Transcription
What is it? What's the role of consciousness in, um, let's say, even just the thing we've been talking about, which is the formation of this interface? Um, but any kind of ways you want to kind of start sure tracking and talking about it?
Well, let me say first what most of my colleagues say, 99% are again assuming that space-time is fundamental, particles in space-time matter is fundamental, and most are reductionist. And so the standard approach to consciousness is to figure out what complicated systems of matter with the right functional properties could possibly lead to the emergence of consciousness. That's the general idea, right? So maybe you have to have neurons, maybe only if you have neurons, but that might not be enough. They have to have certain kinds of complexity in their organization and their dynamics, certain kind of network abilities, for example.
So there are those who say, for example, that consciousness arises from orchestrated collapse of quantum states of microtubules in neurons. So this is Hameroff and Penrose have that's kind of. So it's, it's, you start with something physical, a property of quantum states of neurons, of microtubules in neurons, and you say that somehow an orchestrated collapse of those is consciousness or conscious experiences. Or Integrated Information Theory, again, you start with something physical, and if it has the right kind of functional properties, it's something they call Phi, with the right kind of integrated information, then you have consciousness.
Or you can be a panpsychist, Philip Goff, for example, where you might say, well, in addition to the particles in space and time, those particles are not just matter, they also could have, see, a unit of consciousness. And so, but once again, you're taking space and time and particles as fundamental, and you're adding, um, a new property to them. See, consciousness. And then you have to talk about how when a proton and neutron, where proton and electron get together to form hydrogen, then how those consciousnesses merge to or interact to create the consciousness of of hydrogen and so forth.
Um, there's Attention Schema Theory, which again, this is how neural network processes representing to the network itself, it's attentional processes that could be consciousness. Um, there's Global Workspace Theory and Neuronal Global Workspace Theory. So there's many, many theories of this type. What's what's common to all of them is they assume that space-time is fundamental. They assume that physical processes in space-time is fundamental. Panpsychism adds consciousness, says an additional thing, is almost dualist in that regard.
And my attitude is, our best science is telling us that space-time is not fundamental. So why is that important here? Well, for centuries, deep thinkers thought of earth, air, fire, and water as the fundamental elements. It was a reductionist kind of idea. Nothing was more elemental than those, and you could, you could sort of build everything up from those. When we got the periodic table of elements, we realized that, um, of course, we want to study earth, air, fire, and water. There's combustion science for fire, there's, you know, there's sciences for for all these other things, water and so forth. So we're going to do science with these things, but but fundamental? No, no. If you're looking for something fundamental, those are the wrong building blocks. Earth has many, many different kinds of elements that project into the one thing that we call earth. If you don't understand that there's silicon, that there's iron, that there's all these different kinds of things that project into what we call earth, you're you're hopelessly lost. You're not fundamental, you're not going to get there.
And then after the periodic table, then we came up with quarks, leptons, and gluons, the particles of the standard, you know, the standard model of physics. And and so we actually now know that if you really want to get fundamental, the periodic table isn't it. It's good for chemistry, it's just wonderful for chemistry, but if you're trying to go deep fundamental, what is the fundamental science? That's not it. You're going to have to go to quarks, leptons, and gluons and so forth.
Well, now we've discovered space-time itself is doomed. Quarks, leptons, and gluons are just irreducible representations of the symmetries of space-time. So the whole framework on which consciousness research is being based right now is doomed. And for me, these are my friends and colleagues that are doing this. They're brilliant. They're absolutely, they're they're brilliant. I, my feeling is, I'm so sad that they're stuck with this old framework because if they, they weren't stuck, like with earth, air, fire, and water, you could actually make progress. So it doesn't matter how smart you are, if you start with earth, air, fire, and water, you're not going to get anywhere, right?
