Episode 21: Evolution Outside the Genome

• Links to this episode: Spotify / Apple Podcasts

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

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

Transcript

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

[00:00:14]  Cameo: It is going great. Bruce, how are you?

[00:00:17]  Bruce: I'm doing fairly well. We were just discussing that I'm a little bit of pain, but I sometimes have to go through painful moments because I've got a bit of arthritis. And that's actually relevant to today's discussion. I'm kind of excited to talk about it. The thing that kind of led up to this was I had come across somebody I'm connected to on Twitter, put this up on their Twitter feed, and I looked at it. It's a TED Talk. TED Talk came out a few days ago. It's by Michael Levin, and it's called The Electrical Blueprints That Orchestrate Life. And I got really excited when I watched it. I went back and I did some research afterwards, and it looks like a lot of these ideas are things he's been putting out there for a couple of years. So it isn't, the TED Talk was brand new, and I think it was going to a larger audience for the first time. But it's clear he's actually been kind of trying to advertise his research for a couple of years.

[00:01:18]  Bruce: And so perhaps this isn't like some sort of immediate breakthrough, but it's just starting to get out into the media and people are starting to find out about it. And it's kind of an exciting, interesting breakthrough that he's come up with. So I sent it over to Cameo. And then Cameo, what was your reaction?

[00:01:37]  Cameo: Well, it's very, very fascinating. I watched it and then read the transcript and then talked about it with my husband. And it's very interesting because it does talk about the potential for being able to leverage these electrical blueprints that we're gonna talk about here to really change what happens for us at the cellular level. But I also think it has some really interesting implications across maybe some intelligent design concepts. Like where do these blueprints come from? How are they part of our cellular makeup? And what does it mean for kind of our place within the universe?

[00:02:18]  Bruce: Yes. Okay, so let's talk about that because just by coincidence, I had actually been reading a paper and the paper is called, Was the Watchmaker Blind or Was She One-Eyed? And it's by Raymond Noble and Denis Noble. I assume they're related in some way. It's a really interesting paper. I actually highly recommend that people go read this, particularly if they're fans of neo-Darwinism because the paper, I want to be careful how I say this. In a sense, it refutes neo-Darwinism. Really, it only refutes a certain original formulation of neo-Darwinism, probably not its current formulation. But I think a lot of people would be very interested in what it has to say about classical neo-Darwinism. So I actually, when I was reading this paper, I got really excited because it tied into another paper I had been reading by Donald Campbell, who was talking about the hierarchy of evolutionary processes.

[00:03:18]  Bruce: And it was a direct relationship to that article, which is a much older article back from the era of Karl Popper himself. Suddenly, I see Michael Levin's work, and I'm realizing this is like a really solid example of what they're talking about, more solid than the ones that were in the papers themselves. And so I want to kind of tie all three of these ideas together, of these papers together, and talk about how they relate. So to do that, let me first explain to you was the watchmaker blind or was she one-eyed, what it's about? And then I'll explain how it kind of ties in. They claim in the paper that neo-Darwinian theory usually assumes all knowledge is held either in the genome or it's in the mind or nervous system. So neo-Darwinians would admit there is a source of knowledge outside of the genome that's in human minds, or for that matter, in animal minds.

[00:04:17]  Bruce: You know, you can classically condition a dog, it can learn to beg for a treat. That's not something that anybody believes is knowledge in the genome. Classically, they've tended to assume it's either one of the two. The knowledge either resides in the genome or it resides in the mind, and that's it. Modernly, maybe not all of them believe that. These theories kind of mutate over time, they quote evolve over time. You could easily make the argument that modern neo-Darwinian theory doesn't make that claim. However, the authors of the article, they give examples of where that is the assumption. So clearly, at least in the past, neo-Darwinian theory has had this assumption that there's really only two sources of information, the genes or the mind. And it's been kind of at least tacitly assumed from the way they speak and the way they talk, sometimes not tacitly assumed, sometimes directly stated.

[00:05:14]  Bruce: So for example, they quote Richard Dawkins. He says, in real life, the criteria for selection is always short-term, either simple survival or more generally reproductive success. So Dawkins is making the claim there's no theology [apparently meaning teleology], no purposefulness in evolution. And this is something you will hear over and over again from neo-Darwinians, that evolution has no purposefulness. Now, again, they know there's at least one exception. Like if humans go out and breed dogs, there's clearly purposefulness in evolution at that point. And they wouldn't deny that. But they do tend to assume that that's the one exception. And what this paper is saying is, no, that's actually not the only exception, that there's actually many, many, many exceptions at many different levels of the evolutionary hierarchy, which we'll talk about what that is in just a second.

[00:06:07]  Bruce: They go on to say, Classical neo-Darwinism was formulated by August Weismann and others in the late 19th century to expunge the inheritance of acquired characteristics from Darwin's theory. Blind variation followed by natural selection was claimed to be entirely sufficient. Okay, then they admit that it's been redefined since then, so that's not necessarily true of everyone who would call themselves a neo-Darwinian today. But then they go, refining a term does not, however, change the fact that the original theory using that term is no longer the complete story. Our position can therefore, to some degree, be seen to return to Darwin's multi-mechanism viewpoint though with vastly extended empirical evidence. So they're setting themselves up at least as saying, okay, neo-Darwinian theory, at least classically has thought of things as genes are sort of a gene-centric view of the world.

