Watch Episode Here
Listen to Episode Here
Show Notes
This a ~1 hour 20 minute conversation between Aastha Jain Simes (https://www.livelongerworld.com/), Pamela Lyon (https://scholar.google.com.au/citations?user=1oY1J5kAAAAJ&hl=en), and I. Pamela is a key figure in the development of a biogenic approach to mind, and we talked about her journey to understanding cognition. A few of her papers:
https://aeon.co/essays/the-study-of-the-mind-needs-a-copernican-shift-in-perspective
https://journals.sagepub.com/doi/full/10.1177/1059712319871360
https://link.springer.com/article/10.1007/s10339-005-0016-8
CHAPTERS:
(00:00) From Buddhism To Cognition
(11:18) Autopoiesis And Vitalism
(20:09) Microbial Cognition And Cooperation
(35:36) Slime Molds And Bioelectricity
(41:55) Stress, Creativity, And Evolution
(55:28) Rethinking Biology And AI
(01:10:31) Eastern Views On Cognition
PRODUCED BY:
SOCIAL LINKS:
Podcast Website: https://thoughtforms-life.aipodcast.ing
YouTube: https://www.youtube.com/channel/UC3pVafx6EZqXVI2V_Efu2uw
Apple Podcasts: https://podcasts.apple.com/us/podcast/thoughtforms-life/id1805908099
Spotify: https://open.spotify.com/show/7JCmtoeH53neYyZeOZ6ym5
Twitter: https://x.com/drmichaellevin
Blog: https://thoughtforms.life
The Levin Lab: https://drmichaellevin.org
Transcript
This transcript is automatically generated; we strive for accuracy, but errors in wording or speaker identification may occur. Please verify key details when needed.
[00:00] Aastha Jain Simes: Pamela and I were talking about this offline. A good place to start would be understanding how you got into the field of cognition, because it explains a lot of your work on cognition research. You were doing a PhD in something else, and you stumbled upon the fact that the current models of cognition are not correct. So I'd love to understand your foray into cognition research, and also what you think is incorrect about some of the current theories of cognition.
[00:33] Pamela Lyon: Okay, thank you. I started out in a PhD doing a cross-cultural comparative analysis of four Buddhist propositions. This is Buddhist philosophy. Now, behind those propositions is a concept of mind, cognition. What mind is, what cognition does. And I realized that I actually couldn't do the comparison unless I could come up with consensus views of what cognition was on both sides, on the Western scientific side and the Buddhist side. Now, Buddhist side, no problem. They had a very, very clear understanding of it. They'd been working with it for 2,500 years. Meditators had been developing extraordinary states of mind using this concept. But when I went to the Western side, I couldn't find anything. Now, people say that's really uncharitable, but what I was looking for was something that you could do something with. That you could actually say, well, this is the way the mind works. This is what it is. The mind is the brain. The brain does a whole lot of stuff other than just being a mind. The mind is computation. What does that actually buy you when you're trying to figure out what a mind is, what it does, what it's for? It didn't seem like there was enough there there. At the time I was looking at this, this was cognitive science. There was really very little biology, almost no biology. And evolution was trotted out as something to indicate something about biology, but it really wasn't pursued in any way, shape, or form. Understand at this point, in the late 1990s, coming up to the 21st century, people were only just beginning to talk about embodied cognition. They were starting to talk about animal minds, animal cognition. The seminal book on animal cognition was published in 2001. That means there was a cognitive revolution going on for practically 50 years before we got to that point. And I was just dumbfounded. I didn't understand how we got here. And I couldn't find anywhere a T-shirt worthy definition of cognition on the Western scientific side that I could compare with the Buddhist one. There just wasn't one. There were lots of them.
[04:07] Pamela Lyon: So I thought, this is the problem. Forget the Buddhist propositions. Forget cross-cultural comparative analysis. This is the gap. And so for me, the question was, how on earth do you go about filling this gap? For me, it meant it has to be based in biology because it has to be based in evolution. We know that living things started out small and developed over time. And there was lots of debate and argument about where on the living spectrum cognition jumps out. Genuine cognition. I thought I can't get involved in those arguments because I think the basic premise of that argument is wrong. It's not "where does it jump out?" It's "what is it doing? What is it doing for an organism?" You have to ask those questions first. I did a quick dash to the microbiological literature just to see what was going on there. I had no idea, and I'd been told by everyone I was in discussion with that there was nothing going on there. It was all just reflexive, hardwired, programmed—nothing to see here. Luckily, I had been trained by my experience of writing about the molecular work of two Nobel Prize winners in the year that they won the Nobel, Joseph Goldstein and Michael Brown, who won for cholesterol. I was working in the publications office of what is now Southwestern Medical School. They won the Nobel. Somebody had to write about what their work was, and it was down to me to do it. When I went to ask them questions, they handed me a stack of primary material and said, "You're a smart girl. Come back with questions." I just went, "You're kidding." As it turned out, I took to it like a duck to water. I came back with questions, some of which were open questions in the field. They said, "That's a very good question. We don't have an answer to that yet, but there are people looking at that." They gave me confidence to do it. When I went to the microbiological literature, I had already been trained to go to that literature and work it out.
