 Hello and welcome everyone. This is guest stream number 53.1 at the Active Inference Institute. It's August 23rd, 2023. We're going to be having a presentation followed by a discussion around the recent work of Alexei Tolchinsky, a case for chaos theory inclusion in neuropsychoanalytic modeling. So if you're watching live, please feel free to add any questions into the live chat. Otherwise, we will now have the presentation followed by a discussion with our excellent panelists. So thank you everybody for joining and Alexei to you for the presentation. Thank you, Daniel. Thank you Mark for finding the time. I'm very grateful and excited to talk about this. And thank you Andrea, Brynn, Albert, I hope maybe Mike will join us. And thank you Daniel and Active Inference Institute for organizing this meeting. So can you see the slides? Yep, okay. So I'll just proceed then. This is the outline of the talk. I'll share some preliminary comments and talk briefly about assumptions in the current models, summarize the data, talk about free energy principle formulations as compatible with chaos and talk about some future developments. And hopefully we can discuss that. I'm very much looking forward to have more questions than answers. I'd like to start with what Michael Levin taught me, I guess, is just why it's important. To me, the most important thing is the clinical applications. I think that with your work, Mark and with Yuck Pongsep's work, we have been moving away from dualism over the course of some time. But I think the other big D determinism is still highly prevalent and dominant. And I do think it influences our clinical work, diagnosis and psychotherapy and assessments. And I think that it may be beneficial for some of the harder conditions we work with to consider including non-linear framework. That's my main focus. This is the link to the pre-print and the papers published fully in your psychoanalysis journal. And some of it is a discussion with this foundational paper that you, Mark, wrote. The New Project for a Scientific Psychology General Scheme, which is a perhaps revision, a revival of Freud's famous paper, Project for Scientific Psychology. And I'd like to give you a credit mark that you have been opening conversations between disciplines that hadn't been talking to each other prior to your efforts, such as psychoanalysis and neurosciences. You've been doing it for more than 20 years with great success. In the same spirit, I wanted to mention that Society for Chaos Theory in Psychology exists since 91, which is when their first conference happened. About less than 20 years than the first conference on chaos in Lake Italy, around Lake Italy, Lake Colmo in Italy in 1976, which was mostly physicists at that time and then psychologists joined. And I've attended the recent conference in Toronto, which was absolutely wonderful, and I met extraordinary clinicians and scientists, and I think that it would be beneficial to open that dialogue as well, and maybe if there is a wall between us to consider demolishing the wall. In terms of intent for the paper, you, Mark, pointed out to me that you've had an exchange with Robert Gallitz-Rulevi. For those of you who may not know him, he's a psychoanalyst from Chicago, a psychoanalytic writer and teacher who wrote a book on nonlinear psychoanalysis, and deserves a lot of credit for his advocacy and he's sharing his thoughts. He posed a question to you, Mark, possibly considering nonlinear components in that seminal paper, your project for scientific psychology, and to which you responded that the key point in such discussion was whether the mental apparatus was indeed nonlinear. So I took your comment as a task, as a challenge, if you wish, and decided to write a paper about it. I don't know if the data presented is sufficient, but I tried. And very clearly, I don't intend to contradict or criticize, you know, there would be plenty of me to do. I'm hopeful maybe to add a branch on this tree that you're growing, clinical and theoretical tree, and should you consider it worthwhile to go in this direction, we wouldn't have to change very much because this is the paper by Friston, Lancelot, DeCoste, and others called Stochastic Cares and Mark of Blankets. Friedrich's principle is compatible with the chaotic attracting set. What I thought was not present just yet was a formulation where the attracting set experiences a phase transition, and I took liberty to ask Karl about it and he shared some very nice thoughts that I will, with his permission, share at the end of the talk. I'd like to cite Maxwell Ramsted, whom you know, and his colleagues from the Paypoint Basin Mechanics of Physics often buy beliefs, where they talk about dynamics, mechanics and principles, and in physics, Kepler's work is descriptive. It shows, you know, what the orbits of heavenly bodies are. Mechanics are the how of the dynamics. This is Newton's laws, and principles are prescriptive. They are about why things work this way. There's a certain hierarchy in physics where you could derive the second law of Newton from principle of least action, and then if you have initial conditions, you could calculate the dynamics. The reason I'm mentioning that is, I think, Mark, your new project for scientific psychology is a high road. It seems to start from first principles such as the free energy principle, and you seem to derive things down from there. So I thought, why not try to see how the low road may look like, if in addition to these derivations and deductions and declaration of what entropy should be in the brain, what if we measure it in animal studies and human studies in health and pathology and see what the data shows, and whether or not the low road meets the high road in the middle, or perhaps something may need to be added. I also like to clarify why I spend time in the paper talking about obvious things for everybody here, the axioms and postulates. And I'm quoting here Edward Frankel, who is a mathematician at Berkeley. And, you know, Michael will forgive me, but you know, a viewpoint is that one can say that some parts of biology can be seen as based on chemistry, which can be seen as based on physics, which can be seen as based on mathematics. So how do things work in mathematics? We know that Carl Friston likes to translate nearly everything he works on into the language of mathematics. Classically, there's something called a formal system, which we can metaphorically compare to building a house. So the foundation of the house is a set of axioms that has certain qualities. And then we build things up using the logic of Aristotle and we derive things like theorems. An example would be Euclidean geometry, which is on the plane that is based on five axioms of Euclid. That's the foundation. And then you could derive things like Pythagorean theorem. Now, if we were to change just one of the axioms in the foundation, the entire house changes. An example would be the fifth axiom of Euclid that parallel lines essentially don't intersect. If you change that, then we could have a house on the sphere, which is Lobochevsky-Riemann geometry and the radians on the earth are parallel and they intersect. Or you could have parabolic shape where parallel lines also intersect. So we'll have an entirely different framework. And it may be naive to expect the theorem to look the same. The Pythagorean theorem, if we put it on the sphere, is different. The formula is different. I'm mentioning that is because I love multidisciplinarity and I think that's the way of the future. With that, sometimes when we cross over into another discipline such as psychologists, taking mathematicals or physics ideas, then we may occasionally forget this formal system thing. For example, when we say things like human beings resist or violate the second law of thermodynamics, then the second law was formally related for closed systems in the state of equilibrium, such as a sealed chamber of gas. And human beings are neither closed nor in the state of equilibrium. So it doesn't quite apply. It's a different formal system. One needs physics of open systems to make a statement like that. So again, in mathematics, because you mark seem to have taken us deeply in the made of mathematics in your project, you have differential equations for action, perception, precision. You have, you know, mark of blankets, random dynamical systems, you know, equivalence as partial derivatives. And in mathematics, people fix the set of axioms, they fix the definitions, and then they build a theory, right? So I want to be clear, I want to clarify what are the axioms in the theory that is built? What is the foundation? And the effort of time, maybe I'll talk about just one of them, which is Freud's psychic determinism. And this is a question for me, not an answer. I wanted to ask you, because I am