Audio Player
✓ Using synced audio (timestamps accurate)
Starting at:
Michael Levin Λ Karl Friston Λ Chris Fields on The Meta Hard Problem, Consciousness, and Babbling
June 29, 2022
•
2:08:26
•
undefined
Audio:
Download MP3
✓ Synced audio available: Click any timestamp to play from that point. Timestamps are accurate because we're using the original ad-free audio.
Transcript
Enhanced with Timestamps
300 sentences
17,738 words
Method: api-polled
Transcription time: 125m 57s
The Economist covers math, physics, philosophy, and AI in a manner that shows how different countries perceive developments and how they impact markets. They recently published a piece on China's new neutrino detector. They cover extending life via mitochondrial transplants, creating an entirely new field of medicine. But it's also not just science they analyze.
Culture, they analyze finance, economics, business, international affairs across every region. I'm particularly liking their new insider feature. It was just launched this month. It gives you, it gives me, a front row access to The Economist's internal editorial debates.
Where senior editors argue through the news with world leaders and policy makers in twice weekly long format shows. Basically an extremely high quality podcast. Whether it's scientific innovation or shifting global politics, The Economist provides comprehensive coverage beyond headlines. As a toe listener, you get a special discount. Head over to economist.com slash TOE to subscribe. That's economist.com slash TOE for your discount.
This is Marshawn Beast Mode Lynch. ProjPic is making sports season even more fun. On ProjPic, whether you're a football fan, a basketball fan, it always feels good to be ranked. Right now, new users get $50 instantly in lineups when you play your first $5. The app is simple to use. Pick two or more players. Pick more or less on their stat projections.
anything from touchdown to threes and if you write you can win big mix and match players from any sport on prize picks america's number one daily fantasy sports app prize picks is available in 40 plus states including california texas
Florida and Georgia. Most importantly, all the transactions on the app are fast, safe and secure. Download the PricePix app today and use code Spotify to get $50 in lineups after you play your first $5 lineup. That's code Spotify to get $50 in lineups after you play your first $5 lineup. PricePix. It's good to be right. Must be present in certain states. Visit PricePix.com for restrictions and details.
Listen up, people. Over the years, I've learned how important hydration is in my workouts and everyday life. It's the key to helping my body move, recover, and just have a good time when I'm exercising and staying active. Things go even better when I'm well hydrated before I even start moving.
Noon Hydration doesn't want you to wait to hydrate. They want you to start hydrated. Noon Sport Hydration tablets hydrate better than water alone. Just drop them in your water, dissolve, and enjoy. They contain five essential electrolytes and come in crisp and refreshing flavors like strawberry lemonade, lemon lime, and many more.
They also have non-GMO vegan and gluten-free ingredients and only one gram of sugar. Noon hydrates you so you can do more, go further, and recover faster. And that means you can have a heck of a lot more fun. Since hydrated humans have the most fun, head over to ShopNow on NoonLife.com and pick some up today so you're ready for anything. Because anything can happen after Noon.
I won't let my active psoriatic arthritis joint symptoms define me. Emerge as you. Tremphia gousselkumab is proven to significantly reduce joint pain, stiffness, and swelling in adults with active psoriatic arthritis. Some patients even reported less fatigue as assessed by survey one week prior. Results may vary. Tremphia is taken by injection six times a year after two starter doses at weeks zero and four.
Serious allergic reactions may occur. Trimphia may increase your risk of infections and lower your ability to fight them. Before treatment, your doctor should check you for infections and tuberculosis. Tell your doctor if you have an infection or symptoms of infection including fever, sweats, chills, muscle aches or cough. Tell your doctor if you had a vaccine or plan to.
Emerge as you. Learn more about Tremphia, including important safety information, at Tremphia.com or call 1-877-578-3527. See our ad in Food and Wine magazine. For patients prescribed Tremphia, cost support may be available.
As always, you can click on the timestamp in the description as well as here to skip this longish introduction. The Theories of Everything channel is back with a new season, starting with last week's AMA, which will be in the description. From this point forward, there will be new content approximately once a week, sometimes even more frequently. Today's guests are Carl Friston, Michael Levin, and Chris Field. The former two have been on this channel and mentioned several times prior, whereas Chris Fields is making his debut.
Chris Fields is a researcher in topological quantum field theory as well as information theory and publishes work on what constitutes an observer. Carl Friston is a professor of neuroscience at the University College London and is the inventor of the free energy principle. Michael Levin is a developmental and synthetic biologist at the Tufts University. The links to all their previous podcasts
on the Toe channel will be in the description. My name is Kurt Jaimungal. I'm a Torontonian filmmaker with a background in mathematical physics dedicated to the explication, quote unquote, of the variegated terrain of theories of everything, primarily from a theoretical physics perspective, but as well as trying to understand
What role does consciousness have as constitutive reality? Is it emergent? Is it fundamental? Now the goals of theolocutions is to interject seldomly, if at all, allowing each guest to give their perspectives on one another's thoughts, essentially giving us the experience of a fly on the wall for the sorts of academic conversations that would ordinarily occur behind closed doors, spurring research in real time. One of the central issues of today's podcast is the concept of babbling, which I haven't heard discussed
Virtually anywhere else, outside of child development or language acquisition, though it can be generalized metaphorically, perhaps even literally, to the vacuum fluctuations, to interpret what vacuum fluctuations are doing as an indicator of the universe's proto-consciousness. That is, the universe babbling to understand itself, but I won't spoil the surprise. If you'd like to hear more podcasts like these, then do consider going to patreon.com slash Kurt Jaimungal as the patrons and the sponsors are
The only reasons that I'm able to do this full time. With regard to sponsors, there's one for today's episode, and that is Brilliant. During the winter break, I decided to brush up on the fundamentals of information theory, which is what constructor theory is heavily based in, and I'd like to do an episode on that. So I took Brilliant's course on knowledge and uncertainty and random variables, and after taking that course, I could finally see why entropy is defined the way it is,
why the formula for it is extremely natural. There are plenty of courses. You can even learn group theory, which is what's being referenced when you hear that the internal symmetries of the standard model is U1 cross SU2 cross SU3. Those are Lie groups. Visit brilliant.org slash toe to get 20% off the annual subscription. I recommend that you don't stop before four lessons. Just keep pursuing until you've accomplished at least four. And I think you'll be greatly surprised
At the ease at which you can now comprehend subjects you previously had a difficult time grokking. And now on to today's episode, which is one that I'm extremely lucky to be able to present to you all. You'll be able to see the mutual respect that the guests have for one another. We start off actually by asking, what is it that you respect about each other's work? Enjoy this theolo-cution between Carl Friston, Michael Levin, and Chris Fields. I've been extremely excited about this.
Professor, what do you find unique about Carl's and Chris's work and what is it that you respect about them? Well, what a great question. I think all discussions should start with that question. There's a lot. I mean, I think the most basic thing I can say is this. I'm often asked by conference organizers and students as far as who they should follow, whose work they should keep track of.
And so I keep lists. And Carl and Chris are at the very top of two specific lists for me. And I'm just always amazed about this is that the first is that both of them are experts in at least three different fields, but probably more. So they have the ability to kind of merge really deep understanding of different fields. I mean, there's many people that are experts in one thing,
But I'm just always incredibly impressed at how both of them can weave together really deep knowledge from diverse disciplines. And that leads to the kind of the second thing that I'm really inspired by, which is just the kind of the sheer density of new ideas. You know, it's hard enough to push things in one just kind of one known direction and do science and so on.
Both Chris and Karl's work consistently make me think in new ways, give me ideas I never thought of before. It's just amazing. I'm super excited to be able to work with both of them and to have this discussion. Great. Chris, same question but toward Michael and Karl. Well, I can say that one thing I certainly admire is that
Michael and Carl are both top-level biologists who are interested in theory and deeply involved in theory and have a deep understanding of theory. And something that I enormously admire is that both of them have that interest and yet are deeply engaged in practical applications.
Carl in mental health and Mike in regenerative biology. And that's a rare combination. Carl, what is it that you find unique about Michael's and Chris's work and what do you particularly respect about them as people perhaps? I deliberately haven't rehearsed an answer. So what I say comes from the heart. So with Mike,
I first met him vicariously through a friend Giovanni Pizzullo and didn't really know very much about his work until I visited his website and then I realised how influential and important he was. And then I recognised all the little bulletins you get on social media and emails. I started to recognise his name and realised that he was quite a mover and shaker out there.
What I like about both of them, they both think really, really fast and they both think out of the box. So I think what characterizes both of their thinking really is just to ask what is a thing in and of itself without making any prior assumptions and then putting it back together from basic principles and coming up with some sometimes counterintuitive
conclusions. But for me, almost universally, exactly the same counterintuitive conclusions I've come to after about 10 years of thinking about a particular problem, but never dare tell anybody. But they do. They write it down very eloquently. They're very, very productive.
Chris I've known for less long and I've met him via Mike and in addition to that in the same way that many people I meet have physics envy and I've got quantum envy about Chris. He just seems to know so much and think about the world in a way which I don't have that sort of fluency
What would be an example of this out of the box thinking? You mentioned thinking about what is a thing in and of itself. What else? There are numerous examples. The first thing that just came to mind was just putting together
fundamental questions about self-organization. If you were Varela, you know, sort of autopoiesis, self-assembly, self-construction, if you're a chemist, it would be self-assembly. If you're a theoretical biologist, you will be wondering about why on earth is it that multicellular organisms form? Because there's a deep paradox there. If it is a case that I have to have a surface as a little organ,
Then some of my cells have to stop replicating, and yet that is in direct contradiction to the principles of natural selection. I'm going to turn off my reproductive capacity, my adaptive fitness, so defined in a theoretical context. So just really sort of grasping the nettle and thinking about, well, these things exist.
what principles could possibly explain the paradoxes when looked at from sort of the unilateral, the monothematic view of natural selection for example, or the monothematic view of self-organization in physics. So thinking out of the box literally in this instance entails being able to take multiple perspectives on a particular problem and seeing the contradictions and seeing how they can be resolved.
Now, that would be one of many examples. There was a paper, and I can't remember the specific one, perhaps Mike and Chris will be able to enlighten me, but there was one paper.
written I think about a year or two ago where they actually listed I think were 15 basic conclusions about the the fundamentals and the nature of self-organization and self as distinct from other in any system and this listing 15 really interesting points and predictions all of which either I'm sure will be formally demonstrably analytically true
There was a paper in Neuroscience of Consciousness arguing for an approach to consciousness that spanned phylogeny and we were very interested in that paper
specifically to look at bacteria and single cells and facultative multicellular such as microbial biofilms and ask what do these systems know about the world? How do they see the world? How does E. coli perceive the world, for example?
And we suggested that the answer is in terms of the sorts of things that we call taste and viscosity, but that experiences of taste and viscosity are perfectly good kinds of experiences. And that if we think about how these kinds of systems deal with their worlds,
and solve problems within their worlds, that that would give us a more unified understanding of what it meant to be a system that was aware. And so you're trying to use E. coli as a means of understanding yourself or humans. By the way, is this a project? Do you feel like your undertaking is an attempt for you to understand yourself? Are you trying to understand humans?
I'm wondering, is this more philanthropic or it's selfish and then it bleeds into the philanthropic? Well, I actually also want to understand E. coli. We don't have a great understanding of the lives of other organisms. Professor Levin,
What puzzle do you find most important? What do you think about on a semi daily basis or even perhaps multiple times a day? And then we'll go around the table here with Chris Fields and Karl Fursten next. And then it's essentially me taking a backseat and allowing you all to free flow just for the audience to know. I know you all know this template already. Yeah. So the thing I think about many times a day has to do with the scaling of cognition.
I want to understand two major things. One is how it is that some collection of competent parts comes together to form an emergent self with preferences, goals, memories, cognitive capacities that belong to it but not to the individual parts. I want to understand how that emerges, how the goals of humble
Simple kinds of systems scale up into much more grandiose goals that we see during development, during behavior, during culture, and so on. And I'm also very interested in the left side of that spectrum. Where does it begin? Is there really a zero on the spectrum? I think that all of us here would agree that there is a spectrum for these things. It's not a set of binary categories.
But what happens at the very left side of the spectrum? And so one of the first questions when we get to that, that I had written down to ask both Chris and Carlos to sort of comment on what does the Venn diagram look like of the set of things that are alive versus the set of things that are cognitive? How do those two categories relate to each other? Do they overlap as one a subset of the other? And what really happens at the very beginning?
Can we can we develop a kind of which which I think I think both of them have been working on a kind of basically a kind of panpsychism that doesn't doesn't sort of paint on new cognitive mysteries on top of a physics that works perfectly well, but instead to try and to try to view physics from the bottom up as having already a useful cognitive lens on it. And how does that help us to build up cognition? So that's that's something that I think about every day. Chris,
I thought every day or almost every day for a long time about why we humans see objects embedded in space time. So why do we see things that we treat as independent of each other? And why do we
see them, operate on them, interact with them, think about them as embedded in this coordinate system that we call space. And equally important, why do we see them as maintaining their identities as things over this other coordinate that we call time? Carl.
I have to confess I spend recurrently most of my time thinking about me, but in an academic sense, you know, how do I work? It's a curious mixture of introspection and trying to understand why I can introspect from the perspective or through the lens of a physicist. So I spend most of my time
dissembling preconceptions, the gifts that sentience has given us, and try to reassemble them in relation to physics, density dynamics, that has to be articulated in terms of the kind of maths that a physicist would use, which would be effectively differential equations and the calculus of variations.
And having done that, there are so many, there are a sufficient number of moving parts, but not that many. But the combinatorics then lead you to the kinds of questions that Mike was talking about. So what different ways could this physics of sentience be manifest? What possible ways could it be manifest? And how does that address the distinction between things
A KFC tale in the pursuit of flavor. The holidays were tricky for the Colonel. He loved people, but he also loved peace and quiet. So he cooked up KFC's $4.99 chicken pot pie.
And to my mind an important aspect is planning.
This would disambiguate them from other natural kinds or particular kinds that don't have that kind of facility and what underwrites them. So that's what I spend most of my time doing. I have to say in conversation both with Mike and Chris and other people that you can easily get distracted in a rightful way.
by thinking about that mechanics in exactly what Chris was talking to, which is a sort of scale-free way. But in my rhetoric, that scale-freeness speaks more to the coupling between different scales that Mike was alluding to.
So how does this mechanics, this physics of sentience, which I understand largely in terms of probability theory and effectively Bayesian probability theory. So what you have is a Bayesian mechanics. How does that Bayesian mechanics apply at one scale and another scale? And then of course the big question
is something that I find myself increasingly pressed upon, largely through conversations and collaborations, but that does mean that I spend a lot of my time thinking about that and writing demos and mathematical equations trying to try to get to the underpinnings of it. Mike, I see you nodding, so what is it about what Karl said, sorry, Professor Levin, I apologize, and what is it that Professor Friston said that has you
feeling like you agree with it? Well, as usual, I mean, he said it perfectly. I couldn't say it better myself. I think that's exactly the kind of research agenda that I'm interested in. I mean, that really is kind of a very good description of what I think we're all looking for in a certain sense. Is there a difference between cognition and consciousness? Is one distinct? Is there overlap? Is one a subset?
I actually can't conceive of cognition without consciousness, which I prefer to call awareness, not just to be perverse, but because in many cases, consciousness is used in a way that implicitly means self-consciousness, whereas awareness is often used in a way that does not make that implication.
And I find it difficult to conceive of cognition without any kind of awareness at all. So, I'll ask a question based on what Chris just said and what Carl has to say about it. So, what do you think about the so-called hard problem? Is there in fact a hard problem at all? I would take Andy Clark's view on this, who has worked closely with David Chalmers.
I don't think so. No, I don't think that the hard problem in and of itself is the interesting focus of inquiry. I think the move that I see in philosophy has been to the meta problem or the meta hard problem, which is, as Andy Clark puts it quite succinctly, is why do we spend so much time puzzling about why we are aware?
Just asking that question speaks to, I think, the nature of the hard problem, which is a sort of very much a second order, a meta capability that we can make sense of our own sense making. And we associate that with selfhood. So one has to ask the question, what kinds of systems, again, I think, particles or creatures would have the capacity to represent selfhood
And furthermore, represent the counterfactuals that would enable you to ask the question, why am I conscious? That immediately just logically presupposes that there is an alternative hypothesis of other counterfactual that I'm not conscious. So imagine now you're talking to a creature, you're talking to me, that has this capacity to imagine counterfactuals that cannot exist.
So that's a remarkable capacity to have from my point of view a sort of internal world model or generative model where counterfactuals can exist is quite remarkable. I don't imagine that a thermostat or a virus or an E. coli will have a sufficiently sophisticated set of electrochemical
dynamics or kinetics that would enable them to physically represent counterfactuals of this sort. And I could go on, the usual line of argument is, well, why do we have models? Why do we have the capacity to represent counterfactuals? You end up with saying, yes, you have to have them just if you contemplate things that plan. So if things plan, they have to have
an ability to simulate the future because they have to simulate or represent at some elemental level the consequences of their action upon the world. And if they have now the capacity to represent something that has not yet happened, namely in the future, then they now have the capacity to represent counterfactual outcomes. So I think the meta-hard problem inherits simply because we have a sufficiently
Temporarily deep generative model that can represent the consequences of our actions and of course using the word our implies some kind of elemental agency and some kind of self
Can I pursue that just a little bit? I think you raise an extraordinarily interesting point here, Carl. And so I want to ask you,
From a phylogenetic perspective, do you think that organisms that can plan necessarily are able to represent multiple counterfactuals? So let's give an example. Suppose you're a cheetah chasing a gazelle.
You're planning your moves while looking at the gazelles moves and your objective is success. Do you think that the cheetah worries about the other counterfactual condition where the gazelle isn't caught or does it just represent this one counterfactual condition that is in a sense the goal state? That's an excellent question. I personally think that
And when I answer these questions, I think, how would I simulate this? How would I put a teacher in silico?
and integrate according to the differential equations that must be present under certain assumptions and my guess is that the cheetah chasing the gazelle would as you say just be pursuing a path of least action and it would not have an explicit representation of alternative paths so it would be responding in the moment in a reflexive habitual way so that's a really important question because if we don't have or all of our behavior
is just a manifestation of pursuing those paths of least resistance or mathematically those paths of stationary action, then there is no counterfactual to select from and there is no notion of planning in the sense of I am going to do this as opposed to that.
that doesn't mean to say biological organisms don't have the capacity to plan multiple futures and I'll just give you one real simple example of our brains and I would imagine even the brains of reptiles to a certain extent being able to plan multiple
actions in the future and that's in eye movements. So if I just take one of the simplest problems in terms of planning or perhaps not the simplest but one of the problems of planning that have to be resolved within
several hundred milliseconds, given that we have to choose where to look every 250 milliseconds in order to gather the right kind of information to construct the scene in which we are situated. Coming back to your preoccupation, Chris, how on earth do we then explain these sparse, saccadic samples within some frame of reference, some space-time sort of frame? How do we sense-make with reference
The point I'm making here is that the anatomy of many brains, many phenotypes
does have the capacity to represent the consequences of looking everywhere at the same time at once before they have looked. So these are sometimes referred to as salience maps and we know that there are good candidates for these salience maps, for example the deep layers of the superior colliculus and one could argue possibly some anatomy of the pulvina. So if you look at the neuronal encoding of the salience of where
the salience of the next place to look, what's going to grab my attention and make that manifest overtly with an eye movement to go look over there, then you can interpret, I think, the anatomy, the ocular motor system, and especially that which underlies the control of the saccadic eye movements as entertaining a whole range, possibly millions of potential
plans of action and the consequences. And then the brain literally selects the one that's most likely, which you can simulate by doing bumper tractors on these sort of salience maps, representations of the salience or the epistemic affordance or the expected information gain if I looked over here, looked over here, looked over here, looked over here, you select that, your eyes jump over there, and off you go again.
So I think it's a great question and in a sense speaks, I think, to a certain extent to what Mike was alluding to before, that there are certain setups, there are certain sort of neuronal electrochemical infrastructures out there that can be read as just pursuing paths of least action in a reflexive way. And there are others that may have this richer structure, this deeper structure that could be support the, you know,
Mike, is that what you had in mind in terms of breaking this sort of panpsychism trap, different kinds of
I agree with Chris's gut feeling on this. I find it very difficult to imagine cognition without some sort of awareness or simple consciousness.
What strikes me as different than real about the hard problem is that unlike all the quote unquote easy problems, and maybe unlike the meta hard problem, we kind of know the format or the shape of answers to those questions, right? They're either their numbers or their equations or their lists of capacities. You know, we know what a proper answer would look like. And yet, I find it very hard to imagine if somebody claimed to have a
a good theory of consciousness and I were to ask them, okay, well, what is the prediction of your theory in this particular case? I don't know what the format of the answer looks like, because numbers and the typical things we get don't do the trick, they, you know, they're sort of third person descriptions. And so, so is the answer, does the answer come in the form of a poem? Is it art? Is it, you know, something you have to literally plug yourself into to then have that experience? Like, what's the, I find it different, because I don't know what if we had a true, a correct theory, I don't know what that theory would output.
Well, that's a better answer to the hard problem then. I'm sure the answer is quantum physics. That's what it's going to look like. If we could understand it, the answer will be in quantum physics. Chris, do you agree? Well, I'd like to back up a little bit.