Can I actually just, uh, because the word doomed is so interesting, let me give you some options. Multiple choice quiz: Is space-time, we could say, is reality the way we perceive it, doomed, um, wrong, or fake? Because doomed just means it could still be right, and we're now ready to go deeper. It would be that. So it's not wrong, it's not a complete deviation from a journey toward the truth, right? It's like earth, air, fire, and water is not wrong. There is earth, air, fire, and water. That's a useful framework, but it's not fundamental, right?
Well, there's also wrong, which is they used to believe, as I recently learned, that George Washington was the president, the first president in the United States, was bled to death, uh, for something that could have been easily treated because it was believed that you can get, actually, I need to look into this further, but I guess you get toxins out or demons out, I don't know what you're getting out with the bleeding of a person, right? But so that ended up being wrong, but widely believed as a medical tool. So it's also possible that our assumption of space-time is not just doomed, but it's wrong.
Well, if we believe that it's fundamental, that's wrong. But if we believe it's a useful tool, that's right. But he could see, but bleeding somebody to death was believed to be a useful tool. It wasn't just not fundamental, right? It was very, I'm sure there's cases in which bleeding somebody would work, but it would be a very tiny, tiny, tiny percentage of cases. Uh, so it could be that it's wrong, like it's a side road that's ultimately leading to a dead end, as opposed to a truck stop or something that you can get off of.
My feeling is, is not the dead end kind of thing. I think that the, what the physicists are finding is that there are these structures beyond space-time, but they project back into space-time. And so space-time, when they, when they say space-time is doomed, they're explicit. They're saying it's doomed in the sense that we thought it was fundamental, it's not fundamental. It's a useful, absolutely useful and brilliant data structure, but there are deeper data structures like cosmological polytube. And and space-time is not fundamental. What is doomed in the sense that is wrong is reductionism, which is saying space-time is fundamental, right? Right. The idea that that somehow being smaller in space and time, or space-time is a fundamental nature of reality, that's is a, that's that's just wrong. It turned out to be a useful heuristic for for thermodynamics and so forth, and in several other places, it, reductionism has been very useful, but that's in some sense an artifact of how we use our interface.
Um, yes, you're saying size doesn't matter. Okay, this is very important for me. Ultimately, ultimately, right? It's useful for for theories like thermodynamics, and also for understanding brain networks in terms of individual neurons, and neurons in terms of chemical systems inside cells. That's all very, very useful. But but the idea that we're getting to the more fundamental nature of reality, no. You, when you get all the way down in that direction, you get down to the quarks and gluons. What you realize is what you've gotten down to is not fundamental reality, just the irreducible representations of a data structure. That's all you've gotten down to. So you're always stuck inside the data structure.
So you seem to be getting closer and closer. I went from neural networks to neurons, to neurons to chemistry, chemistry to particles, particles to quarks and gluons. I'm getting closer and closer to the real. No, I'm getting closer and closer to the actual structure of the data structure of space and time, the irreducible representations. That's what you're getting closer to, not to a deeper understanding of what's beyond space-time.
We'll also refer, we'll return again to this question of dynamics, because you keep saying that space-time is doomed, but mostly focusing on the space part of that. It's very interesting to see why, why time gets the bad cred too, because how do you have dynamics without time? Is the thing I'd love to talk to you a little bit about, but let us return. Your brilliant whirlwind overview of the different theories of consciousness that are out there. What is consciousness if outside of space-time? If we think that we want to have a model of consciousness, we as scientists then have to say, what do we want to write down? What, what kind of mathematical model are we going to write down? Right? And if you think about it, there's lots of things that you might want to write down about consciousness. A lot was fairly complicated subject.
Um, so most of my colleagues are saying, let's start with matter or neurons and and, um, say what properties of matter could create consciousness. But I'm saying that that whole thing is out. Space-time is doomed. That whole thing is out. We need to look at consciousness qua consciousness. In other words, not as something that arises in space and time, but perhaps it's something that creates space-time as a data structure. So what do we want? And here again, there's no hard and fast rule, but what you as a scientist have to do is to pick what you think are the minimal assumptions that are going to allow you to boot up a comprehensive theory. That is the trick.
