Theories of Everything with Curt Jaimungal - Joscha Bach on Intelligence, Existence, Time, and Consciousness

Episode Date: October 7, 2020

YouTube link: https://www.youtube.com/watch?v=3MNBxfrmfmI0:00:00 Introduction 0:00:17 Bach's work ethic / daily routine 0:01:35 What is your definition of truth? 0:04:41 Nature's substratum is a "quantum graph"? 0:06:25 Mathematics as the descriptor of all language 0:13:52 Why is constructivist mathematics "real"? What's the definition of "real"? 0:17:06 What does it mean to "exist"? Does "pi" exist? 0:20:14 The mystery of something vs. nothing. Existence is the default. 0:21:11 Bach's model vs. the multiverse 0:26:51 Is the universe deterministic 0:28:23 What determines the initial conditions, as well as the rules? 0:30:55 What is time? Is time fundamental? 0:34:21 What's the optimal algorithm for finding truth? 0:40:40 Are the fundamental laws of physics ultimately "simple"? 0:50:17 The relationship between art and the artist's cost function 0:54:02 Ideas are stories, being directed by intuitions 0:58:00 Society has a minimal role in training your intuitions 0:59:24 Why does art benefit from a repressive government? 1:04:01 A market case for civil rights 1:06:40 Fascism vs communism 1:10:50 Bach's "control / attention / reflective recall" model 1:13:32 What's more fundamental... Consciousness or attention? 1:16:02 The Chinese Room Experiment 1:25:22 Is understanding predicated on consciousness? 1:26:22 Integrated Information Theory of consciousness (IIT) 1:30:15 Donald Hoffman's theory of consciousness 1:32:40 Douglas Hofstadter's "strange loop" theory of consciousness 1:34:10 Holonomic Brain theory of consciousness 1:34:42 Daniel Dennett's theory of consciousness 1:36:57 Sensorimotor theory of consciousness (embodied cognition) 1:44:39 What is intelligence? 1:45:08 Intelligence vs. consciousness 1:46:36 Where does Free Will come into play, in Bach's model? 1:48:46 The opposite of free will can lead to, or feel like, addiction 1:51:48 Changing your identity to effectively live forever 1:59:13 Depersonalization disorder as a result of conceiving of your "self" as illusory 2:02:25 Dealing with a fear of loss of control 2:05:00 What about heart and conscience? 2:07:28 How to test / falsify Bach's model of consciousness 2:13:46 How has Bach's model changed in the past few years? 2:14:41 Why Bach doesn't practice Lucid Dreaming anymore 2:15:33 Dreams and GAN's (a machine learning framework) 2:18:08 If dreams are for helping us learn, why don't we consciously remember our dreams 2:19:58 Are dreams "real"? Is all of reality a dream? 2:20:39 How do you practically change your experience to be most positive / helpful? 2:23:56 What's more important than survival? What's worth dying for? 2:28:27 Bach's identity 2:29:44 Is there anything objectively wrong with hating humanity? 2:30:31 Practical Platonism 2:33:00 What "God" is 2:36:24 Gods are as real as you, Bach claims 2:37:44 What "prayer" is, and why it works 2:41:06 Our society has lost its future and thus our culture 2:43:24 What does Bach disagree with Jordan Peterson about? 2:47:16 The millennials are the first generation that's authoritarian since WW2 2:48:31 Bach's views on the "social justice" movement 2:51:29 Universal Basic Income as an answer to social inequality, or General Artificial Intelligence? 2:57:39 Nested hierarchy of "I"s (the conflicts within ourselves) 2:59:22 In the USA, innovation is "cheating" (for the most part) 3:02:27 Activists are usually operating on false information 3:03:04 Bach's Marxist roots and lessons to his former self 3:08:45 BONUS BIT: On societies problems

Okay, I think it's a good day for 2020. The sun is out again. The sky doesn't look like Mars as far as I can see. Air is so-so, but the meteor didn't hit yet and we're still waiting for the big one in the Bay Area. What's your work ethic like? What do you do on a usual day? Usual day. There are no usual days. So in some sense, every day is its own and every day has its own demands and so on. And I do block out times and do plan things ahead. But for the most part, I get up, I have breakfast, I work, I spend time with the family in between.

And at some point, I maybe read a few pages in a book or watch a movie with the family and go to bed. Do you tend to work in solid blocks, uninterrupted, because you need that for focus? It depends on the type of work. So when I write a quote,, I need long solitary times alone. If I want to write a book, I need to have solid weeks for myself. So I don't write a book at this point. Maybe I figure out how to do this anyway while being locked up with the family. And when I do conceptual work and so on, I can do this in short intervals. And sometimes I do something else, interrupt it, because I realize I need to write something down.

We're going to get into some fairly deep questions right off the bat. So what is your definition of truth? Do you take a correspondence theory of truth, for example? I think that truth is defined in a mathematical paradigm, which means it's defined within a language and it's a certain value that you set on variables that have the property that they can be true or false or have a truth value that varies in degrees. And in some sense, truth is a predicate that you can calculate in this context. And you can translate this into the languages that our mind is typically using which models and in these models, we can have some kind of truth definition,

which means that the model depending on the type of model that you have, can conform to any of your mathematical truth definition. So it can be something that is can be reduced to a set of axioms, for instance. And this means that it can be, in some sense, compressed to the set of axioms or expanded from the set of axioms into a certain state of that descriptive system. And it's difficult to apply truth to an outside world. So I don't believe in reference theory of truth. These references can only exist between different models. On the other hand, we normally never talk about the outside world because it's this weird quantum graph that is not accessible to us and that we take to be the system that generates patterns on our retina and

on our systemic interface to the universe and All these patterns are great models and different. The primary one is an integrated model of the entire universe of a perception is inspect or the perception of external things a plus proprioception, right so we in some sense have Res extensa and the rest cogitans to speak with the cart and res extensa is not the universe itself It's our model of the universe. It's the idea that everything that we perceive corresponds to a region in the same three-dimensional space that is dynamically changing as a temporal extension as well. And res cogitans is everything else.

Our ideas that we have about that, the anticipations that we have, hypotheses that we have, memories that we have, intentions that we have, and so on. And these two interact, but there are several types of models that coexist in our own mind. And when we refer to something in the world, we refer to something in the integrated model of the universe that we have, that is changing.

And it's not static, it's not reality. For instance, my model of the universe at some level contains colors and sounds. And there are no colors and sounds in the physical universe. The physical universe does not offer them. But colors and sounds are functions that our brain computes to interpret certain types of patterns in the universe. So you're saying that when we talk... Sorry, so you're suggesting that the underlying reality is physical and it's a quantum graph. What do you mean by it's a quantum graph?

What I mean is that there is an outside pattern generator. And the physics is exploring the idea that this pattern generator can be explained by a causally closed set of rules. Somewhere out there, there is a system that generates us and that generates our experiences. And the big insight of computation is that a computational system is the necessary and sufficient means to produce arbitrary patterns. And we don't have alternatives to computational descriptions that are able to do that. So it turns out that computation is a way to frame language. And if we want to have languages that describe systems that can produce patterns and are self-consistent and can reduce the first principles, these are computational systems.

computational systems. And then when I say it's a quantum graph, it's a graphical representation, one that disassembles a system into nodes that can hold, for instance, state and links between them that translate the state between the nodes. This is a very general computational description. And so in some sense, we can describe everything, especially extended things that play out in a space as a graph. The space is basically, if we talk about something like a geometric space, it's a very, very regular graph that happens if you zoom out far enough. So basically a graph is so many nodes and so many links between them and so regular ways of translating information in them that you can describe the

function of the entire thing in the limit by operators that give rise to geometry. All languages can be summarized in mathematics and then there's a subset that is computation and this is the one that corresponds to our world. Is that what you're suggesting? So what I would say is that a way of looking at what mathematics is, it's the domain of all languages. And mathematicians are starting out with the simplest languages and exploring them. And these are the formal languages, those where you already know all the properties, or you define them in such a way that you have a good chance to explore most of the properties step by step, and especially you can define them in such a way that you can make proofs in them. And the reason

why we explore mathematics are multiple. There's one that we want to understand the things that we only intuitively understand, for instance geometry, right? We have an intuitive understanding of geometry and that's because our brain makes geometric models of the world that we are embedded in. And so we want to have a way to talk about these models in a way that makes them explicit and allows us to debug them and allows us to express them as formal systems, to teach them, to check up on whether our minds did the geometry right and so on, right, and to have all these models of models these formal descriptions of geometry. That's one reason why we need mathematics and it's one of the reasons why our present

mathematical tradition started. It started very often as geometrical descriptions and the algebraic descriptions that we have those in terms of formulas and so on are often a specification for the geometry. And in some sense, we teach mathematics often the wrong way around in school. We teach kids algebra first as an extension of counting or generalization of counting. And then you have something like x, y equals, say, x squared. And this is an algebraic description, right? And then we learn how to graph these functions. And then we notice, oh my god, this graph of this function looks a lot like a parabola. What a coincidence. And now we see if we can use this algebraic description to describe a parabola. But I think that the invention

was the other way around. Like the world is full of parabola. But I think that the invention was the other way around, like the world is full of parabolas. Whenever you throw stuff, you see a parabola. And then you ask yourself, what's going on here? How can I systematize that? And then you realize, okay, I can make a specification that makes this thing computable because I can compute the algebraic description. There's an algorithm for computing y equals x squared. And this allows me to compute the trajectory of an object that I throw. This is one of the ways that we discovered mathematics. The other one is that all languages that we are using, so logic is a mathematical principle. It's in some sense, a subset of the natural languages that we use,

but we can extend it in such a way that it's encompassing the natural languages in a way. And there was always a big hope that we could make logic so general that we can make natural languages so precise that they become the same thing, that these languages in which we refer to facts can express things that we can prove in a formal strict sense in a way that we can build machines

that can perform language and that can make statements that are true or false and so on. We look at the world and we see some patterns and we try to model those patterns and mathematics is a great way of modeling those patterns. Is that essentially what you're saying or is it something else? So mathematics allows us to build arbitrary languages, not just natural language, but it also allows us to build languages that for instance start out only with bits or that starts out with bit vectors. And you could say that, for instance, a machine learning system is using a language that as an input, as a bit vector, for instance, the bit vector that comes in from a camera sensor, and then it maps this to other bit vectors. And just by finding order

and the patterns, it's able to gradually make some sense of these patterns. It's able to relate them. And it's, for instance, it's able to discover that if you feed it enough bit vectors, all these bit vectors refer to objects in a three-dimensional space because that's the best compression that you find over them. It's the best way to make sense of these patterns to build a model of these patterns that is most predictive of their structure. And mathematics allows us to construct and understand the languages in which that takes place. And you alluded to another thing that is the difference between computation and mathematics. The classical definition of mathematics that we use in our tradition is mostly time and stateless. It's basically everything is taking place in a single state. So if you want to

express a temporal sequence in mathematics usually use an index and this index can be discrete or continuous and You just basically define something like a loop Only it's not like a loop in a programming language Which goes to this literally step by step to compute the sequence you have some instance an integral operator and this integral operator states that god makes a loop in one moment and this loop can integrate over infinitely many elements because god can do that right it's an idealized version of what mathematics can do and it turns out that there are no gods that have

the power to take infinitely many elements and really integrate them. What we can do is we can take a very, very large number of elements that to us is almost indistinguishable from infinitely many elements and then we can integrate them in a pretty, were a pretty long time span of however long it takes, but not over an infinite amount of time. But if we had really infinitely many elements that we would try to throw together, to integrate, to put into one function and execute that function over these elements, this is not possible in anything that is implemented in the physical universe. And it's also not possible for anything that is implemented in mathematics itself

without leading into contradictions. So in some sense, what Gödel discovered, I think, is that into contradictions. So in some sense, what good discovered, I think is that mathematics, this infinities that actually uses these infinities to compute things is leading into contradictions. And the thing that doesn't do that, that only uses states that we can, in some sense, count up and actually execute on this is computation, the subset of mathematics that can be constructed, constructive mathematics. And I would say that constructive mathematics is not a subset of mathematics, unless you say that mathematics also encompasses those things that don't work. Right? You could

say mathematics is all languages, also those languages that are utter nonsense because they are contradictory. That is one way of looking at it. But the part of mathematics that works is computation. The part of mathematics that can be implemented. And I think for something to be real, it needs to be implemented. Something that is not implementable, something that cannot be realized as a system

that is executable by anything in mathematics or in physics is not real, right? So constructive mathematics is the part of mathematics that has a chance of being real. Right, okay. Now, when you say real, do you mind defining? I know you just subtly defined it. Do you mind explicitly defining what real is?

Because the way that I'm thinking about this is, you just mentioned there's classical mathematics. Let's just call it classical mathematics. And it leads to contradictions, like Gertl mentioned. And one way of interpreting that is, there are limits to our knowledge. There are limits to mathematics. Another way is that this is the wrong mathematics. This is not what is describing our universe.

What's real is something else. So Gertl, you actually said that this is a blind alley. Don't go down that alley. Rather than, this is the alley and here's our limitations. You're like, no, no, no, no, no. no here's constructive mathematics and you're saying that this is real. Okay help me understand what you mean when you say real. Just as a small tangent I think it's unfair to Gödel to pretend that this is what he thought he said. Gödel actually believed in this god of mathematics that can do these infinities.

And it was a big shock to him when he came up with this proof. And so in some sense he discovered something about the mathematics that he believed in that seemed to be real to him that he could not really get square. And it's also a thing that a lot of mathematicians and physicists struggle. For instance, I think it's part of what motivates Roger Penrose when he rejects the idea that computers can be conscious because he thinks that human minds

can do non-computable mathematics. In the same way as Gödel did to him, Gödel has proven that computers cannot do all of mathematics, but human minds can. For instance, Penrose has discovered these Penrose tilings. These are, they're tilings which have an infinite variation, right? There are infinities. How can you claim that there are no infinities? But you can, you're only looking at a function, I think, that basically has an open-ended result. It's similar to the function that computes pi. Pi is not just a value. Pi is a function or a set of functions that give you, for instance,

if you translate this into the decimal system, digits, it's an infinite sequence of digits. And you can get as many digits as you can afford. But you can never have a function that relies on having known the last digit of pi. There is nothing where you feed in pi and you get a result. You can only feed in a number of digits of pi and you get a partial result. And this result has the property that for it converges because the digits that you get denote smaller and smaller fractions of pi. And so the results tend to converge to something, but ultimately you don't know the end of it. And so we have to be fair to Gödel to, he did not believe in computation as the solution. He's, he was strongly buried to this pre-constructive mathematics

or non-constructive mathematics. And he basically thought that mathematics somehow has a big problem in itself. And it doesn't. It was just a problem of the formalization that mathematicians have been using for a long time. Okay, so let's get back to your original question. Yeah, what is real?

Let's have it up again. What is real? Can you sum it up again? What is real? Also, while we're talking about this, pi. Does the number pi exist? I don't know. This sounds a bit odd, but to some people, it's like, well, the Pythagorean theorem was discovered. It wasn't invented. The Pythagorean theorem, for example, is true regardless of if humans are around.

Okay, then we can say, does one exist? Did we discover the number one or did it exist? And then we can similarly ask, does pi exist? So what do you think? Does pi exist? What does it mean to exist? I think that pi is being constructed and there is a procedure for this construction that we have discovered and pi is as real as the mandible tractal they are real in pretty much the same sense right there's this pattern that is uh self-contained it is uh there whether you are looking at it or not and it's a pattern that cannot physically exist at least not at an infinite resolution, because it's a pattern that is defined with a procedure that gives you more resolution the deeper you get into it. And you will never know the last details of the Mandelbrot fractal and same way you will never

know the last details of pi. Right, but there's a difference between whether or not we can know it, whether or not it's computable, whether or not it's computable, and whether or not it exists. I think that pi, in the sense that you say it's defined with the last details, does not exist because it cannot be implemented. There can be no system built into the universe that is expressing the last digits of pi. So the last digits of pi do not exist in the sense that they are real, that they are there somehow, that they are out there and influence anything,

that they have a causal influence on something else. And for something to exist, I think it is implemented in such a way that it has a causal influence on other things and can be consistently described with a model that is has a causal influence on other things and can be consistently described with a model that is describing that causal influence. So do the integers exist? Not all of them. Right, there's a large number N.

So I would say that the integers are a model. They're a way of talking about things that are real. But it's pointless to say that this model exists, because it's a structure that is being constructed. The implementation of the structure in your own mind is being constructed, the realization of Piano's arithmetic in a computational system exists in a way. It's implemented to such a degree that for a certain amount of time,

that system can be stable enough to allow us to perform computations with a certain accuracy. And at some point, these computers are going to fall apart. What I would say exists is one way of looking at this. And it's basically a thesis that I don't know how to prove is that there is a causally closed lowest layer that exists. And this is basically the mechanics of the universe, some kind of automaton that uses everything that happens. And there seems to be something, right?

