Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Consciousness is not Computation — Christof Koch
0 Plays2 seconds ago
Christof Koch is a pioneering neuroscientist and one of the most prominent advocates of a scientific approach to consciousness. He has spent decades working at the intersection of neuroscience, philosophy, and computation.
Christof is one of the foremost proponents of Integrated Information Theory (IIT) — a radical proposal that attempts to explain consciousness in terms of causal structure. IIT begins not with the brain, but with experience itself. It takes as its starting point what is undeniable: that something is happening right now — that experience exists. It then looks at the features of conscious experience, for example, that is unified yet composed of parts, and contentful. From there, it builds a theory describing which physical systems support conscious states.
In this conversation, Christof and I explore what a scientific theory of consciousness might need to achieve, and why behavior alone — even the impressive feats of AI is not enough. Nor indeed is any computational account of consciousness: consciousness is about structure, two structures may lead to the same outcomes, but their form might mean that one is conscious and the other not. But we also touch on experience beyond theory — Christof’s reflections on psychedelic experiences and the dissolution of the self.
This is a conversation about what it means to be a whole, what makes a system truly unified — and what it might take to understand, and perhaps even expand, the field of consciousness. Christof’s latest book is Then I Am Myself the World: What Consciousness Is and How to Expand It.
Recommended
Transcript
The Easy and Hard Problems of Consciousness
00:00:00
Speaker
David Chalmers famously split the problem of consciousness into two parts. The easy problem is the what. What's going on in the world when we're having a conscious experience? The hard problem is the why.
00:00:12
Speaker
Why should there be a conscious experience at all, given that state of things? But of course, even the easy problem is hard enough. My guest this week is Christophe Koch.
00:00:24
Speaker
He spent the 90s and early 2000s working with Francis Crick, co-discoverer of DNA, looking for the neural correlates of consciousness, working on this problem of the what. And fact, in 1998, he had a bet with David Chalmers that within 25 years, the scientific community would have accepted a theory of consciousness. we'd We'd know what the neural correlates of consciousness are.
Integrated Information Theory (IIT) and Koch's Perspective
00:00:49
Speaker
He lost that bet, but huge progress has been made through a series of experiments where we look at what's going on in brains when people are having conscious experiences. Christoph is also one of the main proponents of the integrated information theory of consciousness.
00:01:04
Speaker
This was proposed by Giulio Tonini. And it says that consciousness is not about having some certain computations going on in your brain. It's about the structure of things.
00:01:16
Speaker
And when things are in a certain structure and that structure is able to influence itself and receive influences from the world and influence the world in a particular way that makes it very hard to tease apart the different parts of that structure, then you have consciousness.
00:01:32
Speaker
And they can even put a a number against the amount of consciousness that a particular system has. We also talk about Christoph's book, Then I Am Myself the World, his latest book, where he talks about psychedelic experiences.
00:01:47
Speaker
I'm James Robinson, you're listening to Multiverses.
00:01:58
Speaker
Christoph Koch, welcome to Multiverses. Thank you very much for having me, James. um it's It's my pleasure.
What Should a Scientific Theory of Consciousness Explain?
00:02:05
Speaker
So i wanted to start by asking, well, we don't have a scientific theory of consciousness yet.
00:02:11
Speaker
um There's a few contenders out there and there's lots of philosophical theories. But what is it that we're looking for for in a scientific theory? What what would it give us?
00:02:22
Speaker
What the mental is, how it is different for, well, it's obviously different from the physical. So how does it relate to the physical? you know why is my brain conscious, but presumably this bottle doesn't feel like anything?
00:02:38
Speaker
And then you have to explain why my brain can sometimes feel bored and sometimes have an experience of pain or pleasure or eating you know a slice of pizza or love or orgasm or whatever it is.
00:02:53
Speaker
And it also has to explain this curious fact that that assuming, let's, ah James, when we're doing this interview in the same room, so why are there two and only two conscious entities in this room?
00:03:08
Speaker
And where's the borders between the two? And that's non-trivial because, for instance, I could be a split-brain patient. You could be speaking to me as a split-brain patient. in which In some so ah version, it depends the exact etiology and what exactly they have.
00:03:25
Speaker
But assuming if I have a regular brain and a surgeon an evil surgeon, just to make a point, cuts my corpus callosum, the 200 million fiber separating my left brain and my right brain,
00:03:37
Speaker
Then as far as we can tell from a split brain um ah experiments, there are two conscious entities inside this one skull. One can talk, typically the left hemisphere, but the right one, particularly if you silence the left one by anesthetizing it, can hum, can answer simple yes-no questions, and it's also conscious, but doesn't have that andly that linguistic ah capacity.
00:04:02
Speaker
Furthermore, potentially i could be a woman and pregnant and I carry another brain around with me that at least sometimes, i when it's awake, may also have consciousness.
00:04:14
Speaker
so So we need a theory that tells us, okay, this is one consciousness, over here is the second one and potentially over here is the third one, right?
Detecting Consciousness in Non-Responsive Patients
00:04:24
Speaker
And no one even thinks about this.
00:04:28
Speaker
You say explain, um and and that's kind of a that's a word which can mean many things. Would it be enough for for the theory to predict when there are these different patches of consciousness? Or does it need to go, do we want more from it than that? To be able to say, well, this is why there's something mental there, not merely that we believe there is something mental there.
00:04:57
Speaker
Yeah, I mean, it's a good point. So science in general doesn't isn't doesn't do very well with why question. Why there universe? Why do we have two eyes, you know, rather than four or compound eyes like flies have or eight like some spiders?
00:05:14
Speaker
And yeah, usually you can concord some sort of explanation, you know, evolutionary psychology, all of that, but that's all very iffy. But science does very well with mechanisms and with prediction.
00:05:25
Speaker
So, for instance, I started a company, Intrinsic Powers, that does exactly that for bodies that harbor brains that are severely injured. Let's say you have a cardiac arrest or ah you're involved in a massive car accident.
00:05:40
Speaker
you come to the ICU, you're behaviorally non-responsive. So I ask you, James, James, can you hear me? Can you track your, you know, with your eyes, can you track my finger? If I pinch you, do you have a pain response?
00:05:52
Speaker
Well, if if you have none of those things, and typically... until very recently, people would assume that you're unconscious. But now we know that up to 25% of these patients are actually conscious.
