Watch Episode Here

Listen to Episode Here

Show Notes

This is a ~2 hours 4 minute discussion among contributors to the Platonic Space Symposium (

CHAPTERS:

(00:00) Strong anticipation analogy

(08:16) Relational Platonic patterns

(25:06) Finite brains, infinite grammar

(40:05) Beauty and compression

(51:17) Searching for grammar

(01:01:58) Patterns and selfhood

(01:14:44) Patterns claiming territory

(01:29:28) Markets shape selves

(01:43:57) Scaling collective selves

(01:55:47) Testing Platonic feedback

PRODUCED BY:

https://aipodcast.ing

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.

Main Episode

[00:00] Ben: There was something that I wanted to talk about, if everyone's okay with that. And thanks very much for having us here. Nice to meet all of you. Let me go ahead and share something here. I want to show you all a couple of slides. Not very long at all.

[00:14] Elliot Murphy: Go for it.

[00:15] Ben: But if everyone's seeing this stuff about strong anticipation, so basically there's this phenomenon I want to share with you all that I think is structurally analogous to Platonic space and Gresham, but something that happens in a totally mundane, familiar way. So strong anticipation is a process where anticipatory behavior emerges via the coupling of elements within a system rather than through some kind of internal model. That's like doing computations and making forecasts. So a classic example of this is a baseball fielder catching a baseball. One way that they might try to predict where the baseball's going to go is to have some internal model of the physics that says this is the baseball's computed trajectory. So that's where we need to run to catch the ball. And basically the problem with this is that that's really very hard to do in real life. There's a lot of factors to consider. It's hard to understand what that internal machinery would look like. So another much simpler, much more natural way for a baseball player to catch a baseball is to visually couple to the ball, where if you run in a way that maintains a certain visual relationship with the ball, you just hold your visual relationship with it constant, doing whatever actions are necessary to make that happen. You will end up at the place where the ball is going to end up. And this is anticipatory behavior because the only stable way to maintain that relationship is to be basically phase shifted ahead of the ball because then delayed feedback because it takes you time to see what the ball is doing and to figure out how that should update your actions. That's going to then push your phase back so you end up coupled to the ball. So the only stable way to track the ball and end up where it's going to go is to adopt an anticipatory relationship with respect to it. So it's this much more computationally simple, perceptually natural way of trying to catch a ball. And what this produces is a process that in dynamic systems is called enslavement. There's way too much text on this cause Chachi BT made this and it likes to go kind of crazy with it. But if you just look at this fourth point here on enslavement, which it has misspelled, basically what happens is this term gets used where basically the degrees of freedom that the person has as to how they're going to move their body end up being governed by the trajectory of the baseball. If you just know what the baseball's going to do, you can just see the person's movement as a function of that. So they don't like literally lose their free will. But effectively you can think of them as just being controlled by this other pattern that the so the ball controls the person, but the person of course cannot control the ball, only their relationship with it. And so my argument is that if baseballs were invisible, then fielding would look like pattern ingression. So here's the symmetry. Platonic space. You have whatever patterns like prime numbers, these can influence the physical world in the form of cicada cycles and so forth. But the physical world cannot influence Platonic space back. So Platonic space, you know, controls the physical world, but not vice versa. And this creates this apparent dualism. But this is analogous to the trajectory of the fly ball. You have this pattern that enslaves the person who controls what the person does, but the person cannot influence the fly ball back. And so, you know, if the ball was invisible, it would all look the same. We have this evolving behavior in the physical world. And we're like, why is that? There must be some kind of external pattern that we can't quite perceive but somehow is affecting our behavior. The strong, the anticipation version of this with the fly ball and the baseball player basically goes in the opposite direction of dualism though, where it says that this stuff happens precisely because the organism and the environment are coupled parts of one system. So the kind of dynamic system perspective that this comes from the work of people like Michael Turvy and Esther Phelan, explicitly rejects a dualism between organism and environment.

[04:14] Ben: Rather than seeing these as separate systems, we can see them as coupled old elements of a single larger system. And that's how you can get these kinds of natural dynamics. We get this anticipatory behavior. And so basically, that's exactly this argument where Platonic spaces influence the physical world, the physical world can't influence them back, which is structurally identical to a baseball player tracking a fly ball. The ball enslaves the player's trajectory, but in the baseball case there's no dualism. The player's anticipation isn't some kind of internal model predicting an external world; rather, it's an emergent property of the coupling. It's precisely because they're all in the same system that this is able to happen. And so the same logic, because this appears to be structurally similar, may apply to Platonic space. Pattern ingression might be, in fact, an example of dynamic enslavement occurring between coupled elements of a single larger system. I've been blogging about this on my website, interesting essays where I've been writing a whole bunch of things about strong anticipation lately. So that's just been something on my mind because I think it's been super interesting to observe something in just a normal, mundane thing that seems to have the same dynamic dynamics as pattern aggression. So I wanted to share all of that with you and get your thoughts. Ask a question. Well, and for just a quick comment, the first point of that list, I mean, some people believe that that isn't my model. So just to say that I do think it's a bi-directional interface. But my question is this: When you say you couldn't see the baseball, you mean we as third-person observers couldn't see it? Or you mean the guy trying to catch it can't see it? Yeah, if we were scientists studying the system and we had no way to directly perceive a baseball, but we can just see if this guy keeps moving this way as if he's trying to catch a baseball. But we don't know really what baseballs are, where they come from. We might say there's some kind of baseball realm or there's some kind of pattern that the guy's body is coupled to, even though we don't really know how that happens. I mean, fair enough, but he needs, I mean, the ball has to actually be physically there. Maybe, you know, maybe he has a different visual system or something, but physically the ball has to be there for him to see it. Even if we don't, even if we don't see it as third-person observers. Yes, perceptual coupling is absolutely key to how this process works. And so if this analogy is going to hold, there would have to be this whole big empirical question about how exactly do we perceptually couple to these Platonic patterns. And regarding the question of the bi-directional influence, there is a kind of bi-directional influence in the strong anticipation baseball catching where you can't control the trajectory of the ball, but you can control your relationship with the ball. So in that sense you can exert a direct effect on what you're seeing about the ball, which is, in fact, key to how the catch the anticipatory behavior occurs by maintaining a certain relationship. So in that sense, there's a very clear bi-directional effect, which is actually crucial to producing the whole effect. I'll say on the case, on the bi-directional issue, I mean we've spoken on that. I'm not 100% sure as to why there couldn't or wouldn't be an option for influence the other direction. I think there's a lot of implications if you suggest that there is not, if we go that road. But there's a lot of important stuff here, especially about the big issues of whether or not we're really talking about dualism, is the best approach generally, or idealism, or what I think there's a lot of, I guess maybe difference of opinion probably or lack of clarity amongst the many people involved. So I'd love to hear what people's thoughts are on these sort of big issues.

[08:16] Elliot Murphy: I guess for me, I agree. The functional force of your example is really useful to me because it kind of helps tease apart and interrogate that notion of what is a pattern to begin with. What do we count as a possible pattern? If we're going to talk about all possible mathematical cues, right, that are guiding some kind of physical instantiation, which of those mathematical structures are gonna be counted as platonic, quasi platonic patterns and which ones are gonna be potentially not discounted, right? That's kind of a boundary dispute that I don't personally, I don't have a good intuition about.

[08:54] Ben: Yeah, for sure. I mean, the whole point of basically wanting to share this in this conversation is because I don't really know where to go with this in terms of the Platonic, how to answer these questions. Like you have something that very much looks analogous, then I'm like, OK, I don't know what the next step is. So in the obvious case of numbers, right, as a prototypical Platonic example, right, the perception would be my ability to think about them or to calculate or things like that. So instead of visual perception, it would be like a conceptual perception, right? And then maybe how different people use different visual or cognitive metaphors to do calculations, right? So it could be like different ways to shift your relation to the ball here would be using different calculation techniques in order to see that. And then it seems like in that case, the heavy work is done by this, how I orient myself or how I relate myself to the numbers and to the ball. And in some sense, people could say that the way I think about numbers actually creates numbers. So they're not pre-existing. It's my perception of them or my thinking about them that gives them existence. And so the way I relate to the ball actually is the crucial piece that makes it bi directional rather than ingression, right. So I guess that would be the analogy. And have any of you seen what Lisa Feldman Barrett's put out about relational realism? Yeah, so she's a.

[10:27] Elliot Murphy: I watched the video about that the other day. Where she's saying that you can go ahead and summarize it, but it's something to do with the fact that all of these traditionally realistic claims can only be grounded through not just the ball itself, but the relationship between my hand and the ball, right? It only becomes real through the relation, kind of similar to what Mike Gilchrist has been saying, right, about relations and relative pre-existing each other, that type of thing. But yeah, go ahead.

[10:54] Ben: Yeah, yeah, yeah. No, exactly. I mean, she's a neuroscientist and she's basically two levels of the argument. One is in the context of neuroscience specifically, the meaning of internal signals is entirely relational. It's like you get one signal, that means this. Another signal means that, and the brain's job is to add it all up. Rather, no individual signal means anything except in the context of the larger ensemble. And by extension, this argument that that's how everything works. It's always relational realism everywhere, all around. And so that may well be what we're dealing with with Platonic space if there's some kind of relational coupling, maybe as a generalization of perceptual coupling. But the question is how to get that to cash out. But I think it's super interesting. I'll share a link in the chat. I mean, one thing to say about one particular version of this model that I like is that I don't see this as here we are as primarily physical beings and then we somehow connect to these forms once in a while. I think we're forms too. So the whole point is that that space is extremely wide in the range of patterns that inhabit it. And so you get simple, low, simple things that may not be able to be changed, like the value of E and things like that. That's not going to change. And then you have these much more complex higher agency forms, which are us. And so in the case, I think, as Elliot was saying, in the case that you are perceiving mathematical structures, that may be some sort of direct lateral interaction, right? This, to, I don't know, resonance or some kind of lateral interaction between the forms that doesn't have to go through the physical interface, unlike most of the perception that we do here. So that's another piece of this. Yeah, all of this to me, all of this is fundamentally an interaction between these patterns. And then some of it also reflects into this world that we can see in third person. I guess when you think about Platonic space, what sort of scope do you give it in terms of, you know, is it numbers? Is it more than that? Is it every single independent conceptual object, like the minimal form of it or the reduced form of it? Or is it a narrower set because they're already sort of even Michael with morphogenic space is clearly a subset of the broader possibility space, just like assembly space effectively of chemical analogue. And if you've got Platonic space ingressing, I sort of sit with Michael that it has to be enmeshed in one integrated map. And that the ontology or sort of the metaphysics, I think, is quite good, or at least has history, is sort of a dual aspect monism where this realm of ideas or this Platonic space is kind of constituted as information. Like you can't really say what's the currency of it, at least in some coarse way. And I think it's how you compose those things together. So what does Platonic space do in this concept? Or it bounds the possibility and you get upward feedback.