Something seems to be real. And why is there something rather than nothing? I don't know why that is. In some sense, it's the most obscene thing that something exists rather than nothing. It's tremendous. It's much more confusing than everything else I'm aware of. And the easiest explanation for that

is that existence is the default. So perhaps everything that can exist, which means implemented without contradictions, exists. And so you have the superposition of all these computational operators and some of these regions of the resulting fractal contains us. Right. So I'm sure you've heard of the multiverse. This sounds similar to the multiverse that anything that can happen happens, it exists in some way shape or form we had a power outage i'm sorry and the and the worst thing the the audio is gone okay okay that's okay i hope that you're recording on your side yes yeah yep i am

good okay yeah okay he is conspiring against us yeah that's all right it doesn't want the secrets revealed okay okay so so yosha i was asking what's the difference between what you just mentioned and the multiverse idea which is something akin to whatever can exist whatever is possible exists in some way shape or form in some other universe? I think that the multiverse in the Everett-Villagram version, this idea that there is basically at every decision of the quantum collapse, the collapse of the wave function, the universe splits into copies of itself, is a slightly different conception. It's basically a mathematical paradigm that describes that the universe is branching in ways that parts of the description of the universe no longer causally influence the

other parts of the description of the universe. Imagine that you are describing the world as something like a cellular automaton. And in the cellular automaton, you have particles interactions, which might be like gliders in the game of life, like these regular patterns that might influence each other and change state and move through the topology of a space that you define in a certain way. of a space that you define in a certain way. And this thing is defined in such a way that the computation of the entire thing is very inefficient because at every step, the thing splits up into many, many subtopolities in which you have copies of these gliders

or variants of these gliders moving about in different ways, but only a subset of the gliders is going to influence the others. So from the perspective of the future of any kind of system, the only things that are real to you are those that can causally influence you. And if something moves away from you in a way that it can no longer causally influence you, and because it's no longer occupying the same space in a way. We basically have a space with a dimensionality that increases and increases and increases. And this might look like it's extremely wasteful, but that's only from the perspective of somebody who is outside of the system

and cares about one of the timelines of the system. And we only exist within one of those trajectories inside. And for us, we would only experience a smaller and smaller fraction of the resulting computation at every moment, or in the future as in potencia. And this is in some sense what this multiverse idea describes. It's a particular mathematical formalism. It's not exactly the same as the thing that I just described, because

that is independent of the idea of such a multiverse. So what I described is a way to look at the universe as something like an evolving fractal. So you have a generator function that produces all the possible generator function by just enumerating them and just executing them in parallel. And as a result, you get time, space, and matter, and matter is basically the structure that is evolving in this space and is propelled along it.

And space is a set of locations that you can discern that can contain information that is discernible and the ways that the information can travel between these locations right uh yeah so you're saying that we can produce some subsets that are causally closed and then there are others that don't influence us at all and to us whoever lives within this causally closed place so you know what i mean they start constantly looping back to us so basically you send a signal into the universe this signal is going to influence certain things and as a result you you get feedback from this right you push a certain thing and you see the results of what you pushed because photons are bounced off that thing that you are pushing and there is a limit to that that is

the visible universe right they if something goes beyond the region from which light can reach us, it no longer has a causal influence to us. And the multiverse theory is that there is not just this boundary in a very, very large distance, but there is a boundary next to us where we do things that lead to information flowing away from us and not coming back to us, where we do things that lead to information flowing away from us and not coming back to us. Still, there is a conservation of information. This conservation of information is that we can always basically, in some sense, figure out what we did, because all the influences that matter get back to us. And of course, the there is a little bit of a tautology

in there, if we are producing things in the universe with our actions that in some sense would generate new information that goes away from us and doesn't come back, we would never know about it. And in some sense, the multiverse is an inevitable description of such a universe where the collapse of the wave function can happen in multiple ways. Where are the other ways happening? Where are they going? Why is it that we only experience this particular collapse of the wave function? And the multiverse theory is a possible answer to that. And there are other possible answers to that. One could be that, for instance, the collapse of the wave function is deterministic in a certain sense, and that means that it's influenced by things that are non-local and that we cannot pinpoint. Do you believe that the universe is ultimately deterministic?

I don't really have beliefs about that. I think that the conservation of information seems to imply that the universe is deterministic. It also depends on what you see as determinism or indeterminism. A deterministic function is one that gives the same result every time. So the universe being deterministic means we can find a function that describes everything using the same function. that describes everything using the same function. And if it's an indeterministic one, you get a different result every time as a transition between adjacent states of a system.

And you can express this by taking a deterministic function and adding some infinitely long string of random numbers or a string of random numbers that is longer than you can observe. And in this case, you will have an irregularity that is not predictable. And so you can always find in some sense a deterministic model to describe a system unless you can completely see how it's made up. And this problem is that we cannot understand the makeup of an indeterministic system, you cannot open a box and see a true

random number generator, because a true random number generator cannot be constructed. And you can construct a function that gives you the digits of pi, but you cannot construct a function that gives you an infinite number of random numbers. What do you think is giving rise to the initial conditions? You mentioned that maybe existence is the default. What gives rise to that? And what gives rise to the rules?

Maybe there are no initial conditions. So the Big Bang is not in some sense initial condition of the universe itself. It's initial condition from the perspective of an observer. of the universe itself, its initial condition from the perspective of an observer. When we look at the universe, we notice that while there is it's not symmetrical, there is an asymmetry in the universe, but it's reversible in the sense that every state seems to have exactly one preceding state.

This means that there are parts in the universe that remember other parts of the universe. If you are, for instance, taking a billiard ball and you send it through the universe, then there is a reversibility in this, because you could, if you would trace all the interactions that the billiard ball enters directly and do this in the inverse, you could basically restore the state that the billiard ball enters directly and do this in the inverse, you could basically restore the state that existed before the billiard ball entered it. But you also see this fundamental asymmetry. And this is the stuff behind the billiard ball remembers the billiard ball passing, and the stuff before the billiard ball doesn't yet know that it

comes. It doesn't remember that the billiard ball will soon cross through this area and wreak some havoc and change things, right? And this means that there is this entropic arrow in the universe. And this describes in one direction that in which way information gets dispersed through the universe, in which way locations begin remembering the state of other locations, in which way information gets smeared around between locations. And there is an ideal state where this hasn't happened yet, where all the locations have information that is self-contained, that is not correlated with other information in other locations. And this is the Big Bang state. And if you go past the Big Bang state

in the opposite direction, so if you would move to the time before the Big Bang state, the entropic arrow points away from the Big Bang. So basically, all the directions that point away from the Big Bang in time are a future of the Big Bang. Because they are all ones that remember a trajectory that ends in this ideal original state where the information is perfectly correlated with location. But this doesn't mean that the Big Bang state has ever existed. It's just a

mathematical description of a singularity in our entropic arrow of time. I'm not clear what you mean by that there's a memory of the state. But either way, when you say that there's the Big Bang and then there's an entropic arrow, there's a huge mystery, a mystery in physics about the arrow of time. Like what is time and why does it move forward? Is it static? Like you mentioned, if we conceptualize the world in classical non-constructivist mathematics, then there's no room for time. What is time in your model? Is time fundamental? I think that if you want to have an observer, you need to have a system that is multi stable. And it's moving between these states, right, it needs to be able to be influenced by the

environment, and as a result, form some kind of memory. The memory means that as a result of an observation of the system made its environment, it's changing its internal state in a particular deterministic way. And the way that we describe our universe, we notice that we can describe our universe in terms of states of which some explain this current state by being its past. You could also say, imagine you enumerate all the possible universe states and some of these states will look like they contain the memory of past states because they can be the result of a state transition where you are using a

permutation on previous bits, basically the laws of physics, and you get to the next state. And as if you go backwards you get a timeline. And all the possible timelines in the space of the enumeration of universe states, these are all the possible temporally extended universes. From the perspective of an observer, time is, in some sense, the rate of change in the environment as observed by the observer, which means it's relative to the rate of change in the observer. So from observer perspective, time is intrinsically relativistic. From the perspective of outside of the system, which we cannot take,

time would be state transition. So there is a way in which you would enumerate the states by defining a function that orders them. Okay, you just mentioned from the perspective of an observer outside the system. Now, in general relativity, an observer is what follows the time-like curve. Okay, so you're in the cone. It can't be space-like, it can't be null. Now, if you ask, what the heck is the experience of a photon? We can't do that because an observer is defined as what's in the time like cone the thing is the

photon is not multi-stable the photon does not change state right that's why the photon is not able to observe the photon can be absorbed and re-emitted with different properties but by itself the photon doesn't change and only a system that is able to change state can observe and so whenever you have a particle system like you and me right like a single one a single one though what about a single particle it depends what kind of particle but uh so in some sense you can describe every uh particle system as some kind of a compound particle. But it basically depends on whether you can change the property of that particle without that particle changing its identity, so to speak.

So as soon as this model of the particle no longer applies, you use a different description. You used to write for Edge. I don't know if you still do. But in the last question, you wrote, what is the optimal algorithm for finding truth? I'm curious to know, what is the optimal algorithm? I don't know that. So in some sense, the this is a question for people that work in artificial intelligence.

That is, when you exist in circumstances that are learnable, you try to learn about these circumstances, you make a model of them, what's the best algorithm that you can employ to discover a model of what state you are in? And you can show that for the general case, that's not possible. Not every system is learnable. But it seems that the set of universes that contain us has certain limitations. Not every possible universe can contain us. The universe that contain us must be essentially a controllable universe. From our perspective, it seems that, in some sense, atoms control elementary particles and molecules control atoms

and cells control molecules and organisms control cells and societies control organisms and so on, right? So we look at a hierarchy of control. It's one way of looking at the complexity that we're seeing. And a system that is being controlled, that has some kind of control structure, implies that the controller implements a model of that what it is controlling, otherwise it couldn't control it.

If you can implement a model of something, it means that the model is discoverable, which means that the system is learnable. A controllable universe is a learnable universe. The non-controllable parts of the universe will look random to us. They're not controllable. But what we can see is largely controllable, right? There are some zero-point fluctuations of the universe that we cannot control. We don't know where they come from. We don't know how to influence them. And the particles that we are looking at are those parts of the fluctuations that are regular enough to be described in some kind of good control model. It's tempting to think of the universe like something like, say, the surface of an ocean that is in constant fluctuation. And these fluctuations look random

to something that is swimming on that surface. But on the surface, there are also regular patterns, some things like water vortices. And these water vortices need to have particular properties to be stable. And when they have these properties, they're almost indestructible. Basically, there is a certain number of molecules that need to be involved in a certain way or at a certain speed. And then suddenly you get a vortex that is so stable that it can only be broken up

by hitting other vortices that have compatible or properties that are in some relationship to that, right? And this is roughly, I think, a model of the particle dynamics that we're looking at. There are some patterns in the overall dynamics that are so stable that they can be described by control models and can be described by control models and can be exploited by higher level control structures. So they give rise to complexity.

And the ones that can't, we just don't observe them or we observe them as random or we observe them as noise? Yes, they're not predictable, right? In some sense, the meaning of information is its relationship to change in other information. So if you see a blip on your retina, the relationship of that blip to other blips on your retina is the meaning that this blip has. This meaning that you discover is a function that describes the relationship of blips on your retina to each other, to these different changes. on your retina to each other, to these different changes, as, for instance, people in a room that is lit and the sun shines on, and these people walk around, exchange ideas, and the room is three-dimensional, and so on and so on, right? This is the function that your brain discovers to describe all the blips on your retina. There are other blips on your retina that do not fit

into this function, and these blips are noise. And there's a lot of noise on our retina. And in some sense, this is also how we interpret the universe. Everything where we discover a relationship to the other things, this is what we can model and the rest is noise. And the amazing thing is that physics is clarifying the universe to such a high degree and there's so little noise left. Do you think that ultimately there will be no noise left that we'll be able to characterize everything?

I think that we will always be able to construct a function that behaves as if there was no noise, that basically explains everything. But it doesn't mean that this function is necessarily predictive. It doesn't need to be the correct function. And it's not the only function that can explain it. It's like, imagine you live in a computer program like Minecraft, and you observe all the patterns around you. You can always construct a computer program that will work like Minecraft

and will explain all the patterns around you, even the random patterns, right? You can always come up with some pseudo random number generator that produces this, but you will not necessarily be able to discover the truth of the matter, except if the world that you live in is so simple, that it suggests itself that there is only one simple function, or a class of simple functions that can be mapped on to each other with a simple transformation that gives the same result every time, right. So imagine that you discover yet you live in a Mandelbrot fractal, the Mandelbrot fractal is like two

lines of code. And you can express these lines of code in many, many different ways. So the many ways of expressing the same function, the same sequence, but there are mappings between all of those. And so if you discover the Mandelbrot fractal, you can basically say this is the simplest function that explains it, this is the reality that you're looking at. This simple sequential definition of how to calculate these pixels on that plane. And it's conceivable that we would

find such a function for our universe. But if the universe is very complicated, we can still find a very complicated function. In some sense, the quest of physics is to find the shortest function. And the current function that we have that explains most stuff, not everything, but most, is the standard model. It's like half a page of code. And it's already very short, but physicists keep hoping for something that is much shorter. Because half a page of code is still very complicated. And people ask themselves where does all this complexity come from? Do you think that ultimately the code is short?

complexity come from. Do you think that ultimately the code is short or do you think that like Feynman was quoted saying that it might even be an onion where you just keep unveiling the layers and it's more and more complex, less and less complex, and it doesn't follow necessarily a pattern? Maybe there's not even a center. Do you believe there's a center and do you believe that center is simple? It's a weird metaphor. It's mostly ways that we think about the world that we should deconstruct before we trust them, right? What does it mean for something to be a center? It's inside of something and the onion is outside and it's spatially aligned. So what this describes is probably a hierarchy of models. And the question is, does every subsequent layer of modeling that we discover become simpler and simpler? Imagine you take a

microscope, and you look at a cell, and you zoom in and every, every level of resolution, but you discover a new structure. The question is, does the structure become more simple or more complex, right? And does the model converge to something ultimately? Yeah. And I think that it's very likely that it does converge to something from what I understand, but I cannot make such a proof at this point. I think that it must converge to something because there are no infinities, things need to be

constructed. There's also this weird properties that for instance, if you look at the particle generations, they are integer fractions that describe how they differ and their properties. So it could be that there are smallest building blocks of information that make up the particles that we're looking at. There is no infinite division between them. And so it could be that the causally closest, lowest layer is somewhere inside. It's something that we can still construct. 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- I N G.com slash everything and use the code everything. Is it cool if we get into the nitty gritty? I have some questions about your PowerPoint slides.

Yes. Just one, one more thing. I didn't answer the question for, for the optimal algorithm to discover. Yeah. Great. Let's get back to that. Right. So let's, let's get back to that.

Sorry. I went onto a tangent there, but I thought it was necessary. So in some sense, when you look at artificial intelligence systems, So let's get back to that. Sorry, I went onto a tangent there, but I thought it was necessary. So in some sense, when you look at artificial intelligence systems, there was a succession of ideas. And in classical AI, people have been looking at problems like playing chess and constructing algorithms to solve that problem, mostly by hand. And when they looked at new problems, they had to construct a new algorithm and then we had this idea of finding more general algorithms that can work over a very large class of problems general problem solvers and there is the difficulty that if you have a description that is so general that it works on many problems then the description is typically so general that it's too long for a concrete problem.

It basically takes too long to explore the space using this general description. You get an explosion of complexity. There are too many possibilities that you would have to look into if you enumerate them all with your general description. So you end up needing a targeted exploration of the space of possibilities. And this is what the current wave of AI is doing. It's looking for algorithms that discover solutions for problems. So instead of implementing a solution for chess,

we give the system a specification for chess, and then we let it explore the solution space, right? It's discovering an algorithm to play chess using an algorithm that we give it. So we construct a learning algorithm. And the next stage could be that we just describe an algorithm that discovers a learning algorithm for this. This is meta learning. So we don't build a system that learns, we build a system that learns how to learn a given thing. Right? And then the question is, is this the generally best solution already?

Or if not, maybe we need to get one step above and we need to discover a general theory of search. A general mathematical theory that says how to optimally search given certain boundary conditions. The way that I'm conceptualizing what you said is in the first wave of artificial intelligence, it's almost like if-then statements, extremely structured, here's how to play chess, I'm gonna implement the rules myself. Then the second wave is look at a slew of chess

and learn the rules and learn how to play well. Then you're saying that, okay, well, that's great. It's like Watson. Watson's wonderful at one task. But then when you wanna generalize, Watson, can you also move your arms? Okay, well, that requires, and also be regular Watson. That's a bit difficult.