assuming whether you kept the axiom or you changed it. So Freud said in the psychopathology of everyday life that there is nothing arbitrary or undetermined in the psychic life. There's different versions of the statement, and it has been repeated by generations of psychoanalysts over 100 years, and it became very entrenched. It's sort of a cultural meme and a major tenant of psychoanalysis. If I were to give an illustration, perhaps, those of us who are not in psychoanalytic therapy, then let's imagine a analytic therapist hearing a patient who intends to say same with an S, but they use the word shame, right? So such a phenomenon, you know, first of all, I think psychodynamic, psychoanalytic therapist would probably say, that's not an accident, okay? Accident is not a used term in psychoanalysis. And secondly, they will probably infer of what it might mean. For example, infer along the lines of unconscious or emotional dynamics, emotional conflict. Somebody else may infer that it may have neuropsychological reasons. Somebody else may infer that it probably is a migraine, but what you might will not hear in psychodynamic culture is that that was an accident, that it was due to chance. I think that chance, accidents, randomness, stochasticity, chaos are not used, they're bad form in psychoanalysis. And that is the illustration of, I think for its assumption, ultimately, it's an axiom. It's a belief that is not questioned. If we were to translate that assumption in the Friston language, this is the Langevon equation from which Friston starts, which has notations by Daniel, thank you Daniel. And we have the change in the state of X equals the flow plus the omega. Omega is the stochastic component. It's the, you know, in Daniel's metaphor, we have a wave and then we have ripple. So we would have to change the stochastic component if we were to go with the Freud's axiom. And then if we do that, we would find ourselves squarely in the domain of cold physics, Newtonian physics. Laplace is demon. Laplace famously said, give me a position and momentum of every particle and I'll tell you exactly where they have been before and exactly where they will be in the future. Now, Freud didn't make statements like that, but this is what would happen if you kill the stochastic component. I wanna be clear that it's not for its fault at all. Well, first of all, you need axioms to build a theory. He chose that axiom, that's the theory he built. Second of all, it wasn't exactly mathematics. And finally, the zeitgeist of the time was deterministic, you know, perhaps there were, you know, initial derivations of stochastic mechanics and quantum mechanics, but Freud, when he was learning the world of deterministic physics was dominant. And there's an important nuance in this discussion pointed out to me by the first peer reviewer of the paper who said, you know, listen, if you have something that changes like the stochastic thing and then you apply statistics to it, such as variance, then the variance of a changing variable will be deterministic. And then if we look at things like words, which are used a lot in talk therapy, like apple, and apple is a semantic memory, it's a category. It's a coarse-graining already from green apples and red apples and all other sub-times of apples. So in that regard, words are deterministic things. So perhaps when we have a dream analysis session and the patient shares with an analyst, you know, the movie that they saw, the messy, you know, irrational thing with images and tunes and anything and maybe words even, if they narrate it, then they slow things down, they average things out and they put it in deterministic tools, which is words. So with some objects, we can apply deterministic tools, but I wanna point out that the foundation that everything and anything in the psychic life that all the mutations have no stochasticity randomness is a very strong assumption. And I'm curious what would happen if we were to allow ourselves to see things on the sphere or in the parabolic shape. That's not a bad exercise because the general theory of relativity would not have existed between without these other geometries. In terms of definitions, this is a familiar formula that you mark use of Shannon's entropy. He originally wrote about telegraph and ascender and receiver in a fixed alphabet. But here if we draw an attracting set which happens to be Lorenzo tractor in face space and we look at these, I don't know if you can see them, there's all the white points, that's the initial condition. And then they spread out and evolve as time goes by. But you can see the density of the attracting set. It is a higher density and lower density. So we could calculate probability distribution. A slightly different question is this, what would happen if you were to change the initial condition a little bit, one millimeter in face space? Different things can happen and one of them is shown in this diagram that would be divergence. In the formula here is initial condition multiplied by exponent of lambda t. So when lambda is positive, you have divergence. And it's called lepons of exponent. And that's one of the criteria for case. When it is negative, we have convergence. And when it is zero, we have equidistant trajectories. We have Newtonian mechanics. This is an illustration of a chaotic system, which is a great red spot on Jupiter. And it fits perfectly the Friston's paper about the Markov blankets as stochastic chaos because that thing is stable. We have been observing it for 300 years. And one could say that this elliptical shape is the Markov blanket around the chaos that maintains its integrity. It's not a contradiction because in one spatial scale, you have Markov blanket and on another spatial scale, you have a chaos within Markov blanket. This is a convergent system that is a bit trivial to all of you. This is a hole in the ground and then surrounded by heels in all sides. And if you were to put a ball in the heel, it's gonna roll down in the potential well. And typically in face space, we'll call the bottom of the hole an attractor and we'll call the heels repeller. Repeller is a term less frequently used, but it's the highly unstable point in face space from which things move away from and they move toward the attractor. Now that's just one attractor. Now interesting things start happening if you've got two of them. Imagine a second hole immediately to the right and the boundary between them may very quickly become key on it. Here's another illustration on a convergent system from the active inference textbook that I hadn't read when I was working on the paper. I discovered it later. The chart on the left is the density. And again, we see high density in the center and we hate lower density in the periphery. And the chart on the right is surprise. We're very not surprised to find a ball in the center and we're very much surprised if it stays on the heel. And I'll read you a citation. Active inference is restricted to systems like that one on the left, which counter random fluctuations with their average flow and thereby retain their form over time. It may sound like the statement contradicts what I said earlier that FEP is compatible with KMS, but it doesn't because you could have different spatial scales and convergence on one scale and divergence on another. I also historically think that we have been staying converging systems mostly so far. Now, if we move from things like that, which is very simple, to a mind or consciousness, then I think that what we could be looking at is this. That's a landscape with attractors and repellers. That is very complex. And I'm gonna quote the former president of the Care Society for Psychology, is that landscape is dynamic, it moves. Perhaps some things on it are stable if we were to imagine a clinical scenario where there's a patient who is very prone to shame and very averse to express anger, then we could say that shame is an attractor and anger is a repeller. But we work with pattern things in dynamic therapy, so they could be relatively stable until the work in therapy makes them more shallow distributions. But certain things change overnight. Some things change in a week or during the course of therapy. So I mean, strictly speaking, that's a dynamic landscape. And I don't think we model things this way. I'd like to also quote Maxwell-Ramsted on the same paper that free energy principle is a tale of two densities. It would be incomplete to say that we just minimize one variable. What happens is we minimize the Shannon's entropy of probability distribution over sensor states while at the same time we maximize entropy of internal beliefs, which is very important. That's adaptation. If somebody has rigid internal beliefs with sharp probabilities and then COVID happens, they will have lower probability of survival. They're