So what do I want? So what, what I chose to do was to have three things. I said that there are conscious experiences, feeling of headache, the smell of garlic, um, experiencing the color red. There are those. So that's a primitive theory. And the reason I want few primitives, why? Because those are the miracles of the theory, right? The primitives, the assumptions of your theory are the things you're not going to explain. Those are the things you assume. And those experiences, you particularly mean there is a subjectiveness to them. That's what's the thing when people refer to the hard problem of consciousness. Is it feels like something to look at the color red? Okay, exactly, right? It feels like something to have a headache or to feel upset to your stomach. It feels like something. And so, though, so I'm going to grant that in this theory, there are experiences and they're fundamental in some sense. So conscious experience, so they're not derived from physics, they're not functional properties of particles, they are sui generis. They exist just like we assume space-time exists. I'm now saying space-time is just a data structure. It doesn't exist independent of conscious experiences.
Sorry to interrupt once again, but should we be focusing in your thinking on humans alone, or is there something [Music] about in relation to other kinds of organisms that have a sufficiently high level of complexity, or even, uh, or is there some kind of, uh, generalization of the panpsychist idea that all consciousness permeates all matter outside of the usual definition of what matter is inside space-time? So it's beyond human consciousness.
Human consciousness, from my point of view, would be one of a countless variety of consciousnesses. And even within human consciousness, there's a, there's countless variety of consciousness within us, right? You have your left and right hemisphere, and apparently, if you split the corpus callosum, the the personality of the left hemisphere and the religious beliefs of the left hemisphere can be very different from the right hemisphere, and their conscious experiences can be disjoint. One could have one conscious experience. They can play 20 questions. The left hemisphere can have an idea in this mind, and the right hemisphere has to guess, and it might not get it. So, so even within you, there is more than just one consciousness. It's lots of consciousnesses. So I, the general theory of consciousness that I'm after is not just human consciousness. It's going to be just consciousness. And I presume human consciousness is a tiny drop in the bucket of the infinite variety of consciousnesses. That said, I should clarify that the black hole of consciousness is, uh, the the home cat. I'm pretty sure cats lack, uh, is the embodiment of evil and lack all capacity for consciousness or compassion. So I just want to leave that on the table, but that's the theory I'm working. I don't have any good evidence. It's just a shout out. Uh, sorry to distract.
So that's the first assumption. First assumption. That's right. The, the second assumption is that these experiences have consequences. So I'm going to say that conscious experiences can trigger other conscious experiences, somehow. So, so really, in some sense, there's two basic assumptions. There's some kind of causality. Is there's a chain of causality? Does this relate to dynamics? I'll say there's a probabilistic relationship, okay? Um, and then, so I'm trying to be as non-specific to begin with and see where it leaves me. So what I can write down are probability spaces. So probability space, which contains the conscious experiences that this consciousness can have. So I, I'll call this a conscious agent. This technical thing. Now, Annika Harris and I've talked about this, and, and she rightly cautions me that people will think that I'm bringing in a notion of a self for agency and so forth when I say conscious agent. So I just want to say that I use the term conscious agent merely as a technical term. There is no notion of self in my fundamental definition of a conscious station. There are only experiences and probabilistic relationships that of how they trigger other experiences. So the agent is the generator of the conscious experience. The agent is a mathematical structure that includes a probability measure, a probability space of possible conscious experiences, and what in a Markovian kernel, which describes how if this agent has certain conscious experiences, how that will affect the experiences of other conscious agents, okay, including itself.
But you don't think of that as a self? No, there, there is no notion of of a self here. There's no notion of of really of an agent. But is there a locality? Is there an organization? So this is, this is, um, these are conscious units, conscious entities, but they're distinct in some way because they have to interact. Well, so here's the interesting thing. When we write down the mathematics, um, when you have two of these conscious agents interacting, they, the pair satisfy the definition of a conscious agent. So they are a single conscious agent. So there is one conscious agent. Yeah, but it, it has a nice analytic decomposition into as many conscious agents. That's a nice interface. It's a very useful, uh, scientific interface. Yeah, it's, it's a scale-free or if you like, a fractal-like approach to it, in which we can use the same unit of analysis at all scales in in studying consciousness.
