reconcile the two and this is why I’m making a big deal deal about this I want people to sharply understand the problem that is being proposed here and not just gloss over it I want them to take it seriously as a problem okay that’s my that’s my view at this point tentative view I don’t know
[01:19:30] Red: what else to say at this point I definitely feel like I only kind of sort of kind of sort of understand Dipper Mayo’s theory and so I I’m hesitant to proclaim myself as understanding it yet I will try to understand it I will do a future podcast on it and I will try to really make it more clear in my mind and everybody else’s mind but at this point I’m not there
[01:19:51] Blue: okay well Bruce I’ve appreciated your thoughts here you are on your own unique path in life and I I like that this is officially not a short episode I don’t know if that’s possible but we just it just might not be in the cards for us but thank you Bruce I hope you enjoy this day
[01:20:13] Red: thank you
[01:20:20] Blue: hello again if you’ve made it this far please consider giving us a nice rating on whatever platform you use or even making a financial contribution through the link provided in the show notes as you probably know we are a podcast loosely tied together by the Popper Deutsch theory of knowledge we believe David Deutsch’s four strands tie everything together so we discuss science knowledge computation politics art and especially the search for artificial general intelligence also please consider connecting with Bruce on x at B Nielsen 01 also please consider joining the Facebook group the mini worlds of David Deutsch where Bruce and I first started connecting 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.