Professor Mike Levin: Agential Materials #biology #science #philosophy

Hey, good to see you again.

Well, just to say thank you. I appreciate that. A lot of people like it. I mean, I think Kurt does an amazing job and he does something slightly different than what I do, but I think he does great and other people do as well. But for me, you know, we have such a richness of incredible collaborators. That's first. And second, you know, so this doing the open science, you know, and getting letting other people see at least some of what's under the hood, I think is important. And also I try to use it to shed light on some of the junior scientists. We have a bunch of ones there and many more going forward. You know, very promising young people who I think need more attention for their work. So this is, I think, part of the open science. So thanks. I'm glad it resonates.

Yeah. Well, a few things. So first of all, you know, you're right and most of the people, I think most of the people on that symposium do not share my specific view. They're interested in pieces of it and some are compatible with various things, but most of those people would not subscribe to my current view and that's completely fine. My goal is not to amplify things that I say, but it's to have a good productive discussion where everybody learns something and new things come out that I haven't said. So that's great. My view of it is several things. I think the most important, if I had to say number one most important thing that I think people get wrong about it is that the assumption by people who hear about it for the first time and don't really read carefully the things that I've said about it assume that it's some form of mysterianism, that basically what I'm saying is, look, there's a bunch of stuff we don't understand now, therefore it must be a big mystery that we're never going to understand and it comes from some weird other realm where we can't reach. That is 180 degrees opposite from what I'm saying. I'm saying the exact reverse, which is that I think the current emphasis on emergence as an explanation, and I know people say, well, emergence isn't supposed to be an explanation, but that's how it's often treated. So if there is some particular outcome that is surprising, it doesn't have a straightforward story between cause and effect, there's some surprising outcome, well, people say, well, then it's emergent. And I say, what does that mean? And they say, well, these are just facts that hold in our world. In other words, why this thing emerged versus this other thing, and we can talk about some examples of that, and they'll say, well, they're just facts that hold in our world. And so on the one hand, no one really likes, and I've gotten more, you know, sort of hate mail and also like attempts by people to talk me out of it over Platonic space than almost anything else that I've written. And I've said a lot of, you know, counter-paradigm things throughout my history in science. But this one really like triggers a lot of people.

And so no one wants another realm. And then I say, look, these things, the specifics of what emerges, does it have any kind of a structure to it or is it random? Because you've only got, as far as I can see, you've only got two choices. You can either say it's completely random. There is no more story to be told about why you got this rather than that. That's just it. You just have to sort of add it to your big book of emergent surprises and that's all there is to it. You could say that. Or you could say that, well, it's random. There's no structure behind it. Both of those things I think are very pessimistic and kind of antithetical to the assumption of science. We do science because we have the belief, and you can't prove it, but the belief that the world is understandable. And therefore I would much rather think that there is a structured, ordered space of possibilities, a latent space from which these surprises are drawn, and that we absolutely can and should a) study the contents and the structure of that space and b) try to understand under what circumstances and why the contents of that space impact things that happen in the physical world. And the minute you say that, okay, no, these things aren't random, that's it. Now we're in the same camp. Now all I'm saying is you don't have to, if you don't want to call it a realm, don't call it a realm. That's fine. But you've already said that there is some structure behind these things that happen. And as soon as you've said that, if you accept this optimistic idea that no, we're going to study it, right? This is a research program. It is not a mysterian attempt to say, wow, we're never going to know. It's the exact opposite. I propose a very specific research program around it. I have somewhere between five and six people in the lab currently doing experiments and working on these things. So that's the kind of number one thing that I want people to know is that I think it's actually far less a mysterian approach than what I think is currently happening.