[14:32] Ben: And it's sort of like Ian McGilchrist of right brain, left brain. The whole is a distinct thing, and then the parts kind of try to build up to the whole, and we get a thin wedge of that whole space here in our bodies. But there are other types of information that are accessible to us, like imagination, where things don't have to be physically possible, or those ideas also sitting in your form of Platonic space, and are they ingressing in a minimal way? And just a thought, how real are those things in that context? No clue. But let me show you one other thing, which is a totally made-up, evidence-free cosmology that I just think is kind of suggestive, so that I think will just maybe help indicate what a perspective could look like here. So this is another lovely AI-made image. And basically what this is sort of presenting is a picture of the world in which all of the sciences, chemistry, physics, and so on, are not natural kinds. They're just ways that humans kind of break up the world. And they're sort of within these patterns of constraints that end up defining what kinds of patterns can form and what kinds of patterns won't form, which in mathematics would be like your axioms that basically end up defining all the rules of logic as to what kinds of things can actually end up being true and what kinds of things can't be, what kind of patterns can perpetuate and what can't. Which in economics, which is my background, is analogous to scarcity, which is the fundamental concept in economics, that some things are possible, but some things aren't, either trade-offs: if you do this, you can't get that. So you have this pattern of constraints that ends up defining the dynamical trajectories of all the things that we could possibly observe. And then outside of that would be a larger space of just truly everything. So contradiction has the opposite of logical consistency, abundance as the opposite of scarcity in economics, where there are no limits, there are no rules. And so you have everything all at once, like in the principle of explosion: without logical consistency, everything is true and false simultaneously. And so this is all totally made-up. It's not anything I believe is true, and certainly would have no empirical evidence for, but just sort of this perspective I've been playing around with a little bit. So if I were to try to offer some kind of realism, dualism, idealism, whatever kind of take, this is my kind of take. It's something about webs of constraints that end up defining what patterns can evolve and what can't. Is it? I mean, that sort of form of argument is very classically theologian. So it goes back thousands of years, and the metaphysics of pretty much every single persistent religion, that exact structure with different categorical grounding, is used to describe that sort of infinite ground or that unbounded ground restricting itself down to categorical things that we understand. And the category, they do this in different variations with different levels of sophistication. So Hinduism has sort of 26 or 7 different subcategories, and they basically say each category has more restrictions on state within the category. So there are sort of cardinality-less objects or choices, and that applies across every consistent system, these constraint-based dynamics. A tool that we use to make sense of objects, I think throughout history, is like a thinking mechanic because the nature of observers, or the people sampling these spaces, is that they are bounded and finite. So economics naturally sits at the edge of that structure because there's a scarce amount of computational power. At the bare minimum, it's like a model that even if you're just saying you're doing non-computational stuff, you're reporting on it computationally. So if you've got computation, you've got bound in this, then you instantly have scarcity when faced with an Internet ground, an infinite groundswell of every possible thing.

[18:11] Ben: So you have to narrow it to make sense of the world. And so these sorts of ontologies, you can do them in lots of different ways, but they're very, very common throughout history. Sam, are you referencing Sankhya philosophy when you mentioned 26, when you said I'm not a Hinduism X? But I think it's probably a pedantic reference because it means enumeration, I think. And they do have 26. But what's interesting is Sankhya is notoriously dualistic in that particular, as opposed to, say, Advaita Vedanta, which is very... So I'm not a great expert on Hinduism. We have to apologize for the... I think someone muted you. So there was a problem with him. But that wasn't just me, right? No. Yeah, sorry, Sam, but the audio got very wonky for a minute there as you're talking. Yeah, it still is. I think you should mute yourself or something. Yeah, Sam, you might need to just leave and come back or something. Every time you seem to unmute, there's this repeated noise that blocks whatever else you're saying. That Zoom is going to be the new printer, where we have this amazing AI technology and it still never works right. Glad that wasn't just me hearing that. Got to be careful talking about nondualism. You know, I'm trying to take this to be more experimental in that sense. You said that there is no communication between the platonic space to the other way, but maybe there is a way to do some communication. If there is a communication like that, how can we test it? Because again, if I interpreted correctly, the neuroscience, there are people that did split brain surgery and they're reporting on two different personalities, one in the left hemisphere and the right hemisphere, and each part of the brain receive different sensory input and also shared input. If the split brain have no, there is, the corpus callosum is completely severed. So they don't have shared input. So they have completely split and at that point they create or form different personalities. But in normal people we do have connection. So in a way we do have two personalities, but have a shared connection. Yeah. I mean that's, that's people argue about that. There are some people that think that the second personality is formed at the time you cut right, that it wasn't there before. But I don't think that's the right, I don't think that's the majority view.

[21:50] Elliot Murphy: Was gonna say, the majority view is the other option, right? That it's a commingling that's already existed.

[21:57] Ben: So you cannot claim that it's a shared Platonic space in that sense. What I'm inclined to say about that in terms of shared personalities, are any of you all familiar with the A-not-B experiment, experiments done by Esther Thelen and her group back in the 1990s and early 2000s? So what she was able to do is basically, she was the one who started this, but came to, I think, the best understanding of this is there are experiments that show these kind of split-brain results. We have these multiple competing personalities, but rather than being in some brain-damaged person, they actually happen in normal healthy infants. So you can take a normal healthy infant. This is originally Piaget's experiments. You have them play with a toy, and then you hide it in one of two locations right in front of them, two opaque boxes, and you show them what you're doing. There's no tricks. You show them, "I'm hiding it under box A here. Where is it?" And of course the baby reaches for box A. And then once they've done that a bunch of times, you then hide it under box B, showing them exactly what you're doing. They see exactly where you've hidden it. They clearly know how to reach to boxes to get what they want because they've just been doing it successfully. And yet instead of reaching to box B now to get the toy, the infant continues to reach to box A. And that's, it's called the A-not-B error. And then there was this whole flurry of experiments over decades basically showing when this would and wouldn't happen under various conditions. And it's all kinds of, it's very sensitive to a lot of different things. And what Thelen and her group basically were able to show experimentally is that there's always these multiple competing dynamics inside of infants basically. And there's very particular reasons for this. It's not just because they're less developed; like the dynamics are kind of particular, but they're bad at resolving those internal competing dynamics compared to a mature adult. So the kind of potential inference to make from that is there's always split-personality stuff going on inside of us, but just our systems are better at resolving those dynamics to cohere to some system-level plan. And it's entirely, if the strong anticipation analogy to Platonic space is correct, then somehow these competing dynamics get resolved into the kinds of relational patterns we see. But I don't really know where that goes. Is that like internal family systems? Yeah, it's related to stuff like that. I've been trying to push these ideas farther because you see the same things in the economy, I think, and in neuroscience as well. And so, but the families, I don't like how the dynamics are much more ever-present. Like there's always a lot of things going on in all systems. And so I don't know, somehow some of the system's dynamic stuff somehow takes it a little too seriously. So just thinking about how processes resolve, that's kind of where I am on that.

[25:06] Elliot Murphy: Since we’re talking about the brain, I had a couple of notes to maybe present, if that’s OK, Mike. So I don’t have any slides, but I just made a couple of notes this afternoon that offer what I’m trying to think of as maybe an antidote to the common, pretty confused objection that finite brain and infinite process, namely something like human language grammar, are incompatible, right? Because I think that kind of hits at the center of how the themes of the Platonic Symposium are relevant to the neurosciences. That’s what I’m trying to think about the last year or so, like how the general themes that have been talked about in the symposium could be useful to people working within departments of neuroscience and neurosurgery. So one of the things that I care about is whether the formal properties of language, so, you know, Ben mentioned the example of the baseball player, whether or not there are other formal properties that you might say that human language has, like constituent structure or the fact that language doesn’t string words together in a linear sequence, but rather in a kind of recursive hierarchical manner. Whether or not that process emerges from some system akin to a deep learning architecture, like a fully connectionist modern connectionist neural nets approach, or whether there is some kind of inductive bias that helps constrain connectionist architectures from the very beginning that happens to be more tailored towards how human infants acquire their grammars. I currently lean much stronger in the second direction for the reasons that I’ve written a lot about. But I’m interested in using these formal or quasi-Platonic models of linguistic knowledge. Like the fact that human grammar is driven by a compositional operator that via category theory we can formalize in a very specific way. For example, a free commutative non-associative magma operation, sometimes called Merge in the linguistic literature. Basically recursively combining structures together, some way of forming tree structures, and using that formal set of instructions to see if there are neural processes that comply with and hence might be able to instantiate and derive these aspects of language. So we might not be looking for a kind of crude isomorphism, like we likely won’t get neurons and nouns to meet together in a totally sympathetic formal space. But I think over the past decade, the field of theoretical neurolinguistics, which is where I inhabit, has been able to move towards some candidate neural mechanisms like phase-amplitude coupling, phase synchronization, other forms of well-documented neural measurements. And just a quick note here, like the algebraic recipe that I just mentioned to formally characterize human language is chosen because each property corresponds to a load-bearing empirical generalization about natural language. And dropping any of these algebraic formal, you know, bullet points either leads to overgenerating or undergenerating possible human grammars, like possible forms of human language, essentially. So the argument is essentially that this is the weakest algebraic structure, minimal, consistent with the core observations, which is exactly the one that you want to understand the methodological norms of minimal commitments. So that’s the kind of background of my points here. But the reason why I say all this is because I don’t want to necessarily move the conversation towards linguistics, but I kind of want to use this as an interesting case study to make a different point that’s maybe more relevant. So in the literature, some authors when critiquing the cognitive neuroscience of language, which is how the human brain neurally enforces the properties of language, they’ve incorrectly assumed, and this is very common in the field, kind of one of my chief bugbears, that any object constructed by a physical system must also be physical, right? So hence speaking about the biology of language is kind of nonsensical by definition. So the claim by many of these figures is that language is either physical or it’s Platonic, right? It can’t be both. And of course, as some of you guys talked about over the last few months, this is a false dilemma. The rules of arithmetic are also multiply realizable, right from the analog abacus to the digital computer to the brain. Same for, you know, functions and sets and things like that. And also the same for abstract objects definable as mathematical at the proper level of analysis, like a computer program, right? So think of how a computer program explicitly representing the Euclidean axioms and codes only a finite number of bits. It does not and cannot encode the infinite number of bits that could be derived from the postulates. But it would be a bit obtuse to say that such an infinity is implicit or compressed in the explicit axioms. So while it’s true that physically we can’t perform indefinite computation, we are endowed physically with a competence. We are seemingly endowed merely with a competence that does generate a set that could be produced by indefinite computation. So the idea, this goes back to a lot of von Neumann’s claims as well in the ’50s and ’60s, so early ’50s. So the idea in linguistic theory that humans are endowed with the ability to recursively generate an unbounded array of hierarchically structured expressions is not incompatible with the idea that neural tissue can directly execute a sympathetically targeted operator, right? So Turing’s model of computation already showed famously why a finite physical system can encode a rule that generates an unbounded domain.