not adaptable. So adaptability means maximization of entropy of internal beliefs while maintaining the form means the minimization of Shannon's entropy of probability distribution over sensory states. Here's another term, Kolmogorov-Sinai entropy, that is not intuitive and it's from topology. The picture you see is the photograph of the statue called intuition outside the Fields Institute of Mathematics in Toronto. And here's a graphical illustration. If you were to draw the trajectories in face space and surround it by a certain volume, say a circle with a radius epsilon and ask yourself what happens to this volume as they evolve? Well, when it expands, then you can have chaoticity. When it becomes more compact, you could have convergence. When it stays the same, you have periodic system, which is indicated on the chart on the right where you have a correlate of the Kolmogorov-Sinai entropy on the y-axis and you have this epsilon on the x-axis. So the flat line at the bottom where the circle doesn't change is the periodic system. Kolmogorov-Sinai entropy stays zero. The curved line in the center that converges on a key is the proper chaos. This is theoretical chaos. And the upshifts from it is what Friston calls stochastic chaos. This is chaos plus noise. And then you can have pure noise and then you can have pure linearity. So all of these things exist. They're all on the table. You can have geometry on the plane, geometry on the sphere, geometry on the porous bulk shape. So what is the data show? And when we look at EEG entropy, people use approximate entropy, permutation entropy and spectrum entropy, which are all derivations of Kolmogorov-Sinai. And the first two are in different phase space from the last one. I believe, Mark, correct me if I'm wrong. You cite Gerserys study in your book, Hidden Spring, when you talk about entropy and phases of sleep. And they use spectrum entropy, which is a derivation of Kolmogorov-Sinai in the frequency domain analysis phase space. So the data shows that systems, as small as the axon of a squid isolated in the lab, to single neurons, to coupled neurons, to EEG, we have evidence of positive lack of level of exponents in some regimes of operation. With EEG, we can be more specific that alpha-rhythm seems to be weakly nonlinear, while things below alpha like delta seem to be more periodic with noise and things above alpha, such as gamma and high gamma seem to be chaotic. So if we were to summarize this data, that it would be, we can say that Kolmogorov-Sinai entropy appears to increase as generalized arousal increases from delta to gamma-rhythm. Kolmogorov-Sinai entropy seems to be higher in healthy alert brain-mind functioning than in the states of Kolmogorov-Sesho deep sleep, and it can stay close to zero in a Kolmogorov-deep sleep. The interesting anecdote here is that a simple act of closing your eyes when you're alert decreases Kolmogorov-Sinai entropy. Clinical applications are vast and developing fast, for example, except for clinical psychology, I must say, and psychiatry and psychotherapy, but in neurology, the seizure detection on the EEG is right now can be done with the eye with about the same accuracy as a human neurologist and that algorithm does use theoreticity assessment. There's models for Parkinson's, neuropathic pain, and I have other examples that I'll skip for right now. This is a graphical illustration of phases of sleep, and again, it has Kolmogorov-Sinai entropy as a phase of sleep. It shows wakefulness and then shallow sleep, progressively deeper sleep and REM, and that's the time evolution. If you move REM in frequency to where it belongs to wakefulness, then you'll see the same progression that Kolmogorov-Sinai entropy appears to increase as a function of generalized arousal. This is seizure onset at about second four. Teasure is a reduction in chaoticity. Healthy functioning is higher chaoticity than a seizure. This is a coma, which is relatively invariant theta and delta frequency. This is a nearly periodic system with some noise and death is a linear system. It doesn't change. This is persistent vegetative state, minimally conscious state, where researchers were asking a question on prognosis of survival, and they've concluded that diversity and variability of the EHE was the predictor of survival. Higher entropy means life. So the hypothesis is that chaotic stochastic and linear processes as well as hybrid ones, such as primarily chaotic functioning with noise can be present concurrently at different scales of the brain mind or at the same scale, but in different places or at different times. I realize it's a mouthful, but it fits the data presented in the paper. This is again a picture from the textbook Enough to Inference by Thomas Parr, Karl Friston and Giovanni Pezzullo. It's a cortical column and the errors you see here ascending and descending connections. If by chance you're not familiar with the framework, they correspond to the ascending errors are prediction errors and descending ones are predictions. This is what they say in the book, that there's an asymmetry in message passing. And the reason for that is that the operations required to compute prediction error from expectation are nonlinear. This nonlinearity is due to computation of predictions using nonlinear functions that tend to increase the frequency of the signal. I'd like to talk a little bit about reduction. It became very fashionable and sometimes silly to use this like, oh, your reductionist, no, your reductionist, and the second most popular term is oversimplified. You oversimplified, now you oversimplified. To me, they're not bad words. They're necessary tools to simplify and to reduce. The map that equals the territory exactly is useless. Nobody's gonna use it. It must kind of, I'm quoting James Glick, simplify and abstract, but I'll illustrate the specific term. If the car is moving with a fixed velocity in a straight line, and if I know exactly where it is right now, I can tell with certainty where it has been an hour before. That is a reduction in time. And I can do that because the formula is linear. It's distance equals velocity multiplied by time and velocity is constant. If the watch is broken and I disassemble it and I find a faulty part and then I put it back together and it works, then it's a reduction in space. And I can do that because watch is not an end colony to use Daniel's terms. Watch is equal to the collection of its parts. So I wanted to talk about Mark, your mention of the predictive hierarchy composed of billions of homeostats in your work. This is approximate picture. This is not the exact picture of what you talk about in a new project. And this is from your hidden spring. It's a predictive hierarchy where predictions flow from left to right and prediction errors from right to left. And one of the statements in the book in the hidden spring, you say that brains many complex functions really can ultimately be reduced to a few simple mechanisms like this. So it appears that you call it reduction. But I just wanted, it is a powerful model. It is a flexible model. I wanted to describe some features of it. Well, first of all, if homeostat is present at every single scale, then we have this homogeneity of mechanism throughout the scales. And I want to say that in physics, things don't look this way. Quantum mechanics is the micro physics, statistical mechanics is molecular level. Newtonian mechanics is heavenly bodies. And then we have general theory of relativity. There's qualitative shifts that isn't a homogeneity of mechanism all the way throughout. In fact, when people say, take entanglement and they apply it to human beings, that's sort of a misunderstanding of entanglement. The second comment is from Michael Levin, who talks about these scales. And his comment is that you could have different functional spaces at every scale, where you could solve a genetic task on one scale, electrochemical on another, morphological on the third one, and how you do message passing. What is the common language? It's not a trivial question. Along the same lines, I've already mentioned that a message between the scales can be non-linear if you do prediction errors. Now, if you take one non-linear message and then you multiply it by billions, you're gonna have a full on turbulence, which is a chaotic system. It's a dynamical system. And turbulence, you cannot reduce the homostat or anything else. Saying that water consists of H2O does not explain turbulence. So I don't know, that's an open question, but I wanted to maybe share some perceptions of this model. This I talked about, and this is the Tristan's thought about phase transition and the attracting set. He said that if you have a hierarchy of Lorenzo tractors, a slower one and a faster one, and a slower one governing the transitions of the faster