00:06:04
Speaker
So we now have devised a ah procedure where essentially you knock the brain using magnetic pulses and then you measure the reverberation, the response of that brain using EG. Well, we can predict that this patient, although they're not conscious,
00:06:18
Speaker
behaving, they're not speaking, they they are there. there's some There's someone inside that brain who's conscious but cannot signal. So those things science does extremely well rather than saying, well, why is yeah why is there and anything rather than nothing or why why is consciousness?
00:06:38
Speaker
Yeah, it reminds me of, I think it was um Nelson Goodman said something like, well, if we take everything as understood, then there's nothing to explain. But if we refuse to sort of take anything as as obvious, then we can't offer any explanations.
00:06:53
Speaker
I think science sort of sometimes allows us to walk down a ah chain, ah a kind of mechanistic chain of explanation. But at some point, we just end up with bare facts. And it seems that with consciousness, we may not be able to walk very far until we say, okay, well, this is...
00:07:07
Speaker
even if we have a scientific theory of consciousness, it may go very quickly from this is this is what some area of space-time is doing to, oh, and there's consciousness because of that.
00:07:19
Speaker
And you can't really offer more than that, I wonder.
IIT's Axioms and Causal Power
00:07:23
Speaker
Well, so the the the theory I've been most closely associated with is, of course, integrated information theory of consciousness.
00:07:30
Speaker
And that doesn't try to that doesn't try to squeeze the wine of consciousness out of the water of the brain or turn the the water of the brain, you know, via a little miracle into the into the wine of consciousness, right? so And that's what many theorists try to do. they They take the brain, squeeze it really hard and say, oh, it's got to be 40-inch oscillation or it's got to be biochemical synapses or it's got to be global workspace or it's got to be...
00:07:58
Speaker
planning or working memory or any of those things. You could take a different approach, which is what IIT does, and you can say, let's start off with the only thing that I'm absolutely certain of. I mean, this goes back straight back to 1632 or 38 René Descartes, Cogito Argosombe. The only thing I'm truly...
00:08:20
Speaker
100 certain of is that i exist i might be deluded about my body i might believe i'm incredibly attractive to women but that you know could be a complete delusion but i'm not deluded in that i have conscious states and so you could just start with that brute fact and furthermore these conscious states have certain properties they're intrinsically they're very definite whatever my current state is of experience it is something very specific very definite ah it is it is a single experience but it also has structure like my visual experience of space there's left there's right there's up there's down etc etc so i could start with those five fundamental aspects of any conscious experience and then say okay now what is the character of a physical system
00:09:08
Speaker
that can sort of instantiate these five properties. That's how IOT goes about it. yeah so there you don't regress So there you don't have this infinite regress because you start out with consciousness.
00:09:21
Speaker
You don't have this infinite regress. So IOT is is and saying, well, here are some things that if we just reflect on it, like philosophers, like We can sit in our armchair because we don't need to go anywhere to experience consciousness.
00:09:37
Speaker
And we know that you know it has a content. We know it's intrinsic. We know um it sort of doesn't mix with other consciousnesses, if that makes sense. It kind of coalesces.
00:09:48
Speaker
And I can't remember all of the the five kind of tenets. but It's structured. It's structured. And it's one. And it's, okay, you know it's a kind of unity. Right.
00:10:00
Speaker
And... And it's enough just to reflect on those properties to sort of um build up a ah theoretical ah model of what we think consciousness must be like. I mean, I think that's really fascinating.
00:10:15
Speaker
ah But of course, then you need to say, does it actually match what we see in the world? um And so then there's also the empirical aspect of of of looking at the theory. So um maybe you can comment on both.
00:10:28
Speaker
Yeah. So the theory makes specific... So it starts out with... It calls it axioms. It's very axiomatic, which is pretty rare in science. Sort of more closely more closely related to quantum mechanics than anything else, come to think of it.
00:10:41
Speaker
it it says that these five or maybe six axioms, together with the zeros axioms, consciousness exists. And then the other five are, it exists for itself. In other words, it's it's an intrinsic, it's definite, it's specific, it's unitary, and it's it's structured.
00:11:00
Speaker
And then it says, well, so let's look at physical things. How do I know that anything exists? It's this notion of causality, of something exists to the extent that it can exert causal power. This goes back to a comment by Plato.
00:11:17
Speaker
um in ah in the in the dialogue, this ah in one of his early dialogues. And to the so then you use this calculus. So IIT uses this power calculus. Anything that exists, exists by the dint of power. It extends either onto others, like charge, you know, electromagnetic field. If you're a charged particle, you'll be attracted or repelled.
00:11:40
Speaker
if you you know If you have math, you'll be you know you go down the curvature of space-time, of space-time, etc. So that's external ah or extrinsic causal power.
00:11:53
Speaker
And IIT says, well, a causal power upon itself. You have a system, like your brain, it's in a particular state. this system has a certain tendency, a certain probability to go into the next state, it exerts some causal power over its next state. If it wouldn't, your next state would be totally random, just subject to thermal fluctuation.
00:12:13
Speaker
Then it has no causal power. To the extent that it deviates from those causal powers, it has causal power to to go into a particular state and to come from a particular state in its past and to move into a particular state in ah in its future.
00:12:28
Speaker
That is what what intrinsic causal power is. And and the the central sort of theorem of IET says essentially consciousness is nothing but ah system that has maximal cost um maximized intrinsic information whose all of its causal relationships unfolded. And the unfolded causal relationship together, they jointly represent the conscious state of this system at that particular time.
Human Brain Complexity vs. Computers
00:12:57
Speaker
Now, this sounds very abstract, but for any one system, you can explicitly calculate it In fact, there's Python code you can find on the web for to calculate this. And it makes very specific prediction, at least in principle, can be falsified.
00:13:10
Speaker
So it's ah it's it's it's a it is a falsifiable theory. ah So there's a lot to unpack there, and i think worth pausing a moment. So if if I understand the kind of intrinsic causal powers, particularly the part that...
00:13:26
Speaker
ah something is able to act upon itself. and If I kind of contrast that with a ah leaf that's just blowing in the wind, right? it's It's not really determining its past and and and future. It's just being randomly um bashed by particles on either side. And it's got sort of no very low intrinsic causal power.
00:13:47
Speaker
Maybe there's some kind of structure to the leaf that keeps it together. But um whereas I certainly feel that I have some kind of ah control over my future and um ah and therefore some kind of intrinsic causal power. And also, I feel like I'm able to reflect on myself as well. um So not only is my power able to exert itself externally, but I can sort of look inwards and affect the state of my own mind, if that makes sense.