And in a sense, play off what Carl was just saying about alternatives and the role of alternatives in the hard problem. I mean, Chalmers drives his arguments about the hard problem with the alternative of unconsciousness. And this alternative makes great sense against the background of a
particular set of assumptions about physics, which is the assumption that nothing is going on cognitively in the physical world, that there's no reason to talk about experience or awareness or anything like that when we're talking about
of rocks or billiard balls or planets or electrons or anything else. And in that case, there is this cut that is not really a cut in scale, spatial scale or temporal scale,
but is, in a sense, a cut in some measure of complexity that suggests that below some cut point, which may be a bright line and maybe a fuzzy area, who knows, there's the possibility or below some cut point, there's no awareness at all, none, zero. And above that cut point, there's some possibility of awareness.
And in that case, one's forced to say, uh, what, what could be added by complexity other than complexity to produce awareness. And it's that intuition, I believe that drives the heart problem. Do you mind explaining that a bit more of what can be added by complexity? That's not complexity that sentence. Well, um,
Yeah, and one way to put it is complexity, as complexity increases, some magic has to happen. And you suddenly get awareness. And it's not clear quite how much complexity one has to add. But as soon as one reaches the right amount, then out of the blue, you've got awareness.
Whereas without crossing that threshold, you had no awareness whatsoever. So there's a zero point and the zero points somewhere on the scale of complexity, and it's nowhere near the bottom. So I think one can view a lot of research on consciousness as a way to escape that argument.
Consider integrated information theory. In integrated information theory, the criterion for consciousness or awareness is actually very simple. You need to have an internal feedback loop. And if you've got an internal feedback loop, you're at least a candidate for being aware, as long as you're not embedded in something bigger that has bigger internal feedback loops.
And if one thinks about the work of Peter Strawson, for example, from a philosophical perspective, his arguments toward panpsychism are all arguments of the form, there's no way to draw the line in the scale of complexity.
and get a place where magic basically plausibly happens. So I think Chalmers did a real service in posing the problem in that way because it forced us to think about this idea of magic, or you could call it emergence if you wanted to, of something completely new that was awareness.
Now, I do agree with Karl that quantum theory can help us dispel the problem by being, in effect, a theory that's about awareness. But we can get to that farther down the road. Do you have any reason to believe that we're embedded in something larger? You mentioned in IIT, there's the feedback loop that is necessary for consciousness. And you said,
that one is conscious as long as one is not embedded in something else that's of which that is conscious, like there's a larger feedback loop there. Do you see there being some larger feedback loop? Some people say the universe itself is conscious or that societies can act as some larger level consciousness, like with each of us acting as neurons in some sense. Well, I would certainly wouldn't want to rule that view out a priori.
I mean, it's obvious that we're embedded in much bigger systems. And I think we understand essentially nothing about the cognitive capacities of those systems.
Yeah, I think that's a really important question and I think a lot about the perspective of, let's say, if you were a neural network or something like this being trained and so on, what would the perspective of a sub-component of that be? Let's say if you had the capacity as a neuron to sort of look around and ask yourself, do I live in a cold mechanical universe that doesn't care what happens or
Is there some sort of agency in my environment? When I'm as a piece of that network, am I learning from my environment or am I being trained? Because when you're being trained, the real question is how many agents are there in that interaction? Is it just you learning as you will, you're the boss and you're sort of learning whatever from your environment?
Or are you actually being trained because the environment is an agent with an agenda that is training you for some particular purpose? And so this question of how would you know? So you look around and you say to yourself,
If you're a piece of this network and it's being trained for some image recognition task, you would be, I think, wrong to come to the conclusion that you live in a mechanical universe. You should come to the conclusion that it's clearly rewarding you and punishing you for specific things. It's not neutral with respect to what you do. And for some reason, it really likes it when you find pictures of dog eyes or something like this. And you have no idea that what it does is recognize dog faces out of all the other
all the other inputs. So no doubt there's some kind of Gordelian limit to figure, to be able to actually understand what the larger system is doing. But I wonder if we can even have some evidence that just for the fact that yes, there is a greater, you know, sort of a gentile dynamic going on in which I'm caught up, even if you're not able to, you know, sort of comprehend what that's going to be. I don't know. I don't know what that's going to look like. But I think it's pretty important. And I think, as Chris said,
We really don't have a very good science at all of trying to predict or control.
the cognitive capacities of systems made up of parts. So we can know many things about parts, cells, robotics, whatever, and we routinely make these larger systems and then get surprised about what it does or doesn't do or what the goals are going to be, what goals is it capable of pursuing, what preferences does it have. This is probably an existential level job for society is to get a good science of that going.
You have done that though, haven't you? Well, we've started anyway. I would go further than that. I think your work would represent a substantive and established formal framework to address exactly those issues. And you can see embryonic versions arising in many different fields. I just came back from a meeting
of economists and financiers. They are trying to make the move from behavioral economics to cognitive economics. And they were exactly addressing these kinds of issues. You are understanding the
the notion of distributed cognition in a market. So, you know, ultimately, they'll come knocking at your door. I can see that. I can see exactly the sequence of people, you know, searching around for people to talk about, illuminating that realization that the, I mean, it comes back to what we were talking about before, where you started, and I was picking up on the
on the link between different scales of self-organization, where every scale has, you know, complies with the same principles. But how does that contextualize? And I think one really interesting example of that is to think about an individual in an ensemble or a society. And just to wrestle the argument back to pick up on something that Chris was saying about, you know, if there is a bright line between the kind of mechanics that would
qualify as conscious in the sense of being self-aware. And I'm distinguishing that from the kind of consciousness that would just entail qualitative experience and a loss of phenomenal transparency. So I'm talking about now self-awareness as one of the bright lines that may be very blurry and vague, but certainly one which is induced by the hard problem
Just asking a simple question, what kind of explanation for my world, whether I'm a single neuron or a single person, would enable me or justify the notion that I am a self? And the obvious answer is when I have to disambiguate between the consequences of my action and your action, if you are very similar to me. So if I
managed to survive on an alien planet where there are no structures like me that could cause the same sensed consequences, then there would be no problem inferring, did I cause that or did you cause that? So just having the existence of a population of conspecifics in some sort of formal structural sense,
suddenly induces the inference problem, did you cause that, do I cause that and of course that naturally calls for generative models or internal dynamics representations on the inside that entertain the hypothesis or the notion that that's me as opposed to not me. So you get for free in and only in this context where you've put lots of these particles together, the license or the motivation to have
a hypothesis, a representation of selfhood that from a statistician's point of view is exactly the justification for the increased complexity that Chris was talking about. So I was listening to that really interesting sort of
not to IIT and all it brings to the table in terms of this sort of, you know, commitment to some threshold crossing or traversing a cut with increasing complexity. But why? I mean, one simple answer is that, you know, if you ask a statistician, what is complexity? They'll tell you, well, it's basically the degrees of freedom you're using up to explain some sensory data.
You could articulate it as a KL or relative entry between a posterior and a prior if you're a Bayesian, but in essence it's just the degrees of freedom you're using in order to accurately explain these data. So mathematically the evidence is equal to the accuracy minus the complexity. So why would you need, why is that useful? Well an increase in complexity is only licensed by an increase in the accuracy. So the simple argument would go
If I am obliged to model and predict and explain a world that is constituted by other things like me, then the accuracy of my predictions will be greatly enhanced if I have a representation of me as distinct from you. If that entails an increase in the degrees of freedom, namely adding in this extra kind of hypothesis and everything that it entails, then
I'm going to have a greater complexity, but it's a complexity that is more than paid for by the increase in the accuracy. Just coming back to my original point, I think a lot of the work that you've done at the cellular level
would be very gracefully translated to the societal level and to things like economics and to eco niche construction in ethology. And I know Chris is probably wanting to say this, but language as one way of facilitating that notion, that ability to efficiently with minimum complexity in this instance, do this
Hear that sound?
That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the internet's best converting checkout, making it 36% more effective than other leading platforms.
There's also something called Shopify Magic, your AI powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level.
Join the ranks of businesses in 175 countries that have made Shopify the backbone of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklynin. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com
Well, again, what I'd actually like to do is back up a little bit and
put some of what you just said into very simple active inference language. If I see something happening on my Markov blanket, on my interface with the world, then I always have the question, did I do that? Or did the world do that? Where the world means everything outside me. And
In a sense, the answer is always the world did it. So the question becomes, did the world do that in response to something I did to it? Or did it just do it? Not in consequence of any of my actions. And so one gets immediately to this kind of babbling scenario that we've talked about many times, in which
An infant or a robot or some system is trying to figure out by measuring correlations, whether the world's inputs to it have anything to do with its inputs to the world. And just asking that question requires enormous representative capacity.
Because one has to represent one's actions and represent them pretty well in time. And one has to have a good memory to represent enough actions to get any kind of statistical support for drawing an inference about correlation. And that memory has to be represented as a memory
not just part of one's current input. So I think this question that you posed is really the key question faced by any agent at all that's trying to get a model off the ground, which in a sense gets back to the question that Mike asked early, early on about how does this all start?
Maybe it starts with babbling in very simple systems. You know, I was thinking recently, this whole issue of how many agents are there and where is the border between the agent and the world and how do you self model that border is a fascinating topic. And there's an amazing developmental model for this, which is that, you know, we often talk about one embryo and the embryo does this and the embryo does that. But actually,
What happens at the beginning, let's say, for example, an amniote embryos is that there's a flat blasted disc, which has just a few cell layers thick. So it's kind of think of it like a frisbee and it just has a few cell layers. And normally what has to happen is that one point in this disc breaks symmetry and then organizes the primary axis of the first embryo and basically tells all the other cells, don't do it because I'm doing it. And that's how you end up with one embryo.
Now that process is very easily perturbed and many people have done it. I used to do it in my graduate work. And what you can do is if you perturb that process, that initial blasted disk, that undifferentiated sort of pool of cells, which are these sort of proto
low level proto-agents, that pool can break up into not one embryo but actually multiple. I did this in bird embryos and you can have twins and chicken and duck and things like this. Humans have exactly the same structure. You can get them head to head, you can get them side by side, you can get all sorts of geometries.
you can get triplets, you can get multiple individuals emerging by different partitions of this really kind of medium, this particular medium where you have a bunch of cells and you don't know ahead of time how many individuals at the level of how many larger individuals, so embryos, are going to arise from this medium because the dynamics by which in this, you know, local activation and long-range inhibition and things like that,
the dynamics that break up that undifferentiated ocean of potential selves into one, two or three or more selves is actually, it's very dynamic, it can go different ways. And then you get interesting things like this. So for example, you might know that human conjoined twins that are sort of stuck together side to side, one of the twins often has left right asymmetry defects.
It's because when you have two twins side by side, the cells in the middle, both twins can't quite agree on who they belong to. Are they the right side of this twin or are they the left side of that twin? Both twins think they belong to them, but in fact, they're overlapping. They're the same cells. One side will have correct left and right. The other side will have two rights, for example. This ends up giving one of the twins laterality defects with respect to heart and gut pattern.
And so their models, each twins as the collective of cells tries to compute things like where things are and what's left and what's right and so on, their models can disagree with each other. They can draw the boundary between self and world in different ways. And you can have these sort of disputes over certain areas as to who they actually belong to.
And so I'm just incredibly interested in this process of individualization, so to speak, out of this like ocean of potentiality, these cells, you know, 50,000 cells and some number of individuals at the embryo level will be formed. And each of them will have specific goals in morphospace. Each of them will try to achieve very specific morphologies. And you don't know ahead of time how many there were going to be, you know?
Carl, what occurs to you when you hear that? Ambiguity, again, just thinking about the imperatives for self-organization.
When I hear that was a fascinating story, and don't let me forget, we ought to explain to the viewers what babbling means just in this context as well, because that's quite illuminating. Perhaps we can, I'm not sure that the cells do motor babbling, but they certainly resolve the same kind of problem. Chris, correct me if I'm wrong, but by babbling, all we mean is that when you're first born into any universe,
You've got to work out or test the hypothesis that I caused that or the world caused that. So you don't know, you don't have a self-model, whether this is sort of a declarative model or completely sub-personal, just hardwired into the synaptic efficacy and connectivity of your brain.
So the first thing you have to do is just to work out what you can control and what you can't control and the idea is that you engage in what I would call epistemic or respond to epistemic affordances or epistemic plans that reveal knowledge, they resolve uncertainty and in this instance it's the uncertainty about whether I was the cause of this rattle rattling
So if you imagine motor babbling as manifest in a little baby rattling its rattle generating both the sensations from the muscles and the skin but also the visual and the auditory sensations all co-occur providing definitive evidence that there's something special about this process and this event that
that provides the basis for the hypothesis there's a unitary cause there's a unitary cause which is me shaking the rattle but of course it may take several months if not years to actually get to actually realize that cause is me so one can imagine sort of you know robots learning about the the manipulander that they can articulate or the the way in which they can move around and you know
But I think the idea, bringing it back to sort of where does selfhood come from, it would rest upon testing the hypothesis, which has to be physically represented with a deep, more complex generative model of sense making, that in fact, it's me that's actually caused this single cause of all these proprioceptive motor sensations, visual sensations, auditory sensations,
And that will be especially prescient when you're starting to realize that some of these causes, which you thought were you, of the sort associated with nurturing and suckling, were actually due to mum. And we come back to this argument to actually have a good hypothesis which explains why I am not in charge of mum, because she is now not always responding to me when I cry.
To explain that I have to now develop a hypothesis. Yes, it was me causing all of this but sometimes there's something else out there that's not me but very much like me and that's mum and then you can see how there would be a pressure in terms of finding the best explanations for your sensations to have that. So if you take that notion now think of the same problem from the point of view of a cell
a bunch of cells that have to work together or will ultimately form an embryo via this process of symmetry breaking. You have to ask what are the underlying imperatives, how could it be any other way and of course it could be lots of ways but the ways in which it goes wrong which would be another way of saying
those rare occasions or the ways that it doesn't happen in consequence of their rarity involve this ambiguity again, literally in the context of some cells not knowing whether they're belonging to one twin or another twin from the cell's perspective. So I was just thinking about the nature of ambiguity and of course it is exactly the same. One can account
for the simple observation that self-organization does not tolerate ambiguity, it does not tolerate uncertainty. It's only manifest in the context of accurate and well-evidenced definitive exchanges. So come back to Chris's notion of sense-making projecting onto my Markov blanket or my
that I need to resolve all the uncertainty, as much uncertainty as I can. But put that another way, in a slightly more deflationary way, stuff which we see the way the universe seems to work can be described as realizing processes that minimize this kind of uncertainty and ambiguity.
and simply maximise the neutral predictability of what's being projected or impressed upon my surface or my holographic screen or my Markov blanket. So, you know, that's what was going through my mind. I thought it was a beautiful example of, you know, when it goes wrong.
There's uncertainty and ambiguity in the game, and that tells you something quite fundamental about when it goes right. And when it goes right, it's just basically a statement of what exists and what perseveres over time. Chris, did you have anything to add to that? Well, I would be interested to see how actually Mike responds to
that discourse with respect to the example of this embryonic sheet where the cells are each trying to figure out what they're supposed to do and hear that sound.
That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the internet's best converting checkout, making it 36% more effective than other leading platforms.
There's also something called Shopify Magic, your AI-powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level. Join the ranks of businesses in 175 countries that have made Shopify the backbone.
of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklyn. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com slash theories, all lowercase.
Go to Shopify.com slash theories now to grow your business no matter what stage you're in Shopify.com slash theories. Razor blades are like diving boards. The longer the board, the more the wobble, the more the wobble, the more nicks, cuts, scrapes. A bad shave isn't a blade problem. It's an extension problem. Henson is a family owned aerospace parts manufacturer that's made parts for the International Space Station and the Mars Rover.
Now they're bringing that precision engineering to your shaving experience. By using aerospace-grade CNC machines, Henson makes razors that extend less than the thickness of a human hair. The razor also has built-in channels that evacuates hair and cream, which make clogging virtually impossible. Henson Shaving wants to produce the best razors, not the best razor business, so that means no plastics, no subscriptions, no proprietary blades, and no planned obsolescence.
It's also extremely affordable. The Henson razor works with the standard dual edge blades that give you that old school shave with the benefits of this new school tech. It's time to say no to subscriptions and yes to a razor that'll last you a lifetime. Visit hensonshaving.com slash everything.
If you use that code, you'll get two years worth of blades for free. Just make sure to add them to the cart. Plus 100 free blades when you head to H E N S O N S H A V I N G dot com slash everything and use the code everything. There's there's still. They still certainly believe that they should be reproducing, so they they definitely do that and they signal to each other.
But somehow they're the collective effect of all of that signaling to each other is to organize different roles for the for each for themselves. So how does the symmetry breaking occur?
Yeah, so that's really interesting. So as I was hearing Carl talk about this, I realized a couple of things that, first of all, as we were saying before, feedback loops are absolutely central to this process because the easiest way to prevent any of this from happening and to end up with a kind of a mono, you know, sort of a very featureless sheet where there are no embryos is to block the positive feedback loop.
that short range inhibition, long range activation that says to one cell, I'm going to now be the organizer. I'm going to make this access. Everybody else don't do it. Those are both feedback loops. And so if you break those feedback loops, you get nothing. So the feedback loops are right at the beginning of this process. And the other thing that I thought was really interesting that Carl just said is about the cells babbling. So what I think it's a really good name for what we see when you take a cell and you put it out into a dish,
What you'll see in any new environment, you will see two things that I think are probably babbling in different spaces. One is that in physical space, you will see that it's incredibly active. It's constantly putting out and pulling back these
The other babbling takes place in transcriptional space.
So gene expression is also never just sort of still, and this is, you know, these are the genes that are expressed and I'm just going to sit here and do that. All the genes are constantly going up and down within specific ranges as it sort of wiggles in this transcriptional space. And I think both of them, probably physiologically, metabolically, I bet this is going on in all these other spaces. And I think babbling is an excellent framework for understanding what it's actually doing. It's actually taking these little actions.
looking for evidence of specific things that it can then make use of to start to draw boundaries. And as Chris was saying, I think what happens is that in this blastoderm, what will happen is that once that individuation starts where there's a specific cell that starts to go down this road of
being the organizer of this new larger collective, it immediately begins to distort the action space for all the nearby cells. It starts putting out all kinds of signals, you know, reward signals and physical forces and all these other things that now are going to bend the option for all the other cells. And the best example of that we have are these xenobots where these things are just made of embryonic skin cells.
And if you look at a standard old embryo, you get this idea that, well, what do skin cells want to do? They want to be a passive two dimensional layer on the outside of the animal. They do nothing except sit there, keep out the bacteria, you know, nice and a boring two dimensional life. But what you find out is that that's only true because the other cells are basically bullying them into it.
left to their own devices in the absence of these instructive interactions with the rest of the embryo. What the skin cells actually want to do is get together into a three-dimensional kind of ball-like architecture. They are self-motile. They'll run around and move and have various behaviors, including make copies of themselves if provided with materials. And so that is completely obscured by standard development, where what you're seeing actually is
is cells in a space that was really deformed by all their neighbors, right? And that's, you know, that kind of process that starts to make those distinctions where the embryo can tell what parts are inside and what parts are supposedly outside of itself. And that gets reinforced by all these early activities. Carl or Chris, do you want to jump in on that? I think we've been describing psychos that says basically that it's curiosity all the way down.
Every system is trying to figure out what's going on. I was struck with that with this image of the little cells putting out fingers, this is expiration, sort of true blue expiration and mathematically it's simply as Chris says it's what artificial intelligence
research aspires to, which is artificial curiosity, going out there, getting the right kind of data that's going to maximally resolve your uncertainty and paint the way forward. And of course, it's intimately related to planning.
In the sense you have to realise that palpation, whether it's a little cell palpating with its filial or podial, whether it's me palpating my visual world by moving my eyes around, or whether it's the little baby babbling and palpating its cot, this palpation has to be planned.
It could be planned in the sense of the gazelle, sorry, because it's not the gazelle, the cheetah chasing the gazelle. So it doesn't have to be very sophisticated cognitive conscious planning, but certainly has to has to be planned. It has to have this sort of curious behavior. It strikes me that that is such a fundamental aspect that I think would qualify behaviors
that have that aspect as cognitive in some sense. And it's so fundamental because of course, it's just an expression of dynamics that apply to action upon the world that underwrite everything that we do as certainly as scientists and one could imagine as human beings, but certainly as scientists, you know,
The maths of curiosity was actually worked out by Dennis Lindley in the 1950s in terms of expected information gain and then reintroduced by people like David McKay in the context of active learning. So bringing this active notion that you can learn by actively acquiring the right kind of data that optimizes that kind of learning. And it became known as
the principles of optimum Bayesian experimental design. So it's exactly the same kind of curiosity that we as scientists use all the time whenever we design an experiment. It's basically configuring, actively configuring some process to generate something that can be sensed or measured that maximally resolves our uncertainty or forwards the greatest amount of information.
So the three of us as scientists have become experts at this kind of formal curiosity simply just by acquiring knowledge about, in the particular paradigms or setups and fields that we find ourselves in, the right way to do experiments. But you could argue that that's life. In a sense, that is the infant bubbling.
It is that kind of sense making and actively getting the right kind of data to work out your place in the world, to work out what you should do next is one of the most existentially important imperatives. The mechanics that Mike was talking about in terms of the
the little skin cells left to their own devices, having a party and forming balls and wandering around. One could apply that kind of mechanics to people, couldn't you? Or even cultures and countries. We're all trying to find our place. We're all trying to work out how to respond to those constraints. People around us at many different scales put on our behavior and try and infer, well, how am I meant to behave in this situation?
