BI 239 Nedah Nemati: Naturalistic Neuroscience and Lived Experience

[00:00:03] Speaker A: When I got really interested in this, it was always from that frame of, where's the person in all of this? Right. Like, that is actually essential to the explanation and story that we tell. We're now trying to automate science. And so there are all these efforts now to actually just have AI Become the scientist and do the science. And so is that still science? Oh, the neuroethologists of the past weren't, you know, using these really sophisticated tools that we have now. And we're doing a great job. We're building on it. And I think that is just what that's getting wrong is the idea that the behavior that was being studied then is the behavior that you're thinking about now. So this is. This is the issue. The tool theory stuff comes from this tradition of thinking about sciences where the theories were really specified. And I'm not really sure that in neuroscience, we're navigating theories in that way. Right. We have these ideas about how things work, but they're not as solid. And sometimes they come also from the local research communities in which we're situ. [00:01:27] Speaker B: This is brain inspired, powered by the transmitter. Hey, it's Paul. What are you doing, like, right now? What are you actually doing? Are you on a run? Are you riding a bus somewhere? Are you working out? Maybe you're laying down, listening to the sweet, sweet sounds of my voice. Seriously, though, what are you actually doing? What behaviors are you doing? What is a behavior? Neuroscience studies, in part, the relation between brain activity and behaviors. But what is a behavior? It's a simple question, but there is no simple answer. For example, you know you're behaving right now, whatever you're doing, even if you're not doing much when you cross the street, how many behaviors do you use when you sleep, what behaviors do you do? Hopefully, these simple examples will make you think a little bit about how difficult it can be to call something or some set of things a behavior. Neda Namati is a philosopher of neuroscience at Columbia University. I met Neda at a workshop a few months ago where we chatted about the growing trend in neuroscience toward what is sometimes being called naturalistic neuroscience, which really means varying levels of allowing organisms to behave more freely, less constrained than traditional neuroscience experiments that seek to minimize everything that's unrelated to the behavior or cognition that you want to isolate, to study and explain. In more extreme cases, researchers will try in the lab to emulate as much as possible the ecological world that a particular organism has evolved to exist in or even perform experiments out of the lab. In the wild, so to speak. So a good part of our discussion today revolves around this trend and what counts as a, quote, unquote naturalistic behavior and how the tools that we use to perform experiments shape the experiments themselves and the scientific questions being asked. Neda has a neuroscience background, but in her philosophical work she has embedded herself into various neuroscience labs to better understand how the experiences of the researchers themselves, called their lived experience, shape the assumptions and questions in their science. As an example, we discuss her work looking into the neuroscience of sleep from over 100 years ago to today. So when a modern neuroscientist studies sleep, are they studying the same thing that a scientist claimed to be studying 100 years ago? Or as Netta argues, are these two different versions of of sleep incompatible in the scientific sense? Along those same lines, we briefly discussed Netta's newer project exploring the role of AI as a tool in neuroscience. I linked to her website in the show notes@braininspired co podcast239. I also linked to a piece in the transmitter that she co wrote related to the AI topic. That piece is called beyond the Algorithmic Oracle Rethinking Machine Learning in Behavioral Neuroscience. And I link to all of the papers that we discuss along the way. Thank you for being here, whatever it is that you are doing. Seriously though, what are you doing? Here's Netta. Maybe the first thing. First of all, hello, nice to see you again. I don't know when that Population Doctrine workshop was that we met, but maybe a year ago. Was that right? [00:05:08] Speaker A: I don't think. I think it was in October, maybe. [00:05:10] Speaker B: Oh, is it October? [00:05:11] Speaker A: Yeah, sometime in the fall, so. [00:05:13] Speaker B: Okay. [00:05:14] Speaker A: Yeah. [00:05:14] Speaker B: Well, it was nice to meet you then. And we got to talking. We ended up at the same table and we got to talking about naturalistic, quote, unquote, neuroscience, among other things, and some of your other work with lived experience, et cetera, that we're going to talk about today, but I thought that we might start like, selfishly, because in my future lab, should I have a future lab? I'm really interested in having some sort of philosophical component to it, and I don't know if that means bringing in a philosopher into the lab, which is what you do. And so I don't know what that would even look like. And I don't even know what that's called. What is it called when a philosopher goes into a lab to learn about the lab practices? Is there a name for that? Kind of. [00:06:01] Speaker A: I'm not sure there's a formal name for it. Yeah, there should Be. Yeah, there probably should be some kind of designation, but, yeah, so I just say I'm a philosopher in a lab, so I don't really have a term of art for it yet. [00:06:15] Speaker B: Well, that's what I want to know. Like, what would I be getting into? Right. Because I can imagine a scenario where I bring in a philosopher and they're nothing but trouble because they tell me all the things that they nitpick and tell me all the things that I'm doing wrong, or they give me the historical context of why I shouldn't be pursuing this or that or asking the wrong questions. Right. So what is it like being in a lab, and how did you come to be a philosopher in a lab? [00:06:40] Speaker A: Yeah, so for me, it kind of started a bit organically. It didn't start as though I went in intentionally thinking, oh, I'm trying to be a philosopher in a lab. So I was really. I was trained in a certain area of philosophy called continental philosophy, and then I then received training in neuroscience, and I wanted lab experience. So I started working in a lab. And I really was in the lab as a scientist. So I wasn't trying to go in as a philosopher when I started out. And one of the things that I got really interested in was just kind of recognizing in lab meetings and in conversations with my peers when we were doing the science, how much of the science was being driven by things that were not being codified in the actual process of recording what we were doing. And that, to me, is when I kind of started this, as, you know, just privately, I would take notes. It wasn't. I like to say that I had, like, a separate lab notebook just documenting what was going on. But it more was like voice notes on my phone and scraps of paper and things here and there in conversations