 So I've been, you know, I've been a provocateur in the past and this shall continue. I'm gonna propose that the brain is basically consists of something slightly larger or more elaborate than just the neocortex. I mean it would be sort of interesting perhaps to to ask the question. I mean we have high throughput now technologies but while I did see this striatum though it wasn't really discussed in your slides Mike and I did see you know a bit of I think I thought I saw some some thalamus potentially at least listed which are nice nice structures not cortical but I would like to throw this question out there you know let's say that my belief system is that the cortex is the umbrella of the actual brain. Take it from there you know take me down tell me why I'm wrong. How would you take this one? Well what strikes me is of course that yes there is a challenge in neuroscience sometimes that it's it's about the cortex and then one one cortex. Well hippocampus also is very you know popular and amygdala and few other structures. And then you have a cerebellar community who owns most of the neurons right in the brain and we have a lot of very strange things going on in terms of how the brain is organized how it is connected what parts are connected with very very dense projections and surprisingly thin or weak projections between enormous regions. And these are very puzzling questions that I think you probably are thinking about in this regard. What we really are missing is the brainstem because there are very few research on the brainstem and this is such a crucial structure for homostasis and keeping us alive that if you talk with neurologist they say okay the death is not just the cardiac arrest it's just the brainstem stopping functioning. That's how they define the deaths. Okay so yeah so in we are too much focused on the external part. To defend the cortical people I guess we can say when your cortex makes us human so but at the same time the brain has a number of functions that needs to be understood as a whole. But it also makes the macaques macaques and marmosets not quite true actually because as we know the this limbic structure are crucial for the typical behavior of animals. You've probably seen that you heard of these results of these really remarkable studies that people do that you know that without input from lower brain structures you can't really make decisions right. You can't in fact you may have met people like this they probably don't have those you know but that you can't you can generate alternatives but you in can't in fact break ties and that requires a form of emotional lower kind of thing and that's quite remarkable I mean. There are also a few papers showing that people who don't have practically cerebral cortex actually still can behave in many ways which just when you look from outside you think that they're normal people. Although many functions are gone. Zombie agents. I also mentioned in Japan then for the the large-scale brain simulation projects they're interested in not just the cortex but the interaction between cortex subcortical structures cerebellum and in terms of data structure this is also another point so if you can if you consider that you can lay out the cortex into a flat 2D surface this is very nice for distributed computations but how you deal with them and also because of the the large number of neurons in the cortex and cerebellum so how do you actually organize the data structures I guess some subcortical structures are nuclei they're less neurons this is not so much of a challenge but you also have to consider that so that's just from the simulation perspective. I might add to that that in terms of map transformations in the brain as you say the cerebral cortex and the cerebellum can be dealt with in a kind of 2D fashion whereas the intermediate stages are 3D and we worked on that with local coordinate system as being very helpful and I think these type of mappings are also explaining a lot about what happens. And then we have a lot of small nuclei in the brain stem which are actually providing lots of neurotransmitters really changing the way the brain works so lots of things to be done. More questions please. Thank you very much. I would like to ask you about this quite old structure by Urya it means you know this name Urya who divided a brain into three blocks and this theory is quite well because we someone mentioned about hippocampus someone mentioned about neocortex and we still I think we still should work in this scheme because there is a so-called neuropsychology and also this neuropsychology is a neuropsychology linguistic it means this is a linguistic psychology on the level of a brain coming for example from Chomsky idea of grammars and all of these linguistics problems because really brain human brain is in the moment where we are able to construct a language it means we have a linguistic for example in Chomsky idea and what is the now what is the state of this investigation in the simulation of brain working also well I my general comment to the question would be that there are many such schemes during history that are used to let's say at a very high level create theories about brain regions and how things are connected and how they work and you use different languages to to then tell a story about the brain I think that's that's fine but what's probably now interesting is to look at how all the different observations the data we have map on to such schemes and then maybe some are useful some are not I think most of them are absolutely totally unusual or no use in my opinion I usually do not think it's helpful to parcel it brain into two strict regions it's basically made up of the networks and and very distributed functions there is a lot of work on neuroscience of language of course and people are doing really very nice research also looking at neuroimaging etc so but I haven't seen anyone really linking that to Chomsky but I don't think that the Chomsky is relevant to neuroscience in a way I think it is because Chomsky idea is really a Chomsky linguistic is supported by psychology by these grammars and also about something which is called a fundamental grammar yes well in popular literature there is lots of you know descriptions of the sort I'm talking about neuroscience literature and what we are talking about here I haven't seen any paper in neuroscience journals linking the neuroscience of language to Chomsky so that's the fact right the fact that psychologists talk about that and other people talk about that doesn't change that yeah next question please that is also neuro psychology linguistic by Uria there is a quite big that that was a century ago we're not discussing so so the brain develops over several weeks or months from a neural tube and expands out and you've gradients of gene expression that determine cell identity based on like modeling of that process about gene networks and and and gradients expression is there any theoretical or simulation based result that would suggest that there should be clusters issue belt identified I mean because like it seems like it's a it's a question