You hear that sound? That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the Internet's best converting checkout, making it 36% more effective than other leading platforms.
There's also something called Shopify Magic, your AI-powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level. Join the ranks of businesses in 175 countries that have made Shopify the backbone.
of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklynin. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com slash theories, all lowercase.
Go to shopify.com slash theories now to grow your business no matter what stage you're in shopify.com slash theories. So even if we all start off with the same genetic code and the same sort of model of your how people behave.
The context in which we find ourselves now needs to be inferred in order to know how to behave in this context, either as a child or as a mother, as a politician or as an aid worker or first responder, whatever you want, you have to make that inference. And of course, that inference, inferring that context requires the curious behavior that Chris had picked up on in Mike's example.
Wow, you know what I'd love to hear is each of you talk about what is the simplest, most basic thing. So, you know, hopefully going down to like physics prior to life, but wherever you like, what is the most basic thing that is able to do that? Because that's what I find in talking about this stuff to other communities. The thing that people are the most resistant to is this idea that these kind of dynamics, this kind of exploration, curiosity, prediction, all of these things that we're talking about,
that people think it's brains first, and then you can do all these great things. I would love to hear what you think is the simplest physical system that can do that kind of thing so that we can talk about how far down it actually goes.
While Chris is thinking about that with a profound answer, I'll give you a not a trivial answer, but a formal answer. I think you'd have to identify the depth of planning that underwrote the dynamics of the behavior of the system in question. And that would be basically
If you were putting this in a sort of path integral formulation it would be the sort of the amount of time over which you are integrating your paths and you're evaluating the best way forward. Once you move from very short time scales then the depth of that planning I think eludes physical realizations that could be written down in terms of
representations in terms of concentrations and depolarizations and the like and one has to move to a quantum discrete time representation. So you know I guess I'm basically making the difference between the sort of the
the kind of dynamics you'd find in a in chemotaxis or a thermostat that could be written down as differential equations. In fact Mike, you're the kind of differential equations we used in one of those early pattern formation papers of knowing your place approaches where everything could be articulated in terms of differential equations so that there is a kind of
But it's not of the curious kind that we talked about. It's just following the most likely paths of least action. So I would say that they're not cognitive in the sense that they have planning. So I put chemotaxis and thermostats and possibly most viruses probably at a precognitive or protocognitive level. But as soon as you get to
generative models that can roll out deep into the future and I repeat just when you try to implement this in silico you know you're using computer simulations you really have to move to a discrete time representation which I think is a non-trivial thing I mean just for example remember the sort of the eye movement example you know we sense make by getting little snapshots of the world every 250 milliseconds
It's not like I've got, I'm a thermostat and I'm taking a continuous record of the temperature. That is a continuous, I'm actually getting discrete quantal observations. And I think once you move to that kind of generative model or biophysics that entails the gradient flows under that generative model, then I think you're in a position to talk about sort of planning and cognition and curiosity of the kind that we're talking about.
So I, you know, I'd literally put time constants on this, I'd say about, you know, 300 milliseconds, if you've got the ability to represent the future, 300 milliseconds in, you know, sorry, represent the consequences of your actions, more than 300 milliseconds in the future, then I put you in, you know, having crossed that loop, one of those cuts and moved from being a sort of, you know, a virus to, you know, to, I don't know, perhaps an insect, I'm not sure.
Are you seriously putting the 300 milliseconds forward or is that facetious? It's facetious but not as facetious as you might think. I think there's one way of celebrating the
escapes from panpsychism that we started off with when Mike was saying what he spends his time talking about. One of the ways is to commit to vagueness and how many grains of sand do I need before this ceases or becomes a pile and try and work out what the quantity is that defines that spectrum that we were talking about before, along which there is vagueness at some point according
of a certain kind arises. But I also think there are natural kinds that actually have bright lines between them. And, you know, what I'm saying here, and I don't think I've said it before, and I probably wouldn't say it again, unless I've simulated it, but there's certainly my experience
qualitatively different kinds of generative models that represent time in very very different ways. You can get so mathematically things like Kalman filters that can be cast or linear quadratic control robotics and the like that can be cast purely in terms of futures that are represented in terms of velocities and accelerations and higher order motions that are just in the dynamics and the equations of motion.
And you can realize and build these things, you can simulate them and they behave in a very plausible way. You can get them to do handwriting, you can get them to sort of self-organize and do all sorts of interesting things a la thermostat. But that's as far as it goes. These things don't show curious behavior. To do curious behavior, you've really got to roll out much further into the future.
And in running out further into the future, it's very difficult to represent the future just using differential equations unless you move to very, very high orders of motion.
at which point you start to sort of to minimize that complexity, start to articulate things in discrete time, and then the numerics become much simpler and doable. But then you've got a qualitatively different kind of generative model, you move from a continuous to a discrete time, a quantized generative model.
Um, so that for me is not, uh, does not allow a vagueness, you know, there is actually a natural kind that will represent things in the moment in terms of flows and dynamics and other natural kinds that will represent things, um, in terms of discrete, not universal clock time, but discrete iterations. Um, and I'm saying that the, you know, for, for certainly our scale of interactions and the way that we sense make, I would have thought 300 milliseconds was not a controversial
I say that because that's the cognitive moment. It's also the time interval between the way that we sample the world, not just with our eyes, but when we sniff, we sniff at a frequency, which means that each new sampling of that world, that palpation with our chemoreceptors and our factory receptors is a game every 200 milliseconds. So just to me, there's a sort of discrete belief updating, and it is a discrete belief updating that is
underwritten by the way we act. The plans have to be in these discrete chunks of time. So that's where the 300 milliseconds come from. You see it everywhere. Mice whisking, they whisk at that frequency. So they touch the palpate every 300 milliseconds. The way I'm speaking now, every phoneme is reaching your ears every 300 milliseconds or so, or possibly less. So wherever you look in our biological scale,
Well, it's interesting because in physics, the number three has an interesting role in fundamental physics because there's three generations of
Well, I was just having a visual impression actually when Mike was talking and of babbling and
The visual impression was of our sort of model of the quantum vacuum. And when we think of when we think of the vacuum, or at least I think of it visually as little things popping in and out of existence all the time. And so
One could think of that as a kind of babbling, right? The field itself is exploring what's going on outside of itself. That's extremely interesting. By forming a little entity to go off and explore and then report back. So this is a very different picture from the one that Carl gave in that
I suppose it also has a similarity or two that when we write down a model of that activity, it is discrete. It is operating in a discrete kind by generating discrete events, even though it's modeled in a continuous space time, which once one digs a little bit deeply, the continuous space time itself becomes discrete.
But this raises a question for me, which is another of the things that I think about a lot, which is the question of what we mean by randomness. And certainly to take a Bayesian point of view where probabilities are all subjective,
It becomes very difficult to say what randomness might mean if used in a way that's meant to be objective or to refer to something objective. And we model things like fluctuations in the vacuum as random, which from a Bayesian point of view just means it's uncertain to us. It represents information that we
don't have and quite possibly can't get even in principle. But if we think of that system as an agent, then of course, we're expected to be exploring its world, whatever it is, whatever exists, that's not it. And maybe we can really think of babbling all the way down.
as a model of what's happening in these phenomena that we persist in thinking of as random. But we also don't typically think of these sorts of systems as agents. This is incredibly interesting.
What I heard Carl saying, among other things, is that there is a qualitative transition when things go from analog to digital in a certain sense, from continuous to phased.
And based on what Chris just said, if you just remind us to think about the kind of the quantum world, where at least to my understanding, a lot of things are in fact, you know, sort of discrete. Does that mean that really could it be that that is the sort of base state as Chris just outlined with that kind of proto cognitive exploration all the way down?
that these continuous models that we put on top of it by coarse-graining, you know, tiny events into the kind of macroscopic things that we see here, that's an abstraction that sort of loses the essential mindfulness of it and then we have to try to rebuild it again by the time we get to brains. But actually, all we've done is obscure the fact that all the way from the beginning, it was already digital to start with,
And we just sort of put some Vaseline on our lens here to make everything look sort of continuous, which then made it look like there was nothing going on at the lower levels. And then somehow we're shocked that we have to recover it from somewhere later on. That's what I just heard from the combination of those two explanations. A very interesting idea, and it connects nicely to the history of mathematics, or at least the history of mathematics post Descartes and Leibniz.
Carl, did you have any thoughts to add to that? Yeah, I'd probably
This is why I wanted to hang out with Mike and Chris to learn about quantum theory, because I do think there's something fundamental about the way that we carve up the world or the way the world can only be defined in a carved way into tiles or quanta or discrete things. So I think that those observations that we just heard I think are quite fundamental and it would be great to understand
everything from a sort of a quantum perspective, a discretized perspective, and how continuous constructs like, you know, space and time emerge from that, you know, if necessary. Having said that, I have to say that as a physicist, I would bring something else to the table. So I'm quite comfortable dealing with random fluctuations. So for me, random fluctuations are just variables that change very, very quickly.
So if I want to write down as a physicist, so now I talk about a physicist that doesn't know anything about quantum mechanics. So I just want to write down and launch that equation to describe my universe. I write it as a mixture of states and random fluctuations and somebody asked me what license is that? I just say random fluctuations are very, very quick and the states are very, very slow.
And this is just a natural consequence of some kind of renormalization group. You know, I could make them even slower by going to the next state scale and throw even more fast stuff. So there's nothing really, if you like, problematic from the point of view of that kind of physicist in the random fluctuations. It just means that all the
stuff that determines flows and dynamics of the kind that we can that we engage with and we can plan has a particular form. That particular form does actually bring a quantile aspect to the table or a discretization aspect to the table that may or may not be if you like
But what I'm talking about here are the mathematical image of life cycles and biorhythms and oscillations. And I mean that right from the perspective of a dendrite in my hippocampus oscillating away at a fast gamma frequency, you know, sort of say 80 Hertz through to the motion of the heavenly bodies. Wherever you look, you get this
solenoidal flow that is characteristic of things that hang around and don't dissipate almost immediately and these are not part of the random fluctuation but they become more evident as you go away from the fast random stuff into the slow dynamics in the sense that you know clearly the orbit of the moon around well the orbit of the earth around the sun for example
This deals with very, very massive, large things where all the random fluctuations are effectively averaged to zero. Whereas if you go down to the quantum scale, then you're not licensed to do that. And these solenoidal flows become much less apparent. But our scale, if life is, on this reading of physics, sufficiently classical to have
oscillations, biorhythms, things that would you need for feedback that didn't go off to plus or minus infinity, for example, just saying that you're circulating in some state or phase space in a way that you keep yourself to some attracting set. And as you move through this, you are naturally going to be oscillating in a highly nonlinear possibly chaotic way, but still the essence is life cycles.
If that's the case, then the Poincare section does indeed permit a description that is necessarily discretized in the sense you're going to pass through the neighborhood of various states at discrete points in time recurrently. So that's what I was thinking from the point of view of a physicist who does not have the fluency of quantum mechanics to get back to a discrete thing, but it also I think highlights the
The deep connection between things recurring discreetly or occurring discreetly.
being one way of defining another characteristic of biological self-organization, which is basically, I repeat, biorhythms and life cycles that have this sort of self-contained itinerancy. Of course, in that self-containment, you've got this natural resolution of ambiguity and uncertainty because you always know you're on the same itinerant path.
repeatedly visiting the states that you've once occupied. So that's what I would say from the point of view of a physicist. I may say something very different in a few years' time when I've learned more about quantum theory. So let me ask Chris, could you ever describe that kind of physicists' treatment of dynamics in pure quantum information theory at that kind of scale? Well, I think if we try to think of scaling up from
Descriptions of things like quantum fields, then this looks impossible. But if we think of it informationally, maybe it doesn't look impossible. I'll go back to John Wheeler's notion that a bit is fundamentally the answer to a yes or no question. And in
The information theory is about bits and systems that exchange bits. But... Do you mind repeating that last part? But what? What Wheeler's aphorism doesn't specify is the complexity of the question that has a yes or no answer.
So in your case, Carl, you have an oscillatory system that passes after a discrete time of particular boundary. So the system may well be asking the question, have I passed this boundary? And the answer is yes or no. Or am I passing the boundary right now? Well, that's one bit.
That's a yes or no question, but it's an extremely complicated question. So it may be that we can develop a theory that leaves the complexity of the generative model unspecified and concentrates just on the bit flow across the boundary that
counts as evidence in one direction and perturbative action in the other direction. And then I think the two pictures fit together rather nicely, because one can imagine the generative model being the model that generates the questions, operating at many different scales.
Where one could think of the scales as computational complexity. And this gets back in a sense to your question about language. Language is a way of posing questions. It's a way of babbling in the world. This process of babbling, it seems to me, as far as I understand, it's like poking a prodding to test your own self-model to see what's the difference between me and the world. It sounds like that
never stops, even though we call it babbling as if infants do it. Does it ever stop? Are we not engaging in this right now? And I like this phrase babbling all the way down. I'm going to put that on a t-shirt by the way, Chris, or, and it also, by the way, seems like babbling all the way up too, because we're constantly engaging in this. So firstly, is that true? Does babbling have an end? And, and also does it not presume a self in order to test the self?
What I mean is, Carl, when you were speaking about the process of babbling, it's the infant doing something on the world in order to make a model. But that first part was it is doing something to the world to me already presumes the external world. Whereas the way that I'm understanding it is that the self is a model, but it's also in your point of view, it's something objective. And I believe from my readings of Chris, I believe Chris is more on the non-dual end
Right, well, I mean, just to dispel the notion that the self is objective. No, I didn't mean that. I meant it exactly the same way that I think both Chris and Mike have been intimating that, you know, from the inside of a cell or a person or the baby's skull,
Self is quite a sophisticated construct, fantasy, hypothesis, that is physically realized, possibly neuromally, in some lucky creatures and some lucky children. You may actually find some children never get that far. I'm thinking of people with severe autism, for example, who don't have a theory of mind, simply because they never developed the notion that I am a person.
And, you know, what comes along with that is that, you know, if you've got very, very severe autism, or, you know, if you like, sort of a theoretically idealized autism, that means that personhood doesn't exist. So you would be, I would regard you as an animate fringe or, you know,
some interesting autonomous vehicle, but you would never be a person because I don't have the hypothesis that there are people out there and that people share certain sort of phenotypic characteristics and have intentions in the same sense that I have intentions. I'm not aware of that because I just don't have that hypothesis. So it's a long-winded way of saying that I didn't mean to imply
any objectivity here. I was more going the other way, really. I'm just saying that selfhood, that would be one part of consciousness. And if you want to associate cognition and consciousness, then possibly that kind of minimum selfhood would be necessarily part of cognition. This is a gift that you have to get by crossing one of those bright lines.
and it's just a fantasy that you bring to the table to make sense of all the impressions on your sort of sensory surface or your holographic screen or your cell surface or also the receptors you are equipped with to sense what's going on out there. Does that make sense? Yeah, I know that Mike and Chris, you may have a response to that, but I just want to quickly respond to what Professor Friston was saying.
And this is to both of you. How do you then prevent an existential crisis as you start to study this? So for example, Carl, before you and I were speaking, and I told you about some experiences I had where I've deeply felt what you're saying is being true, that the self is a fantasy, and this oneness with all this world being generated in one's mind. And I felt like that was akin to a psychotic episode. And it was extremely destabilizing. And to be quite frank, I'm still recovering from that.
And so I'm unsure how any of you all do this work without constantly being in a debilitating state of existence. Well, existence is the question here, and it's an existential crisis which I'm inquiring about. So how is it that you prevent yourself from getting into that? How could you say these words of, well, the self is not objective, the self may be a fantasy, quote unquote, but not then be able to not speak and look around you in awe and
I think there is a good deal of existential uncertainty involved in doing what we're doing. Just par for the course. Right. How do you get over it? I'm not sure when gets over it.
I think it's more being comfortable with it. You know, I think you raise a really important point and I think this is in a class of a number of really destabilizing
uh ideas that one picks up as as one gets older and studies things and there are many of these. I remember as a as a child being completely freaked out by the fact that I realized that you know we grow up and we become these adults and these adults bear some resemblance to the children but they're definitely not the same thing and so in some sense you could sort of see it coming it's almost even even before you realize that you're gonna you know you're gonna you're gonna die biologically someday before long before that you realize that you're going to undergo some transition
that's going to, you know, turn you into a completely different sort of creature, which really, you know, has some continuity, but maybe not, you know, certainly not you persisting in the same way as you are now, right? So some kind of weird butterfly, you know, caterpillar butterfly effect where you're just not going to be around in that same way as you are now, right? So that's kind of the first
thing you realize as a child. And then there's many others that sort of, you know, you learn about Boltzmann brains and all the different, you know, Humean arguments of it, you know, all this stuff that, you know, I had to, I did, I did a philosophy class with my son when he was little, we were doing homeschooling. And so I kind of did this. And I realized very quickly that I have to be very careful with this stuff, because if you write somebody who has the capacity to understand it, it's extremely destabilizing.
And I think that at this point for myself, I sort of go with Descartes a little bit in the following sense. I'm not depressed by it in the sense that whatever we discover is a truer model of the self than we may have had. I don't think there's any way in which it makes us less valuable, less interesting, less engaging.
All of that is still true. It's just that now we've realized that a different set of mechanisms can now implement that magic than we thought. So you may have had some sort of idea that, well, there has to be some sort of
different kind of unique process that makes this permanent self and therefore I'm this individual and I have true beliefs and preferences and I exist in some way. I don't think you can ever convince yourself that that isn't true and that you don't exist. It's just that now you find out that actually the way you get to be that is by these sets of mechanisms and they're very interesting. It's the sort of thing that Karl and Chris have been talking about.
So that's what gets you to be this kind of individual, and it's kind of amazing that that works. But to me, that doesn't diminish all of the things that I thought were true about what the individual actually is. The impermanence is there, but we knew that from the time you were a kid, you knew that it wasn't going to be permanent.
So that that part, you know, you sort of have to deal with from the start. After that, it's all you're learning after that is just that you were really mistaken about what it takes to be that sort of creature. You thought maybe it had to be some magical whatever. Now you think that, wow, it's a set of feedback loops and whatever else. So that's surprising. But that's OK. We learn about, you know, we learn about different underpinnings to things all the time. So that's OK. But I don't I don't think any of it diminishes the sort of just the incredible awe of being a sentient being that can think about these things.
Yeah, just to reiterate that. So I didn't in any way mean to demean our selfhood, or your selfhood in particular, by calling it a fantasy, I actually meant it, and indeed have written about it, in terms of something quite fantastic. So we've written papers called the fantastic organ, the brain is a fantastic organ, just because it can produce these fantasies is absolutely fantastic. And of course, they
in relation to our earlier conversation, the capacity to have these fantasies which have incredible explanatory power, if you can attain them and you can maintain them, is the part of the complexity which allows us to cross these bright lines or these gaps or cuts and become closer to the kind of creatures that can even entertain the notion of the heart problem.
Mike, I just love that meta-answer. We don't know what the answer will look like. I'm going to tell Andy Clark about that. He'll tell David Choms. To come back to the point, I didn't mean to demean that. Phrases that I remember coming across and having to learn when I was a student, things like existential angst and ontological security, all of these things rear their
face if you are worried about these things but the very fact that you are able to entertain the notion that you are not real is quite remarkable and I repeat it is a gift of a fantastic sort it is the kind of fantasy that very few creatures have and sometimes even few people have so there are people who actually will go out there to try and
experience that and to become and treasure the the gift of being able to make sense of the world as me by taking drugs or doing mindfulness therapy in order to actually get to a state where almost a Freudian oceanic state where selfhood is dissolved so they can experience it and certainly if you've taken sort of psychedelic drugs in sufficient quantity
You may have that kind of experience, that's a derealization or dissociative experience, which is the realization that your construct, your fantasy of selfhood that is engaged and engaging continuously for as long as you live, then
you know, if you want to appreciate that, then you have to experience that, you know, the alternative hypothesis. So, you know, experiencing derealization, or this is depersonalization, is actually, I think, one really important step to self understanding. So I don't see it as at all flattening nor diminishing. It's just part of self understanding, which not everybody gets.
I mean, in some sense, the most destabilizing thing of all ought to be just the bare facts of developmental biology. When you learn that you are in fact not some sort of indivisible monadic whatever, which I'm not even sure what that would entail if you were, but the fact that all intelligences are collective intelligences, we are all made of parts, and the idea that you used to be this quiescent oocyte, this little bag of chemical reactions,
that did very little and then very slowly sort of step by step you turned into whatever you are now. If that's not destabilizing I don't know what is because just knowing nothing knowing nothing else about evolution or I mean to me developmental biology is sort of like the you know the queen of all the sciences because just that fact that you go from chemistry and physics to mind in nine months plus to however many years it takes and that you are literally a collection of cells that you could see in pond water more or less
is just if you know you're a bunch of cells in the trench coat you know like the in the cartoons like if if that's not destabilizing i don't know what i don't know what else is so quickly just i know we have to get going and i'd like to wrap with a set of brief questions but quickly carl and perhaps well anyone who would like to answer what is it that separates those who are depersonalized or go through certain psychotic episodes from
What is it that separates those who are fortunately like myself where it's abating and I'm able to function and I'm able to have friends and so on versus those who go into a ward? Because obviously those issues can, like you mentioned, it's a gift.