with people that I collected over time. And that's sort of how it started for me, was thinking about things a little bit differently from how we were just kind of typically talking about how ideas were represented in these kind of more formal contexts. And then that evolved. So now I have a kind of procedure for it in some way. [00:08:21] Speaker B: And that modified your own. [00:08:23] Speaker A: Yeah, exactly. And so that kind of started when I started my postdoc. So I was hired to kind of, you know, enact this practice of being a philosopher in different scientific communities. And I will say that it doesn't take a single form. It really differs depending on the PI and the culture of a lab. So if you were to hire a philosopher, probably it would look very different than another kind of lab. And I would say, you know, I like to think of what I'm doing as not necessarily being voyeuristic, like observing things and then just taking them off from my own little projects, but actually trying to contribute meaningfully to the questions that are happening on the ground. So even if there is this aspect of critique, I try to think of it as something more constructive than simply like, oh, you're doing something wrong, and here I'm going to point that out to you, [00:09:30] Speaker B: but you're not, like, sitting there whispering into your voice notes, like, and now he reached for the pipette, but he did not wash his hand. You know, whatever. There's gotta be a little bit about. Of that going on. But I mean, from my perspective, right, if I wanted to bring in someone like you, it would be for a generative, collaborative kind of role. It wouldn't. Although I guess there are some people in philosophy that's. So maybe there are two ways of doing it. Right. You could be like me, if I'm a PI, and seek out that. But then from the philosopher's perspective, they might want to integrate themselves, and so they might actually seek out the lab itself. And so that dynamic might make a difference. Which are you? Which have you been? Both, perhaps? [00:10:21] Speaker A: Yeah. I would say, again, it's sort of transformed. So the first couple years in my postdoc, I was thinking about it in terms of the lab as a whole, my interaction with the lab as a whole, and not necessarily thinking about it in terms of, hey, I have this. You do have a set of questions that you go in, trying to think about and answering. But it is also an experiment for yourself, like how this is going to play out. And the. The thing that you want to be really aware of, I think, is also respecting the hierarchies and the relationships that people have with one another. And one thing to say about how I do this is it's interesting that you say, I'm not whispering in a voice recorder. Somebody's moving their hand and doing this thing with a pipette. But I do actually take field notes. I do write down stuff while things are happening. And I recognize that my presence is also mediating changes in these environments. Right. So it's really. You spend a lot of time navigating that, understanding what role you're positively playing in these contexts. And so I was doing a lot of that in the first couple years, but over the past year, I've transitioned a bit out of that because I started a different project where I've been looking at different Tools used for behavioral analysis there. I have to go into lab environments that I can't really embed myself in for a very long period of time. I will do semi structured interviews with different members of the lab there. I don't also discriminate. I try to interview anyone I possibly can, but then that takes on a more formal relationship because I don't have time to get to know people as people before spending time with them. If I want to learn something about their experimental setup, it has to be very, you know, quick. I have to come in, sort of lay out what I'm trying to do, set a time to meet with them. It has this sort of. Yeah, it's inorganic in a way. So, you know, depending on your goals, these, the ways in which you're going to do this interaction are going to really differ, I think. [00:12:47] Speaker B: Yeah, yeah. I'm imagining like if I brought you in today, when we're done here, I'm going to go into work and I'm going to run one of my experiments and in a vacuum by myself. I don't, you know, I just, I do what needs to be done. I troubleshoot whether the reward is coming out and you know, all these different parameters. But, and it takes a lot just to get the damn thing to work right and to get good, good data. And then I can imagine if I brought you in with me that that would maybe alter, I don't know, it wouldn't alter. It has the potential to alter what I'm doing. But more than that, it might just, I might just be questioning every step I'm taking, you know, like, what is Netta thinking? You know, so, and I don't know that that would cause like friction, but like you said, you are mediating a little bit by, by the very presence in there. But, but I mean, is there ever any like, friction? And I promise we'll get, we'll move on to more interesting things. But I'm actually curious, like, is there friction? Like is the dynamic always. Is everyone happy to include you going back away from the formal kinds of interviews? [00:13:53] Speaker A: No, no, no, not at all. So again, I think it depends. I've never, Okay, I will say this. I've never been met with straight out hostility or anything like that. So I think there is this wariness before starting any kind of engagement, which is are you going to take time away from what I have to do and. [00:14:13] Speaker B: Right, right. [00:14:13] Speaker A: You know, that's, that's one thing I think like they're having trained in neuroscience myself actually plays to my advantage. Which is I know exactly when to come in and when to step out in a certain sense. And I'm not just trying to get my own agenda off the ground. So I think there's some kind of, I don't know, shared understanding, mutual understanding that helps in that respect. But one thing that really caught me off guard when I started doing this work was not realizing that. So neuroscience has really kind of. Even in the. Since I was practicing in the lab, since that time, it's been only like a decade or so, but it's changed quite dramatically in terms of technological expertise that people will bring into these spaces. And one thing that really surprised me was that sometimes that expertise is your ticket to security in a way that I wasn't really thinking about in, you know, when I was going into these observations or the. No, no, not my ticket, but the ticket for somebody who's in a lab, like, their expertise in a particular area is their ticket to, you know, their security in that space. And, you know, if I'm coming in and trying to get all the nitty gritty details of what they're doing, I didn't realize sometimes that can actually come off as quite threatening because they're also incentivized to keep some knowledge to themselves, you know, because they. They want to be invaluable for doing this particular kind of thing, whether that's, you know, local to the lab that they're working in, or more broadly in terms of something that they're trying to put