Well, Watsonatson can you move your arms and talk to people it requires too much time because your your function is too general but yet our brains do it are you saying something like that or am i completely wrong almost almost uh starting with watson it's a family of things basically it's a brand that ibm has been using to label its ai efforts or part of its AI efforts after the jeopardy thing became famous. So Watson is not one thing, it's many things. And it's slightly different from AlphaGo because AlphaGo is an algorithm that is specified in a particular paper and DeepMind is not renaming everything that it does into AlphaGo because AlphaGo got famous.

It has AlphaZero and has a number of other things that are somewhat related. And eventually, when you talk about the algorithms, you can also use the technical name. You could look at, for instance, DeepQ learning. And DeepQ learning is a particular small class of algorithms that can solve certain problems very well and others not very well. At the moment, the most interesting class of algorithms a lot of people are talking about, or one of the most interesting ones, is transformers, which look for embeddings of features based on similarities over many layers

using an attention-based algorithm. And the fascinating thing is that the same algorithm that discovers structure and language also discovers structure in images. And now the tempting question is, is there an optimal algorithm that can discover structure everywhere? And that maybe is recursive in the sense that it's starting to explore what kind of strategies of discovering structure are the best ones. And then it settles for those. Is there a universal recipe? And I don't think that there is a hope

in the sense to say this algorithm doesn't exist and humans will always be better. Because if humans are better than the algorithms that we discover, even the most general one, it means that humans are implementing a more general algorithm than the most general one. And means that humans are implementing a more general algorithm than the most general one, and we can discover it, right? There is no reason why humans can do

something that an algorithm that we write down somehow cannot do. We can also write down the formula for evolution, and we can, in the worst case, evolve the algorithms that we need to. And everything that we do is in some sense an optimization of brute force evolution where we do a blind search. We do try to find directions in which we can optimize the search. And for me this question is there this optimal algorithm to discover choose the optimal learning algorithm that would mean we can stop doing science because as scientists we can only now execute on this algorithm. And of course, we can leave this to a machine now and we should go to a beach and surf instead.

And there's a factor of time because we could implement evolution and let it run for a billion years, and then it would discover something that's greater than us in terms of general, but you probably wouldn't need to because the evolution that we are looking at is only slow for multicellular organisms, because in multicellular organisms, you need to bootstrap the entire organism before you can evaluate it, which in our case takes very long. For us, it's also necessary to train this, the new instance of the algorithm for a long time

before it becomes functional again, right? If you want to breed the optimal scientist, you cannot just vary our genome and look at the outcome. You also need to expand this. So you need to incubate for nine months and then you need to raise this until it's in its 20s or 30s or 40s. And then you get an evaluation and then you can decide which ones of those you should put into the next generation, right?

This is something that is a result of the way that biological evolution works based on cells. If you just evolve single-celled organisms, it goes very, very fast, right? In a few hours, you can have quite substantial changes. And the microorganisms that, for instance, we breed in our gut are often quite specific, trained for tasks. Basically, our gut is breeding organisms for its purposes. And that is done in reactors. Our guts are, in some sense, breeding reactors for microorganisms. And it's also a substantial part of our nervous system is duty-bound to deal with this breeding task, with farming these microorganisms.

All these gut neurons are mostly dealing with, I think, maintaining this extremely large farm that has specific organisms in it. And this works because it's such a quick thing to breed single-celled organisms. But if we build AIs in this way way we would not have to reinstate the entire phenotype based on a genotype over a long time and retrain it. We can probably just change the parts that we need and leave everything else intact. So the evolutionary research that we could do in our technical systems can be many many orders of magnitude faster. It can also be much more directed because often we know what we're looking for. So we can define a fitness function that is very close to the solution or that is narrowing the solution space dramatically. I remember you said that

artists are tuned to their loss function, something like that. That's what they're obsessed with. Now, the way that I understand that is what you're saying is artists are interested in their behavior and what incentives and rewards they have, which are their values. And they're trying to replicate that or represent that in some level. Is that correct or is that wrong? Close. So what I try to say is that art is in some sense, a dysfunction. Sorry, what function art, a dysfunction,

dysfunction? Yes. Because so they are basically two different ways of looking at art. A non artist, a normal person, healthy person sees art as a tool. It's instrumental to something. It might be a tool for education, for entertainment, for signaling status, for ornamentation. And an artist, and I am from an artist family and totally identify with this stuff, is a person that thinks that the purpose of art is to capture conscious states. Full stop. This is the purpose of art. It's this observation for the sake of observation, because the is this is the purpose of art it's this observation for the sake of observation because the conscious state is the important thing that needs to be

conserved an artist is somebody who eats to do art and a craftsman is somebody who might produce artifacts but they do art to eat right it's it's a very different way of looking at things. For them, the art is instrumental to doing something. The art is an artifact. And for the artist, it's a service. It's a service to something that is more important than all the other things that you could be doing at the same time.

And so if you see the artist as the metaphors of artificial intelligence, you could say that the mind of an artist is a system that has fallen in love with the shape of the loss function. It's no longer optimizing the loss function itself. It's trying to figure out what it looks like. It's trying to capture a certain structure for its own sake, no longer for the reward that is going to increase as a result of applying what you've learned. Do you see yourself as doing that? Yeah. I think it's a deformation and I can retrace it in a way. It's an identification that happens at a certain level.

For instance, my mind is a very conceptual mind. I perceive myself as something that thinks, that solves problems, that reflects. And I perceive my emotions and my body as being outside of that for the most part. Emotions come into me and they disturb me and I need to deal with them. And I need to make sure that they don't distract me or that they don't overwhelm me or they don't kill me.

But I don't identify for the most part as this emotional being. Of course, sometimes I go over in that state and I realize that state in which you are this emotional being that is motivational and that is embodied and experiencing. That's the normal state that we are supposed to be in. And a lot of scientists and philosophers are identified in a similar way as me in a way. I think that a scientist or an artist or a philosopher is born when a child discovers that it trusts its ideas more than its feelings.

And it happens often because you are wired in a slightly different way than other kids around you. And as a result, your social interactions fail and you can't explain that. So you act like every other child based on your intuitions, on the 99% of what your mind is doing and is training. And these intuitions are wrong. What do you mean when you say that your ideas are different than your feelings? Because obviously, your ideas are somewhat influenced by your feelings and your perceptions. Yes. How do you disentangle them? My ideas are stories that I construct. I have agency over my ideas. I don't have agency over my feelings before I'm in doubt.

When I'm a child, my feelings are the result of the interaction between the model that the mind maintains of the universe and the model that it maintains of the self according to the needs of that self, right? So when things happen in the universe, the mind evaluates them as good or bad, which means their frustrate needs or their satisfy needs. And as a result, the self represents joy or suffering. And when the mind doesn't have the correct intrinsic model, an innate model that you're born with, and how to interact with the environment, then your needs are going to be constantly frustrated. So for instance, I grew up in a forest far remote from other

villages. And when I got into first grade and met other kids pretty much for the first time, I had difficulty relating to them. They were not interested in the same things as I was. I was a nerd. I was reading a lot. I was interested in math and physics and science fiction and history and stories and so on. And other kids were interested in soccer. And I couldn't get myself to be interested in soccer. And as a result, I was excluded from many of the games. And later, the same thing happened with respect to politics. It was Eastern Germany. You were expected to pretend that you conform

with the prevailing ideas. And if you didn't, you were punished. And even if these ideas were illogical. And at the same time, your teachers told you to be critical and don't take in all the ideas from your environment without criticism, because you know, this is how fascism happened in Germany. People took in bad ideas from the environment without criticism, because you know, this is how fascism happened in Germany. People took in bad ideas from the environment without criticizing them.

They did not develop moral agency. So I thought, the thing that I'm doing, that I question everything, that I want to know why something is the way it is, and you need to explain it to me before I believe you. That would be a virtuous thing. And so there was an apparent hypocrisy that my teachers told me one thing and I behaved according to that and they punished me if I did this. It was difficult for me as a child to make sense of that. And as a result, I basically decided at some level, it was not something that was deeply reflected, but it was inevitable to distrust my emotions. I had to, because they were wrong. They were not pointing me in the right direction. I had to form theories on how things work. And I think in a healthy mind, this development is temporary. You are a being that is directed by its intuitions. And these intuitions are something that is trained. It's not something that is random or superstitious. It's your intuition that tells you whether this is going to be a good relationship or not, whether you should marry that person or not, whether this

person should be your friend or not, whether you should take this job or another. Because if you try to make proofs about this, you're not going to get anywhere. It's way too complicated. Your thinking, your ideas are way too brittle. And science is in some sense the part of our mind that is meant to deal with our darkest and murkiest emotions. With those where we don't have solid intuitions. So science has marginal value only for the individual. Largely you need good intuitions. It has also only marginal value for society.

Not every, ought to be a scientist, society wouldn't work like that. Because science is too brittle. The ideas that scientists come up with don't tell you what kind of relationship you should enter. And if you overvalue science, then your society is going to go astray. You need to have solid common sense. You need to have good intuitions, good understanding of how things work. And only in those areas where these intuitions break down where you need to make proofs this is where science really helps and shines you mentioned that your intuition is trained trained by who and what by your life by your experiences so you start out yourself or trained by society or both no i think that society is is uh often seen as too

big of an influence or uh i think that society is a small part of the physical universe and the thing that for instance trains your intuition of how large an object is when it has a certain distance from you and you see it's so and so large on your retina that's not given by society right that's given in some sense by some innate intuitions but eventually it's given by learning and you learn it by being embedded into the universe. When you have an intuition of how many steps you should take to catch a ball that is flying to you, it's not society that teaches you that. It's your interaction with the environment that teaches you that. And so the same thing is also true for social interaction. Most parts of

the social interaction are not taught by society. It's you being immersed in the environment that teaches you what to do. And it does so based on mostly innate impulses that make you interested, for instance, in the mental states of others, and what they think about you and how they react to you. And in some sense, society is the result of those things. It's not the cause of those things. I recall you saying that in a repressive environment, like, let's say, Eastern Germany, that artists flourish in a sense because they have to constantly define themselves. I don't see that as necessarily salutary because you then have to define yourself with regard to the society. So it's almost like you're a contrarian. But then without society, you don't have an identity. You know, for example, when someone says, I'm counterculture.

Okay, but that means you're defining yourself in terms of the culture, not within yourself. In other words, why is it positive for an artist to grow up in an environment that is intolerant or inflexible? It's often there's a motivating force between the art and the artist. And so there is a topic that is motivating the artist to talk about. And there can be many topics that artists are on about. In a simple case, the topics that the artist is obsessed with is the imagery that is possessing the artist. It could be just the overwhelming force of, for instance, musical patterns or of visual patterns, just the aesthetics itself that wants to have an expression. In that case, the society is irrelevant. It could also be that what's important to the artist is the discourse with other artists,

where there is a history of art and there are certain movements in art and artists are engaging with this movement. And of course, these are the artists that by definition are the most influential ones. But not all artists care about being influential. There are many artists which only care about their own inner imagery and this imagery has only an incidental relationship to what happens around them. And there is also a lot of art that is directed on the political or the social and so on. It's not necessarily activism, but it's the relationship that the self experiences in the

contrast or in the conflict with the environment which itself gives rise to the observation to the object of the art and if you live in a society like eastern germany it's a very interesting a point in history because eastern germany had a weird economy we guaranteed everybody a job we guaranteed everybody health insurance and a pension and a home and it sounds great yes and i think it objectively was great in some sense the productivity was very low because people were not incentivized to work very hard. Because as a result of working very hard, you didn't get better food or a better home, or something like that, right. So people had maintained roughly the same productivity as they did in the 1950s. And it was not the fault

of the individual necessarily, it was the fault of how the entire system was set up. So for instance, the factories in Eastern Germany were communally owned or nationally owned. This means nobody had skin in the game. There was no single individual that stood to profit if the factory was more productive. And if an individual was more productive in a largely unproductive factory, it didn't have a big result on the global outcome of society. It would only have a massive result on the well-being of that individual because it was working very hard without having a good result while everybody else around them was

slacking off. So that was one of the big issues. And I think that ultimately it was the economy that killed the east the ability the inability to set incentives for innovation and this thing that you have need to have a factory where somebody has skin in the game so somebody owns the factory and profits directly from the results of the factory leads to a large inequality and as a result to injustice but we had the control group western germany had this amazing injustice where you have billionaires that own factories and you have lots and lots of people that work for the billionaires and own a fraction of what they do and have a life that is arguably perhaps a fraction as good but the point is the life of these workers in the factories in the west was way better than the life of the workers in the factories in the West was way better than the life of the workers in the factories in the East.

Because the productivity was so much higher. There were better consumer goods and there was better protection of the environment as a result of the better productivity. And there were more civil rights. You could take holidays at grander beaches. You could travel the world and so on and so on. Why does increased productivity go hand in hand with civil rights? Because there is more to go around. You have more room to ask for things if you are better off.

Once you are better off, it's much harder to oppress you because you have alternatives to what is being given to you. And in some sense, once you have a society where the individual is not terrorized into compliance, then this individual will try to take as much freedom as they can get for self-actualization. And so Western Germany gave, in some sense, more room for self-actualization to the working class. But at the same time,

if you were not interested into having goods, if you were not interested in having a big expensive car, if you're not interested in making traveling the world and so on, then Eastern Germany gave more room of self for self actualization to artists, because at a certain baseline, you didn't have to worry about existential issues, you will never see French, right, you wouldn't have to worry about existential issues. You never had an empty fridge, right? You wouldn't have fancy stuff in your fridge, but you would never go hungry. You would never have to be afraid that you wouldn't be able to pay your rent.

Right. So the hardest part, just so I can recapitulate, the hardest part for an artist is to make a living because what you're doing is you're making the art not to eat. You're eating for the art, like you mentioned before. You want to have space so you can do it. Yeah. You want to have space for the non-economical thing the the economical thing is uh the thing that the artist doesn't like the to making art instrumental for something that you can sell

is something that most artists don't like most artists just want to be left alone and do their art and uh of course they need to eat and they need to have some sort of support structure and so in some sense eastern germany uh gave you all these things if you were willing to resist the political pressure and the social pressure of playing along with the system if you were willing to say um you know you cannot do anything to me as long as i have something in the fridge and i just do my art uh that was amazing right so you could uh if you were willing to um, I'm not part of this worker collective, I don't want to have a job in the factory, and so on. You could do what my

father did. And he was something it's how a child of 1968. He bought a house in the countryside, a watermill, because he didn't get along very well with society. He was a nerd, in some sense, without water mill, because he didn't get along very well with society. He was a nerd in some sense without knowing what that was and decided to have his own life, built his own kingdom where he wouldn't have these conflicts with the political reality and the social reality of the society around him and could just have the life that he wanted to,

which was painting and sculptures and whatever crossed his mind. Once you characterized fascism as a superorganism that doesn't care about the individual cell, and if you're not contributing to the whole, then you're excised. And I was curious, what's the difference between that characterization of fascism and communism? Communism is tricky because it didn't really exist, right? Eastern Germany didn't call itself communist. It was real existing socialism. Communism was a utopia that we aspire to have in many, many generations, but it was basically our promise of the afterlife

that justified the present injustices and inaccuracies and mishaps of the system. And so I think that there was a big difference between the socialist country that I lived in and fascism. Fascism defines the value of the individual exactly as its contribution to the group. Which means if you are a disabled person, you should probably die, because your value is not negative. Right? If you are a disabled person, you should probably die because your value is not negative. If you are a person that is not identifying as part of the group, for instance,

if you are Jewish and you have your own community and your own values and you are more cosmopolitan and bound to a cosmopolitan culture and not to the idea of supremacism of white Aryans. You are an enemy, right? You are a diffecist. You are something that lives inside of this superorganism, and you should be removed by its immune response. And so this extreme brutality of fascism that is destroying everything that is not itself,

and that it doesn't perceive as valuable, is unique to fascism that is destroying everything that is not itself and that is it doesn't perceive as valuable is unique to fascism in a way and especially when you do this at an industrial scale if you do this with modernist principles there are other societies of course which do the same thing as fascism does which eradicate all the individuals that do not have the warrior tribe for instance or that eradicate everybody who's a little bit different, but they don't do this at scale. And the socialism was also a modernist society, so it worked at scale,

but it did not eradicate individuals for being disabled or being different. There was eugenicism, but the eugenicism existed at the same scale, the same amount of roughly the same time span as it existed in the West. Right. So in the 1970s, disabled people were often sterilized, because scientists decided, society decided or some group within society decided that they probably shouldn't have offspring because most of

these conditions were heritable, and would create liability in future generations. So the trade-off was sterilized them and something that we think now is immoral. But by and large, you were not being eradicated because you were different. Eastern Germany didn't have gulags as Stalin did. The gulags were, I think,

arguably as bad as the katsets of the fascists. But they targeted people more or less randomly. Like Stalin killed everyone. Hitler killed those that he thought were not detrimental to the state, that were enemies of the state. And Stalin killed people on a whim. There was no safety in Stalinism. There was some safety if you were a proper member of society in German fascism. And in Eastern Germany, it was a rather civilian society.