Well, I'll answer from the point of view of a psychiatrist with a commitment to sort of a mechanistic or theoretical understanding
I think it's the inability to resolve the uncertainty that attends entertaining the dual hypothesis that I may exist or I may not exist as me. But as soon as you increase your hypothesis space, you naturally induce an uncertainty, an ambivalence and a kind of ambiguity. So it is exactly the same, if you like,
We're talking about the same imperatives that we've been talking about all along, which is to sort of resolve uncertainty, resolve ambiguity, to make the most sense of things. So if you've entertained this hypothesis that you are not you and you are you and you can't fully resolve that, then you will find yourself in a state of chronic uncertainty.
That unfortunately does have pathological consequences. I'm not sure that being in a psych ward is necessarily too pathological. I mean, you know, there are ways to live very nicely and constructively and fantastically in psych wards. And I've done that. I spent two years in a therapeutic community. But there will be other sequelae of an inability to resolve uncertainty. They're expressed clinically and physiologically in terms of allostatic load and the like.
chronic stress and the like, which can be very unpleasant and ultimately can have physical consequences from the point of view of the effects on your body. So the question now is why can't some people resolve that uncertainty? I think it's then a question of the biophysical processes that enable you to update your beliefs and build good models of your world.
Again, this is a very pragmatic answer from the point of view of a clinician. Probably the answers lie in the way that your brain has assembled itself to coordinate and attend to various sources of evidence and it looks at the mechanisms that underwrite that ability to deploy attention rest very much upon
Neuromodulating neurotransmitters are the kind you know you have heard about in the popular media and will have read about things like dopamine, serotonin and indeed oxytocin. All the drugs that are either used or the targets of psycholytics or psychedelics or psychomimetics or drugs that are used to try and control some of these feelings and some of these signs and symptoms specifically that attend
anxiety disorders and depression. So, you know, understanding the, this is not the feedback loops that do self-organization in a developing embryo, but the same kinds of mechanisms that do the feedback loops and the self-organization of different structures within the brain that can be construed as blankets within blankets or holographic screens within holographic screens or cells within cells.
or complying with the same principles, but in the particular instance of the brain doing this and understanding the neurochemistry and the biophysics of this fundamental sense-making self-organization. It turns out that these are certain, these neurochemical messages
are particularly important in resolving that kind of uncertainty. And it may be that you have either taken drugs or you have just grown up to not be able to deploy these neurotransmitters properly. Again, leading us to, well, how would you gain control of it? Well, you might want to do mindfulness training, meditation and the like, just internal attention state training basically to mentalize
how you deploy your attention.
Mike, I know that you had a set of questions that you're eager to ask both Carl and Chris. Did you already get to all of them?
That would be impossible, but I got to many of them, most of the important ones. I learned a massive amount today. This has been just amazing. But I will be pestering both of them with additional questions in years to come. Great. To end, I have a question that I didn't email you that just occurred to me as you all were speaking. It's something that Professor Levin that you said, what are some aspects of your field, maybe their theories or concepts that
your peers are extremely resistant to, but you are not. Wow, I don't know if we got time for all that. That's a lot of stuff. I mean, the most basic, there are many, I mean, the most basic thing is this notion of profound symmetries between these fields. You know, this idea that you can basically import
concepts and strategies from cognitive neuroscience or behavioral science and deploy them in other spaces, developmental biology, metabolism, that it even makes sense to talk about, you know, goals, memories, planning, navigation, all of these kinds of things outside of what we're used to, which is some, you know, rat in a maze or something like this. So that continuum, this idea that something important goes all the way down or maybe a long way down, even if not all the way down,
That is in a few circles, that's okay. In most circles that I deal with, that's just a complete heresy. Yeah, there are many others, but that's kind of a basic one. Chris? Well, I think that Mike really nailed it with that answer. And I would say
Yes, it's not a popular idea that physics and biology and psychology are actually all more or less the same science. But I think that they are. Professor Friston. Yeah, well, that would be my favorite heresy. I'm just going to agree with Mike and Chris. You know, physics, biology, psychology, they're all the same thing.
And I just wanted to end by saying, heresy, I'm not so sure anymore. The number of times, for example, I see Mike Levin cited in high-end philosophy journals and treatments suggests to me that, in fact, it's going to be the minority, at least in a few years' time, that don't subscribe to this more holistic encompassing, I think,
And I'll end with one that I see from particularly Chris and Michael is it sounds like
without using the word, it sounds like what you all are doing or trying to understand God or trying to come up with a model of God. And the reason why I say that is because with looking at a neuron and saying, okay, how can a neuron look around and have some model and know that it's embedded in something higher? That sounds to me like, well, if we apply that to ourselves and continually do so, we would reach something akin to God. And then Chris with the quantum babbling, if one can think of it like that. Well, what is babbling?
I know that maybe you all don't want to use the word God, but for me, what I have that many of my peers are extremely against would be to philosophize about God or to explicate God or to take God seriously as not an old man in the sky with a beard. And so I see you all as doing something similar in them. I'm happy to be in your company. I'm extremely lucky.
all science starts with an act of faith, with a very important act of faith, which is that the world is in some way understandable, that we are not just a bubble in this, you know, random universe that just happens to, you know, as a total random string of coincidence looks like it's got laws, and then tomorrow somewhere, you know, while you sleep, the whole thing falls apart, and we go back to a random, you know, sort of random distribution of events.
So we all there's no proof of any of it of any of that right we take we have to take that on faith and that may be for a scientist that may be the most destabilizing thing if you can't get yourself to believe that that the world isn't for some weird reason understandable and amenable at least to some extent to our logic and probing then then you can't do science you can't do any of the things that that we do so
So being useful and hopefully productive in what we do starts with that act of faith that is just completely, to my knowledge, it's an axiom. It's not based on any evidence that you could possibly have. So I do think that sort of underlies everything that we do, just this idea that we can in fact squeeze some understanding out of what's going on around this.
I'd also think you'd be surprised. I know that you said that it doesn't comport with the traditional definition of God, but it depends on what tradition and how far back one goes because there are mystics, let's say Christian mystics and Muslim mystics or Islamic mystics and so on. And obviously then the eastern end is much more mystical. So, okay. Well, thank you all so much for coming out to this and I've learned a tremendous amount and I hope you all have too.
I'm blessed and thank you so much. Well, thank you very much. This was this was a very interesting conversation and thank you very much for putting it together. Yeah, absolutely. This is this is great. Great, great idea. Thanks for having us on.
Bye everyone. Bye. Take care.
The podcast is now finished. If you'd like to support conversations like this, then do consider going to patreon.com slash C-U-R-T-J-A-I-M-U-N-G-A-L. That is Kurt Jaimungal. It's support from the patrons and from the sponsors that allow me to do this full time. Every dollar helps tremendously. Thank you.
▶ View Full JSON Data (Word-Level Timestamps)
{
"source": "transcribe.metaboat.io",
"workspace_id": "AXs1igz",
"job_seq": 9922,
"audio_duration_seconds": 7556.78,
"completed_at": "2025-12-01T01:35:05Z",
"segments": [
{
"end_time": 20.896,
"index": 0,
"start_time": 0.009,
"text": " The Economist covers math, physics, philosophy, and AI in a manner that shows how different countries perceive developments and how they impact markets. They recently published a piece on China's new neutrino detector. They cover extending life via mitochondrial transplants, creating an entirely new field of medicine. But it's also not just science they analyze."
},
{
"end_time": 36.067,
"index": 1,
"start_time": 20.896,
"text": " Culture, they analyze finance, economics, business, international affairs across every region. I'm particularly liking their new insider feature. It was just launched this month. It gives you, it gives me, a front row access to The Economist's internal editorial debates."
},
{
"end_time": 64.514,
"index": 2,
"start_time": 36.34,
"text": " Where senior editors argue through the news with world leaders and policy makers in twice weekly long format shows. Basically an extremely high quality podcast. Whether it's scientific innovation or shifting global politics, The Economist provides comprehensive coverage beyond headlines. As a toe listener, you get a special discount. Head over to economist.com slash TOE to subscribe. That's economist.com slash TOE for your discount."
},
{
"end_time": 89.189,
"index": 3,
"start_time": 66.169,
"text": " This is Marshawn Beast Mode Lynch. ProjPic is making sports season even more fun. On ProjPic, whether you're a football fan, a basketball fan, it always feels good to be ranked. Right now, new users get $50 instantly in lineups when you play your first $5. The app is simple to use. Pick two or more players. Pick more or less on their stat projections."
},
{
"end_time": 104.565,
"index": 4,
"start_time": 89.189,
"text": " anything from touchdown to threes and if you write you can win big mix and match players from any sport on prize picks america's number one daily fantasy sports app prize picks is available in 40 plus states including california texas"
},
{
"end_time": 126.237,
"index": 5,
"start_time": 104.838,
"text": " Florida and Georgia. Most importantly, all the transactions on the app are fast, safe and secure. Download the PricePix app today and use code Spotify to get $50 in lineups after you play your first $5 lineup. That's code Spotify to get $50 in lineups after you play your first $5 lineup. PricePix. It's good to be right. Must be present in certain states. Visit PricePix.com for restrictions and details."
},
{
"end_time": 142.739,
"index": 6,
"start_time": 126.459,
"text": " Listen up, people. Over the years, I've learned how important hydration is in my workouts and everyday life. It's the key to helping my body move, recover, and just have a good time when I'm exercising and staying active. Things go even better when I'm well hydrated before I even start moving."
},
{
"end_time": 161.63,
"index": 7,
"start_time": 142.739,
"text": " Noon Hydration doesn't want you to wait to hydrate. They want you to start hydrated. Noon Sport Hydration tablets hydrate better than water alone. Just drop them in your water, dissolve, and enjoy. They contain five essential electrolytes and come in crisp and refreshing flavors like strawberry lemonade, lemon lime, and many more."
},
{
"end_time": 186.186,
"index": 8,
"start_time": 161.63,
"text": " They also have non-GMO vegan and gluten-free ingredients and only one gram of sugar. Noon hydrates you so you can do more, go further, and recover faster. And that means you can have a heck of a lot more fun. Since hydrated humans have the most fun, head over to ShopNow on NoonLife.com and pick some up today so you're ready for anything. Because anything can happen after Noon."
},
{
"end_time": 211.084,
"index": 9,
"start_time": 187.585,
"text": " I won't let my active psoriatic arthritis joint symptoms define me. Emerge as you. Tremphia gousselkumab is proven to significantly reduce joint pain, stiffness, and swelling in adults with active psoriatic arthritis. Some patients even reported less fatigue as assessed by survey one week prior. Results may vary. Tremphia is taken by injection six times a year after two starter doses at weeks zero and four."
},
{
"end_time": 230.009,
"index": 10,
"start_time": 211.084,
"text": " Serious allergic reactions may occur. Trimphia may increase your risk of infections and lower your ability to fight them. Before treatment, your doctor should check you for infections and tuberculosis. Tell your doctor if you have an infection or symptoms of infection including fever, sweats, chills, muscle aches or cough. Tell your doctor if you had a vaccine or plan to."
},
{
"end_time": 245.316,
"index": 11,
"start_time": 230.009,
"text": " Emerge as you. Learn more about Tremphia, including important safety information, at Tremphia.com or call 1-877-578-3527. See our ad in Food and Wine magazine. For patients prescribed Tremphia, cost support may be available."
},
{
"end_time": 277.21,
"index": 12,
"start_time": 248.797,
"text": " As always, you can click on the timestamp in the description as well as here to skip this longish introduction. The Theories of Everything channel is back with a new season, starting with last week's AMA, which will be in the description. From this point forward, there will be new content approximately once a week, sometimes even more frequently. Today's guests are Carl Friston, Michael Levin, and Chris Field. The former two have been on this channel and mentioned several times prior, whereas Chris Fields is making his debut."
},
{
"end_time": 299.343,
"index": 13,
"start_time": 277.21,
"text": " Chris Fields is a researcher in topological quantum field theory as well as information theory and publishes work on what constitutes an observer. Carl Friston is a professor of neuroscience at the University College London and is the inventor of the free energy principle. Michael Levin is a developmental and synthetic biologist at the Tufts University. The links to all their previous podcasts"
},
{
"end_time": 316.869,
"index": 14,
"start_time": 299.343,
"text": " on the Toe channel will be in the description. My name is Kurt Jaimungal. I'm a Torontonian filmmaker with a background in mathematical physics dedicated to the explication, quote unquote, of the variegated terrain of theories of everything, primarily from a theoretical physics perspective, but as well as trying to understand"
},
{
"end_time": 345.657,
"index": 15,
"start_time": 316.869,
"text": " What role does consciousness have as constitutive reality? Is it emergent? Is it fundamental? Now the goals of theolocutions is to interject seldomly, if at all, allowing each guest to give their perspectives on one another's thoughts, essentially giving us the experience of a fly on the wall for the sorts of academic conversations that would ordinarily occur behind closed doors, spurring research in real time. One of the central issues of today's podcast is the concept of babbling, which I haven't heard discussed"
},
{
"end_time": 373.626,
"index": 16,
"start_time": 345.998,
"text": " Virtually anywhere else, outside of child development or language acquisition, though it can be generalized metaphorically, perhaps even literally, to the vacuum fluctuations, to interpret what vacuum fluctuations are doing as an indicator of the universe's proto-consciousness. That is, the universe babbling to understand itself, but I won't spoil the surprise. If you'd like to hear more podcasts like these, then do consider going to patreon.com slash Kurt Jaimungal as the patrons and the sponsors are"
},
{
"end_time": 397.125,
"index": 17,
"start_time": 374.036,
"text": " The only reasons that I'm able to do this full time. With regard to sponsors, there's one for today's episode, and that is Brilliant. During the winter break, I decided to brush up on the fundamentals of information theory, which is what constructor theory is heavily based in, and I'd like to do an episode on that. So I took Brilliant's course on knowledge and uncertainty and random variables, and after taking that course, I could finally see why entropy is defined the way it is,"
},
{
"end_time": 425.555,
"index": 18,
"start_time": 397.125,
"text": " why the formula for it is extremely natural. There are plenty of courses. You can even learn group theory, which is what's being referenced when you hear that the internal symmetries of the standard model is U1 cross SU2 cross SU3. Those are Lie groups. Visit brilliant.org slash toe to get 20% off the annual subscription. I recommend that you don't stop before four lessons. Just keep pursuing until you've accomplished at least four. And I think you'll be greatly surprised"
},
{
"end_time": 451.305,
"index": 19,
"start_time": 425.742,
"text": " At the ease at which you can now comprehend subjects you previously had a difficult time grokking. And now on to today's episode, which is one that I'm extremely lucky to be able to present to you all. You'll be able to see the mutual respect that the guests have for one another. We start off actually by asking, what is it that you respect about each other's work? Enjoy this theolo-cution between Carl Friston, Michael Levin, and Chris Fields. I've been extremely excited about this."
},
{
"end_time": 475.879,
"index": 20,
"start_time": 451.561,
"text": " Professor, what do you find unique about Carl's and Chris's work and what is it that you respect about them? Well, what a great question. I think all discussions should start with that question. There's a lot. I mean, I think the most basic thing I can say is this. I'm often asked by conference organizers and students as far as who they should follow, whose work they should keep track of."
},
{
"end_time": 498.473,
"index": 21,
"start_time": 476.032,
"text": " And so I keep lists. And Carl and Chris are at the very top of two specific lists for me. And I'm just always amazed about this is that the first is that both of them are experts in at least three different fields, but probably more. So they have the ability to kind of merge really deep understanding of different fields. I mean, there's many people that are experts in one thing,"
},
{
"end_time": 520.094,
"index": 22,
"start_time": 499.07,
"text": " But I'm just always incredibly impressed at how both of them can weave together really deep knowledge from diverse disciplines. And that leads to the kind of the second thing that I'm really inspired by, which is just the kind of the sheer density of new ideas. You know, it's hard enough to push things in one just kind of one known direction and do science and so on."
},
{
"end_time": 549.036,
"index": 23,
"start_time": 520.316,
"text": " Both Chris and Karl's work consistently make me think in new ways, give me ideas I never thought of before. It's just amazing. I'm super excited to be able to work with both of them and to have this discussion. Great. Chris, same question but toward Michael and Karl. Well, I can say that one thing I certainly admire is that"
},
{
"end_time": 579.087,
"index": 24,
"start_time": 550.213,
"text": " Michael and Carl are both top-level biologists who are interested in theory and deeply involved in theory and have a deep understanding of theory. And something that I enormously admire is that both of them have that interest and yet are deeply engaged in practical applications."
},
{
"end_time": 609.906,
"index": 25,
"start_time": 580.299,
"text": " Carl in mental health and Mike in regenerative biology. And that's a rare combination. Carl, what is it that you find unique about Michael's and Chris's work and what do you particularly respect about them as people perhaps? I deliberately haven't rehearsed an answer. So what I say comes from the heart. So with Mike,"
},
{
"end_time": 634.326,
"index": 26,
"start_time": 610.52,
"text": " I first met him vicariously through a friend Giovanni Pizzullo and didn't really know very much about his work until I visited his website and then I realised how influential and important he was. And then I recognised all the little bulletins you get on social media and emails. I started to recognise his name and realised that he was quite a mover and shaker out there."
},
{
"end_time": 660.316,
"index": 27,
"start_time": 635.162,
"text": " What I like about both of them, they both think really, really fast and they both think out of the box. So I think what characterizes both of their thinking really is just to ask what is a thing in and of itself without making any prior assumptions and then putting it back together from basic principles and coming up with some sometimes counterintuitive"
},
{
"end_time": 680.009,
"index": 28,
"start_time": 661.015,
"text": " conclusions. But for me, almost universally, exactly the same counterintuitive conclusions I've come to after about 10 years of thinking about a particular problem, but never dare tell anybody. But they do. They write it down very eloquently. They're very, very productive."
},
{
"end_time": 702.244,
"index": 29,
"start_time": 680.401,
"text": " Chris I've known for less long and I've met him via Mike and in addition to that in the same way that many people I meet have physics envy and I've got quantum envy about Chris. He just seems to know so much and think about the world in a way which I don't have that sort of fluency"
},
{
"end_time": 731.049,
"index": 30,
"start_time": 706.698,
"text": " What would be an example of this out of the box thinking? You mentioned thinking about what is a thing in and of itself. What else? There are numerous examples. The first thing that just came to mind was just putting together"
},
{
"end_time": 757.5,
"index": 31,
"start_time": 731.749,
"text": " fundamental questions about self-organization. If you were Varela, you know, sort of autopoiesis, self-assembly, self-construction, if you're a chemist, it would be self-assembly. If you're a theoretical biologist, you will be wondering about why on earth is it that multicellular organisms form? Because there's a deep paradox there. If it is a case that I have to have a surface as a little organ,"
},
{
"end_time": 777.517,
"index": 32,
"start_time": 758.148,
"text": " Then some of my cells have to stop replicating, and yet that is in direct contradiction to the principles of natural selection. I'm going to turn off my reproductive capacity, my adaptive fitness, so defined in a theoretical context. So just really sort of grasping the nettle and thinking about, well, these things exist."
},
{
"end_time": 801.954,
"index": 33,
"start_time": 778.148,
"text": " what principles could possibly explain the paradoxes when looked at from sort of the unilateral, the monothematic view of natural selection for example, or the monothematic view of self-organization in physics. So thinking out of the box literally in this instance entails being able to take multiple perspectives on a particular problem and seeing the contradictions and seeing how they can be resolved."
},
{
"end_time": 812.432,
"index": 34,
"start_time": 802.363,
"text": " Now, that would be one of many examples. There was a paper, and I can't remember the specific one, perhaps Mike and Chris will be able to enlighten me, but there was one paper."
},
{
"end_time": 842.21,
"index": 35,
"start_time": 813.08,
"text": " written I think about a year or two ago where they actually listed I think were 15 basic conclusions about the the fundamentals and the nature of self-organization and self as distinct from other in any system and this listing 15 really interesting points and predictions all of which either I'm sure will be formally demonstrably analytically true"
},
{
"end_time": 870.862,
"index": 36,
"start_time": 842.602,
"text": " There was a paper in Neuroscience of Consciousness arguing for an approach to consciousness that spanned phylogeny and we were very interested in that paper"
},
{
"end_time": 901.613,
"index": 37,
"start_time": 872.227,
"text": " specifically to look at bacteria and single cells and facultative multicellular such as microbial biofilms and ask what do these systems know about the world? How do they see the world? How does E. coli perceive the world, for example?"
},
{
"end_time": 930.486,
"index": 38,
"start_time": 903.148,
"text": " And we suggested that the answer is in terms of the sorts of things that we call taste and viscosity, but that experiences of taste and viscosity are perfectly good kinds of experiences. And that if we think about how these kinds of systems deal with their worlds,"
},
{
"end_time": 959.053,
"index": 39,
"start_time": 932.398,
"text": " and solve problems within their worlds, that that would give us a more unified understanding of what it meant to be a system that was aware. And so you're trying to use E. coli as a means of understanding yourself or humans. By the way, is this a project? Do you feel like your undertaking is an attempt for you to understand yourself? Are you trying to understand humans?"