their name on and say that I'm the one doing this. And so I think there is some masking that people, you know, they're not intentionally shutting me out or anything like that. You know, it's not. It's not open hostility, but there is this kind of obscuring of information that I sometimes find. And I think about it not as, like, I don't take it personally, but I think about it in this broader sociological sense, which is, why are you, you know, incentivized to share information with me in this way rather than that way? Right. And not kind of give me all the details. And it's kind of. There's a dance there that also others expect. You know, when you go into lab meetings, it's very similar as well. Like, people are not going to be giving all the details. And I find that interesting, like, what's left out sometimes. [00:16:50] Speaker B: Yeah, that's true, though. Even, like when I meet with my. With the PI in my lab right now, I have to think, which plot should I show him? And how much should I tell him about how I got that plot? Because I know it's going to quickly go. Because anytime you put a picture up in front of someone, for example, they're going to have their own agenda and start asking questions. And it's not at all necessarily what you. How you wanted to summarize what you're showing them. Right. So it's going to quickly devolve. So there's even that dance. So it's a. It's at all levels. [00:17:22] Speaker A: Yeah, yeah. [00:17:23] Speaker B: Why neuroscience? [00:17:27] Speaker A: I have the cheesiest answer. So probably so many philosophers of science who are doing neuroscience are going to give you some answer like this, but I was really interested in consciousness and like I said, I was trained from a more continental tradition in philosophy. And there I was question, you know, interesting questions about the mind and consciousness. And um, there was at that time that I was kind of getting trained, a lot of excitement about, around neuroscience and answering these questions. Um, and I was very confused at the time just what neuroscience had to do with the mind. And so then I went. [00:18:06] Speaker B: Continental perspective can maybe just say what continental is versus analytic, I guess, for the audience. [00:18:12] Speaker A: Yeah, I, I really actually shouldn't because I'm pretty critical of there being such a sharp division between these two traditions. But yeah, I mean, I would say that what's probably relevant for our conversation maybe at some later point is just to say that the biggest kind of way in which they come apart is probably the emphasis on this kind of post linguistic turn that happens in philosophy and where analytic philosophy really takes off. People start thinking about things in terms of concepts and how language is structuring thought. Whereas I think in the continental tradition things are, they are, you know, attending to questions about the mind in a very different way that are critical of some of the naturalistic assumptions maybe that are taken on in this analytic, you know, movement. And I, I was of that tradition, you know, I was, I was skeptical that doing science was going to be the answer to these questions about mind mentality, you know, experience and that, that was something that I tried, you know, talking to other people about when I was doing my training. They said, well, if you want to critique something, you should learn it, so you should learn it first. And so that's when I went and started doing neuroscience training. And that to me was, you know, it became this is, this is where it all started for me is asking this question of what is neuroscience giving us in terms of thinking about consciousness and mind? And then I started to work with people who were answering those Questions by studying behavior. And then behavior became my object of interest. So I would say that, you know, it's a. It's kind of a cheesy answer. We all. But even a lot of people in neuroscience, I think, get into it because they're. They're quite. [00:20:12] Speaker B: I would imagine, the vast majority. It's not cheesy at all to me. I mean, that's what. What could be more exciting than to study, like, one of the biggest mysteries? Right? [00:20:20] Speaker A: Yeah. [00:20:21] Speaker B: Okay, so I asked you, why neuroscience? So then why sleep? [00:20:28] Speaker A: So, yeah, I think so. I mean, so many of my questions are coming from, you know, interest also in philosophy, like, in a very, very deep interest in the way that we philosophically frame conversations and things about the mind. And one of the. I think sleep has just canonically been ignored when we think about experience. So that was something that was a bit bizarre to me, working in neuroscience, where sleep is such an active state, so to speak. So people who do sleep research, there's just so much happening in the brain. And I think that it was a bit of a. I mean, it was that first disconnect to me that caught my interest, that when we talk about conscious states, when we talk about experience, when we talk about whatever, we talk about wake states, and we don't talk about sleep unless it's dreaming. But that's totally not the experience of the researchers who actually studied this. And one of the other things I think is interesting is the way that sleep plays such a fundamental role in your cognition. You know, in terms of if you don't sleep, it affects quite a bit, not just your ability to think clearly and these kinds of things, but we know that your whole system will shut down in a very important way. And so it has a regulatory role. It synchronizes a lot of different parts of your body in order to allow for those cognitive processes to work optimally. Using that language is a bit weird, but, you know, this is what we've come to know more over the years, and I think that that just was really fascinating to me. It's a very complex kind of phenomena, and that, to me, is just interesting. Yeah. [00:22:21] Speaker B: Yeah. Well, part of what you have discussed in a few papers is maybe I'm jumping the gun here a little bit, but. And this is related to consciousness. It's related to naturalistic neuroscience, whatever that is, that we'll get to and. And sort of on the ontology almost of behavior. Like what. What counts. How do we operationalize a behavior? How do we define the. What is a behavior? And what isn't? What is this kind of behavior versus another kind of behavior? And I think sleep in this example might have been a really good. Just how you described might have been a really good topic to study because, yes, it's not ignored, but it's less studied, perhaps, than other sorts of cognitive behaviors. But also, you point out that. And I want you to describe this to me more that our definition, our operational definition of quote, unquote sleep, as if it's one thing has changed. And you even say that it's incommensurable now the way we define sleep and study it with the way that ethologists of yore, of days of yore defined it and studied it. So maybe you can describe a little bit about that. And then I also want to talk about how lived experience, which is you'll have to say what lived experience is, but how lived experience comes into play in terms of refining