[00:07:01]  Bruce: And I think you'll see just as a layman yourself, that you certainly kind of had in your mind this idea that there's all this knowledge in the genes and you really haven't ever thought too much about where knowledge might else reside besides the genes or the mind. Precisely because neo-Darwinian theory has kind of gotten out into the culture and it has sort of been imbibed in that way, even though it's not entirely accurate. Does that make sense so far? Are you following me?

[00:07:30]  Cameo: Yeah. And I think you're right. Like one of the things, the way they teach things like DNA in school is here's all of the information that's going to make up the person that you are. It's this combination of these building blocks at the cellular level.

[00:07:47]  Bruce: Right. And that's not entirely untrue, right? It's, you have to kind of just understand the nuance that's being suggested here. They go on to say in the paper, genomes and phenotypes are far from being equivalent. The mismatch works both ways. The same genome can be used to generate many different phenotypes and the same phenotype can evolve through many different genome variations to such an extent that the sequence variations may even be unpredictable. Okay, so they're trying to get the idea across that evolution actually works on the phenotype. When I say phenotype, that's the actual body or organism that gets produced. That's what evolution actually works on. It doesn't directly impact the genes, only indirectly insofar as the genes helped create the phenotype. But once you realize that the phenotype actually is impacted by many different things, not just the genes, then you start to understand that evolution is gonna be impacting all the inputs to the phenotype, not just the genes.

[00:08:48]  Cameo: Right.

[00:08:48]  Bruce: Okay, so now they go on and they define something called natural purposiveness. They're trying to make this not sound too much like an intelligent design or something like that, because it isn't. It's still a very natural process that they're talking about. But they're saying, the gene-centric view of evolutions are incorrect. Evolution is a high-level forming process, not simply a matter of genome informatics. It is rather the development of the conceptual interpretation of the process of evolution that differs from neo-Darwinism in implementing the principle that there is no privileged level of causation. We believe this is the novel concept, conceptual advance that they're making. It differs radically from views of evolution that privilege the role of DNA sequences in the intergenerational transmission of inheritance. Now, when I say this, you can probably think of exceptions because we're starting to hear about exceptions.

[00:09:45]  Bruce: Everybody kind of knows about epigenetics now. And epigenetics is a non-genetic form of knowledge that gets transmitted. And they mentioned this in their article, and they give another example. And this first example is a really good example, but it's a well-known one though. So the immune system, you think about how the immune system works. The immune system doesn't come with knowledge of every single possible pathogen and how to stop it. What it actually does is it comes with a learning algorithm.

[00:10:19]  Cameo: Right, absolutely.

[00:10:21]  Bruce: What the learning algorithm says to do is it says, okay, so remember your DNA, it's in every single cell in your body. And so the cells that are gonna produce the antibodies, they use DNA. The algorithm that they have says, oh, look, we're being invaded. Let's hyper mutate our DNA, but only in this one area that affects the antibodies, nothing else because that would break us. And so it starts going through hyper mutation of its own DNA. And it then produces these antibodies that have random components to them that then go out and the ones that are successful are the ones that then start to get replicated. And then the ones that aren't successful, they just kind of die off. Based on that, you wind up with an immune system, which is able to not have to have knowledge in it from the beginning about how to fight every pathogen, but instead can generate that knowledge for how to fight the pathogen in real time without having to give birth to another generation of puppies or something.

[00:11:31]  Bruce: And so again, this is something that's well known, right? And so it's interesting that this hasn't really, people haven't really understood. This is an example of how the genetic view of evolution isn't really quite right. Even though we knew it wasn't quite right, the full implications of it haven't been thought through all the way.

[00:11:54]  Unidentified speaker: [Uncertain: “If that makes sense.”]

[00:11:55]  Bruce: This is then an example of how, in this case, we're still using DNA. You don't have to use DNA. And that's what we're gonna get to with Levin's information. In this case, the DNA has a learning algorithm that tells itself how to modify itself. The other example they give in the paper is jumping genes. Apparently, it was discovered before they had actually like sequenced the genome, they had discovered that, quote, working with Indian corn, this researcher showed that in response to stress, genetic material can move around between different chromosomes. They had discovered that the genes in chromosomes could hop from chromosome to chromosome if the genetic material was under stress to try to protect itself. At least at the time, they didn't have any idea how it was doing it. How was it detecting that there was a stress, that there was a problem?

[00:12:49]  Bruce: How was it then knowing to try to take that gene and go relocate it somewhere else that's not under stress? And yet this is again a clear example of how the strict genomic informatics view of evolution isn't correct, right? That somehow the genes had some sort of learning algorithm that they had instantiated that allowed it to move their genes so they didn't have to stay static to try to protect themselves. Okay, and again, we're still dealing with genes here, but in a way that's at a different level of hierarchy. So now what's this idea of a hierarchy? Well, this is actually an idea that dates back to probably before Donald Campbell, but he's the one that I first discovered the idea from and he explains it quite well. So Donald Campbell was a huge fan of Karl Popper, who obviously inspired a lot of this podcast. And Donald Campbell came to this realization, apparently at the same time as Popper, but he published first, and Popper was still in the middle of writing his stuff according to Popper.

