 All right. Hello, everyone. Welcome to Acton Flab, live stream number 39.2. It's March 9th, 2022. Welcome to the Acton Flab. We're a participatory online lab that is communicating, learning, and practicing applied active inference. You can find us at the links here on the slide. This is a recorded and an archived live stream. So please provide us with feedback so we can improve on our work. All backgrounds and perspectives are welcome, and we'll be following video etiquette for live streams. Check out activeinference.org to learn more about anything that the lab is up to. Today in stream number 39.2, we're in our third discussion around this paper, morphogenesis as Bayesian inference, a variational approach to pattern formation control in complex biological systems by Franz Kutschling, who's here with us, as well as Firstin, Georgiev, and Levin. So we're going to have a fun discussion in the .2. The .0 was some background and context, hopefully over viewing the paper. The .1, we had a discussion that opened up several threads, and today we will in the .2 say hello on the introductions and then just jump in to a blank slide or to a not blank slide. So I'm Daniel, I'm a researcher in California, and I'll pass it to Dean. Morning, I'm Dean, I'm in Calgary, and I am going to say that I have one thing I want to look at down the road, and which is how long is a person a stem cell, which I think this paper might help us parse a little bit, and I'll pass it down to blue. Hi, I'm blue. I'm a researcher in New Mexico, and I also, similar to Dean, have questions about when a cell becomes a cancer cell, and like how does it develop its own generative model, and just in general, like when you're looking at complex systems, how these small perturbations can lead to like a totally different effect downstream, which I think is kind of related to the stem cell question that Dean had, and I'll pass it over to the first author, Frantz. Hi, I'm Frantz. I am a PhD student in the Level Lab in Tufts, met with near Boston, and yeah, I study part of active inference in my theoretical work, and I do a lot of experimental work as well, looking at information processing in organisms and cells. So where does stemness or how do stem cells play into this paper and this line of research, Frantz? Definitely something we've touched on, but just to kind of start generally, and then we'll zoom in somewhere, so take it away. Sure. The idea is that the cells initially, when they're in the simulation, when they're being initialized, they have, they're randomly initialized, and they have very low beliefs about what cell type they are. They can be one of four different cell types in this particular simulation, and they all share a model, a type of morphology they could achieve that they essentially are programmed to achieve at some point, but none of the cells are any of the given cell types initially, at least not with any significant precision. And as they move along signaling to each other and sensing each other and secreting certain types in response to each other, they then slowly by slowly infer a cell type, which essentially is differentiation. So cells starting with low precision about what kind of cell they believe themselves to be, and then there's a process of differentiation. So, Dean or Blue, any thoughts on that sort of stem to differentiated continuum, and I'll pull up an image from a non-active perspective that also shows that kind of a continuum. So I have just something to add that ties into the paper where there's aberrant signaling. So a stem cell differentiates, but then during the process of carcinogenesis, like retains some kind of or D differentiates into something that's more similar to a stem cell than it is to a different kind of cell. And so this plasticity is important to retain like so that we can repair ourselves, but also it can lead to kind of an aberrant signaling pattern. Yep, like this notion of having a terminal cell revert to a state and regain the ability of having stemness is one of the characteristics of cancer because it can then sometimes differentiate into other cell types that are like normal and downstream of a precursor or abnormal in some other way. I was going to say it's also something that happens and especially in higher organisms, right? Not every organism keeps absolute stem cells in all kind of populations. One example that I used to work on is an NZeperfish where you have an always developing eye and there's certain parts of the retina essentially that can de-differentiate some extents. They're not entirely stem cells. They have a stem cell niche as well, but there's also especially in all also in higher organisms, while they keep those populations of either stem cells or of already posh differentiated cells that can de-differentiate. So it's an important part in regeneration. And ironically, this is something that's not been studied for 4-4-1 of this. There's older research that was done where highly regenerative organisms when they undergo, they did this in salamanders and they injected them with carcinogens and they basically had it. You know, saw a tumor progressing and then they curve a lymph in these salamands highly regenerating. They actually renormalize the tumors essentially. So initiation of regeneration in those organisms seems to actually counteract cancer progression and even kind of remove the tumors, which is very interesting. And it kind of, this feeds a little bit into the idea in this paper as well that regeneration, right? What happens is you certainly now have a much higher flow of information, things that we arrange to de-differentiate. And that is kind of like what I was thinking as well with the simulation here of the cancer rescue phenomena where you rescue this formation of aberrant cells and aberrant signaling by having a stronger contact between cells as well. I'm not saying I'm a modern generation here, although you can. I think in the original paper, Carl has done some weights cut, I think, part off and let it grow again. I think that has been done in the original paper in the annoying ones place. But that's something to kind of keep in mind. This is something that was experimentally highly studied to an extent. So one general note and then a question. This question, this idea that the organisms that have the highly regenerative capacity and there's a continuum from regrowing just the tail to regrowing a whole limb to kind of an extreme or a highly regenerative case like the planaria that was studied in this paper. So it's all kind of on a continuum from total ability to regenerate from any adult somatic cell can fully recapitulate the full population all the way to a total opposite of that. And then you pointed out that it might be related to cancer, progressions and phenotype. And that speaks to a blue brought up about the like system level effects. There's the individual cell and then there's the system level because clearly there's something about the system level in the regenerative situation that would lead to this being different. Or maybe it could be explained by just something internal to the model. Anyways, you mentioned that higher information was associated with rearrangement. Could you explain a little bit about that or how does it play out in this paper or in general? Yeah, it's basically the idea that and it's not so much the total information content but the flow of it. And I'm talking a little bit about this last time, right? It's hard to really kind of quantify this in a more meaningful biological sense. But the idea is that it's information flow. I think Dina asked this question like is more information always good. And my kind of answer to that was it's within context of what the participating organisms can receive and can process. So in the sense that if you have cells that are processing certain information with a certain kind of rates, time constants, then kind of getting information in within with that timeframe in mind is what makes it more likely for them to kind of understand better their environment and react to it. As opposed to a cell kind of thing of it also cancer cells that you know have the whole sensory part completely perturbed that do not react to the environment anymore than everything else becomes noise and it gets ignored. That's kind of the idea. So there is a lot of differences in how the cancer cells because they have changed ion profile, ion channel profiles have changed machinery of what they're actually transcribing. So all of that leads to the difference in how signals are being processed. And understanding that and then reacting to that is I think is a key to understanding and treating this disease. Awesome. Thanks, Dean. I don't know if I have the answer to my question which is how long is a person a in finger and scare quotes a stem cell. But one of the things you mentioned, Franz, in the white one was I think Stephen asked the question about how much volume were in the flow. And what I've heard you say back was it's always a state space dependency, which which I when I'm reading the paper was I think you made a really good sound case for that. But then one of the things that when you had to when you had to sign off one of the questions that I brought up hopefully that maybe we can discuss a little bit today is they configure four and five of the paper. All that black area around when the sort of cells move into into place. That's a crude way of describing what's going on there. But all that black area around that. Am I is it fair to say that that's kind of the the state space dependency or is that just assumed? And how would we because Daniel did a good job of sort of saying there's some invisibility aspects to this when we do multiple layering. And so how do you give people a sense of what's going on in that in that black that blacked out area? Right? I think I think black is a good a good color representation because it's not really a color. I don't think but it but it kind of speaks to this idea that we've got these low beliefs. But how long or how long as a person whereas cells as a cluster of cells is that beneficial? Right? We're talking we're talking of course here under the the big umbrella of morphology. So that was kind of where I was wondering today. Maybe you could you could explain it because you the focus the eye tends to focus