[30:05] Elliot Murphy: And similarly, you can consider a computer capable of simulating a game of chess. A chess-playing program is abstract, but it's instantiated in silicon, radio waves, memory states, etcetera. So a recursive linguistic function can be abstract and biologically realized in the brain. So the fact that a rule is mathematical doesn't mean it can't be physically implemented, right? So one of the interesting arguments that even figures like Dawkins have recently argued actually as being reasonable is the idea that natural language grammar has within it a kind of all-or-nothing component to it. So there are some aspects of language that we share with other species and are associative and sequential and graded. But other components, like the capacity for digital infinity or unbounded, recursive infinity, seem to be a kind of all-or-nothing capacity, right? You can't really have half of that or one-third of that. It's hard to think of how you can capture that in existing gradualist accounts. And in fact, any of these existing gradualist accounts of language evolution have had a hard time dealing with these sorts of invariant, mathematical formal properties of language. They kind of get brushed under, or they get claimed as just being emergent, right? It's just something that surprises us, and we don't really know how to explain it from the actual first principles of our models. So the issue is not that every aspect of the broad language faculty appears all-or-nothing, but that certain formal properties, as I just said, unbounded recursive formation, look less like a mere increase in associative capacity and more like an entry into a different generative regime. So we arrive at concerns of complexity science here, right? Anderson's more is different. People like David Krakauer have obviously spoken about this quite a lot. In physics and chemistry, it's taken for granted that more can be different through different types of, you know, if you want to call them and so on up. But in cognitive neuroscience, one of the things that frustrates me is that there's still a strong bias, and you could even call that dogma, that all explanations for human behavior must be yoked to the same reductionist level of explanation. Everything from touch to music perception to consciousness to language must be reduced to the same model, 20th-century-style model of neural function, right? And of course, a lot of these assumptions are derived from 20th-century models of motor function, cat visual and processing, and all the rest of it. But there's no reason why the same strictures will be wholly relevant for neural theories of language and higher cognition, especially in parts of the brain that are clearly outside of primary, you know, centrometer cortex. So one of my main gripes and the reasons why the Platonic themes are important here is because a lot of contemporary neuroscientists kind of have that, I think, problematic assumption that all of what we've learned from the brain in 20th-century, late-20th-century textbooks about cortical computation have to be explanatory, relevant for all of our human cognitive capacities, including our perception and our understanding of linguistic and even mathematical geometric reasoning, right? So people like Stanislas Dehaene in France has done some really good work where he's trying to figure out how the human brain processes core knowledge functions like mathematics, number, morality, music, and language and all these sorts of symbolic or seemingly discrete symbolic domains. But that may not be incompatible with other models of biology. So neuroscience has kind of privileged these forms of explanation, which have been inherited from central motor systems like coding, prediction, and things like that. But it's not obvious that these explanatory templates are sufficient for the algebraic or generative capacities of human language. For these reasons, I think the thesis that human unique intellectual capacities emerge purely from an expanded capacity for information processing can't be the whole story. That's a common explanation: human evolution, expanded information capacity. But for example, in one of Stephen Piantadosi's recent papers, he has a discussion of relational reasoning, and they acknowledge in this paper that symbol training pragmatically helps non-human primates succeed at quaternary tasks. But this is actually a problem for the pure capacity story because if it were just about holding more items in working memory, then why would an externally provided symbol unlock performance so dramatically? The symbol seems to be doing representational work, like compressing or reformatting the problem in some way. I'm not sure exactly what the presence of the symbol does, but it does seem to change performance in an interesting way. It's not just offloading memory. So again, I think this kind of speaks to the importance of these Platonic symbolic concerns of intelligence, like within human neuroscience, that is. And I'm not talking about the rest of our physics and biology. I'm just focusing narrowly here.

[35:05] Elliot Murphy: So I just want to briefly, I’ll just finish by connecting this with some of Mike’s ideas too, right. So Mike’s suggestions about temporal mathematical laws ingressing into biology and offering universal constraints on forms of life is not entirely dissimilar from some of these ideas in linguistics about enabling constraints for natural language, like Professor Chomsky’s idea about efficient computation having a causal role in the formation of linguistic grammaticality judgments and inferences. So the idea that there exists deep organizational motifs for organic forms strongly speaks to this kind of traditional concern of linguistics, right? So Mike’s assumed some kind of latent space that structures in the natural world seem to draw from and be constrained by the universe, as Ben was just saying, right, the universe draws from this latent space and not the other way around. So are there laws of mathematics that seem to be thoroughly invariant to physical law, except for these sorts of questions? And these sorts of questions encourage folk like me who may be partially sympathetic to broadly Platonic theories, not in the traditional sense, but also in this modern sense in philosophy of mathematics, to assume that life has evolved through being sculpted not simply by selection of logic, but by a latent space of mathematical forms. And for me personally, I’m interested in thinking more in my work about to what extent these former properties of language that I just vaguely mentioned do seem to be rather unique, at least in the cognitive science space. So, contrary to many misconceptions, I think the notion that language as a system of discrete infinity is not incompatible or inconsistent with the assumption that language has a finite biological basis, right? The consilience is affected by Turing’s proof of the coherency of a finite theory procedure, right, i.e. defining the computable numbers as low as determinable by his mathematical machine generative of infinite sets. Many neuroscientists in the present area have either limited the operational definition of language, and not just language, by the way, but also with the forms of cognition to the sorts of algorithms and structures that are safely neurobiologically plausible. So that’s kind of a common theme, right? We have to figure out what is neurobiologically plausible, and if there are certain theories of cognition that don’t seem to be neurobiologically plausible, then we query them and potentially dismiss them. But of course, that has a large presupposition about what exactly is neurobiologically plausible in the same way that physicists used to debate and still do what is physically plausible, that those sorts of open debates are less common in the neurosciences, right? At least in my kind of view, right. Even if there’s no regard to whether or not those neurobiologically plausible structures, often predictive sequential processing, bear any principle of relation to the behaviour or forms like linguistic forms under investigation, right. So language is plainly not just about next token prediction, right, despite what Sam Altman might tell you. So we don’t yet know enough about how the physical and computational properties of neurons make any strong preliminary conclusions. And I would say about the formal properties of language being incompatible with matter, essentially. So I guess what I’m trying to say is that biology may implement operators whose formal profile is best described at an abstract level, like recursive composition, etcetera, but that doesn’t mean that biology itself is incompatible with these abstract notions, right? From mathematics, biology repeatedly seems to realize, as people like Sam have written about very nicely, abstract spaces of possibility, morpho spaces, behavioral space, coctive space, and in language, syntactic spaces. And the natural sciences need accounts of how physical systems navigate those spaces and how the brain neurally enforces these algebraic properties of intelligence. So the formal properties of language are not rendered biologically suspect. I’m kind of arguing not simply because they are abstract, right? Like a finite neural system may implement a finite generative procedure whose consequences define an unbounded space of possible structures. The question is not whether language is physical or platonic, but it’s how a biological system comes to realize, approximate, or exploit a particular region of formal possibility space. That’s kind of the topic of formal neuroscience, right, like theoretical neuroscience. And I think this makes language a really interesting case study for broader questions about morpho space, cognition and biological organization, because with language, you have a really rich formal recipe for the forms and the behavior that you’re interested in. And you happen to be the object of study happens to be the human brain. So it’s a really nice mix of both of these formal concerns and these biological neurochemical types of concerns as well. Of course, like a neural system doesn’t need to literally contain nouns and verbs and things like that, as I said, right? But it may need to respect certain invariants like hierarchy, non-associativity, variable binding and things like that. What kind of biological organization allows neural tissue to become sensitive to such abstract invariance? That’s the question that my ROSE model of Language of the brain attempts to address. And it’s kind of what my contribution to the Platonic Symposium was all about. But yeah, that’s kind of what I wanted to offer as a general set of questions and concerns. I wanted to just.

[40:05] Ben: Go back to your example with the monkey being taught a letter or a symbol and how that improved the performance generally. One of the things I'm thinking about within this network or this ad-style model or computational model of this possibility space is that certain objects admit more morphisms, more potential embeddings than other objects. So language or letter or symbol for the observer or for the thing that's sampling it could admit more possible reductions, more possible compression. So that symbol in this sort of formalism frees up computational space for the monkey to do other things. And that extra computation, because he's been able to basically reduce a bunch of objects or a bunch of things that it comes across regularly to that symbol more than one, or it has more variability than, say, a stick on the ground or a tool that might be one-on-one potential mechanic or explanatory option, fits within this idea of computational boundedness as a sort of core drop. And in that computational model of how like a sort of selection criteria for Platonic space. So the idea is that these ideas are competitive and, you know, embedding or observers working with them or being directed or pulled by them. That ingression works both ways. So observers sculpt the idea like this idea of co-creators or interactions or relaters. You're enforcing some carving around the edges of an idea to try and get it down to that minimal hole. And maths is very good at that. And so it's science because it's effectively gradient from you're finding lots of invariants as you go up the curve to more complex objects in latent space. Maybe it's an emotional, maybe it's the way a book feels when you read it because of the rhythm that is not reducible in the same way. So when you're using language to kind of carve the edges of that concept, there's more variability, there are more potential ways of expressing that concept. So that's kind of one thing I'm playing with at the moment, trying to figure out how to design experiments to do that. So I'm not that sort of thing. I'd say it's not very good. I'm not very good at it, but it's something I'd love to try and explore because I think it's a rational mechanism that fits with a lot of the core things that are true about computation in the very abstract of how it works, particularly in Stevens model, but also in things like Church-Turing, and also applies to things like constructive theory in a bleak second-hand way.

[43:01] Elliot Murphy: Yeah, it definitely connects to constructive theory. I totally agree with that. I would recommend the work of Stan Dehaene. He's done some really cool experiments on minimal description of auditory working memory, visual, linguistic, and tone, the different kinds of linguistic and nonlinguistic modalities. And he's showing that using integrating recordings in humans and also extra training recordings that the patterns that they may be formally distinct across music and language and what have you, right? But they still adhere to, they comply with some of these broader Fristonian concerns about minimal description length, etcetera. And the ones that do are the ones that are recognized most saliently and arise to consciousness and also derive a lot of other aesthetic preferences, which one thing that I've been thinking about recently, which I think you might like some actually, given your background, is there's a debate in linguistics, which is more of a philosophical debate, but really about how the Great Leap Forward happened. And if the evolution of language was linked to some of these other higher order capacities, right. So Chomsky had an idea that language might be causally linked to the emergence of the successive function and then the natural numbers, right. So they both fell off all at the same time. I'm not sure if I necessarily agree with that, but there's some other ideas that when language emerged, it seemed to have a functional use, of course, for communication and sharing ideas with other, you know, conspecifics. But it also seems to be very linked for, like, formally linked to our notion of beauty. So Chomsky also had this idea that language evolved for beauty in a very vague sense. And that's kind of hard to cash out and explain in ways that we could hash out here maybe, but there's some kind of intuition there. The formal characterization of language is compatible and helps facilitate the creative expression of other nonlinguistic forms of beauty and expression. And there's been some pushback in the literature because some people from the stroke side and the neuropsychology side say that language is functionally segregated in the brain from other systems like music and morality and domain-general cognition and the salience network and the tension frontoparietal attention network. So therefore, language has to be some kind of separate system. But again, that's a there's a lot of controversy about just because a certain part of the brain might be a certain function might be in a different part of the brain, it could still be doing a similar computation. So the obvious example that people always ignore is motor cortex, right? So dorsal versus ventral motor cortex are different parts of the brain, but they are still subserved by the same beta phase synchronization neurocomputational mechanism, right? Whether or not you're doing motor coordination for your hands or your face, it's the same kind of algorithm essentially, even if it's a different part of the brain. Sort of strange inferences have been made about, oh, because this part of the brain is, you know, this functions over here and there's sort of slightly different functions over here. They have to be subserved by wildly different modular systems and etcetera. So that's a, that's a, that's a real ongoing topic. But yeah, I do, I do like this idea and I can't really explain why I like it, but the idea that language evolves for beauty and is itself a form of, you know, a beautiful form of object, it seems to cash. And, you know, it gels with me for reasons that I think people who have a similar instinct working in this Platonic kind of space will probably recognize, even if it's, you know, same aesthetic preference but different kind of domain, right?