The number of political prisoners was quite comparable to a number of Western countries. So I would say that in terms of civil rights and so on, it's far inferior to what existed in the West, in Western Germany, but it's still one of the most livable countries in the East. And just by being different, by being an artist who didn't play along, you didn't run the risk to get into prison.

Let's get to some of your PowerPoint slides. Is that okay? Mm-hmm. Okay. Let's see if I can do some screen sharing. Okay, so it says in Architecture of Consciousness. Let's see. There were a few points in here that I wasn't sure exactly about.

Okay, so Construction Process C changes the brain state. So this is the brain state, Xi, at some point. Okay, so construction process C changes the brain state. So this is the brain state, Xi, at some point. Then based on the brain state, changes it from Xi1, minus 1, which is the prior brain state. Cool, okay. Okay, so that's almost like an evolution in physics where you have a state and then you evolve it by one step. Okay, so then you have an attentional process that scans the current brain state and records... Okay, what are the generative parameters of Xi?

So Xi is your brain state, but what are the generative parameters? I think this explanation is way too technical. By now, I think I have a better way of explaining all this than these slides. It's 2017 after all. So imagine that your organism is a control system. So there's a big control agent that is regulating your body temperature and it's moving your limbs about and it's doing all these things in the pursuit of food and social interaction

and all the things that are important to you. And it's basically like a big elephant. And your consciousness is like a monkey sitting on top of that big elephant and is prodding it. And the purpose of consciousness is basically a control model of the attention of the system. And attention exists for learning. So what does the elephant learn? The elephant learns in two ways.

It learns in some sense by repetition and force. So if you just repeat things often enough in the environment of the elephant, the elephant might pick it up. And the other thing what it learns is what it pays attention to. And paying attention to means you single out some features, and you explore the relationship between these features. And these features can be far remote to each other. And they can be very abstractly constructed.

So in some sense, to be able to learn how to dance, you need to relate extremely complicated features in the world. You need to relate the expression of your partner, the music, the movement of your body, the social context in which the dance takes place, and lots of things, right? And to do all these things well, you need to combine them all into a unique gestalt that is very hard to express in a simple specification. So what you need to have is a way of singling out all the aspects that go into the dance and singling out the way in which you have currently related them and how you would need to change them to improve that. And you need to figure out

whether this model of your attention where you should direct it, it was a good model, and whether you should change it. And so you should change what you pay attention to, right? Attention is more fundamental than consciousness. Yes, and it comes before. So consciousness is the expression of attending. If you don't attend, if there is no attention at all, then there can be no consciousness. You can only remember these things as having been conscious that you attended to. Right.

And also, it's not sufficient that you attended to them. You also need to store the things that you attended to in such a way that you can recreate them later in the context of having attended to them. So you basically need to have an attentional protocol that integrates over all these experiences. And so you have an attention agent living inside of the control agent. And this attention agent is basically living inside of your mind. Okay, wait, wait, sorry, sorry. So hold on. So you have a control agent, and that control agent controls the attention? no uh you have a control agent that controls the organism at large the big

elephant okay okay let me you're just you're teaching me right now this is these are office hours there's a control agent then there's attention and the control agent controls the organism yes and the mind is the expression of attention the mind is uh in the way i use the word is basically the uh the software that runs on your brain or that emerges over the activity of your brain so it's the entirety of the mental processes and uh you could think of it as the operating system okay and there's a difference between mind and consciousness yes so the mind is the whole including the unconscious yes okay okay and yes i would say unconscious mind makes sense as a term right so uh most of the correct of what happens in the mind is is not conscious yeah and there's also uh maybe we should

go into this separately but maybe we should do it now there is this very big issue uh what does it mean to be conscious right it's it's unimaginable that a physical system, like clockwork could become conscious, how would the physical system become conscious, something that Leibniz describes, for instance, with a metaphor of a mill, imagine that your mind was like a big mill, of all these moving parts, and so on, and you enlarge it in such a way that you can enter it and you see all these mechanical parts pushing and on, and you enlarge it in such a way that you can enter it and you see all these mechanical parts

pushing and pulling against each other. There is nothing in there over which feelings and emotion and perception consciousness could arise. And this is a very strong intuition that also drives so Chinese room argument that I think is a reformulation of Leibniz intuition pump here. Daniel Zeiss- So what do you think of that experiment, the Chinese room? So I think the difficulty is that it's a category mistake to think that consciousness exists at that level.

So imagine you are trying to build an artificial agent that is conscious in the same way as you and me are. We cannot do this with a physical system, right? A physical system cannot be conscious apparently. So you would have to simulate it, you have to make it as if to pretend to be conscious. How would you do this, you would need friends to have qualia qualia are the features of experience, there are things like colors and sounds and the relationship between them and surfaces and the experience of

information flowing through space and being hindered there are pressure and lightness and heaviness and so on, right? And valence, something feels good or bad. And good or bad means that it forces you to leave that state or attracts you to that state and so on. And all these dimensions, these feature dimensions, we can of course implement as basically geometrical models that we can compute this formal systems of our choice. So there is technical difficulty in how to get all these computations right, but there is nothing mysterious about how to make a system that behaves as if it would see colors, right? It's going to not just measure the wavelengths, but it's going to generalize

the wavelengths in such a way that it normalizes over arbitrary lighting conditions, as for instance, red is something that passes as red under arbitrary lighting conditions on the surface of an object that is red. And that you generalize this over all the red objects, and then you also get the associations to blood and the flag of the working class and to roses and love and heat, right? All the rednesses are now one step away from this abstraction of red that can be translated as a feature dimension

as part of the simulated experience of a virtual system. And so, of course, this experience is not real. There is not really any physical system that experiences anything. It's only being represented as if it was. It's basically like a multimedia story. It's not written in words in a natural language or in a logical language, but it's written in something like a machine learning language. And then this model is being used to drive the behavior of the agent. It's a control model that

is being used to inform what the system should learn and how it should interact with the environment. And as a result, it produces new steps, it produces new model contents. And some of these model contents are, for instance, thoughts. These are not real thoughts. These are as if thoughts, which means they are conceptual and linguistic and ideatic and imagined configurations of such features that give rise to the next set of features, to the next set of thoughts and experiences, right? Again, these are virtual. They're not real. Here is what if the system would feel something? What if there was a person? What if there were social interactions that would matter to the system? What if that would experience red in a way that corresponds to heat and so on, right?

So you build all this into the system and you let it drive behavior, including speech and self-reports, and it gives rise to the next set of things. So this system would be indistinguishable from us, right? Because it's also thinking the same things now. It's producing the same thoughts, it's producing the same story. And it would have emotions and experiences, it would experience phenomena for the same reason that a character in a novel experiences it, because it's being written into the novel by the author. In our case, the author is the mind. And I think the answer to the big

question, how is it that we can be conscious in the physical universe is, we are not in the physical universe, that we only exist in that story, we only exist in the dream that's written by the mind. So this is the way it works, it's virtual, consciousness is virtual. And the experience of realness that we have is not the realness of the physical universe because the physical universe doesn't feel like anything, right? Reality doesn't feel real. What feels real is only simulacrum. Only a simulation can feel real to a simulacrum. property consciousness is a simulated property of a simulated system a physical system cannot be conscious only a simulation can be conscious outside of the dream there is nobody so we're conscious because we're part of the story that the brain tells itself yes we only are conscious inside of the story and we are inside we are characters in a story that the brain is telling itself and we can talk across stories so here we are right now here story that the brain is telling itself. And we can talk across stories.

So here we are. Right. Now, here's where I'm having trouble. Because if I write a story, Lord of the Rings, for example, is Frodo Baggins actually feeling something? Or is he just scribbles on a paper? Just because it's a story doesn't mean it has to have experiences associated with it. So why do we? The question is, what do you mean by is?

Right. So this is the the story it sounds like it sounds like bill clinton it depends upon what the meaning of the word is yes yes so it has to do with the question of uh what's being taken as the uh of of reality i think that a reality that cannot be experienced is very unsatisfying, right? It's not the reality that most people would refer to. The experience of reality is something that is virtual. It's something that's, it's the experience of the VR generated by your mind, right? It's a virtual reality that you inhabit as a non-player character. And the non-player character is generated by the mind as well to describe the interactions of the organism with the world. It's a

story about what the organism does in the world. It's the best story that the mind can come up with. And this story is being used to inform the behavior of the organism. And you cannot break out of the story. That's why the story is real to you. So the thing is to Frodo, his feelings are real in the story. To us, they're not, because we see this from the outside. We see how this thing can be constructed. We can even change the story if we have a pen and paper

and can make him feel something different. But to Frodo himself, it doesn't make sense that his feelings are real to him unless we write into the story that they suddenly don't feel real to him anymore. Right. Give me a scenario where you can write and make a conscious being from your writing of a story. Do you just write Frodo now feels so-and-so and then Frodo actually does feel so-and-so. So it doesn't work with natural language. You have to use a functional language that is basically not just giving rise to a description of what Frodo is doing, but you need to have a functional implementation of an agent in an environment, right? So you need

to have something like a representation of Middle-earth, and you need to have something like a representation of Middle Earth and you need to have a control agent inside of Middle Earth that is being controlled by Frodo's self. And Frodo's self is a model of what Frodo is. And it's a model of basically the affordances of that agent and the state of that agent that is driven by that control model. And so this control model is going to contain thoughts that are the result of Frodo's interaction with Frodo's environment and these thoughts when Frodo is implemented properly will reflect on the needs that this agent has and the needs are in the agent because they're being programmed into it right it behaves according to these needs and if if the model is adequate, and is conforming to

those needs, then this model is representing pleasure and pain in such a way that photo would describe them similar to the way that we would describe pleasure and pain. And he would describe them as his thoughts, because that's the best implementation of its control model that photos mind can come up with. So the language that is being used is not a natural language is not English words or something. It's a functional language. It's a programming language in a way. So can you then take that and simulate a small consciousness?

Let's because we mentioned that it's extremely complex to do something like a human, at least at this, at least in 2020. Can we make a small conscious agent? I think that we can. But if the agent is too small, it's very difficult to ascribe an interesting type of consciousness to it. And the biggest difficulty is if you have a conscious system that is not able to attend to anything meaningful that we can relate to. How can we say that it tends to anything? And in a way, the biggest unsolved problem of artificial intelligence is not consciousness, it's understanding. And understanding means that we map everything

that we perceive onto a unified model of the universe, more or less unified. But it's, we basically explain something by creating a relationship to a unified meaning and the unified meaning is our model of the world and everything that we perceive we are able to integrate into this model of the world and this is something that our ai systems are so far incapable of doing i think we are getting there but yeah is understanding predicated on consciousness or do we not need consciousness for understanding

it depends a little bit on how we would define understanding so to understand something in the it's often seen as a verb that means there is a relationship between the one who understands it and the thing that is being. That means there is a relationship between the one who understands it and the thing that is being understood,

which means you have to have a self-concept and a system that can attend. And the self-concept needs to have a representation of the fact that it's currently attending and what's in the focus of attention. And I think if the question of whether a system is conscious comes down to the question

of whether it has a functional model of what it's attending to and the fact that it's attending to it. So you have to have these two aspects. There's contents of attention and the fact that you are currently aware of them using this attentional process. Once you have that, I think it makes sense to ascribe consciousness to the system. Have you heard of the other theories of consciousness, like the integrated information theory and Daniel Dennett and so on? What do you think of them? Well, let's go IIT. I think that IIT is several things. So it starts out, if you look at the axioms of IIT, with the phenomenologist description of consciousness. It describes what consciousness

feels like from the inside. So for instance, you have this impression of a here and now. And this here and now is distinct from the physical here and now, I think. Something that IIT, as far as I'm aware, doesn't really emphasize, but it's pretty clear that from the perspective of the here and now in consciousness the physical universe is is not in that here and now and cannot be because we often construct the conscious experience after the fact or predictively which means the the here and now of physics is smeared out right and we are able to experience things consciously that don't happen physically, not just because we are simplifying them, but because we feed emerge the features of our

models in a way that is not compatible with the physical universe, but that is useful to the control of the physical universe. So basically, the contents of our consciousness are determined by what makes up useful control model, not by what's physically possible and what's physically happening. The other aspect of IIT is its denial of functionalism. So IIT in some sense makes metaphysical assumptions. And these metaphysical assumptions, I suspect, amount to panpsychism,

which means that consciousness is, in some sense, inseparable from matter or from the background of the universe. Therefore, it must be an intrinsic property of matter itself. You could say, imagine you cannot determine what color is. And if you look with a microscope, you cannot see what color is made of. Color doesn't have components, so color must be an intrinsic part of matter.

Color is not made by matter. It's inseparable from matter. Every matter contains color, right? Right. This is almost correct, but it's not because if you zoom in at a certain point, there is no more color. Color only makes sense as a kind of interpretation of what we sense about matter. And I think that's also true for consciousness.

Consciousness only exists within minds, not within the physical universe. And so in some sense, IIT, I think, is putting the locus of consciousness into a domain where it doesn't belong. And then there are the technical aspects of the implementation of the IIT theory, IIT, that is this factor phi, the measure of integration. And there are some good aspects about this. So in the sense that your own neocortex is integrating information in such a way that when a lot of it is synchronized and is stored in an intentional protocol, then you will probably have a larger focus of your attentional awareness than what you have when your consciousness is highly fragmented and your

this will be reflected in the fragmentation of the cortical contents, which means that a lot of it is not firing in synchrony. And this is being described to some degree by phi. But this is a relatively small aspect of phi. And I think if you go deeper and try to make more out of phi, then it falls apart because it's no longer necessary in sufficient condition. You don't know what gradual states of phi mean and so on. So I don't think that I buy IIT at this point. It's not explaining

how this phenomenon comes about. And it's descriptive. And it has metaphysics that cannot be evidenced that are not predictive. The next thing was Donald Hoffman. I buy Donald Hoffman's first part of the theory, which is typically, the world looks nothing like what we experience it as. That is obviously the case. And the second half I don't follow. It depends on which way you look at what he writes and spells out. So the second part to me is the thing that computers are an inadequate representation of what our brains are doing. Our brains are not computers, they are something else. And I think this results from a misunderstanding of what a computer is.

So he's driving intuition here that a computer is a digital von Neumann machine similar to a PC. intuition here that a computer is a digital von Neumann machine similar to a PC. Right, right, right. And our brain is not organized in the same way as our personal computers are. Our brains are self-organizing systems, they are oscillators, and they use very different ways of translating the information between the different parts and achieving coherence and so on than our digital computers are. But this doesn't mean that they cannot be described as Turing machines. The Turing machine is too general for that thing.

Basically, you're looking at a finite state machine that can be used to compute representations and execute control functions. And the brain is clearly in that category. And what we can show mathematically is that the different instances of that category are equivalent to each other, which means there are mappings. You can take a digital computer and create a simulation of the brain, a virtual brain in that digital computer that behaves in the same way in principle. And also in principle, vice versa. So you can create a simulation of a digital computer in a brain. behaves in the same way in principle. And also in principle vice versa. So you can create a simulation of a digital computer in a brain, but it's much, much more

difficult because our brain is largely indeterministic. So it's hard to get enough determinism from the brain to keep a model of a large digital computer stable in it. And so we can only run relatively small digital computers in our brain. But using external tools, for instance, if you're allowed to use pen and paper, we can use our chaotic and antagonistic brains to run pretty large simulations of digital computers, but externalizing the memory and determinism. What about Douglas Hofstadter's idea of consciousness, the strange loop? I'm not sure if you're familiar with it,

but what do you think of it? And do you mind explaining to the audience the idea as far as you understand it? I don't think that I can adequately explain it. To me, Hofstadter is bound to a tradition of computer science that is taking Gödel as talking about the property of the languages that he still tried to use. So he's not a proper computationalist. He is using classical semantics to describe computational system and it leads to

contradictions in his descriptions. And I think that the strange loop is in the class of these contradictions. And I think that our consciousness is not strange as a loop. There is some loop going on, but it's not a strange loop at all. It's a loop that goes between the contents of the attention and paying attention to the fact that we still pay attention. So we notice that we haven't drifted off. So the strange loop, sorry, the strange loop doesn't exist. There's no such thing as a strange

loop because it's predicated on false mathematics or mathematics that doesn't apply. Yes, but I am not sure if I represented adequately and I would have to reread it and properly formulate it. So I am reluctant to talk about it because I cannot properly translate and define it here. I would have to look it up again. Yeah, no problem. How about the holonomic theory, the holonomic brain theory? Oh, it's a long time that I stumbled into this one. I didn't quite understand how it was explaining consciousness. So there is a certain way in which it is pointing at the hierarchies of perception that we have, and that's accurate.