I think that's right. There's a lot of misunderstanding. I think that if you don't want another realm, that's okay. But you can't assume monism. That isn't science. You can't have it on the cheap. You can't just say, well, I want a single realm. Therefore, there's a single realm. No. If you would like a single realm, you have to show how everything boils down to a single realm. And what I think is not useful is just to take everything that exists. We're going to take, for example, the objects of mathematics. We're going to take the objects of physics and we're just going to say all of that is one realm. It's the universe with a big U. Great. But I'm not interested in labels. You're still going to end up with two subrealms, each of which is accessible to completely different approaches, techniques, and so on. So no matter what you call it, they're not the same. And so I think it is true that there are important facts that are not facts of physics. And what I mean by that is you're not going to discover them by doing physics. In other words, you can't just fire the math department and hope the physicists are going to tell you why it's the four-color problem and why the primes are distributed in a certain way. The physicists aren't going to tell you that. You can't just do that. That's a. b. What you also can't do is at the beginning of the Big Bang tweak some of the fine, you know, tweak some of the dimensionless constants that define our universe, however many, say half a dozen, however many there are, and I'm going to make the natural logarithm be something else than the value it is now. That isn't going to work. And the more you know, the other thing that I would point out is imagine if I tell you that, you know what, it turns out that the speed of light actually changes very, very slowly. It changes, it's not actually constant. There are physicists who have suggested this. And if I say that, you say, wow, that's wild, but I guess I can live with that. If that's what happens, then that's what happens and we can figure it out in physics. If I tell you, and I'm working on some fun stuff around this currently, if I tell you that actually the value of e or the digits of pi out beyond, you know, a few hundred digits that most people ever look at, yeah, some of them are starting to float off too. They're different than they used to be when it was first calculated. Everybody says that is insane. That's ridiculous.

You know, you started out with a minimal assumption, something like, you know, the empty set and the successor of the empty set, or maybe, you know, logic or something. And pretty soon you discover the value of e and the value of pi and you can't just, as you could, you know, math is a human invention, great, let's say you change some of those things in ways that you would like, right? Like, yes, there are other kinds of mathematics where things look different, but having made some simple assumptions, there are things you cannot change. And the things you cannot change, even though you would like to, are the things that you did not create, invent, or any of that. You discovered them. And so I think that is right, that idea that the outrage that you would feel if I made the claim that e is sort of changing and whatever, that's like, we all already buy into this, this intuition that yeah, these are facts that are to be discovered, not things that we just invent. And so that's it. So from there, I only say a couple of new things after that. A, that we already know that some of these facts that are not facts of physics, that can't be changed by things you do in the physical world. On the contrary, in fact, they do impact the physical world. So I think they're causal in the way that causality is the most useful, again, to me as an engineer. I don't want to, you know, people ask, what's the ontological status of the Platonic space? I have no idea what that question really means. I have very little interest in like labels of ontological status. To me, it's real if as an engineer I need to worry about it and I can exploit it. Either of those things is true, it's as real as real is going to get for all of us finite observers. So, yeah. Right. So these facts of mathematics impinge upon things that happen in biology, things that happen in physics, and so on, because they are the actual, the most insightful reason for why things are the way they are. What I mean by that is if you start with any fact of physics, you know, why did the fermions do this or that, or you start, you know, why did the cicadas come out at, you know, 13 years and like a 5-year-old, if you just keep asking why, eventually you end up with the mathematicians. Like you just keep asking why, why, why, why, and eventually it's okay because this thing has certain symmetries. Well, why is that? Go talk to the mathematicians. They'll tell you why. And that's it. So everything eventually ends up there. And to me, an interesting definition of causality is the best explanation for why something happened. That's the cause of it. If you want to know the cause of why something happened, the best explanation is not the one with the smallest, you know, particle, whatever, or it's not the lowest kind of, the one with the lowest agentic level or something like that. It's the most insightful explanation. It's the one that tells you why you got this rather than that. And I would argue that these facts serve as that for a very wide range of phenomena. And the final thing I will say is that so far I don't think I've said anything that's like, I feel all of these things that I just said are pretty strong. I realize that there are people who are not Platonist mathematicians and so on. Like I know all that, but I think these are pretty strong. But now we can make a hypothesis. And the hypothesis is this. As much as until now, most people have pretty much assumed that even if there is a latent space of Platonic truths, these are only relevant for mathematics. These are the truths of mathematics. They are limited. They have very low agency. They're static. They don't change. And that's it. And everything else kind of has conventional explanations with conventional causality plus this magical emergence. So the hypothesis I want to make is this. Let's relax that axiom. I don't know why we would have it as an axiom. I don't think we should assume it. I think we should treat it as an experimental outcome. Let's ask what implications for experiment and for applications and so on can we have if we say that in that latent space of non-physical facts we have the kind of static, low-agency things that are amenable to our mathematical kinds of techniques. But what if there are also patterns that we would recognize? They're more complex, higher-agency, active patterns that we recognize as forms of behavior, aka kinds of minds, right? Behavioral propensities. Maybe they are patterns of shape that we might be interested in in developmental biology. Maybe there are patterns of physiology, you know, things like that. So what if it's the case that actually that space has something to inform all of the disciplines, not just mathematics, but maybe all of the disciplines have something to learn from it. And then that makes a very clear set of predictions for a research program, and that's what we're doing.