Lorenzo tractor, then you could have a FEP formulation with the attracting set experiencing changes in transitions. And that's brilliant, I think, and it matches the anatomy because the brainstem and the subcortical structures appear to be slower Lorenzo tractor that could be seen as governing the transitions of the cortex, which is an astral Lorenzo tractor. I know I'm oversupporting things a great deal, but it sort of, this is his thoughts. I think him and Lancelot de Costa are working on that. Perhaps Maxwell-Ramsted's group is working on that as well. The clinical applications, the landscape that I've mentioned before would be absolutely lovely to see if it were to consider going there. Borderline personality disorder, another term for that could be complex trauma. The metaphor that we use for these patients is they're predictably unpredictable. That the only thing we know on Wednesday that we don't know what's gonna come up, suggesting chaoticity. Acute trauma has a symptom of hyperarousal, suggesting possibly higher level of chaoticity. An onset of mania can be seen as a transition into a higher energy state. And addictions, particularly with stimulants, can be seen as phase transitions into a chaotic state. It may sound like what I said now is contradictory to what I said before, when I suggested that, you know, lower chaoticity could be a problem such as an onset of a seizure. And here I'm saying higher chaoticity can be a problem. But that would only be a contradiction in the linear world. In the real world where linear is very small percentage of what's happening, I'm saying both that too little and too much chaos could be a problem. Just enough chaos in balance could be perfectly fine. This is a graphical illustration of what can be done. This is from my colleague and friend, David Pinkers, who is a professor in Chapman University and I think also former president of the Society for Chaos and Psychology Society. These are trees and we sound like completely unrelated to what we do, except if you look at a human lung or dendrites in a neuron, and you can see a healthy tree on the right from me and a nearly dead tree on the left. And if you do the fractal assessment, fractal dimension assessment on them by counting the thickness of branches and count the number of branches at each level, then you can quantify it. And then you can have approximate quantification of health and resilience versus pathology. And again, as exotic as it may sound, this is very similar to what people do when they detect a seizure on the EEG. They measure CD, Correlational Dimension, which is approximately fractal dimension. And I think my 30 minutes are up. Thank you so much for your attention. Thank you, Lexie. Great presentation. Well, looking forward to this discussion. First, Mark, please feel free to add your reflections. Then Andrea, then we'll continue on. Thanks very much. When Lexie asked me if I would join this panel, the first thing that I said to him was yes, coupled with the second thing being, but please be aware that I know next to nothing about chaos theory. So I'm not sure how much of a contribution I can make to a technical discussion of this kind. So I'm just repeating that before I say anything else. If I say anything that sounds amateur, that sounds as if this guy doesn't know what he's talking about when it comes to chaos theory, it's because I'm an amateur and I don't know what I'm talking about when it comes to chaos theory. That said, the second thing I want to say is I'm not entirely sure what is at stake here. It seems as if Lexie feels that it's really important that I and our colleagues get our head around chaos theory and bring some of what it has to offer into our field. Because I don't know much about it, I don't know what's at stake. You see, I don't know why that's important. So perhaps somewhere along the line in the next just less than an hour that we have, perhaps, Lexie, you can clarify that for us. The third thing I wanted to say is, and perhaps in order to illustrate the first thing is that as Lexie was talking now, it seemed as if he was talking about chaos as if it were synonymous with stochasticity or randomness. And that's not my understanding, my very limited understanding of chaos theory was that it's not a synonymous chaos as used in that science or in that theory is not synonymous with randomness. It's rather synonymous with unpredictability. Because we don't know the initial conditions of the system or we can't specify all the variables with sufficient precision in their initial conditions, it becomes impossible to predict. And because of the nature of the interrelationships of the variables, one tiny little error in your estimation of the initial conditions can lead to massive consequences. And this is why it's so hard to predict chaotic systems. So if that is what is meant, then I would say, sure, of course, this applies to the way that the mind works and the way that the brain works. But that leads me back to my second point. I'm not sure what's at stake, but I would say if that's what you're asking, if that's what you're claiming rather, Alexi, then I would say that surely must be so, that it is exceedingly difficult to predict mental and neural events at any level of complexity. Now, I think I've said three things. Let me now say a fourth thing. I think that a crucial, well, let me put it this way. Of course, what the brain and mind are trying to do, it's a matter of trying to predict what's going to happen. So what I've just said about how difficult it is to predict what's going to happen in the brain or mind, it's almost the same thing as to say that the mind and brain are trying to model the world. That's the main thing that we're trying to do is have a good model of the world so that we can get some sort of grip on this chaos so that we can act rationally in the sense that we can act in order to preserve ourselves, that we do not dissipate by just exploring all possible states and no longer existing as a system because we as systems need to remain within certain bounds across a great many variables, as you said, we can't afford to just let shit happen and throw ourselves to its fates. We need to try to predict in a very unpredictable world what must I do in order to meet this need given how the world works. So the brain is trying to model the world. Our predictive model is trying to synchronize itself with the world and for that reason, it's going to be very chaotic. But the crucial point is it's trying to reduce the unpredictability. With that said, I'll make a fifth point or maybe it's another part of my fourth point. No, I think it's a fifth point, which is that the way that I conceptualize the predictive model, and it's not me alone who does this, is that it is a hierarchy model and I think that's very important. And perhaps here's where things do start, where it starts to, at least in my mind, I start to see some of what's at stake. I think that the more peripheral layers of the predictive model, I see the predictive model as a, not as a triangle, but rather as an onion. So the core is the, the core of the predictive model is the deepest layers of the model. That is to say, the layers that have the greatest spatial and temporal generalizability, and they're the most stable layers of the model. And as we move toward the periphery, which coincides with the periphery of the nervous system, in other words, the encounter with the outside world, so it becomes more chaotic. We are able to, we can tolerate more unpredictability at the periphery and we're trying to, and that's, and that also allows for greater accuracy because the world is so unpredictable. We have a more chaotic outer layer of our predictive model, where things also, there's less that's at stake. Is that car gonna turn left or right? Well, I thought it was gonna turn left, turn right, so what? It doesn't matter, but it is the car going to obey the laws of gravity, that matters more. That's not something that's a much more consequential surprise if it does not. And perhaps that's a silly analogy, but at least it illustrates the general point. Our most heartfelt beliefs and that is to say, the deepest layers of our predictive hierarchy, there we find the same patterns over and again and we find ourselves, we are more certain of our predictions and we act more stereotypically in accordance with them, whereas at the periphery, that does not apply. So I think at one point in your presentation, Alexi, you were saying something like, one size doesn't fit all, it's not all just one mechanism all the way down. Of course I completely agree with that. It's not just one mechanism. The different homeostats that you were referring to on their different scales are also serving different needs and there's not a common currency. I think that's a very important part