[07:41] Pamela Lyon: And what I found was dumbfounding. Not only sensing, but sensorimotor activity and memory and non-associative learning and making decisions under uncertainty and error correction and anticipation or prediction. It was like there's communication. There was an entire world there that nobody was touching. Nobody was touching and everybody thought there's no there there. So once I discovered that there was something really basic that involved cognition, I said, okay, well, what is biology got to say about cognition? So I started looking at self-organizing complex systems, that entire body of work. I was looking at Maturana and Varela. I looked at Stuart Kauffman. I looked at Walter Elsasser. I looked at Robert Rosen. I looked at Ludwig von Bertalanffy. I have this strange mind; I hoover up everything and then feed it to myself and my subconscious runs the numbers and comes up with something. I really don't know how it works. And so I started seeing patterns in self-organizing complex systems. That gave me the first set of what I call biogenic principles. When I say principles, it means whatever explanation you come up with for cognition, it must accord with these principles. What I found in cognitive science was that there really weren't very many principles that held anybody to account. People could make up almost anything they wanted to and say, well, I've got a thought experiment that says blah. Dozens and dozens of papers would be written on blah, but it actually didn't move the dial at all on what cognition actually is and what it does. It wasn't that it was wrong; it's just that there was an elephant in the room that nobody was looking at because everyone was concerned that any definition of cognition had to apply to machines as well as living systems. Noam Chomsky wrote a brilliant takedown of this entire ethos in an essay on the 1990s side of the turn of the century. When I read that, I went, right, that's exactly what I'm seeing. I was constantly triangulating what I was finding with people who were leaders in the field to make sure that I wasn't saying something off. That's because I used to be a journalist. When I was a newspaper journalist, everything I wrote had to pass muster with people who read it and all kinds of people read stuff in newspapers. We used to have a more common view of how the world works. I was paranoid about being wrong. As we say, I covered my *** really well, because I had to. And that's how I got into this field.
[11:18] Aastha Jain Simes: When you say "self-organizing complex systems," what does that entail? What's the full spectrum of what that is?
[11:27] Pamela Lyon: The thing about living systems is not simply that they self-organize or that they maintain themselves. As Maturana and Varela pointed out, this was Maturana's original insight. But it's that living systems produce themselves. They don't just—they're complicated and they're complex, meaning you really can't predict what systems you're going to get from the underlying dynamics necessarily. That's where I think complexity comes in, although I am definitely not an expert on complexity by any stretch of the imagination. They self-organize, they self-maintain, and they self-produce. So that is the entire spectrum. They are systems. Von Bertalanffy was the first one to generate an idea of general systems, a theory of general systems. He introduced systems. That was picked up by Walter Elsasser and Robert Rosen. Robert Rosen, who was one of the first major people, was trained by Nicholas Rashevsky, who was the father of mathematical biology. Rosen was a brilliant student, and they were dealing with the complex aspects of these things, which would include how these systems work. But then Maturana and Varela threw this massive spanner in the works called autopoiesis, self-production. If you had a jetliner that was autopoietic, it wouldn't just be organizing itself and maintaining itself. It would be gathering the materials to synthesize the parts of the jetliner and synthesizing the parts and getting rid of the parts that don't work while it's flying. That's what living systems do. They do something that is extraordinary. Even though you can find precursors to certain kinds of elements in the non-living world, living systems just do something that's quite weird. The more I looked at it, the weirder it looked. Then I realized there was a term because somebody leveled it at me while I was doing my PhD. This was a professor whose opinion I really valued. He said, "What you're talking about is vitalism. You're a vitalist." I went, "What in the hell is that?" That's when I basically bracketed that. I just said, can't go there, don't want to go there. But now that I'm older and I've been around the block a bit, I've looked at it. I've looked at the debate, and the term vitalist was devised by people who, with the best will in the world, saw that there were elements of the living state that could be modeled like machines. Certain things that machines did, we could see happening in the living world, and you could build an automaton duck or bird that could do certain things that ducks and birds do; superficially they looked the same.
[15:45] Pamela Lyon: Now, this view comes from Descartes. And Descartes was a genius. But he did make some mistakes. And one of the mistakes that he made possibly was about the nature of mind. Certainly he was wrong about animals. And there are perfectly good Catholic reasons for getting those things wrong when the Catholic Church is burning people like Giordano Bruno and even people like Galileo are trying to walk back elements of their thought because they're afraid of being burned at the stake as well. So there are really good reasons for doing what he did. Right from the beginning, there were people like Immanuel Kant who saw that there were things about living things that seemed very, very different from machines and from the non-living world. Those people came to be called, by the mechanists, vitalists. They were never a school. They were never a univocal anything. They were just people who came to similar seeming conclusions about certain things. And so Walter S. Elsasser, who was working on self-organizing complex systems and who influenced Stuart Kauffman, got labeled as a vitalist. Ludwig von Bertalanffy was called out by no less a person than Jacques Monod, the Nobel Prize winner for co-discovery of the operon; he called von Bertalanffy, a systems thinker, a vitalist. Francis Crick eviscerated Elsasser as a vitalist because he was saying, as a physicist, we actually can't get to what living systems are doing from the principles of physics as we know them. Francis Crick, the co-discoverer of DNA and Nobel laureate. This was a serious battle in the 20th century. And developmental biology was put back decades by this battle. Because anybody who raised their head and said there's something really quite strange going on here was called a vitalist. Jakob von Uexküll, the guy who gave us the idea of Umwelt, that every single organism that exists has a sensory world that is specific to it, was labeled a vitalist, and he was doing serious marine biology because he said there's this sensory world. Galileo bracketed everything sensory or qualitative right at the beginning of his physics because he didn't want to have to deal with it. We completely forgot that was a methodological decision. It doesn't mean that qualities don't exist, that sensing doesn't exist, but we forget. We make these methodological decisions, and we forget what we did. I'm in my 70s now, so I forget a lot. I can't even find my glasses half of the time. But in science, these things happen.