Well, I do think it's correct, and so I want to be very careful here because there is a lot of tendency to, what I've noticed is that there's a tendency to take some of the things I've said and run with them in directions where lots of people have different agendas, right? Where they want this kind of thing to go. And so I want to be careful and not commit to anything other than the things I specifically say here. First of all, I do think it's true that the kind of formal models we have for clockwork, so mechanisms, machines, algorithms, these kinds of things, I do suspect that these formal models fully describe probably nothing in our world. That's my guess. Now, let's just remember, first of all, this is not a wild claim. I think philosophers have nailed this down long ago that no formal model should be taken as fully describing any phenomenon. The question is how useful is your formal model in specific cases. I think we can say that. And so now the question is, okay, we have this formal model of machines, of mechanisms, of clockworks that lack certain functions that we supposedly have. Of course, once you take that view, then eventually you run into the question, wait a minute, my stuff obeys the laws of chemistry. So how am I escaping all of this? Well, we can get to that. But at least whatever the mystery there is, at least we can say that like these simple machines and these dumb mechanisms, they're just clockwork. They don't have these other properties. So I think that's fundamentally mistaken. I think it's mistaken philosophically and I think it's mistaken empirically as a matter of experiment. I think that's wrong. And I think that, yeah, I think it's the case that these kind of formal models really need to be updated. I think they're missing a lot. They really need to be updated. I don't think at this point there are probably any truly dumb machines and clockworks. I think there are lazy observers and observers that haven't learned to detect the... I mean, the bubble sort example you bring up, people have been like, every CS 101 student has been studying bubble sort for how long? 80 years or something like that, I guess. Nobody had noticed this, and it wasn't because they weren't smart. There were plenty of smart people looking at it. It's because most people had assumed that it only does what you've asked it to do, does exactly what you've asked it to do. Reasonable. That's how we think about it in computer science as an algorithm that has specific steps for doing specific things, and those are the things you look for. So I think I'm not sure in the end when we really get this diverse intelligence research program mature, and it is only at this point the field I think is at a very early stage, when we get as mature, I'm not sure there are going to be any clockwork dumb machines left in the universe. Now, that's not to say that for some purposes that isn't a useful frame, right? So if you're, you know, just as an extreme example, if you're doing Archimedes' simple machines, you're probably fine for the purposes of leverage and things like that. You're probably fine taking that framing and not getting into, well, does it have minimal anticipatory capacity and all of that kind of stuff. So there will be scenarios where this is perfectly useful, but there will be other scenarios where it misses everything of importance. And we already know about this. If you in your house, if you have an issue with your heating and cooling system and you call an HVAC technician and they show up and they say, yeah, I'm a reductionist. I don't believe in goals. I certainly don't believe in machines that have goals. There's just atoms and the void. And okay, if you get an HVAC technician that doesn't understand that your thermostat is a goal-seeking system, if they think that goals are magic instead of cybernetics, you've got a bad HVAC technician. So we already know that there are certain framings that, not because anybody's interested in mystifying anything, but it's simply no longer magic to have systems with goals that are quite, quote-unquote, simple. So I do think that perhaps there is no purely inanimate matter, dumb machines. I'm not sure there really are any of those things. But the hard work now is to see, first of all, how many of our frameworks in cognitive science and behavior science are applicable in weird substrates that we wouldn't think they would apply. And then the harder problem, that's actually the easier problem. We know how to do that. I've shown many examples of this in our lab. We've done it. We know how to do that. Other people have too. Then there's the harder question of saying, okay, so now we know behavior science is at the bottom, not at the top of the, you know, how they do this pyramid, right? Well, there's physics and the chemistry, and so I actually think it's upside down. I think behavior science is more fundamental. But the next question is the harder question is going to be, have we got the major categories in behavior science or are we missing some? Because we may, you know, porting it is one thing, making sure it's relatively rich and complete is, I mean, it'll never be complete, complete, but making sure we've got the major categories that we need as a mature species, that's harder. But nevertheless, we have to take that journey, I think.