of how consciousness works, why we have quality at all is because we can't reduce things to a common currency. We're dealing with categorical variables at the very least in relation to our bodily needs. There are categorically distinct needs and there need to be treated as such and this is why we have different feelings in relation to these different categories of free energy. So those are some of the main things I wanted to say just to get the ball rolling. I'm mindful of the fact that Andrea Clarici only has 15 more minutes with us. So maybe I can pick up on some other points later, but at this stage, let me hand over to Andrea. Thank you. Can you hear me now? Yes, sorry. I will try also to articulate something about the very fascinating paper of Alexiabin. I must confess, since I'm also a psychoanalytic psychotherapist as psychiatrist, it's not my field, I'm not a physicist, so I can really say few words. Some is, as I said, already fascination because in general, I hate to stick to rigid theories. So if there is a paper like Alexiabin who can chaoticize my theories, even in psychoanalysis, in the workings of the mind and the brain, I'm happy to dive into it. At the same time, I don't have the skills to discuss the math part of this paper or even the physical formulas and so on. But I was reminded of it very early and I think you all know that book, the Jacques Monod Chance and Necessity, which is a classic. And I think it's a book in which you can, from scratch, I mean here that our aim in the world, and I know I'm sounding theological, but we have theological aims in order to stay alive, we have to face chance and chaos. So I think this is, in order to survive and reproduce, we have to have a mathematical system to reduce chaos and to bring up again determinism or psychic determinism. Since one of the points that Alexiabin, I think raised, is can we still hold the point of Freud's about psychic determinism? And I think if you apply not to the physics of just materialism, of things actually, but to the living organism, I think we do, this is my point and I would like to hear what Alexiabin says. In the end, the chaos theory mathematics, as far as I understand, it must be a mathematics who reduce chance or chaos to a linearity, at least in the living organism. And I wonder if, and what Alexiabin thinks about it, because he seems to disagree with that. Want to give a direct response, Lexi, or then Brain and Albert would be awesome to hear your takes as well. Some of the thoughts from Mark and from Andrea. Should I start by responding to Andrea's points or Mark's, sir? Maybe I'll start by, because I think the key point that Mark, you brought up is what's at stake is still unclear. My impression is, and please correct me if I'm wrong, I was very influenced, Mark, by your paper. I forgot exactly the journal where you talked about psychiatry world, how these sort of convention-based diagnosis of fear is a major depressive disorder. Here's a bipolar disorder, nothing of which has to do with etiology. Convention and statistics. And then you brought up this revolutionary thing with effective neuroscience, where you talked about the manic state and the depression state, and tied into the punk subsystem, I think, is extraordinary. What I wanted to say is, if we look at how we diagnose and think about what's happening, and we look at the language of how we do it, it is still a deterministic language at the end of the day. So when we moved from DSM-4 to DSM-5, we brilliantly discovered that continuous variables exist. We used to have buckets, everything was in a separate bucket, and then we said autism is a spectrum, and everybody applauded. But mathematicians were quietly laughing in the corner, because of course, continuous variables exist in, but we don't do dynamics. We're still very static. Here's this clear-cut thing, and this is the clear-cut continuous thing. And my impression, I may be wrong, is that even in our neuropsychedelic developments, we still do things deterministically. We assess which bucket things belong to, maybe two buckets at the same time, or three buckets. But we don't do things like how the chaotic systems are qualitatively different from static things, I'm quoting William Sulis. They have transience, they have contextuality, and they have emergence. We don't do that, you know? And I think that if we embrace the chaos theory, if we do things like correlational dimension assessment, or fractal dimension assessment, starting from the system of classification of emotions, there's still buckets. Here is a clear-cut thing called fear, and that's a clear-cut thing called rage, buckets, determinism, okay? And, you know, they look differently in the dynamical systems world. And then the system of psychopathology could look like, you know, correlational dimension changed, and not that somebody fell into a bucket. You know, if we look at this complex, you know, landscape of attractors and repellers, and examine psychodynamic things like, please forgive me if I'm incorrect, but one of the possible definitions of repression could be seen as not letting things into consciousness. Maybe we can call it suppression, but what I'm trying to say is, I don't wanna think that right now, it hurts. It's a painful thing. That, in the terminology that I shared, is a repeller. And perhaps when you work through things in a psychodynamic psychotherapy, perhaps you put a scar tissue on the wound, it becomes less of a sharp, you know, repeller, and it smooths out. You know, we don't talk, we don't diagnose like that. We don't treat like that. We don't think like that. One paper that I'm working on that is in peer review right now, and I can't share too much, says that, you know, we think like storytellers. I'm quoting Daniel Kahneman. Our, you know, think in mind of a clinician is not a statistician. We don't think in probabilities. We think in labels, and we think in stories. And I know you had this powerful quote, Mark, from Freud, who said that that's the nature of the mind. And I most respectfully disagree that that's the entire nature of the mind. I do think that we have things on the mind that are not stories and not words. So if the nature of the phenomenon is hectic and stochastic and periodic, then we can use appropriate mechanisms to diagnose and classify that fits the nature of the model. The second thing you said, Q, since stochasticity, I absolutely agree with you that they're different. You know, the chart that I've shown, if you converge to a stable stochastic, chaos is deterministic, forward deterministic. So if initial conditions are changing, they will diverge, but you can predict things forward, not back, but there is no formula in a stochastic thing. You know, you throw the coin, you don't know, you have no predictability whatsoever, short-term or third or long-term. So they are different. And in the sense that you're talking to sensitivity to initial conditions, this is music to my heart to hear you to say that you acknowledge that curiosity clearly is present in the brain. On which level of the hierarchy in your model is a very important thing. But again, when we draw this picture, whether it's a triangle or an onion, with a core and layers, it's a static model. And when you acknowledge that there's higher unpredictability toward the periphery, it could be, but there could be dynamics about the structure of the model. And perhaps what you're saying is survival, you know, like, I must breathe and I must breathe air and I'm not in the water. And that is a very stable prediction, absolutely. But when we deal with psychopathology and we deal with things like shame and pride and awe and other things, then I wonder if we should continue to use this metaphor of a homeostat. I really don't see how homeostat applies to gamma, rhythm in the brain or to shame. So a homeostat is the convergent systems so that you minimize entropy by keeping things in the rage. However, if you dive inside the homeostat, if you look at the temperature variation inside our narrow range, you may see chaos, which is exactly the picture that I saw. So that I showed. And to Andrea's point is about the Freud's principle of psychic determinism. Yes, I almost say this thing that we haven't stand goodbye to that. We haven't tried to say that chance exists in the mental life, accidents exist. So we don't do that. There's a firm tradition in psychoanalytic psychotherapy to say there's no accidents. It's all determined. That's a major key point of the discussion. I don't know if I said too much. Maybe Andrea, if you wanna give any... Yes, please. Or how do you want things to continue in the coming minutes and years? Okay, thank you. Thank you, Daniel. Just adding a word that I didn't use because I think it's very important about trying to defend the point about psychic determinism, which