[00:13:52]  Bruce: And he said, oh, hey, it seems to me, so Karl Popper's theory is this idea that evolution applies at the level of human ideas and at the level of science. And so he's taken this idea of evolution, which at the time was understood as survival of the fittest. That's probably not the best way to describe. I've taken some criticism over calling it that because that brings to mind kind of a false view of evolution that the strongest animal survives or something like that. It's really the fittest at doing replicating itself.

[00:14:27]  Cameo: Well, and error correcting. Yeah.

[00:14:31]  Bruce: Well, actually error correcting is relative. Are we talking about the gene? Are we talking about the organism? And those aren't the same.

[00:14:38]  Cameo: I know, but we're talking about the organism itself.

[00:14:41]  Bruce: Yeah. Campbell says, you know what? It's not just that evolution applies at the level of biology or human ideas, but it applies all over the place. There's actually this, it's a generally applied principle that happens in a hierarchy. And so first he had this idea that it happens in a hierarchy. Okay, so he goes, human knowledge processes, when examined in continuity with the evolutionary sequence, turn out to involve numerous mechanisms at various levels, hierarchically related and with some form of selective retention process at each level. He gives examples of this for human knowledge. He gives examples of how we might think about how our actions might impact others and then edit our actions before we actually have to go try our action out and then find out how people react badly to us. Okay, and that's like a different level of evolution than the other levels we've talked about.

[00:15:44]  Bruce: But it also might be, it might also be that, the very fact that you see something, you don't even have to think about it, but you immediately decide to do something different. That's itself, the seeing substitutes for you having to go move or act because you're able to see. So he argues that's a different levels from the human ideas level. And yet it's still a level of evolution that's taking place.

[00:16:10]  Unidentified speaker: Interesting.

[00:16:11]  Bruce: So then after pointing out that Popper's theory applies or evolutionary theory, I should say, applies at all these different levels, he points out that it's also outside of human knowledge. He says turning to higher levels, the model can be applied to such dramatically teleological achievements. This is an important word here as embryological growth and wound healing. He goes on to explain that the way a leg grows back on a salamander is not something that's determined by the genes directly, but there's kind of an algorithm that represents what environment it thinks the salamander is going to be in and then detects when to stop growing that leg.

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

[00:16:55]  Bruce: Now you can see how this directly relates to Michael Levin's stuff. Yeah. So Campbell talks about this. And then he says, the different levels of hierarchy in evolution, they relate to each other. So if I have, let's say that I produce some knowledge at one level in the hierarchy, that knowledge can then be used as a heuristic at the next level down in the hierarchy, so that it doesn't have to search so far to be able to find variance anymore. Okay.

[00:17:24]  Cameo: So that's intriguing.

[00:17:26]  Bruce: Now, what that means is, is that evolution is teleological. It means it's purposeful. Not ultimately, now Campbell points this out, every single thing that in evolution has purpose itself at some point, otherwise it would be an infinite regress, has to have just been blind variation at some point, right? So, but once you've evolved that learning algorithm and the new level of the hierarchy, it can be purposeful from that point forward. Okay. Now this is actually the point Raymond Noble and Denis Noble are trying to make. Okay. They're trying to say, there's this natural purpose in evolution because there is no privileged level of causation in evolution. So once you've evolved a learning algorithm, it then can evolve according to a purpose. And what they're saying is, is that this is far more important to how evolution actually works than what the classical neo-Darwinian version of evolution seems to imply.

[00:18:30]  Bruce: Because evolution itself, this is quoting them, evolution itself evolves because you're actually evolving purposes that then in turn evolve from there. Which think of it as like targeting in on what you want to evolve. And so you've now constrained or narrowed what you have to select between.

[00:18:52]  Cameo: Sure. It's a pretty intriguing concept. Which will ultimately make so much sense that how could it be any other way because-

[00:19:07]  Bruce: I agree.

[00:19:08]  Cameo: Because how could you have a system designed to constantly adapt if the system itself couldn't adapt itself? Right.

[00:19:17]  Bruce: Okay. So now how does this relate to Michael Levin's stuff? So let's maybe let's just describe what is in the TED Talk. I highly, we'll put a link in the show notes. I highly encourage people to just go watch it and see for themselves. And this is the typical warning. We're not experts. We're just laymen that are interested. We're doing our best to describe what we saw and we're doing our best to understand it and we're likely wrong on all sorts of things. And you know what? We don't even apologize. It's just the way it is. You don't need to completely understand the subject to start talking about it and to start interacting with it. If you had to do that, no one could talk about anything, right?

[00:19:58]  Cameo: No, right. Life would be pretty boring then.

[00:20:00]  Bruce: Right. I want to get away from the idea that I have to be an expert to talk about this. I am not an expert and I'm going to talk about it and go look into it for yourself if you disagree with our interpretation. And quite frankly, we're probably mangling it for all I know, right? So actually, Cameo, why don't you describe what were the things that you found most interesting in the TED Talk that he brings up?

[00:20:28]  Cameo: Well, one of the things that's very intriguing is the way they were able to isolate the electrical signal or watch the electrical messaging happening between the cells or almost to the cells. I don't know if I could say, I believe the cells were talking to each other. It's more like he's describing a set of instructions that are being applied across the cellular level through the electrical signals.

[00:21:05]  Bruce: So interestingly, Dan Elton, who a friend of mine on Twitter, who's a PhD, that was the thing he found, if I'm understanding Dan correctly, the thing he found the most maybe misleading.