[20:05] Michael Levin: Good. That's what you want.
[20:09] Aastha Jain Simes: Mike, question for you. When you think about some of these self-organizing complex systems and the biogenic principles and cognition, does this also apply to, say, anthrobots and xenobots and cells and tissues and some of the work that you do?
[20:29] Michael Levin: I agree with everything that Pamela just said. There absolutely is some weird things going on with these systems. I take it weirdly further because we are now finding some of those same things happening in very minimal systems that nobody would call alive. I think that these vitalist/organicist thinkers are on the right track. I just think they stopped too early. I think they didn't follow their convictions far enough. I get yelled at way more by that crowd than I do by the molecular biologists. They don't like the cognitive stuff that we're saying about gene regulatory networks and whatnot, but they're actually less mad than the other folks because it's pretty important for a lot of people to keep a very strict separation between the majesty of living things and the machines, these boring sort of mechanical things. I think they should take those views and put them on steroids because I see this stuff going all the way down. We're just not very good at noticing it in some of these weird, really minimal systems. You have to pay a lot of attention to find it. Arthropods, sure. I don't think you have to be too imaginative to see it there. It hits you in the face when you're watching these things. The creative problem solving that they do, you see it. By the way, I first came across Pamela's stuff in 2006, the biogenic approach to cognition, which I thought was amazing. I would love to hear just a few of your favorite examples from the microbes, from the microbial war. You ran through some, but what are some of the amazing things that you found in that world?
[22:45] Pamela Lyon: Thank you very much. Why didn't you reach out in 2006? You didn't contact me until 2015.
[22:55] Michael Levin: I wasn't reaching out to almost anybody at that time. I had just started my lab. I was trying to get the bioelectric stuff off the ground. I was keeping fairly focused. And there was some time before we got big enough that I could start.
[23:17] Pamela Lyon: Yeah.
Michael Levin: Branching out.
[23:19] Pamela Lyon: Bacillus subtilis is now my current favorite microbe. In biofilms they are such a wonderful example of bioelectricity and how bioelectricity is used communicatively between individual living cells that come together as a collective. I love the fact that the centers of biofilms, when they start to starve, send out an electrical signal to the periphery, which basically says, stop eating and reproducing and let some of those nutrients diffuse into the center so we can eat too. They basically timeshare scarce nutrients because biofilms only form under declining nutrients. That was the original view of them. I'm not sure that's true anymore. I think that it's a good survival strategy no matter what you're doing. But bioelectricity is also used for different colonies of Bacillus subtilis to timeshare scarce nutrients between each other. But the first microbe that captured my heart and still has a huge, huge, huge role in my thinking is Myxococcus xanthus. It's a golden microbe. It is slow. It can't do much on its own. It is relentlessly social, and it's predatory. It is predatory to the point where C. elegans, which is so much bigger than these single cells, if they sense that Myxococcus is around, they will flee. They will go somewhere else. They avoid Myxococcus xanthus, because they are such profound ********. They collectively can prey on colonies of other bacteria, but they also can prey on larger things. They do so by coordinating their activity, surrounding them and then releasing enzymes that will break apart cell walls, which means the guts of the cells start breaking out. And C. elegans' cells are not as highly protected as bacterial cells are, so breaking apart a bacterial cell wall is a harder ask than something softer. When I found out that C. elegans were afraid of Myxococcus xanthus, I thought, "How is this even possible?" And what Myxococcus xanthus does when it feeds on a colony of something is engage in this oscillatory behavior, which basically sends nutrients back further down the line so that these cells are also feeding and reproduce.
[27:24] Pamela Lyon: Their colony gets larger. When there's a step down, we used to think it was starvation, but now it's just a step down in nutrients. When nutrients decrease, they start aggregating. They aggregate into fruiting bodies, and when they aggregate, in one of the best descriptions I ever read by Shimkitz, he said it's like the great animal migrations of the Serengeti: you've got tens of thousands of cells aggregating in herds that then come together to form these mounds that then differentiate. In order to form the mounds, all of them have to be competent to create these structures. Some of them aren't. Somehow the cells that are competent can transfer their competence across the cell membrane to make others competent. This is called complementation, the mere existence of which made me start questioning everything that I'd read about the evolution of cooperation. If this is happening at the microbial level, cooperation is really built in at the bottom. You can attribute this to kin selection to a certain extent, but in Bacillus subtilis cases they're much more diverse, much more heterogeneous than kin selection by itself would suggest when looking at biofilms grown from wild populations. So there's a lot more going on in the living world than we know. When they build these structures, at the top you have myxospores. There's a lot of differentiation going on. Sporulation is massively energetically costly. A certain proportion of the cells will lyse themselves to provide nutrients to keep this going. After the fruiting bodies are formed, there are cells that remain as sentinels to keep fruiting bodies free from predation. Now I will tell you about something that Jim Shapiro described as the most interesting paper that nobody's ever read, and that is by Martin Dworkin, who was the one who tamed Myxococcus xanthus for laboratory study in the first place. He was just messing around in the lab and seeing what might happen if you do this. They put inert glass beads, no chemical signal whatsoever, onto a medium in which Myxococcus xanthus was growing and moving. Dworkin — this was a 1983 paper — and his colleagues observed Myxococcus xanthus migrating to the glass beads, crawling all over them, and then leaving. What on earth were they migrating toward? There was no chemical signal. They did this experiment over and over again to make sure that there was no chemical signal. On that basis, Dworkin mistakenly assumed that what Myxococcus xanthus does isn't chemotaxis. They do chemotax. Based on this experiment, he said something else is going on here. The only thing that he could say was that it probably has something to do with deformation of the medium; it might be a mechanical signal. That means Myxococcus xanthus has more sensory modalities that mean something to it than we had any idea. So its umwelt is richer than we knew. When you put this entire picture together, the umwelt of Myxococcus xanthus is just extraordinarily complex, because there are all these signals that do things. There are all these signals, many of which we still don't know what they do. That's my favorite microbe, even today, even though we don't know how bioelectric it is yet. I'm sure it is.