00:14:19
Speaker
um So I feel like those intuitions kind of make sense. um Let me see. ah Can you perhaps explain a bit more about this unfolding of the of the causal powers?
00:14:34
Speaker
What does one mean by that? Well, so IIT does something that goes back all the way to Aristotle. It tries to define exactly, it's it's the age-old question of meriology, what defines the parts of a system to make up a whole?
00:14:55
Speaker
it's ah It's a non-trivial question. What defines a whole? So for example, what defines my body? Do my glasses belong to my body? Well, during the day they do, at night they don't, and of course I can take them off. What about my what about the tooth filling? Is it part of my body? Well, yeah, sort of it is.
00:15:12
Speaker
you know what What about food that I ingest? Well, it's briefly. So so defining, um it defines a whole, which is sort of ah as a ah is defined by the maximum of of intrinsic,
00:15:25
Speaker
um causal ah power, which is quantified by this number that people often bandied about called phi, the Greek letter phi, which is it's a pure number. It can be zero or it can be positive. If it's zero, the system doesn't have integrated information.
00:15:41
Speaker
The bigger it is, this number is, the bigger the integrated information of the system, the more the system is irreducible. So it really tries to say, let's look at a physical system, your brain,
00:15:55
Speaker
a pond out there, a star, any physical system. And we can define all the causal powers that the various constituents have on on each other, the molecules or the atoms or the synapses um or the cells or the the um their columns. You can do it at all possible scales, spatial scales, temporal scales, and organizational scales.
00:16:18
Speaker
And IIT says, well, the whole, that's really the substrate of any conscious experience, is the system that is the maximum of causal effect. So let's say if I take my brain, particularly in my cerebral cortex, it has a particular...
00:16:35
Speaker
ah ah boundary. So it says there is a boundary in my brain, a conceptual boundary and of all the neurons that are maximally integrated. If I want to add more and more neurons to them, well, they're connected, but you're really requiring the entire system to be connected. So it's a little bit like an empire that keeps on growing. And then, of course, the periphery but doesn't hold anymore, famously, as in the poem.
00:17:00
Speaker
The periphery doesn't hold anymore because it's lost its sort of strong central connection. So there's always a trade-off between a system that that has high complexity, but it cannot go indefinite because the the distal parts are much less tightly connected to the central parts.
00:17:17
Speaker
if the system becomes too large. So for any particular system that you can describe, whether be neurons interacting with wires or neurons, sorry, transistors interacting with wires and neurons interacting with axons, you can precisely define if you have a description, an exhaustive description of the system,
00:17:35
Speaker
At what scale is the does the system exist the most? What is the most irreducible part of that system? And that system, per IIT, that system is the one that it is a subset of consciousness.
00:17:51
Speaker
Now, you asked me the intrinsic power. So within that system, you have to look at all the possible interactions of two neurons onto those five neurons. And those three neurons, what happens if they turn on?
00:18:03
Speaker
What effect do they have on those 10 neurons over here? And those 50 neurons over there, if they turn on or turn off, what effect will they have on the rest of the system? So in principle, you have to evaluate all those.
00:18:16
Speaker
So it's combinatorial. It scales as two to the two neurons. you know double exponentially. So it's a vast number of possible relations that you have to that you have to consider. So even a simple system that has, let's say, 10 neurons that can be either on or off has two to the two to the n. In other words, two to the two to the 10.
00:18:39
Speaker
So that's two to the thousand different neurons. different relationships you have to consider. So it very quickly grows hyper-astronomical. But that's not and that's not an ontological problem. That's an epistemic problem. Yeah, it's it's not easy to calculate all of that, but we know in thermodynamics it's not very easy to calculate the partition function of a molecule of gas, right? No one has ever done it with 10 to the 23 molecules, right?
00:19:06
Speaker
um So, that's what the intrinsic causal power. it's It's the sum total of all the interactions within this system.
00:19:17
Speaker
And that gives rise to this vast web of causal interaction. And that is, this is a central assertion of IT, t that is what your conscious experience is identical to with nothing left over. Okay, I think I have it. and the The analogy that's coming to my mind actually is is with computer code and how well-written computer code is very modular.
00:19:42
Speaker
And you've got these different parts that are kind of i easy to separate and consider um apart from the other each each other. And they're interacting through nicely define defined interfaces.
00:19:56
Speaker
and We're saying that for something that's conscious, it's very hard to do that kind of ah separation into different parts. You find the neurons within the brain, which are somehow all working together, though they may be you know separated all over the place, in such a way that it's very hard to... um it's very hard to consider them separately.
00:20:19
Speaker
it Is that kind of the bones of it? Okay, that that's entirely correct for cortex, for the neocortex, the outermost layer of the brain that's maximally expanded in us humans.
00:20:30
Speaker
For the cerebellum, for instance, so it turns out we have roughly 86 billion neurons, at least four middle-aged Brazilian men, 86 billion neurons. But 80% of those are in the little brain at the back of the head called the cerebellum.
00:20:47
Speaker
And we know empirically from clinical literature, so if a small number of people are born without a cerebellum, they have great trouble doing fine movement, they have trouble learning speaking, but they can eventually accommodate to that.
00:21:01
Speaker
In fact, there's a movie made about them. ah But they are they theyre they're clearly conscious. but We know from the clinical literature, if you have a ah glioblastoma or some tumor in your cerebellum, the surgeon has to remove remove it.
The Limits of AI and Simulations in Achieving Consciousness
00:21:16
Speaker
You become ataxic. You become, you you know you can't do this you know, you can't do this very precise anymore. You can't do, you know, you're shaking, etc. But... people don't complain that they lost consciousness.
00:21:29
Speaker
Why not? Well, because the organization of the cerebellum is very, is quite simple. It's only feet forward. There's no massive excitatory feedback that's such a characteristic feature of cerebellum.
00:21:40
Speaker
And the cortex, then as as the cerebellum is organized in these two-dimensional slabs, exactly as you said, modules. In fact, The cerebellum is really a good approximation to a lookup table, to what in computer science would be considered ah a lookup table, which is great for doing you know sensory motor control, but it's not terribly integrated and therefore doesn't appear to contribute to consciousness.
00:22:06
Speaker
so yeah Exactly the reason you mentioned. It's a feed-forward network, whereas the other parts of the brain, the information's going both ways. um And I mean, that in itself is fascinating that such a large part of our brain doesn't seem to be involved in consciousness at all. And I think it's just... Neither does your spinal cord.