I think that, well, you know, there are plenty of smarter and other open-minded people out there. I think it's a couple of things. I think one is we are in my group, we are able to span the gamut from really deep philosophical, and that's what I consider to be my job in the group is to try to paint the roadmap from deep philosophical questions to computation models to actual at-the-bench experiments. So on our best day, we take things that people have argued about for thousands of years and we figure out what the implications would be of specific models for that, and then we try to test them. I do think that we've made some progress that is very exciting. But again, just the beginnings, just the beginnings of it. And I think that a huge part of it is, and again, I'm in no way claiming that I'm the only one that does this. Lots of people do it. But I think maintaining a kind of a sense of childish wonder about things is really critical because there is a tendency to move the goalposts. So the first time people see something, I can give you a few examples. The first time they see something, it's kind of shocking. But after a while, that becomes the new floor, the new normal, and we don't think about it much anymore. And when that happens, you say, well, I guess that's just how it works. And I am always very skeptical about that. And when I see those kinds of things, I don't adjust my floor very easily. And so when I see those things, I think, yeah, no, I don't think we can just say that's how it works. I think there's something buried here. And the biological or biomedical version of this is the placebo control trials. So the way that works is you test some sort of drug and then of course you have a placebo control. And let's say you find out that, well, the placebo group, you know, 20% of them had an improvement, in a drug group, you know, 60% of them had an improvement. Awesome. I'm going to subtract it. I'm going to say my drug is 40% effective. That's fantastic. But once you've done that, you've lost the first amazing thing. It's like, wait a minute, why did 20% of them get, why is the placebo 20% and not zero? So by redefining the floor, by saying that my floor is now 20, so everything sort of gets adjusted relative to that. We're not going to look into why that happened. Like this is baseline. That I think is part of the problem. And there's a bunch of discoveries going back, you know, well over 100 years where progressively we just get kind of habituated to things and we just sort of assume it's, you know, sort of it's fine. But the more sort of impressive these things get, the more I think that's a mistake. You know, the more I think that, no, actually we should be asking for why that's happening. We can't just say, well, these, you know, this is how things happen in our world. So there's a whole sort of ladder of these things that I'm writing a talk right now where I, you know, kind of go down all these carefully, both in biology and in computer science and math. And then once you kind of look at it as a pattern, then you realize that, yeah, actually I think we're missing something very important. And, you know, to your previous point, I mean, I'm certainly not going to venture a guess as to what, you know, what the large-scale purpose of things is. You know, there are plenty of people throughout history and now who talk about that stuff. I won't, but I will say that the basic capacity to have goals, and in some cases, the capacity to find new goals when your prior goals aren't working out or when you're bored with them or whatever, that whole spectrum, that stuff, I think, is baked into our world. I don't know if it's possible to have a world without that, but in this universe, I think it goes all the way down.

Yeah. So there's a few things we can say. So I'll talk about placebo. Remind me about the brain thing because I'll forget.

