 In terms of the math of the conscious agents, we're running what John Maynard Smith, this evolutionary game theory style, evolutionary simulations, and there's several components to it. And correct me if I'm wrong, and then if not, we'll have you elaborate on it. Would the first component being a category theory? And that basically being like the agents themselves and the multi-agent reinforcement learning we could say maybe. Then we have Markovian morphisms, the Markovian kernels. We have the explorations that are happening. We kind of called this the earlier you were talking about like a Monte Carlo tree search and a model-based reinforcement learning. Style for the agents to explore and morph. So the actual sets of the agents, those sets, those category sets they morph over time based on the experiences that the agents undergo. And then there's a replicator equation for procreation. So it's three things. So replicator creation, procreation. And I believe you said that you don't have the quasi-species model yet, which is for mutation, but that these are kind of, would you say that these are kind of the abstract components as the category theory, the Markovian morphisms, the replicator equation, the quasi-species model. Is that approximately in the right ballpark? Well, certainly that's true when we're doing our simulations just of evolution by natural selection, right? So John Maynard Smith transformed Darwin's idea into beautiful mathematics back in the 70s. And this is called evolutionary game theory. And so in fact, I use, I and my graduate students, Justin Mark and Brian Merrin used John Maynard Smith's evolutionary game theory to first run simulations. Before we had the theorems, we ran the simulations just to see if it was worth our time, right? And you just wanna see what's gonna happen before you spend your time on the math, especially when you have limited math talents like me. So we ran the simulations first and there we did find that, you know, organisms that saw reality, virtual organisms in our simulations that saw reality went extinct when they competed against ones that of equal complexity that saw none of reality and were just tuned to the fitness payoff. So that's what led me to then go to a real mathematician, Chaitan Prakash, and pursue the theorems. Now, in the case of the agent dynamics, I'm not wedded to evolutionary game theoretic kind of model of the dynamics of consciousness. I have to be open to a much wider range of possible dynamical systems, right? Interesting. It's in general going to be simply some kind of dynamics on graphs. So the general mathematical arena is these graphs and dynamics on graphs, which is a fairly recent and quite complicated branch of mathematics. Now, one constraint will be that whatever dynamics I come up with the conscious agents, when I project it into the headset into space time, I need to show why it looks like we get John Maynard Smith's specific kinds of evolutionary dynamics in the headset, right? That has to come out, but I don't have to put it in to begin with in the level of conscious agents. And that would actually be more impressive if I had some more general kind of dynamics or some kind of different dynamics. And that then very interestingly, when I project it, looks like evolution by natural selection and evolutionary game theoretic. So right now I'm just pursuing, but category three, I think will be a big part of it, right? Right now we're using Markovian dynamics on graphs, but we're thinking that for the bigger picture of how agents interact and combine to form new agents, we may have to go to a category theoretic representation that's more general than the Markovian dynamics that we're looking at right now. So maybe monoidal categories to begin with. It's so interesting. So the conscious agent theory, the conscious agent dynamics are occurring on graph. And that's kind of the, and then the conscious agents get ascribed that Alan conscious agent gets ascribed a category set of, for example, like your DNA may be in there, something like that. Maybe the morphisms that that agent undergoes may be like we described earlier, this most common morphism of drinking the water, which then enables the agent to live another day because they've imbibed water, versus having the, you know, the Rube Goldberg transcendent idea of a business that they build incessantly obsessively over five years actually gets them a very serious long-term fitness payoff. So, and which gives them a better stance on the hierarchy for in the replicator equation, their genetics will mate with the better partners genetics. Is it taking in that level of processing? Eventually, of course, we're going to, I mean, that's what we see in the interface, right? We see this kind of evolutionary dynamics and we see genes, we see DNA, we see the replicator equation, we see reproductive success and failure. And so those are things that we'll have to show at least are the way from the point of view of our space-time headset, our interface, but that's the way this dynamics looks to us through that. But it may be that it's that that point of view deeply misses a lot. Yes. It really misunderstands what's really going on with this whole realm of conscious agents. So I don't want to in any way restrict my imagination about the dynamics agents to anything, even about sexual reproduction and fitness, the whole notion of fitness and reproduction and so forth may be only an artifact of our headset, just like space-time, space-time itself, the very structure of space-time is an artifact of our headset. It's not a deep insight into reality. So that's the weird thing. And the fun thing about the challenge of this is, so when I go after this theory of conscious agents, and then you're asking exactly the right questions to really expose what it means to try as a scientist to go to a level of theory that goes beyond space-time, right? You have to, on the one hand, be constrained ultimately by empirical things that we can measure inside our interface, like DNA and fitness and survival and so forth. But our ideas can't be constrained necessarily in this deeper realm. We really have to let our imagination go and ask the really tough question, which is first, what kind of deeper theory do we want to go after? So I'm going after a theory of conscious agents and consciousness. Others that I should mention tried to go a different way. I mean, so Seth Lloyd was looking for something beyond space-time and he proposed quantum bits and quantum gates. So that's mathematically precise. And there's this abstract world outside of space and time with these quantum bits and quantum gates. And he was able to show how with each quantum gate, the action of the gate corresponded to the curvature of little patch of relativistic space-time, gravitational general relativistic space-time. And he could get, he could sort of boot up general relativity from this deeper reality. Now you could of course ask him, why in the world should it be that this deeper reality is quantum bits and quantum gates? Who ordered that? And why? Why should that be the deepest