[00:21:16]  Cameo: So interesting.

[00:21:17]  Bruce: He felt that Levin was making it sound too much like the cells were intelligent and talking to each other. And he felt like that wasn't really the truth. And I don't want to put words into Dan's mouth either. Maybe I'm misunderstanding Dan. But the way you just said that I think is accurate, right? There's communication going on. That's probably been known for a while, that there's communication between cells. I think the thing that I didn't realize is that that communication is what determines what the phenotype is. So the genes may give the knowledge to the operating system. Think of it as like a software operating system. There's this electrical operating, bioelectrical operating system that runs between the cells, that itself contains knowledge. The genes may tell it how to instantiate that learning algorithm. So that the knowledge in that sense comes from the genes.

[00:22:12]  Bruce: But it is what actually forms the phenotype, not the genes directly. For example, by giving a signal to frogs or tadpoles, they could say, grow extra legs. And suddenly you'll have this tadpole with growing these extra legs, going into a frog with six legs or something like that. All without ever editing the genes.

[00:22:38]  Cameo: Right, no manipulation at the genetic level at all.

[00:22:41]  Bruce: Yeah. And then the one that was really cool was the flatworm. Now, flatworms don't sexually reproduce. They reproduce by splitting. But they split the flatworm in half and then told it to grow back with two heads. And so you have these flatworms that have a head on each side. And then after that, every time you split it, it always grows back with two heads until you reprogram the operating system again. Because that is now, its phenotype is now, the information on how to form its phenotype is now in the communication, the bioelectrical signals in the cells. And so even though the genes haven't been changed, you get this vastly different phenotype, exactly like Raymond Noble and Denis Noble were explaining, that the genes do not ultimately determine the phenotype by themselves.

[00:23:37]  Cameo: Right.

[00:23:38]  Bruce: Okay, so that was a beautiful example of what the article was talking about. And now, are you familiar with CRISPR? You've probably heard about CRISPR, right?

[00:23:47]  Cameo: Yeah, sure.

[00:23:48]  Bruce: So CRISPR is this very cool technology that has a lot of promise that they're gonna be able to put things into the body that change our DNA. So if you have a genetic disease, arthritis, a genetic disease, they can someday inject me with something that will say, okay, stop, that gene is, let's inject you with a different gene, it'll replace throughout my body and I'll end up being healed of the disease, right? That's very cool. The fact that they can do that at all is amazing. But what Levin is saying is, that's more complicated than it needs to be.

[00:24:27]  Cameo: Right.

[00:24:28]  Bruce: Is that actually we can figure out how to just simply tell the cells in your body, here use a different bioelectrical signal, here's the different software statements that you need, using an analogy here, but it's not that far off because it is software we're talking about. And my cells will automatically, without ever having to use CRISPR to change my genes, will automatically know, don't do arthritis, just stop.

[00:24:56]  Cameo: Just stop doing that, that's silly.

[00:24:59]  Unidentified speaker: Yeah.

[00:25:01]  Bruce: Presumably, it's caused by, there's different things that can cause arthritis. That's kind of a general term. But what I have in mind here is like some sort of autoimmune thing where your body's reacting too strongly when it doesn't need to. Or like allergies or something along those lines, where you have some sort of response. You can just tell your body, stop, don't do that anymore. There's not even a need to change the genes. The individual cells will simply perpetuate that information at the cellular level without having to have a gene change, which presumably is cheaper and easier and less scary. And it seems like it would be a fairly substantial breakthrough if we could actually use this like that.

[00:25:45]  Cameo: Right, if we could figure out how to, well, how to speak the, how to isolate and speak the language of that electrical signal and how to make sure that we're applying the right, essentially reprogramming, you know, I think about it almost as a firmware update.

[00:26:03]  Bruce: Yes, it's like a firmware update, yes.

[00:26:07]  Cameo: Yes, a body firmware update. I mean, conceptually, yes, it seems a little bit better to do a firmware update. Than to, you know, have to completely replace the hardware at the genetic level. It's a very fascinating, you know, and so much of when our bodies start to fall apart, it's cellular failure, failure, cancer, you know, autoimmune disease, allergies, those things, aging, you know, our cells stop doing the right things.

[00:26:37]  Bruce: It's interesting that he actually talks about that really briefly, like you almost miss it in the TED talk. But he basically, strongly implies, actually using these bioelectrical signals, we're going to learn to be able to just tell your cells to stop aging, you know, or we're going to be able to learn to just tell your cells to go heal, when to stop, when to start healing, when to stop healing, to avoid cancer, we're able to tell your cells, hey, stop that cancer. That's the wrong thing to do. That's very exciting, right? That there's lots of potential here that's huge, that comes from understanding what this biological language is that they're researching.

[00:27:18]  Cameo: So you say he's been talking about this concept maybe for a while. Like, all right, he's discovered this very cool thing. Is he struggling getting people to pay attention to it? Is there, you know, there's always a huge gap between concept and like, how do we actually turn this into a treatment I can go have, you know? Like I could go have a, you know, an oxygen treatment that's going to promise X, Y, Z thing. You know, how can I go have a cellular reboot treatment that will, you know, and how can I as an entrepreneur build, you know, a spa where somebody can go and have a cellular treatment that will apply some electrical signals to their cellular makeup?