[46:30] Ben: Yeah, I mean, like they tend to agree. I think these sorts of psycho technologies like language and as you move to writing are all tools to increase our ability to access, if you think this sort of Platonic space in whatever format you decide to have it is real, then every one of those technologies is like an exponential increase in the amount of that space we've got access to us. Therefore, the amount of possible objects that we can bring into this world, and we search for objects that are coherent across whatever sort of piece of that we sample. And that coherence where we discover those, I think it's linked to aesthetics and beauty and symmetry. Objects evolve. Lots of things in this world involve broken symmetries, and symmetries in some respects seem to be a big component of that. And you're talking about sort of an autorecursive function. Recursive function is effectively like operation operating on itself, is as close to symmetrical as you can get, with different ratios you can get to like that's how you get to the golden ratio, right? So these are things that seem baked into the structure that we're sort of sensitive to, or sensitive to the finding. And you see it passively in other forms of life like, you know, peacocks like things, things trend to beauty to signal things are good. And that can be in birdsong or anything, but we just have much more of a range of ways and languages like we can now do that much, much cleaner. And I think there's all those technologies. And now we're sitting at the point where, you know, we've got the next one, which is basically like, you know, might be inside all of the things we know. But there are lots of connections between the things we know already that are very hard for us to map that these systems are quite good at. And my guess is that it's going to be, you know, I think the function of people in this new world where you can search, you know, effectively LLMs, that you do a very lossy search of Platonic space. You direct it to a gradient, you direct it to a space. And you say, and your taste is knowing which of those sort of sinks or attractors in that lost landscape are actually reasonable to link together and the mechanisms to do it. And what's going to happen is as the systems get better, they might get better at mapping the inside of that space more and more. And my guess is that humans will go to the edge of that space where they compose new things from the inside. And that's sort of that recursiveness in language being sort of accelerated by something that's recursively generating it internally, now feeding something that's at the edge of that space, recursively generating new things or finding new things, if you think space is permanent. And that's sort of quite, you know, romantic. You know, it's like romantic or a romantic relationship with that idea because I think it's quite a meaningful idea. And we're sitting in a world where these systems seem to be, you know, there was an article today that like, is in Nature, like scientists going to be doggone in five years or 10 years or something like that. And that just seems, one, wholly stupid and two, they're not understanding what these systems are or what the component parts are. And you know, you have really brilliant people like Carl Friston and like, because it's not like it's not out there. These things don't do the explore part. We're very good at that. And these things are very good at the gradient part. And I think that that sort of break is going to be how these things kind of work together and why how these sort of things will kind of come together through that aesthetic relationship between that process. Like you can use it to check out a lot of crap, which is what most of the Internet does, or you can do something a bit more interesting. Hey, Elliot, does the way that you've described the language, does that help us or what tools exist to try to look for it in unconventional data sets? So if we, whether it be a neural data set or the actual output that may or may not contain language, what I mean, I guess it's a subset of the setting problem sort of right? But what tools are available to ask whether, like if you have a system, let's say we have two systems, they're clearly communicating. We can do the mutual, you know, the transfer entropy or whatever. Like we know, we know that we know they're talking to each other. But now the question comes, OK, are they using some sort of, some sort of language? Yeah. Is there a grammar? Whatever are there tools available to analyze it and make a claim on that?

[51:17] Elliot Murphy: Yeah, the field of NLP, like natural language processing, has done a lot of work trying to figure out, for example, if you give some transformer architectures the ability to train and learn some basic grammatical, very basic grammatical properties like the way you form questions, the way that pronouns relate to the nouns, et cetera, really basic relational stuff. And it turns out that they can often approximate some of these basic structures, but they do so much better and much cleaner if you add what's called a transformer grammar. So you add some kind of explicit grammatical rules inside the, you know, you feed the transformer some explicit kind of prior, an abductive bias, I guess, about the grammatical rules. And so with transformer grammars, you get a lot, a lot closer in the space that's been explored right now to figuring out and sourcing out relevant linguistic patterns that's out there. The issue that I have, though, is that a lot of people try and do this for this approach and they look in the non-language, non-human space, right? So whales, there's lots of papers about do whales have language, right? Do other forms of species have language? And I guess the question you're asking would be a similar type of search, like which other kind of structures in nature could have a possible language? The main, the real problem here is that the focus is there's been too much focus, I think, on the kind of sequential and probabilistic properties of passing, which human language does do for sure, and not in a focus on the kind of more invariant. So Sam mentioned broken symmetries. One of the core algorithms in human language is symmetry breaking, where you form two objects together into a set and then one of them kind of becomes the head or the label of the phrase. So every time you put two words together, one of them seems to lead and kind of do semantic inference. There's never kind of outside of compound noun phrases like swordfish. I guess swordfish is a rare example, but most phrases in human language, when you put things together, there's like the tree, I have a leans left, Orleans right, right, the kind of branching left or right. There's some kind of forced symmetry breaking. So that's kind of the core algebraic component that I would look for, I guess, in other systems, like if you'd say, OK, do these things have language? Do they have a similar type of symmetry-breaking algorithm that would be predicted based on the algorithm that all human beings have? So like the fact that, you know, there's some examples are the fact that different languages have different surface orders like SVO, SOV, OSV, etcetera, you know, subject-verb-object, but the underlying kind of semantic instructions in conceptual space are invariant. So, you know, when I talk about a whiskey glass or a big brown bottle of whiskey, the linear order of those words is different from the way a Japanese-speaking person would say those words. But the underlying conceptual instruction is the same.

[54:26] Elliot Murphy: And the way that the symmetry is broken is also identical. So the underlying kind of semantic configuration, if you like, is identical, but the surface form is different. So there's been too much focus on the patterns in the surface form in human language, which is why if you do that, you end up concluding that whales have language too. And you end up concluding that dolphins have language because they also adhere statistically to some of the similar surface patterns like Zipf's law, for example, and other sorts of foundational laws in linguistics and cog science. And you say, well, because they have the same kind of statistical compressions, or they adhere to the same statistical constraints on sequential chunking rules, then therefore whale communication systems are just a different form of language. The risky move there is that you conflate the expressive medium with the underlying kind of state. So I assume that whale language may be, on the surface, from complex, and it may have some of the same statistical regularities that human language does. But I would assume, and we don't know for sure because we can't get, you know, test this for sure, but we assume that at least I assume that a whale's conceptual apparatus is not being regulated in the same complex ways that human conceptual apparatus is being regulated by human linguistic instruction, right? So that when I hear a sentence, I'm not just passing the linear order of words. I'm having a very specific revision to my world model based on, or I'm navigating my world model, I should say, based on the kind of instructions that I pass together. So it's hard to, it's often easy to conclude that nonhuman species do have a form of language because, oh, it looks like they, you know, the expressive modalities adhere to the same kind of statistical biases that we do. Could there's been some work on pigeon communication systems and how pigeon communication also adheres to some of these patterns too? But like I said, and this is one of the problems that large language models have too, right? Because, you know, I've published some papers with Gary Marcus and some other people showing that large language models will often generate a particular structure because they think it's the most common or likely kind of construction to a particular linguistic question, though it may not adhere to grammatical rules. So there's a lot of work showing that large language models don't, they can't pass. They can pass, I'm sorry, they can pass impossible languages that would be impossible for a human being in ways that are pretty much just as good for them. So they would learn an impossible grammar just as much as a human an impossible grammar, right? So whereas human beings would never, you know, adhere to these kinds of impossible constraints. So the real Searchlight I think should not necessarily be on the surface distributional forms, but on this very simple underlying algorithm, a rhythmic set of instructions that you then, you know, lead into this more expressive kind of landscape. I don't know if that helps with the question, but it's generally where I kind of would land here.

[57:35] Ben: Can I ask a question just do with animals? Is it like a composability point, like human language has composability? Animals' calls, or however they speak, they're not composable, or does that have anything to do with it? It's not outside of anything I don't know about. So it's just.

[57:55] Elliot Murphy: So it's this is where I, it's very common in the philosophy literature for people to talk about combining concepts, right. Humans combine concepts when you, in fact, let me just pull up an example that I was working on.

[58:12] Ben: Today I mean more simply the composability of the single letter. I mean down to the letter level. It’s not the same chunk ability as we have more ability to make more distinct things, and therefore we get to this thing that’s kind of what I’m getting at.

[58:32] Elliot Murphy: Yeah, monkeys can do minimal morphological composition. They can say a call like crack or a call like hawk for an airborne or a land-born predator. And then they can emphasize it, like, oh, so let's say crack, oh, or haku. And they'll emphasize it based on pragmatic considerations, like whether or not the predator is close by or far away. So that seems to be the full expressive range, the combinatorial expressive range, at least externally. Internally, it is possible that they are combining in conceptual space a much more rich array of features, but they're just not externalizing that. Whereas humans, we externalize fairly complex expressions, and then we map it to an even more complex underlying space, underlying state. There's been some work by Tecumseh Fitch in Vienna in the last decade showing that. Just a very vague summary. I can't remember the details exactly, but nonhuman primates seem to have all of the requisite acoustic articulatory expressive apparatus, and indeed not just the apparatus, but a lot of the neural control of the expressive apparatus to produce virtually all of the 35 or so, 36-odd phonemes that human beings can produce in English. So they can physically produce those sounds; they just can't link the expression of those sounds to a underlying, you know, a commutative device essentially. So I don't think it's a limitation in the physics of the acoustics, but I do think it's a limitation in the symbolic space that we feed into. So it's often said that, you know, humans, you know, some linguists kind of sometimes exaggerate that human language is like the main thing that makes us unique. I think it is broadly true, but there's also a representational thinking, which is of course separate and prior to language, like the fact that humans can just do representational cognition and symbolic forms of cognition, right? The Vlascal cave paintings. And that seems to be preserved in patients who've lost their language faculty, like stroke patients. So representational thinking is really at the core of all this, like this symbolic representational reasoning ability. And I just, I think that is the main kind of limiting factor, and less about the statistics of how you parse the symbols and the statistics of how you kind of construct them. So like I mentioned pigeons, you know, a few minutes ago, there is some work on, I think, the pigeon brain that they can pass a lot. Well, they can't pass, right? But they have a lot of the similar statistical biases for passing orthographic information, right? So you give humans orthographic information, you read things, and you kind of are constrained by your visual, you know, anterior mid fusiform gyrus in your ventral temporal cortex. Pigeons seem to have a similar constraint in terms of how they pass the statistics of orthographic information. They can't read those, they can't read what they mean. But when they visually pass those symbols, like letters and numbers, they seem to be adhering to the same distributional constraints. So again, I think it's more in the symbolic space where the explanation of novelty lies, less in the kind of, you know, statistical space, I guess. Yeah.