[31:29] Pamela Lyon: I'm sure that the collective fields of Myxococcus xanthus are really important in its collective behavior. I have no doubt of it. And all of that's due to your work, my dear, which changed my view of everything. I used to be very dedicated to the view that we can't go down to the genetic level. That's just taking it too far down. Because I thought if we do that, then we are vitalists, basically. I had been trained not to go there. Once I had seen evidence because I came across Ladislav Kovac's work on cognitive biology, that led me to Brian Goodwin. That's a completely different area, and we can talk about that. But that convinced me that you had to be right, and now I have absolutely no doubt that you're right that it goes all the way down to the molecular level. We have no idea how any of this works, why it does this. But I think that we have to look to self-production. Autopoiesis. Autopoiesis demands it. Once you get a system that is self-producing, not just self-organizing or self-maintaining, I don't think that we can do without it. I think these kinds of scare-quoted 'cognitive' processes — but it's cognition all the way down. It's cognition all the way down. The sooner we realize that, the brakes will suddenly be off a lot of science, and people won't have to dance around and caveat this and that because they're worried about the V word. Let's just put it to the side and say we'll deal with this later, because it's actually not important. What's important is understanding what's going on in these systems. That's what's important.
[35:36] Michael Levin: Your point about the beads and no chemical gradient. We had this — Nirosha Murugan in my group did this — we had this Physarum, the slime mold. You put it in the center of a Petri dish. We have these tiny glass discs. I think they're about 10 milligrams a piece. They're just glass. There's no chemical on them. And you put three over there, you put one over here, and it grows out a little bit for about four hours. It tugs on the agar. And it turns out that all that time, it's building up information about its environment. Afterwards, boom, it just shoots for the three consistently. No chemicals, no gradients; it's just from that, and the sensitivity is insane because we did all kinds of stuff to try to fool it. So you can stack the three on top of each other and you can rock it. We put it on a rocker and we did all these things to confuse it, but it's sensing the strain angle on the order of 10 centimeters; it's this large thing and it has this idea of what's going on around it. It's amazing, this idea of the Umwelt and these things pulling together a representation of what's going on before it acts, long before it actually does anything.
[37:06] Pamela Lyon: One more thing about Myxococcus xanthus: this is another rarely cited paper that I absolutely love—there is evidence to suggest that Myxococcus xanthus will extrude cAMP as a signal to draw E. coli into its orbit because it's too slow to follow the flagellated bacteria. It's too slow. So it needs to lure them in close so it can then surround them, lyse them and eat them. A microbe emitting a biologically relevant chemical as a signal to draw in prey? Give me a break. That's why the Xenobots are so important, because they show really clearly what's built into the system. It's not all in the genome. The genome's the hardware; it's the software that runs the damn thing, which is why all of the work that you've done on bioelectricity is so important. The first time I heard about it and I heard you explain it, I just went, my God, this guy's gonna win a Nobel Prize. I knew him when. I know you're not supposed to even think that way, but seriously, the whole bioelectric field that's opened up, with you basically sticking your neck out on the chopping block and letting people shoot arrows at you and try and cut it off and all that stuff, was absolutely fundamental to our understanding of a really fundamental biological capacity. I am so grateful for that work. Thank you.
[39:39] Michael Levin: That's amazing. I got my inspiration from people like Harold Burr, who in the 1930s had nothing but a good volt meter and a crystal ball. There are a few other examples of somebody who sees the future so clearly with very limited actual ability to really prove any of this stuff. He just measured things. He measured trees and embryos and psychiatric patients' skin potentials. On the basis of all of this, in his little book he laid out pretty much—we're probably no more than 3/4 through all of the things that he found that we've actually now documented. He was visualizing all these things many decades into the future. There were others, people like Lionel Jaffe and all of his folks. They caught so much crap in their day from all the molecular biologists. It was amazing.
[40:52] Pamela Lyon: Thank you for that example, because what you've just described is that the creativity of the human mind interacting with nature is essential to our understanding.