But if I want to talk about, so there's no notion of learning, memory, problem-solving, intelligence, self, agency. So none of that is fundamental. So, and the reason I did that was because I want to assume as little as possible. Everything I assume is a miracle in the theory. It's not something you explain, it's something you assume. So I have to build networks of conscious agents. If I want to have a notion of a self, I have to build a self. I have to build learning, memory, problem-solving, intelligence, and planning. All these different things, I have to build networks of conscious agents to do that. It's a trivial theorem that networks of conscious agents are computationally universal. That's trivial. So anything that we can do with neural networks or, you know, automata, you can do with networks of conscious agents. That's trivial. But, but you can also do more. The events in the probability space need not be computable. So the Markovian dynamics is not restricted to computable functions because the very events themselves need not be computable. So, so this can capture any computable theory. Anything we can do with neural networks, we can do with conscious agent networks. But, but it leaves open the door for the possibility of of non-computable, um, interactions between conscious agents. So we have to, to, if we want a theory of of memory, we have to build it. And there's lots of different ways you could build. We've actually got a paper, Chris Fields took the lead on this, and he, we have a paper called Conscious Agent Networks, where where Chris takes the lead and shows how to use these networks of conscious agents to build memory and to build primitive kinds of of of learning.
But can you, uh, provide some intuition of what conscious networks, network of conscious networks of conscious agents helps you? First of all, what that looks like, and I don't just mean mathematically, of course, maybe that might help build up intuition, but how that helps us potentially solve the hard problem of consciousness, right? Or is that baked in? That that exists? Can you solve the hard problem of consciousness, why it tastes delicious when you eat a delicious ice cream, with, uh, networks of conscious agents, or is that taken as an assumption?
The standard way the hard problem is thought of is we're assuming space and time and particles or neurons, for example. These are just physical things that that have no consciousness, and we have to explain how the conscious experience of the taste of chocolate could emerge from those. So that's the typical hard problem of consciousness. Is that problem, right? How do you boot up the taste of chocolate, the experience with the taste of chocolate from neurons, say, or the right kind of artificial intelligence circuitry? How do you boot that up? That's that's typically what the hard problem of consciousness means to to researchers. Notice that I'm changing the problem. I'm not trying to boot up conscious experiences from the dynamics of neurons or silicon or something like that. I'm saying that that's the wrong problem. My hard problem would go in the other direction. If I start with conscious experiences, how do I build up space and time? How do I build up what I call the physical world? How do I build up what we call brains? Because I'm saying consciousness is not something that brains do. Brains are something that consciousness makes up. It's one among the experience, it's an experience, an ephemeral experience in consciousness. I look inside. So to be very, very clear, right now I have no neurons. If you looked, you would see neurons. That's a data structure that you would create on the fly, and it's a very useful one. As soon as you look away, you garbage collect that data structure, just like that Necker cube that I was talking about on the piece of paper. When you look, you see a 3D cube, you create it on the fly. As soon as you look away, that's gone.
When you say "you," you mean a human being scientist right now? That's right. More generally, it'll be conscious agents, because as you pointed out in my asking for a theory of questions only about humans? No, it's, it's consciousness, which human consciousness is just a tiny sliver. So, but you are saying that there is, that's the useful data structure. How many other data structures are there? That's why I said, you, human, if there's another earth, if there's another alien civilization and doing these kinds of investigations, would they come up with similar data structures? Probably not. What is the space of data structures? I guess is what I'm asking.