[1:01:58] Ben: So I'm trying to track everything that's been said and integrate us. I'm having some trouble, but let me see if that works. So there's something about beauty and also about spirituality and theology as constraints, or science as being like about the boundaries that cut up the whole into pieces. There's something about beauty often being linked to one of the core feelings related to somehow spirituality. So it seems not unrelated that we have spirituality and beauty and constraints as cutting that up into pieces, into symbols specifically, right? So we're trying to understand, like, what is this reality we live in? How does it work? How do we perceive it, right? Then it seems like there is the whole, which is just the undivided actual world, and then us carving it up into patterns, finding patterns and noise, creating symbols out of that that then have meaning, right? And then we sort of ascribe that meaning that is inherently and deeply linked to our survival, right? In terms of at least homeostasis, maintaining, avoiding predators and things like those are the symbols that become relevant to us over evolutionary time scales, right? So the sort of patterns that we really attempt to, that we find, that we carve the world up to are the patterns that help us survive, right? So in terms of this creation of these kind of platonic forms out of this just noise of sensory input, it seems to me like there's this sort of feedback loop, right? We have this input, we carve it up into symbols that we ascribe meaning to. And in particular, we ascribe these representational higher kind of platonic patterns perhaps to them, right? Which is sort of an interpretation map, right? That, like, if I see a green light as I'm driving, I interpret that as a specific sort of meaning, right? Which then the interpretation map sort of helps me to survive in the physical reality. So it has to have that feedback loop into being meaningful for physical reality, right? And so it creates sort of this feedback loop between what are the platonic patterns that are relevant? And then how do they actually matter in terms of consequences in physical world. So to me, it feels very much like relevance realization from Verbaiki is sort of core to this, to what time patterns are. And what you were saying about, yeah, also I guess interestingly in spirituality often, right, we see like in the beginning was the word. So they sort of ascribe language and symbol and symbolic and representational reasoning as the source of reality as we know it at least. And that's also true in Hinduism with sort of Vaach farh Vaach. And yeah, I think that's roughly. And I guess I also had a kind of question to what Michael was saying earlier, if I'm also a pattern, right? That it's not that there are patterns somewhere, but we are also patterns, right? This was in relation to this baseball I think we're talking about. And if everything is patterns, then how is platonic space or reasoning different than just materialism, if everything is the pattern interacting? And so that was sort of the piece, but I'm still, I don't know, I'd almost like, I'd really love to find a way to integrate this discussion of all the pieces that people are saying into like some, like, OK, this is the model of reality we have. And so we're struggling with being the edge of that.

[1:05:40] Ben: That's sort of my question as well about how, how is it that whatever is happening in the seemingly physical is not having any sort of impact back onto the Platonic realm or Platonic space? Is it somehow already an infinite complete space, or like we kind of touched on it, or is it capable of growing, right? So I'd love to hear everybody's thoughts on that. I mean, I would just make one modification to what Pavel just said, which I'll say in addition, which is all of that is reasonable. But that whole story was told from the perspective of the humans that pick patterns. But I think to various degrees, the patterns in that space, just like as we are, the patterns themselves have agendas. They are agentic to different degrees, and they have agendas. So it's not just us who pick patterns that suit us. The patterns to different degrees in different media have incentives. I'm not exactly sure if the vocabulary is right here, but I don't, you know, I don't, it's the thoughts and thinkers symmetry, right? Like, yes, these patterns can exist. They can exist within our cognitive systems, unless they are thoughts that we have, anywhere from fleeting thoughts to obsessive compulsive thoughts to like whole personality fragments and so on. But the patterns themselves, I think, have various degrees of motivation and competency to inhabit various media. And there's a bunch of examples of that that we could talk about. But I think, you know, I think you don't always choose the patterns you get, is what I'm saying. And depending on what you build as an interface, whether that's a bio bot or an embryo or a language model or whatever, it will become, you know, sort of occupied, invested, whatever, by patterns, some of which you knew about and wanted and some of which you did not want and probably don't know about. So that's, you know, that's one thing that I would put in. And again, to Darren's point, I would say, yeah, I absolutely think that they are in some way modified by their experiences projecting through the physical world. And so not maybe not all of them, right. So I don't know that you can do anything here to change E and things like that. But other kinds of patterns probably are modified by their experience. And that probably means that at some future point when they ingress somewhere else, they'll be different. You know, there will be differences that are probably detectable. The idea of the patterns having their own agendas that we don't really get to pick is basically, I think, a core principle of the dynamic systems approach to motor behavior. And I can actually show you all a quick little example of that. So I'm going to share my screen again to do this is from a different presentation, but turns out to be relevant. So this is Esther Phelan. She's a big influence of mine. This is an infant, not sure how old, but maybe somewhere between five to seven months old. So not old enough to be walking on his own. But what you're about to see is that if you set up the situation right by supporting him on a treadmill, he's going to start walking. And so I'll show you that here. What? See why it's placed on the treadmill? A stepping behavior comes back. So one key component of walking, stepping, does seem to be involved.

[1:09:25] Elliot Murphy: Think it probably.

[1:09:26] Ben: Is built into children from the start. The stretch of the legs on the treadmill is enough to start that Walter Megan process going. So basically what's going on there is that motor behaviors like stepping. As that kind of experiment shows, it's not something the brain just plans and says, OK, here's how I want the legs to move. That's a baby. Their brains never tried to walk before, has not had time to hypothetically download some ostensible walking program from the genes. Like that's not what's happening. What's happening is that the anatomical structure of a human being just naturally enters that mode when you get the situation right. So it's another example of this kind of strong anticipation phenomenon where when the organism couples to the environment to maintain a certain kind of relationship, which in this case is a phase relationship between the legs, that just produces this kind of stepping behavior. And so instead of motor behavior just being controlled by the brain or just being this prescribed stuff that evolved and is somehow stored in our genome, it's something that just emerges dynamically through this relational system.

[1:10:40] Elliot Murphy: Yeah, I just wanted to say as well, I like what you said, Pavel, about the connection to theology and spirituality and language. I thought that kind of resonated with me quite a lot. Obviously I'm a linguist, so I'm biased, and I like the idea that in the beginning was the weird. That's good, and not in the beginning was economics as the tragedy, I might say, right? That's the other alternative. But no, I also like what Sam said as well about linking this back to some of the theological and philosophical ideas. I kind of like what the early Wittgenstein said about how the world is a totality of facts, essentially, right. And if the evolution of language afforded humans the ability to make these novel structural inferences about, so for example, here's a very trivial example, right? A lot of people in philosophy of language, including me, kind of think that truth is a linguistic kind of notion, right? As in you can only really get to truth judgments at a certain level of grammatical complexity, right? So if I just say lion or laptop or the laptop, that can't be true or false, right? It's only when I get a full subject-predicate argument structure, then it can be truth or false, right? So the laptop is on the table, that can be true or false, right? But if I just say laptop hungry, you know, outside, or do these sorts of random commands and requests that are not fully subject-predicate enriched, then you don't have a notion of truth. So this kind of evolution of language being linked to the evolution of truth, which is very not 1,000,000 miles away from what we were talking about, about mathematical truths, so to speak, maybe not massively dissimilar. And then I think we do get a new notion of world, right? Like having language gives us an entirely new notion of what the world is because we can construct it. So I often think like there's been some ideas that if language allows you to, like the notion of the self, for example, is very well connected to language, right? The concept of self is probably the most complex concept we have. It's like insanely rich because it's connected to so many different cognitive systems and modules and inferences, right? It's not just its own isolated, encapsulated thing. It's really enriched by basically every neural system and faculty that we have, right? Where it's attention, affective states, memories, etcetera. It's a very richly constructed notion of self. But then once you're linked back with language, you come to the possible conclusion that, you know, other animals have memories, but they probably don't have a past, so to speak, right? You got lots of subtle distinctions about how the sense of self that humans have is a linguistically informed notion. It's a kind of a linguistic concept in the same way that the world itself, it's not to sound like Wittgenstein too much, but like obviously the world itself is not a language game, but there is a certain part of which how we think about the world is being strongly linguistically inflected. It's really causally constructed by our language faculty, which is why I'm kind of so keen on using language as an example to think about the negotiation between forms and the brain, essentially, yeah.

[1:13:53] Ben: I like your point about truth because that's like the primal dualism, right? Once you have notion of truth and falsehood, that is the primal duality that then creates the notion of self and not self and all this. So very much in those spiritual. Can I ask a question? Sorry about that. This thing where if an animal or a human loses vision, then let's say auditory perception starts to take over, right? Some of the real estate does. Does language processing ever do that? Do you ever see language processing taking over new territory like that?

[1:14:44] Elliot Murphy: Yeah, there are some really nice ideas by David Ilkeman, right, who's at Baylor just down the road here, which I don't completely agree with, and, you know, most people don't. But he's got some nice ideas that, as you said, when he's got this idea that different parts of the brain are eager to take over territory from other parts of the brain, right. And as soon as they get the opportunity to do that, they will proceed to do that. In the case of blind individuals, for example, the best example I know of is there are people who are born blind but who develop a fairly rich language system. Depending on when they became blind, whether they're born blind or if they become blind during childhood, will determine how much of occipital cortex is used for language function. You can predict very, very, very cleanly how much of the, you know, set when you give them sentences versus lists of random words that don't have any linguistic structure. The contrast and activation between those two conditions will be predicted by it. The contrast of activation in visual sites, so parts of the brain that they just don't need because they don't have that sensory modality, will be predicted by the language. But there's a slightly more nuanced point to that. Obviously neuroplasticity, neurosurgeons depend on it. They need it to do what they're doing because often you'll resect tissue from tumor sites or epileptic tissue from perisilvian language areas, in which case you're going to risk potential language deficits. And in fact, you can induce aphasias like we've had some cases in our hospital where patients unfortunately become effectively fully aphasic. But then the cost of that is either to remove a very problematic tumor or solve otherwise intractable epilepsy, which is like not allowing them to drive or hold the job down, etcetera. So it's a very, very delicate issue. And in fact, our lab has done some way of trying to figure out other sites that should potentially be more respected during surgery because they often have less obvious terms of language deficits once you resect them. But there's really good work by John Krakauer in Johns Hopkins, who's written some fantastic papers. One of the best papers that I know of is a paper called Against Cortical Reorganization, which was published in eLife a few years ago. And he says that yes, it's true that language cortex and other forms of, you know, other brain networks will take over territory when that territory is kind of free real estate, so to speak. But there's very specific constraints and predictable constraints on how and when different combinations of takeovers will happen. So it's not as plastic and as much of a free-for-all as you might expect from a kind of extreme plasticity perspective, but it's also not entirely impossible. There's some kind of interesting, very complex and yet to be figured out middle ground, which there isn't a huge amount of work on about how and when, you know, which order of takeover of cortical territory will happen based on existing deficits. That's a very complicated question. But yeah, I recommend that the Krakauer paper. He's really elaborated very, very nicely.