what a behavior is, et cetera. [00:23:59] Speaker A: Yeah, yeah. So, I mean, there's so much in this, and I think in some ways it's hard to even talk about, to get clear on how to approach these problems. So I think. Yeah, so the boundary around sleep is not something that is clear. But again, as you've said, I've tried to also discuss how historically that's shifted. And that's something that I actually tried to get into a bit later. The first piece of work that I published on this topic was to talk about, you know, the way that we operationalize sleep. People think that you sort of invoke your everyday experience as part of that, you know, operationalization. But then as soon as you get that operationalized definition and you start testing it, those kinds of experiences that are associated with the state that is oftentimes seen as, you know, something that you don't experience, it just drops out. And so the point is to say that we've really kind of under theorized the role of experience in scientific investigation and sort. And I think sleep is a really good case for that, for reasons that we haven't actually talked about. So you may think of it like this. You know, your experience, your first person experience is. May shape the way that you think about sleep, but as you kind of scientifically investigate it, that comes back around to the experiences that you have of it as well. And I think that sleep is a particularly good case because you don't claim to experience it. So one example of this is people have been surveyed on their sleep quality and things like that for so long, especially in recent years, because people are Trying to create technologies around optimizing sleep. There's a lot more qualitative. And one thing they'll do is they'll ask people, how did you sleep? And they'll say things like, oh, well, I didn't sleep so great. And they'll ask why? And now they have these ways of recording the amount that they sleep. These are proxy measures. And they'll say something like, well, my smart watch or whatever it was, told me I only slept like five hours or something like that. And so then you've got this idea of sleep as this thing that you need to do for a certain amount of time. That's, again, part of your experience, and that's kind of feeding into the explanation you give for what's going on for you personally. So anyway, the point here is just to say that I think there are things that we can do to talk about sleep as a behavior. How do we. What is a behavior in neuroscience? It's an interesting case for that. But there's also. I think it serves as a really interesting example, too, for thinking about the study of behavior more broadly and what is sort of, again, left out of the picture when we talk about experimental practice. I think it's a really good example for demonstrating that something that we don't tie our experience to actually invokes quite a lot of experience for its study throughout investigation. Those are two different things. When it comes to the question of the boundary around what a behavior is here, I think it gets a bit complicated because when I turn to the historical work, it's to say that something about our material practices plays a really fundamental role in terms of how we think about the ontology. [00:27:50] Speaker B: By material practice, you mean the tools that we use to survey. [00:27:55] Speaker A: Yeah, the tools that we use. The kind of physical conditions that we occupy. You use the word ethology here in that paper. I talk about these kinds of naturalistic practices that were happening for sleep studies that predate this kind of formal discipline of ethology, but are mirror it in some really important ways in terms of just physical environment. The way that people were thinking about what a lab was. What a laboratory was, was very different. And so I think all of that stuff is playing a very crucial role in terms of what the boundary around the phenomena is that we're investigating in the first place, how it gets represented, the way that we discuss it in light of the organism that occupies this state or these processes. So I think these things can be really different. Yeah. [00:29:01] Speaker B: So going back to the lived experience aspect of this, you were Saying that this is an under theorized aspect in this case neuroscience, because. And I'll summarize this and you can tell me what I get wrong. So I want to study sleep and I have my personal individual experience of sleep and what I think it is and what I think it isn't. So I'm going to go study sleep in elephants or something. Right. And so I go out in the wild and I can assess whether they're asleep. I record how long they sleep and their behavior after that sleep versus if they sleep for a short time, different behaviors. I record different behaviors and I infer something about cognition related to sleep. But this is solely based on my own personal operational definition of sleep, which is sort of the common. Let's say. Well, I don't even know where to start, where sleep was defined in the first place. Right. Because even early on when the EEG was discovered, alpha rhythms were discovered in the EEG losing sleep with Berger. Right. Well, anyway, so thus far, did I get that approximately correct? [00:30:19] Speaker A: Yeah. So I would say that the operational definition isn't a personal definition. It is something that we have empirically arrived at. But the point is to say that the way that we operationalize sleep today still is historically contingent. Right. So it may not be your subjective definition of what's going on, but it doesn't mean that it's also not tied to a particular moment and what was possible for thinking about sleep and how you get situated there. So the incommensurability conversation is related to this because the way that we operationalize sleep today doesn't match the sort of way that we would have operationalized sleep in these kind of pre. Practices. I don't know, I don't want to call it the before times, because there are people who actually think about sleep in these ways actively. Currently, they're stuck in the past. Yeah, yeah, it's not. Well, not necessarily. It's just a very different idea of what sleep is, what the boundary of the phenomena is. And so the kind of current operational definition doesn't fit that. That's sort of canonically used in neurobiology. So there's a definition that kind of comes about and is exploited still. [00:31:39] Speaker B: Well, yeah. Okay. So the word incommensurable. Right. So I don't know if it's the molecular revolution where you demarcate maybe where this kind of happened, but. So these days, right, we can use all sorts of tools to study how genes are being expressed, the local field potential amplitudes across the scalp. There are all sorts of markers and one of the points that you make in your paper is that a hundred years ago, right, maybe it was a hundred years ago thereabouts, when people were studying sleep, they just. They didn't operationalize it the same way. They didn't look at molecular neurobiology to. To demarcate like when sleep is happening and what kind of sleep it is, et cetera. And these days we use those as the operational definition. And correct me if I'm