My my guess is that if consciousness is fundamental, consciousness is all there is, then the only thing that mathematical structure can be about is possibilities of consciousness. And that suggests to me that there could be an infinite variety of consciousnesses, and a eventually small fraction of them use space-time data structures and the kinds of structures that we use. There's an infinite variety of data structures. Now, this is very similar to something that Max Tegmark has said, but I want to distinguish it. He has his Level IV multiverse idea. He thinks that mathematics is fundamental, and so that's the fundamental reality, and since there's an infinite variety of endless variety of mathematical structures, there's an infinite variety of multiverses in his view. I'm saying something similar in spirit, but importantly different. There's an infinite variety of mathematical structures, absolutely. But mathematics isn't the fundamental reality in this in this framework. Consciousness is. And mathematics is to consciousness like bones are to an organism. You need the bones. So mathematics is, is not divorced from consciousness, but it's not the entirety of consciousness by any means. And so there's an infinite variety of of conscious consciousnesses and and and signaling games that consciousnesses could interact via, and therefore worlds, your common worlds, data structures that they can use to to communicate. So space and time is just one of an infinite variety. And so I, I think that what will, what we'll find is that as we go outside of our little space-time bubble, we will encounter utterly alien forms of conscious experience that we may not be able to really comprehend. In the following sense, if I, if I ask you to imagine a color that you've never seen before, does anything happen? Right? Nothing happens. Nothing happens. And that's just one color. I'm asking for just a color. We actually know, by the way, that there are apparently there are women, um, called tetrachromats, who have four color receptors, not just three. And Kimberly Jameson and others who've studied these women have good evidence that they apparently have a new dimension of color experience that the rest of us don't have. So, so these women are apparently living in a world of color that you and I can't even concretely imagine. No man can imagine them. Yeah, and and yet they're real color experiences. And so in that sense, I'm saying, now take that little baby step. Oh, there are women who have color experiences that I could never have. Well, that's shocking. Now take that infinite. There are consciousnesses where every aspect of their their experiences is like that new color. It's something utterly alien to you. You, you'd have nothing like that. And yet these are all possible varieties of conscious experience.
When you say there's a lot of consciousnesses, it's a singular consciousness basically, the set of possible experiences you can have in that subjective way, as opposed to the me, the underlying mechanism? Because you say that, you know, having, uh, extra color receptor, extra, you know, ability to have new experiences, that's somehow a different consciousness. Is there a way to see that as all the same consciousness, the subjectivity itself?
Right, because when we have two of these conscious agents interacting in the mathematics, they actually satisfy the definition of a conscious agent. So, in fact, they are a single conscious agent. So, so in fact, one way to think about what I'm saying, I'm postulating with my colleagues, Jayton and Chris and others, Robert Brentner and so forth, there is one big conscious agent, infinitely complicated. But fortunately, we can, for analytic purposes, break it down all the way to, in some sense, the simplest consciousness, which has one conscious experience. One, I like this, this one agent can experience red 35. That's it. That's what that's what it experiences. You can get all the way down to that.
So you think it's possible that consciousness, whatever that is, is, uh, much more, is fundamental, or at least much more in the direction of the fundamental than is space-time as we perceive it? That's the proposal. And therefore, what I have to do in terms of the hard problem of consciousness is to show how dynamical systems of conscious agents could lead to what we call space and time and neurons and brain activity. In other words, we have to show how you get space-time and physical objects in entirely from a theory of conscious agents outside of space-time, with with the dynamics outside of space-time. So that's, that's, and I can tell you how we plan to to do that, but, but that's that's the idea.
Okay, the magic of it, the chocolate is delicious. So, so there's a mathematical kind of thing that we could say here, how it can emerge within the system of networks of conscious agents. But, um, is there going to be at the end of of the proof, why chocolate is so delicious? Or or no? I guess I'm, I'm going to ask different kinds of dumb questions to try to sneak up. Oh, well, that's the right question. And when I say that I took conscious experiences as fundamental, what that means is, in the current version of my theory, I'm not explaining conscious experiences, where they came from. That's the miracle. That's one of the miracles. So I have two miracles in my theory: there are conscious experiences like the taste of chocolate, and that the, there's a probabilistic relationship when certain conscious experiences occur, others are more likely to occur. Those are the two miracles. It's possible to get beyond that and somehow start to chip away at the miracle-ness of that miracle. Chocolate is still delicious. I hope so. I've got my hands full with what I'm doing right now, but, but the, I can just say at top level, how I would think about that, that would get at this consciousness without form. This, this is going to be really, this is really tough, because it's consciousness without form versus the various forms that consciousness takes, for the experiences that it has. Right, right.