[00:28:09]  Bruce: Yeah, good question. You know, I don't know the answer to that question, right? Because there is a world of difference between some professor, you know, somewhere doing research.

[00:28:20]  Unidentified speaker: With worms.

[00:28:22]  Bruce: Yeah, with worms. And then someone actually getting excited enough about the potential to make money with it, that they put up money to go try to turn this into an actual treatment. And as far as I can tell, no one's even attempting to do that at this point, right? What do I know? Maybe there's talks going on behind the scenes or something, right? I don't know that he would announce that if he was, had a deal in the works.

[00:28:48]  Cameo: No, but like there is, there's a really, really big gap on, you know, he might not even, well, he probably is not even the type of scientist who could figure that out, right? I mean, he's, he does a thing and he's got a thing like, that it's, it's a whole different ball of wax to figure out.

[00:29:10]  Bruce: It is.

[00:29:11]  Cameo: And, and talk about [unclear] money. Well, you know, you could spend, you could spend multiple people's careers trying to solve that problem.

[00:29:20]  Bruce: Yeah. So here's something I found interesting about this. So do you remember the part about the Picasso frogs? So in tadpoles, they'll take the tadpole and they'll force it to develop where the eyes are in the wrong spot or something like that, right?

[00:29:35]  Cameo: Right.

[00:29:36]  Bruce: This is what the Picasso tadpole.

[00:29:37]  Cameo: Painting, yeah.

[00:29:38]  Bruce: Then it develops into a frog and it still turns into a normal frog face. And the reason why, and he uses the word error correction, this is one of the things that caught my attention, is because this operating system we're talking about, this biological operating system, it does error correction. It says, oh, how do I, it doesn't have direct program instructions, take this eye and move it here when you turn a tadpole into a frog. Instead, it understands what the goal is. Here's another example of goal-oriented evolution, right? And [it] error corrects towards that goal. So this is a really strong indicator. This operating system, we probably could have predicted this, but we're still gonna call it out, that it is a learning algorithm, right? That it has in it various types of means by which it can evolve solutions to problems, purposefully.

[00:30:34]  Bruce: It says, I want the frog to have a face that looks like a normal frog's. And so I'm going to figure out how to error correct to get it into the right spot. So this is a really good example of both what Campbell and the Nobles are talking about, where it seems like, I'm making some assumptions here that may not be true, but it seems like we're talking about some form of evolution that's taking place outside the genome, but not in the human, not in the mind either.

[00:31:04]  Cameo: That was the implication I took away also.

[00:31:08]  Bruce: Anytime you say the word error correction, that seems like what you're talking about, although words are just words. He might have something else in mind and I'm misinterpreting. But yeah, I thought that was very interesting. So it'd be interesting if you could then give new program instructions and I don't know if this is possible or not, because I don't understand the technology, and you could give it something more sophisticated, how to not create cancer, but also not age. Because one of the theories that's out there is that cancer is the natural outcome of, aging is the way the body stops cancer, is basically the theory. The idea is that cells are programmed to die after a certain number of generations so that they don't become cancerous. It's not too hard to change the cells so that they don't age by having longer telomeres. But if you do that, then you'll die of cancer instead.

[00:32:04]  Bruce: Evolution, if this theory is true, I don't know if it's true or not, but if this theory is true, that would seem to indicate that evolution could not find, had to find a sweet spot between cancer and aging. So what we would really want is neither. We would want something that's more sophisticated than that, that allows us to-

[00:32:26]  Cameo: My initial reaction to that is, well, then why the hell do kids get cancer so much?

[00:32:33]  Bruce: Keep in mind that it's at the cell level.

[00:32:35]  Unidentified speaker: I know.

[00:32:36]  Bruce: But yes, it's-

[00:32:38]  Cameo: It just seems like a cop-out.

[00:32:41]  Bruce: All right, fair enough. But it should still be a soluble problem either way.

[00:32:47]  Cameo: Yes, I agree.

[00:32:48]  Bruce: And that might even be something where- How do I explain what I'm thinking here? And I'm grasping here, and I admit. But it seems like genes are more rigid. They've evolved, they're messy, they have to work in certain ways. This seemed almost more like it was easier to reprogram to give it new instructions that the genes never would have conceived of.

[00:33:12]  Cameo: Well, and you know, in a way, like if you look at- It is easier to have kind of a firmware infrastructure where you can have a piece of hardware that you can continually update with new information that will allow as you are learning things and adapting to things, you can go back within the confines of the original hardware and apply a new set of instructions. That is a more adaptable way to build things and a way to build things faster and kind of more componentized. Building in more components makes things be able to grow and change better. Maybe at the end of the day, it's the only way that something that you need at both a hardware component that is the genome and then a software component that is this electrical thing that has maybe always been there and we just didn't understand that it was happening.

[00:34:12]  Bruce: Anyhow, very interesting presentation. And at least it surprised me. Maybe other people have more knowledge of this and this isn't as big a breakthrough as it first appears to me. But it seemed like this was kind of a big deal. And I really hope that more people will discover about this work. And even if it is not quite right, error correction will take place, right? Well, parts will filter through and the parts that are maybe exaggerated won't. And we'll end up with something that is useful here.