wrong here, and by doing so, we're kind of redefining what sleep is relative to the way it was defined in the past. Is that the incommensurability that we're talking about. [00:32:41] Speaker A: Yeah, yeah. So, like, kind of loosely incommensurable here just means, you know, lack of a common measure. So you can't really compare these two because they. They don't measure the same way. And so I would say that, yeah, it's. It's not even. And one of the things that's kind of tricky if you look at today we have an operational definition of sleep that we use. And we say that this is like a good definition across, like animals. You know, this is how we're going to study it. [00:33:10] Speaker B: And we should say in the papers, I mean, you really focus on sleep in flies. Drosophila. [00:33:15] Speaker A: Yeah. In the paper, the cases that I look at are, yeah, it's not just flies, it's actually cross invertebrate. So this is trying to understand sleep in a system that's pretty different from us. And that I think there's a good amount of distance temporally to kind of see how things shifted in terms of defining sleep in the early 1900s versus how we're studying it since the 2000s. I wouldn't say that in the past. This past case that I use gives us a formal definition for how to operationalize sleep. We can only look at the practices. And I actually think that that's part of what's interesting is that it doesn't think that you can actually give a single way of defining and operationalizing sleep even within the same kind of organism. And so you see much more room for flexibility in a behavior as a concept. Then you get in kind of later studies of sleep that are trying to standardize it because the experimental goals have shifted also over time in terms of, like, what. What qualifies as a good explanation for sleep to begin with in these animals. So I think, yeah, you know, we can talk. I talk about this kind of inward turn. Yeah, we move. We develop better tools for understanding, like what are some processes that are correlating with this change in the state of the animal that is observable. And then that sort of becomes canonified in a way as something that's very important. So today, if you talk to sleep researchers, I actually went to a museum not long ago, and then they were asking. Somebody asked this question about whether some kind of animal sleeps. And they said, oh, yeah, we got EEG measures that look like there are these changes, and it looks like sleep, even though outwardly it doesn't look like it's sleeping. That can sometimes substitute the thing that you observe, whether it meets your intuition or not. But that wouldn't have qualified in the past. Not because molecular information wasn't available either. I try to make the point that there was a lot of physiological information. It might not have been as precise as what we have now, but its role in mediating our understanding of what the behavior is was really different as well. So it, in some ways is part of this dance on triangulating on multiple kinds of information and not this sort of reductive procedure that we take today where we say that like, oh, here are the markers, therefore it must be sleeping. And even that, I would say, you know, that's a very caricatured version of what's happening. But still there is this contrast that makes it hard to compare these two things. [00:36:21] Speaker B: Which is better? Modern reduction or a bright. A wider birth of what counts or a wider variety? Yeah, what is better? [00:36:34] Speaker A: I think it depends on your goal. So, you know, there is a lot we don't know. And maybe this will actually, like, take us into talking about some of the natural behavior stuff, because I think there's quite a lot we don't yet understand in terms of what we think a good explanation is at the moment in neuroscience. And yeah, so in terms of what's better, I think it will highly depend on that, you know, and I'm not, you know, as a philosopher of science, I'm really interested in kind of documenting this. [00:37:15] Speaker B: Often in these sorts of issues, you know, there's. There's raging debate about what counts, what doesn't, how to what makes a good explanation, et cetera. So you said since the early 2000s, this sort of molecular reductionist approach to operationalized sleep and that very inside specific way. Way. There are two kind of big papers, I think, that came out at the time that heralded this sort of change. But what's the state now? Are people settled, happy? Is the sleep neuroscience world settled on a definition everyone agrees with, with what counts as quote, unquote, sleep? Or is it is. Are the debates raging on how. What's the state right now? [00:38:00] Speaker A: I guess it depends on who you ask. I wouldn't say, you know, from the outside, not as, you know, practicing sleep. Sc I think we have a pretty good operational definition and people are pretty satisfied with that for at least getting their experiments off the ground. I wouldn't say that the debate is settled by any means. Again, I think here things get a bit complicated because the more neuroscience is so technologically driven and so the more tools that come out and then the expectations shift on what is required to kind of explained this thing that was previously explained and so on, that could definitely shift the landscape again. So I don't think of anything as settled in this respect in the field because we're going to constantly. I wouldn't even use the word improve. It's just change in terms of what our standards are for explaining is refine [00:38:59] Speaker B: is the proper term also. [00:39:01] Speaker A: Wait, refine? Well, it can be, I think. So how do you see refine here? [00:39:11] Speaker B: Well, I mean, refining in terms of it seems with the reductionist approach or just maybe the scientific approach in general, you take a phenomenon of interest. I'm going to study sleep. And then you realize sleep is actually divided into multiple stages. And then you start studying a stage and you realize, oh, that is, you have this gene turns on early during that stage and then there's a, you know, fading tail of that gene expression which leads into a different stage. And where is the boundary? So you start, there's a cascade of minutiae that you discover and it's just smaller and smaller minutiae. Right. And that's what I mean by kind of refinement. And then, and then you end up with, oh, sleep is actually 6 million things. [00:39:59] Speaker A: Yeah. So if you. So, so it's a kind of reduction then is what you. What you see. [00:40:04] Speaker B: I suppose, Yeah. I just. You use the word refinement and in terms of like refining what, what behaviors are which, you know, we'll get to. So I was wondering if that's a proper term here. [00:40:14] Speaker A: Yeah, I mean, maybe the way I use it is different than just reduction, but I would say that I don't think of it as just reduction either. I do think that some of the tools we build also move us in a. They can increase the complexity of our understanding. Right. [00:40:29] Speaker B: Conceptually. Right. These conceptual shifts. Yeah, go ahead. Sorry, I interrupted you. [00:40:35] Speaker A: Oh, no, no, no, it's fine. That. Yeah, do. [00:40:38] Speaker