on the object within the frame as opposed to sort of the area around the what we would consider to be the object. So maybe you can kind of help me understand this state space dependency piece. Yeah I mean so if you have a figure two so and what I didn't plot in the figure four and five and two I did is the back of concentrations right so there are of course some levels of signaling molecules in the background as well which will also of course be the case in the actual simulations down the line four and five. The blackness so the one interesting part about the simulation which was done mainly for simplicity and lack of necessity is to not actually have an external environment per se like there's nothing right there's no external force like outside of base everything in the simulations is done by the cells where the cells they make the environment right so there is essentially there's an external environment by the signaling molecules that they put out but we don't have any external limit like limits around it we just basically that's the space we plot the signaling molecules don't go faster than that and there isn't any external force which we could do right we the part of the simulations we do this asymmetry we essentially get we get sort of an implicit environment by changing the response to the signals around that which you can imagine that you put something in the in the medium that that flips whole signaling being processed and that's it so back to your question so this is about this blackness and when is that beneficial if I answer that question correctly it's there is of course like a hype like a precision component to it and kind of more less hand less hand wavy if you know if your environment is very volatile and constantly changes and you keep right you have a not not this essentially is a closed system right if you had an open system constantly you would get signals also in the environment this will also become a lot messier and especially initially right in the in the first frames of that figure four and five where you get the simulations they all started together I mentioned last time there's certain smoothness requirements where like if we had to put this dampening on the sensitivity a long time so inverse dampening it will get more sensitive with time which is because initially they're very unstructured they have you know very low parable least and they will jump crazy if they have high sensitivity to each other because they're in such close proximity and that is something that we of course do see in biology right we do be most especially the higher organisms right they start off usually in close environments now of course you can you can argue this you don't need to go to any information that means argue well it's just you know thermally insulated they have their own food supplies it makes you know just makes like a more safe zone like just physically energy speaking but I do think there's a big component to it it's like shutting this highly volatile you know itself inferring organism often influenced from the environment at a stage when all these cells are figuring out what to do each other so they need they kind of need to have a kind of a barrier between signals from environments which normally you know to us adults turn into to a mature organism are not really dangerous but to end developing organism full of some cells can be problematic and will basically would would perturb its natural kind of progression in this state space on this sulfate acquisition there is a cool paper I can reference to you from Chris Fields who I am from disclosure I'm working with and very much in Miami who wrote a paper about modus celerity as exactly the consequence of what I'm talking about I just said the modus celerity basically emerging in a as a way to shut some cells or you know cells that are in an initial solution of modus celerity cells that are going through a solution towards complexity needing some kind of barrier from the environment basically making themselves making creating a mock-up bank around them that makes it so they're not constantly you know being perturbed by the environment and that's the idea of how my modus celerity arose in the first place is to insulate niches which will become stem cell niches from the environment so that cells that are undergoing you know very very dynamic changes in short amount of time that therefore need to be kind of sensitive to the environment cannot be need to be isolated from an organism environment that would perturb this careful inference of cell type thanks Dean yeah so in this in this figure too you can see that there's we're talking about ranges because we've got values on the horizontal and the vertical sides of the box so so one of the things that I think is really interesting is that once you so you we've all spelled it out if we're dealing with active inference we want to avoid something surprising right the whole the whole uh variational aspect of this so again if we're talking about going from something that is has low beliefs to something that gains in in I'll say gains in sensitivity without hopefully killing what I want to say next if if it's about signal adaptation to these top-down conditions which influence the final form taken is there some way based on what you researched with this paper is there some way to know as you're approaching that place of energy exhaustion like is there something that the cell tells you that it's it's finding this too surprising because one of the things we again we talked about was that rate of change being so rapid say like the UN report on global warming where we simply don't have the machinery to be able to adapt is there something that you found out through this paper that kind of said to you as you were looking at these things oh wait a second here we're just gonna we're gonna we're not it our intent is not to kill this thing off but we're approaching that place where the environment is just too changing too fast yeah i'm trying to give you a more need to answer than hand waving us but let's start with the general light answer first so yes you do it's you can of course you can like one way you can quickly see it is and that kind of like something that you get with the picture in an error plot that you also get with the free energy plot if you right in a perfect world you'll get like a nice smooth like you know an exponential decay of the free energy function if that keeps changing bouncing up and down that will be very problematic i did some simulations afterwards they're not included in this paper but based on exactly the same stuff where i was including a time based on top of this time sensitivity you know increase slow um that they are used for basically you know increase in the sensitivity all over the time but having a low in the beginning on top of that i put in a an impulse thing like a sinusoidal or rectangular function which was we were just trying to simulate what i'm doing in my in my experiments in a lab where basically you know i pulled certain signaling inputs to my organisms i wanted to see basically if i randomized that or if made that in a certain certain pattern how well would the simulation still be able to cope with that and in in the case where i had like a nice regular pattern of that um you're in the in the free energy land in the free energy decay you hardly saw that you know initially initially when the sensitivity was very high it was it kind of you know you can see it there there was like a big jump in this and fuzziness you could see it also in the belief updates you know basically they were they kind of emerged kind of like they diverged initially very strongly and diverged in terms of their beliefs that they had about themselves but then kind of afterwards the when the sensitivity overall was kind of acclimating but was kind of varying around this acclimation point you didn't see that much anymore in the in the in the in the regular pattern but in the ones where less regular um that was very much perturbed and the and the free energy also didn't actually quite uh did not only did it not um come to kind of like uh to this asymptotic behavior but actually started increasing and getting basically um just was really perturbed and you didn't see that so to answer the question like um how do you see if if you are not adapting correctly even though you don't actually know yet what what the what your external status i would answer it's how much are you switching back and forth between your beliefs and uh i guess more precisely also like how certain are you of that right this is something that we talk about the medic cognition which i've recently for my last people have been getting into and reading about medic cognition and rats and primates and other animals saying the first step to medic cognition is as far as i understand it is um having a certain confidence about your results related to rats where they basically um gave rats the option in this kind of cognitive task um that they could also just not answer the task like not to not not push level it was i think it was levels um i forgot where the task was but it was based on levels and they had an option to not push a lever um for which they would get a lower reward than if they um a much lower reward than if they got it correctly but more than if they got it wrongly and so what they saw is that they actually would insert in certain cases when you know when they hadn't learned it when they hadn't learned the task correctly they would just not push the lever and that is used as like a first sense of of the medical condition a sense of they must have been unsure about what you do and then chose not to do it so i think that's kind of like if you it's one thing if you kind of lean back and forth because you always think oh you know i'm this and that um being unsure is normal but kind of uh kind of uh constantly going back and forth between certain states of sureness is also problematic so that's a generalized answer to this um yeah i think i think the interesting thing there is the emphasis on the back and forth metaphorically speaking we tend to uh give a lot of attention over to the balance piece but what what you're talking about when you're talking about the environment is is not the plank as much as a moving fulcrum so how do you how do you adjust right the