00:22:27
Speaker
Neither does most of your body. Yeah. Yeah. but It's weird that, I mean, one thing that just strikes me is why do we need consciousness, right? we can We can do, why has this thing evolved when it seems like, you know, all our fine motor schools and so many other things.
00:22:45
Speaker
um I mean, we can talk about AI, for instance. I don't think AI is conscious. I think you agree with me. um And yet it seems so capable. How is it that, you know, we've got this experience of our abilities and not just the abilities themselves. But we evolved and they didn't. And I'll tell you why.
00:23:04
Speaker
We evolved so we ah natural selection doesn't doesn't select for modular design. So it doesn't select for simple, easy, feedforward design.
00:23:15
Speaker
And if you ever do, I've done this quite a bit at Caltech, when when you simulate brains through natural evolution, you know you you play artificial life. And you get these very, very complicated brains that do simple things like, you know, I simulated this with a ah little and ah ah little creature, a little vehicle running through a maze that had a ah ah brain that was wired up through genes. And then I sent 10,000 of these creatures into this maze and I randomly selected a subset of them, perturbed their genes and selected always ah the the top 10% performers.
00:23:53
Speaker
in terms of how far they can run through the maze. After a while, you get these creatures that just evolved by by natural selection that are very good at recognizing where they have to go.
00:24:03
Speaker
But if you look at their brains, they're incredibly complicated because, you know, The selection is is haphazardly. It searches you know in a very high-dimensional space randomly. right There's no overarching designer that says, well, you have to go this way.
00:24:19
Speaker
okay So this gives rise to this enormously complex network. And a feature of these complex networks is that they will have high integrated um information.
00:24:30
Speaker
So in that narrow, in the sense stricto, in the strict sense, consciousness is not what's selected for. Because consciousness, ultimately what what's selected for is behavior, right? in In this simple computer simulation that that I played, what was selected for, how far does every generation of these little animates, these little computer simulations, how far did they make it through the maze?
00:24:52
Speaker
What gets selected for in real life, do you leave descendant, you know? you will have enough DNA? At the end of the day, do you have offspring, right? That's what gets selected for. So ultimately, it's it's behavior. It's not conscious experience.
00:25:05
Speaker
But the feature of the universe that we live in is that highly complex system that have these highly nonlinear ah causal interactions that they are, that this is exactly what is expressed as as conscious experience.
00:25:21
Speaker
Now, computers are very different because we build them, we design them, we want simplicity. Otherwise, we can't figure it out. So, in fact, if you look at the at the at the CPU, at the, you know, the actual, the place where the rubber of computation hits a physical road, you know, you're you're looking at one output of one gate, you know, being connected to the gate to a gate of a couple of other transistors. And this is, you know, and ah repeated across billions of transistors, right? So, and we can cascade them. So ultimately they're very powerful and they give rise to LLMs.
00:25:54
Speaker
But at the level of the wiring, it's very, very simple, right? It's modular. It's highly simple. It's very low input. fan in but the the number of inputs to a transistor in the part of the brain of the computer that does a computation is three or four or five. And the number of of transistors any one transistor outputs to also three or four or five. In the brain, it's like 50,000, 100,000. So that's why, unless we build very different computers, today's computers, although they can...
00:26:28
Speaker
soon can do everything we can do, they will never be what we are, which is conscious. I think this point is really worth emphasizing because it it will seem to many people, um perhaps because it's against the prevailing opinions, that
00:26:47
Speaker
very odd that a computer could do everything that we can do and could conceivably behave exactly as we behave and yet not be conscious. So can you perhaps explain how IIT predicts and makes sense of that?
00:27:06
Speaker
You can imagine a computer simulation of a human brain. right Sooner or later, we'll get there. i mean, right now, we can simulate a tiny quinoa grain piece of brain.
00:27:18
Speaker
Ultimately, in the fullness of time, we can simulate a human brain, you know including all the neurons, the synapses, et cetera. And of course, this computer, if it's done correctly, will wake up and and speak, because that's what humans do. So if you connect it to a synthesizer, of course, it'll speak.
00:27:38
Speaker
But again, in in illrit it'll speak like you and I and will speak, particularly if it's been trained on you know all the world's literature, like an LLM is. Of course, because it's been trained by us to read Tolstoy, War and Peace, and every possible novel, of course it's going to talk about love and hate and despair and and all the other you know existential conscious ah experiences that and that people talk about.
00:28:04
Speaker
But that's very different. So that's all behavior. that But that's all function. But that's not consciousness. Consciousness is more than just function. I can sit here not doing anything, not talking at all, but I can be highly conscious.
00:28:17
Speaker
You can do experiences when you sleep tonight. You're going to go to sleep. You'll be paralyzed. Your brain will partly paralyze your body, yet you have conscious experience, right? You dream. You can be tripping.
00:28:28
Speaker
You can be meditating where you sit on a cushion, on your meditation cushion, Without moving at all, then you can have profound experiences. So for us, it's not about doing, it's about being.
00:28:42
Speaker
So the the computer consciousness right now, it's all a deep fake. Unless we build computers in the image of a human brain. So this is called neuromorphic engineering.
00:28:54
Speaker
Very few companies are trying to do it. It's much harder and it's not clear why you should because you know digital computers are just so powerful. But in principle, you know this brain is a piece of furniture of the universe like any other.
00:29:08
Speaker
So in principle, you know a computer could also be conscious. it's it's not There isn't anything supernatural like a soul, like we used to think. I don't think there's a soul that's uniquely attached to us.
00:29:21
Speaker
But you need the right you need the right hardware if you want. You need the right substrate. And digital computers, as currently conceived, don't have that substrate. I do think this is a real strength of IIT that...
00:29:36
Speaker
you know there is this intuition that consciousness and behavior should be separable. And many people, I think, in recent times have kind of tried to deny this and say, well, all there is to consciousness is is behavior.
00:29:48
Speaker
And yet, I just need to think about my myself and understand that's not the case, right? it's I can behave very differently from the way that I'm i'm feeling or I can you know move on autopilot and not be really experiencing things to the fullest or I can be dreaming and not behaving at all and and yet have these very rich experiences as as you point out.
00:30:10
Speaker
um There is this kind of internality to to consciousness that you know it's certainly possible to conceive of someone who has no or a thing that has no conscious experience and yet behaves like one.