[00:34:46]  Cameo: Yes, because I mean, well, very few things are so completely wrong like cold fusion, where as you dig into the science, maybe there wasn't anything there. You know, and so like you say, there's probably elements of correctness and elements of error correction that will happen. But that will start to and maybe none of it in our lifetimes, sadly, but maybe so. I don't know.

[00:35:15]  Bruce: Yeah, you know, that's the most frustrating thing to me is how long it takes to take these ideas. I mean, like, do you remember a while back, probably decades ago now, there was a news article that went around about these mice that they had engineered that could regrow limbs. Do you, at the time, I had the thought, oh, wow, I wonder, like, I wonder if, what if they turn that into something that's medical, right? Where you could like regrow a finger or something, right? Amazing if they could do that, right? But that was decades ago. And far as I know, no one's even working on that as a medical technology.

[00:35:56]  Cameo: Right, to be able to regrow things.

[00:35:58]  Bruce: Right. And yet there's no particular reason why you shouldn't be able to. Right. I mean, it's, we know that that's a perfectly normal thing to, for an organ, for some organisms to do. It's not, I mean, like, you know, the lizard—you cut off its tail and it grows a new tail.

[00:36:14]  Cameo: Right.

[00:36:15]  Bruce: And it's regrowing limbs isn't an impossibility. It just so happens that with the more complex and organisms like ourselves, it's not the most useful thing to do evolutionarily, but it would be really useful to us. Right. I mean, like if someone who lost a limb could regrow it, that would be amazing. That seems like that would be something we would want. And then it seems like that might even have, like they didn't know at the time if the mouse would live a longer life or not, but it seems like it would. Right. If you've got this super regenerating mouse, why wouldn't it live a longer life? It seems like a lot of these ideas you hear about them, they come out and then yes, they'll probably eventually turn into something. You know, they go into the knowledge in the background somewhere, but they're not going to, they don't happen fast enough to actually do me any good.

[00:37:09]  Cameo: So part of that, I think, is because of the way that theoretical things and research things are funded, where you have, you know, individual people who develop interest or specialty within a field and start to come up with concepts and want to pursue concepts. But, you know, academia is funded the way it's funded and it's publish or perish. And it's, there are probably ways that that money flows through, through research in general that really keep, in a way, probably breakthroughs from happening and collaboration a little bit. Maybe it's better than it used to be because you have people able to, with the, you know, with the rise of the Internet across the world, you have the ability for like people to find each other and start to kind of build off each other's knowledge. But then there's also still kind of protective competitiveness within research where people don't want to, I don't know that people are necessarily trying to solve the world's problems.

[00:38:14]  Cameo: You know, that's not really what we're doing yet. We're.

[00:38:17]  Bruce: Yeah.

[00:38:18]  Cameo: You know, and, and, and coming back to my statement about Levin, not probably knowing how to turn this into something that's a sellable. He's a research scientist who's good at, you know, tracking and watching worms and, you know, seeing what, what is a path from a kind of R and D standpoint from where he is at to something even conceptual that you could test on a mouse or test on a, you know, that's, that's a big path with a lot of science between it, you know? And if you could, if you could leverage, you know, if you were the head, the research scientist head at a big university that had a whole, a whole, you know, maybe, I don't know, you almost would need to develop a new branch of science with people. But I think it's just hard for us to build big, complicated things because the problem is big and complicated.

[00:39:19]  Bruce: So there's several things here that you just said that probably could be an episode in and of themselves. One of them is that we split up knowledge into buckets, right? We, there's the biology department and there's the physics department and there's the chemistry department, and there's no real, there's no, they're not set up, they're set up to be, to make it easy to teach and easy to administrate. And they're not really set up to make it easy to share knowledge between these different areas.

[00:39:55]  Unidentified speaker: Yeah.

[00:39:56]  Bruce: Knowledge doesn't fit into buckets in real life, right? These are completely arbitrary categories. And I'm not against this, right? I've seen some people get like upset over this and I'm not, the need to administrate things is a real need, right? Yeah. The need to make things so that we can make it easy to have certain types of classes is a real need. There really is a lack of sharing and of knowledge between the different buckets, between the different areas of knowledge. And that's where the breakthroughs come from, right? You've got people who are thinking outside the box because they come from a different area, but I shouldn't say it's not like all breakthroughs come that way, but that's a really common way for breakthroughs to happen. So clearly there's something to this idea that what you really want is you want to respect the fact that knowledge knows no boundaries.

[00:40:46]  Cameo: Right.

[00:40:47]  Bruce: And then we do that at several levels. It's not just that at a university, we split things up into departments, but we understand there to be a difference between a research scientist and business.

[00:41:00]  Cameo: Right.

[00:41:01]  Bruce: And there's all sorts of different levels where we have these kind of artificial barriers.

[00:41:08]  Cameo: Well, and to your point, we seem to think of problems as having a very... It's the Wright brothers inventing the airplane. All sorts of people who specialized in this concept of flight were working really, really hard at this problem. A lot of people were really focused on how can we break through and actually get ourselves into the air. And it took people who knew nothing about that and kind of had their own set of physics and their own understanding of the world to come in and not just break through, but figure out execution on a concept.