[1:17:47] Ben: Cool. I think, yeah, you know this business of patterns and not just static patterns, but computational patterns. Do you know different types of compute taking over new real estate? I think might be, I'm not sure yet, but it might be a good model system for this business of looking at some of the agendas that these patterns have of trying to be embodied. And we see this. I mean, so there's the neuroscience cases that you talked about. There's the developmental biology version. We see this in a number of cases where there's literally a battle that goes on, a competition between existing patterns for real estate for the cell collectives to see which one is going to take over. And we see it in our ectopic eye induction. We see it in the transplants of the two-headed tissues and stuff like that. You can actually watch now. Now we detect it electrically, but you can actually see these patterns compete for who's going to establish that. And once they do get in, then the materials just snap into line. So then the cells just do whatever the pattern says and then you get whatever you get. So, yeah, I don't know. That's one of the things we're developing, some assays to look at this and to see what it looks like from the perspective of not just from the perspective of the excitable medium, supporting various patterns, but actually from the perspective of the patterns, trying to have a bigger impact on things in physical space. I mean, what's really interesting about this discussions of patterns and sort of agendas of these patterns, right? So, you know, intuitively throughout history we have folk tales and, you know, even religious stories or apocryphal tales around religions like, you know, "the demon made me do it" is the classic from like, you know, the witches. Maybe these are all pointers at some object in concept space in one of you. They're like, and what's happening is when you think about how to observe as computationally couple or join up, what are they doing? Well, one observer is offering a proposition, a potential rule or a potential choice that can be integrated into the other observer's internal model or their world model. And that will be weighted on, you know, what impact on boundedness given their historic like home, given their tails. It's like a minimal form of that model. And they'll go, oh, well, my model is locally estimatable and I have a rough idea of my direction. So I'm going to use some FEP-style, you know, measure to get to there or Epiplex T to figure it out. And I'm going to be maybe slightly wrong. And when you have these sort of objects, all these patterns, maybe it's a type of behavior or something, they're offering a proposition that's giving a score to that internal model to that, like left-brain part of you that's lossy and inaccurate. And then if accepting that proposition, you bring it into your world model. But that attractive or that sort of pattern in, you know, this latent space may well be quite embedded and may not be just the proposition. It's like, why do people get addicted to things? Because the pattern at the front offers massive reward and then has a deep, you know, a deep gradient and you can't get out of that sort of valley without extreme noise, which, you know, messes up. And my sense is that human beings, more than other sort of observers or, you know, animals, plants, rocks, we are in many more basins at the same time in this space and those basins, because this is, you know, computationally modelled as like an infinite group point, right? It's like a big mesh network and all those attractors are connected. And it's like a bit of noise from one can really knock you in different directions. And I feel like these attractors, as they get more people in them, that basin gets like more the course analogy is gravitational pull towards it. So more obser. It's why you see sort of rushes of crowds and economics.

[1:21:53] Ben: It's like Howard Marks' idea, and my background is finance. So it's very boring to have to, but these sorts of ideas of how things speed up. You have these recurrences. You talked about recursiveness and language. Is that recursiveness and embedding quality where my interaction with that pattern? But it could just be another observer. If you're going to treat patterns as agentic, well then they're different types of observers. Maybe they're not physically in our space, or they have to be hosted by something that's in our part of that space. But that loop around embeds that idea more and more in that space. And it's like we have, you talked about sort of how what are we doing to these ideas, or if this space is uniform, infinite, and already exists, and has all these ideas in there, and we're identifying them by paying more attention, we're effectively deepening that structure. Maybe we are actually mining those values partially, like some value is very computationally efficient if you want that model, or has high valence to us or high fitness value, and we plow into that value. And we only later discover, because of our boundedness, that that value is actually not the best gradient. It's a local maximum. And these are gradient-following heuristics. But I think it's a useful one because then you can say, well, does the space work evolutionarily? And can we pull evolution into a Platonic space and have that as a driver and informational driver of evolution? It's like something that Susan Blackmore writes about quite a lot in her memetics work, which I think is pretty excellent and much more sophisticated than me. But I think that's sort of the direction of travel that you can go for with how to explain how these agentic patterns work. I think it's quite an interesting way of thinking. So let's use the tools of network theory, tools of small-world networks, and these are all tractable, particularly with the technology of an LLM, which is an N-dimensional network. It's like you get this sort of architecture for free, and these sorts of things can potentially be unlocked by it, so you can model them. So I've got a question here for everyone, kind of building on some of these things: hypothetically, near future, when if we were to bring in more advanced sort of cyborg-type technology and we merge more and more with technology, maybe we have access to more of this quote Platonic space. But in terms of selfhood, where would some of you stand on, is there a risk effectively? Like does it create a new dominant member at a certain point that relegates our former self to something lesser than, or is it almost have the potential to be wiped out? Curious where people take on that sort of thing. I heard a really interesting. I don't know if this has already happened or if it's a near future thing. But basically it's an AI prosthetic for people with degenerative brain disease that are losing their memories. And so at first this thing is like, your car keys are here and you know where you put them, and little stuff like that. But as the thing progresses, the AI learns more and more about what you've got. And as the human loses track of it, with progressive loss of memory, the AI picks up the slack. And so eventually, you know, you sort of take that to its logical conclusion, and the spectrum shifts right along. So you start off with a disbalance, and then you're pretty even together, and then the spectrum shifts the other way. So I could imagine, and I mean that seems to be a very real therapeutic scenario that's sure going to play out if it already isn't.

[1:26:02] Elliot Murphy: There was a conference that went on today and yesterday at Rice University that I went to here in Houston about neuroengineering initiatives using brain-computer interfaces, right. One of the things that I was involved in is working in a space of brain-computer interfaces for language and speech. And there's a really cool project being done by some of the guys at Baylor University, Baylor College of Medicine, where they're trying to implant a Utah array, like a chip, in a dorsal frontal site called area 55B, which is just a little bit the posterior end of middle frontal gyrus. So kind of north of Broca's area and language cortex, but still kind of ahead of primary somatosensory motor cortex. One of the cool things about this part of the brain is that it seems to represent, it seems to be active for a lot of speech and language content, but it's also very close to motor cortex. So their goal is to try and use it for BCI purposes, right? An implanted chip in this device to help as essentially as a language-motor interface BCI. So it would decode speech and language from this area 55B part of the brain, because of course area 55B is not part of primary visual cortex. Primary visual cortex is actually part of the upper part of STG, which is carved into the gyrus. So it's hard to have an implant there. 55B is luckily a gyrus, so it's easier for folks to, for neurosurgeons to, you know, do a craniotomy and implant a Utah array there, so it's more accessible. It's up around the dorsal frontal cortex, which is a major site of current BCI implantation, but anyway, for all sorts of different disease states. So the goal would be to decode language semantics as a way of getting to these fine motor instructions. So instead of implanting a device right over different portions of motor cortex, you would simply implant it in 55B and then using your linguistic inner speech, you can construct a sense of like, OK, I'm going to grab the hammer in order to smash this object. And that linguistic instruction would then essentially take over all of the attendant fine motor instructions. So the BCI would be optimized to kind of dip its toe into some of the motor instruction signals, but also some of the language signals. So it kind of acted like this dual interesting interface between language and motor coordination, which is a really interesting way to kind of, first of all, address a broader patient population, I guess, right. But also in terms of this notion of interface is really important in, in, in, and will become more important in the brain-computer interface world, I think, because it's not just about addressing single discrete encapsulated deficits. It's about the interface. So why do people want to recover their language? Again, it's not just to think to themselves, it's to be able to use their language function to provide refined instructions to different areas of conceptual space and behavior, essentially, right. So you use language for a particular functional goal, which is very different. That's the applied engineering sense. That's obviously a very different goal from the goal that I have, which is just to look at these formal algebraic properties of language and really understand how the language system is orchestrated in the brain. The goal there is to actually use it for a clinical purpose. So yeah, there's a lot of really exciting work going on here.

[1:29:28] Ben: If I'm understanding this correctly, I'm thinking that the price system might be an example of what's being discussed, where basically prices are this interface that do goal regeneration and novelty absorption, where whenever there's some surprising event in the world that might be too distant and too complex to deal with, like say some storm in China that has like a big effect on things. What's going on with that? How should you react? That would be really hard to figure out. And then what prices do is they say, OK, that's too hard for you to figure out. So all we just need to do is adjust the route system of relative prices. And what that does is it conveys the shock to you in a way that you know how to deal with, which is just to go shopping based on a given set of prices. And what that does is simultaneously creates that enslavement process, which is, you know, from strong anticipation where you follow the economy's invisible hand and like you're kind of obeying its processes. But at the same time, the way that the invisible hand gets constructed is from the behaviors among the people; how they interact produces the order parameters that end up in slave for everyone. So it's the circular causality thing. We have this interface that people create that basically helps them deal with things that would be too hard for them to deal with on their own. What's interesting is the dynamic and economics when you look at like what pulls things. One of the ideas I'm thinking about is these very coordinated small-world networks of agents or observers that come together and then have, for a brief period of time, outsized influence on that overall object. So in your invisible hand scenario, it's like the core of the idea. You could use the financial crisis of 2008 as an example. The core of the CMBS idea was actually a very good idea, and it was very good for a few years. And Gillian Tett's written about it at length. But that pulled the whole system, and suddenly you had the whole of that invisible hand now pointing at this area where they could exploit a lot of effectively very cheap gain. There was a lot of excess fat in this part of the system. This gradient was very exploitable for the limit. That gradient was interlinked with many others and effectively had to shift back and shunt back. And that was the get too far into one of these attractors. You get kicked back out by the noise quite aggressively. Kind of think of it as like a steep valley wall. And as you approach things like booms and bust cycles, that's like running into the end of the line where it's not composable anymore. And the next choice is always computation inefficient. So you kind of have to go to the fastest path or try and get out of that track. There has to be a big, big shock to move you. And these are dynamics I'm starting to think about in sketch around some of the tensor basic, very basic tensor mathematics around how observers do that proposition sharing and how they have control. How those patterns, which are composed of agents reinforcing them, which agents drive those patterns. If you're in a company, you can imagine that as one mind; the node at the top, the CEO, has more control over the direction of that whole object. But in Platonic space, that's just a network of nodes and relations. And there's one bit that pulls it left and right in the space and a bunch of people in the middle of the shape doing the exploit work of finding efficiencies, of finding more structure to make that object more persistent through time. And I think those two dynamics can be used when you think about economics, you think about language, you think about generation and novel structure.