the balance i guess is the is the outcome but it's the the process itself is the back and forth so yeah i wondered about that because some things you do want from like from an epigenetic standpoint you do want certain cells to die you want their energy to expire absolutely i mean i've right that's part of the battle for it as opposed to the balance you know there's no balance metaphor you want there you want them to die and they're not dying so that's really interesting thank you and i i wanted to add one one more kind of a attempt to answer this more specifically is there's a paper by i'm probably pronouncing his name wrong droughily which was um just google valence and uh active inference um they basically looked at you know valence in terms of like um judging if your actions are you know certain like are they but better or worse basically is that um are you you be signing a sign positive or negative to to outcomes um and they were doing that uh in the active inference schemes and they were defining and you know there's other definitions um i almost can't read about this but they were looking at it in the active inference scheme about the rate of change they were looking at the first and second or derivatives um of um how i'm saying this correctly um about how the prior beliefs were updated to the precision that how how important like how how sure they were about what cell type they were and so basically in the second and um order derivative basically how much would they change at any given time in their beliefs that's what they um what they plugged in valence into different mode that gets more complex than that but that's the gist of it so um that i think is a uh in the within the framework of active inference is a more specific answer to your question that's look at the the rate of change by which your beliefs are are updating and that with respect to you uncertainty of course um there's this whole um cars written about this as well um i'm very interested in as well from the proudest standpoint of you is that stress make stress sensing mechanisms like stress is a universal um driver um in biology it's actually much more important than rewards that kind of find that we always talk about rewards but um biology is much more focused on on stress um because it's more informative and it's more important to deal with than than than what maximizing but of course you know always say it's just negative but um the the point they're being is uh stress as in as in terms of uncertainty like not having not minimizing your energy effectively and stably um that will lead to stress and there's paper that doesn't humans i think mentioned last time as well what they do experiments with humans and mice as well were they very ethical where they shock um the gift like the electric shocks i think they're both from mice and humans um and then they showed that the quarterly level so they were more stressed out not just by this like the shocks themselves they couldn't predict right um the electric shocks so it's that uncertainty of what's coming next i think that is the most if i had to like say one thing that is most negative of uh of you're going in a bad place so if you are stressed because you have no idea what's happening and you cannot even if you know something that's going happening you know when it's happening that's a lot less bad than if you don't know at all um one last really dark example to really make it a dark place is there is um i'm from um i was born east germany um i guess i was only when you opened the wall fell so um but we you know my parents were very much and there was uh this um now it's a museum but it was a prison for the the basically political between east and east germany and east poland and uh he went there as a class once and um they showed us all the thing all the techniques they used to interrogate prisoners including torture and one of the torture devices um was i believe in a japanese device where you would it's free time you you put your head down on this device and there's a water that's dropping from the top onto your neck so you think at this point like uh that doesn't sound too bad right the problem was is that the drops you know there's um there's a lot of viscosity in there so um you know a lot of kind of fluctuations so the drops wouldn't always come at the same time and they kept repeatedly um happening all time and time again so what the and the prisoners actually people that used to be in prison they were actually ones giving the tour so they were telling us yeah if you want a device after a couple of hours or so these drops will feel like hammers on your on your neck and apparently a big component of that was the um not the the body not being able to adapt to it because you know the drops don't come at a perfect five-second interval it happens kind of at semi-frase randomized um intervals so um that i think there's a huge part in and not just looking at stress over time but how you can adapt to the stress and in order to adapt to stress adaptation to me always involves a component of prediction you can't predict the source of the stress when it happens it makes things a lot worse this has been shown extensively in humans and also in lower organisms and i believe very strongly that there's something fundamental to any biological system to fundamental drive if you if you view if you prescribed if you describe to the point of view that minimizing uncertainty in your environment is a fundamental drive of life then with that you prescribe that minimizing stress and the source of stress over time and being able to predict it is also fundamental and lack of makes a very dangerous system blue thanks so all that's super interesting um just okay so let me i'm going to start at the front and work backwards so in terms of uncertainty and um how that causes stress uh i'm really kind of i don't know reluctant to use the word stress in a biological system like i think about stressing a biological system is like pushing it out of equilibrium um but but maybe that's just a stimulus right like i want you to rebalance and go to like a new equilibrium like or like turn on heat shock protein or you know there's a variety of different things like many ways you can view stress as a stimulus to do something new or to perturb the system in a new way um but but in terms of uncertainty i think that there's a lot of truth in um how uncomfortable people are with uncertainty like at a cognitive level and you can even see it like in the stock market before like the election is in before the election results like the stock market goes crazy and because nobody knows i mean it's not that one candidate is preferable over the other but it's just the unknowing is very um like people freak out so and that speaks i think to metacognition also um and then so i wanted to back up a little bit if i can um to your discussion of information flow um and how that might be difficult to quantify and daniel i don't know can you flip to the section for me please on um the generalized flow section i think we started to get into that a little bit last week and we did discuss it in the doc zero but i have like i i know that this is very um mathematically technical and so i just have a little bit of of questions about um this and maybe you can help explain it in a way that's perhaps less technical but specifically information flow um and chris fields is coming uh to discuss the fpp for generic quantum systems next week and so i can i can similarly interrogate him about about some of of this stuff and i'm looking forward to it but um you know we have a discussion there's a discussion in the paper about the probability flow um and information like a probability current i think is actually what is used and getting into information currents like if you look up what is an information current i think um there's like the idea of the von neumann information current and um that like is explicitly related to quantum systems so is there any similar like do we have a way to measure information current because even when mike came to the live stream and talked about um expanding cognition biological cognition computational boundary of itself there it was sorry um but when you expand biological cognition like there's uh you know you're basically expanding your informational awareness and so is there anything like an information current or how can you best relate that within a biological system is it like an expansion of of your computational boundary or does it look like just more information goes in and comes out i think it's uh it depends on what level you're looking at i would say it's both um so the the the second part you mentioned that the information fluid comes in out that's something we just have more access to biologically speaking experimentally right um and you can absolutely you know you can you can look at you know you have all these uh sensors for for sensors for receptors being activated for um genetic um transcription in response to that so so that's something that we can you have the best chance of quantifying experimentally and in terms of computationally it's something you can monitor and then also couple that with what chris fields um good thing you mentioned that because then i can relate to him that's basically what what he's also very interested in right like how does physical energy of course played all all all into this in the biological metabolism right now this comes for free that's something that biologists often kind of um my brawl is actually no theoretical biologists often kept under under the table because the idea is that energy is abundant and in in life which is to make sense true right there's plenty of sunlight that's plenty of of things to just kind of fool around um but at the same time locally just with entropy there is a strong competition there so just because you are living in a reservoir that has plenty of food and sunlight um that doesn't mean that you're locally are gonna i'm gonna survive because there's people locally competing with you the same is true right i mean this is always like when you talk to um when you talk to people that people that you know have problems the idea of um why does entropy i don't want to say names but if you that saying that entropy is always increase if if k r if k r's in the universe increases how do