},
{
"end_time": 980.862,
"index": 40,
"start_time": 959.394,
"text": " I'm wondering, is this more philanthropic or it's selfish and then it bleeds into the philanthropic? Well, I actually also want to understand E. coli. We don't have a great understanding of the lives of other organisms. Professor Levin,"
},
{
"end_time": 1008.387,
"index": 41,
"start_time": 981.647,
"text": " What puzzle do you find most important? What do you think about on a semi daily basis or even perhaps multiple times a day? And then we'll go around the table here with Chris Fields and Karl Fursten next. And then it's essentially me taking a backseat and allowing you all to free flow just for the audience to know. I know you all know this template already. Yeah. So the thing I think about many times a day has to do with the scaling of cognition."
},
{
"end_time": 1031.51,
"index": 42,
"start_time": 1008.677,
"text": " I want to understand two major things. One is how it is that some collection of competent parts comes together to form an emergent self with preferences, goals, memories, cognitive capacities that belong to it but not to the individual parts. I want to understand how that emerges, how the goals of humble"
},
{
"end_time": 1054.957,
"index": 43,
"start_time": 1031.92,
"text": " Simple kinds of systems scale up into much more grandiose goals that we see during development, during behavior, during culture, and so on. And I'm also very interested in the left side of that spectrum. Where does it begin? Is there really a zero on the spectrum? I think that all of us here would agree that there is a spectrum for these things. It's not a set of binary categories."
},
{
"end_time": 1079.258,
"index": 44,
"start_time": 1055.316,
"text": " But what happens at the very left side of the spectrum? And so one of the first questions when we get to that, that I had written down to ask both Chris and Carlos to sort of comment on what does the Venn diagram look like of the set of things that are alive versus the set of things that are cognitive? How do those two categories relate to each other? Do they overlap as one a subset of the other? And what really happens at the very beginning?"
},
{
"end_time": 1108.677,
"index": 45,
"start_time": 1079.753,
"text": " Can we can we develop a kind of which which I think I think both of them have been working on a kind of basically a kind of panpsychism that doesn't doesn't sort of paint on new cognitive mysteries on top of a physics that works perfectly well, but instead to try and to try to view physics from the bottom up as having already a useful cognitive lens on it. And how does that help us to build up cognition? So that's that's something that I think about every day. Chris,"
},
{
"end_time": 1138.507,
"index": 46,
"start_time": 1111.561,
"text": " I thought every day or almost every day for a long time about why we humans see objects embedded in space time. So why do we see things that we treat as independent of each other? And why do we"
},
{
"end_time": 1168.2,
"index": 47,
"start_time": 1139.036,
"text": " see them, operate on them, interact with them, think about them as embedded in this coordinate system that we call space. And equally important, why do we see them as maintaining their identities as things over this other coordinate that we call time? Carl."
},
{
"end_time": 1193.473,
"index": 48,
"start_time": 1170.06,
"text": " I have to confess I spend recurrently most of my time thinking about me, but in an academic sense, you know, how do I work? It's a curious mixture of introspection and trying to understand why I can introspect from the perspective or through the lens of a physicist. So I spend most of my time"
},
{
"end_time": 1217.329,
"index": 49,
"start_time": 1194.275,
"text": " dissembling preconceptions, the gifts that sentience has given us, and try to reassemble them in relation to physics, density dynamics, that has to be articulated in terms of the kind of maths that a physicist would use, which would be effectively differential equations and the calculus of variations."
},
{
"end_time": 1245.811,
"index": 50,
"start_time": 1217.637,
"text": " And having done that, there are so many, there are a sufficient number of moving parts, but not that many. But the combinatorics then lead you to the kinds of questions that Mike was talking about. So what different ways could this physics of sentience be manifest? What possible ways could it be manifest? And how does that address the distinction between things"
},
{
"end_time": 1262.108,
"index": 51,
"start_time": 1245.998,
"text": " A KFC tale in the pursuit of flavor. The holidays were tricky for the Colonel. He loved people, but he also loved peace and quiet. So he cooked up KFC's $4.99 chicken pot pie."
},
{
"end_time": 1284.206,
"index": 52,
"start_time": 1262.108,
"text": " And to my mind an important aspect is planning."
},
{
"end_time": 1307.5,
"index": 53,
"start_time": 1284.599,
"text": " This would disambiguate them from other natural kinds or particular kinds that don't have that kind of facility and what underwrites them. So that's what I spend most of my time doing. I have to say in conversation both with Mike and Chris and other people that you can easily get distracted in a rightful way."
},
{
"end_time": 1322.722,
"index": 54,
"start_time": 1307.807,
"text": " by thinking about that mechanics in exactly what Chris was talking to, which is a sort of scale-free way. But in my rhetoric, that scale-freeness speaks more to the coupling between different scales that Mike was alluding to."
},
{
"end_time": 1345.845,
"index": 55,
"start_time": 1323.166,
"text": " So how does this mechanics, this physics of sentience, which I understand largely in terms of probability theory and effectively Bayesian probability theory. So what you have is a Bayesian mechanics. How does that Bayesian mechanics apply at one scale and another scale? And then of course the big question"
},
{
"end_time": 1380.111,
"index": 56,
"start_time": 1350.179,
"text": " is something that I find myself increasingly pressed upon, largely through conversations and collaborations, but that does mean that I spend a lot of my time thinking about that and writing demos and mathematical equations trying to try to get to the underpinnings of it. Mike, I see you nodding, so what is it about what Karl said, sorry, Professor Levin, I apologize, and what is it that Professor Friston said that has you"
},
{
"end_time": 1407.585,
"index": 57,
"start_time": 1380.52,
"text": " feeling like you agree with it? Well, as usual, I mean, he said it perfectly. I couldn't say it better myself. I think that's exactly the kind of research agenda that I'm interested in. I mean, that really is kind of a very good description of what I think we're all looking for in a certain sense. Is there a difference between cognition and consciousness? Is one distinct? Is there overlap? Is one a subset?"
},
{
"end_time": 1438.814,
"index": 58,
"start_time": 1408.831,
"text": " I actually can't conceive of cognition without consciousness, which I prefer to call awareness, not just to be perverse, but because in many cases, consciousness is used in a way that implicitly means self-consciousness, whereas awareness is often used in a way that does not make that implication."
},
{
"end_time": 1469.206,
"index": 59,
"start_time": 1440.179,
"text": " And I find it difficult to conceive of cognition without any kind of awareness at all. So, I'll ask a question based on what Chris just said and what Carl has to say about it. So, what do you think about the so-called hard problem? Is there in fact a hard problem at all? I would take Andy Clark's view on this, who has worked closely with David Chalmers."
},
{
"end_time": 1499.787,
"index": 60,
"start_time": 1470.657,
"text": " I don't think so. No, I don't think that the hard problem in and of itself is the interesting focus of inquiry. I think the move that I see in philosophy has been to the meta problem or the meta hard problem, which is, as Andy Clark puts it quite succinctly, is why do we spend so much time puzzling about why we are aware?"
},
{
"end_time": 1529.514,
"index": 61,
"start_time": 1500.299,
"text": " Just asking that question speaks to, I think, the nature of the hard problem, which is a sort of very much a second order, a meta capability that we can make sense of our own sense making. And we associate that with selfhood. So one has to ask the question, what kinds of systems, again, I think, particles or creatures would have the capacity to represent selfhood"
},
{
"end_time": 1556.51,
"index": 62,
"start_time": 1529.514,
"text": " And furthermore, represent the counterfactuals that would enable you to ask the question, why am I conscious? That immediately just logically presupposes that there is an alternative hypothesis of other counterfactual that I'm not conscious. So imagine now you're talking to a creature, you're talking to me, that has this capacity to imagine counterfactuals that cannot exist."
},
{
"end_time": 1579.957,
"index": 63,
"start_time": 1557.193,
"text": " So that's a remarkable capacity to have from my point of view a sort of internal world model or generative model where counterfactuals can exist is quite remarkable. I don't imagine that a thermostat or a virus or an E. coli will have a sufficiently sophisticated set of electrochemical"
},
{
"end_time": 1610.043,
"index": 64,
"start_time": 1580.213,
"text": " dynamics or kinetics that would enable them to physically represent counterfactuals of this sort. And I could go on, the usual line of argument is, well, why do we have models? Why do we have the capacity to represent counterfactuals? You end up with saying, yes, you have to have them just if you contemplate things that plan. So if things plan, they have to have"
},
{
"end_time": 1638.251,
"index": 65,
"start_time": 1610.401,
"text": " an ability to simulate the future because they have to simulate or represent at some elemental level the consequences of their action upon the world. And if they have now the capacity to represent something that has not yet happened, namely in the future, then they now have the capacity to represent counterfactual outcomes. So I think the meta-hard problem inherits simply because we have a sufficiently"
},
{
"end_time": 1652.125,
"index": 66,
"start_time": 1638.763,
"text": " Temporarily deep generative model that can represent the consequences of our actions and of course using the word our implies some kind of elemental agency and some kind of self"
},
{
"end_time": 1683.507,
"index": 67,
"start_time": 1655.503,
"text": " Can I pursue that just a little bit? I think you raise an extraordinarily interesting point here, Carl. And so I want to ask you,"
},
{
"end_time": 1713.899,
"index": 68,
"start_time": 1684.241,
"text": " From a phylogenetic perspective, do you think that organisms that can plan necessarily are able to represent multiple counterfactuals? So let's give an example. Suppose you're a cheetah chasing a gazelle."
},
{
"end_time": 1743.882,
"index": 69,
"start_time": 1714.974,
"text": " You're planning your moves while looking at the gazelles moves and your objective is success. Do you think that the cheetah worries about the other counterfactual condition where the gazelle isn't caught or does it just represent this one counterfactual condition that is in a sense the goal state? That's an excellent question. I personally think that"
},
{
"end_time": 1751.271,
"index": 70,
"start_time": 1745.401,
"text": " And when I answer these questions, I think, how would I simulate this? How would I put a teacher in silico?"
},
{
"end_time": 1780.657,
"index": 71,
"start_time": 1751.681,
"text": " and integrate according to the differential equations that must be present under certain assumptions and my guess is that the cheetah chasing the gazelle would as you say just be pursuing a path of least action and it would not have an explicit representation of alternative paths so it would be responding in the moment in a reflexive habitual way so that's a really important question because if we don't have or all of our behavior"
},
{
"end_time": 1797.961,
"index": 72,
"start_time": 1781.203,
"text": " is just a manifestation of pursuing those paths of least resistance or mathematically those paths of stationary action, then there is no counterfactual to select from and there is no notion of planning in the sense of I am going to do this as opposed to that."
},
{
"end_time": 1819.155,
"index": 73,
"start_time": 1798.524,
"text": " that doesn't mean to say biological organisms don't have the capacity to plan multiple futures and I'll just give you one real simple example of our brains and I would imagine even the brains of reptiles to a certain extent being able to plan multiple"
},
{
"end_time": 1835.845,
"index": 74,
"start_time": 1819.565,
"text": " actions in the future and that's in eye movements. So if I just take one of the simplest problems in terms of planning or perhaps not the simplest but one of the problems of planning that have to be resolved within"
},
{
"end_time": 1864.753,
"index": 75,
"start_time": 1836.118,
"text": " several hundred milliseconds, given that we have to choose where to look every 250 milliseconds in order to gather the right kind of information to construct the scene in which we are situated. Coming back to your preoccupation, Chris, how on earth do we then explain these sparse, saccadic samples within some frame of reference, some space-time sort of frame? How do we sense-make with reference"
},
{
"end_time": 1876.493,
"index": 76,
"start_time": 1866.971,
"text": " The point I'm making here is that the anatomy of many brains, many phenotypes"
},
{
"end_time": 1904.326,
"index": 77,
"start_time": 1876.852,
"text": " does have the capacity to represent the consequences of looking everywhere at the same time at once before they have looked. So these are sometimes referred to as salience maps and we know that there are good candidates for these salience maps, for example the deep layers of the superior colliculus and one could argue possibly some anatomy of the pulvina. So if you look at the neuronal encoding of the salience of where"
},
{
"end_time": 1934.309,
"index": 78,
"start_time": 1904.701,
"text": " the salience of the next place to look, what's going to grab my attention and make that manifest overtly with an eye movement to go look over there, then you can interpret, I think, the anatomy, the ocular motor system, and especially that which underlies the control of the saccadic eye movements as entertaining a whole range, possibly millions of potential"
},
{
"end_time": 1959.616,
"index": 79,
"start_time": 1934.735,
"text": " plans of action and the consequences. And then the brain literally selects the one that's most likely, which you can simulate by doing bumper tractors on these sort of salience maps, representations of the salience or the epistemic affordance or the expected information gain if I looked over here, looked over here, looked over here, looked over here, you select that, your eyes jump over there, and off you go again."
},
{
"end_time": 1988.899,
"index": 80,
"start_time": 1960.026,
"text": " So I think it's a great question and in a sense speaks, I think, to a certain extent to what Mike was alluding to before, that there are certain setups, there are certain sort of neuronal electrochemical infrastructures out there that can be read as just pursuing paths of least action in a reflexive way. And there are others that may have this richer structure, this deeper structure that could be support the, you know,"
},
{
"end_time": 2018.49,
"index": 81,
"start_time": 1989.172,
"text": " Mike, is that what you had in mind in terms of breaking this sort of panpsychism trap, different kinds of"
},
{
"end_time": 2035.06,
"index": 82,
"start_time": 2019.121,
"text": " I agree with Chris's gut feeling on this. I find it very difficult to imagine cognition without some sort of awareness or simple consciousness."
},
{
"end_time": 2060.418,
"index": 83,
"start_time": 2035.452,
"text": " What strikes me as different than real about the hard problem is that unlike all the quote unquote easy problems, and maybe unlike the meta hard problem, we kind of know the format or the shape of answers to those questions, right? They're either their numbers or their equations or their lists of capacities. You know, we know what a proper answer would look like. And yet, I find it very hard to imagine if somebody claimed to have a"
},
{
"end_time": 2090.981,
"index": 84,
"start_time": 2060.981,
"text": " a good theory of consciousness and I were to ask them, okay, well, what is the prediction of your theory in this particular case? I don't know what the format of the answer looks like, because numbers and the typical things we get don't do the trick, they, you know, they're sort of third person descriptions. And so, so is the answer, does the answer come in the form of a poem? Is it art? Is it, you know, something you have to literally plug yourself into to then have that experience? Like, what's the, I find it different, because I don't know what if we had a true, a correct theory, I don't know what that theory would output."
},
{
"end_time": 2120.811,
"index": 85,
"start_time": 2091.323,
"text": " Well, that's a better answer to the hard problem then. I'm sure the answer is quantum physics. That's what it's going to look like. If we could understand it, the answer will be in quantum physics. Chris, do you agree? Well, I'd like to back up a little bit."
},
{
"end_time": 2150.896,
"index": 86,
"start_time": 2121.527,
"text": " And in a sense, play off what Carl was just saying about alternatives and the role of alternatives in the hard problem. I mean, Chalmers drives his arguments about the hard problem with the alternative of unconsciousness. And this alternative makes great sense against the background of a"
},
{
"end_time": 2181.118,
"index": 87,
"start_time": 2151.152,
"text": " particular set of assumptions about physics, which is the assumption that nothing is going on cognitively in the physical world, that there's no reason to talk about experience or awareness or anything like that when we're talking about"
},
{
"end_time": 2207.517,
"index": 88,
"start_time": 2182.073,
"text": " of rocks or billiard balls or planets or electrons or anything else. And in that case, there is this cut that is not really a cut in scale, spatial scale or temporal scale,"
},
{
"end_time": 2237.5,
"index": 89,
"start_time": 2208.148,
"text": " but is, in a sense, a cut in some measure of complexity that suggests that below some cut point, which may be a bright line and maybe a fuzzy area, who knows, there's the possibility or below some cut point, there's no awareness at all, none, zero. And above that cut point, there's some possibility of awareness."
},
{
"end_time": 2266.596,
"index": 90,
"start_time": 2238.558,
"text": " And in that case, one's forced to say, uh, what, what could be added by complexity other than complexity to produce awareness. And it's that intuition, I believe that drives the heart problem. Do you mind explaining that a bit more of what can be added by complexity? That's not complexity that sentence. Well, um,"
},
{
"end_time": 2295.094,
"index": 91,
"start_time": 2268.575,
"text": " Yeah, and one way to put it is complexity, as complexity increases, some magic has to happen. And you suddenly get awareness. And it's not clear quite how much complexity one has to add. But as soon as one reaches the right amount, then out of the blue, you've got awareness."
},
{
"end_time": 2324.821,
"index": 92,
"start_time": 2296.254,
"text": " Whereas without crossing that threshold, you had no awareness whatsoever. So there's a zero point and the zero points somewhere on the scale of complexity, and it's nowhere near the bottom. So I think one can view a lot of research on consciousness as a way to escape that argument."
},
{
"end_time": 2355.162,
"index": 93,
"start_time": 2325.282,
"text": " Consider integrated information theory. In integrated information theory, the criterion for consciousness or awareness is actually very simple. You need to have an internal feedback loop. And if you've got an internal feedback loop, you're at least a candidate for being aware, as long as you're not embedded in something bigger that has bigger internal feedback loops."
},
{
"end_time": 2385.23,
"index": 94,
"start_time": 2358.336,
"text": " And if one thinks about the work of Peter Strawson, for example, from a philosophical perspective, his arguments toward panpsychism are all arguments of the form, there's no way to draw the line in the scale of complexity."
},
{
"end_time": 2414.019,
"index": 95,
"start_time": 2386.681,
"text": " and get a place where magic basically plausibly happens. So I think Chalmers did a real service in posing the problem in that way because it forced us to think about this idea of magic, or you could call it emergence if you wanted to, of something completely new that was awareness."
},
{
"end_time": 2447.415,
"index": 96,
"start_time": 2418.097,
"text": " Now, I do agree with Karl that quantum theory can help us dispel the problem by being, in effect, a theory that's about awareness. But we can get to that farther down the road. Do you have any reason to believe that we're embedded in something larger? You mentioned in IIT, there's the feedback loop that is necessary for consciousness. And you said,"
},
{
"end_time": 2473.985,
"index": 97,
"start_time": 2447.858,
"text": " that one is conscious as long as one is not embedded in something else that's of which that is conscious, like there's a larger feedback loop there. Do you see there being some larger feedback loop? Some people say the universe itself is conscious or that societies can act as some larger level consciousness, like with each of us acting as neurons in some sense. Well, I would certainly wouldn't want to rule that view out a priori."
},
{
"end_time": 2490.418,
"index": 98,
"start_time": 2475.367,
"text": " I mean, it's obvious that we're embedded in much bigger systems. And I think we understand essentially nothing about the cognitive capacities of those systems."
},
{
"end_time": 2514.753,
"index": 99,
"start_time": 2491.954,
"text": " Yeah, I think that's a really important question and I think a lot about the perspective of, let's say, if you were a neural network or something like this being trained and so on, what would the perspective of a sub-component of that be? Let's say if you had the capacity as a neuron to sort of look around and ask yourself, do I live in a cold mechanical universe that doesn't care what happens or"
},
{
"end_time": 2534.753,
"index": 100,
"start_time": 2514.753,
"text": " Is there some sort of agency in my environment? When I'm as a piece of that network, am I learning from my environment or am I being trained? Because when you're being trained, the real question is how many agents are there in that interaction? Is it just you learning as you will, you're the boss and you're sort of learning whatever from your environment?"
},
{
"end_time": 2545.845,
"index": 101,
"start_time": 2534.753,
"text": " Or are you actually being trained because the environment is an agent with an agenda that is training you for some particular purpose? And so this question of how would you know? So you look around and you say to yourself,"
},
{
"end_time": 2574.002,
"index": 102,
"start_time": 2546.459,
"text": " If you're a piece of this network and it's being trained for some image recognition task, you would be, I think, wrong to come to the conclusion that you live in a mechanical universe. You should come to the conclusion that it's clearly rewarding you and punishing you for specific things. It's not neutral with respect to what you do. And for some reason, it really likes it when you find pictures of dog eyes or something like this. And you have no idea that what it does is recognize dog faces out of all the other"
},
{
"end_time": 2600.623,
"index": 103,
"start_time": 2574.002,
"text": " all the other inputs. So no doubt there's some kind of Gordelian limit to figure, to be able to actually understand what the larger system is doing. But I wonder if we can even have some evidence that just for the fact that yes, there is a greater, you know, sort of a gentile dynamic going on in which I'm caught up, even if you're not able to, you know, sort of comprehend what that's going to be. I don't know. I don't know what that's going to look like. But I think it's pretty important. And I think, as Chris said,"
},
{
"end_time": 2607.944,
"index": 104,
"start_time": 2600.623,
"text": " We really don't have a very good science at all of trying to predict or control."
},
{
"end_time": 2635.111,
"index": 105,
"start_time": 2608.404,
"text": " the cognitive capacities of systems made up of parts. So we can know many things about parts, cells, robotics, whatever, and we routinely make these larger systems and then get surprised about what it does or doesn't do or what the goals are going to be, what goals is it capable of pursuing, what preferences does it have. This is probably an existential level job for society is to get a good science of that going."