[1:33:37] Ben: Those dynamics are always intention. The thing I'd like about them is they are core dynamics of evolution. And I think that you can combine this sort of evolutionary idea with this recursion to get a language for concepts or ideas or, you know, agents in this type of space and how to define that behavior. And so it's really interesting because I thought the paper Ben wrote with Michael was really like the sort of initial, initial direction on that was really, really, really interesting. Yeah, I find it interesting how I guess a lot of these questions, like to your question, Darren, are just different self-construed, right? There is a self-construal that can be like a cyborg sort of thing. There is control in this market sort of price organization where we are kind of a collective and there's a self there. And I almost feel like that the notion of self is already very culturally conditioned and like the notion of self that I am familiar with in the West, I think is very conditioned by the sorts of markets and capitalist structures and individualism that we have. And so I have the idea that there's me and there's another person and I have ownership of, you know, my things, private property or, you know, even like my body kind of belongs to me, things like that, right? Whereas I think in different cultures, it can look very different. There is a notion of maybe my notion of self is much more identified with my immediate family than it is with this individual body and things. And so I almost would say to your question, Darren, is kind of like, yes, I think it would be very different, but also like it's already the notion of selves is a very diverse landscape, like a much more diverse landscape than we typically assume in the sort of Western academia sort of way of thinking. And so yeah, it'll change drastically, but also that's not that surprising and that's not that new. And I guess the question I also have for everyone is what exactly is the relation between the notion of self-construal and self and the sort of Platonic patterns and the agency of these patterns. I guess self is one pattern that is agentic in some sense. But no, I don't know that relation. I guess clearly. Yeah, I totally agree with you in terms of the differences in the notion of self already culturally or just even, I mean, it's been such a major sticking point even between religions, right? Whether you have essentially like Othman or non-Othman or something, right? Like, and that's a massive sticking point, right? So yeah, I don't doubt it'll be completely different. But I just, I guess I'm just maybe somewhat concerned if basically being replaced pretty quickly when you talked about that, I guess a little bit. So yeah, it's just a concern. Well, but maybe it's like, well, thank goodness, right? I mean, all of Buddhism talks about, you know, the deconstruction of the sense of self, right? And so maybe it's like finally we can replace our stuck Western notion of individualized lonely self with something. Well, yeah. I mean, sounds like you have quite a background on it. Like what are they called? The Skandas or whatever? I'm trying to remember the aggregates that come together within Buddhism that at every given moment sort of make your experience as we bring in. I don't know how that might need to be modified as we bring in tech into a much more integrated way. Or maybe it doesn't even need to be modified. But yeah, I know religious studies is probably not everyone's thing here, so I don't want to derail us. But yeah, that is my background, by the way. Curious. I'm the odd man out completely of the group. Okay, that explains it, because I'm learning much more about Buddhism and Hinduism than I expected.

[1:37:46] Elliot Murphy: In this conversation.

[1:37:48] Ben: Yeah, I'm actually running a community on complexity science and contemplative studies, synergy and consortium. So we're a research community. We're just creating that, so maybe I'll drop a link in the chat if you want to join. I wanted to mention corpus callosum, the split brain, right, that they somehow measured. I don't know how reliable that is, but the sort of bandwidth, right, like certain, I think it's not even that high. It's like some sort of megabytes or gigabytes per second, right? Communication between the two hemispheres. And so if we can get brain-computer interfaces that have that sort of bandwidth just over Wi-Fi, then we can conceivably imagine that the sense of self, just as we don't have clear two senses of self, as long as it's functioning right, the connection, the sense of self could integrate across the Internet and all of that. And so also have that tech-self conglomerate that creates collective reality and dissolves the individual sense of self, which to an individual would feel like death, right? So it's like this great thing that Buddhism always. And that's the other thing, right? Enlightenment and Buddhism, they often say that, well, from the perspective of the self, it is as scary as death. That is why we avoid enlightenment, right? Just once you're there, you're like, oh, it's actually way better, but the self has to die. So yeah, I guess that's what I was asking is if it would feel like a sense of death right at that moment or not.

[1:39:22] Elliot Murphy: Yeah, I know. So the brain works on like 20 watts of energy, and it can compute about roughly 1 exaflop of information based on all the connections in the brain. Current LLMs are slightly above that. But I think we have to be careful about treading into any potential IIT-type domain here of kind of equating the source of informational complexity with any representational format, so to speak, right? Because then if you were to match that equation, then LLMs could have multiple cells all working together, right? I don't know. But yeah, I like what you said about replacing or updating the kind of more traditional westernized self in some way. Like I'm writing something at the moment about worker co-ops, and one of the most interesting places in the United States where you'll find kind of socialist-leaning economic enterprises is in red states, like the public banks of South Dakota. And a lot of the southern states in the United States have a lot of very interesting experiments in economic democracy and worker self-management and forms of economic organization that are not necessarily thoroughly associated with typical westernized, you know, LLC-type structures. But you find it. I don't know if you want to translate this into some kind of a tractor landscape language or whatever, but like you'll find these sorts of pockets of innovation happening in places where you often kind of least expect it. And I find that really interesting, where you'll see moments that you'll otherwise associate with some kind of, you know, non-capitalist, anti-capitalist kind of framework bringing itself into the deep, dark recesses of the Bible Belt, which is really, and not necessarily in, you know, Brooklyn barista, you know, self-owned coffee shops or whatever. It's much more traditional sorts of modes of economic organization. And I don't know if the notion of the self is as causally relevant here in grounding these in negotiating between these economic. Again, I'm using the word economics in just a very, you know, loose sense here. I don't know if the notion of self is as causally relevant when talking about these things in terms of like, does your notion of self have to be... you know, there's all kind of Marxist idea, right, that sound will be would have been familiar with, like your mode of consciousness is like economically, materially determined, right. Is that true to the extent that we would want to connect it to things like agency? Like if I go work for Amazon versus going to work for a workers co-op in South Dakota, would that necessarily change my sense of self, or would it or would it know? Like I don't know if these conversations are as connected as we often think they should be, right? Like you could be an incredibly open-minded, spiritual, wonderful human being working for Amazon, or you could be a total selfish, you know, piece of crap working for a workers co-op or a public bank in South Dakota, right, or whatever. Like I don't know. Or, you know, I'm just trying to say like I don't know if the connection there is as tight as it has to be. Yeah.

[1:42:36] Ben: Well, I think we do have something of an answer to the questions because I think the economy is a really convenient example of this. I'll throw a preprint into the chat here about the scaling of the self across different subjects like economics and biology and stuff. And basically, I mean, what we argue is like a self is a coordinated system that's aligned to some shared goal, like a virtual governor, if you've heard that term. And what seems to be the case, you know, we have this other, it's not published as it's not a preprint anymore, but paper about the economy as a collective intelligence. And so the argument is when you join that you do massively scale up yourself in a sort of very distinctive sense where the price system allows your cognitive light cone to handle an entire class of problems that you wouldn't be able to deal with on your own. And yet the first-person experience of that is basically like nothing. Like theoretically when you go shopping, you join this massive collective intelligence, this incredible hive mind with problem-solving powers that are way beyond anything any individual human can manage. And yet what it feels like is going shopping. So the third-person effects are enormous. The first-person effects at the first order seem almost non-existent, which is very curious.

[1:43:57] Elliot Murphy: That's really interesting. I'll definitely read that.

[1:43:59] Ben: And that makes sense. But it seems like that's an example that makes total sense. I think that would be quite distinct though if we're talking about radically altering the architecture of your body or brain or these interfaces. And then what might occur compared to just engaging with the economy. Do you not think that that would be a fundamental distinction? I don't know how you'd predict that ahead of time. Like as humans who have gone shopping, we understand that its first-person effects seem pretty minimal. But if I was an alien and you looked at a human going shopping and said, "What do you think's going on with this guy's sense of self?" I don't know how we'd know if it's a big effect or not. So maybe that's some kind of useful control or comparison. We know there's a sense in which you can massively scale your cognitive light cone and link to a collective intelligence in a way that basically doesn't seem to affect you personally at all. Yeah. That's interesting. Yeah, it's interesting. Probably counterexample to certain classes of hypothesis. Yeah, in a couple of weeks, I'll be giving a talk about some of this other, the dynamic system stuff I'm talking about with strong anticipation and that example with the baby on the treadmills about how those kinds of ideas show an argument about how cells get scaled across different substrates and problem spaces. So that might have more to share at that point, but I don't have a ton to say about it right now because I'm still working out the slides and stuff. So it seems like still the majority of you, I know we're almost out of time, but maybe the majority, half of the people seem to be of the mindset that whatever we're talking about with this latent space or Platonic space, is it somehow infinite or complete or is that not where most people are at?

[1:46:17] Elliot Murphy: I just had a quick question for Ben. Actually I was interested, like, you know, these theories of consciousness often cue, like, the self versus the content of consciousness versus the state or whatever the different components of the umbrella sense of consciousness, right. To what extent do you think in your work, right, is the relevance of the representational content of experience, right, versus modulations like in the shopping example, right? It's, I would presume that you could execute the same computational demands and goals and instructions for searching for, at the very basic level, searching for items, sifting through a set of options and then choosing it. You can obviously shop online at home where you're not in a social setting, right? The kind of differences between doing the underlying minimally necessary instructions for shopping, quote unquote, versus shopping in different environments like a busy supermarket versus a less busy supermarket versus online shopping where it's just you and your friends maybe or even on your own, right, that kind of spectrum of social expansion. Are you trying to investigate how the notion of self would be modulated by the content, right? Or more the actual, more of a social type of milieu, right?

[1:47:41] Ben: What we've just been doing is a fully third person take on the self. And so the question of your individual sense of self isn't something that I've intentionally tried to research. But it does happen to be the case that because we do all know what it's like to live in the economy and go shopping and things, that it does certainly at least falsify some hypotheses about. Like if you're scaling up your cognitive life, must your sense of self necessarily change radically? Apparently not. If you connect to a collective intelligence, must you share consciousness or even have any awareness you're part of a larger system? Not necessarily, but that's as far as it's gone for me so far. But what I will do is share this other paper on the self. This is a paper also taking a third person perspective, but you might think it's good food for thought, more of this dynamic system stuff, where they basically argue that you can see the formation of an agent in real time. Like you can see a system going from not an agent to an agent as it assembles a basically coordinated pattern of behavior. And that's the kind of third person perspective on selfhood. So what does that feel like on a first person level? But you can study a lot of other things about the formation of a selfhood. And that's what I've been trying to do, yeah.