B: Well, do. Okay, so what do scientists today look Back on the early studies that you write about and say, well, we can't even use that data to. Because our definition is so different that, like, I don't even know if it counts, you know, Is there that sort of. [00:40:54] Speaker A: No, no, I think exactly the opposite. I think that there's this idea of continuity that people just maybe are implicitly committed to. I mean, in some cases explicitly, they're committed to it because they'll directly cite older research as sort of like motivating. The way that they're thinking about it's [00:41:11] Speaker B: motivating, but it doesn't mean you could still be motivated and. And think that, well, they just didn't have the right tools. They didn't have the tools that we have, so they couldn't understand it as well. Right. [00:41:20] Speaker A: Yeah. Yeah. So. So. So I do think that's how they think. I think that, oh, they were studying the same phenomena in the past. They just didn't have the resources we have now. We're better at understanding what that thing was, and we're building upon it. And that's what I mean by. There's this kind of idea of linearity there in progress that they're making. And. And again, this is why I think that this makes contact with some of the literature on natural behavior, is that that assumption is there when people turn to these other traditions that were probably more popular for studying behavior, and they think, oh, well, I think this is wrong to even characterize it this way because we still have people doing neuroethology, but it's like, oh, the neuroethologists of the past weren't using these really sophisticated tools that we have now, and we're doing a great job. We're building on it. And I think that is just what that's getting wrong is the idea that the behavior that was being studied then is the behavior that you're thinking about now. So this is. This is the issue. And I don't think that they think of it as something different. I think that they only think about it as maybe coarser in some sense, but not necessarily a different kind of thing. [00:42:37] Speaker B: Yeah, but your claim is it is a different kind of thing. [00:42:41] Speaker A: Yeah, absolutely. Yeah. [00:42:43] Speaker B: And that's because of the. And why is that? Rather than. I'll just ask you. So why is it a different kind of thing? [00:42:50] Speaker A: Yeah, so in the paper, I try to lay out that the practices around, again, the kind of material conditions for studying those behaviors are not just studying, you know, the same behavior with different kinds of tools. It's that what the Behavior is, is understood through the ways that you're investigating it. And so there is where you get that kind of incommensurability emerging. And the way that I kind of take on this idea of incommensurability is really borrowed from philosophers of science and practice, like people actually looking at how the tools and the instruments that we use are changing the phenomena and shaping those phenomena that we study. So this idea of incommensurability is really borrowed from Ian Hacking tradition. And that's how I see it as different from maybe talk of incommensurability that abstracts away from those kinds of practices. Yeah. [00:43:49] Speaker B: Is this related to the concept of semantic drift, which I bring up too frequently on this podcast? I use this phrase like it's a super common thing, but. But I mean, there is semantic, conceptual. Semantic drift in science as well. I mean, is this. Does how. How is what you're describing related to that concept? Just how our concepts in science change over time? [00:44:14] Speaker A: Yeah, well, definitely. I mean, this is something that the discussion on incommensurability, like the most popular discussions of it, happen through people like Kuhn. And so if you're talking about this in, with that kind of idea in mind, where there are real shifts in the paradigmatic ways of thinking about things, then this is very similar. So one thing I don't know, and I'd be curious if you want to say more about it, about this idea of drift, is whether or not you're retaining something and just certain dimensions of it are changing, or whether you think of it as like a wholesale, you know, shift that I think is really important to think about. And again, I think I'm less interested in these discussions from an abstract point of view. I want to know, is the notion of sleep that we were using then, within these kind of material practices something that doesn't serve the way that we're thinking about them within our current frame. Right. So really, again, comes back down to the practices like, what is it doing for us when we make reference to those kind of procedures for thinking about what sleep was? And this is where I am worried about the discussions on natural behavior. When people look at these things and say something like, okay, well, we're just improving on an old way of thinking about this with better methods. Is that really what's happening is what I'm asking. [00:45:55] Speaker B: Yeah, well, let's get into that and we can revisit what we've been discussing thus far as well. So, okay, so like I said, okay, so naturalistic, quote, unquote, neuroscience has exploded we talked about a book by Nachim Ulovsky. Is that his. [00:46:14] Speaker A: Yeah. Nahum Ulanovsky. Yeah. Who's. Yeah, yeah. [00:46:19] Speaker B: Which. And you know, so this is the past, I don't know, 10 years or maybe less than that, I suppose. But everything's naturalistic and it's kind of glorified because of our modern tools. We can record a lot more data, but we can also record data in less restricted, reduced experimental settings. So when organisms are behaving more freely, less constrained in the lab. And the idea is that by allowing them to perform these more quote unquote naturalistic behaviors and studying it without all the constraints of reducing it to like a single arm reach or a single eye movement with the head fixed still, these traditional like reduce and control everything that you can because you want to study the precise mechanism of the thing that you're. Of the eye movement, for example, and you want to remove all of the other, all of the other behaviors because they're potential confounds that would then dirty up your explanation that could serve as confounds to your explanation of the actual eye movement or reach or whatever it is that you're studying. Falling asleep, for example. So you want to loosen those constraints and study it in the most naturalistic setting possible. And this term naturalistic here has started, it's popular, but it, it has started to garner some criticism. And I think that, that you have some criticism as well. Like. Well, a lot of people are saying, well, what is naturalistic about that? So, so for example, let me describe to you my experimental setup. Okay. And you can tell me where I am perhaps without. You don't have to judge, but it'd be great if you did. But. All right, so I have a, let's say later today I'm going to go take a mouse. I'm going to put it in like a 2 foot by 2 foot space. It has electrodes in parts of its brain which is attached via head stage to its head. And so there's like a