},
{
"end_time": 2665.043,
"index": 106,
"start_time": 2636.852,
"text": " You have done that though, haven't you? Well, we've started anyway. I would go further than that. I think your work would represent a substantive and established formal framework to address exactly those issues. And you can see embryonic versions arising in many different fields. I just came back from a meeting"
},
{
"end_time": 2678.695,
"index": 107,
"start_time": 2665.538,
"text": " of economists and financiers. They are trying to make the move from behavioral economics to cognitive economics. And they were exactly addressing these kinds of issues. You are understanding the"
},
{
"end_time": 2704.053,
"index": 108,
"start_time": 2679.07,
"text": " the notion of distributed cognition in a market. So, you know, ultimately, they'll come knocking at your door. I can see that. I can see exactly the sequence of people, you know, searching around for people to talk about, illuminating that realization that the, I mean, it comes back to what we were talking about before, where you started, and I was picking up on the"
},
{
"end_time": 2734.65,
"index": 109,
"start_time": 2704.701,
"text": " on the link between different scales of self-organization, where every scale has, you know, complies with the same principles. But how does that contextualize? And I think one really interesting example of that is to think about an individual in an ensemble or a society. And just to wrestle the argument back to pick up on something that Chris was saying about, you know, if there is a bright line between the kind of mechanics that would"
},
{
"end_time": 2762.739,
"index": 110,
"start_time": 2734.991,
"text": " qualify as conscious in the sense of being self-aware. And I'm distinguishing that from the kind of consciousness that would just entail qualitative experience and a loss of phenomenal transparency. So I'm talking about now self-awareness as one of the bright lines that may be very blurry and vague, but certainly one which is induced by the hard problem"
},
{
"end_time": 2791.391,
"index": 111,
"start_time": 2763.985,
"text": " Just asking a simple question, what kind of explanation for my world, whether I'm a single neuron or a single person, would enable me or justify the notion that I am a self? And the obvious answer is when I have to disambiguate between the consequences of my action and your action, if you are very similar to me. So if I"
},
{
"end_time": 2815.009,
"index": 112,
"start_time": 2792.142,
"text": " managed to survive on an alien planet where there are no structures like me that could cause the same sensed consequences, then there would be no problem inferring, did I cause that or did you cause that? So just having the existence of a population of conspecifics in some sort of formal structural sense,"
},
{
"end_time": 2843.916,
"index": 113,
"start_time": 2815.503,
"text": " suddenly induces the inference problem, did you cause that, do I cause that and of course that naturally calls for generative models or internal dynamics representations on the inside that entertain the hypothesis or the notion that that's me as opposed to not me. So you get for free in and only in this context where you've put lots of these particles together, the license or the motivation to have"
},
{
"end_time": 2863.166,
"index": 114,
"start_time": 2844.275,
"text": " a hypothesis, a representation of selfhood that from a statistician's point of view is exactly the justification for the increased complexity that Chris was talking about. So I was listening to that really interesting sort of"
},
{
"end_time": 2890.418,
"index": 115,
"start_time": 2863.985,
"text": " not to IIT and all it brings to the table in terms of this sort of, you know, commitment to some threshold crossing or traversing a cut with increasing complexity. But why? I mean, one simple answer is that, you know, if you ask a statistician, what is complexity? They'll tell you, well, it's basically the degrees of freedom you're using up to explain some sensory data."
},
{
"end_time": 2920.828,
"index": 116,
"start_time": 2891.084,
"text": " You could articulate it as a KL or relative entry between a posterior and a prior if you're a Bayesian, but in essence it's just the degrees of freedom you're using in order to accurately explain these data. So mathematically the evidence is equal to the accuracy minus the complexity. So why would you need, why is that useful? Well an increase in complexity is only licensed by an increase in the accuracy. So the simple argument would go"
},
{
"end_time": 2948.49,
"index": 117,
"start_time": 2921.203,
"text": " If I am obliged to model and predict and explain a world that is constituted by other things like me, then the accuracy of my predictions will be greatly enhanced if I have a representation of me as distinct from you. If that entails an increase in the degrees of freedom, namely adding in this extra kind of hypothesis and everything that it entails, then"
},
{
"end_time": 2969.991,
"index": 118,
"start_time": 2948.66,
"text": " I'm going to have a greater complexity, but it's a complexity that is more than paid for by the increase in the accuracy. Just coming back to my original point, I think a lot of the work that you've done at the cellular level"
},
{
"end_time": 2994.718,
"index": 119,
"start_time": 2970.333,
"text": " would be very gracefully translated to the societal level and to things like economics and to eco niche construction in ethology. And I know Chris is probably wanting to say this, but language as one way of facilitating that notion, that ability to efficiently with minimum complexity in this instance, do this"
},
{
"end_time": 3004.292,
"index": 120,
"start_time": 2997.568,
"text": " Hear that sound?"
},
{
"end_time": 3031.391,
"index": 121,
"start_time": 3005.247,
"text": " That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the internet's best converting checkout, making it 36% more effective than other leading platforms."
},
{
"end_time": 3051.203,
"index": 122,
"start_time": 3031.391,
"text": " There's also something called Shopify Magic, your AI powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level."
},
{
"end_time": 3080.811,
"index": 123,
"start_time": 3051.203,
"text": " Join the ranks of businesses in 175 countries that have made Shopify the backbone of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklynin. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com"
},
{
"end_time": 3108.439,
"index": 124,
"start_time": 3080.811,
"text": " Well, again, what I'd actually like to do is back up a little bit and"
},
{
"end_time": 3137.773,
"index": 125,
"start_time": 3108.78,
"text": " put some of what you just said into very simple active inference language. If I see something happening on my Markov blanket, on my interface with the world, then I always have the question, did I do that? Or did the world do that? Where the world means everything outside me. And"
},
{
"end_time": 3167.346,
"index": 126,
"start_time": 3139.053,
"text": " In a sense, the answer is always the world did it. So the question becomes, did the world do that in response to something I did to it? Or did it just do it? Not in consequence of any of my actions. And so one gets immediately to this kind of babbling scenario that we've talked about many times, in which"
},
{
"end_time": 3196.271,
"index": 127,
"start_time": 3168.012,
"text": " An infant or a robot or some system is trying to figure out by measuring correlations, whether the world's inputs to it have anything to do with its inputs to the world. And just asking that question requires enormous representative capacity."
},
{
"end_time": 3226.288,
"index": 128,
"start_time": 3197.381,
"text": " Because one has to represent one's actions and represent them pretty well in time. And one has to have a good memory to represent enough actions to get any kind of statistical support for drawing an inference about correlation. And that memory has to be represented as a memory"
},
{
"end_time": 3256.988,
"index": 129,
"start_time": 3228.524,
"text": " not just part of one's current input. So I think this question that you posed is really the key question faced by any agent at all that's trying to get a model off the ground, which in a sense gets back to the question that Mike asked early, early on about how does this all start?"
},
{
"end_time": 3284.343,
"index": 130,
"start_time": 3257.722,
"text": " Maybe it starts with babbling in very simple systems. You know, I was thinking recently, this whole issue of how many agents are there and where is the border between the agent and the world and how do you self model that border is a fascinating topic. And there's an amazing developmental model for this, which is that, you know, we often talk about one embryo and the embryo does this and the embryo does that. But actually,"
},
{
"end_time": 3309.667,
"index": 131,
"start_time": 3284.599,
"text": " What happens at the beginning, let's say, for example, an amniote embryos is that there's a flat blasted disc, which has just a few cell layers thick. So it's kind of think of it like a frisbee and it just has a few cell layers. And normally what has to happen is that one point in this disc breaks symmetry and then organizes the primary axis of the first embryo and basically tells all the other cells, don't do it because I'm doing it. And that's how you end up with one embryo."
},
{
"end_time": 3324.735,
"index": 132,
"start_time": 3310.026,
"text": " Now that process is very easily perturbed and many people have done it. I used to do it in my graduate work. And what you can do is if you perturb that process, that initial blasted disk, that undifferentiated sort of pool of cells, which are these sort of proto"
},
{
"end_time": 3343.916,
"index": 133,
"start_time": 3325.282,
"text": " low level proto-agents, that pool can break up into not one embryo but actually multiple. I did this in bird embryos and you can have twins and chicken and duck and things like this. Humans have exactly the same structure. You can get them head to head, you can get them side by side, you can get all sorts of geometries."
},
{
"end_time": 3368.37,
"index": 134,
"start_time": 3343.916,
"text": " you can get triplets, you can get multiple individuals emerging by different partitions of this really kind of medium, this particular medium where you have a bunch of cells and you don't know ahead of time how many individuals at the level of how many larger individuals, so embryos, are going to arise from this medium because the dynamics by which in this, you know, local activation and long-range inhibition and things like that,"
},
{
"end_time": 3391.527,
"index": 135,
"start_time": 3368.37,
"text": " the dynamics that break up that undifferentiated ocean of potential selves into one, two or three or more selves is actually, it's very dynamic, it can go different ways. And then you get interesting things like this. So for example, you might know that human conjoined twins that are sort of stuck together side to side, one of the twins often has left right asymmetry defects."
},
{
"end_time": 3415.145,
"index": 136,
"start_time": 3391.886,
"text": " It's because when you have two twins side by side, the cells in the middle, both twins can't quite agree on who they belong to. Are they the right side of this twin or are they the left side of that twin? Both twins think they belong to them, but in fact, they're overlapping. They're the same cells. One side will have correct left and right. The other side will have two rights, for example. This ends up giving one of the twins laterality defects with respect to heart and gut pattern."
},
{
"end_time": 3436.237,
"index": 137,
"start_time": 3415.435,
"text": " And so their models, each twins as the collective of cells tries to compute things like where things are and what's left and what's right and so on, their models can disagree with each other. They can draw the boundary between self and world in different ways. And you can have these sort of disputes over certain areas as to who they actually belong to."
},
{
"end_time": 3463.814,
"index": 138,
"start_time": 3436.886,
"text": " And so I'm just incredibly interested in this process of individualization, so to speak, out of this like ocean of potentiality, these cells, you know, 50,000 cells and some number of individuals at the embryo level will be formed. And each of them will have specific goals in morphospace. Each of them will try to achieve very specific morphologies. And you don't know ahead of time how many there were going to be, you know?"
},
{
"end_time": 3481.425,
"index": 139,
"start_time": 3465.503,
"text": " Carl, what occurs to you when you hear that? Ambiguity, again, just thinking about the imperatives for self-organization."
},
{
"end_time": 3507.432,
"index": 140,
"start_time": 3482.056,
"text": " When I hear that was a fascinating story, and don't let me forget, we ought to explain to the viewers what babbling means just in this context as well, because that's quite illuminating. Perhaps we can, I'm not sure that the cells do motor babbling, but they certainly resolve the same kind of problem. Chris, correct me if I'm wrong, but by babbling, all we mean is that when you're first born into any universe,"
},
{
"end_time": 3530.265,
"index": 141,
"start_time": 3507.875,
"text": " You've got to work out or test the hypothesis that I caused that or the world caused that. So you don't know, you don't have a self-model, whether this is sort of a declarative model or completely sub-personal, just hardwired into the synaptic efficacy and connectivity of your brain."
},
{
"end_time": 3558.763,
"index": 142,
"start_time": 3530.674,
"text": " So the first thing you have to do is just to work out what you can control and what you can't control and the idea is that you engage in what I would call epistemic or respond to epistemic affordances or epistemic plans that reveal knowledge, they resolve uncertainty and in this instance it's the uncertainty about whether I was the cause of this rattle rattling"
},
{
"end_time": 3588.251,
"index": 143,
"start_time": 3559.019,
"text": " So if you imagine motor babbling as manifest in a little baby rattling its rattle generating both the sensations from the muscles and the skin but also the visual and the auditory sensations all co-occur providing definitive evidence that there's something special about this process and this event that"
},
{
"end_time": 3615.452,
"index": 144,
"start_time": 3588.456,
"text": " that provides the basis for the hypothesis there's a unitary cause there's a unitary cause which is me shaking the rattle but of course it may take several months if not years to actually get to actually realize that cause is me so one can imagine sort of you know robots learning about the the manipulander that they can articulate or the the way in which they can move around and you know"
},
{
"end_time": 3646.22,
"index": 145,
"start_time": 3616.544,
"text": " But I think the idea, bringing it back to sort of where does selfhood come from, it would rest upon testing the hypothesis, which has to be physically represented with a deep, more complex generative model of sense making, that in fact, it's me that's actually caused this single cause of all these proprioceptive motor sensations, visual sensations, auditory sensations,"
},
{
"end_time": 3671.032,
"index": 146,
"start_time": 3646.544,
"text": " And that will be especially prescient when you're starting to realize that some of these causes, which you thought were you, of the sort associated with nurturing and suckling, were actually due to mum. And we come back to this argument to actually have a good hypothesis which explains why I am not in charge of mum, because she is now not always responding to me when I cry."
},
{
"end_time": 3696.237,
"index": 147,
"start_time": 3671.357,
"text": " To explain that I have to now develop a hypothesis. Yes, it was me causing all of this but sometimes there's something else out there that's not me but very much like me and that's mum and then you can see how there would be a pressure in terms of finding the best explanations for your sensations to have that. So if you take that notion now think of the same problem from the point of view of a cell"
},
{
"end_time": 3723.609,
"index": 148,
"start_time": 3696.578,
"text": " a bunch of cells that have to work together or will ultimately form an embryo via this process of symmetry breaking. You have to ask what are the underlying imperatives, how could it be any other way and of course it could be lots of ways but the ways in which it goes wrong which would be another way of saying"
},
{
"end_time": 3753.575,
"index": 149,
"start_time": 3723.814,
"text": " those rare occasions or the ways that it doesn't happen in consequence of their rarity involve this ambiguity again, literally in the context of some cells not knowing whether they're belonging to one twin or another twin from the cell's perspective. So I was just thinking about the nature of ambiguity and of course it is exactly the same. One can account"
},
{
"end_time": 3781.237,
"index": 150,
"start_time": 3754.462,
"text": " for the simple observation that self-organization does not tolerate ambiguity, it does not tolerate uncertainty. It's only manifest in the context of accurate and well-evidenced definitive exchanges. So come back to Chris's notion of sense-making projecting onto my Markov blanket or my"
},
{
"end_time": 3808.985,
"index": 151,
"start_time": 3783.626,
"text": " that I need to resolve all the uncertainty, as much uncertainty as I can. But put that another way, in a slightly more deflationary way, stuff which we see the way the universe seems to work can be described as realizing processes that minimize this kind of uncertainty and ambiguity."
},
{
"end_time": 3827.961,
"index": 152,
"start_time": 3809.326,
"text": " and simply maximise the neutral predictability of what's being projected or impressed upon my surface or my holographic screen or my Markov blanket. So, you know, that's what was going through my mind. I thought it was a beautiful example of, you know, when it goes wrong."
},
{
"end_time": 3857.824,
"index": 153,
"start_time": 3828.251,
"text": " There's uncertainty and ambiguity in the game, and that tells you something quite fundamental about when it goes right. And when it goes right, it's just basically a statement of what exists and what perseveres over time. Chris, did you have anything to add to that? Well, I would be interested to see how actually Mike responds to"
},
{
"end_time": 3877.449,
"index": 154,
"start_time": 3858.541,
"text": " that discourse with respect to the example of this embryonic sheet where the cells are each trying to figure out what they're supposed to do and hear that sound."
},
{
"end_time": 3904.462,
"index": 155,
"start_time": 3878.353,
"text": " That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the internet's best converting checkout, making it 36% more effective than other leading platforms."
},
{
"end_time": 3930.606,
"index": 156,
"start_time": 3904.462,
"text": " There's also something called Shopify Magic, your AI-powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level. Join the ranks of businesses in 175 countries that have made Shopify the backbone."
},
{
"end_time": 3956.357,
"index": 157,
"start_time": 3930.606,
"text": " of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklyn. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com slash theories, all lowercase."
},
{
"end_time": 3985.213,
"index": 158,
"start_time": 3956.357,
"text": " Go to Shopify.com slash theories now to grow your business no matter what stage you're in Shopify.com slash theories. Razor blades are like diving boards. The longer the board, the more the wobble, the more the wobble, the more nicks, cuts, scrapes. A bad shave isn't a blade problem. It's an extension problem. Henson is a family owned aerospace parts manufacturer that's made parts for the International Space Station and the Mars Rover."
},
{
"end_time": 4013.695,
"index": 159,
"start_time": 3985.213,
"text": " Now they're bringing that precision engineering to your shaving experience. By using aerospace-grade CNC machines, Henson makes razors that extend less than the thickness of a human hair. The razor also has built-in channels that evacuates hair and cream, which make clogging virtually impossible. Henson Shaving wants to produce the best razors, not the best razor business, so that means no plastics, no subscriptions, no proprietary blades, and no planned obsolescence."
},
{
"end_time": 4030.077,
"index": 160,
"start_time": 4013.695,
"text": " It's also extremely affordable. The Henson razor works with the standard dual edge blades that give you that old school shave with the benefits of this new school tech. It's time to say no to subscriptions and yes to a razor that'll last you a lifetime. Visit hensonshaving.com slash everything."
},
{
"end_time": 4059.155,
"index": 161,
"start_time": 4030.077,
"text": " If you use that code, you'll get two years worth of blades for free. Just make sure to add them to the cart. Plus 100 free blades when you head to H E N S O N S H A V I N G dot com slash everything and use the code everything. There's there's still. They still certainly believe that they should be reproducing, so they they definitely do that and they signal to each other."
},
{
"end_time": 4073.507,
"index": 162,
"start_time": 4059.872,
"text": " But somehow they're the collective effect of all of that signaling to each other is to organize different roles for the for each for themselves. So how does the symmetry breaking occur?"
},
{
"end_time": 4100.35,
"index": 163,
"start_time": 4074.548,
"text": " Yeah, so that's really interesting. So as I was hearing Carl talk about this, I realized a couple of things that, first of all, as we were saying before, feedback loops are absolutely central to this process because the easiest way to prevent any of this from happening and to end up with a kind of a mono, you know, sort of a very featureless sheet where there are no embryos is to block the positive feedback loop."
},
{
"end_time": 4128.2,
"index": 164,
"start_time": 4100.35,
"text": " that short range inhibition, long range activation that says to one cell, I'm going to now be the organizer. I'm going to make this access. Everybody else don't do it. Those are both feedback loops. And so if you break those feedback loops, you get nothing. So the feedback loops are right at the beginning of this process. And the other thing that I thought was really interesting that Carl just said is about the cells babbling. So what I think it's a really good name for what we see when you take a cell and you put it out into a dish,"
},
{
"end_time": 4141.954,
"index": 165,
"start_time": 4128.66,
"text": " What you'll see in any new environment, you will see two things that I think are probably babbling in different spaces. One is that in physical space, you will see that it's incredibly active. It's constantly putting out and pulling back these"
},
{
"end_time": 4162.961,
"index": 166,
"start_time": 4142.5,
"text": " The other babbling takes place in transcriptional space."
},
{
"end_time": 4189.206,
"index": 167,
"start_time": 4163.183,
"text": " So gene expression is also never just sort of still, and this is, you know, these are the genes that are expressed and I'm just going to sit here and do that. All the genes are constantly going up and down within specific ranges as it sort of wiggles in this transcriptional space. And I think both of them, probably physiologically, metabolically, I bet this is going on in all these other spaces. And I think babbling is an excellent framework for understanding what it's actually doing. It's actually taking these little actions."
},
{
"end_time": 4210.776,
"index": 168,
"start_time": 4189.394,
"text": " looking for evidence of specific things that it can then make use of to start to draw boundaries. And as Chris was saying, I think what happens is that in this blastoderm, what will happen is that once that individuation starts where there's a specific cell that starts to go down this road of"
},
{
"end_time": 4236.886,
"index": 169,
"start_time": 4211.101,
"text": " being the organizer of this new larger collective, it immediately begins to distort the action space for all the nearby cells. It starts putting out all kinds of signals, you know, reward signals and physical forces and all these other things that now are going to bend the option for all the other cells. And the best example of that we have are these xenobots where these things are just made of embryonic skin cells."
},
{
"end_time": 4253.37,
"index": 170,
"start_time": 4237.108,
"text": " And if you look at a standard old embryo, you get this idea that, well, what do skin cells want to do? They want to be a passive two dimensional layer on the outside of the animal. They do nothing except sit there, keep out the bacteria, you know, nice and a boring two dimensional life. But what you find out is that that's only true because the other cells are basically bullying them into it."
},
{
"end_time": 4278.148,
"index": 171,
"start_time": 4253.37,
"text": " left to their own devices in the absence of these instructive interactions with the rest of the embryo. What the skin cells actually want to do is get together into a three-dimensional kind of ball-like architecture. They are self-motile. They'll run around and move and have various behaviors, including make copies of themselves if provided with materials. And so that is completely obscured by standard development, where what you're seeing actually is"
},
{
"end_time": 4308.08,
"index": 172,
"start_time": 4278.353,
"text": " is cells in a space that was really deformed by all their neighbors, right? And that's, you know, that kind of process that starts to make those distinctions where the embryo can tell what parts are inside and what parts are supposedly outside of itself. And that gets reinforced by all these early activities. Carl or Chris, do you want to jump in on that? I think we've been describing psychos that says basically that it's curiosity all the way down."
},
{
"end_time": 4330.981,
"index": 173,
"start_time": 4308.558,
"text": " Every system is trying to figure out what's going on. I was struck with that with this image of the little cells putting out fingers, this is expiration, sort of true blue expiration and mathematically it's simply as Chris says it's what artificial intelligence"
},
{
"end_time": 4346.937,
"index": 174,
"start_time": 4331.305,
"text": " research aspires to, which is artificial curiosity, going out there, getting the right kind of data that's going to maximally resolve your uncertainty and paint the way forward. And of course, it's intimately related to planning."