[1:49:05] Elliot Murphy: In answer to Darren's question, I kind of oscillate between this kind of Timothy Jackson-style driven constructivist account of the space of possible forms and the more classically Platonist, as in actual Plato, position on the space of forms. And depending on which of those you navigate towards will constrain the number of plantaric forms, right, in terms of the scope or size of them and which ones will be relevant across different domains. My instinct is to, so I'm an internalist about philosophy of language and mind. Like I think the language is a mind, internal construction. And when we communicate with each other, we do so not because we share a language, quote unquote, any more than we share a visual system or a digestive system, right? We have our own internal linguistic competence and we communicate because those are sufficiently overlapping, not because we share them and the same of self. Same with agency. Like I'm very much an internalist, which is uncommon in the literature. Like most people tend towards externalism and many people I've collaborated with tend towards externalism, in which case they're a bit more pluralistic about the full landscape of Platonic forms then, essentially, right? They would be much more open to it. I tend to be a bit more reserved there. And I'm trying to think of ways to internalize mathematics as well, right? I don't know if it's possible to do that. I think there's some very convincing arguments from Platonic philosophy of mathematics that would convince me otherwise. So I'm not, I don't have a strong opinion there, but I think that the answer to that question will dictate and constrain what those possible, what the possibility space of forms is. It's either something that is hallucinated essentially from your own world model in the same way that we hallucinate the English language, right? It's not a real thing. It's just there is no English language which is not out there. You can't find it. You can't, you know, touch it, right? It's a convenient abstraction. Other people think that consciousness is to some extent a convenient abstraction, right? All of these things may reduce to convenient tricks that we pull. Mathematics does seem to be unique though. And I agree with a lot of what Mike said in his previous talks too about this, right? It does seem to be a unique case.

[1:51:23] Ben: I agree with Mike and you about the mathematical objects and sort of their necessity, eternality, and whatnot. But I guess what I'm wondering is, what is the purpose of the ingression of all these various things if there's not, I guess, makeup? I don't know if it's really a term, but X aggression from our space back, providing novelty, even if you will, back into this photonic space. I don't really see what the point would even be behind any of it if there wasn't some sort of, I guess, novelty being driven, both like we was talking about bidirectional at the beginning. I don't know. There's this idea, right, that somehow Platonic forms somehow inform or maybe even create what language is for us, right? That they create this set of concepts that we then can share, right? And through that shared language or overlapping language, we can kind of, I don't understand each other, and it's not really communicating information because whatever I say, I never know what you receive, but I can somehow entrain your brain with my brain in some way, right? And then it's interesting when that doesn't work, when we don't have enough of this shared Platonic context, right? We have to sort of go back down a level to the more immediate, to the more basic, right, which, like for apes, right, they have gestural communication or something somehow more embodied, right? That like when I'm just using my tongue to make weird sounds, you have to have a lot of shared cultural context to understand what I'm saying. If I just use my body motion, it's, you know, a little more maybe immediate, more obvious, more, right. And so it seems like, you know, war, for example, right, when you have some negotiations and they don't work because you don't share enough common language or common cultural context, or you don't understand each other, right, you regress to this more primitive form of language of like something we definitely both agree on is we don't want to die and it hurts to get hit. So we can share that context, now we can use that language. So the language of violence being as another language, which I guess also has its own Platonic space associated with it of, you know, the abstract concepts that are communicated through this language of physical violence. And to me, it's almost interesting to view the problem of war as a problem of language, right? I don't know, maybe either breakdown of language, but more interestingly as just like it's another kind of language and maybe it's like manipulating the vocabulary of the language and especially in context of how that informs this abstract, like Platonic space. So let me ask you on bring up war and violence and all this. Presumably there was some first moment where somebody saw a rock and realized, okay, there's the forms associated with the rock, but there was some novel possibility, right, comes to play where this rock could become a weapon, right? That has happened at some point. Originally, did the possibility of there being a weapon, was that necessary 13.8 billion years ago or something like this? Or, you know, was it just like there already? Are we saying it's a possibility, or is that something that is generated somehow in the process? I think it's only violent when you already have a notion of self, and therefore it's violent to the self, right? I think without a notion of self, it's sort of just communication. It's like one cluster of atoms communicating something to another cluster of atoms in fact, right? There's no a priori way of saying that, like what I'm doing now with just saying words is less violent than what, you know, an animal can do to another one with a rock, right? Because from a perspective of atoms, it's just one cluster of atoms influencing another cluster of atoms, right. And so, I mean, what about your shirt? Your shirt looks red. I'm wearing red as well somewhere you probably can't see it, but like redness again, is that something that is, it is redness? Does redness need to be the same way? We're talking about the mathematical objects. That's I guess what I'm asking.

[1:55:47] Elliot Murphy: I think your question is a good one though, Darren, because I also don't know how would we even tell if there was bi directional inference? How would we even know if something that we had done or our universe had done then impacted and shaped or updated or revised the Platonic realm of patterns, right? We wouldn't know because we don't have direct access to it, right. So this it's almost like we kind of set up to fail to answer that question already. I'm not sure how you can even address it.

[1:56:16] Ben: I guess maybe the fact of our continued existence, that's not a good answer, is it? But the might still be useful. I mean, yeah, but you're right, I don't really see how you would.

[1:56:28] Elliot Murphy: Well, you know that this is getting off a little bit here, but the end theory, right? The idea that there's natural selection being applied to universes, the multiverses, different universes survive based on their fitness. The four fundamental forces are calibrated in the same way and ones that are calibrated in a problematic way collapse on themselves or expand exponents, etcetera, in a really problematic way for infinity. So there's a kind of possibility space in that landscape, as the constructor theorists would say. But I don't know if that form of selectional logic is being imposed from some kind of Platonic realm, right? And if it is, I don't know how the success or unsuccess of that multiverse project, like every time universe dies or survives, would that update or revise this Platonic space or would it or would it not? Presumably not, because that's what makes it Platonic.

[1:57:28] Ben: I can see there being more symbiotic and parasitic type universes. Are you saying a multiverse scenario where define such that the ones that are more productive of novelty are less parasitic or something like this as well? Yeah. I mean, even if there's a multiverse, it still doesn't mean it has to be, I guess, complete, so to speak, or infinite. But it begs a lot of the question too, of like if there's a multiverse, when you make something like Michael's making in his lab, something unusual or whatever, are we pulling down forms that were already created in some other multiverse and then updated into this Platonic space? Or is it already this Platonic space? Like we said, it's somehow complete and infinite. I mean, it's, again, can't really answer that. I don't say you would. But they're questions that are interesting. Yeah. Like if you're asking them as questions like what is ultimately true, I guess I don't really believe in such questions more in the like what is observable, what is measurable type of camp. And it seems like there it's sort of a matter of perspective, right? If I divide the universe into this is myself and this is not myself, right? And I have the boundary, then there is a way in which the matching, right? Like the symbols that help myself to survive are going to be the ones that become stabilized in my. So for that, that would give the sense of like Platonic space is just emergent from that, which allows me to survive, right. On the other hand, if I kind of view it from, I don't know, outside perspective without this boundary of self, not self, right. And then which like that's impossible in a way. So I wouldn't, I almost feel like the objective existence of Platonic space that is abstract is kind of contingent on having a view from nowhere which does not really exist. So I don't know. For me, there's like transcendental perspective on that Platonic space existence dependent, independent of physical reality. And I have a hard time understanding how that. Yeah.

[1:59:45] Elliot Murphy: I guess I would say that I kind of think of science as a chance convergence between properties of the mind and properties of the external world, right? They kind of meet in a sweet spot where there's a mode of understanding that we allow ourselves, right, because of our cognitive scopes and limits, right? We can't understand everything. So for me, the platonic stuff could potentially just be where parts where our brain cannot link or interface. Modes of understanding, i.e. mathematical concepts, with other modes of understanding, right? Other modes of explanation, other cognitive modules, right, or language. Language is language of the thought, etc. So maths is encapsulated in the same way that colour is encapsulated for language. So for example, there's no human language that morphologically marks colour features like hue or brightness like we have words for colour. Even though colour is a really important feature of daily life, it's a very salient property of human existence, the language system doesn't seem to care about colour, right? We only have individual words for colour: green, red, yellow. But we don't grammatically inflect colour. But we do grammatically inflect things like tense, right? Future tense, past tense. That's a, that's a architectural property of the language system. It pays attention to tense because it cares about inventive information, but it doesn't care about colour, right? But that doesn't mean that colour is a Platonic thing that we have to explain in some additional way. It just means that these two systems in our brains are not explaining each other. They're not interfacing with each other. So it's one possibility that I would kind of offer and entertain would be that whenever we encounter something that looks like a platonic form, that has existed before the universe came about or whatever, could potentially simply be explained in the sense that it cannot be interfaced or made compatible or reformatted or communicate with other aspects of our science-forming faculty. You know, the ways that our brains do science, right: induction, abductive reasoning, deductive reasoning, whatever, all these different forms of reasoning. Mathematical properties are just unique. And it's not that they're unique because of what they are themselves. They are unique in relation, going back to McGill trick as I did, right? They're unique in relation to other modes of explanation, like different aspects of major movements in 20th century, 19th century science tools that evolved, right? You know, non-dynamics, evolution, computation, these sorts of things that have emerged in the last 200 years that we are just simply able to negotiate between these elements more carefully. And that's good. And then we say, well, we have an explanation because we make principal connection between them. Same aspects of mathematics just might never be able to be grounded or embedded or cashed out in those other systems. They might simply be a unique currency, a unique computational neural currency that we have. And by the way, I want to remind you guys of one very important point. Mathematics has been like a dormant faculty for us for all of human evolution. It was only in the 17th, 18th, 19th centuries that most of mathematics was discovered and figured out, right? You know, and then you got John Conway discovering the surreal numbers and all these sorts of crazy abstract properties of mathematics, precisely because they don't seem to have any applied use cases, that kind of pure math, right? But then it turns out later down the line, a lot of things that Conway explored in the mid-20th century do actually crop up in nature, and you can use them for functional explanatory purpose. So in that case, they seem a little bit less platonic than we had assumed. And there's a certain space of mathematics that is more platonic because it's so disconnected and disjointed that it can't possibly, it has to be from some kind of extra dimension. Maybe not. Maybe time will tell. And if time will not tell, that doesn't mean that there's, we'd have to entertain some kind of dualism. It just means that there are properties of, there are certain mathematical features that our minds can be attuned to because of this convergence between our brains and the actual real world. And that's where that's the real sweet spot, right? I don't know if that makes a huge amount of sense, but that's kind of how I'm trying to wrestle with this very difficult question from my kind of perspective of linguistics, right? And I don't know if that's the right way. I don't know if it's misguided, but that's kind of my intuition.

[2:04:16] Ben: Can see that it's interesting, yeah. I mean, we're all from such different backgrounds that it's fun to even bring everyone together because we're speaking of language, right? We're all speaking slightly different, but that's fun. Different languages, yeah.