reality? And we could scratch our heads about that. And that's a legitimate question. But I'm just saying that once you go and say space-time isn't fundamental, there's countless ways that we can go as scientists to look. And so for me, so it's a non-trivial choice to say, I'm looking for consciousness and my motivation was simply because I think I'm conscious. I think you're conscious. If I'm wrong about that, and I could be wrong about that, then why I'm wrong about pretty much everything. And so I decided I wanted to propose a theory in which consciousness is fundamental and see if I couldn't do the hard work of showing that space-time emerges as just a headset to represent all these, think about it as a big social dynamics, like the Twitterverse. So we're proposing this big Twitterverse of conscious agents. But since it's like the Twitterverse, there's tens of millions of Twitter users, billions of tweets. What do we do is that you can't interact with all the tweets or the Twitter users. We need a visualization tool to see what's trending in New York to LA. Zoom out. What's happening in the United States? What's happening in Europe? What's happening in China? Zoom in. What is Alan tweeting right now? So a good tool would let me zoom in to just Alan, zoom out to what's happening at the whole United States. That's what our headset is. Right now, my headset is representing most of the conscious agents with really dumb symbols, like tables and chairs and so forth, not because tables and chairs are conscious. Just like if I have it in my headset, I've got something that's representing what's trending in the United States. Some red objects are doing something over here in green objects. It's not that red objects and green objects are conscious or Twitter's. That's just how my interface codes the stuff because it has to give up. But I'm not giving up so much when I interact with you. I'm actually interacting with a specific consciousness and you're interacting with my consciousness. And so our headset, in many cases, will just use really crude symbols that we call just inanimate physical objects. That doesn't mean anything deep. It just means that our symbols have to give up. Right, there's no, in this point of view, there's no fundamental distinction between living and non-living. We think it's a fundamental distinction in reality. It's just where our interface symbols are more or less insightful about the conscious agents that we're interacting with. So there's no fundamental principle distinction between living and non-living. And there's no way that life just boots up from the unconscious ingredients of our interface. So this changes the whole point of view. But to get right back to your question, what we have to do in this theory of conscious agents and this dynamics is not assume John Maynard-Smith's theory. We have to get it coming out as a projection of a deeper dynamics. And so we have to ask, and this is the really fun part and the creative part for scientists, you have to ask every theory has certain assumptions. What are, and we want to have as few assumptions as possible, right? Because everything you assume, you're not explaining. So every time you put something into an assumption, you're saying, can't explain that. So that's an assumption. So you want as few assumptions as possible. And so there's no theory of everything, right? There is no scientific theory of everything because every theory has some assumptions that will not be explained by that theory. So there are theories of everything except my assumptions. That's what science had. We have theories of everything except my assumption. And of course, then you can get a deeper theory which tries to explain those assumptions, but it'll have its own new assumptions, right? And so science always will have this limit that we only have theories of everything except my assumptions. And so my theory of consciousness is no exception. I will have a theory of consciousness except my assumptions. And so my assumptions are that there are conscious experiences like the taste of chocolate, the smell of garlic, the headache and that those conscious experiences can affect other conscious experiences. So they inform actions that affect other conscious experiences. That's it. I just want to say those two things. So I mean, so notice I'm not saying anything about a self or self-awareness or self-consciousness. I'm not saying anything about memory or learning or intelligence or problem solving. There's all these things that I'm not saying, they're not part of the kitchen sink that I'm throwing into my theory. These are all things that I'm saying I will have to explain. The only thing I'm going to let my self have is two things. There are experiences and experiences inform actions and the only actions are to affect other experiences. That's it. With those, so that minimal structure, my hope is to say, I can then give you a theory in which we build selves as data structures. We can build theories out of networks. So this is going to be the dynamics of graphs. It's going to be graph dynamics in which we show how to compute things like memory, learning, problem solving, intelligence and so forth. And by the way, it's a theorem from our mathematical statement of the assumptions that I just gave you. There are experiences and they can affect other experiences. By writing it down as Markovian kernels as we have, it's a trivial theorem that this framework is computational universal. And what that means is in principle, there is no limit. Any theory that we could build of learning, memory, problem solving or intelligence using computers, I can build as my theory of conscious agents. If you can do it on a computer, I can build a theory in this new language of conscious agent networks. So I know from the get go that I have an adequate but minimal set of assumptions that lets me build all these theories. I can build a theory of learning, memory, problem solving, a self and so forth out of this. The two assumptions that you made about experience being fundamental and also about experiences impacting each other agents that then impact their actions that those are pretty fundamental to ancient spirituality, perennial wisdoms around the planet which we'll talk about also later. So I appreciate how you say that there's many different ways to up the source code mountain. And okay, yeah, yeah, I like that a lot. There's a lot of like Brian Keating just had Stephen Wolfram and Eric Weinstein on and they were talking about the nature of mathematical reality and they were talking about the hopfibration and they were talking about how that vibrational data is a big deal and it's happening infinitely far away and then the emergence of it. And to me, that style of thinking and visualizing I think is also important. I don't know where that lands on the mountain towards the source code and where that would incorporate a conscious agents or if that would happen downstream or whatnot. So I really appreciate how there's many different ways up the source code mountain and that there's different ways to speak about and visualize it.