I'm very fond of. I mean, the word is adaptation. I have difficulties in... Because when you see a psychotic patient, when you see a borderline patient, well, I don't see chaos or total unpredictability. I think adaptation, their disruptions I try to understand what are the... Even the mathematical part behind that, in my opinion, has to have a tendency toward an order which is adaptive. In this way, I see the mathematical part of chaos maybe of use, I mean, but bringing to an explanation, maybe not to an order or to a determinism, but to a point in which there must be this adaptive reason, even in the chaos. Do you see my point? I do, but I think that there's a deeper issue about how we think. Say there's a clinician, psychoanalytic psychotherapist. We have a work in memory and the Miller's Law that Mark you quote a lot about the seven units of information in the order verbal work in memory. I think in the Miller's paper, he actually said seven chunks. So I'm sharing something from the paper and review right now. We can not think probabilistically because it overloads the work in memory. This is why when I'm a clinician, I think about psychosis, I come up with a story, a formulation that fits in the work in memory. And this influences our work. We don't have the mathematics in here to think in the terms of probabilistic things and chaos theory and other things. We need tools to do that. And in mainstream neurology, they use tools. They do talk to the patient that examined the patient, but they have MRI, they have EHE, they have PESCEN and all the other things. In psychoanalytic psychotherapy, we'll say we don't need any tools. We'll just talk to a person and that is sufficient for everything we do. But that's not necessarily true. Beatrice Beebe, who is a psychoanalyst works with a mother infant diet and there's zero words exchanged to zilch. But she looks at this movie that she records on a video camera and then she throws it, shows it in a slow motion. So she takes the flow, which is dynamic flow, not words, not discreet things. And then she slows it down and she shows the level of attunement between the mother and the child. We don't do things like that. We talk and we use these deterministic things to understand the psychopathology of the patients and that psychosis and that schizophrenia and that is bipolar one. This is panic system and this is seeking system. We determine everything and that is a limitation because inside I believe, what is a fear cascade? How does fear start? It's a flow. It's a turbulent thing. Now we put one label and it's a fear. We've classified it. So for some things it's appropriate, but for some things it may not be. It may be, you know, as Einstein said, as Mark, you quoted Einstein, as simple as possible, but not simpler than that. So at which point we are missing the phenomenon by using deterministic tools in our work. Thank you, Alex. Thank you. Thank you Daniel and Mark to everyone. Thank you. Have to go. Awesome. So Albert and then Brennan will continue on this topic. Yeah, thank you for the presentation and the very interesting discussion so far. One thing that comes to mind is the, well, the thing we're talking about right now, the psychic determinism as depicted as being a foundational law in psychoanalysis. Well, I'm thinking about the authors, the psychoanalytic authors who came, the prominent ones who came after Freud, who seem to have room for this chaos. Maybe they didn't use this term explicitly, but I'm thinking, for example, Wilfred Biong who's got the notion of the O layer, which has connotations of, well, the thing that cannot be known or categorized who has indeterminate elements of the mind. And that's one of his major contributions to psychoanalysis. On the other hand, more the French schools, I'm thinking of Jacques Lacanon, who's got the notion of the real, which is also one of his major contributions to psychoanalysis where notions of rupture, of absence, or of lack, are very prominent. I'm not sure, I'm not an expert here, but I think Lacanon explicitly challenges the notion of Freud's mechanistic worldview in one of his seminars. So I'm wondering what you think about this, Alexei. I'm not an expert in Lacanon at all. And I know Wilfred Biong in a limited fashion. So perhaps Mark is a better person to talk. And I like Freud, particularly in Mark's interpretations and translations, but maybe Mark can comment on Lacanon beyond. I am not sufficiently knowledgeable about their work. If you want to, Mark or Bren, you can share some thoughts. Yeah, please. Well, I think that the point that Albert's making, that is we must remember that psychoanalysis is not synonymous with Freud, that not everything in psychoanalysis stands or falls by whether or not we accept this or that tenet of Freud's. And in neuro-psychoanalysis, we try very much to draw upon all of the traditions within psychoanalysis. And we sort of fondly believe that we'll be able to find more common ground because there are greater empirical tools available in neuro-psychoanalysis than in psychoanalysis by itself. You know, we are hoping to resolve some of these theoretical disagreements, trying to move beyond just different systems of belief, trying to move more toward a psychoanalysis in which we can decide questions in an ordinary scientific way. But it will always be very difficult for exactly the reason that Alexi's whole talk is about, I suppose, which is the extreme complexity and therefore unpredictability of the phenomena that we're dealing with. So I think Albert's points are valid that the Beyonce approach, Lacan's approach, they add a great deal, they don't have as much to lose if we throw out the principle of psychical determinism. But having said that, let me just quickly say a word about psychical determinism because this has been a big part of what Andrea picked up on in what Alexi said. I think it comes down to the question that I asked at the outset. It's not a matter of is the behavior of the mind random or it's a question of how predictable, it's not a matter of is there no determinism, it's a matter of how easily can we trace effects back to their causes. So the fact that there's a great deal that goes on where we can't do that. There are many things our patients say, whether they be slips of the tongue or anything else of that kind and not only to do with parapraxis, but anything that they do. There's certain things that we see with regularity. I mean, the very phenomenon of transference is all about regularities. That there's certain things that we see, yep, this I can understand. This happens repeatedly. I think I can explain that with reference to certain aspects of the patients in a world and their past, but there's a great deal else that we can't. So I don't think that it's a kind of, either there's determinism or there's stochasticity. I think that this is why I was talking also about the layers of the predictive hierarchy, that the deeper layers are the unconscious ones. Their things are pretty stereotyped, pretty predictable. In fact, too predictable. I mean, in other words, it's over generalizes that it's too confident about its beliefs, but as we head toward the pre-conscious and conscious peripheries, there things are much less predictable. So that's, I won't say more, there's a couple of things in what Alexi said in his comments after mine, that if we have time, I'll come back to you. But I think let's rather hear from Brynn before I say anything more. Yeah, thank you so much. I'm perhaps the least qualified here to comment on your paper, Alexi, but I do have one or two thoughts that come from a place of being very fortunate. That's why my life to have been a patient for many, many hours, sometimes on the couch, and also to be a clinical psychologist. I failed to mention that in my introduction. So I'm a clinician, I was until recently when my foray began into more research-based areas and by the active inference institutes and Mark's work and Carl Pristin's work, I was a pure clinician. I had it in my mind that I was going to treat psychopathology to the best of my ability, but really it was being a patient first and foremost that brought me into closer contact with the theme of your paper and certainly the themes of Mark's work. And I think I've encountered it from both areas and it certainly, it matters to my mind, it matters less the inner workings of the degree to which a stochastic, all these big wonderful terms that I'm only beginning to even think about. But what matters to me is that how do you help somebody who is in a state, who comes in disheveled and how does one go about helping, aiding somebody in that state? I fell out of love with psychoanalysis for personal reasons, but it was in returning to it, I've become more kind of a better acquainted with its strengths and its limitations. And I'm