But I haven't understood how the holonomic theory explains consciousness, in the sense that I didn't see a specification that I can implement and end up with a conscious system. How about Daniel Dennett's idea of consciousness? Or at least his explanation as to how consciousness arises and what it is. I think that Dennett, as far as I understand Dennett, has nothing wrong with what Dennett says ever. The main thing is that the things that I read from him, I don't have any objections except that he seems to ignore the part that most read from him, I don't have any objections

except that he seems to ignore the part that most people struggle with which is phenomenal experience. So basically he does make fun of the people that explain how Mary proves that machines cannot be conscious and he's justified in making fun of them. And it's a very good read, but he is not going to convince any of these people because he is not deconstructing the thing that they are struggling with, which is phenomenal experience.

He's mostly seems to be ignoring it. So it's incomplete. It's correct, but incomplete. I, yeah, I suspect that he doesn't think that there's that much to explain because he may not have that

much phenomenal experience. He probably has aphantasia. I'm only speculating here. It could be that basically, Daniel's mind is so conceptual that he doesn't think that there is that much to explain. And there are people which are rarely visiting the conceptual realm, and they have the experience that this logical language is unfit to describe anything of consequence in real everyday life so how could it explain something that is so fundamental as a phenomenal experience right you need art to describe that and uh danny is so far removed in the way that he speaks, talks, things, operates, that his own mind operates, that it's very hard, I think, to see for the phenomenologists how Dennett is actually talking

about the same thing. But having said that, I do think that Dennett is right. It's just he is not giving us a specification that we can implement at this point. And again, he's a philosopher, so he's not dealing with the specifications very much. And so when I read Dennett, I don't object to anything that I read so far, but I also learned very little. It's basically trivially true what he says. How about the sensory motor theory? So what does it mean to have a sensory motor theory of consciousness? I suspect that there are a number of people that would refer to sensory motor integration as at the core of consciousness. And you could say that there is crucially a notion of agency that results from performing actions and sensing the result of the actions and making sense of the relationship between actions and perception. When I was confronted with this notion, I thought, well, how do you know that there is action?

What exists is eventually just a notion of perception. Have you heard of Bergson's theory of consciousness? Wait, let's go. Yeah, sorry, sorry. Let's continue with the sensory motor thing, if you're still willing to go. Yeah, no, no, I apologize. I took your pauses as you were finished.

No, no. My bad. By the way, just in case, I'm a bit loopy. I haven't had enough, I haven't had sleep, proper sleep in like three days, and I've only eaten one meal in 72 hours, approximately, because I'm in the middle of a fast. So, if I seem a bit drowsy or slow, please forgive me. I have no difficulty forgiving you. Okay, why don't you continue about the sensory motor? I have no difficulty forgiving you.

Okay, why don't you continue about the sensory motor? So I suspect that we discover our body, the motor aspect, only through the senses. So it's basically, we discover the control architecture. We discover that there is a relationship between the environment and certain states, which we notice as being intentional states, and the body that is the instrument of translating intentional states into changes in the environment. And this control hierarchy of intention originating in motivation and needs, and then this leading to the initiation of motor actions and the initiation of motor actions leading to physical actions or mental operations, and then as a result to changes in the world, the perceived external world or then the perceived mental world. And this again leading to changes in our needs and motivations and so on.

This discovery of the loop is only possible because we have the entire loop given. This discovery of the loop is only possible because we have the entire loop given. If any element of this loop was missing, we would not be able to discover our own agency. We only discover our bodies as being that instrument. We wouldn't discover that we have a body outside of volitional states and environment. We wouldn't discover volitional states in the absence of a body and the environment. And the body could also be a mental body in which it is able to perform mathematical operations. That's probably sufficient for this. But you will need to see some outcome of your actions and some way to affect these outcomes to arrive at the notion of your own agency.

affect these outcomes to arrive at the notion of your own agency. And the type of intelligence that makes us distinct is the ability to conceptualize ourselves as an agent embedded into the universe. Our, the generality of our intelligence is given by us having to solve control problems that are so general that we need to model ourselves, that we need to reverse engineer us, right? Imagine you start out with the thermostat. And a thermostat is a system that controls the temperature in a room based on a measurement of the temperature and a control impulse that turns a heating element on and off.

And now imagine that this thermometer is very close to the heating element. So there is some feedback between the heating element and the measurement that you make. And if you want to get the temperature in the room right and don't want to run into wild fluctuations, you might need to have a second order control loop. And the second order control loop is basically correcting the measurement that you're making with your sensor for the activity of the heating element. that you're making with your sensor for the activity of the heating element. Which means now your second-order control loop will have to implement a model of the interaction between heating element and heater.

And it might have to implement a model of the temperature of the heater itself and how much this contributes to the temperature in the vicinity. And the second-order control group is what's conscious in this example? No, it's just a second order model. No, it's so basically, we are now looking at a nested system, a nested cybernetic loop. And if you have a room where you, for instance, have a changing volume of air, because you sometimes open the window or not, you might need to have a third order control loop that is now measuring how the heating element is changing the temperature of the room, depending on that third hidden variable.

And you try to guess at this hidden variable. And now if you also have temperature fluctuations on the outside, because maybe you have a change of season and the air that comes in and out, now that might require more complicated loops, right? And eventually you will also need to have a model that describes the sensitivity of the heating sensor and the inaccuracies of the heating sensor. And maybe at certain temperatures, the switch doesn't operate at the same rate as it would happen at your default temperature, right? So you need to allow for the quirks of your own control architecture.

And this means that at some point, the control loops have to model the system itself. So you get to a system that is modeling its own place in the environment, its own relationship to the environment. Consciousness is not yet related to this. Consciousness is a tool to discover this. Imagine you have this vast multitude of possible measurements and possible hidden states that

you cannot directly measure, but you have to construct to explain the data that you're measuring and the relationship between them. And because you have this vast conflagration of possible relationships between them, you cannot just do a blind search. Instead, you need to have some kind of a directed search, something that is structuring your search and telling you which of these parameters you should single out and relate to each other

and try to change the relationship, see what the outcome is, and so on. And that is the purpose of consciousness. It's the direction of attention over this multitude of possible states. Okay, so the purpose of consciousness is the direction of attention over this multitude of possible states. Okay, so the purpose of consciousness is the direction of attention. But also at the same time, you said that it was the story that the brain tells itself. So in this thermostat example, what is the story that's being told?

So the story that could be told about the thermostat would be that there is a thermostat or a measuring system that is regulating a parameter in the world that relates to other parameters in the world, for instance, volume of air, outside temperature, frequency at which a window is being opened, maybe other agents which open and close the windows, maybe other agents that change my settings depending on what's happening and so on, right? And the more data you can integrate, the more complete your model of the universe will get, the more parameters of the universe you will have to integrate. And at some point, you'll get to a model that is complete in the sense that you

have a smaller set of causal laws, basically physical laws that are sufficient to explain the conditions of your existence. Right, the basic principle that explains how everything relates to everything, how functions are being computed, how a system can exist in the universe, how a universe can be generated, and how a system can change itself as a relationship to that that is computation. So computation gives you a language to talk about first principles.

And what is intelligence in this? Is it just the ability to model for the generation of more models, whether or not they're predictive or whether or not they fit? Yeah, what's called intelligence, the ability to make models. But of course, this makes sense in the context of a control task usually. So there is a certain control task that is being fulfilled. And the intelligence of the thermostat would be measured by the complexity of the models that the thermostat can build as a result of the interaction between the data and the measurements and its interactions. Is there a relationship between intelligence and consciousness such that the higher your IQ as an individual, let's say as a human, you are more conscious? I would say that consciousness is a solution for creating certain types of models. So it's basically an aspect of a certain class of algorithms, of algorithms that require the direction of attention in a particular way. And so the fact that we can recall having attended to something while being

something that was attending and while being aware of the fact that we were attending instead of drifting off. This is what determines our type of consciousness. And it's not necessarily the case that every system has all these degrees of freedom or needs to have these degrees of freedom to be able to solve the same problems as we do, because we could just fixate these things. Maybe we can have an attentional system that is itself unconscious. When you say to solve the problems, are you referring to just propagation of existence, that they don't die?

No, our death is unrelated to this. Our death is a concept that exists. It's an idea that our world line is broken in a certain way, that it ends. And we are not continuous in the first place. Our idea of existing in a continuous fashion is a construction that we have over our memories to explain memories that we seem to have of past states. Where does free will come in?

Free will is a representation, it's in the system that it's made a decision and the decision is being made on the best understanding of what's correct. And free will is basically the outflow of this control task. It's the outflow of the control algorithm being executed in the right way. The opposite to free will is not determinism. If you are indeterministic, you cannot have free will. If you behave randomly, there is no will involved, right? It's just random.

And the opposite to free will is also not coercion, because you are deciding that you are giving in to the coercion. You wouldn't need to be coerced if you wouldn't have a degree of freedom. But the opposite to free will is compulsion. It's basically when you do something despite knowing better. The opposite of free will is compulsion as well as randomness? Randomness is the absence of will at all, right? The system that is random has no will.

So the will cannot be free or not. But so we have to look at the opposite of the freedom. And the opposite of the freedom is not the coercion, it's the compulsion. What's the difference? It's when the system, the compulsion means that you have a model of what you should be doing, but you don't find yourself acting on it.

You find yourself acting on something else you're acting on based on some impulse or some addiction. And that is basically the true impingement on your freedom. But it's important to realize that freedom is not an absolute notion in the physical sense. It's, it's a reference that we make to certain internal states. So when I refer to my own decisions as being the result of my free will, it depends on the context in which I use this. And when I talk about the experiential context, I experience my will as free when I have the impression that I made the decision based on parameters that are the right ones,

that are in the proper order with respect to the control structures that my mind currently implements. And not because of some glitch in the matrix, of some glitch in the system that implements me, or of some erroneous programming or some external force that is spreading in my mind. So when people have the impression that they act out of a compulsion,

for instance, because they, say, for instance, have anorexia, they might decide to, or bulimia, they might decide not to throw up after eating, but they cannot help themselves. They just have this enormous urge to throw up or make themselves throw up. There's nothing that they can do about this. And it's a very disturbing experience, because it impinges on your freedom, there's one thing that you want

to do. And another thing that you find yourself to be doing. And this is a very big existential disturbance that happens in that case. Okay, so freedom is like, you have a model, then you execute on the model based on the parameters, and it's salutary, and it's positive. What does positive mean? That it fits your goals. It typically does, right?

So imagine you have your Frodo in your Middle-Earth world, and it's a story. And we imagine we implement this as a computer simulation like a minecraft middle earth and you have your frodo agent in there and the frodo agent is acting based on models that frodo is creating then uh frodo would uh probably conceptualize his actions as being the result of his own free will if uh he has the impression that everything happened in the way that it was supposed to in his own mind. That is, he is perceiving certain things, there are certain things he wants as a result of his physiological, social and cognitive needs, and spiritual transcendental needs, maybe,

which I think may be understood as a class of social needs. And as a result, he is doing certain things, he's making certain decisions, because they increase the likelihood that he is reaping the anticipated rewards with respect to his needs. And if this all happens in the proper order, then his mind will represent, I wanted this. The intention is being represented. And I wanted this because of a mechanism that was only determined by what I needed and what I consider to be the right thing, which defines my own freedom.

So it's in some sense a paradox. The more you know what you ought to be doing, the more agency you have, and the more freedom you have subjectively, but the fewer degrees of freedom you have. And the less you know what you are doing, the more degrees of freedom you have, but the less do your actions mean anything, which means you have less objective freedom or because you have less will.

So is free will a story that we tell ourselves? Part of that? It's a model, right? In a sense, it's a story that we tell ourselves. But it's not we who do this. It's the mind who tells it to the self. It's upstream from the self. Your mind cannot control

what experiences at its own will except in certain states at one point you mentioned that the dalai lama can effectively live forever in the sense that he identifies with the government and as long as that government is instantiated and not dissolved then he lives in some way shape or form okay i didn't quite get that do you mind explaining what do you mean that he identifies with so-and-so so uh what i mean is not this i said i meant as so uh most of us identify as a person in the sense that we live for a certain time span uh we have certain organismic needs. We have a physiology. We have social relationships to our environment. We have relationships that we serve. We have a greater whole that we serve

that gives rise to our spirituality and so on. And all these things define what we try to keep stable, what we perpetuate, the thing that we try to control, the control system that we are for. This is where we are the thermostat for, right? All these dimensions of needs. A few hundred physiological needs, a dozen social needs, a handful of cognitive needs. And keeping all these in balance gives rise to our identification. The identification is a result of us making models of how these needs relate. And so we create a hierarchy of purposes. The needs themselves are not sufficient. We need to have a model of what is going to give us pleasure and pain. And this is what we would call a purpose.

And the purposes need to be compatible with each other. And this hierarchy of purposes that we end up with is, in some sense, our soul. It's who we are. Or what we think we are. What we think of as ourselves. And can we change this hierarchy of purpose yes of course we can we do our in our course of our life it changes so for instance um for most people it changes radically when they have children right and what i mean is can we consciously

direct it yes it's not like it's the mind there's something behind us that's producing us and we're just players in this game yes we have the feeling that's producing us. And we're just players in this game. We have a feeling that we're controlling it, but we're actually just being told what to do. So we can control it in such a way that we identify pathways in which the models that are being created in the self or as contents of the self inform future behavior. And of course, the self itself is not an agent. It's a model of that. But you can experience that from the level at which your self is constituted, you can change the identification of the self. This is basically Keegan level five, where an agent gets agency, not just over the way it constructs its beliefs, but also agency over the way

an agent constructs its identification. And colloquially, we talk about these states as ones of enlightenment, because we realize that the way things appear to us, that these appearances are representations. Things are not objectively good or bad, but that there is a choice that happens at some level in the mind, whether these things are being experienced as good or bad, and that we are responsible for our reactions to things. And the way that we react to things is instrumental to higher level goals that we might have. And once this happens, we can learn a number of techniques

in which we change how things appear to us. So for instance, when you do the dishes, you might find it horrible to do the dishes because it takes time away from you. It makes your fingers wet and sticky and it's annoying and so on. You could also realize it's time out for you where you just do a very simple physical task that itself is pleasant because it's nice and warm on your hands your body doesn't hurt while you do it and you get some time to contemplate and you need to do it anyway and you can turn this into a time that you enjoy right and you can get agency over the simple thing so this sounds like in self-development where they would say, just reframe your problems into something positive.

So let's say you have to run. You hate running. You just say, well, I'm doing something that's good for my body. I like it. Yeah. So the question is, are you just telling yourself a different story consciously or do you experience the story as being different. And so the intended result is that something happens upstream of your experience, which means you now suddenly experience doing the dishes as pleasant, intrinsically pleasant, it's not just you're

talking yourself into some kind of delusion that makes you pretend that you like, right? So how do you cross that barrier? Because if you just tell yourself i like this task i like this task even though you hate it you feel like you're being self-deceptive and it doesn't work so how do you actually get it so that you experience positive emotion from it in that case it's super simple you just focus on those aspects of the task that are that for instance contain sensory pleasure and there is And the aesthetic pleasure of being able to follow your own thoughts where you do something that does not bind your attention very much and is not directed on, say, work goals or family goals or something else.

So you can enjoy the mental freedom that you get and you can enjoy the pleasant aspects of the sensation of the warm water and the soap and the movement of the hands and the softness of the cloth that you use for cleaning and the hardness and of the things that you are cleaning and so on and the sense of cleanliness that you are creating in the world and the aesthetics that are involved in that process in the the same way, if you don't want to do the dishes because it takes attention from you, you can focus on the negative aspects. And by emphasizing this in your attention,

you basically put a spotlight on this part or that part of reality, and you emphasize the parts that you experience in there, right? So you can get pleasure, aesthetic and sensory pleasure from a task and you can get sensory horrors from it and aesthetic displeasure from the same task if you focus on different aspects of it. If it's a matter of changing one's focus

from the negative to the positive, how can we seldom do that? If it's so positive, I mean, if it's so net positive to look at a task and just focus on what's bringing you sensory pleasure, why don't we do that? I suspect that we don't have intrinsic attention on this for

the most part, because it would not be useful if you protect ourselves in this way. Maybe there is a reason why we don't like doing the dishes or we like doing the dishes that we are not wise enough to discover. And if we could just reprogram our reaction to things before we understand that reason, maybe that would be premature and we would end up in the local optimum in the way that we organize our life,

where we end up being a dishwasher when we should instead be a lover or an artist or an explorer or an intellectual worker, right? So maybe it's too early to reprogram your experience before you know what you're actually doing. I see. So you have to understand yourself because there could be an evolutionary reason for why. Yeah. I suspect evolution would have given us the ability to reframe our experiences fundamentally if that would have been useful. And the fact that it's not is if you cheat yourselves into experiencing whatever you

do as pleasant too early, it might make you very happy, but also dysfunctional. You also mentioned once that your theory of consciousness is something that we intuitively know and that when you tell people, they had a suspicion. And for some people, when they stumble upon this insight that you also elucidate, that they get depersonalization disorder. It can go two ways, where they feel liberated, or they feel distance from their body and it's net wretched and ruinous for them. What do you say to those people who don who feel disidentified from their, who don't feel identified with their body, and it's not a positive experience? What advice would you give them? I would advise them

to go to a real therapist, because I'm sadly not a therapist. I'm not competent of doing this. Don't listen to me. If it gives weird experiences in your body, I have cannot take responsibility for what's happening to you. So I'm just a cognitive scientist. That would be super dangerous, I think, and irresponsible if I just try to create weird experiences in your body as a result of my theories. So I would say this is a side effect. And it might be a side effect of you trying to answer a question that I have myself. And me answering this question doesn't lead to weird experiences in my body. I have weird experiences in my body because I exist, right? Existing is

weird. And I want to explain why is it possible that I exist, that I experience, that I have a body? Why is it that this body sometimes feels very big and unwieldy? Why is it sometimes extremely small? Why does it sometimes disappear? How is it possible that I can have an out-of-body experience? I want to explain all these things, and I find plausible explanations that all make sense and are much more logical than the inverse of these explanations, right? So I find them helpful. And what I can offer people is suggested solutions to similar questions that I have. And I might be wrong with my answers. They're the best answers that I can give at this point in my life.