[00:41:46]  Bruce: So actually, the Wright brothers is a really interesting example. So we typically say the Wright brothers invented the airplane. And from a certain point of view, that is true. But there were airplanes before the Wright brothers. There were people before the Wright brothers that had created flying machines, which would rightly be called airplanes, that worked. What the Wright brothers really invented, if you really want to get technical, is they invented a control [for] an airplane that could be used while you were on the airplane.

[00:42:19]  Cameo: Oh, interesting.

[00:42:22]  Bruce: That was the big breakthrough that they came up with. Because they already knew how to make a flying machine.

[00:42:29]  Cameo: They didn't know how to control a flying machine.

[00:42:31]  Bruce: That's right. No one knew how to control a flying machine while you were in it.

[00:42:35]  Cameo: That's a pretty big breakthrough.

[00:42:36]  Bruce: It is. And of course, an airplane is not really an airplane if you can't be in it. So the statement, the Wright brothers invented the airplane, is still a true statement, but only if you understand the word airplane in a certain way.

[00:42:50]  Unidentified speaker: Interesting.

[00:42:51]  Bruce: There's a lot of things like that in life, where the issue here, and this is again probably a subject in of itself, the problem is that there are more concepts than there are words. And so just the way human natural language has to work, words have to point at multiple concepts. One word has to be reused for multiple related concepts. And because it wouldn't be possible to actually create one word per concept. In fact, according to Karl Popper's Theory of Knowledge, that should actually be impossible because every individual human has to evolve their own understanding of that word separate from each other, using each other for feedback.

[00:43:39]  Cameo: We should have a whole episode just on that.

[00:43:42]  Bruce: We should. Actually, there's a scientist who did considerable research on this, Douglas Hofstadter, who I love his work. I know a lot of my AGI friends don't like his stuff. And maybe his stuff is just interesting on its own, regardless whether it's a good path for AGI or not. He documents a ton of this, and it was all part of his research. He was trying to research AGI, and he ended up documenting a lot of stuff about language instead because he thought that that might be a hint to how the human brain actually did stuff. And so he ended up documenting how the human mind abstracts things. So how do we get from the point that a desktop on a wooden desk is the same thing in some sense as a desktop on a computer desk inside of an operating system? And yet we somehow do it. We somehow make some sort of connection that there's enough similarities between the desktop of a wooden desk that when Windows comes along or Apple comes along or Macintosh comes along, that we can call what's on their screen a desktop and you suddenly have knowledge about how it works.

[00:44:58]  Bruce: And it's imperfect knowledge because it's not quite the same. And so he documents this. He documents how we abstract things. One of the reasons why this seems to be just necessary is just the economy of words, right? Words absolutely will take on lots and lots of different meanings. And what we actually do in the human mind is we determine by context what that word means in that context. And we don't even realize we're doing it, right? So let me give you an example that he uses, which is a really clever example. Let's take the word grow. What does the word grow mean?

[00:45:39]  Cameo: It's almost not an answerable question, Bruce, because if you look it up in the dictionary, it has 72 entries for what the word means.

[00:45:48]  Bruce: It's very common for people to say, it means to make something larger.

[00:45:52]  Unidentified speaker: Uh-huh.

[00:45:53]  Bruce: And that's actually a good answer. So then he would tell people, Hofstadter, he would tell people, he would say, actually, it sometimes just means a change of size, whether larger or smaller. People will challenge him. They'll actually say, that is not true. In the English language, the word grow means to, to change larger in size. That is what the word grow means. So he'll say, okay, in Alice in Wonderland, she takes the whatever, and she grows smaller and smaller and smaller. The fact is, is that English-using people use that use of the word grow all the time to mean just a change in size.

[00:46:30]  Cameo: So we have a favorite family game that is a word game that we play where you're supposed to think of a word, and the ideal word has just vast amount of meetings—I mean, meanings, not meetings. Because some words, grow is a pretty good example, has lots of different ways that we use it in the language, that we have an emotional meaning for grow. It means to advance or to...

[00:47:06]  Bruce: You point that out. Yes.

[00:47:07]  Cameo: Yes, right. And you could, if you had a group of people who were all trying competitively to think of yet another meaning of a way that we use grow to mean a variety of different things, you would find almost a constantly expanding set of explanation for that particular word.

[00:47:27]  Bruce: Yes. Okay. And that's his point is that actually the meaning of the word is in constant evolution based on how people actually use it. The way I've tried to explain this to people is words don't actually have definitions. Definitions are a made up abstraction that don't exist in real life. They're useful, but they're just approximations. Words actually have uses. And so when a person says, you know, the stock market was growing smaller today, everybody knows what that means. Nobody even stops and thinks about the fact that the word grow has a definition that means to get larger. And then sees it as some sort of contradiction. It's because we just simply know the word can be used in that way, period, end of story. And it communicates an idea. If it happens often enough, you don't even feel awkward about it. You just use it. And he points out that over time, this means that words will start off with some sort of meaning.

[00:48:30]  Bruce: And like he uses like the word handsome, like what's handsome got to do with hands and some, you know, at some point in our history, they had some sort of relationship to those words, right? It had some sort of relationship to those words, but it's lost it, right? There's all sorts of words like that, where when you try to pick them apart, he points out that they either have lost their meaning entirely or they've partially lost their meaning. Like you can still make out what the subwords are.

[00:48:59]  Cameo: What happened?