[41:10] Michael Levin: That's exactly right. I'll ask you towards the end what you think the future of all this is; to me, one of the biggest open areas is the creativity aspect. Because we can see there are parts of what these things are doing that are algorithmic, that are "turn the crank" — the process: you turn the crank, you get your answer. Then there are other parts that there is no better way to explain than some kind of creative process. It's not algorithmic. There are no steps that you can do to solve some of these things. I think it's baked into the bottom of what we're all interested in. To me, that's the big open frontier.
[41:55] Pamela Lyon: I think that you've been dancing around this a bit, and I've been dancing around it a bit. And I think that when you look at what it takes to be alive, to maintain life, that whole self-production thing, autopoiesis plus self-organization and self-maintenance. This is why we have so many responses to stress. Stress responses are, I think, on one level, the engines of creativity, because you have to get out of this problem. Now, microbes, once again, provide a very interesting entree into this. There are these factors called ECFs, which, unfortunately, it's been so long since I've written it out that I can't remember. It's environmental something factor. We don't know exactly what they do, but they seem to be wild cards that come into play when an overwhelming stressor is encountered. It is believed that they provide some degrees of freedom beyond the standard stress responses that are built into the genome and the hardware and software of the cell. They're wild cards. That may just be because we don't know what they do and they seem to activate a lot under stress. But they're an additional tool for dealing with stress. I wrote a chapter years ago that nobody read, but that's okay. That happens in chapters in books. They don't get read. Papers get read. I was advocating for a co-evolution hypothesis of cognitive capacity and stress response. These two things ratcheted each other. They scaffolded each other. What convinced me of this, first of all, was that in microbes, the most sophisticated, cognitively demanding behavior they engage in is stress related. It's all in anticipation of stressors that are existentially challenging. When we started recognizing in the early part of this century that the brain really wasn't an immune privileged organ, that immune system elements like cytokines were working everywhere in the brain, not only activating under stress but involved in normal memory, normal learning, normal cognitive activity — what do you mean tumor necrosis factor alpha is involved in the scaling of synapses that lay down a memory? What in the hell's that? What do you mean IL-1B and IL-6, which are notoriously heavy-duty actors under stress conditions, are involved in normal cognition? That doesn't make any sense unless you start looking at what it takes for something to live, and all of the gigantic things that can go wrong. Some of them are predictable, but so many of them aren't. So I think that stress responses are the source of our creativity. I think that they had to be in the very beginning.
[46:40] Pamela Lyon: How on earth did we get photosynthesis when exposure to UV radiation breaks DNA? Think about it. You have to be close enough to the sun for long enough to photosynthesize, but you have to be able to protect yourself against DNA breakage, which is the one thing that's going to make sure that you can't reproduce effectively or reliably. The evolution of circadian rhythmicity, of partitioning physiology so that it can be done at times that aren't dangerous—there's so much going on early on in the origin of life. When in the hell do ion channels show up? When do they show up and how do they show up? Okay, so we're pretty clear now that stromatolites, fossil stromatolites in Western Australia and other parts of the world date back to 3.85 billion years. What in the hell are we talking about when there's basement rock off the coast of Greenland, your homeland, that indicates organic life at 4.2 billion years? What on Earth is going on here? Because 4.2 billion years, we are being pummeled by stuff from space. There is water on this planet that is older than the planet. It's older than this planet. How did that get there? It's all coming in from space. It's all coming in from everywhere. And we've got Haitian conditions, which means Haiti's not nice, not really peaceful and easily synthesizing life. What on earth is going on here? We haven't got a clue. But the good news is we haven't got a clue. There's this entire area that really needs exploration. Until we can get a handle on what the basic kinds of functionality are that you need to keep the show on the road, we can't even begin to trace what might have been the original cognitive contribution to this state. Until we can get there, we're not going to understand the rest of it. Because things had to be sensed pretty damn quickly in order for these things to be partitioned, for circadian rhythms to evolve, apprehending that proximity to a particular kind of wavelength is going to damage you such that you need to move away from it when it's really, really high, or you've got to have other protective modes of being. There is a program already extant on the planet called Major Transitions in the Evolution of Cognition, but most of that starts at nervous systems. We're not talking about the basement level of all this stuff. Until we understand the basement level and the kinds of transitions that we're looking at, the development of mechanosensitive ion channels is really critical in that.
[51:24] Aastha Jain Simes: Mike, I'm curious, how do you think some of this ties into your work on stress sharing as a cognitive glue?
[51:33] Michael Levin: Pamela, if you've seen this, we have some stuff. I agree 100%: stress is critical. We now have a paper on stress sharing, in particular a mechanism for binding competent subunits into a larger whole.
[51:50] Pamela Lyon: Can you send that to me?
[51:51] Michael Levin: I will. I absolutely will. But we also have experimental work, which is not published yet; we're still in the middle of it, looking at how evolution has reused some of the molecules that initially were about failures to protect your DNA, protein misfolding, all of those things. They got repurposed for these large-scale things like my anatomy isn't right. And there's some amazing stuff on the role of stress in binding these anatomical intelligences that are trying to keep their navigation in morphospace correct. I 100% agree with you, even that you could take all the way down and I say these things and then people laugh, geometric frustration is...
[52:43] Pamela Lyon: Geometric frustration.