00:30:23
Speaker
um And so I think it's really refreshing to have a theory which speaks to that and says, yes, consciousness is not just um functions or behavior. um And if I understand it correctly, it's It's because IIT says, well, consciousness is a future of the structure of the system. It's not just the input-output. It's causal powers. Correct. Nothing to do. In fact, IIT says there is, in principle, no Turing test.
00:30:48
Speaker
So Turing test is famously a test for intelligence. You know, for for can you fake it, right? That was the original Turing test, right? The imitation game of Alan Turing, right?
00:30:59
Speaker
Can I put some entity, you know, ah in ah in a room and then by teletype, that was the dominant technology, they interact with it. And after a while, I can't say it's a human or machine. Then at least i have to admit, well, it seems to be doing things like humans do, right?
00:31:14
Speaker
And clearly LLMs can do that. we We can agree that and they're getting better and better. Yeah. And so there there are two reactions to that. So some people still believe, and I think the dominant ethos in Silicon Valley and among technology is that, of course, machines will be conscious,
00:31:33
Speaker
we we will be conscious right? and Anything else is just it's just organic chauvinism, right? because they can do everything we can do and just listen to them, right? They can pontificate endless.
00:31:47
Speaker
Well, of course, because they've sort of read everything that humans can can can ah you know have produced. um But then there's also now a growing unease. You see it in philosophers even where they say, wait wait a minute, now that these LLMs are around,
00:32:03
Speaker
Yeah, they just seem to behave like it, but they aren't really conscious. So now, in fact, you can see in the field, there are some people who say, well, we must go to first principle. Maybe they need metabolism.
00:32:16
Speaker
Maybe they need something like life, you know, sort of almost back to vitalism. Because people i like you, you're saying, well, yeah, it can't just be behavior.
00:32:29
Speaker
um and But IT has always been there. It's not about behavior. it's not Consciousness is not just a clever hack. Nothing do with computation. Intelligence, that is about computation. And that, as we see in LLMs, you can instantiate in Turing machines.
00:32:48
Speaker
But states are being – that's why it doesn't get wet inside a simulation of a rainstorm.
00:32:55
Speaker
That's why you can simulate a black hole at the center of our galaxy and you do not have to be afraid that once you turn on the simulation, you'll be sucked into into this simulated black hole.
00:33:07
Speaker
I think that's right. I mean, I think it's interesting because most simulations are like that. They're just a simulation that we understand models the thing that misses out the essence of it.
00:33:19
Speaker
um There are some kind of simulations, like if I simulate a calculator, what I have is still a calculator. The simulation is itself a calculator, but that really only works if you buy into this kind of computational functionalism.
00:33:34
Speaker
Yes. Yeah. yeah Yes. It's an abstraction. You abstract it to its its function. And then, yeah, the function could be instantiated in in different ways. But ah pain is not just an abstraction.
00:33:48
Speaker
You know, my particular toothache, that's not an abstraction. That's as concrete as it gets. Just like a black hole is not just an abstraction. it can actually bend space-time around it, right?
00:34:01
Speaker
Yeah, so you can simulate particles inside the computer simulation will be sucked into it. But that's all internal to the computer simulation. Here in the real world, it does not have causal power because you can't simulate causal power, right?
00:34:17
Speaker
That's what it comes down to. You cannot simulate causal power.
00:34:22
Speaker
when we're talking about ah While we're talking about simulations, um this does IIT sort of... either lend credence to or perhaps reduce the the force of Boston's simulation argument. if i think of this and There's kind of different ways you can run the simulation argument, um but maybe one way of running it is to say, well, you know some higher intelligence has decided to simulate a human civilization or particular civilization. Yeah, yeah.
00:34:53
Speaker
And for for me, if I were to simulate a civilization, I would, as you say, um prioritize kind of computational efficiency. And I wouldn't design in consciousness.
00:35:05
Speaker
If I wanted to know you know how the stock market was going to perform, in my I would you know build a sort of mini model of society. And I would build it as simply as possible. And presumably, consciousness would not be um a feature of that model.
00:35:19
Speaker
Simply because, as you said, the most direct route to building intelligent behavior when designed um leads to lead to a system which doesn't have high integrated information.
00:35:32
Speaker
Look, I don't consider the simulation hypothesis to be worthwhile of serious debate. I mean, we can talk about, and people did this in the 12th and 13th century scholastic, how many angels can can can you know can be located on a pinhead.
00:35:48
Speaker
ah This used to be a discussion among scholars. And this is about as useful as that, yes. Okay. because And then, you know, then everything happens at the next universe up. Then you just regress to problems.
00:36:02
Speaker
And then, of course, you can see famously, well, it's simulation all the way down, right? Rather than then a turtle standing on an elephant and then it's elephant or turtles, I forgot which one, all the way down. You say, oh, it's simulations all the way down.
00:36:15
Speaker
I think it's turtles all the down to elephants. yes Or maybe that. I mean, yes, these are sort of fun arguments you can have, but they're totally useless in the ah in the real world. So I've never really considered to any extent the simulation a hypothesis very seriously.
00:36:33
Speaker
I simply don't find it interesting. I think um philosophy has got to be allowed to roam over ah onto subjects which aren't of any ah importance to the real world. Oh, yeah, that's fine. If they want to do it, that's that's perfectly fine to keep themselves busy and and you know paid and all of that. But, yeah, I don't find it a serious problem.
00:36:54
Speaker
At least when it claims to intersect with anything talking in this universe about this particular physics and this particular state of consciousness, I find it utterly uninformative.
00:37:06
Speaker
Yeah, it's fun. I do find, though, the and so your your comment earlier that um you know if we had been designed, um we wouldn't have consciousness, which is just... um I think that's that's that's really interesting. Well, if we would have been designed... Well, it depends what design criteria you use. I mean, if you want to mimic...
00:37:26
Speaker
In neomorphic architecture, you could go do it, but engineers, they want things that are simple, that you can replicate a trillion times, that are highly reproducible, where you can understand it intellectually, you can simulate it.
00:37:37
Speaker
And all of that predisposes us to build simple things that we can then cascade, you know, like a Boeing 777 triple that has, what, two million parts or something like that, right? Otherwise, we couldn't build it.
00:37:53
Speaker
yeah So that argument also holds and and for computers. And I want to also come back to, you mentioned earlier that you in in practice, it's very hard to calculate um the integrated information of a system because there's just so many ah interactions that one needs to consider.