And they are compatible with the answers that most of the other thinkers give that have looked into this. They're also largely coextensional with the answers that a lot of meditators give. looked into this. They are also largely co-extensional with the answers that a lot of meditators give. It's only that they are using a language that seems to be incompatible with the language that we have established since the Enlightenment in our own culture.

It's basically there is a disconnect between what we experience as being real and how we talk about reality in our culture. And this makes it so hard to make sense of consciousness and feelings and phenomenal experience and identity and transcendence and so on. And the goal that I have when I give these explanations to myself and others

is to explain how we can get what we experience as being real and what we also observe as happening around us in our interaction with the environment and the way that we reason about reality how we can put this into one how we can construct a metaphysics that is compatible with our scientific worldview and that allows us to make sense of what we experience. What about people who fear losing control? That as soon as they have this insight that the mind is just, that you are just a story within your own mind,

that you don't have control. Now you say, well, we have control, but I'm unclear as to how, as to who has control. It seems like the mind has control and there's some relationship between the story that you're in and what's telling the story i don't see that connection so they never had control in the first place yes and i mean you see this every day right there are things that you

do that you would prefer you wouldn't do at least most of us are in that state and unless there are perfect sages in a daoist tradition where everything that they do is in a complete union with their perception of what the universe needs to have done at that point. And most of us don't get to this point. It's very hard to cultivate your mind to this union where you have an identity between you perceive what needs to be done and what you do. And this discrepancy is something that we have to explain. And we can only explain it as us not being in control. And meditators describe this as the monkeys trying to

prod the elephant and the elephant just walking its way as it wants to. And sometimes it's aligned with what the monkey sees and thinks is important. But by and large, the elephant follows its own wisdom. Do you have any advice for someone who finds that disconcerting, that the elephant is more in control? I think it's try to not take the monkey that seriously and try to sense what the elephant is doing and realize that it's a much, much larger dance than the dance of the monkey on top of the elephant.

And it's also a larger dance than the elephant itself. It's the entire forest, right? And the elephant being part of it and being in resonance with it, interacting with it. And there's only very few decisions in proportion to the entire thing that is happening that can actually be controlled. And they can only be controlled in certain ways and what you can explore is in which ways you can create a coherence between what you perceive that needs to be done the global aesthetics of the universe that you prefer the way the forest should look like and this is a part of your task is to figuring out these aesthetics how can this universe be coherent and consistent and as a result beautiful and what is the things that can be controlled as part of it locally and you are the

result of that you are not the cause of that you are the result of the local control you're not the this thing that causes the local control now the local control you mentioned intuition earlier that it's important to follow your intuition and i assume that intuition a component of that is your conscience your heart is that a part of the elephant so in some sense you should stop trying to direct the elephant and follow the elephant as the monkey that you need to do both but ideally you want to have a state where the elephant is treating the monkey as one of its most useful tools. But of course not its only one. And the monkey needs to be able to shut up from time to time.

And it can give the elephant feedback, especially when the elephant needs it. the elephant feedback, especially when the elephant needs it. And so in some sense, there should be a friendship between the monkey and the elephant in the sense that the monkey waits until it has its task and its time and the elephant is actually asking it for something. The purpose of reason and analytical thinking is to repair perception. And when I mean intuition, I mean perception. What do you mean that the purpose is to repair perception? perception is the part of our mind that is integrating information in a way that is not linguistic and not conceptual. It means the integration is not discrete. It's an integration

that happens over many, many features, often in an irreducible way, or in a way that we don't yet understand. Imagine the way that you integrate information when you try to catch a ball. You see how the appearance of the ball changes in your visual field, and as a result, you learn how to move to catch it. And if you try to do this analytically, if you try to compute the model with the capacity that your mind has you're not going to be as effective than when you are using in perceptual model which we would call intuitive right you train your intuition of what movements you should be making to get the ball and when this is systematically not working okay i said you can use your reason

to figure out what's going on okay or. Or when you are already very good, and you try to figure out is there a way in which I could do it better, right, then you can use your reason to construct a system that is measuring your movements and using camera and optimizing your technique or a simpler thing like finding a better trainer. Right, right, right. Okay, so there's a mismatch between what you want and then what you get. And the repairing is yes is what reason is for it's basically for dealing with these edge cases

is there a way to falsify or test your model of consciousness i think that there is a way to test it in the sense that we can at some point build a system that will explain that it's conscious to others, that that would be the ultimate proof to itself, and that there would be nothing left to be explained. Wait, wait, sorry, sorry. If it just says that it's conscious, then it's conscious? So the question is whether you build it in such a way that it cheats. So you could, of course, make a chatbot

that pretends that it's conscious without being conscious. But this would mean that at some point you will see a functional difference. There will be a difference between the behavior of a conscious system and the behavior of a system that is not conscious. And I think currently that the difference that you would observe is that the system does not have a control model of its own attention. It's not aware of the fact that it's attending and what it is attending to. So, for instance, the question, is a cat conscious, I think is a decidable question.

It's a question that comes down to whether the cat can be best explained as being aware of what it attends to. And based on this criterion, I would say cats are clearly conscious. And if you look at a sleepwalker, a sleepwalker is a person that is unaware of what they attend to. They can attend to things, but they don't know that they do this. And as a result, they cannot question their actions. They cannot redirect their attention. They behave in a way like an automaton, because this attention loop is missing that would be able to reflect on what they are doing

and learn something from that. Yes, but they cannot learn. They cannot change their behavior as a result of reflecting on the interaction with their environment. They cannot direct their attention in this sense for this attention learning. They can perform all the automatic autonomous behaviors that the elephant has been trained into. There are perceptions taking place, right?

They can open a door. They can even make dinner. But they are unable to learn something. So they're able to coordinate actions. But it's like an orchestra without a conductor. So capacity to learn as well as attention is what's required for consciousness. And somehow this becomes a test of consciousness that you can falsify it from this. I think that the ability to learn is neither necessary nor sufficient.

You have people that are conscious and that have lost the ability to learn. And you have systems that can learn and clearly they're not conscious. learn. And you have systems that can learn, clearly, they're not conscious. But I think the purpose why we have this attention is largely to enable us to do a targeted recall of index memories for the purpose of learning. Is it ever possible to get a continuous perception from a discrete phenomenon? So what I mean is, let's say we're just bits, zeros and ones, and it's binary

discrete. Yet we perceive continuity, smoothness. Okay. How can smoothness come about from discreteness? The trick that our brain is using, because our neurons don't act continuously, right? The neurons involved tend to fire at rates of like 20 hertz. How is it possible that we see a continuous movement? And the trick that our brain seems to be using is that it uses key frames and vectors that tell it how to compute the next key frame. And you can see some evidence for that in two ways.

One is there are optical illusions where you have a static image that seems to move on the page. And if you have such an optical illusion, it shows that there is a difference between the appearance of movement and the change of location. If something was moving continuously, you would expect it to change location. But if something can move without changing location, it means that your brain is representing the movement separate from the change in location. The change in location is the difference in the keyframe. There is only one frame, right?

You don't look at different keyframes when you look at a static printout of an optical illusion that moves. If you see it moving, it's because you only perceive the vector of movement. This means it's a static representation that applies to the single frame and tells you where you would expect the thing to be if it was a changing location. And the second evidence is that there are people which have brain lesions that lead to a stroboscopic representation of reality, which means they only perceive the keyframes, but not the movement between the keyframes. Okay, so let's forget about external sensory experience. And what if you close your eyes and you visualize in your mind's eye a circle? So you see that as smooth. Now, are we just wrong in our perception? The circle is actually not smooth. We tricked ourselves somehow. It's actually jagged like pixels. If we were to zoom in in how is it that we can get smoothness even

internally i would have to look at a real circle because i have a fantasia so i have a circular light up here and i can see it as a smooth in the sense but the smoothness is mostly the absence of a detectable non-smoothness right so basically I can use a function that describes the progression of the line using this smooth circle and I don't notice features that go away from that simple function. And I would have a more complicated function to describe an object that has jaggies. So in some sense the smoothness is a decision surface between features. It tells me where to expect more sensory data. It tells me where to expect certain blips on the retina or on my mental retina, so to speak, on my mental stage

when I imagine that object. So it's basically some kind of generator function that tells me in which way I expect the features to fall. But it's not something where I can only see the jaggies if I expect to see jaggies. And this can get confirmed in some way, right? And the jaggy function, the one where I see aliasing or corners and so on, needs to have an explicit representation in a certain way. If that representation is absent, then I will only see the smooth surface. How has your view of consciousness changed in the past few years, let's say four years? I think that I focus, it's basically a shift in focus,

it's the shift that goes away from the phenomenon experience itself, to what gives rise to the family experience, and especially the way attention is implemented as in as opposed to control in general. And so I would say that I get closer to an implementation, my view of the phenomenology of consciousness, I don't think has changed in the last few years. So my phenomenology of consciousness is the result of observing consciousness, of zooming in at different layers of resolution and observing altered states, for instance, the dreaming states at night and when I do lucid dreaming or the hypnagogic state between dreaming and

waking in the morning and so on. Do you practice lucid dreaming? I did this in the past, but I don't do it systematically anymore. I suspect it's not functional for the brain because it's in some sense like inducing a trip in your brain, similar to taking drugs, because you are forming long-term memories of things that you are not meant to form long-term memories about. Basically, there are several modes of learning. One is a simple conceptual learning, where your perception doesn't change, but the way you relate your perception is changing.

And there is another one where you change the construction of reality itself, the way that you construe reality. And if you go to this level, if you also change the way you relate to the environments you need to deconstruct or suspend yourself and your agency and the boundary between self and universe and i suspect that's one of the reasons why we have these dream states in which we don't react interact with reality itself right i remember you saying that dreams might be something akin to

generative adversarial networks and i'm curious curious to explore that. What do you mean? How does that come about? So in some sense, we are producing hypothetical realities that can predict sensory patterns. And we have a system that acts as a discriminator that tunes these generations of degenerative functions to see whether they are able to explain sensory data. And the most important discriminator is your perceptual apparatus that is connected to your sensory input, that is your retina, your cochlea, and so on. So by and large, your thalamus, which is the big switchboard that connects the different brain regions and your sensory input. So in some sense, your imagination is being used to predict sensory data.

And the set of functions that is closely predicting the next batch of sensory data is what you experience as reality. But it's a dream. Every experience of reality is in some sense a dream state. And the dream states at night are different from the dream states during the day, mostly in two aspects one is in if you unless you do lucid dreaming you don't have a consistent sense of agency which means you cannot recall who you are and you cannot really direct your attention in any way there is no subject involved there might be a story about a subject

but the subject is not doing things that can be controlled by the subject. And in a lucid dream, you bring this agent online, you gain a sense of agency and a sense of control, it can direct your attention according to control parameters, right? This means now you have a system that is exerting control based on the expectation of maximizing some kind of reward in some of the dimensions that your mind cares about. But at night, the second thing that happens beside the agency is you are no longer in touch with your sensory apparatus. So you have no way to access what's happening on your retina anymore. You mean the external world? Yes. So everything that plays out in your cortex is now originating in your cortex, place out in your cortex, is now originating in your cortex. There is some

slight interference with the sensation of color or whether you need to urinate or smell. Smell translates relatively well into dreams for most people. But by and large, you cannot sense what happens in the outside world. And this is not going to enter your dreams in any consistent way. Instead, you are only going to use your dreams in any consistent way. Instead, you are only going to use your mental representations to make sense of other mental representations. If dreams are for learning, why is it that we don't remember them? Why does it go away? It's largely because dreams play out as situations of things that never happened. But we do that all the time in our own head

when we're just thinking about speaking to someone like a boss, like, what am I going to say to that boss? How do I get a raise? How do I not get fired? Yes, but all these things are prefaced as this doesn't happen. This is an imagination of an imaginary situation. I'm playing out the following things. These things will happen in reality in the following way.

And then you can compare them with reality and you can use this to tune your imagination to make it better next time. Whereas in a dream, your construction of reality itself is changed. For instance, you see objects from perspectives that you've never seen them from. You might have a flying dream as a result, right?

You see the world from a top-down perspective. And as a child, I think many children have flying dreams for that reason that basically your brain is generating new perspectives of known objects. And so you can recognize them from these new perspectives. That's, it's very useful, but it's not very useful to remember that you can fly because you can't. Do you know of any studies that have been done about people who

can recall their dreams versus people who can't and if they report higher life satisfaction if any of those groups no i'm unaware of that so i i don't know how if people that we can recall their dreams are happier than people that can't and i suspect that it should be possible to change the equilibrium of most people that cannot recall their dreams in such a way that they can or to wake them up at the right moment and that they will be able to remember their dreams if you wake them up during REM phases. But I'm unaware of these studies. I'm not a sleep scientist. If all we did was dream, is that real? Well, all we do is dream in a way, right?

So every perception of reality is a trance state. It's a dream state. There is no reality that can be sensed. It's only this VR that you are entranced to believing that it's real. It's a movie that your mind is showing to the self, and the self is recording in some sense what happens at its boundary with this attentional protocol.

And we can partially recreate these binding states later on as the memories of states that you think you have been conscious of. And this is all there is. This is only this dream. Let's end this on a positive note with you saying,

with you telling people, like almost instructionally, how is it that from your insights, from what you've said, how do you get from that to then changing your mind so that you experience positive emotion or at least a negation of negative emotion? There's none, an absence of it. I don't think that you should sort emotions into positive and negative ones. I think that you should look whether your emotions are helpful or unhelpful. And you should have the most appropriate and helpful emotions that you can have,

not the most positive emotions that you can have. The purpose of life is not to be happy in the sense that you should be in a state of constant bliss. You should be able to achieve the things that matter to you. And the emotions help you for that. So you should check whether your emotions, for instance, express ruminations, which means you might be caught in a loop that is unproductive and you're just veering a groove in your mind rather than making progress. You should see whether you are suffering,

which is usually the result of you trying to change something that you cannot actually control, at least not in the way that you're currently trying it. Right? So this is what you should be monitoring, you should monitor the trajectory of your emotions, and see whether they are still helpful. But they're your tools. And just turning your tools into something that only gives you one site is not helpful. So how do you control your tools to make them helpful? It seems like you do this. Or do you struggle with this? Oh, I struggle with this. I'm probably not the best person to ask. Well, here's one of the reasons why I ask. When I see you, you're extremely positive.

And most cognitive scientists that I talk to, they're neutral. Neutral to positive. And you're almost always happy. It seems like you're not perturbed. You're not easily perturbed. You're equable. No, I think it's a useful state for communication. And a lot of people that have to maintain an academic position find it extremely useful to

look like a professor, right? It's a culture, you have to maintain a certain gravitas. If you come across as a friendly person, or as a humorous person, that might limit your impact with certain audiences, right? Are you really the person that deserves this funding if you are goofy in some way? Right. Yeah. So, of course, I don't want to be goofy, but I also don't want to scare you. And I want us to have a straightforward, friendly and maybe even loving conversation. So I'm trying to open myself to you and I try to build a personal relationship.

And I find that kindness and friendliness and humor are useful tools for that. I also find that humor is often a useful tool to deal with your own suffering and kindness and friendliness. So instead of basically using roughness to enter your suffering, that is typically something that pushes you out of the area, which you would need to deal with. And humor is sometimes a tool that allows you to make an area of your mind that you have to explore because you need to repair it more bearable. Is there any principle that's higher that's worth dying for?

So for example, like it sounds vague, but for example, you mentioned that the machines will likely win, or let's imagine there's a scenario where they win. Let's just hypothetical. Then merging with them if we want to survive is what we should do. But is surviving what we should do? What if the machines win and part of the machine's goal is to take over the entire planet? I know we've done that, but let's say they do it in a way that causes excruciation for most of life. But we say, well, we want to live, so let's merge with them. Should we do that? Is there something else that's more important than simply propagating?