},
{
"end_time": 4370.52,
"index": 175,
"start_time": 4347.705,
"text": " In the sense you have to realise that palpation, whether it's a little cell palpating with its filial or podial, whether it's me palpating my visual world by moving my eyes around, or whether it's the little baby babbling and palpating its cot, this palpation has to be planned."
},
{
"end_time": 4397.363,
"index": 176,
"start_time": 4371.305,
"text": " It could be planned in the sense of the gazelle, sorry, because it's not the gazelle, the cheetah chasing the gazelle. So it doesn't have to be very sophisticated cognitive conscious planning, but certainly has to has to be planned. It has to have this sort of curious behavior. It strikes me that that is such a fundamental aspect that I think would qualify behaviors"
},
{
"end_time": 4421.476,
"index": 177,
"start_time": 4397.705,
"text": " that have that aspect as cognitive in some sense. And it's so fundamental because of course, it's just an expression of dynamics that apply to action upon the world that underwrite everything that we do as certainly as scientists and one could imagine as human beings, but certainly as scientists, you know,"
},
{
"end_time": 4448.677,
"index": 178,
"start_time": 4421.783,
"text": " The maths of curiosity was actually worked out by Dennis Lindley in the 1950s in terms of expected information gain and then reintroduced by people like David McKay in the context of active learning. So bringing this active notion that you can learn by actively acquiring the right kind of data that optimizes that kind of learning. And it became known as"
},
{
"end_time": 4474.121,
"index": 179,
"start_time": 4448.985,
"text": " the principles of optimum Bayesian experimental design. So it's exactly the same kind of curiosity that we as scientists use all the time whenever we design an experiment. It's basically configuring, actively configuring some process to generate something that can be sensed or measured that maximally resolves our uncertainty or forwards the greatest amount of information."
},
{
"end_time": 4499.275,
"index": 180,
"start_time": 4474.497,
"text": " So the three of us as scientists have become experts at this kind of formal curiosity simply just by acquiring knowledge about, in the particular paradigms or setups and fields that we find ourselves in, the right way to do experiments. But you could argue that that's life. In a sense, that is the infant bubbling."
},
{
"end_time": 4524.36,
"index": 181,
"start_time": 4499.667,
"text": " It is that kind of sense making and actively getting the right kind of data to work out your place in the world, to work out what you should do next is one of the most existentially important imperatives. The mechanics that Mike was talking about in terms of the"
},
{
"end_time": 4554.616,
"index": 182,
"start_time": 4524.855,
"text": " the little skin cells left to their own devices, having a party and forming balls and wandering around. One could apply that kind of mechanics to people, couldn't you? Or even cultures and countries. We're all trying to find our place. We're all trying to work out how to respond to those constraints. People around us at many different scales put on our behavior and try and infer, well, how am I meant to behave in this situation?"
},
{
"end_time": 4583.131,
"index": 183,
"start_time": 4555.111,
"text": " You hear that sound? That's the sweet sound of success with Shopify. Shopify is the all-encompassing commerce platform that's with you from the first flicker of an idea to the moment you realize you're running a global enterprise. Whether it's handcrafted jewelry or high-tech gadgets, Shopify supports you at every point of sale, both online and in person. They streamline the process with the Internet's best converting checkout, making it 36% more effective than other leading platforms."
},
{
"end_time": 4609.241,
"index": 184,
"start_time": 4583.131,
"text": " There's also something called Shopify Magic, your AI-powered assistant that's like an all-star team member working tirelessly behind the scenes. What I find fascinating about Shopify is how it scales with your ambition. No matter how big you want to grow, Shopify gives you everything you need to take control and take your business to the next level. Join the ranks of businesses in 175 countries that have made Shopify the backbone."
},
{
"end_time": 4635.009,
"index": 185,
"start_time": 4609.241,
"text": " of their commerce. Shopify, by the way, powers 10% of all e-commerce in the United States, including huge names like Allbirds, Rothy's, and Brooklynin. If you ever need help, their award-winning support is like having a mentor that's just a click away. Now, are you ready to start your own success story? Sign up for a $1 per month trial period at Shopify.com slash theories, all lowercase."
},
{
"end_time": 4653.814,
"index": 186,
"start_time": 4635.009,
"text": " Go to shopify.com slash theories now to grow your business no matter what stage you're in shopify.com slash theories. So even if we all start off with the same genetic code and the same sort of model of your how people behave."
},
{
"end_time": 4680.674,
"index": 187,
"start_time": 4654.48,
"text": " The context in which we find ourselves now needs to be inferred in order to know how to behave in this context, either as a child or as a mother, as a politician or as an aid worker or first responder, whatever you want, you have to make that inference. And of course, that inference, inferring that context requires the curious behavior that Chris had picked up on in Mike's example."
},
{
"end_time": 4711.596,
"index": 188,
"start_time": 4681.766,
"text": " Wow, you know what I'd love to hear is each of you talk about what is the simplest, most basic thing. So, you know, hopefully going down to like physics prior to life, but wherever you like, what is the most basic thing that is able to do that? Because that's what I find in talking about this stuff to other communities. The thing that people are the most resistant to is this idea that these kind of dynamics, this kind of exploration, curiosity, prediction, all of these things that we're talking about,"
},
{
"end_time": 4725.64,
"index": 189,
"start_time": 4711.886,
"text": " that people think it's brains first, and then you can do all these great things. I would love to hear what you think is the simplest physical system that can do that kind of thing so that we can talk about how far down it actually goes."
},
{
"end_time": 4756.323,
"index": 190,
"start_time": 4729.377,
"text": " While Chris is thinking about that with a profound answer, I'll give you a not a trivial answer, but a formal answer. I think you'd have to identify the depth of planning that underwrote the dynamics of the behavior of the system in question. And that would be basically"
},
{
"end_time": 4785.589,
"index": 191,
"start_time": 4756.732,
"text": " If you were putting this in a sort of path integral formulation it would be the sort of the amount of time over which you are integrating your paths and you're evaluating the best way forward. Once you move from very short time scales then the depth of that planning I think eludes physical realizations that could be written down in terms of"
},
{
"end_time": 4806.715,
"index": 192,
"start_time": 4786.084,
"text": " representations in terms of concentrations and depolarizations and the like and one has to move to a quantum discrete time representation. So you know I guess I'm basically making the difference between the sort of the"
},
{
"end_time": 4832.022,
"index": 193,
"start_time": 4807.073,
"text": " the kind of dynamics you'd find in a in chemotaxis or a thermostat that could be written down as differential equations. In fact Mike, you're the kind of differential equations we used in one of those early pattern formation papers of knowing your place approaches where everything could be articulated in terms of differential equations so that there is a kind of"
},
{
"end_time": 4857.005,
"index": 194,
"start_time": 4832.483,
"text": " But it's not of the curious kind that we talked about. It's just following the most likely paths of least action. So I would say that they're not cognitive in the sense that they have planning. So I put chemotaxis and thermostats and possibly most viruses probably at a precognitive or protocognitive level. But as soon as you get to"
},
{
"end_time": 4882.176,
"index": 195,
"start_time": 4857.261,
"text": " generative models that can roll out deep into the future and I repeat just when you try to implement this in silico you know you're using computer simulations you really have to move to a discrete time representation which I think is a non-trivial thing I mean just for example remember the sort of the eye movement example you know we sense make by getting little snapshots of the world every 250 milliseconds"
},
{
"end_time": 4911.288,
"index": 196,
"start_time": 4882.619,
"text": " It's not like I've got, I'm a thermostat and I'm taking a continuous record of the temperature. That is a continuous, I'm actually getting discrete quantal observations. And I think once you move to that kind of generative model or biophysics that entails the gradient flows under that generative model, then I think you're in a position to talk about sort of planning and cognition and curiosity of the kind that we're talking about."
},
{
"end_time": 4941.493,
"index": 197,
"start_time": 4911.698,
"text": " So I, you know, I'd literally put time constants on this, I'd say about, you know, 300 milliseconds, if you've got the ability to represent the future, 300 milliseconds in, you know, sorry, represent the consequences of your actions, more than 300 milliseconds in the future, then I put you in, you know, having crossed that loop, one of those cuts and moved from being a sort of, you know, a virus to, you know, to, I don't know, perhaps an insect, I'm not sure."
},
{
"end_time": 4965.742,
"index": 198,
"start_time": 4942.363,
"text": " Are you seriously putting the 300 milliseconds forward or is that facetious? It's facetious but not as facetious as you might think. I think there's one way of celebrating the"
},
{
"end_time": 4992.278,
"index": 199,
"start_time": 4965.947,
"text": " escapes from panpsychism that we started off with when Mike was saying what he spends his time talking about. One of the ways is to commit to vagueness and how many grains of sand do I need before this ceases or becomes a pile and try and work out what the quantity is that defines that spectrum that we were talking about before, along which there is vagueness at some point according"
},
{
"end_time": 5014.462,
"index": 200,
"start_time": 4992.278,
"text": " of a certain kind arises. But I also think there are natural kinds that actually have bright lines between them. And, you know, what I'm saying here, and I don't think I've said it before, and I probably wouldn't say it again, unless I've simulated it, but there's certainly my experience"
},
{
"end_time": 5039.019,
"index": 201,
"start_time": 5014.923,
"text": " qualitatively different kinds of generative models that represent time in very very different ways. You can get so mathematically things like Kalman filters that can be cast or linear quadratic control robotics and the like that can be cast purely in terms of futures that are represented in terms of velocities and accelerations and higher order motions that are just in the dynamics and the equations of motion."
},
{
"end_time": 5063.302,
"index": 202,
"start_time": 5039.753,
"text": " And you can realize and build these things, you can simulate them and they behave in a very plausible way. You can get them to do handwriting, you can get them to sort of self-organize and do all sorts of interesting things a la thermostat. But that's as far as it goes. These things don't show curious behavior. To do curious behavior, you've really got to roll out much further into the future."
},
{
"end_time": 5071.766,
"index": 203,
"start_time": 5063.524,
"text": " And in running out further into the future, it's very difficult to represent the future just using differential equations unless you move to very, very high orders of motion."
},
{
"end_time": 5095.981,
"index": 204,
"start_time": 5072.346,
"text": " at which point you start to sort of to minimize that complexity, start to articulate things in discrete time, and then the numerics become much simpler and doable. But then you've got a qualitatively different kind of generative model, you move from a continuous to a discrete time, a quantized generative model."
},
{
"end_time": 5126.101,
"index": 205,
"start_time": 5096.374,
"text": " Um, so that for me is not, uh, does not allow a vagueness, you know, there is actually a natural kind that will represent things in the moment in terms of flows and dynamics and other natural kinds that will represent things, um, in terms of discrete, not universal clock time, but discrete iterations. Um, and I'm saying that the, you know, for, for certainly our scale of interactions and the way that we sense make, I would have thought 300 milliseconds was not a controversial"
},
{
"end_time": 5154.582,
"index": 206,
"start_time": 5126.271,
"text": " I say that because that's the cognitive moment. It's also the time interval between the way that we sample the world, not just with our eyes, but when we sniff, we sniff at a frequency, which means that each new sampling of that world, that palpation with our chemoreceptors and our factory receptors is a game every 200 milliseconds. So just to me, there's a sort of discrete belief updating, and it is a discrete belief updating that is"
},
{
"end_time": 5181.834,
"index": 207,
"start_time": 5154.923,
"text": " underwritten by the way we act. The plans have to be in these discrete chunks of time. So that's where the 300 milliseconds come from. You see it everywhere. Mice whisking, they whisk at that frequency. So they touch the palpate every 300 milliseconds. The way I'm speaking now, every phoneme is reaching your ears every 300 milliseconds or so, or possibly less. So wherever you look in our biological scale,"
},
{
"end_time": 5210.828,
"index": 208,
"start_time": 5182.398,
"text": " Well, it's interesting because in physics, the number three has an interesting role in fundamental physics because there's three generations of"
},
{
"end_time": 5238.422,
"index": 209,
"start_time": 5211.288,
"text": " Well, I was just having a visual impression actually when Mike was talking and of babbling and"
},
{
"end_time": 5260.64,
"index": 210,
"start_time": 5239.155,
"text": " The visual impression was of our sort of model of the quantum vacuum. And when we think of when we think of the vacuum, or at least I think of it visually as little things popping in and out of existence all the time. And so"
},
{
"end_time": 5290.23,
"index": 211,
"start_time": 5262.824,
"text": " One could think of that as a kind of babbling, right? The field itself is exploring what's going on outside of itself. That's extremely interesting. By forming a little entity to go off and explore and then report back. So this is a very different picture from the one that Carl gave in that"
},
{
"end_time": 5320.52,
"index": 212,
"start_time": 5290.879,
"text": " I suppose it also has a similarity or two that when we write down a model of that activity, it is discrete. It is operating in a discrete kind by generating discrete events, even though it's modeled in a continuous space time, which once one digs a little bit deeply, the continuous space time itself becomes discrete."
},
{
"end_time": 5350.299,
"index": 213,
"start_time": 5322.193,
"text": " But this raises a question for me, which is another of the things that I think about a lot, which is the question of what we mean by randomness. And certainly to take a Bayesian point of view where probabilities are all subjective,"
},
{
"end_time": 5379.582,
"index": 214,
"start_time": 5351.049,
"text": " It becomes very difficult to say what randomness might mean if used in a way that's meant to be objective or to refer to something objective. And we model things like fluctuations in the vacuum as random, which from a Bayesian point of view just means it's uncertain to us. It represents information that we"
},
{
"end_time": 5409.48,
"index": 215,
"start_time": 5380.23,
"text": " don't have and quite possibly can't get even in principle. But if we think of that system as an agent, then of course, we're expected to be exploring its world, whatever it is, whatever exists, that's not it. And maybe we can really think of babbling all the way down."
},
{
"end_time": 5428.08,
"index": 216,
"start_time": 5409.684,
"text": " as a model of what's happening in these phenomena that we persist in thinking of as random. But we also don't typically think of these sorts of systems as agents. This is incredibly interesting."
},
{
"end_time": 5442.278,
"index": 217,
"start_time": 5428.933,
"text": " What I heard Carl saying, among other things, is that there is a qualitative transition when things go from analog to digital in a certain sense, from continuous to phased."
},
{
"end_time": 5466.203,
"index": 218,
"start_time": 5442.722,
"text": " And based on what Chris just said, if you just remind us to think about the kind of the quantum world, where at least to my understanding, a lot of things are in fact, you know, sort of discrete. Does that mean that really could it be that that is the sort of base state as Chris just outlined with that kind of proto cognitive exploration all the way down?"
},
{
"end_time": 5490.452,
"index": 219,
"start_time": 5466.544,
"text": " that these continuous models that we put on top of it by coarse-graining, you know, tiny events into the kind of macroscopic things that we see here, that's an abstraction that sort of loses the essential mindfulness of it and then we have to try to rebuild it again by the time we get to brains. But actually, all we've done is obscure the fact that all the way from the beginning, it was already digital to start with,"
},
{
"end_time": 5519.753,
"index": 220,
"start_time": 5490.452,
"text": " And we just sort of put some Vaseline on our lens here to make everything look sort of continuous, which then made it look like there was nothing going on at the lower levels. And then somehow we're shocked that we have to recover it from somewhere later on. That's what I just heard from the combination of those two explanations. A very interesting idea, and it connects nicely to the history of mathematics, or at least the history of mathematics post Descartes and Leibniz."
},
{
"end_time": 5548.985,
"index": 221,
"start_time": 5520.401,
"text": " Carl, did you have any thoughts to add to that? Yeah, I'd probably"
},
{
"end_time": 5577.91,
"index": 222,
"start_time": 5549.787,
"text": " This is why I wanted to hang out with Mike and Chris to learn about quantum theory, because I do think there's something fundamental about the way that we carve up the world or the way the world can only be defined in a carved way into tiles or quanta or discrete things. So I think that those observations that we just heard I think are quite fundamental and it would be great to understand"
},
{
"end_time": 5606.749,
"index": 223,
"start_time": 5578.234,
"text": " everything from a sort of a quantum perspective, a discretized perspective, and how continuous constructs like, you know, space and time emerge from that, you know, if necessary. Having said that, I have to say that as a physicist, I would bring something else to the table. So I'm quite comfortable dealing with random fluctuations. So for me, random fluctuations are just variables that change very, very quickly."
},
{
"end_time": 5629.804,
"index": 224,
"start_time": 5607.09,
"text": " So if I want to write down as a physicist, so now I talk about a physicist that doesn't know anything about quantum mechanics. So I just want to write down and launch that equation to describe my universe. I write it as a mixture of states and random fluctuations and somebody asked me what license is that? I just say random fluctuations are very, very quick and the states are very, very slow."
},
{
"end_time": 5655.572,
"index": 225,
"start_time": 5630.145,
"text": " And this is just a natural consequence of some kind of renormalization group. You know, I could make them even slower by going to the next state scale and throw even more fast stuff. So there's nothing really, if you like, problematic from the point of view of that kind of physicist in the random fluctuations. It just means that all the"
},
{
"end_time": 5678.37,
"index": 226,
"start_time": 5655.913,
"text": " stuff that determines flows and dynamics of the kind that we can that we engage with and we can plan has a particular form. That particular form does actually bring a quantile aspect to the table or a discretization aspect to the table that may or may not be if you like"
},
{
"end_time": 5707.688,
"index": 227,
"start_time": 5678.814,
"text": " But what I'm talking about here are the mathematical image of life cycles and biorhythms and oscillations. And I mean that right from the perspective of a dendrite in my hippocampus oscillating away at a fast gamma frequency, you know, sort of say 80 Hertz through to the motion of the heavenly bodies. Wherever you look, you get this"
},
{
"end_time": 5736.476,
"index": 228,
"start_time": 5708.029,
"text": " solenoidal flow that is characteristic of things that hang around and don't dissipate almost immediately and these are not part of the random fluctuation but they become more evident as you go away from the fast random stuff into the slow dynamics in the sense that you know clearly the orbit of the moon around well the orbit of the earth around the sun for example"
},
{
"end_time": 5765.776,
"index": 229,
"start_time": 5736.681,
"text": " This deals with very, very massive, large things where all the random fluctuations are effectively averaged to zero. Whereas if you go down to the quantum scale, then you're not licensed to do that. And these solenoidal flows become much less apparent. But our scale, if life is, on this reading of physics, sufficiently classical to have"
},
{
"end_time": 5792.671,
"index": 230,
"start_time": 5766.408,
"text": " oscillations, biorhythms, things that would you need for feedback that didn't go off to plus or minus infinity, for example, just saying that you're circulating in some state or phase space in a way that you keep yourself to some attracting set. And as you move through this, you are naturally going to be oscillating in a highly nonlinear possibly chaotic way, but still the essence is life cycles."
},
{
"end_time": 5821.544,
"index": 231,
"start_time": 5793.183,
"text": " If that's the case, then the Poincare section does indeed permit a description that is necessarily discretized in the sense you're going to pass through the neighborhood of various states at discrete points in time recurrently. So that's what I was thinking from the point of view of a physicist who does not have the fluency of quantum mechanics to get back to a discrete thing, but it also I think highlights the"
},
{
"end_time": 5832.21,
"index": 232,
"start_time": 5822.159,
"text": " The deep connection between things recurring discreetly or occurring discreetly."
},
{
"end_time": 5856.886,
"index": 233,
"start_time": 5833.166,
"text": " being one way of defining another characteristic of biological self-organization, which is basically, I repeat, biorhythms and life cycles that have this sort of self-contained itinerancy. Of course, in that self-containment, you've got this natural resolution of ambiguity and uncertainty because you always know you're on the same itinerant path."
},
{
"end_time": 5886.903,
"index": 234,
"start_time": 5857.312,
"text": " repeatedly visiting the states that you've once occupied. So that's what I would say from the point of view of a physicist. I may say something very different in a few years' time when I've learned more about quantum theory. So let me ask Chris, could you ever describe that kind of physicists' treatment of dynamics in pure quantum information theory at that kind of scale? Well, I think if we try to think of scaling up from"
},
{
"end_time": 5916.186,
"index": 235,
"start_time": 5888.012,
"text": " Descriptions of things like quantum fields, then this looks impossible. But if we think of it informationally, maybe it doesn't look impossible. I'll go back to John Wheeler's notion that a bit is fundamentally the answer to a yes or no question. And in"
},
{
"end_time": 5940.486,
"index": 236,
"start_time": 5916.408,
"text": " The information theory is about bits and systems that exchange bits. But... Do you mind repeating that last part? But what? What Wheeler's aphorism doesn't specify is the complexity of the question that has a yes or no answer."
},
{
"end_time": 5967.449,
"index": 237,
"start_time": 5942.125,
"text": " So in your case, Carl, you have an oscillatory system that passes after a discrete time of particular boundary. So the system may well be asking the question, have I passed this boundary? And the answer is yes or no. Or am I passing the boundary right now? Well, that's one bit."
},
{
"end_time": 5995.725,
"index": 238,
"start_time": 5968.592,
"text": " That's a yes or no question, but it's an extremely complicated question. So it may be that we can develop a theory that leaves the complexity of the generative model unspecified and concentrates just on the bit flow across the boundary that"
},
{
"end_time": 6020.64,
"index": 239,
"start_time": 5995.964,
"text": " counts as evidence in one direction and perturbative action in the other direction. And then I think the two pictures fit together rather nicely, because one can imagine the generative model being the model that generates the questions, operating at many different scales."