Neural Connectivity and Consciousness
00:38:12
Speaker
um Nonetheless, do we kind of For a given system of neurons or of nodes or whatever, um do we know what the ah structure is that would produce maximum integrated information?
00:38:27
Speaker
Or is it that there's not a single structure, but many ones? How does that look? Okay, so yeah, you can use just like you did in physics, ah you know, the turn of 19th century when they had these, um you know, they had all these functional approximation to thermodynamics, all these Taylor expansion series.
00:38:45
Speaker
Yeah, you can you can make approximation, you can exploit symmetry, for instance, symmetry consideration. You can connect and ah exploit certain types of bipartition where you could put up upper boundaries on this.
00:38:58
Speaker
Yeah, so, and Tononi has done some explicit work on this. So, for instance, If you have a connectivity that is um topographic, where you have lots of connections in all 2D or maybe even 3D to your neighbors, and this type of connectivity falls off as you go farther and farther apart from any one particular neuron.
00:39:19
Speaker
okay It's a topographic connectivity of the sort that you see in visual cortex and auditory cortex and somatosensory cortex because it reflects the structure of the environment.
00:39:31
Speaker
There's lots of interactions locally, but the farther apart you have, the more sparse the interaction is. If you compare this sort of grid-like topographic interaction with something that's more typical that you...
00:39:43
Speaker
ah find up in the front part of the brain, which is a little bit more random access, where you seemingly have one connection here and then another connection there, another connection there.
00:39:54
Speaker
You can do more abstract connectivity and more abstract computation. The topographic ah connectivity ah yields much higher phi values than sort of this random access um connectivity.
00:40:10
Speaker
And I think that would explain why Even the cortex proper, in fact, there's a big nature paper coming out a couple of months from now about this.
00:40:21
Speaker
Even in cortex proper, it seems to me that the back of the but the back of cortex that we call the posterior heart zone is much more closely tied to conscious and conscious experience than the front, which again is more closely tied to intelligence, to abstract reasoning, to moral reasoning, to social interactions, e etc.,
00:40:43
Speaker
So certain types of connectivity, even if you give me, okay, you say I give you 50 wires for each neuron, right? And I can now ask, well, I can put these 50 wires either just to to um to connect nearby neurons or I can use them to connect random neurons.
00:41:00
Speaker
Then in terms of phi, the the local connectivity gives much higher rise, ah ah it gives rise to higher ah values, much, much higher values of integrated information than the much sparser long distance connectivity.
00:41:17
Speaker
And I guess that makes sense because with the kind of long distance connectivity, you can sort of, you can kind of imagine splitting the the network into two parts, which are kind of, you know, goes loosely, more or less loosely linked.
00:41:29
Speaker
um You mentioned earlier as well that um networks with feedback rather than just, you know, feed forward have much higher integration information.
00:41:41
Speaker
What about- In principle, so this is this is an interesting argument that has been used against IIT. So in principle, for any fixed architecture, you can unfold ah ah ah feedback network into series of purely feedforward networks.
00:41:56
Speaker
okay i mean, this was already recognized early in the 20th century. you can You can sort of equivalent, you have a function of functions and you can map that out into a simple series of just purely feed-forward functions.
00:42:12
Speaker
what's What's remarkable about that, so now you have two networks that functionally are exactly identical. yet yet So in terms of input-output function, they do exactly, they take the same input into the same output.
00:42:26
Speaker
Yet one is heavily feedback, the other one is purely feedforward. And so per IIT, one would feel like something, the other wouldn't feel like anything. And this really weirds some people really out. They find this impossible to believe and therefore they think IIT can't be right because how can it be that two things do exactly the same thing but supposedly have very different experiences up to this one having no experience whatsoever in that?
00:42:53
Speaker
that one having experience. But that's precisely because consciousness is not about doing. It's not a function. yeah It doesn't map an input into an output like a Turing machine does. That's great if you want to build LLMs or predict the stock market and and all those things.
00:43:08
Speaker
But if you want to if you want to design a system that is pain and pleasure and is hungry, you don't get that. Yeah, that that makes absolute sense to me. It's it's a strength of IIT that... um I think so. But...
00:43:22
Speaker
Right. Couldn't it be more? It just makes sense that this, that the consciousness is some kind of, ah
00:43:31
Speaker
It's how the stuff gets done, right? It's it's not what gets done. It's that feeling that accompanies the things happening. and And yeah, it makes sense that the same things can happen.
00:43:43
Speaker
And in one case, there's no feeling, there's no experience. um And all I can see that could explain that is a different structure that produces the same um sets of behavior. So yeah, I i think that- In fact, there's another counterintuitive.
00:43:57
Speaker
So there's a related counterintuitive prediction. um So, A, people, many functionalists believe, you know, if the system does the same input-output, either they're both conscious or none of them are conscious.
00:44:11
Speaker
There's this other prediction. If you take a piece of brain that is, let's say, a piece of your brain or my brain, in principle, it is quite possible, in fact, I think this might happen during certain meditative states of pure presence, that this piece of brain has no activity whatsoever,
00:44:29
Speaker
So no neurons are firing in the limit, let's see. Yet it is it experiences something. Because, again, experience is not sending information from this part of the brain from that part to that part of the brain that's apparent to an external observer.
00:44:46
Speaker
So I, looking down at this brain, if I were a neurosurgeon or neuroscientist, I'd say, well, there's nothing being signaled. But remember, signaling involves a sender and a receiver, but the brain creates meaning you know for itself. There isn't a sender in the brain that says, hey, receiver, I'm going to send you information, right? No, it has to create the meaning the meaning totally for itself.
00:45:09
Speaker
And now you make this other prediction. Now I take a brain where i point by injecting, let's say, lidocaine, a local anesthetic agent, I can prevent the pain the brain from firing.
00:45:21
Speaker
So now I have two brains. This is a normal brain that doesn't fire. This is a brain that's prevented from firing. So both look exactly the same. I look at the two brains, they're not firing.
00:45:34
Speaker
Yet this one is conscious and this one isn't. Why? Well, because this one could fire, but it chooses not to fire. It's like the Sherlock Holmes stories, you know, the dog um that could bark but didn't bark. And that turned out, Sherlock Holmes tells that to the inspector Lestrade, that is the critical evidence.
00:45:51
Speaker
Well, this one cannot fire because it's prevented. for It has lost its causal power because the anesthetic. This one... has it has a full complement of causal power, but in its current state doesn't fire.