I think that something can only be important if the mind makes it so. The physical universe by itself has no importance whatsoever physical universe by itself has no importance whatsoever. Life by itself has no importance whatsoever. From the perspectives that my own aesthetics give me, I think complexity is valuable. And maintaining complexity over long time spans is desirable. And so basically having a state

in which you see the continuation of life at high complexity on this planet seems to be a desirable thing to me on the other hand life is excruciating for most conscious beings for most of the time in some sense existence is by itself not necessarily pleasant if it's consciously experienced, even though it's a constructed thing by the mind itself. And so I don't have an absolute answer to questions like this. They only need to be framed by a certain context. In the context of having children, I will give you a very different answer than by the context of looking at a planetary ecosystem. So depending on that particular

context, for instance, what do I wish for my children? Or what do I wish for my friends? How they should explore existence? Or what do I wish for my species? Or what do I wish for the ecosystem that I'm part of? These are very specific questions. And for these specific questions... But there are conflicts, no? Yeah, of course there are. And in some sense,

ethics is about the negotiation between these conflicts. But they all are predicated by a choice that you need to make in the beginning, the choice of what is important to you. And initially, we don't make that choice

because we have innate choices that evolution has done for us and that are solidified in our interaction with the social environment and so on. But eventually we get to the point where we get agency over these choices and where we can deconstruct them. And then the answers become complicated. And there is not a single answer because the answer obviously depends on the system that you are right if you become a machine so to speak if you uh identify with a different system for instance in this way that the dharama

identifies as a form of government as an institution a thing that uh or the queen to some degree identifies with the crown which is an institution it limits her actions as a human being the queen is not free to do what you and me are doing the dhala lama is not free to do what you and me are doing because he operates on different constraints right and these constraints give him both more agency and fewer degrees of freedom in a certain way and they also free him from worrying about certain things, because to the degree that you identify, for instance, with your family, you are free to worry about your own individuality within the family. To the degree that you identify with an ideology,

you can disidentify with your personal ideas. To the degree that you identify with being a form of governance for a group of beings, you can identify with this single individual that you are and it doesn't really matter if you die as long as you are reborn, this form of government in a new individual that is performing the role as well as you can or even better maybe, right? So this who am I?

What is this thing that I stand for? What is it that I identify with? It's not necessarily related to the substrate very tightly. In the same way as the software does not need to identify with the hardware that it runs on, you are not identified with your brain. You are identified with the things that you care about. And as soon as you get agency over what you care about, you can say that the choices that you make are more or less consistent with each other, and they are more or less compatible with the choices that others make.

But there is no absolute answer anymore. What do you identify as? It depends. I am not identified as the same thing in every state. And sometimes I identify as a father and sometimes as a lover, sometimes as a partner, as a friend, as a co-worker, as somebody who thinks, as somebody who struggles, as somebody who doesn't want to live anymore, somebody who wants to get some sleep. And nothing is as important as that and so on. Let's explore that one where as someone who doesn't want to live anymore, because I'm curious, what if someone says, I think what's most important is the destruction of life. I don't think it's worth it. I think that the amount of suffering is not worth the pleasure.

And you're saying that objectively, there's nothing wrong with that. So what if they come to that conclusion and they pursue it? Is there, yeah, what if they do that? That just, it sounds like it's objectively wrong, but you're saying, oh, that's fine. Just pursue what you want, as long as it's important to you. So, yes, I can only object to this to the degree that it is conflicting with my own goals. And for most people, that would be the case. And so they legitimately would give opposition. There's also this issue. It's very, very hard to sterilize a planet,

life is extremely resilient. Or humans, let's just say they dislike humans, because humans are the ones that have betrayed them in the past. Yeah, the humans are going to take care of themselves at some point, right? The species is not immortal. We might already be on the way out. And of course, there are individuals which the individual might fall out of this. And Peterson describes this as the school shooter got angry at God. Right. And I think that God, in this way that Peterson has been using it in this sentence, is the platonic form of the civilization that somebody is part of it's the greater whole

seen as a sentient agent the relationship of this individual to this god is the relationship that a cell has to the organism and are you a platonist no but uh so in some sense there is an a practical sense in which i'm a platonist but i'm not a Platonist in the intellectual sense. So I don't think that these categories exist beyond being more or less coherent models. But I think that as a learning system, I need to believe that there is something at the end of the gradient, right? There is going to be a certain model that I can approximate that describes reality optimally well, given my resources and starting point. And in this sense, I have a strong experience of Platonism, that these categories are real. And so in this context, it seems that we are a state-building species. We

are not a species of solitary individuals, unless you are a sociopath that does not have any sense for a greater whole. And most of us have the sense of feeling that we are functionally part of something that is more important than us as an individual. And this would be the implementation that you would need to give a cell if the cell was conscious and was able to make sense of its relationship to the environment, if the cell is part of an organism and not a single-celled organism. Organisms don't actually exist, right? Organisms are a way to think of large groups of cells that act as part of a greater whole. Sorry, when you say organisms don't exist,

do you mean to say that they're just a collection of molecules that move in a particular way and we model them as organisms? An organism is a function that describes the interaction between a group of cells. And it's a function that is different from a bunch of cells, because it says that some of the cells are not helpful to the organism, right? It's a function that describes a control structure. There are cells which don't belong to the organism, even though they are in the same region.

And some of these cells even share the same genome, but there might be tumors, for instance, and the organism tries to get rid of them. And so the organism is, in some sense, a function that describes an order. And the same way a society or a civilization is a function that describes this order. And the organism, if you look closely,

only exists approximately, right? Because you cannot describe everything that happens there using that thing. And if you look at reality, you only can do this in a hypothetical space where you reason about what lots of cells are doing. And the same thing is true for society or for cells. And this notion of the emergence of a society that is entirely coherent, where the behavior of all the individuals has sense with respect to the greater whole. This coincides with the invention of the concept of

gods. And I think they are basically the same concept. It's the idea that individuals can interact in such a way that they form an agent on the next level of description. And this agent is sentient. It has a relationship to the world around it, has goals, and a relationship to its constituents, to the parts. And you can see in the history of the religions that this relationship to the constituent parts changes, right? The old Abrahamic God is really a mean fucker. He doesn't care about the individuals. What he does to Job, just to prove a point to the devil is horrible, right, from the perspective of an individual. But from the perspective of an organism doing something to its cells,

it doesn't matter at all, right? Of course the organism is able to do things to its cells and the cells are not supposed to care about this because they belong to the organism. They are owned by the organism. They only exist by the grace of the organism and for the good of the organism. It reminds me of what you describe fascism as like.

Yes, exactly. And the idea of, for instance, the introduction of Jesus Christ and Catholicism was necessary to deal with a religion that was compatible with the Roman Empire. So you want to have a society where you already have Pax Romana and every individual has something like its own dignity and its own role, regardless of the society.

And you need to make a good offer to these individuals and you need to structure the relationship between the individuals and allow them to grow into the part of the organism. So you introduce humanism. And the entire idea of Jesus Christ, I think, is the introduction of humanism into this Hebrew religion. And of course, the Hebrew religion has changed later on and became more humanist in other ways. But I think that originally the invention or the introduction of the concept of Jesus was exactly this humanization of this resulting hyper-organism. And we have been selected to be part of a hyper-organism. And we have been selected to be part of an hyperorganism. People that didn't play their role in the hyperorganism and were unwilling to subscribe to it, they were often killed, right? They didn't have a lot of offspring. And we did this for a period of many, many

generations, literally over more than a thousand years. And so most of the people that live today on this planet are the result of having grown up in such systems of organization. And so they are selected for these systems of organizations. We are all selected for feeling part of something that is much larger than a tribe or a family. We are part of a transcendental greater whole of some civilization. And the old word before enlightenment for this civilization was God. some civilization. And the old word before enlightenment for this civilization was God. We are reluctant to use this word because it's so tainted by the mythology of the cults that we invented or that were invented to stabilize the civilizations. The religions invented mythologies

that are not possibly true, that don't have evidence going for them and cannot have possible evidence, like creator gods cannot have evidence for them you cannot observe as an inhabitant of the universe neck that relates to its creation and any statement thereof will only be a mental state that the individual has not something that is a valid experiment that tells you something about reality so gods are mental constructs they are about as real as selves which are also mental constructs god is so gods are as real as you in the sense that you are a mental construct yes gods can are basically selves that span multiple minds and the greek gods are good good examples for that so

the greek gods are stable because they are all archetypes. They are all certain extremes of psychology or connections of extremes of psychology that give rise to some kind of a human archetype. And this makes them immortal. You can refer to them across human beings and you can treat them as if they would exist. And then there are the demon gods like Hercules, who exists because he has stories that make him immortal. But this thing that you can live in a mind of another person or of an organism and move from one mind to another one. This is what makes you a god in this conception of the Greeks. It's not the normative force that our religions had. And in our religions, the

identification of a god is not just it's an immortal superhero comic character, but it's some kind of archetype. Instead, it's a singular thing. It's a monotheist god, or it's a subset of it that is like a limb of that god that is describing what our civilization ought to be seen as. And the relationship that we have to that God is established, for instance, in prayer. Prayer is an activity in which we meditate about the properties of God and the relationships that we have to God, and thereby establish God. And in the process of prayers, gods can even become conscious if that's part of their specification, even though they use the hardware of the human brain of the individual human brain so in some sense prayer works even though the god doesn't exist in the physicalist sense it exists in some abstract sense

yes of course the prayer changes the relationships that people have to each other and the identifications that people have they change what people think is right and normal and good and what they want to do. And they change the way they interact, how they share resources. Do you pray? Do you meditate? Yes. And I could say that there is maybe everybody prays in a certain way

in the sense that we try to spend time in establishing our relationship to the greater whole and reflect on that. So I would say that in a very secular way I'm praying. But I grew up in a world where the religious cults were seen as repulsive because they are antagonistic to rationality. And I take issue with the anti-rationalism of the way the mythology of religions is enforced. So somebody who forces me to believe things that are manifestly untrue, just to have a checksum on my mind that distinguishes me from non-believers, is violating me. That's why I cannot be part of such a cult.

At least not in my present state. I can only be part of something that is compatible with being rational. And this means retaining autonomy over my identity and over my thoughts and over my morality and over my beliefs. Without that morality, identity, I cannot be part of something. And there are groups that act like this.

And you could say that they are spiritual, but they're probably not religious, at least not in the traditional sense, because they don't have organized religion for the most part but so to get back to this a person who wants to end all life it tends to be a person who is not disinterested in life it tends to be a person who does take an interest in the greater whole because the actions of that individual affect other individuals right they want to end life and experience for all of them, right? And so this means that there is a relationship that this individual has to the greater whole.

And it's one of disenchantment. It's one of opposition. It's one where that individual decides that the greater whole is not good, that it's not worth it, that it's doing something that is morally unjustified. And this can happen, right? If you, for instance, if you are a kid that is mistreated and bullied by everyone and you don't have a space in the world at all, and you decide it's not your fault and so you should die,

but it's everybody else's fault. And the way that everybody else plays, and maybe it's the way that everybody else is organized by the forces of the universe and evolution, and you can just not make peace with that maybe you radically oppose it and you want to end it and that experientially might uh manifest itself as um running amok so it's a result of not just a loss of meaning but an inversion of meaning it's uh it's an inversion of this spiritual need that people have but it's still a spiritual need wait can you expound on that it's an inversion of the spiritual need

i think that in our society that has lost its future and uh you cannot have a civilization without planning for a future because this is what a civilization is for and about right in the nietzschean sense of the death of a higher value? No, it's basically a civilization is the thing which makes you build a cathedral over 500 years. It's something that allows you to act on long-term plans. It's something that allows you

to organize things in such a way that your grandchildren will have a way of living. And we have given up on that. Our future is changing much more rapidly than our models of the future. And so we have given up on that. Our future is changing much more rapidly than our models of the future. And so we have stopped tracking the future. But what we can track now is that there is not much future left, possibly. At least we don't see how it can play out well. We don't have strategies to deal with the existential problems that our future is bringing.

We realize that the summer is awkward. That's one of the worst summers that we ever had. We realize that the summer is awkward. That's one of the worst summers that we ever had. But we also realize that this is going to be one of the coolest summers of the next 100 years. We are unable to imagine what the summer of 2025 will be like. And when we think about it, we are terrified because we don't prepare for it. We don't have ways for dealing with something that is worse than the status quo. What does that have to do with spirituality?

So basically, we live in a world that has lost its future. And as a result, we have lost our culture. And so our spirituality, our innate need for being part of a culture that is giving rise to a sentient civilization, it has become a phantom lamp. And that's why people are drawn to superstition to to feel that phantom lamp so this phantom lamp attaches itself to ideas of a conscious universe or of immaterial deities that care about you in some magical way and this is all of course bullshit this is uh this is really the expression of that phantom lamp. The thing that is real is life on the planet. It's the ecosystem.

And it's our part in the ecosystem. And it's our civilization that organizes us as part of that ecosystem. And if we cannot maintain that, then everything becomes meaningless. And we notice this loss of meaning. What do you disagree about with peterson you mentioned peterson's conception of the high school shooters as saying they are objecting to god and you said well that's correct in some sense there are many many aspects where i don't really agree with peterson it's i just refer to him because he is one of the few public intellectuals we have left for better or worse.

I'm curious, what ways do you defer? We talked about Dennett and you said you disagree with him in some sense because he's incomplete. What about Peterson? That's more commission rather than commission. I refer to Peterson because he is a common point of reference. We all know about

him, right? No, that's fine, but I'm curious where you disagree. And where you agree. Yes. So, yeah, the interesting thing is for me when I have to talk about disagreement because disagreement is the default state between minds, isn't it?

It's only where we can establish agreement when we are independent thinkers, where we see that we understand things in the same way so peterson thinks that the best way to interact with a cult that tries to be state religion with some kind of agrigore is to make a stand and to expose yourself and i'm not sure if i agree with that so basically he is fighting a cultural war in an ineffectual way, I think. There is also a deeper level. Peterson thinks that growing up consists in

making a sacrifice. And the sacrifice is self-actualization. He would like to be happy, but he cannot afford to be happy because it's incompatible with being an adult and doing the things an adult has to do. And his sacrifice is incomplete. He has not sacrificed his need for self-actualization. And that's why he appears to be so bitter. Right? If he had made the sacrifice, he wouldn't be bitter. What sacrifice?

Sacrifice of self-actualization? Yes. Meaning? He would become a priest he would have that serene state of somebody who has not lost anything because he doesn't need anything what he doesn't have right this if you look at the archetypal priest it's a person that is serene that is smiling because they are at peace with themselves and the world. And they might be suffering momentarily because the unbelievers crucify him. But apart from the moments of acute pain

or the moments of compassion for their flock when they are involved in their dealings with the world and try to help them and fail at doing so because not all things can be helped, right? The priest is supposed to be okay with what he does and peterson is not okay with what he does he is suffering because he has retained his identity that the one that he thinks he has sacrificed he hasn't i'm trying to understand i'm trying to understand this so you're saying that he has an identity

and because he's holding on to it, it's like a want and you should get rid of your wants and then you'll be placid and serene and tranquil. Yes. So he basically he's doing something that he has not fully internalized himself. He is expressing the tension between what he thinks he needs to be doing

and what he does, what he feels would reward him for doing. He feels pain in doing what he does. And there is, of course, this other thing that he is acting on certain incentives in this game. And the question is, are these incentives completely pure? So he is a publicist that is filling a certain niche, a certain vacuum. He is trying to give people values. He's trying to project an authority in a time that needs authority.

And he might not be projecting exactly the right authority or fitting authority, but he's interacting with the fact that the millennials are the first generation since uh the post-war generation that are authoritarian again right every generation after world war ii was liberal why are we authoritarian i think that the millennials became authoritarian because they realized that liberalism has failed them. It has failed at saving the environment, at offering resources and self-actualization to everybody. So now we need to go to some authoritarian system where we control what people think and feel and how they interact.

And the result of the world, the insufficiency of the world, is no longer seen as the absence, the result of the absence of freedom, right? The postwar generation saw that the problems of the world were that we didn't have enough freedom. We needed to free individuals, for instance, we should have the ability to engage freely and love and sexuality to self actualize. And the problems that we had in relationship was because we didn't have enough freedom in our relationships. And now, and now the results of injustice in the world are seen

by many millennials as the result of a surplus of freedom. Well, oppression is the flip side of referring to the that's that's privilege, or the yes, Yes. It's basically the, now the issue that we need to fight is privilege. We don't need, and privilege is a surplus of freedom. And if we can remove the privilege, as a result, we get less oppression in a more just world. Right.