realizing that that when you help somebody and manage to reduce the noise, to reduce the chaos in somebody's mind, when they become the greater degrees of health start returning to them, that they do rely far more whether they know it or not on probabilistic reasoning in their daily kind of functioning. And so I think there is a strong case to be made for a mathematical way of processing the world. It doesn't have to necessarily be made conscious to the person, it can be a bit overwhelming to try and impart that on somebody. But I think a brain that is not working properly regardless of the degree to which there is chaos that can be measured or quantifiably measured in that brain, somebody is tripping over themselves because they are unable to predict what might happen. And that is a terrifying prospect because I think there is some reliance and some staunch dependability on the fact that the sound is going to rise and that it's going to set and that I can then focus on other things. So that's just my comments more from kind of a personal front of being a patient and encountering the limitations of psychotherapy and what I would have done in those moments of need to have not, as you say, had a more deterministic view of what was happening. And I've tried to transfer that or carry that over right or wrongly into my own therapy as a psychotherapist. So a more personal thing of what it means to be a feel chaotic and what I see regardless of how confirmed it is or not, but chaos reigns and it's to try and not compound that is certainly something that I'm interested in. So yeah. Thank you, Bryn. I'll also give my thought on what's at stake and what's on the table as a total non-clinician but what I saw in Alexi's paper and also what I see people coming towards. I see the paper as Alexi making a real stable of concepts or portfolio of topics ranging from the first principles of active inference like surprise bounding to dynamics, stochastic, chaotic which are not the same thing but they sometimes cohabitate. So bringing a set of concepts into play making it matter for the field so that it can matter for the clinician so that it can matter for the patient which is where you totally brought it home to Bryn. And in my totally simplified model I was seeing all this diagnostic machinery from personalized data, EEG recordings all these surveys and samples all the different conversation therapy itself all these different measurements, inference and then it passes through the Markov blanket of the diagnosis and it's like well now we know that that person is X or they have X and then here's where we now branch back out with modalities of treatment and so that bottlenecking has obvious relevance in terms of like getting a categorical grip on what we basically all agree are wet wear, complex systems they're sensitive to initial conditions all of that and what if though like you brought up the example with a scar tissue maybe it's literal bodily scar tissue maybe it's conceptual scar tissue what if that inference were able to be understood in the context of an attractor landscape and it's not just a singular attractor so we have multiple attractors we have a chaotic landscape and so what if we were able to perforate that blanket speaking super loosely to bring in what we have already recognized as diagnostically relevant like the fact that patients who have differential entropy have this or that epileptic outcome or coma outcome bring in that machinery that led to the diagnosis but now take it beyond the diagnosis and maybe even beyond the clinicians concept directly into the hands of the patient Thank you so much perhaps we can hear from Mark and yeah, I'm not eager to talk if there is time I'll say something but if not that's okay Well, let me first of all say something about my screen you might some of you have noticed that suddenly everything went black behind me that's because I have I'm in South Africa where they have power problems and then a kind person in my household thought that they would help and they've put this huge spotlight here which is shining in my eyes and so if I look as if I'm in a rabbit in headlights it's because I am I sent a text message saying please come and remove this big lamp you put here but they don't seem to have received the message Now back to our actual topic to the extent that I'm able to not be distracted from it I think that when I asked Alexi what was at stake his main answer was to do with the diagnosis now I don't mean that he's saying that that's all that's at stake but I think he's saying this illustrates what's at stake So I want to first of all make sure that because I'm not sure who our audience is and who's going to be watching this in case anybody is misunderstands I want to make clear that in psychoanalysis we've long ago moved away from DSM and ICD type diagnosis we always thought that the the psychology the psychiatric methodology was wholly artificial and created divisions where they don't exist in nature and we replaced it with a far more dynamic approach to diagnosis and it's not just a matter of labelling it's a matter also of what kind of thought process lies behind the diagnosis in other words conceptualizing what's going on here rather than saying this is an instance of personality disorder and I'll specify which type you know we don't think like that in psychoanalysis and really we never have in neuropsychoanalysis we build in that same tradition I think that what Alexi was worrying about is he's saying yeah but you know you use you use systems you say that there's a certain number of drives and you use those to pass what you're seeing clinically and yes we do and we have to because you can't just say what I'm seeing is chaos you know what you've got to say is well what I'm seeing is chaos but let me try my best to find the deeper structure of this and so what we do is much the same as what I was saying earlier about how brain and mind work likewise the clinicians mind works that way we have some deeper concepts that organize what we see and then more on the surface we see the complexity you know in other words we don't think that we can reduce a patient just to one saying this patient has problems with this drive or something it's not as simple as that of course but we do start with saying well these are the drives these are the attractors to put it in terms of the attractor landscape that you're talking about these are the these are the preferred states of the of the human phenotype you know and in mental life too they are preferred states there's certain there's certain attractors that we're trying to stay in those in those like for example fear we want to be safe we don't want to feel fear with respect to attachment bonding we want to stay close to our to our attachment figures we don't want to lose them with respect to rage we don't want to be frustrated and have things standing in our way we want to we want to be able to have free access to the objects of our needs and so on you know those are these attractors and you know we we have great difficulty staying in those in those homeostatic bounds and so you know it's it is a dynamic picture and then it's a quick over over the question of the of the different drives there's the highly individualized predictive model that each one of us comes up with in relation to those drives and in relation to the conflicts between those drives and so on so I think that this way of thinking that we use in psychoanalysis and neuropsychoanalysis is already of the kind that Alexi is calling for so I think that what Daniels just said is that therefore perhaps very important to make clear it's that that he's offering us a language an additional language an additional I don't just mean language I should say an additional set of concepts and and a highly developed one which we might find useful and so we should take this on board that sounds very convincing to me if that's what's at stake here that that Alexi is saying this is a language that lends itself to the phenomena that you study and to the way in which you go about studying it then it sounds to me as if Alexi is right so that's the main thing I wanted to say in response to to his comments in relation to what I said so if anyone else wants to say anything else that's that's good with me but Alexi if I've misunderstood and if there's something else that you think is at stake that I'm not getting now's your chance to say it well I just wanted to end the conversation when you said that that if this language might be of any even to consider this language to in our future models I'm a happy man and there's nothing else I need to say and I just wanted to say that these are human things they're occupational hazards some of the things I'm talking about this paper I'm working on about narrative stuff is that of course it is much easier