00:46:06
Speaker
So again, IIT says this is highly conscious. It might, in fact, approximate the state of pure presence in long-term meditator, while this one doesn't feel like anything. Yeah. yeah I mean, I think, as you say, many people will will find that counterintuitive, that just the potential to do something ah it sort of has a strong effect. But I mean, others will feel like, well, actually, this is quite familiar in physics. And I think I mentioned this when when we briefly chatted last night, that the Aranov-Bohm effect, this very weird quantum mechanical effect where you have some charged particle and it's not moving through an electric field, it's not moving from a magnetic field, but there is a magnetic potential that it's going through. And that changes, it's ah you know that can shift the favors phase of its wave function.
00:46:55
Speaker
um And so, yeah, this this seems kind of strangely analogous to this. I mean, not directly, but it's just you get your head around just how weird physics is.
00:47:06
Speaker
um And maybe that's a good thing. which is Which opens another can that we don't have to discuss. And, you know, we can go down ah the rabbit hole here.
00:47:19
Speaker
You know, given some of these very strange effects in and quantum mechanics, the one you mentioned, and also, of course, entanglement, right belts, ah Bell's inequality and ah violation of Bell's inequality.
00:47:35
Speaker
Is there possible, you know, does this is somehow, this is sort of independent of IIT, right? is there Is there a link at really a deep level? Because, you know, it's you know quite it's been difficult to get rid of the observer in in quantum mechanics, right?
00:47:51
Speaker
ah You know, if you think about it conceptually, so Is this an entryway possibly for consciousness? And in fact, there are indeed quantum mechanical versions of IIT that people have proposed. I don't know what the status of that is.
00:48:07
Speaker
Yeah, i think i mean i think i I think you can get rid of the observer from quantum mechanics. I'm sort an everyetian in that sense. but um and And so I don't think there's...
00:48:21
Speaker
ah For instance, I don't think there's non-locality in the sense of you know action at a distance. um ah And and you know Bell's theorem show that you know there's no single feature of the world which kind of determines the state of electrons. But actually, if you permit the world to branch, then you kind of get around that.
00:48:42
Speaker
um Yeah, but then also you get lots of other problems. How do you define probability? Yeah. right yeah I buy into all that. Lots of other things. You don't get an infinite number of universes for free.
00:48:58
Speaker
I'm happy with an infinite number of universes and the price is great. I know, I know, i know, I know, but hetot but you know, anyhow. do think there are some interesting things here. One is, you know, the...
00:49:15
Speaker
The entanglement, ah i can see how that boosts the complexity of something and perhaps you know um i can imagine how the integrated information of a system with lots of entanglement going on with would would really be pretty high.
00:49:29
Speaker
um And the other thing is, I mean, there are While I don't believe there's this kind of non-local action at a distance, I do think ah quantum mechanics shows there is a kind of non-separability to space-time. And again, that's from the Aronoff-Bohm effect rather than the Bohm's other experiments.
00:49:47
Speaker
And so, by the way, this relates to our earlier point about the difficulty of defining truly what a system is, because these entangled pairs, right, they clearly cannot be considered. They're irreducible, right, in some very real sense.
00:50:00
Speaker
yeah Yeah, I think that's right. I think that's right. you can't You can't give a complete description of um space-time um that is sort of decomposable into two separate parts, right? Correct. Or better put, you can't just take description of this bit, description of this bit, and put them together. They're irreducible.
00:50:23
Speaker
Exactly. and so this So one way to think about IoT, it's a way to quantify what exactly a hole is. you know, a whole, an entire system is.
00:50:34
Speaker
And it puts very precise borders on it. ah So in principle, you can have these three neurons, although they also seem to connect it to each other. Those two may be one group, maybe one, maybe a whole, but not the third one.
00:50:49
Speaker
So again, it yeah, it's it's a nice way to think about are these systems interconnected, including entanglement, if you if you have a quantum mechanical version of IoT. Yeah, yeah.
00:51:00
Speaker
yeah
00:51:02
Speaker
Wow. Okay. i I was initially hoping to spend sort of like, I don't know, ah a third of this podcast talk about IIT, but it's just too fascinating. so But I do want to like, um before you go, um i just want to ask you a couple of ah other questions about your kind of personal experiences of expanding consciousness.
00:51:23
Speaker
um This is again, like a completely different route, but It is consciousness related, so I think
Personal and Mystical Experiences of Consciousness
00:51:31
Speaker
it's relevant. and um so yeah Can you tell us just a little bit about what made you interested in not just the science the pursuit of a scientific theory of consciousness, but how to expand your own consciousness um and you know your successes or or otherwise with that?
00:51:54
Speaker
Well, I mean, so when you deal with consciousness, you've got to take the motto of the Royal Society of England seriously and trust no one, take no one's word for it, right?
00:52:07
Speaker
ah And you've got to experience it yourself, right? are If you tell me about it, I hear you, but I don't have that experience, right? It's famously impossible to directly convey an experience, except by saying, well, this experience is like that other experience. But if I never had that class of experience, it's impossible to to convey.
00:52:28
Speaker
so um you know, a couple of years ago, as I got more and more interested in in these, is it really true that you can expand consciousness? What would that mean? How could that be fitted into things like iot I had a number of occasions to try different, attend different ceremonies.
00:52:51
Speaker
And during one of them, i yeah i had a near-death experience, which was absolutely terrifying. I'll i'll never do that again. But there you can, for the first time i ever experienced, well, something that I'd early on experienced. So so i I used to be a passionate dancer, ballet and classical rock and roll. And what I loved about it is when you're really entwined with another person, you know, when you're dancing with a woman, you do this for hours, you can completely lost yourself lose yourself. You can go into these complete states of flow.
00:53:28
Speaker
You know, where you're highly active, you don't you hardly drink anything because you you don't want to. It's not necessary. But you just dance, you know, figures, put the, you know, the the woman on your shoulders or twirl around your hips and all of that.
00:53:42
Speaker
And you completely lose yourself. Same thing. I used to do rock climbing when you're on the wall on the sharp end of ah of of the rope. You know, you go up a big wall. You can completely lose yourself in the here and now.
00:53:53
Speaker
And that chatter, that constant chatter, that constant voice in your head that constantly criticizes and criticizes other people, all of that is totally gone. And I always found those states very addictive and very liberating. And very you you you you come back and you're very happy just having experienced this.