we see much more structure evolving and then um the of course the answer to that is that well just like in a in a in a bath of water if you have certain oil droplets are going to come together because overall that makes that gives more degrees of freedom for all the water molecules so the entire entire system entropy did increase but you manage to find a solution that this does increase up but locally create creating more order that's at the drive of um or i think of life overall the quote that probably the talk that got me into biology is um was from a biophysicist in heidelberg was saying life is not about energy it's about entropy um but the two come together right yes it's it's about how much information is available but then by extent of that once you are subscribed to that once you have to form these localized structures then energy becomes important again um so and that's what i think is needed right now when you have this highly highly space to organize systems energy suddenly does become uh something that is valuable again um yeah i think i kind of lost my train of thought there um the the boundary of self that's right you the other sort of information flow that was the part of my information flow you can measure that um you have good ideas what what's lacking is and a quantum understanding of of of how metabolism is really hooked into the information flow whether or not it's built into the model or whether it's ad hoc just down down the line you give it certain resources and after a while it just exploits them the boundary of self is interesting because and of course in the cell we think what we have a you know there's a membrane the boundary of the cell is fairly fairly obvious it's actually the easiest as i think and of course if you look at two multicellular organisms it becomes um you know tissue scale then that becomes a lot more difficult but even on a cellular level um if you have certain receptors um you know you have certain activations of it then a model energy and a general model not just in the active inference scheme but especially an active inference scheme you also have a model of yourself right and if at what point your own your own what you're what you're secreting on the owner your signals that you secrete also feedback into what you're sensing after all so it is kind of thought that that's that whole system is very much perturbed in cancer cells as well where um the how much like is it is are you actually receiving things from the environment are you basically just uh secreting and secreting and you don't really react to it anymore so the the boundary of self what what Mike if Mike already talked about this he will be have been much more like on i can be about this but the idea is the boundary of self is encoded specifically with time and spatial constraints so like how much your boundary of self what your yourself model be more general about this yourself model depends very much on your sensual on your sensory memory like how how far do you sense and how and spatially how how much around you can you sense and how much back do you have encoded as memory how much are affected in the past um um persisting so that you know so even though if you have a cell that's a certain membrane if that cell does not actually keep any track if that cell is not somehow sensing beyond a certain boundary and then it creates on on boundary by these highly preferred cells that are basically surrounding themselves with each other and suddenly the environment becomes nude um so the boundary of the self has especially expanded i would argue um i think that's kind of the the the thought train here i i'm just hope that was somewhere i would answer your question awesome blue so um just i was curious um you mentioned that there was a talk that um got you hooked into biology was it eric smith and his discussion of biology and entropy because that's also fabulous it was not i'm i'm real just seeing that no was um damn that's really embarrassing that the guy i don't remember the name of the person that got me into biologies it's as i was speaking english right now it's all in german back then and always kind of my brain then has problems switching between those two in content as well as language um i'll come back to me probably within this talk eric smith does a wonderful job though of um you know discussing biology in terms of entropy and so if you are not familiar with this work i would just highly recommend it yeah that really reflects on our um discussion what parts of the self model are important to carry forward and then what are the affordances for direct and indirect self modification so that's one piece like you mentioned the autocrine signaling and that's not even a contentious component it just is being modeled in a different way with active inference rather than just a molecular event that's happening that like a cells secrete molecules that they also possess receptors to measure it's just putting that into a framework of reflexive self modeling and stigmer g and niche modification here um but just one other point that i thought was really maybe useful to highlight was this sensitivity and just like many of the terms in actin for kind of seeing that there is a narrow or a quantitative sense and then there's a broader sense because in any dynamical systems model you might hear about sensitivity of the parameter which is how much changing that parameter changes the outcome of the system and so that's a quantitative sense and then also it was brought up that it's important to model these like second order derivatives like how fast are things changing relative to expectations or how fast are they changing how they're changing and that is what the generalized flow is is all those higher derivatives so we can use that quantitative sense of sensitivity to look at how parameter changes in the generalized coordinates of motion matter but also we can perhaps with cognitive modeling take this discussion of sensitivity in the way that people are usually meaning it like sensitivity of what is seen on someone's emotional state and then model that using cognitive parameters like valence and affect even though that's quite a disjoint use from the parameter sensitivity discussion on the generalized coordinates so it's was an awesome discussion with like what is stress and how does that relate to the parametric sort of neutral perspective on what stress or sensitivity might mean just the model descriptive versus some of these functional or phenomenological even ways that these terms come into play blue so that's super interesting thinking about sensitivity and stress in just in terms of cognition right like so I know like people that have sensory processing disorders get like super aggravated by something like a bright light or like a windy day and just how how sensitivity does play into maybe the flow of of information like because you know I'm suddenly sensitive to this and I'm also sensitized to every other thing in my environment like someone's screaming over here and it's windy and it's hot and it's bright and then like I become just so overstimulated right like people have these these processing things and it's sensory is that something that like has momentum right like so you have enhanced sensitivity and is that just like then all of a sudden you're just you're so sensitive like we hear people say stuff like that like stopping so sensitive just in terms of cognitive processing I thought that that was just super interesting I think that's an excellent point I mean that's something that that that shows exactly how important it is for humans and I again argue low levels as well and in the evolution to be able to model our sensitivity to change our sensitivity right we would never be able to have conversations in low environments if we couldn't change consciously our and sometimes subconsciously as well our sensitivity parameters and I think that's something else that then also fits into metacognition right you'll learn like oh well I'm not paying attention enough so something to pay more attention to you trying to tune this in and I think that's the in the sense of on the cell level I think what where that feeds in is the is by learning over time right by how much are being how many receptors are being activated how does that feed in downstream then there is this component where essentially there's the cell can upper sensitivity by expressing more receptors by if it's not and saying it's it's it's inputs it can increase those proteins are responsible for sensing these things and there's also different if you talk about iron channels specifically and that we work with that there are different versions of iron channels that have different rate constants the same is true for other for active proteins as well which pumps so that becomes a lot more you know explicitly modulating sensitivity over time and their way to you know better react and cope with adapt to its environment again adaptation keep in mind involves prediction in most cases if not explicitly then specifically awesome and by the way the guys Michael Hausmann and Heidelberg before I turn around my great oh that was the person coming to physics and to biology biophysics cool um just to kind of connect that change in sensitivity to some of those mechanisms so if it's the sensitivity to a given hormone or a given circulating molecule um then that sometimes gets accomplished biologically by changes in like the membrane receptor density or changes in downstream signaling pathways so even if the amount of receptor in the membrane stays constant it's like you can release more of the neurotransmitter in the synapse or you could have more or less receptors or you could have more or less of all these regulators and phosphorylating proteins etc in the downstream signaling pathways because there's a a lot of complexity in that space but there's a lot of knobs including rates of change and rates of change and lag effects with transcription factors and um all this other basically anything is possible and then at the organism level just to give one and then anyone else maybe at the some other level like it's kind of like a nest may ant being sensitive to interactions and then a nurse who's a younger ant is like initially more susceptible or more sensitive at due to multiple reasons like at the