[52:45] Michael Levin: It's a term. That's a term from physics and math: geometric frustration. I keep saying it's real frustration. It's actual, real frustration. What they have is these — if you can think about tiling the plane with certain shapes, eventually you can work this off into a corner where it doesn't fit. If you make it fit, then somewhere else it won't fit. Or when you have magnetic domains or these Ising models where they try to make everybody aligned but somebody's not aligned. If you fix that one, it propagates somewhere else. This inability to get all your parts onto the same story of what should be happening — to convince all your parts to work towards the same, to buy into the same vision of what the hell you should be doing — is actual internal frustration. It drives a lot of downstream steps. I think that's exactly what evolution capitalized on in the kinds of things that we think of, where we're internally incoherent. I think it goes all the way, as we said, all the way down.
[53:56] Pamela Lyon: That's a wonderful concept. I love it.
[54:00] Michael Levin: Yeah, Richard Watson and I've been doing a lot on this now, thinking about what does it take to get your parts aligned? to bend their Umwelt and their landscape so that they do the thing that they want to do, but it's serving this global purpose in another space that they can't visualize, but you're bending their space in that way.
[54:27] Pamela Lyon: Nitric oxide is a really important stress product that became the dominant second messenger in multicellular life.
[54:46] Michael Levin: That's a good point. I should take a look. We haven't been tracking it. Don't they have nitrous oxide, nitric oxide sensors, fluorescent reporters?
[54:56] Pamela Lyon: They do. The work that you want to look at is coming out of Nicole King's lab. She had a PhD student. They call him Dr. No because of his focus on this. So some of the really interesting work is coming out of her lab.
[55:22] Michael Levin: Interesting. Okay. Yeah, I'll check into that. Cool.
[55:25] Pamela Lyon: Yeah.
[55:28] Aastha Jain Simes: One thing I want to make explicit is that when both of you speak of cognition all the way down, it's not just that it implies that cognition is not just associated with the brain, which is how most people think about cognition. I know both of you have co-authored a paper on this as well. What do you think are some of the implications when you start thinking about cognition so differently, where it's not just about the brain, it's all the way down? You have cognition in some of these microbes and the beautiful examples that you gave, Pamela. What do you think are some of the implications when we start thinking about it in the future?
[56:07] Pamela Lyon: Mike, do you want to go first?
[56:09] Michael Levin: No, please, you go.
[56:12] Pamela Lyon: As I alluded to earlier, there are all kinds of things that we are finding at the molecular level that people recognize. They try to use language, they try to avoid language. That is cognitive because they know they'll get jumped on. But what they're actually looking at are cognitive phenomena. If they could recognize that it is cognitive phenomena, not like cognition, but in a bare sense — this is seriously cognitive — they can ask different questions, investigate things in different ways, and develop theories that they couldn't develop before. Biology has always been criticized as basically descriptive, and you're just describing this phenomenon, that phenomenon; it's like bean counting. There's not enough glue between the beans that will hold them together. As Mike said in one of the titles of his papers, there is a cognitive glue. If we can start recognizing that these are phenomena that we recognize in ourselves, that we can see them across the spectrum of life... Whether they're in viruses or not depends upon where you think viruses came from. Whether viruses were from cells that then stripped everything down and became just optimized reproducing machines — there are multiple views on where viruses come from. So I'm bracketing viruses entirely. If you can recognize these kinds of phenomena — memory phenomena, non-associative learning, maybe even associative learning — if amoeba can do it, maybe microbes can too. Maybe bacteria can do it too. We don't know. Decision making, error correction, those sorts of things. If that is happening at the molecular level, at the level of what we have always called physiology, which is what Brian Goodwin saw when he was working in developmental biology in the 1970s, he was looking at enzymatic action. That's what he was seeing. He put a label to it, and I think it probably ruined his career. He had a career anyway, because he had a brain the size of a planet, but I think he was living that down for a very long time. I haven't been able to talk to anybody who can confirm that yet; I'm not a biographer. Waddington was dead; he didn't have his mentor and his defender there. He took things a step further than even Waddington would've gone, and I think he just got chopped off at the knees. That's what I think happened. We can ask different questions. We can understand related phenomena much better than we can now, which will allow us to start developing the theories that biology really needs.
[1:00:02] Aastha Jain Simes: Mike, what are your thoughts?