00:54:13
Speaker
And, and, um, but then I had this near death expense where you have something similar, but more profound where, because you can always in those other kids dancing and, and, you know, climbing, et cetera, in principle, you can snap out of it, right? Fairly quickly.
00:54:28
Speaker
Right. And you can stop and, and then you're immediately back with your usual self in your head. But doing this near death expense, there's no one there. No one home anymore, no memories, no body, no external world, no space, no time. It wasn't too slow. It wasn't too fast. It simply was.
00:54:47
Speaker
Yet whatever remains is highly conscious in this case of off this bright, this incredible bright light, this this light at the end of the tunnel or whatever, and this feeling of terror and ecstasy.
00:55:01
Speaker
And that stays with me. I mean, this is not many years ago, but it stays, missed it's ever present. But that wasn't an – mean, it was a radical different state that that affirmed what i always believed, that self-consciousness. So most people, if you ask them, you say, what's consciousness? They say, oh, it's the ability of humans to think about themselves.
00:55:23
Speaker
I think that's wrong. That is clearly a conscious state. I can reflect upon myself what I had for breakfast, where, you know, at some point i'll be in the future I'll be dead, et cetera. but But that's just one out of many states, right?
00:55:36
Speaker
and And animals, my my dog who's around here somewhere, ah my dog is is not terrible self-conscious, if at all. Yet he's, you know, my dog is a highly conscious creature. and And so it affirmed that, yes, you can be highly conscious without there being an external world or any sense of me.
00:55:53
Speaker
There wasn't Christoph, there there wasn't a body, there wasn't, there was none of that. There were no memories, there was none of that. Yet there was still a conscious experience.
00:56:04
Speaker
But that's not sort of expanded. I did have a sort of a different experience sitting at midnight on a beach in in Brazil where I became literally...
00:56:16
Speaker
but and I use that as the title of my last book, I became one with the world. i I had this, what people call a mystical experience, where again, you lose sense of yourself, of your body, and you just experience sort of a world at large. And since the boundary between you and the world is gone, you become the world. The world and you are one and the same thing. And that's truly a complete expansion of Yeah, now you've expanded to encompass everything there is.
00:56:49
Speaker
I mean, that's pretty that's pretty sweet. And is there some sort of reason to think that IIT would regard that as having a higher phi value, you know being more conscious?
00:57:02
Speaker
Or is this kind of pure speculation? yeah No, so that would be the obvious explanation that under certain conditions, drug-induced or meditation or during other states, you can...
00:57:15
Speaker
and that probably leads to a minimum of activity in certain parts of the brain, particularly the posterior cingulate and what some people call the default mode network, where you have representation of self, if all those activity are sort of shut down or dramatically reduced,
00:57:35
Speaker
it mainly because now it may involve that that the the substrate, the one that maximizes phi, may be vastly larger or much larger than before. And given this 2 to the 2 to the n increase in ah in phi, you know, if suddenly you double the number of neurons that are involved in the substrate, 2 to the 2 the 2, you know, that grows...
00:57:59
Speaker
you know ah and that grows so So yes, it's entirely possible that during that state, your phi is vastly larger than your phi in your regular state.
00:58:11
Speaker
Because it is true that many people throughout history and And in fact, it's a common language you hear if you you know if you hear talking about about ah certain trips that people take on mushroom, that I've expanded my consciousness.
00:58:26
Speaker
And so I think this this common observation, I think, it's reflected yeah in could be reflected in this very measurable thing that you can actually measure.
00:58:38
Speaker
and And it's because the sort of the the voice of the ego, I think you describe it as the kind of gravitational pull of the ego. I really like that in your book. um Because that silenced, loosed, silenced,
00:58:52
Speaker
there's sort of a wider network which is involved becomes involved in consciousness. is I think it's absolutely necessary. You need to still that constant me, me, me, me. It's always about me. And of course, in our current government in this country, you see this.
00:59:08
Speaker
It's all about me. um And yeah, when that shuts down, then you can hear these much more widely distributed voices. And you can hear, you can be conscious of everything else and the vast, you I mean, this vast universe that we live in and it's profound beauty and you get this the state of transcendence.
00:59:30
Speaker
It's very magical and it'll stay with me till the day I
00:59:35
Speaker
I've never really considered before that cogito ergo sum, you know, is not the most basic statement that you can make. Perhaps it should be there is experience, um you know therefore there is something.
00:59:48
Speaker
Yes. Yeah. And in some sense, I mean, so ah if you go back to ah to Descartes, he doesn't just mean ego. He doesn't mean self-conscious because he also means and um something that senses, that hears, that sees. So he means all these forms. So it's not just in a sense that me, I have self-consciousness, although that's the the most, you know, Kogan expression we know of, congito a I think, therefore I am. Yes.
01:00:15
Speaker
Ultimately, there is experience. Yes. Well, I think this has just been a wonderful tour and i'm I'm so glad that we, um well, i I hope we did justice to IIT. I know it's a very involved topic, but I think we've given a flavor of it to to people.
01:00:31
Speaker
But I'm also really happy that we've managed to ah to cover this and you expressed your experience so beautifully. Yeah. it As you say, these are transformational things. You can't really... They're hard to communicate if you've not had them. You have to expand them, yeah.
01:00:49
Speaker
It's a little bit like there is this... In philosophy of mind, there is this ah Mary, the color scientist. So it's a little bit like... like Well, or... Another example I like to use, imagine you never dream and no one ever dreams.
01:01:03
Speaker
And then once in your life, you have this dream, you have this amazing dream that you that you fully consciously remember. what I mean, how would that be? You would say, well, yeah, I was alive. I was flying. I was seeing my long dead friend or lover or pet or whatever it is. It was there. It was as real as anything else.
01:01:22
Speaker
But because you do it all the time, people say, yeah, that's just a dream, right? but but But it has, i mean, as i mean William James talks about it, right, the father of American psychology, these experiences, once you've had them, they have this very powerful authority over you, and you can't deny them. You can't just say, well...
01:01:40
Speaker
You know, it's just my brain. Well, no, this state this state that I experienced was so profound, so transformative, ah and associated with such such state of grace. It's really an act of unearned grace. I mean, Huxley says that. He used the word unearned grace, which which I love, to have experienced this state.
01:02:01
Speaker
mean, ultimately, all our knowledge of the world comes via experience, and so it's it's very difficult to dismiss any experience, especially one that that's really profound. Correct. All right. Let's leave it with that, James. Thank you very much.
01:02:15
Speaker
Thank you very much. Yeah. Absolute pleasure.