antenna at the brain in just spatially where the nurse is and then there's this um development towards being a forager being sensitive to forager cues finding oneself in spaces where there are forager cues that are useful um so there's a lot of ways that these changes in the type of sensitivity again are already things that we talk about in developmental biology and so it's so cool to see how that is connected to the molecular mechanism the affective angle and then with the underpinning of the generalized flow yeah dean so you guys can push push back on this hard if you want but it's interesting i one of the terms that i coined was that stress programming or at least programming with stress assumed was basically learning either formally or informally that's what it is because there's a certain amount of stress involved be it positive or negative but i kind of through that through that viewfinder i think it's interesting that back to the back and forth thing stress programming can be safer or it can be riskier and you know as long as it's not sort of killing the learner if safer in the sense of as intervention i think this paper showed that you can intervene but riskier is as distribution like sending something out into the unknown it can be safer as teaching or taught it can also be riskier as unteaching or untaught sort of just sort of throwing somebody into a situation and see whether they can paddle or not the last thing is it can also be seen as both modeling in the safer sense because we reduce some of the variation or it can also be riskier as in terms of coordinating meaning the person has to kind of figure it out as they're flying the plane so i wonder what you guys think did do you think that can we is there some agreement that learning in general be it formalized or informalized has has an aspect of stress programming that kind of is assumed or from my way out awesome question blue or france what do you think um i i don't think you're off it's there's there's definitely of course you know that there's explicit incorporations of stress into learning but your question i think is like is it is kind of always simplicity there even if not explicitly so what we're interesting kind of this blue mention earlier that stress you know can be interpreted as like human perceived stress hormones but can also just be like perturbation from from equilibrium from homeostasis and they actually do kind of this this has been reached this has been combined and homeostatic reinforcement learning where they kind of implicitly explicitly make the reward and reward based you know reinforcement learning about the homeostatic off point and then stress becomes suddenly a lot more explicit in that sense of perturbation from from your homeostatic set point in the interactive inference scheme stress is basically just your you could just define it basically as your you know your overall um difference and again like with those Bayons I was talking about like how much is your is your is your prior belief certainty changing are you minimizing uncertainty over time or not um it is it is hard i think it's hard it's hard both to completely say stress is not involved and without saying that while you're just calling everything stress what kind of it's it's really it depends on what how you define stress but if you think if you go off go of something like as in like stress as you're not right now in an energy minimized state you're not in a an active inference scheme or you're not in a your homeostatic set point in a in a more basic scheme then stress is I would argue you begin just learning yes because without that if you don't have the difference to that if you don't actually if you don't have any drive to learn because you're not you know you think everything's perfect then it's hard to to have learning so specifically if learning is about with the environment in context what is what are you getting from your environment are you trying to adapt to it is that are you trying to learning in a sense of like adapt to it um and find a certain equilibrium with it then I think stress becomes very much explicit and it's ubiquitous in learning okay I just want to add one thing and that's perfect because I always talk about the minimum of two and I think you I think you're absolutely right Franz I think you have to then look at stress as being either voluntary or involuntary I think that's the minimum of two that you have to look at because I think if you have voluntary stress it's it's it's it's measurably different than the involuntary type so yeah you're right I don't think it's I don't think we can just throw under everything under one umbrella and make it a monolith I think we have to partition it right from the very get go and say um all programming be it a curriculum whatever whatever the curriculum is that that those stem cells have that's some of it's going to be voluntary and some of it's going to be involuntary based on the situation that they're they're placed in so again I wasn't trying to pull in that direction but I didn't bring up stress but I really like that it was brought up so okay yeah blue blue sorry so just um from a neuroscience perspective it's interesting at like the voluntary and involuntary stress thing is so different and I feel like maybe Daniel will remember we discussed that at some point like when you're exposed to a stress that like you do to yourself on purpose like your strength training or something like that versus um you know some some externally imposed stress that you didn't volunteer for I think that that has a difference in affect I feel like we looked at a paper I can't remember off the top of my head though um but in terms of sensory processing it's really interesting in neuroscience we actually adapt right so so you can listen to like you're you this is a great example like you turn on the car radio and like you're listening jamming out in the car for I don't know a while 20 minutes 30 minutes and then suddenly it doesn't seem as loud and so you turn it up and then you're jamming out jamming out jamming out and still it doesn't seem as loud so you turn it up because the same level of stimulus produces a decreasing sensation um and that's true in terms of you know visual like when you first turn on the light you're like oh it's so bright but then you adapt and so in in terms of sensory processing we have a mechanism for adaptation um and I wonder about actually cells like not neurons because this is like I'm specifically talking about like a sensory processing way but I wonder if cells do the same thing I mean I guess like you know the receptor shuts or the receptors triggered and then it can't get activated again for some certain amount of time so even just receptor binding in and of itself is like an adaptation mechanism like you can't constantly like dump the chemical into the receptor open the channel or whatever like that's not like biochemically possible but but I wonder if um you know cells have some other adaptation mechanism outside of this um like channel opening or closing or just this is there more than like the mechano dynamics of the channel that um enables cellular adaptation um I can think of a couple of ones um but no one expert on this but um there's what you can think of first like on the epigenic on the epigenic level if you basically just um you know recruit more histones basically have a more tightly packed chromosome you're going to have less genetic expression no matter what kind of signaling input you get to the nucleus that's one that I can think of it um another one is uh mechanically so the right most of the I hope I'm not over generalizing but I think most of the you know signaling against that that receptors are sensing and that has to be transported to the nucleus somehow to to cause transcription that is usually directed right that has to correct of across cytoskeleton so if you change the mechanical properties of the cytoskeleton itself um and uh not again here I'm out of my league but um that's America very well imagine um I do ask someone that's my that topic um lastly it's if you talk in the intercell intercellular context then you know um things like uh you know what we and you know in the normal context you have these you know the synapses in the in the non-neural context you still have gap junctions and other complexes that basically allow transport between cells and those can also be modulated um by the cell so then there you can also get habituation and definitely no examples where where people have shown habituation and uh even other um there's active research on on on showing that there's more just habituation but also actually the learning of a certain parent as well um I think that's true so I'm really interested in the cytoskeleton um because a lot of you know I mean the receptors are in the membrane embedded in the membrane but we don't really there's not a lot of work that looks at like the cytoskeletal dynamics with respect to um like modulation of receptor dynamics right right so like that's a field that I think or I haven't seen very much work um in terms of how does the underlying cytoskeleton contribute to the properties of the membrane and how a cell might be behaving and just in terms of like you know things having to get into the nucleus I don't know I feel that one um I think and it's just because I did very early work like transfecting cells in a variety of different ways and you know to get a gene into the into the cell to get the cell to express the gene you have to get it through the nucleus right how does it get into the nucleus like nobody can answer you so like if it doesn't have a nuclear target on it like how is it going so is it just like osmosis itself through or like like just how is that exactly happening and so you call the vendor like well I'm trying to use your product whether you're electroporating or you know using some kind of lipid mediated transduction they don't know they have no idea actually so it's just a super interesting um prospect and I think that in terms of both that and the cytoskeleton there's a lot of like dynamic remodeling that occurs during cell cycles right like during the phases you know like when a cell is dividing and growing and becoming two cells um there's and so like the nuclear membrane but in terms of terminally differentiated