you're asking with data which is which is which is great but is there is there any like good like model that would suggest what you should find when you look at that there's not much I don't think much theoretical work there that that's thing I don't I don't also have a I have a sort of a I'm a skeptic kind of in a way of things in general I mean I kind of like like all we you see groups okay you can do all kinds of fancy things you can justify groups statistically but the question is do they really do something and I mean I have this weird feeling that I wanted I wanted to kind of see if I could the idea occurred to me to see if you could prove it some kind of a theorem like this that in a sufficiently complicated here I'm gonna this is very informal here's a challenge that I think will be funds right in a sufficiently complicated system with some sort of birth death process or whatever etc. and certain rules I don't know what they are yet that the system if you and you evolved it using that we're gonna bring back genetic algorithms you evolve it just like a gene we're using some kind of genetic algorithm and I'd like to argue that that the types that result that inherently there's going to be a bunch of pathological types that don't do anything in the system I want to claim that any system evolving according to natural selection would produce such a thing now maybe that's nonsense I don't know but anyway I think that's what do you think about that well one step back I think the question is the right type of question to ask with regard to all the data we are collecting and for example now with Mike systems you you can hopefully go back and actually ask questions along these lines you can investigate over time just some other topic where development plays a role consolation of the brain areas sub areas and so on development helps a lot also topography projection systems popularly now given all kind of fancy names such as connectoms whatever you have but they also follow very often very developmental principles and if you trace back how they are formed you get back into the story you're aiming to and I think again the data sharing part of this having access to data doing the dinosaur story over again that that's what we want to have in the in the in the near future all right any other questions around yes please so I mean this is exciting work and as someone who's always kind of done several base type models so it's really exciting to see all this data coming out and so I mean I guess what I've read and seen before I think it was in Sebastian songs article about the cell types have to be associated with some function or you know circuit and this is a challenge right so at least if we kind of could be clear about as it's going to come out of this data then you know mapping it on to function right so you might you know see a VIP positive cells but there's you know ten different types of those so you know but certain types might be active on the elements moving and stuff so it has to kind of experiments have to kind of go along with you know being able to know that you're talking about that particular cell type that can be identified during during you know whatever behavior and then those are really challenging experiments right but at least if we have the data that you guys are producing and sort of classifying then you can sort of ask these questions more specifically so that's what sort of in my mind really exciting right so I mean the cell type will come from that combined work is you know so any of these sort of development questions new or modulation questions the state of the system right it's all going to depend on that right but at least if we can sort of be clear about you know the genetic you know all those transcriptomics and at least that's kind of like a base start that's kind of how I view it is that fair to say or as someone who's not an expert in you know this is reasonable yeah okay thank you anyone wants to comment on that that was a long comment no opinion I think it's it's well I mean one of the most interesting questions at least for me is at which level do you think we will understand how the brain works in general terms which would be for example useful to well create a better artificial intelligence systems and other things like that because obviously when you look at the cells it's extremely important to pharmacology to our understanding of new molecular medicine etc. I mean what goes wrong with the particular cell may have a very particular expression in all kinds of mental disease etc. but maybe this is not the level which we really need to understand general work of the brain so what do you think will be the proper level for a description that we still could use words for that not just huge datasets right and and be able to have the feeling that we roughly understand what the brains are doing and why they're doing it this way Marianne and I were talking about that something similar the other day is that you know I wanted you know depends what your motivations are right I mean there's at least two big motivations let's say there's many possibly many for studying the brain but one let's say the broadest ones for me is that you want to understand how how people think and you want to know potentially how to build something even more sophisticated let's say that's one motivation and the second is disease you really want to help human suffering and human disease and and I'm not so interested in that well it's just fun motivation I find that it's too expensive and it's too hard to just there's other ways to have fun so I don't really buy that that one so much and so for me the two biggest motivations are probably those two and you know take for example it may be that like I was saying it may be that the I mean the whole molecular context of the brain only gets you so far in terms of understanding how you think right and and it's intimately connected with the disease I mean that's unequivocal but I just don't eat the question is could something much simpler than what we are do as well I mean that's one question so I think that you have to we have to understand what you're interested in the question and acknowledge that that understanding the brain means different things to different people right here so often that we don't understand the brain right and then some of us think well that we do understand yeah yeah understanding at very different levels and I suppose in modeling you refer to multi-scale modeling has as a key aspect there that it's really very different and one of the challenges there perhaps of some relevance to the question is actually to connect those modeling efforts also at the level of tools and simulations that's some not directly what we're discussing but some people think that's a related question okay so on this fundamental note Alex has the last word and but you will just close the conference I was just gonna add that understanding intelligence I think it's not just about understanding the brain so you you can extend this out to understand the entire organism organism and so on so I think yeah that that's also very important