[1:00:04] Michael Levin: I think what you just said is critical because a lot of people think that this kind of talk is poetry. They say, "Oh, well, sure you can say that, but what's the point?" The whole point is, as Pamela just said, the judge of the criteria for all of this is that it has to lead to new discoveries, it has to lead to new research programs, and it has to lead to new capabilities. Otherwise, there's no point. All the verbiage, all the terminology — what does it enable? What does the old view constrain? What does this view enable you to do that you couldn't do before? As an example, gene regulatory networks, just a set of a handful of chemicals turning each other on and off. You look at these things and you say they're deterministic, they look pretty dumb and mechanical. They're important for everything: health, disease, development, evolution. When you make that assumption, it has a very specific implication. It means that if you want to change the way they behave, you're going to use dynamical systems theory and you're going to have to rewire the thing. Gene therapy: you're adding nodes, removing nodes, changing promoters; physically rewiring the hardware is your only option. If you instead, as we did a few years ago, say that you can't just assume anything is dumb and mechanical any more than you can just assume there's a majestic spirit in every rock. Maybe there is, but you have to show that that's helpful in some fashion. If you do the experiment and you say we're not going to assume, we're going to do actual experiments. It turns out these things can do six different kinds of learning, including Pavlovian conditioning. You don't need the cell nucleus, you don't need any of that. From the math alone, they can do it. That opens the opportunity for something completely different, which is drug conditioning. That has huge implications for medicine and the way you use [drugs]. We're now doing all this with real cells. If you don't assume where things are on that spectrum of persuadability but do the experiments, whole new sets of tools open up to you. All of this that neuroscience has come up with — not only tools and techniques, but some amazing formalisms and tools for thinking about how collectives and multi-scale controls work — gives you permission to take those tools and apply them to things that aren't brains. I don't think neuroscience is about neurons in any case. If you don't think that's the case, you will never be motivated to do that. You will never try these crazy things. You will never find the new biology. I think it's extremely liberating because by dissolving these old categories, you get to port tools and test them out and see what you can get. This is why we put embryos on memory blockers and hallucinogens; this is how you discover novel regenerative applications by asking, what does the system know? What does it remember? What does it believe in? How do I convince it to do something different? Not force it at the molecular level and get all kinds of unexpected side effects as it tries to fight you, but actually get buy-in from the system. I think it's going to be massively important in many practical applications. My hope is that Chris Fields has the saying, "only technology settles arguments." We can say this stuff, and people will complain, but when the applications start coming forward, eventually it will all tick over and everybody will say, "Yeah, we knew it all along." I'm kind of seeing this on the horizon already. Eventually people will say, "Everybody always knew this. What are you talking about? Who needed you to say this? Everybody already knew this." It just flips over to that, and then that's fine.
[1:04:16] Aastha Jain Simes: I just want to stress how groundbreaking some of this work is, because for most people, when they think of cognition, they think of neuroscience and they think of the brain, but when you start thinking of cognition in these different systems or cells, tissues, microbes, all the way down, it opens up such a different perspective. You can start, as you said, thinking of new theories. You can start thinking of new applications and new experiments to run, which Mike has done some beautiful work around. And Pamela, you have some amazing work too. One question here is also: how do you start thinking about cognition in artificial intelligence?
[1:05:02] Pamela Lyon: If you guys have thought about that before. That's not my area. I really can't do that. I have to stay with the living. I don't even go there. So Mike will have to speak there. But I will say that if the view that Mike and I are advocating became accepted to the point where it was taught in schools to children, we would not be having the trouble we are having right now dealing with climate change. We would understand so much more about the commonality of everything that lives and how much we rely on other living things. We wouldn't be having this trouble. So I will hand over AI to Mike.
[1:06:07] Michael Levin: As far as I can tell, the architecture of life is what's critical about life. Partially, it's the creative aspect of never being sure of what your memories mean and having to constantly reinterpret your own engrams because you don't have access to the past. What you have access to are the traces that passed you, the messages that previously left for you, and you need to reinterpret them as best as you can. Our computer technology is exactly the opposite. We have abstraction layers, error correction, and redundancy codes to make sure that the data stays fixed. The data is fixed because we are interpreting it, and you don't want your registers floating off and then you don't know what's going on. So life is not like that at all. From that perspective, all of the things that we've made are not the right architecture at all. However, one thing I've now realized is that if we don't even know what the damn bubble sort is doing, which is six lines of code, deterministic, transparent, and we have now found that it does all these weird side quests while it's sorting the numbers, it also does all this other weird ****. If we don't even know what that does, we should be extremely cautious and have some humility around the fact that when you make something, just because you made it, that does not mean you know what it is and what it does. When they say artificial intelligence, I don't believe we make intelligence synthetically any more than we make it biologically. I think what we make are pointers or interfaces that pull down certain patterns. Sometimes if you're a good engineer, you'll pull down the thing you want. But very often you pull down a bunch of other stuff that you did not expect, not just complexity, not just unpredictability, but actual cognition, emergent cognition. We can look at these architectures and say, "They don't even have a homeostatic feedback loop." They're certainly not pulling down anything remotely like a human mind or probably any kind of biological mind. But that doesn't mean that we're not currently fishing in some corners of that space that maybe have never been embodied before, probably not on Earth, but maybe nowhere else. I think we need a lot of caution about that. People have said—I've had debates with people. Somebody who writes these AIs will say, "I make them. I know what they do; it's just linear algebra." They say, first of all, you don't even know what bubble sort does. Second of all, it's just linear algebra and you're just chemistry or quantum foam or something, but of course you're not. Why are we fairly comfortable to say that the story of biochemistry is not the story of the human mind, but somehow think that the story of algorithms and Turing machines is the story of these things that you're now making? Those are our abstractions with the limitations of the formal model. They're not necessarily limitations of the thing itself. That's my story on AI. I think we have no idea what we're making — very, very little.
[1:09:31] Pamela Lyon: Well, you've just scared me to death, Mike.
[1:09:34] Michael Levin: That's my job. I often apologize in advance. A couple of weeks ago, Mark Solms had me at the Psychoanalytic Association. I was on Zoom giving a talk to the psychoanalysts. That's the first thing I said, because they wanted to hear about it. I said, I'm sorry, because what I'm about to tell you, I'm sure it's not going to make you feel better. You're going to have to put your thinking cap on and ask yourself how you're going to offer your services, which are very needed, to a very wide range of clients in the future, things you can't even imagine: hybrids and cyborgs and novel beings of a variety that Darwin's "endless forms most beautiful" didn't even begin to scratch the surface.