cells like a neuron or something that's not undergoing you know constant change in remodeling it's curious to see how that can happen um and I'm interested if anybody has any any good like hardcore mechanistic papers that you want to point out to me or if anybody listening in the live stream wants to drop them in the YouTube chat I would really love to take a look at them. Hill I really like Dean returning to this uh idea that there's the there's a continuum or there's multiple archetype dimensions so not that it's all going to map on to just two but we know it's at least locally going to be at minimum to that there's the kind of uh one mode of interacting in an educational or training setting with a teaching and a taught that could still be in an interactionist or an instructionist way versus the unteaching and the untaught so that's the the difference between the athlete having their form like observed in a tight feedback versus when they're on a sojourn away from that sort of a training context and so there's like an environment or there's a setting where there's some kind of stress as fundamental and then there's another type of environment where the different kind of stress is fundamental and then um Franz it was an awesome point moving from that basic homeostatic framing where stress is usually considered at the first derivative or the state so it's like for a thermo organism thermal stress is going to be when it's hot or cold and then that's going to be like a bowl or it's a v or it's a bathtub or it's some other thermal stress curve it's about temperature moving into the generalized coordinates of motion with temperature and change of temperature and change of change of temperatures all the higher derivatives of temperature that expands the space a lot because maybe it's possible that a temperature that you slowly reach like over 10 years is a different amount of stress than just jumping there instantly time matters and so the rates of change captures that time dependence in an instance with a snapshot vector so that's sort of the formal use of how we can um look beyond just the stress on temperature and then we introduced the whole cognitive stress because he said well for the active inference entity this is sort of like the first resonance and this is the one whose violation of integrity results in physical death but then stress moves to higher or different analogous settings and that could be how fast am I learning and as pointed out by Dave in the chat and by Dean and others like it really matters whether somebody has agency or affordances to change it and so my closing point there with the voluntary involuntary is like grad school and stress because or any research or any educational environment but one that many people experience is like it can be stressful it's also a very often privileged setting to be in and one is taken care of sometimes in a way that even like neighbors won't be and so that is something that like I saw first and second hand and just thinking about the stress it's like why is it stressful to just read a few papers some days or something like that so just kind of interesting with what is voluntary and involuntary how we commit ourselves to different kinds of stress of what kind it's like well I wanted it to be novel but not that novel and how we actually make those action selections in uncertain environments blue so just to add on to the voluntary versus involuntary like grad school is an entirely voluntary endeavor obviously but like high school is not and like I dropped out of high school at 16 like as soon as I was old enough to drop out I was like I'm out like I'm not doing this compulsory regimented program like sorry um no regrets right like I started college obviously have a phd like did well but it's the difference versus involuntary versus voluntary like when it's in when it's compulsory it becomes torture like I mean just my just my two cents we are just in this question about computation group recently about you know the the this whole idea of like how do you quantify any kind of agency and uh uh Josh Bunger in that in that context works with us brought up this idea instead of talking about agency which is how to quantify you can talk about empowerment and in the computational sense empowerment is about how much do your actions how much can your actions influence your sensations again this makes very very clear in an active inference scheme right where you have this marker blanket so to give an example right if you if you don't there's no actions you can take to change it that's you know we can think about as natively stressful but we can also we can also define that fairly clearly in a computational sense using something like empowerment so it's it's uh and that again I'd like that you brought this up because this price home the point that a lot of these fundamental things you understand about stress and human contact with anyone that's ever dealt with depression or people that suffer from depression where it knows that it's never really about the the the objective total amount of stress but but how it's being perceived by that person and I think this empowerment helps a lot this definition of empowerment helps a lot to understand that better how much control do you have of that there's other things that of course we enter that but it also makes them fairly clearly that the same a lot of the fundamental principles must hold true and and lower organism as well because these definitions are of course you know as soon as you prescribe actions of course you can argue that there's no real actions and and and cell biology um then that's okay that's just then don't want to talk about that level but if you do look at active states in a in this other context then you do get then you can fairly simply in an active influence scheme but other ones as well define empowerment measure empowerment over time by looking at how much how actions actually change the sensory states of the of the cell and then definitions of stress from that point of view all it's become fairly fairly explicit can I ask you a question yeah can I ask you a question because you mentioned in the when you were with us in the point one for a bit there and I don't know if you re-mentioned it today you said you went to mark I think and you said I want to do something cool we're going to bring this back to the voluntary and involuntary piece and then he and then you I think you said I think you said I'd never heard of active inference before I was pointed in that direction to Carl's lab so in the in the in the context of what's voluntary and involuntary when you when you when you said I want to do something cool that sounds like you wanted to do something voluntarily then you got then you got this deep dive into active you push that up against your biology background which is something that you've obviously done because you like it right and you wanted to finish off so so going forward now that that voluntary piece that got you into the active inference what do you see the active inference doing in terms of this empowerment question because Daniel's rid it out here with a question mark behind it so how do you feel more agent agental agential sorry I always kill words because I'm not very anyway I'm just curious what you think now in terms of what sort of empowerment so you've sort of become in the in the sense that now that you've got this sense of what active inference could do and what some of these math formalisms provide and what the statistical world bridges yeah what do you what do you what do you like what do you think in going forward in terms of the reapplication of what you what you pulled out of it out of it out of a different bucket and and then it's set down in the bucket that you're most familiar with see me like what we're put out on a personal level like for my life decisions yeah like when you're when you're going to be going forward and trying to and trying to use that active inference sense in terms of some of those decisions and what kind of environments do you think you see yourself projecting yourself into that you may not have projected yourself into prior to sort of engaging or encountering or exposing yourself to active inference that's a question um I think in the end it's uh my wife and I we both love this word word serendipity which is the concept of happy accidents which there's all kinds of connotations to it by an active inference scheme or any any you know when you look at you know what what is the information from you what the idea is that you if you have too much if you have so it's going to be active inference right now i'm sorry that if you have super high precision you're very strong prior beliefs then whatever you if you are confining yourself and you don't allow variances um to to change your mind then then you're almost always going to be upset because you never you know your precision is set too high right there's there's a paper coming going to be coming out soon about this collaborator as well looks about precision in this context as well so that's something that you know any active inference modeler will know that you have to you know keep that in mind when you put your your your your precision for me personally um it it really is an idea of like I think every interest this is not not another idea by myself by any stretch but any any human tell stories about their environment but most about themselves and the storytelling is in my sense nothing else than doing a model an active inference scheme making a generative model and setting that so I think too for me to I'm way too young to be giving life advice but for me to to to achieve any kind of happiness you have to really be careful of uh what story you tell about yourself and how strictly confine that on things and there is definitely advantages of I think personally I've always had better luck personally I think by by not not letting not to find that too too strictly and really looking into what what your environment throws at you and I think that helps and I know personally people that have a very hard time with that and that they're very strong modeled so if you know I'm gonna you know get married at that age and then I do this and that and then when things don't happen that way I think at least needs love and happiness so if you know in that