So one thing is I think placebo research is very interesting. I think it's really fundamental. And Fabrizio Benedetti, who's one of my favorite placebo researchers, has this quote where he says, words and drugs have the same mechanism of action. And it's really interesting because any cognitive system in order to act in some way has to be able to transduce abstract high-level goals in spaces that its parts don't know anything about down to the action of the parts. That's the whole point of being a collective system that does something that its parts don't do. We are all collective, like our bodies are all collective intelligences. And so just as an example, when people say, wow, placebo, it's so weird. It's, you know, the mind is affecting the body. It's like this rare mind-body medicine kind of thing, right? So unexpected. My point is, look, when you wake up in the morning, you may have financial goals, social goals, research goals, whatever they are. These are very abstract, high-level cognitive things. In order for you to do any of those things, you have to get up out of bed and walk. That requires ions to cross muscle membranes. The chemistry literally dances to the tune of these incredibly high-level abstract things that don't exist at the level of physics. What are we to make of voluntary motion, right? Like, yes, you can tell a story that, you know, obviously people who want to remain strict materialists will say there are no high-level goals. What there is is neurotransmitters pushing each other around and so on. That's fine. You can tell that story, but let's not forget you can always tell a story at any, you can always find a perspective that misses the whole point. Like that's not hard. And I think that's what this kind of materialism does because the most insightful explanation of what actually happened, you know, it's like the goal-seeking behavior, airline prices. You know, you're not going to tell an insightful story about why airline prices went up or down in terms of electrons. It's just not going to happen. And that's not a philosophical preference. That's an empirical statement that from certain perspectives it's very hard to make things happen if you don't believe in cybernetics. You're going to have a hard time. You're going to have to really micromanage that thermostat, you know, at the molecular level. It's demonstrably, objectively harder to do things that way, which says to me that you've missed the best perspective. You've not actually coarsened the system correctly. So to go back to placebo, this idea of my body is not some rare thing, you know, tons of yoga, meditation or whatever. It is the everyday reality most of us, and it can break down like with locked-in syndrome and things like that. But it's the everyday reality of voluntary motion, and our bodies are built specifically on an architecture that allows to transduce those kinds of abstract things down into the movement of ions so that you can act on. Right? So this idea of systems that are acting not because they're being pushed and pulled by the outside world like a dandelion seed flying around out there, but actually are being driven by their own internal state, that's very widespread. That's obviously not just for humans. It's not just for brains. That's very, very widespread. Things like perceptual control theory and active inference and so on really are fundamentally all about that. This idea that you act in accordance with your expectations, you know, beliefs and expectations and things like that. And so the simplest example of placebo effects that I know of are our studies of gene regulatory networks. So these are simple pathways, a handful of molecules turning each other on or off or up or down or whatever. And what you can show is that for a large class of these things, they are capable of doing Pavlovian conditioning. They're capable of learning associative relationships between two stimuli. So you stimulate one molecule that causes a response. You stimulate another molecule, doesn't cause a response. You do them together, and much like Pavlov's dog with the bell and the food, eventually even the inert one will cause the effect to happen because the system has changed state to associate them in such a way that you see this is a, and it's very important, this is a software thing. It's not hardware. It isn't that we rewired the network. It doesn't require any changes of hardware whatsoever. It's invisible at the, this learning is invisible at the level of the hardware. If you X-ray it and see what all the connections are, you're not going to see anything. It's a dynamical system property that's taken on because of prior experience. And now the system will react to a stimulus different than it would have before the training. I mean, that's the definition of memory. And so what you can say then is that the system is implementing a kind of placebo effect because that second drug does not actually, whatever actually means, does not actually cause the response. At least it didn't before you trained it. Now it does because the system believes it should. And I'm not saying that gene regulatory networks have human-level metacognition where they can sit there and think about their own beliefs and all that kind of stuff. We have no idea about that in these simple systems, but functionally they learn, they have associative conditioning, and it absolutely is, I think, a valid form of placebo effect because it's acting in accordance to meaning of the stimulus, and that meaning is not set by the physics of the system. It's not set by anything like that. It's set by prior experience internal to the system. The system is such that it assigned that meaning, and now that's how it acts. So I think that's a minimal model of the placebo effect.

Yeah. Right. It's rare. It's not common, it's rare, but it isn't one patient. Karina Kaufman and I published a review recently of these patients. You can see it, we discuss a bunch of the different scenarios. Look, yeah, and the scenario is that what you have is a person that is missing a huge amount of the brain tissue that's normally thought to be associated with these functions, and they have normal or above-normal intelligence. That sort of phenomenon are cousins to other phenomena where you have savants without a history of education in a particular field who are really good at certain things. They call out prime numbers and things like that. And other scenarios, you know, geniuses, you see these little prodigies, somebody sticks a two-year-old in front of a piano and off they go, right? So all of these are related phenomena.

And what they're related by is that it subverts our expectations about what's needed for a particular level of performance. This is actually really important because that's the basis of my research program in the Platonic space. They are outlier cases, and why your friend thought it was ridiculous is because people have expectations for how much effort or computational material or whatever it costs to get a certain outcome, a certain degree of performance. So let's ask, first of all, where do our expectations for that come from? How do we calibrate them? How do we know what to expect? So let's say I show you a chunk of brain tissue, or in fact, let's say I show you a whole human brain, and they say, how many human individuals do you think fit in there? Like when we do computer science, we kind of know if you want to do a certain amount of computation, we know how many logic gates we need to have, registers, whatever. How many do we need to have for brain tissue? What's the density like, cubic centimeters per human personality? How many should there be?

And if you have an answer, I mean, we have an answer, well, roughly one per brain. That's not because we understand what's going on, that's just because usually that's what happened. And so this goes back to my point about not being — you know, a little humility about this stuff is really important because we have to ask ourselves, why do we have a strong expectation? And we do, we have these intuitions. What are they based on? What do we actually know about the density? And so then you have phenomena such as, well, first of all, if you were to — you know, in these experiments where they inject sodium pentothal into one of the carotid arteries, the left language-using hemisphere goes to sleep, and when that happens, you can talk to the right hemisphere. It doesn't usually talk language, but it can draw and it can pick multiple choices and things like that.