So when I write down a probability space for these conscious experiences, I say, here's a probability space for the possible conscious experiences. Right? It's just like when I write down a probability space for an experiment, like I'm going to flip a coin twice, right? And I want to look at the probabilities of various outcomes. So I have to write down the probability space. There could be heads, heads, heads, tails, tails, heads, tails, tails. So you, before as any class in probability, you're told, write down your probability space. If you don't write down your probability space, you can't get started. So here's my probability space for consciousness. How do I want to interpret that structure? The structure is just sitting there. There's going to be a dynamics that happens on it, right? Experiences appear and they disappear, just like heads appears and disappears. So, so one way to think about that fundamental probability space is that corresponds to consciousness without any content, the infinite consciousness that transcends any particular content.
Well, do you think of that as a mechanism? As a thing? Like the the rules that govern the dynamics of the thing outside of space-time? Isn't that, if you think consciousness is fundamental, isn't it essentially getting like it is solving the hard problem, which is like from, where does this thing pop up? Which is the mechanism of the thing popping up, whatever the consciousness is, the different kinds? It was so on that mechanism, and also the, the question I want to ask is, uh, how, how tricky do you think it is to solve that problem? You've solved a lot of difficult problems throughout the history of humanity. There's probably, uh, more problems to solve left than we've solved by like an infinity. But along that long journey of intelligent species, how, when will we solve this consciousness one, which is one way to measure the difficulty of the problem?
So I'll give two answers. There's one problem I think we can solve, but we haven't solved yet, and that is the reverse of what my colleagues call the hard problem. The problem of how do you start with conscious experiences in the way that I've just described them in the dynamics and build up space and time and brains. That I think is a tough technical problem, but some principle solvable. So I think we can solve that. So we would solve the hard problem, not by showing how brains create consciousness, but how networks of conscious agents create what we call the, the symbols that we call brains. So that, that I think.
But does that allow you to, so that's interesting, that's an interesting idea. Consciousness creates the brain, not the brain creates consciousness. But does that allow you to build the thing? My guess is that it will enable unbelievable technologies. Once, and I'll tell you why I think it plugs into the work that the physicists are doing. So this theory of consciousness will be even deeper than the structures that the physicists are finding, like the amplitude. But the other, but the other answer to your question is less positive. I, as I said earlier, I think that there is no such thing as a theory of everything. So that, I think that my, the theory that my team is working on, this conscious station theory, is just a 1.0 theory. Theory. We're using probability spaces and Markovian kernels. I can easily see people now saying, well, we can do better if we go to category theory, and we can get a deeper, perhaps more interesting. And then someone will say, well, now I'll go to topos theory. And then they'll be, so I imagine that there'll be, you know, conscious agents 5, 10, 3 trillion, 0.0. But I think it will never end. I think ultimately this question that we sort of put our fingers on, of how does the formless give birth to form, to the taste, the wonderful taste of chocolate, I think that we will always go deeper and deeper, but we will never solve that. That, that in some sense, that will be a primitive. I hope I'm wrong. Maybe I'm, maybe it's just the the limits of my current imagination. Um, so I'll just say my imagination right now doesn't appear that deep. Hopefully. So I don't, by the way, I'm saying this, I don't want to discourage some brilliant 20-year-old who then later on proves me dead wrong. I, I hope to be proven dead right. Just like you said, a century from now, everything we're saying now, everything you're saying, all your theories will be laughing stock. They will respect the, uh, the the puzzle-solving abilities and how how much we were able to do with so little, but, uh, outside of that, it will all be just, uh, the silliness will be entertainment for a teenager.