for us to think in categories even in the psychoanalytic world we do let's take the McWilliam system and the previous one we used to say three levels of functioning you know neurotic borderline psychotic these are categories they're not DSM categories they're not bipolar one but their category is nonetheless then we have character styles obsessive character style histrionic character style their categories okay so to some degree it's a human condition to put things in categories and I do think that this is this is a very strong historical reasons that we think this way we have generations of thinking this way in categories and yes at the end of the day when we paint a portrait of a patient it's developed and it's like a tapestry all of these things weave into one another and it's complex but we start from basic categories and you know um certainly if we consider that you know some things can be transient there can be emergence they're like that the certain behavior in one context is a compulsion but it's not a compulsion in another context and perhaps I'm saying trivial things to clinician and to clinicians and uh also what what what during you said just to comment when you said as a patient to reduce the noise and reduce the chaos I detected bias in the language you see how we call you know psychopathology in DSM disorders disorder is a bad word for us and I'm trying to say chaos is healthy I spend most of my day in gamma functioning which is a chaotic state it's not a bad thing so in the folk language chaos is a bad word disorder is a bad I want to put things in order and then perhaps we're missing something what comes to mind is I think Mark you had meetings with Michael Levin and Ian McGilchrist who's you know big on talking about the mechanistic you know logical linear straight and narrow left hemisphere type of course it's not left hemisphere but that's the model versus holistic sensorial kind of right hemisphere artistic type and we're missing the second part like Daniel's background is surrounded by art where you know we we spend less time in the right hemisphere maybe maybe appropriately so but I think that there's some things that lend themselves to that level of understanding and we break things we look at psychopathology and a human suffering like it's a carburetor sometimes where we again you know put it in buckets that's my impression but what you said is wonderful and I you know I don't contradict you anything I agree with what you said Mark so Alex if I can just comment you know if I if it's almost like an arc you know the the pot of gold at the end of the arc is is the reduced chaos the reduced noise in terms of the patients from the patients perspective but how you get there ironically is actually by inducing states of tolerable chaos and then kind of reformulating that reconsolidating that chaos so it's not that a that from the you know healing or growth or whatever the patient might present to you for it has everything to do with minimal or marginal chaos induction and then you know reintegration of that so the whole process is is premised on chaos but in bite-sized doses maybe one could could say right and just to finish up and then I'll stop talking Jonathan Shadler who is you all know and is you know widely respected as a psychoanalytic practitioner and teacher and writer he recently said that take a patient X with an analyst Y together the diet has emergent properties that is not the same thing as patient X with an analyst Z right and show me one study where we we take that into account we don't we just we just do random you know a clinical trial then we do meta-analysis of random clinical trial what technique did the therapist use they used you know whatever four letter therapy they applied nobody looks at the emergent properties in a diet the this matrix you know between the two and that's the language of dynamical systems this is Daniel Friedman you know and colony that you know a single end doesn't build bridges and grow mushrooms and colony does all those things we don't talk about emergence in psychology psychology do we it's true it's also a fundamental challenge after all you can't replicate within one dyad or across dyads in a finite world so so it does point to fundamentals in our last minutes if each would like to give some thoughts where do we go as we now the operative exponent away from each other in our paths so maybe Albert, Bryn, Mark and then Alexi what are your overall thoughts or reflections and then where are we heading go for it Albert yeah I was I was about to comment on Alexi about the emergent properties thing it's kind of the same point I made earlier and later psychoanalytic theories do talk about emerging properties but yes I think what Professor Solbs just mentioned I mean this new set of concepts makes it very practical for us to communicate with each other because psychoanalysis is a bad habit of reinventing new concepts every new decade so it's there's a lot of chaos to use that word within psychoanalytic discourse so yeah but overall I thought I'm very grateful I think this was a very interesting talk so yeah thank you thank you yes so by way of closing remarks I I agree with actually it seems to me that there's there's an abundance of agreement in this meeting I don't see any big controversy so that's nice the I think what Bryn said he is really captures it I liked it because it because what he said was so experienced near you know he was just describing what it feels like to be a patient and to be in a state of not knowing and uncertainty as opposed to you know a more stable kind of tranquility and then the the ability to to allow yourself to return to uncertainty if you've if you've found answers that are too simple and that that are that are too static to allow ourselves to go back into this state of of not really knowing that's that I think really does describe what what we try to do for our patients what we have to remember applies to ourselves not only because we are ultimately patients too but because it applies to how the mind works it applies to science how science works you know that we we do need to find order but there is disorder out there and so we're always at risk of coming prematurely to a rigid system of belief and we have to return back to the cold face and back to the chaos but ultimately where we want to be is back in in that valley and so the as much as Alexi's right to remind us that at the that the the highest levels of brain activity in many senses of the word are entropic and that's deep sleep and seizures you know are bad things as much as that's true nevertheless I think that the great trend of the mind one of its fundamental working principles is where we're wanting to be more certain we want life to be more predictable we don't like to be in chaos and I think that that's an important thing to remember you know while we are acknowledging the importance of chaos and of the highly unpredictable nature of the world and of the phenomena we study in our field nevertheless we retreat from it and for very good reasons so those are my closing remarks and thank you and I'll be very brief Alexi yeah just thank you for the paper and for having such a strong focus on the patient and on psychopath well on human the human condition I think that's so nice to have that come out of the paper when at first glance it doesn't seem to be there so thanks for making it so human and then just Daniel for organizing and to spend time with Mark and Albert it's a great pleasure so thank you very much Alexi final words oh I want to thank everybody and Mark and Brynn and Albert and Daniel Andrea thank you this is a great fun and I'm very grateful honestly for your help and openness and flexibility one of the other reasons is kind of the time we live in you've all heard and made that comment that the rate of change is accelerating you know when we were hunter-gatherers uncertainty was norm you wake up there's nothing in the fridge you have to go hunt and we tolerate an uncertainty better we could put our worldly positions on the back and then we started living in the world of iPads iPhones were bothered with information and we must minimize uncertainty we must you know have predictability in everything which is not the natural world for mammals and now you know we live our life and then bang COVID hits and then bang you know certain person becomes the president of the United States you know and everything goes upside down so the black swans are hitting us more frequently so I think having the theory of that you know at our fingerprints and applying it to our work is probably you know beneficial but again thank you all so much I really appreciate it awesome well thank you all for joining the conversation continues so till next time thank you thank you bye bye bye bye thank you thanks a lot bye bye