},
{
"end_time": 6052.91,
"index": 240,
"start_time": 6023.916,
"text": " Where one could think of the scales as computational complexity. And this gets back in a sense to your question about language. Language is a way of posing questions. It's a way of babbling in the world. This process of babbling, it seems to me, as far as I understand, it's like poking a prodding to test your own self-model to see what's the difference between me and the world. It sounds like that"
},
{
"end_time": 6082.193,
"index": 241,
"start_time": 6053.592,
"text": " never stops, even though we call it babbling as if infants do it. Does it ever stop? Are we not engaging in this right now? And I like this phrase babbling all the way down. I'm going to put that on a t-shirt by the way, Chris, or, and it also, by the way, seems like babbling all the way up too, because we're constantly engaging in this. So firstly, is that true? Does babbling have an end? And, and also does it not presume a self in order to test the self?"
},
{
"end_time": 6111.237,
"index": 242,
"start_time": 6082.654,
"text": " What I mean is, Carl, when you were speaking about the process of babbling, it's the infant doing something on the world in order to make a model. But that first part was it is doing something to the world to me already presumes the external world. Whereas the way that I'm understanding it is that the self is a model, but it's also in your point of view, it's something objective. And I believe from my readings of Chris, I believe Chris is more on the non-dual end"
},
{
"end_time": 6140.503,
"index": 243,
"start_time": 6111.476,
"text": " Right, well, I mean, just to dispel the notion that the self is objective. No, I didn't mean that. I meant it exactly the same way that I think both Chris and Mike have been intimating that, you know, from the inside of a cell or a person or the baby's skull,"
},
{
"end_time": 6171.152,
"index": 244,
"start_time": 6141.152,
"text": " Self is quite a sophisticated construct, fantasy, hypothesis, that is physically realized, possibly neuromally, in some lucky creatures and some lucky children. You may actually find some children never get that far. I'm thinking of people with severe autism, for example, who don't have a theory of mind, simply because they never developed the notion that I am a person."
},
{
"end_time": 6188.677,
"index": 245,
"start_time": 6171.425,
"text": " And, you know, what comes along with that is that, you know, if you've got very, very severe autism, or, you know, if you like, sort of a theoretically idealized autism, that means that personhood doesn't exist. So you would be, I would regard you as an animate fringe or, you know,"
},
{
"end_time": 6213.302,
"index": 246,
"start_time": 6189.07,
"text": " some interesting autonomous vehicle, but you would never be a person because I don't have the hypothesis that there are people out there and that people share certain sort of phenotypic characteristics and have intentions in the same sense that I have intentions. I'm not aware of that because I just don't have that hypothesis. So it's a long-winded way of saying that I didn't mean to imply"
},
{
"end_time": 6242.381,
"index": 247,
"start_time": 6213.302,
"text": " any objectivity here. I was more going the other way, really. I'm just saying that selfhood, that would be one part of consciousness. And if you want to associate cognition and consciousness, then possibly that kind of minimum selfhood would be necessarily part of cognition. This is a gift that you have to get by crossing one of those bright lines."
},
{
"end_time": 6269.667,
"index": 248,
"start_time": 6242.739,
"text": " and it's just a fantasy that you bring to the table to make sense of all the impressions on your sort of sensory surface or your holographic screen or your cell surface or also the receptors you are equipped with to sense what's going on out there. Does that make sense? Yeah, I know that Mike and Chris, you may have a response to that, but I just want to quickly respond to what Professor Friston was saying."
},
{
"end_time": 6296.971,
"index": 249,
"start_time": 6270.043,
"text": " And this is to both of you. How do you then prevent an existential crisis as you start to study this? So for example, Carl, before you and I were speaking, and I told you about some experiences I had where I've deeply felt what you're saying is being true, that the self is a fantasy, and this oneness with all this world being generated in one's mind. And I felt like that was akin to a psychotic episode. And it was extremely destabilizing. And to be quite frank, I'm still recovering from that."
},
{
"end_time": 6326.903,
"index": 250,
"start_time": 6297.261,
"text": " And so I'm unsure how any of you all do this work without constantly being in a debilitating state of existence. Well, existence is the question here, and it's an existential crisis which I'm inquiring about. So how is it that you prevent yourself from getting into that? How could you say these words of, well, the self is not objective, the self may be a fantasy, quote unquote, but not then be able to not speak and look around you in awe and"
},
{
"end_time": 6356.852,
"index": 251,
"start_time": 6327.244,
"text": " I think there is a good deal of existential uncertainty involved in doing what we're doing. Just par for the course. Right. How do you get over it? I'm not sure when gets over it."
},
{
"end_time": 6383.2,
"index": 252,
"start_time": 6359.036,
"text": " I think it's more being comfortable with it. You know, I think you raise a really important point and I think this is in a class of a number of really destabilizing"
},
{
"end_time": 6411.903,
"index": 253,
"start_time": 6383.592,
"text": " uh ideas that one picks up as as one gets older and studies things and there are many of these. I remember as a as a child being completely freaked out by the fact that I realized that you know we grow up and we become these adults and these adults bear some resemblance to the children but they're definitely not the same thing and so in some sense you could sort of see it coming it's almost even even before you realize that you're gonna you know you're gonna you're gonna die biologically someday before long before that you realize that you're going to undergo some transition"
},
{
"end_time": 6432.551,
"index": 254,
"start_time": 6411.903,
"text": " that's going to, you know, turn you into a completely different sort of creature, which really, you know, has some continuity, but maybe not, you know, certainly not you persisting in the same way as you are now, right? So some kind of weird butterfly, you know, caterpillar butterfly effect where you're just not going to be around in that same way as you are now, right? So that's kind of the first"
},
{
"end_time": 6457.807,
"index": 255,
"start_time": 6432.551,
"text": " thing you realize as a child. And then there's many others that sort of, you know, you learn about Boltzmann brains and all the different, you know, Humean arguments of it, you know, all this stuff that, you know, I had to, I did, I did a philosophy class with my son when he was little, we were doing homeschooling. And so I kind of did this. And I realized very quickly that I have to be very careful with this stuff, because if you write somebody who has the capacity to understand it, it's extremely destabilizing."
},
{
"end_time": 6484.138,
"index": 256,
"start_time": 6458.217,
"text": " And I think that at this point for myself, I sort of go with Descartes a little bit in the following sense. I'm not depressed by it in the sense that whatever we discover is a truer model of the self than we may have had. I don't think there's any way in which it makes us less valuable, less interesting, less engaging."
},
{
"end_time": 6498.148,
"index": 257,
"start_time": 6484.582,
"text": " All of that is still true. It's just that now we've realized that a different set of mechanisms can now implement that magic than we thought. So you may have had some sort of idea that, well, there has to be some sort of"
},
{
"end_time": 6522.329,
"index": 258,
"start_time": 6498.643,
"text": " different kind of unique process that makes this permanent self and therefore I'm this individual and I have true beliefs and preferences and I exist in some way. I don't think you can ever convince yourself that that isn't true and that you don't exist. It's just that now you find out that actually the way you get to be that is by these sets of mechanisms and they're very interesting. It's the sort of thing that Karl and Chris have been talking about."
},
{
"end_time": 6542.278,
"index": 259,
"start_time": 6522.329,
"text": " So that's what gets you to be this kind of individual, and it's kind of amazing that that works. But to me, that doesn't diminish all of the things that I thought were true about what the individual actually is. The impermanence is there, but we knew that from the time you were a kid, you knew that it wasn't going to be permanent."
},
{
"end_time": 6572.21,
"index": 260,
"start_time": 6542.278,
"text": " So that that part, you know, you sort of have to deal with from the start. After that, it's all you're learning after that is just that you were really mistaken about what it takes to be that sort of creature. You thought maybe it had to be some magical whatever. Now you think that, wow, it's a set of feedback loops and whatever else. So that's surprising. But that's OK. We learn about, you know, we learn about different underpinnings to things all the time. So that's OK. But I don't I don't think any of it diminishes the sort of just the incredible awe of being a sentient being that can think about these things."
},
{
"end_time": 6601.493,
"index": 261,
"start_time": 6573.183,
"text": " Yeah, just to reiterate that. So I didn't in any way mean to demean our selfhood, or your selfhood in particular, by calling it a fantasy, I actually meant it, and indeed have written about it, in terms of something quite fantastic. So we've written papers called the fantastic organ, the brain is a fantastic organ, just because it can produce these fantasies is absolutely fantastic. And of course, they"
},
{
"end_time": 6631.493,
"index": 262,
"start_time": 6602.005,
"text": " in relation to our earlier conversation, the capacity to have these fantasies which have incredible explanatory power, if you can attain them and you can maintain them, is the part of the complexity which allows us to cross these bright lines or these gaps or cuts and become closer to the kind of creatures that can even entertain the notion of the heart problem."
},
{
"end_time": 6657.773,
"index": 263,
"start_time": 6631.886,
"text": " Mike, I just love that meta-answer. We don't know what the answer will look like. I'm going to tell Andy Clark about that. He'll tell David Choms. To come back to the point, I didn't mean to demean that. Phrases that I remember coming across and having to learn when I was a student, things like existential angst and ontological security, all of these things rear their"
},
{
"end_time": 6682.005,
"index": 264,
"start_time": 6658.2,
"text": " face if you are worried about these things but the very fact that you are able to entertain the notion that you are not real is quite remarkable and I repeat it is a gift of a fantastic sort it is the kind of fantasy that very few creatures have and sometimes even few people have so there are people who actually will go out there to try and"
},
{
"end_time": 6707.944,
"index": 265,
"start_time": 6682.824,
"text": " experience that and to become and treasure the the gift of being able to make sense of the world as me by taking drugs or doing mindfulness therapy in order to actually get to a state where almost a Freudian oceanic state where selfhood is dissolved so they can experience it and certainly if you've taken sort of psychedelic drugs in sufficient quantity"
},
{
"end_time": 6727.21,
"index": 266,
"start_time": 6708.166,
"text": " You may have that kind of experience, that's a derealization or dissociative experience, which is the realization that your construct, your fantasy of selfhood that is engaged and engaging continuously for as long as you live, then"
},
{
"end_time": 6754.753,
"index": 267,
"start_time": 6727.5,
"text": " you know, if you want to appreciate that, then you have to experience that, you know, the alternative hypothesis. So, you know, experiencing derealization, or this is depersonalization, is actually, I think, one really important step to self understanding. So I don't see it as at all flattening nor diminishing. It's just part of self understanding, which not everybody gets."
},
{
"end_time": 6783.439,
"index": 268,
"start_time": 6755.316,
"text": " I mean, in some sense, the most destabilizing thing of all ought to be just the bare facts of developmental biology. When you learn that you are in fact not some sort of indivisible monadic whatever, which I'm not even sure what that would entail if you were, but the fact that all intelligences are collective intelligences, we are all made of parts, and the idea that you used to be this quiescent oocyte, this little bag of chemical reactions,"
},
{
"end_time": 6812.995,
"index": 269,
"start_time": 6783.439,
"text": " that did very little and then very slowly sort of step by step you turned into whatever you are now. If that's not destabilizing I don't know what is because just knowing nothing knowing nothing else about evolution or I mean to me developmental biology is sort of like the you know the queen of all the sciences because just that fact that you go from chemistry and physics to mind in nine months plus to however many years it takes and that you are literally a collection of cells that you could see in pond water more or less"
},
{
"end_time": 6839.735,
"index": 270,
"start_time": 6812.995,
"text": " is just if you know you're a bunch of cells in the trench coat you know like the in the cartoons like if if that's not destabilizing i don't know what i don't know what else is so quickly just i know we have to get going and i'd like to wrap with a set of brief questions but quickly carl and perhaps well anyone who would like to answer what is it that separates those who are depersonalized or go through certain psychotic episodes from"
},
{
"end_time": 6855.145,
"index": 271,
"start_time": 6840.282,
"text": " What is it that separates those who are fortunately like myself where it's abating and I'm able to function and I'm able to have friends and so on versus those who go into a ward? Because obviously those issues can, like you mentioned, it's a gift."
},
{
"end_time": 6883.097,
"index": 272,
"start_time": 6856.169,
"text": " Well, I'll answer from the point of view of a psychiatrist with a commitment to sort of a mechanistic or theoretical understanding"
},
{
"end_time": 6908.08,
"index": 273,
"start_time": 6883.814,
"text": " I think it's the inability to resolve the uncertainty that attends entertaining the dual hypothesis that I may exist or I may not exist as me. But as soon as you increase your hypothesis space, you naturally induce an uncertainty, an ambivalence and a kind of ambiguity. So it is exactly the same, if you like,"
},
{
"end_time": 6928.114,
"index": 274,
"start_time": 6909.616,
"text": " We're talking about the same imperatives that we've been talking about all along, which is to sort of resolve uncertainty, resolve ambiguity, to make the most sense of things. So if you've entertained this hypothesis that you are not you and you are you and you can't fully resolve that, then you will find yourself in a state of chronic uncertainty."
},
{
"end_time": 6956.254,
"index": 275,
"start_time": 6928.558,
"text": " That unfortunately does have pathological consequences. I'm not sure that being in a psych ward is necessarily too pathological. I mean, you know, there are ways to live very nicely and constructively and fantastically in psych wards. And I've done that. I spent two years in a therapeutic community. But there will be other sequelae of an inability to resolve uncertainty. They're expressed clinically and physiologically in terms of allostatic load and the like."
},
{
"end_time": 6983.08,
"index": 276,
"start_time": 6956.681,
"text": " chronic stress and the like, which can be very unpleasant and ultimately can have physical consequences from the point of view of the effects on your body. So the question now is why can't some people resolve that uncertainty? I think it's then a question of the biophysical processes that enable you to update your beliefs and build good models of your world."
},
{
"end_time": 7009.172,
"index": 277,
"start_time": 6983.49,
"text": " Again, this is a very pragmatic answer from the point of view of a clinician. Probably the answers lie in the way that your brain has assembled itself to coordinate and attend to various sources of evidence and it looks at the mechanisms that underwrite that ability to deploy attention rest very much upon"
},
{
"end_time": 7039.07,
"index": 278,
"start_time": 7009.462,
"text": " Neuromodulating neurotransmitters are the kind you know you have heard about in the popular media and will have read about things like dopamine, serotonin and indeed oxytocin. All the drugs that are either used or the targets of psycholytics or psychedelics or psychomimetics or drugs that are used to try and control some of these feelings and some of these signs and symptoms specifically that attend"
},
{
"end_time": 7066.442,
"index": 279,
"start_time": 7039.343,
"text": " anxiety disorders and depression. So, you know, understanding the, this is not the feedback loops that do self-organization in a developing embryo, but the same kinds of mechanisms that do the feedback loops and the self-organization of different structures within the brain that can be construed as blankets within blankets or holographic screens within holographic screens or cells within cells."
},
{
"end_time": 7087.21,
"index": 280,
"start_time": 7066.664,
"text": " or complying with the same principles, but in the particular instance of the brain doing this and understanding the neurochemistry and the biophysics of this fundamental sense-making self-organization. It turns out that these are certain, these neurochemical messages"
},
{
"end_time": 7111.715,
"index": 281,
"start_time": 7087.346,
"text": " are particularly important in resolving that kind of uncertainty. And it may be that you have either taken drugs or you have just grown up to not be able to deploy these neurotransmitters properly. Again, leading us to, well, how would you gain control of it? Well, you might want to do mindfulness training, meditation and the like, just internal attention state training basically to mentalize"
},
{
"end_time": 7130.316,
"index": 282,
"start_time": 7112.09,
"text": " how you deploy your attention."
},
{
"end_time": 7152.517,
"index": 283,
"start_time": 7132.892,
"text": " Mike, I know that you had a set of questions that you're eager to ask both Carl and Chris. Did you already get to all of them?"
},
{
"end_time": 7181.237,
"index": 284,
"start_time": 7153.387,
"text": " That would be impossible, but I got to many of them, most of the important ones. I learned a massive amount today. This has been just amazing. But I will be pestering both of them with additional questions in years to come. Great. To end, I have a question that I didn't email you that just occurred to me as you all were speaking. It's something that Professor Levin that you said, what are some aspects of your field, maybe their theories or concepts that"
},
{
"end_time": 7207.022,
"index": 285,
"start_time": 7181.613,
"text": " your peers are extremely resistant to, but you are not. Wow, I don't know if we got time for all that. That's a lot of stuff. I mean, the most basic, there are many, I mean, the most basic thing is this notion of profound symmetries between these fields. You know, this idea that you can basically import"
},
{
"end_time": 7235.623,
"index": 286,
"start_time": 7207.176,
"text": " concepts and strategies from cognitive neuroscience or behavioral science and deploy them in other spaces, developmental biology, metabolism, that it even makes sense to talk about, you know, goals, memories, planning, navigation, all of these kinds of things outside of what we're used to, which is some, you know, rat in a maze or something like this. So that continuum, this idea that something important goes all the way down or maybe a long way down, even if not all the way down,"
},
{
"end_time": 7264.497,
"index": 287,
"start_time": 7235.623,
"text": " That is in a few circles, that's okay. In most circles that I deal with, that's just a complete heresy. Yeah, there are many others, but that's kind of a basic one. Chris? Well, I think that Mike really nailed it with that answer. And I would say"
},
{
"end_time": 7294.531,
"index": 288,
"start_time": 7265.213,
"text": " Yes, it's not a popular idea that physics and biology and psychology are actually all more or less the same science. But I think that they are. Professor Friston. Yeah, well, that would be my favorite heresy. I'm just going to agree with Mike and Chris. You know, physics, biology, psychology, they're all the same thing."
},
{
"end_time": 7324.684,
"index": 289,
"start_time": 7294.957,
"text": " And I just wanted to end by saying, heresy, I'm not so sure anymore. The number of times, for example, I see Mike Levin cited in high-end philosophy journals and treatments suggests to me that, in fact, it's going to be the minority, at least in a few years' time, that don't subscribe to this more holistic encompassing, I think,"
},
{
"end_time": 7348.473,
"index": 290,
"start_time": 7325.043,
"text": " And I'll end with one that I see from particularly Chris and Michael is it sounds like"
},
{
"end_time": 7375.469,
"index": 291,
"start_time": 7348.865,
"text": " without using the word, it sounds like what you all are doing or trying to understand God or trying to come up with a model of God. And the reason why I say that is because with looking at a neuron and saying, okay, how can a neuron look around and have some model and know that it's embedded in something higher? That sounds to me like, well, if we apply that to ourselves and continually do so, we would reach something akin to God. And then Chris with the quantum babbling, if one can think of it like that. Well, what is babbling?"
},
{
"end_time": 7400.247,
"index": 292,
"start_time": 7376.288,
"text": " I know that maybe you all don't want to use the word God, but for me, what I have that many of my peers are extremely against would be to philosophize about God or to explicate God or to take God seriously as not an old man in the sky with a beard. And so I see you all as doing something similar in them. I'm happy to be in your company. I'm extremely lucky."
},
{
"end_time": 7426.032,
"index": 293,
"start_time": 7401.067,
"text": " all science starts with an act of faith, with a very important act of faith, which is that the world is in some way understandable, that we are not just a bubble in this, you know, random universe that just happens to, you know, as a total random string of coincidence looks like it's got laws, and then tomorrow somewhere, you know, while you sleep, the whole thing falls apart, and we go back to a random, you know, sort of random distribution of events."
},
{
"end_time": 7447.619,
"index": 294,
"start_time": 7426.032,
"text": " So we all there's no proof of any of it of any of that right we take we have to take that on faith and that may be for a scientist that may be the most destabilizing thing if you can't get yourself to believe that that the world isn't for some weird reason understandable and amenable at least to some extent to our logic and probing then then you can't do science you can't do any of the things that that we do so"
},
{
"end_time": 7467.978,
"index": 295,
"start_time": 7447.841,
"text": " So being useful and hopefully productive in what we do starts with that act of faith that is just completely, to my knowledge, it's an axiom. It's not based on any evidence that you could possibly have. So I do think that sort of underlies everything that we do, just this idea that we can in fact squeeze some understanding out of what's going on around this."
},
{
"end_time": 7498.609,
"index": 296,
"start_time": 7469.445,
"text": " I'd also think you'd be surprised. I know that you said that it doesn't comport with the traditional definition of God, but it depends on what tradition and how far back one goes because there are mystics, let's say Christian mystics and Muslim mystics or Islamic mystics and so on. And obviously then the eastern end is much more mystical. So, okay. Well, thank you all so much for coming out to this and I've learned a tremendous amount and I hope you all have too."
},
{
"end_time": 7512.108,
"index": 297,
"start_time": 7498.933,
"text": " I'm blessed and thank you so much. Well, thank you very much. This was this was a very interesting conversation and thank you very much for putting it together. Yeah, absolutely. This is this is great. Great, great idea. Thanks for having us on."
},
{
"end_time": 7535.401,
"index": 298,
"start_time": 7512.807,
"text": " Bye everyone. Bye. Take care."
},
{
"end_time": 7556.783,
"index": 299,
"start_time": 7537.602,
"text": " The podcast is now finished. If you'd like to support conversations like this, then do consider going to patreon.com slash C-U-R-T-J-A-I-M-U-N-G-A-L. That is Kurt Jaimungal. It's support from the patrons and from the sponsors that allow me to do this full time. Every dollar helps tremendously. Thank you."
}
]
}No transcript available.