[00:49:00]  Bruce: Right, you can still figure out where the word came from because the words haven't entirely lost their meaning yet. And yet you never really think about it. You just simply know the way we take words into our minds and use them with concepts. We simply use the word and it becomes a signal for that concept, but only within certain contexts. And we're always using the full context of the sentence, the situation, someone pointing, things like that, to be able to resolve what idea is actually being communicated by this word. And you have to do that because otherwise it would be impossible to communicate because words simply don't point to single, they never point to single concepts.

[00:49:43]  Unidentified speaker: Okay.

[00:49:43]  Cameo: So I propose we stop talking about this right now because we could do another whole hour on that. And let's either do it for next time or the one after because I know you also got a subject ready to go for next time.

[00:49:58]  Bruce: Sounds good. Oh, just before we go, I want to bring up one other thing that was in the Michael Levin's thing, the Xenobot. What did you think of that?

[00:50:05]  Cameo: Xenobot. So you're going to have to help me remember the...

[00:50:10]  Bruce: Xenobot was where they took the skin cells of a frog. They let it [unclear]. I can't remember if they told it to clump together, but it clumped together and it formed a little organism. And the organism, the... Remember, the genes aren't being changed. The organism learned to steer itself around and explore its environment, even using little hairs or cilia that were intended to move mucus. So they had no purpose. They weren't... The purpose, if they were inside the frog, would have been completely different than locomotion. And yet it learned on its own to create locomotion and move itself around and to explore its environment.

[00:50:55]  Cameo: I don't really... In a way, it seems like a difference... Like, where is that instruction coming from? His assertion is that still, is that essentially electrical subset of rules that are actually stronger than the genetic set of rules, perhaps.

[00:51:17]  Bruce: That seems to be what he's implying, yeah.

[00:51:19]  Cameo: That has some huge implications for even what our current understanding of... Everything we know about life is really genetically driven. Like, we know that electricity is a thing. But all of our understanding is, that's where we've been putting all of our focus. It means that we have no idea what's going on, honestly.

[00:51:42]  Bruce: You know, I suppose... I mean, I would love to see people who know more about this than me really assess the xenobots. You know, is it possible that it's not actually exploring the environment, it's just by chance looks that way or something like that? I would love to see criticisms about that. But certainly the pictures that he showed of it, it sort of looked like it was kind of swimming around, exploring its environment. And so that would be very suggestive that the operating system had a good enough intelligent error correction algorithm that it was able to take these skin cells intended for an entirely different purpose by the genes and repurpose them and basically turn itself into its own little organism, an organism that doesn't exist anywhere else in regular life.

[00:52:32]  Cameo: In nature, yeah. And that's kind of not scary, but it just means that we really don't understand. And that wouldn't be that big of a surprise, but I think we have a desire to believe, especially as we've made enormous advances in understanding the genome to feeling like we're cracking the code. But if there's this whole entire subsystem that actually has complete override capabilities...

[00:53:03]  Bruce: Yeah. Yeah, I see where you're going with this.

[00:53:06]  Cameo: Yeah, I mean, it means that we've gotten really good at the... We've done a lot of discovery on the hardware, but we had no idea that there was a separate sub software operating system.

[00:53:18]  Bruce: Yeah. You know, I agree with everything you're saying, but if you really think about it, that probably is a big surprise, but it probably shouldn't be, because of course minds have been that way for forever, right?

[00:53:31]  Cameo: Sure.

[00:53:32]  Bruce: You know, clearly the genes somehow give rise to minds, and minds are able to do purposeful, you know, evolution and changes that the genes never intended.

[00:53:45]  Cameo: Sure, sure.

[00:53:47]  Bruce: So in many ways, it's maybe a shock, but maybe it shouldn't be a shock.

[00:53:53]  Cameo: Yeah, you're probably right. Fascinating, though.

[00:53:57]  Bruce: Yeah, that was one of the things that I found the biggest. The two things that really wowed me the most were the two-headed flatworm that could reproduce itself as a two-headed flatworm without using CRISPR, and the other one was the xenobot. Again, I kind of want to see what other scientists say about this. I want people more knowledgeable to first criticize what he's saying, [to see whether] it survives criticism or not. But if it is actually taking skin cells and repurposing it into its own little organism, that really is kind of a big deal, right? That is a surprising level of intelligence that exists at the biological level outside of the genes.

[00:54:39]  Cameo: Yes. And we could talk about it for hours. But great conversation, Bruce. I'm glad that you shared this as a topic.

[00:54:49]  Bruce: Yes. And thank you for talking with me about it.

[00:54:51]  Cameo: Yes. Always happy to.

[00:54:53]  Bruce: Talk to you later.

[00:54:55]  Unidentified speaker: Okay, bye.

[00:54:59]  Bruce: The Theory of Anything podcast could use your help. We have a small but loyal audience, and we'd like to get the word out about the podcast to others so others can enjoy it as well. To the best of our knowledge, we're the only podcast that covers all four strands of David Deutsch's philosophy as well as other interesting subjects. If you're enjoying this podcast, please give us a 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.

[00:55:45]  Bruce: The first is via our podcast host site Anchor. Just go to anchor.fm slash four dash strands, F-O-U-R dash S-T-R-A-N-D-S. There's a support button available that allows you to do reoccurring donations. If you want to make a one-time donation, go to our blog, which is fourstrands.org. There is a donation button there that uses PayPal. Thank you.