[1:10:31] Aastha Jain Simes: Pamela, I know you approach this work coming from Buddhist philosophy, and you still do a lot of work with Buddhist monks and Eastern philosophy. It strikes me that some views on cognition are more accepted by Eastern philosophy. It may not be from a very scientific perspective, but I think people embody that in many ways. Do you think that's true? Could you explain that? And why do you think the East accepts it more compared to the West?
[1:11:09] Pamela Lyon: That's a great question. Thank you. You have to understand first and foremost that I am a heretic. When I undertook this PhD in this project, I had to risk losing my faith, but I didn't. But unfortunately, I turned up things that my particular tradition doesn't accept. They don't accept plants, bacteria, the molecular workings of the body as being cognitive. So there's an entire layer of what I do that is not acceptable to the Buddhist view as it stands right now. It was more labile in the very early years of Buddhism, but in certain Mahayana traditions it really hardened. But the idea that "living"—because they use the term "living"—all living beings, which meant that plants had to be put in another category that they weren't actually alive. So in my lifetime, I have had educated, not Western educated, but educated, highly learned geshes in the Tibetan tradition sit up on the teaching throne and describe how plants aren't alive. I've had to listen to that and try to correct them, but they can't be corrected. Or it's like, let's just not talk about it. But once the Buddhists decided that living things were minded creatures, they were cognizers. There was something about being alive that they saw as associated with cognition, which is why they could see it in insects and really small things like that. There was no problem in seeing cognition in an ant and a human. That came out of the incredibly rich tradition of classical India, of the Rig Veda and classical Hinduism. And then the reactions to classical Hinduism by Mahavira, who was the founder of the Jains, and Buddha, who founded Buddhism. They reacted against those pictures with their empirical experiences. They did an investigation and they reported on what they found, rather than relying on scripture. But all of that came out of a very rich tradition in India.
[1:14:11] Pamela Lyon: And that spread everywhere, as William Dalrymple mentions in his latest book, "The Golden Road," is that Buddhism was probably one of the greatest exports that India ever had. It went to the Roman Empire, to Greece, to China, to Southeast Asia, and it profoundly changed the way people think. People in the East basically have that view because of classical Hinduism and the traditions that arose within that tradition. One of my favorite stories out of this is in 1983. Imanishi, the eminent Japanese primatologist, had been publishing work on Japanese macaques saying, we think we're seeing culture here. We think we're seeing cultural transmission in Japanese macaques and certain ways of doing things that are different between different troops. A scientist, a paleontologist from the UK, was so outraged by this. He flew all the way to Japan and tried to talk Imanishi out of this, that this is insane talk, can't possibly be true. Couldn't convince him. Came back, was given three pages in Nature to trash Imanishi's entire approach to primatology as anti-Darwinian, childish, all this sort of stuff. And we don't know about any culture in animals now, do we? We have culture everywhere in animals. We have that in birds. We have it in cetaceans. We have it in all kinds of animals. Nature gave this guy three pages to trash the entire idea. What was the difference between Imanishi and Halsted? Imanishi came from a Buddhist culture, and Halsted came from Judeo-Christianity. Good question.
[1:17:13] Aastha Jain Simes: I'm Jain, and I grew up with Jainism, so it was part of the reason I asked the question. It's intuitively easy for me to grasp. I know Mike, you have to run, so.
[1:17:24] Michael Levin: I will run in a minute.
[1:17:25] Pamela Lyon: Mahavira believed in plant cognition. The Jains believe in plant cognition, but the Buddhists did not.
[1:17:34] Aastha Jain Simes: That's right, yeah.
[1:17:36] Michael Levin: A quick anecdote along those lines. I recently, I've talked to people at the Vatican, I've talked to rabbis about it, and you get what you expect. But I gave a talk once at a, I'm not going to call him out by name, but it was an Indic tradition. And I gave him a talk and I thought these guys are going to love this because finally, there's a scientist that's saying some of this stuff. Well, they flipped out because they were so mad. They were so mad that I had the non-living things on my spectrum, and they kept saying, No, absolutely not. They said, It's dead matter. I said, What? There is no dead matter. There's just lazy observers and things like this. What makes you think that? And they said, No, absolutely not. They said, Aren't you the guys that talk about the soul incarnating in bodies? And they said, Well, yes, living bodies. And I said, Well, who are you to tell these spirits if you make a beautiful robotic embodiment that somebody's not going to come live in it? I've heard people talk about you might get reborn as a block of wood and all this stuff. And you're telling me now that they're not going to come live in these synthetic embodiments. I say, absolutely not. They were so incredibly attached to this sharp separation. They wanted the dead matter, and then they wanted all this awesome magical stuff. And there was no getting past it.
[1:19:13] Aastha Jain Simes: Fascinating. All right, Mike, any last questions, thoughts.
[1:19:18] Michael Levin: Thank you so much, Pamela. It is always delightful to see you. You're an inspiration. I've said it before. The stuff you've done, way ahead of the curve of anybody in this area. Just amazing.
[1:19:33] Pamela Lyon: Thank you for having me. It's been an absolute pleasure. I clearly love talking about this stuff.
[1:19:42] Aastha Jain Simes: No, this was wonderful. Thank you so, much.