context active inference what can I do for you I mean I think it's it it makes things explicit that we all kind of hear about some level and you know I don't know let me confirm psychology biologists you know folk psychology things just people giving you advice parents saying I think oh now this is I don't want to say that you need any of those insights I think you can get this lots of ways but what I like about that inference scheme that it you know it's it's putting this dimension it makes makes the it makes the also biology the biggest advantages you get a very explicit structure about the information flow for very simple systems and because it's been primarily applied in neurosciences you get in for us intuitive results from that what happens if you have to have precision what happens if you probably don't change what happens if you know if your energy isn't being minimized all these things I can fairly easily explain that in the context of of of human behavior because on this point we all have to understanding but I think there's for the biology point of view there's there's a lot of things we can learn from that if you look at the fun the mathematical definitions then of course a deal of the baggage that you get from the neuroscience which I have to deal a lot with which is fair but for your life I think it just helps you if once you see really how how much study has been put into misinformation you know amount like bad inferences and of course people doing tests in an active inference scheme about you know psychology like behavioral tasks states um then I think that that's one lesson I I can definitely learn from that if I hadn't already is that be mindful of be mindful of your priors be mindful of the rate of change like how what are you really stressed out about right now what's the worst uncertainty coming from and how much should that throw you off right and this is in a sense to kind of being all this home a little bit the biggest limitation of of any algorithm of any model is that you are defining the general model beforehand not just an active inference scheme but any there's there's papers that come out that you know criticize the idea of artificial generalized intelligence because you always have a predefined model and life on that sense but that definition is does not do that life always you know biological systems evolution kind of generates bunch of diversity in its background and from that kind of they pick the best path to constantly evolve a generative model so you know that's kind of like limitation of active inference any other algorithm which can be overcome by you know chain different algorithms together by evolving system by having models make models there's some cool work on that too but that's something you need to take I think both computationally scientifically what's personally can take a mind of like at what point do you change your own model and allow for flexibility that that's what makes human right now still that's one of the stuff that makes humans superior I would say and to any to any kind of algorithm right now is because they all come pre-defined with a lot of the hard models which again it's I don't actually I'm saying this now and I also don't actually read that 100% I don't actually believe in fundamental differences between different intelligences but that's something that at least kind of intuitively makes sense to us we're not that's really true scientifically I have my my my points that but that's going off off the rails awesome wow a few things like we often talk about concordances with different systems and it brings the cognitive apparatus that was developed for humans and starts to at least instrumentally if not from a realism perspective also project that and so it allows as a transdiscipline people to talk about systems and map their analogies and models in a way that would help us find the resonances between cellular metabolism and economics kind of taking a vague feeling like those might have some similarities into using the same notation terms forms etc so that's one very interesting piece about active inference and also the limitations in terms of at what order and how much time and attention is allocated to this structural aspect and how many different families of structures are provided on one and then to your well communicated uncertainty about if and how biology is different from other kinds of processes that's an awesome question especially connecting back to your point about the reservoir of energy versus how locally there can be a different energy or entropy balance so it's almost like there might be little pieces that can be mapped to digital signal processing with super high fidelity and that'll be taken as like a win for the realists and then there might be other parts where the instrumentalism bridges because we're just making a model of this one smaller thing and this one larger thing and then in cases where this is fully computationally defined or fully described then there's like a case for a real what does empowerment look like at maybe like a lower organismal level or even like a cellular level how like is cellular behavior or lower organism behavior like voluntary versus compulsory and like is there a way to measure that like obviously like you can trap an animal in a box and frequently like you know we'll try to try to get out but if you don't try it would it go into the box and hang out there like like does that happen maybe I don't know does anybody know of any examples of like what empowerment looks like at like that's just it's a super interesting concept to me in the in the sense if you take that definition off um the definition of how much your actions change your sensory approach your perceptions then and definitely definitely you I always I would it's if you want to have a definition it's of course built on what your different actions are and that is fairly easy and because you don't have to go into this route what is voluntary not you just basically something external state you're basically even simpler you look at the the the flow of states the mark your mark of blanket which you can show forms naturally which color is shown and then from that point if you want to have active states then you can measure that fairly fairly easily so as example I would argue if you know if cells actually interact with each other and really change other cell types you have here's a great example moho genesis is traditionally much I think it's pretty fairly favor established it's really driven by a lot of organizing centers always a substitute of cells that we try to sell the ones I would argue then that those cells are probably very much empowered because their actions cause differentiation of cells around them which then will very much and they're going to do a certain expression of self well so then that will they have action very much control that whereas those themselves um yet the receiving and have a low impact because they're um their own signaling probably isn't really because they're not they often just don't have that's that they don't make a lot of that signaling so that they it's being organizing all of them so they are not going to have a low that's I think of that you're asking me or anyone yeah what as our guest or anyone else is welcome to do well there to go I personally like to get to where there's more than kind of qualitative like the making making actual like instead of just like making these after models of of active inference uh you know we can measure on hand if you can really map those two together right like in the sense that I mean a lot of work went into this whole SPM software that they developed in uh in London at this point right you actually can get fmri data and then you can do dynamic causal model and you can really you can really postulate different causal models and then do a you know a Bayesian belief updating you can do a hypothesis testing right in this in the sense that which hypothesis is more probably you'll have cost that you see um if we get to that level in a in a morphogenesis exam you can get that high fidelity images of the developing embryo that you would get over time get signal so you can measure and feed back into your model and really make really test the past of what's cause a model like what structure would have cost these sensations that you or the data that you're receiving that will be a huge step forwards and microscopy is really doing some uh having some great advance on that one example I can give us a friend of mine works on that Dimitri Krum and he used to be in Heidelberg and he light sheet microscopes there's a lot of other groups I mean he was a master student phd student um so about uh that from him but there's groups um you can do high fidelity um light sheet from two sides really nice like near confocal resolution um but with high high temporal resolution um of their organism doing more and recapture the entire process of you know from constellation further on that's you could collect where you can really start to make past like our cause and models in an active inference framework and then they're basically we test where's the outcome there and you can retest and uh this this this the idea that we already have as incident I mentioned in in morphogenesis and we test out how much that that fits into our notions of of functional uh delete updating in actions um I would I would love to see that in my lifetime wow super awesome so just to kind of restate that um SPM has been developed in the capacity of behavioral neuro imaging and so it allows a combination of observable features such as like what button somebody pressed or what image somebody was shown and also observable features of the environment like the measurements coming in off of the EEG or the fMRI or the MEG and it allows integrated modeling of those different kinds of parameters along with the dynamical causal modeling on the underlying dynamic system identification and there's a lot more in the SPM documentation on that and so active here along with some advances in for example the ability to look at the process of embryogenesis and track cellular location for a whole organism developing through time then that can be fit as if those are the fMRI measurements of this embryo SPM and then the underlying mechanics of development like a hypothesized