little wire coming up and that's the only thing that's attached to it. But it's attached the whole time. Otherwise the mouse is freely able to just move around. And all we do is let the mouse do what it quote unquote naturally does. So it walks around the arena, it grooms, it sort of sniffs around, it'll rear. So all these quote unquote naturalistic behaviors, right. Meanwhile, we're videotaping up underneath and we're tracking the animals key points on the animals bodies. We feed that to an unsupervised learning algorithm that was built in house. It's called B side for those in the ITRI lab. Great tool. And what we get out is kinematics of the different paws moving in different ways in the tails and the nose and stuff. And then we feed that through the unsupervised learning algorithm, and it spits out behavior labels that we can then go and put English words to. In this case, that is walking, that is paw grooming, that is rearing and stuff. So. And that. That's sort of the pipeline. Okay, so that's our experimental setup. And for a while, we were calling it naturalistic, because that's what you do. That's sort of the term that. That people were using in the neuroscience. Right. So now I've started with these behaviors of. We've started. We were searching for different terms to call it, because the term naturalistic is somewhat problematic because of what people like you have, you know, have started to look at here. So we're calling it, for now, spontaneous, these spontaneous behaviors, because we're not queuing it to do anything. But, okay, so am I studying naturalistic behavior? What am I doing? [00:50:08] Speaker A: Okay, so, yeah, I think it's fine to transition to calling it spontaneous. It's okay to call it naturalistic. I think my criticism in the piece is not. It's about kind of insensitivity to the way that we actually are. So all of these questions are getting indexed to the technological turns that are happening in the field. And what is naturalistic is getting set to that? It's insensitivity towards that. You could call that naturalistic, and that's fine. But I think that what that does then is it generates so many different understandings of what is naturalistic. One example that I give that may be relevant to walk through because of the case you're describing is this kind of move that at Columbia, Gwyneth Card wrote a piece that she coined this term reverse neuroethology. So I don't know if you saw that in the paper, but basically the idea here is that you are not kind of using an animal that's adapted to do a certain behavior really well. Instead, you're using champion model. That is seen as a champion model because it has a kind of history of being a good lab model, something that is standardized. In this case, you said you're using mice, right? Yeah, that would be an example of such a model. It's not that these mice exhibit some unique behavior. They're just doing their own thing, whatever Lab mice do. And then you're kind of trying to study that behavior. So in this case, you're calling it spontaneous behavior, but someone else may call that natural behavior. And I think there that becomes really interesting. So it's not necessarily like asking this question. And again, like, we may go back to this in commensurability conversation, because here it's like, imagine if you were actually interested in what a certain animal can uniquely do, right, in terms of behavior was adapted to do so. Then you may be asking more questions related to its environmental history, like its ecological niche, something like that, the environment in which it evolved, what other organisms it interacts with, that kind of stuff. And then you would be studying behavior from that frame. And then I think that that would look really different than what you're trying to do, which is to try to, especially when you feed it into these behavioral tracking algorithms, you're trying to understand motives that emerge that can be generalized across multiple animals. So the kind of explanation that you're going to give at the end is not going to be something that's unique to that animal in terms of its spontaneous behavior or whatever, but something that can be generalized and fit within a larger story about how these kinds of behavioral changes emerge. So I think there your explanatory goals can look really different. You can both be calling it natural behavior, and that's what is the problem now. It's not just talking past one another. You're doing something, someone's doing something else. If you buy into my argument about the incommensurability stuff that I mentioned before, you could actually run into major theoretical disagreement with one another, but then be calling it the same thing, right? [00:53:54] Speaker B: Yeah, I think that happens all the time in neuroscience and probably in other sciences, but specifically, especially in neuroscience, I [00:54:02] Speaker A: think a lot of the times we talk past one another, I think it's more dramatic when we're really not talking about the same thing. And that is the part about this that I find interesting is if we take seriously that these tools are reshaping what even qualifies as the behavior. I mean, I think that in the case of Bsord and a lot of other interesting unsupervised algorithms that are out there now, one thing that people are really optimistic about is that it's removing us as the observer in this process. Right. And it's giving us categories. So you can then go give human labels to those things. But you weren't the one who kind of like drew them out as what's important. And so it's Supposed to say something about how we're not good observers of behavior, we're not good for classifying behavior, and we want these unsupervised systems to give us an objective read on what's happening. [00:55:02] Speaker B: It's interesting. So you write about this in the transmitter as well, using these sort of unsupervised learning algorithms with the goal of removing the experimenter. Right. So you don't want to inject your own biases in it, and you argue that it should be a collaborative effort. But, you know, it's interesting, even with all these unsupervised tools. I was at a conference and I can't remember who I was speaking with, but he was suggesting that what we really need is, is to bring in expert ethologists to sort of do a supervised learning, maybe on top of the unsupervised, I don't know. But, but to tell us, you know, okay, well, that's. That groom is indicative. That's a different groom than the first groom because of such and such context, et cetera, to help us refine those. And is that sort of what you're arguing in that transmitter piece? [00:55:55] Speaker A: Yeah, sort of. So I co wrote that piece with a friend of mine, Matt Whiteway, who's a research scientist at the Zuckerman Institute, and he also helped develop Lightning Pose, which is another tool. One thing that he and I talk a lot about, the kind of relevant to what you just said, is that there are real differences between the researchers who are kind of using the data sets that have been collected of these animal behaviors who've never interacted with these animals before, versus those who are actually working with them every day and then kind of