But it also means basically we have to limit the self-expression of people. And this is opposed to liberalism. What if they say that hey we just want freedom for ourselves you have freedom all we're doing is trying to promulgate freedom just to those who are on the oppressed end yeah then there would be liberals there would be people like for instance uh the gay movement in the 1970s which said uh you have freedom to marry for instance and we also want to marry.

But social justice is, for instance, telling people if you are heterosexual, you cannot kiss in public because that's heteronormative. I see. And it's insulting those people which cannot kiss in public. And if you have an ability to do a certain thing, you cannot construct a life around this ability because that's ableist. Right?

So instead, we have to level the playing field. We have to build a society that gives a level playing ground to those which have no ability. And this leads into some apparent contradictions, but ideologies have no difficulties with this contradiction. And the main issue is that the rationality of this liberal system is still a rationality that, even though it's internally logical, is a rationality that doesn't serve most people. And it seems to be something that Peterson doesn't seem to understand. You cannot force people to abide by the logic of a system by which they lose. Why should they play a game by which they lose?

Why should you create criteria for getting a job that require the equivalent of an aptitude test? That you fail. And if that job is the only way that you can feed your kids, right? If for instance, becoming a STEM scientist or a machine learning engineer is one of the main ways that you can have social mobility

and late stage capitalism in the US. And you limit this to a certain subset of the population. Isn't that massively unfair to most people in society? And so why should most people subscribe to the criteria by which you give out these jobs? And so you will find yourself with a movement. What's the answer? What's the alternative? I don't know what the answer is,

but I think that, so I suspect that Peterson is not going to solve the problem, right? He is telling these people you are wrong when you try to change the criteria for how we give out jobs in STEM. But he is not addressing the reason why they want to change the criteria.

And the reason is, once more? The reason is social inequality. The reason is that they don't know how to change the criteria. And the reason is, once more? The reason is social inequality. The reason is that they don't know how to feed their children. So is UBI an answer? I don't see a systemic order in which UBI works. So I think that if you want to introduce UBI,

you ought to produce a simulation of the economic environment in which the UBI is sustainable. I think that UBI is the attempt to perpetuate a system under changed conditions. At the moment, wages, salary, is the way that we allocate resources to individuals, right? And they are also a way in which we evaluate the value of the contribution of the individual to society. And they are a way to discipline individuals

and a way to integrate individuals into teams and groups, into society as a whole and measure the value of their contribution. And as soon as you automate things and globalize and outsource, this falls apart.

And this is what we witnessed to some degree. It never worked perfectly well, but now it works less than ever because we have more productivity than ever and people don't live better than ever. And so how do we deal with that? And so the idea is we give people salaries,

but they are independent of what they contribute to society and independent of productivity and this is probably sounds like great for artists yes of course because artists are largely not going to change what they do so in some sense if you give an artist a salary they are still going to do their art because they are intrinsically motivated and it's very difficult to force an artist to not do art at least the artist will suffer a lot and the artist will typically play along with society arguably a society that consists entirely of artists will not work very well because many of the other things will be left undone right and a lot of society requires that people do things

despite not wanting to do them. People that collect the garbage probably need to be paid very well. And they deserve to be paid very well because nobody wants to do this voluntarily. Unless it's your own garbage and nobody else does it. Then, of course, somebody will eventually need to do it. So how can we perform this allocation of resources? How can we make sure that the garbage gets collected? And how can we make sure that people have skin in the game in our larger enterprises

in a system where we have UBI? So I suspect that something like a citizen income, where you have community-based income, and communities decide what kind of labor you perform. And this money can be given out as stipends, for instance, if you want to write a book and the community says, yes, sure, write this book. It's a useful thing to do. But we still have a way to allocate people into nursing jobs or into social interaction or into community management or into education.

I think that would be a good thing to have. I see a big danger in the particularization of society if people no longer feel as part of a greater whole and just see that society is that thing that feeds me, but it's something that I don't need to put things back into. And to think that UBI is going to magically achieve this because people have an intrinsic need of doing that it's probably not going to work out because there as if there is no force ultimately for doing that

over multiple generation there's going to be drift if there's no force to make you contribute to the whole society yes then eventually people uh will stop contributing to society because of the drift our opinions are not intrinsically mobile there is no intrinsically moral power in the universe. It's ecological. If an opinion is possible or if a behavior is possible, it will exist. If it's incentivized, it will be abundant. And if it's not helpful, it doesn't matter.

It will still be abundant. It will just mean that the system breaks down. And so I like the idea of UBI. But in some sense, the artists in eastern Germany and so on did have UBI and we basically had the right to work but we were not really forced to work in eastern Germany and as a result we went bankrupt

our society went bankrupt like literally and the houses that we lived in they still had the pockmarks of the last war because in 40 years we were not able to get enough resources to fix the houses, even in our capital, Berlin. It's ridiculous. We always had a shortage of labor, for instance.

The West had an enormous surplus of labor and often didn't know how to get people into gainful employment, because the productivity grew, but the population didn't shrink. And there was still labor competition, so working hours didn't shrink. And there was still labor competition, so working hours didn't shrink. And as a result, a growing number of people got unemployed because we were unable to allocate labor in an efficient way in the West. And in the East, people just absorbed productivity by being unproductive. And to some degree, this also happens in the US, right?

The healthcare system is the most expensive healthcare system in the world. And it's largely because most of it consists from unproductive things in documenting transactions. And most of the things that people do in the US is arguably documentation of transactions. It's the biggest part of employment, apparently. And so people work very hard, very long hours, and they still live in houses made from Tyvek and plywood, have bad water, and have healthcare that makes them bankrupt. And so the big question would be, how can we change this? How can we implement an architecture of systemic incentives? And I think that UBI is not part of systemic thinking. It's only dealing with a single symptom at a single level. And this is not the right way to comprehend society. You need to zoom out and understand the superorganism.

And is AI the solution or GAI? I think that AI can help. Definitely, it can help in making simulations and models of extremely complicated things. But there is also difficulty if we start to compete with AIs as individuals and as groups and as societies. You're probably going to lose if you succeed in building them. There's no reason why we should not succeed in building them. Right. They need to be our friends in some way. Yeah. But if you teach the rocks how to think, we are not going to share many purposes with them. How do you deal with a nested hierarchy of eyes? So for example, someone says, I want to eat that chocolate, but I don't

want to want to eat that chocolate. So it's like they have different eyes, different selves in your model. How does that work? It happens in every one of our minds, we can see this in children, especially, right. So children cannot establish behaviors that integrate over long time spans. And I think the difference between these different eyes is the time span, the length of the games over which they integrate their rewards.

So you have behaviors which integrate over short time spans and those that integrate over long time spans. And the difficulty is not so much to find that integration and to implement it. The difficulty is mostly attention deficit. Sorry, meaning? is mostly attention deficit. Sorry, meaning? If you are unable to maintain an intrinsic awareness on your long-term goals, then you are in trouble.

If you are able to model the world in deep time, that is the main difference between powers, power between individuals and groups is that they're able to model the world deeply and act on long-term plans. So if we want to have civilization prolonged, then we should identify with the highest I in the nested hierarchy? We basically should act on extremely long-term plans, right?

And we need to implement incentives that allow us to act on such long-term plans. And I think that, for instance, our present US society has foregone this organization. So the idea was here to basically remove structure. And as a result, we have more freedom for innovation. And at a certain level, innovation is indistinguishable from cheating. And the US is a society that basically cheats a lot on all levels. What do you mean that innovation is synonymous with cheating? It means that you play short games.

It means that you try to take shortcuts. Instead of doing the right thing, you do something that creates a little bit more dirt here and sludge and toxic waste, and you hope you're able to deal with it later. Well, you can innovate and build wind farms, no? You can innovate productive technologies. So not all innovation is cheating in that sense. No, of course not. But in some sense, the way in which you comprehend our role in society is to try to move upwards by innovating.

And a society that is well organized should not be focused on moving everybody upwards. It's about moving everybody inwards. Everybody should get better at what they're doing. We want to have, in some sense, the goal is not to make bread cheaper and more abundant because the bread is already abundant. You want to make it better and more wholesome and more healthy. Instead, we invent kinds of yeast that make the bread go up faster, but that give a whole generation problems with digesting it. Right? So having gluten intolerance is now a widespread and ubiquitous phenomenon, despite our civilization having been

adapted to bread and yeast and wheat for a long time. It's because we changed the weeds faster than we could adapt to them. And the way that we would adapt to them would be by evolution, which means selection, which means technically all these kids with celiac disease and the people with mild gluten intolerance should have less offspring. And then after a long time,

we have adapted to the new kinds of yeast. Is this a price that we are going to pay for having the bread being a little bit cheaper? Probably not, right. And so by saying we allow bread like this, or even bread where we suspect that it works like this, this is cheating. Or if you see, for instance, the current increase of disorders that are developmental, like,

Daniel Zeiss- sorry, when you say cheating, you mean it's a net detriment to the society? It's basically where somebody knows that what they're doing is wrong if they take a long perspective. If you say, if you would believe in God, would God want you to do that? That is cheating. If this is what God wouldn't want you to do.

And what God wants you to do is to play a very, very long game, is to do the right thing to the best of your knowledge. And doing a thing that might work relatively well in the short term, but in the long term kills the bees or increases the prevalence of autism spectrum disorder because you put stuff in their foot that doesn't kill rats in three months, but disrupts their angiocranial signaling during developmental periods, right? This is not what you should be doing. And this doesn't mean that simple blind activism is the answer. Activists often know less about a subject than somebody who

has a neutral position, a neutral perspective on the thing. Activism is distorting your perspective on things. And the people that have the most distorted perspective also tend to be the most activist about it, if you think about it, right? Because you are the one who gets most agitated about it. If you are extremely agitated about a subject that is not important, you are going to be the activist.

So it's the loudest voices that are the most emotional? Yeah, well, the other way aroundest voices that are the most emotional yeah well if you are the other way around if you are the most emotional person you tend to be the loudest one i see i see i see and but of course it doesn't mean that activism is per se wrong right uh it just means that uh the certainty that the activist has about things is often not justified and it's uh this is only basically a message to my younger self. Why? What do you mean? Oh, when I was 16 years old, I needed a new exactly what was in the best interest of the working class. Were you a Marxist?

Yeah, of course, I grew up in the system. And it made so much sense. And the crisis, you're an activist, an activist. I was basically willing to be an activist about this and when the wall came down i was very much in favor of not reunifying with western germany but i wanted to have a model that is more like scandinavia and basically a third way one that wouldn't be as all-out capitalist and so on i also thought this idea of keeping the factories collectivist instead of having them owned by billionaires

would be much more just and therefore desirable, right? Similar to how many millennials see it right now, which say nobody should be a billionaire. A billionaire shouldn't exist because it's so unjust. But I thought when our working class voluntarily decided to be exploited by billionaires again, which they did when they voted for reunification under the conditions that were on offer. I thought they were confused and being manipulated by the press. And there was enough evidence that this manipulation took place,

right? There was a lot of propaganda for making that happen. But what I was too stupid to realize is that this idea of justice only existed in my own head. And what was real to people was, under which conditions do you send your kids to school? What kind of food do you have on the table? What's the quality of your yogurt? How nice is your apartment? What's the quality of your carpet? How many days of holiday do you have?

And that matters more than billionaires owning a factory? Of course. It really matters. It really matters what are your living conditions. And if you have a society where Elon Musk can unleash innovation because he is a billionaire and can control resources

in which that state-run committee of functionaries of the party cannot, right, this is the better society. So inequality is not bad as long as the lowest tier has a certain objective level of life satisfaction? Yeah.

I think it's very hard to justify a society that uh you very have equality but everybody lives a shitty life it's much easier to justify in a society in which uh the medium income is very high and uh the poor people live a good life and there is an extremely high inequality. Inequality is not intrinsically bad. The question is whether it's justifiable and or the opposite whether the fight against inequality is justified by creating a world that is intrinsically better and I think most people would agree that a world in which the majority lives better is the better world. would agree that a world in which the majority lives better is the better world. So this was the thing where I didn't understand the systemic relationships. And you were certain about it when you were 16?

Yes, of course. Because I didn't see the contradictions yet. I saw a simple logical connection that flew from the Marxist theory. I saw the antagonism between the ruling class and the working class. I saw the injustice that would result from the system. I saw the limitations that existed within that system. I saw the trend of capitalism to destroy its environment and itself and to use more resources than it could replace and externalize the cost of production to the environment and to people that were not part

of the markets and so on. But I didn't understand that the alternatives, all attainable alternatives were worse. And that the fact that my own society was worse was not the result of lack of trying. I thought it was basically moral shortcomings of our government that led to the fact that socialism, as we experienced it had worse outcomes than capitalism as other people experienced it i didn't understand that the capitalism that existed in western germany was a system that was constructed in a better way than the socialism that existed in the east now the difficulty is capitalism that exists in western germany is also not sustainable in the wrong way it's also going to crash as far as we can see. What about in the US?

Same thing. Yeah, same thing. Only worse because the system is larger and the feedback loops are longer, so they're less effective. So it's better in the short run? If you have a system that is more... We're playing the best level of the game right now.

Democracy works relatively well in cities and city-states, and it's very difficult to get it to work at the state level, and it's almost impossible to make it effective on a level of a large nation-state because the feedback loops are too long, right? It's very difficult to set the incentives for governance, right? So do we need a global government? In some sense we need, I think, if we want to regulate our relationship with the environment properly because otherwise we will have a competition about the things that we don't want to compete about. For instance,

if we don't have a global government, but we have free trade, we might have a competition about who is willing to allow the destruction of the environment locally more than others. Or who is willing to accept worse conditions for their working class.

And so if you had a global government, you would be able to regulate that. But if you, on the other hand, have a global government, you don't have a competition between different governments anymore. So you have no incentive for the government to govern well. How do you deal with that one? And so as a species, or as people that have political theories, we have not found universal answers to these extremely difficult questions. Joscha, it's been extremely pleasurable. Thank you so much. It's probably the most edifying and substantive podcast that I have.

I don't know a subject that we didn't touch on. Many. Thank you so much, man. Thank you too. I enjoyed having this conversation. By the way, with respect to the social justice movement, it's difficult, I think, to say in the long term whether it's a good thing or not.

It's basically an ideological movement that tries to become state religion. movement it's difficult i think to say in the long term whether it's a good thing or not it's basically an ideological movement that tries to become state religion and this seems to be poised to do so and i suspect the reason why it is emerging its part is social media right so social media is creating incentives for egregores to emerge and to possess people. And the other thing is that the mainstream society is not working very well. And this leads to revolutionary movements. And a part of social justice is about redistribution of resources. It's a weird way of being a leftist in which you don't care so much about the economic conditions under which people actually exist, but you care about the identities of people. So you don't care about the contrast between people living in sheds and people living in

palaces, but you care about palace dwelling quotas for your own people. And so it seems to be a movement that is largely driven by the upper middle class trying to get in the lower upper class, something like that. Right. It's mostly academics that are already, you could say, in a privileged position. And I'm putting this into square quotes because academia is more open in a society than it has been for most of the existence of humanity. And so in some sense, the society is very democratic in the sense that everybody in the society is free to become an oligarch and enter the ruling class. And of course, the society is not set up in such a way that everybody can become an oligarch.

It also would not work like this. And not everybody has the necessary traits to become an oligarch. So the whole thing is in some sense rigged, but it's not rigged as it was before, where your birth decided everything else. And when you try to get away from what you were born into, people would go after you and kill you. And most of the previous social movements, for instance, the Bolsheviks in Tsarist Russia, were working against a system of indentured servitude, or the communards in France, which started the French Revolution, were going against the monarchy, which was no longer

able to manage society in the right way, right? People were starving despite an increase in productivity. And this mismanagement of society had to be addressed. And it was addressed in a way that was extremely brutal and led to, by itself, to starvation and to the destruction of a lot of culture and a lot of things that were beautiful and probably deserve to be maintained. But the society itself that was being destroyed was no more sustainable.

And that was the reason why this revolutionary movement came up. And when you have violent revolutionary movements that are destructive, this is often the result of your society not being able to implement mechanisms that reform themselves in a more benevolent way. And the US is stuck in this sense. It's stuck by lots of mafias that take out resources at every level. We are not able to build new infrastructure anymore

for that reason, for instance. Whenever we try to build a new high-speed train, the money just evaporates. And when we try to heal cancer, the person that has that cancer typically goes bankrupt in the process. So, right, something is wrong in that whole system. And we don't have inter-systemic forces that can repair the system. Instead, the system, by virtue of its own intrasystemic forces, is getting worse. So something needs to change the system. Even if the alternative is worse for time being, the alternative eventually will need

to get its shit together after it's taken power. So I suspect that's what's happening. And so if we zoom out far enough, it's very hard to evaluate whether the present revolutionary movements, despite the problems that they are going to cause and already causing, are wrong or right. Eventually, it's just large groups of chimpanzees that tell each other stories about what we do.