morphogen gradient or a known or hypothesized mechanical relationship that's part of the model that's like the underlying neural systems model in SPM and so it's kind of a great dream but if it can be stated if it's likely to exist it sounds like you're taking some actions that will reduce your surprise about it yeah pretty cool so on the microscopy note I'm like getting ready to receive and set up of IO AFM I'm really looking forward to like the multi parametric capability of like I can look structures top a lot also like due fluorescence imaging so I'm really excited to have multiple affordances available on this like the my chances are great like I've used an AFM button you know it gives you the mechanics of her sample but not not it's I've also been able to look at the sample you just the you know the mapping without being able to see it so I'm really excited to be able to see and measure mechanical what's going on in a cell but that's outside of skeleton parameters that was discussing earlier about how that you know manipulates or enables cells to do different things cool let's look at any other things or also we can sort of give a closing round to whatever people feel like so one one thing I'd like if I really about the aberrant signaling figure so I don't think we went into figure five very much last time in the dot one and maybe you could talk about the aberrant signaling rescue kind of unpack that for us in your own words we're not seeing the slides anymore also Daniel I don't know if you know that or I'm not yeah it's also kind of lagged on YouTube so I'm recording it I'll reupload like the high quality one it's kind of like weird it's a normal video chat but the YouTube stream link is like disrupted but say lovey um yeah so actually if you don't see the I'm showing figure five even though you can't see it so just oh yeah or I see okay we can do that too yes that's great thank you awesome yeah so the noise so this is the the top row here's just the norm simulation this reference right you start off with these eight cells from the beginning and then kind of have you know this vague they're very specified by low prior belief of what type they are and as they kind of out that density changes so the the hue tone here makes them means that they're now more the more whether they are the more sure they are of being at this head type type cell or in this case the green that the tape and be where I base one one cell which you can see kind of heating off and with the error points towards that had a reduced sensitivity to element um two six basically had a really hard time basically um first of all you know it was it was much more was longer kind of un-specified than its neighbor dolls were and in the end it achieves actually it gets wrong cell type way it's not not only is it at a wrong position like there's no cell in the original one at that place and even were you expected to be either yellow red or somewhere in between but not blue as it is down here so um in the sense the standard initiation of a of a cell of a cancer is I turn this that way because I think it's basically the first step in cancer initiation is a mis-identification of your environment and your your place in it right the original paper from Carlos called knowing what knowing one's place the cell definitely does not know its place so that this is where it goes in and then here is where you have basically um the same as above but the overall not just the absolute flow of actually signaling like two one cells is to get to that has been up regulated so that each cell basically now drives more concentrations of cells to that cell and basically kind of tries to the idea of to compensate for that that all the sensitivity and what happens I would expect at all is that you'll have um if a cell contracting with it more here so you don't see that so much yet you see the cell that kind of initially we've normally gone gone down more cell you push out in the end here what happened the words is that we're put into more cons I know of of signaling concentration and allowed arrangement allowed this this cell to re-take a place that is conform what you say in the title morphology and what I also tell you I always wait when I when I talk about this figure and both figures really is you do not in the simulation at any point in time change target morphology all the actual encoding was exactly the same so what they're expecting and where they're expecting us was exactly the same and that I think is where I see the the fundamental promise of this flow that you don't always want to change the code that we we know that we have higher value to have in gene therapy we ever had and yet I think anyone really thinks that this is the way forward in all cases and in all you know if if you can do it and you have really complete understanding that's great and you have no other options and an extreme case that is absolutely very warranted I'm not trying to dissuade anyone anyone from advances in gene therapy but I think it's you know there's the genome so complex for me things that can go wrong we don't know all of the details that if you did not have to start messing up people voting I think that would be preferred and I think this is and also from a even just from a capability setting you know that you don't always have the capability not with high fidelity that we think often do so our lab as well I kind of take at the core of my experimental approach to things is that I want to understand how cells interact with their environment and all starting them and work on that manipulation basically control information and sensing and drive them towards different dates again with that Wellington unscathed that we talked last time about you want to direct the the flow I do actually this time have I took my shirt and you can see that right now I do have the one landscape they are my shirt this time I remember that um so that's kind of like what because like you're trying to direct the flow but tell fate um decision-making not try to manipulate the actual basins themselves so what was the most perturbed signal response oh sorry okay blue there yeah the perturbed signal response so what was yeah what was how did you perturb the like did you cut it off did it get less information or did it like was it supposed to like infer an incorrect place like how exactly in that one cell did you manipulate the parameters of that one cell so that it went to a different place yeah basically um let me see I think that was was basically done um the and uh it was a decrease in sensitivity is that what it was I'm trying to make the specific equation right now instead of hand waving the so it's basically in that we also formally induce right so that that's basically the the how how much of a right this the the how much actions are being updated based on sensory inputs um and we can we can put a gray on then the sign on there of how much they they act on that that is the part and I manipulate here so but that basically means is that by altering the unscapability of the cell how much how much sensory update up how many how much to what extent sensations are being updated based on external information flow and what kind of you know how on top of that that by that you basically are being or reducing the sensitivity to the environment you basically you're you're misarranged you're misregulating the again the empowerment essentially but that's that's downstream but you know you may have the the potential of the cell to act with environments and do anything about and also even upstream how much sensing of that so in this case specifically it just wasn't really sensing as much of its environment as it was supposed to be sense yeah um it almost is equivalent to the argument or the dissolution of sort of nature and nurture partitioning by just taking a complexity stance kind of like Evelyn Fox Keller in her book The Mirage of a Space Between Nature and Nurture you were talking about a target morphology but then about how changes in the signal we're about to do number 40 on the quantum FEP paper and blue and our friend Jason and I have been preparing a lot for it and I think separating quantum from the electrons and their behavior and just asking well what is quantum strategy or what does this quantum statistics or quantum information projected onto biology mean not how it's been approached from the mechanistic question which is quantum biology well that's about the synapses and their quantum mechanical properties or about photon tunneling or proton tunneling and like quantum effects but what about just using the statistics instrumentally of quantum and talking about situations with with complex patterns of uncertainty and observation bias and memory and modeling and all of that and this was like a dot too that kind of like opened up a whole new area for us in our discussions with morphology and like the spatial and the embedded and and then cracked a window into just another area we'll go into and you brought so many awesome insights so we really appreciate it yeah that that talk should be a real pleasure how you recommend doing it Chris feels it's great to listen to and it's some exciting work and if you're also if you have seen I'm sure you have seen that the the particular paper from Carl Frist and he has a section there on quantum mechanics and it's a it's a very cool section because he actually derives I think the Schrodinger equation equation from from a essentially from my to inference point of view from just a definition of what your log probability density is very cool kind of uh you know I don't know how much that will feed into other and future work but the it's it's always very exciting to me when you see different fields at least formally being able to be related to each other um that usually means there's something something going on something fundamental we'll have more to say in dot board so and join us yeah and if you're welcome anytime to participate yeah it'll be Chris and Carl and I think Mike I think everybody's coming so if you want to come and join the panel you're more than pleasure all the events and pop in if it works so thanks again everybody um hope that the viewers can bear with the lag if it was there but um say love you so peace out everyone thanks a lot for this great discussion thanks for having me bye bye