And then you find out that, wait, that thing has opinions that the speaking patient didn't have. So is it really one-to-one? Then you find out that, oh wait a minute, dissociative identity disorder, you know, somebody has half a dozen personalities, more, you know, dozens of personalities. What is actually the density that you can expect? How much brain does it actually take to run a full human personality? We don't know. We don't know because we don't really know what the mechanisms of memory are. We know what mechanisms are required for memory to be apparent. We know things about — and I haven't said anything yet about consciousness, but just to say, we know what it takes for the speaking left hemisphere to claim that a particular phenomenon was conscious. We have no idea what that means about the other stuff in your body that may or may not have conscious experience but doesn't talk about it, can't talk about it. We don't know, you know, like they say, history books are written by the winners, right? Neuroscience books are written by the winners, the linguistically capable left hemisphere who says, 'Yeah, that's unconscious learning because I didn't have a conscious experience of that learning.' Well, right, you didn't, but also I'm not talking to everything in the body, I'm talking to one particular module that happens to be good at language. So I'm very skeptical about those kinds of things too. So I think we have to back off from this idea that we know how much brain is required for whatever, and I think we don't. And the thing is that these cases are rare, but they do occur, so we have to deal with them. And typically what people would say in neuroscience is, well, redundancy. So actually, yeah, something else, you know, you're missing some stuff, some other stuff, you know, kind of stepped up and took place. I'm not saying that we can't sort of epicycle these things into our understanding the way that people used to do with planetary motions that don't fit the earth-centric model. You can, but that's not the same as having a framework that actually predicts that this would happen.

And this is important because when you look at an early embryo, for example, I've done these experiments in duck embryos as a grad student. You take an early embryo. It's like a flat disc of hundreds of thousands of cells, and you can cut it into pieces, and each of those pieces will become its own embryo. So how many individuals were actually in there? It isn't fixed. It's not determined by the genetics, right? It could be anywhere from zero, one — usually one — but it could be up to half a dozen in the case of the bird embryo. How many individuals are actually in there? And for that matter, what is an embryo? Do we have a definition? It's kind of interesting. We're actually trying to do that now, mathematically calculate using tools from causal information theory to try and actually say what an embryo actually is. I don't think developmental biology has a definition for what an embryo is. I mean, we can all point at one when we see one, but that's not the same as having a good, an actual definition, specifically in cases where we know that a blastula can generate numerous distinct individuals. And then you have interesting questions if you let the pieces heal up, then they're conjoined twins and triplets and whatever, and then you've got cells in there that have to decide, am I the right side of this guy or am I the left side of that guy? Like, where's the border? Where do I end and this other individual begins? These are all questions that are very live and actionable in biology.

And so we sort of...

So, no, they are identical because — well, identical if you mean genetically clonal, they are.

Yeah.

Yeah. And so what we have here is again this kind of mismatch between what you expect and what you get. And to me that's just a reminder that our expectations are in some cases not based on much of anything at all. And in other cases, in particular like in the computer science cases where actually the expectations are firmly based on strong theory that has been around for a long time — you know, what can you expect out of an algorithm, what do mathematical features do, and so on — in those cases, it's not based on nothing. It absolutely is based on strong theory. But I think what it's telling us is that those theories are incomplete. I don't think anybody should be terribly shocked about this. I think we should assume that all theories are incomplete. And I don't think there are any sacred cows, or there shouldn't be, in any of this stuff. And so if you're getting a mismatch between what standard ways of thinking about things have led you to expect and what actually happens, then that's a break on technological progress. It means you don't understand the universe as well as you'd hope, and then we just look for better formalisms.

Absolutely. Absolutely. Yeah. No, we haven't handed anything over. I mean, anybody can — everything we do is preprinted for free and then published, so anybody can work on them if they want. There's also some pretty good work out of Charlie Ren's lab on a variant of anthrobots.

Charlie Ren, yeah.

Yeah, he's not doing the basal cognition stuff that we do, but he does disease models and things like that. So there are other people working on it. No, anthrobots. Anthrobots are amazing, and here are the broad sketches of our research program currently with anthrobots. First of all, yeah, they can heal neural scratches, which raises the possibility of in-body repair. So biobots made of your own cells, which means you don't need immunosuppressants, right? It's your own cells. We need to figure out what else, if anything else, they can heal. So we're doing those experiments. We're checking cardiac tissue. We're going to look at cancer. We're going to look at immune issues. We're going to look at microbiome issues. So all of that. So that's kind of the biomedical side of things. Then there's the fact that anthrobots, by both epigenetic and transcriptional clocks, anthrobots are younger than the cells they come from. They're younger. That means that the process of morphogenesis, the process of becoming a coherent, higher-level, higher-scale being out of parts, that process itself can somehow roll back the age clock. I have some thoughts about why that would be, but obviously that's very important and we're looking into that as part of our aging and longevity research.