collecting that data. And so you can see the analysis will differ pretty dramatically depending on your experience with the organism. And another thing that I think is interesting here is your question about is this what I'm saying? I mean, I actually think that a lot of these unsupervised methods aren't as unsupervised as we think they are. And so I try to frame some of this analysis that I'm doing on these different tools within the framework of opacity, the problem of opacity in science. So I think that people think, okay, you feed these things into an unsupervised network. That's where the opacity problem emerges. You don't really know what's going on and you get some output and then whatever. And oftentimes these are the kind of neural networks that is causing this opacity issue. Right. But I actually try to show that even in cases where you do have some understanding, it's an unsupervised model, but you have some understanding of, loosely, what the model is representing. So imagine it's not a neural network and, you know, some things about what it's representing with the behavior, you can still end up in cases where you don't really know what's going on. And it has to do with the kind of surround conditions of these algorithms. So the kind of temporal assumptions that were built in or the way that they were designed with respect to the actual, like, physical arena that you're recording information from and who, in fact coded this thing. What the kind of. Maybe there's some Bayesian priors you don't know about. I mean, there are all these things that are kind of feeding in. And so the opacity issue isn't something that's like, you know, this is a technical problem where things are not accessible. It's a kind of material one. And there, I think, we haven't done as good of a job, again, characterizing what are the various stages that we're actually intervening when we're designing these systems and exploiting them. Right. So why is it, you know, going back to the conversations I have with Matt, it's like, why is it that you get such differences with people who know the animal versus those who don't? You know, that shouldn't be the case if this were just something technical that you're supposed to exploit. So the fact that you do get these differences, I think is interesting, and it shows you that. And one other difference I've observed, which may be relevant to some of the stuff you're doing, is that you're in a lab where they develop this method. So imagine, like, I work with people who are also in labs where they haven't developed these tools, but they're using them for their models, systems. The results they get can look really different. [00:59:18] Speaker B: Right, right. [00:59:19] Speaker A: And so that's also another thing that I find fascinating is like, that it should be, you know, it's. It's at least standardized. Yeah, standardized in universal. And the same across context. Yeah. [00:59:32] Speaker B: Well, that's why it's important to replicate these sorts of things, like, across laboratories and, you know, organizations like International Brain Laboratory, where their goal was to, like, let's do the same experiment in a thousand different labs and then collect it all in this big database. And they should all be the same, essentially, because we made it as similar as possible, use the same protocols, but it's important to analyze data from multiple labs because every lab has its own challenges. Even on a day to day basis like this day, the solenoid didn't open as much or something, or it was a little colder in the lab. You know, there are all these like little details that matter. I wanted to read this from your, from the paper. So the paper that, about the naturalistic stuff that we've been talking about is called what is quote unquote natural about naturalistic neuroscience. And so we're going to continue talking a little bit about this. But I like this quote from the paper. Ultimately, I think this comes at the very end in the conclusion maybe ultimately the naturalistic turn in neuroscience is neither a unifying revolution nor a clear corrective to the study of behavior. It is a set of shifting practices built atop powerful but sometimes incompatible visions of behavior. So I think that nicely summarizes a lot of what we already have been discussing. So you say it's not a unifying revolution. And I think that that is sort of the hope in neuroscience, right? By calling it naturalistic neuroscience, we're somehow improving and we're doing it right now because for years before that people would sort of criticize a lab experiment and saying, well, that's not out of the context of what they would do in their ecological setting anyway, like, so it's not even a valid thing that you're studying. So now we in neuroscience have corrected that by becoming wilder in fear. But then here you are saying, well, this is not the unifying revolution that you believe it is. Perhaps. [01:01:45] Speaker A: Yeah, I mean that's a hard thing to get to. But I think the first step is to pay attention to our tools, pay attention to our practices, attention to the apparatuses that we use to measure things. I think again, if they're, I mean, this is the philosopher of science, just like pounding this idea over and over again is that those things matter. They're not just these kinds of things on the outside. And then there's the phenomena that is pure, unaffected out there. And then we're just trying to chip away at it, right? And understand like different dimensions of it. It's like, no, that actively shapes what the phenomena is, what we consider what that phenomena to be. We see through these apparatuses, we see through these devices. And I mean, I'm not by any means the first to say this. This is just, you know, again, the hacking tradition of thinking about things, Hans or Reinberger thinking about experimental systems. Like this is kind of a very old tradition of thinking about these scientific phenomena through the material conditions by which we are understanding them. And that's so important here. And I think that unless we kind of start there, we're not going to know what our kind of goals are. So, I mean, I think one thing that's interesting is like the thing you brought up with the ibl. So Matt is also the co author on that piece. He's also one of the core software engineers in that group. And I find that he has. [01:03:29] Speaker B: Does he have horror stories about like. Sorry to interrupt, but just. I can't. I. Because I've thought about like the people who are in charge of, of developing that, that pipeline that is going to standardize all these things, and I just thought, oh, that'd be a rough job. [01:03:43] Speaker A: But yeah, yeah, yeah, no, I think it's been, yeah, from what I understand you just talking to people who are part of that group, it's been, it's a huge effort, you know, and so I, but again, it's like, I look at that and I think, well, that's not what everybody's in the business of trying to do either. Right. Standardization is their goal. And that is at least like one of the many kind of goals that people may have in the field. But we're also interested in variability. We're also interested in these moments when things don't work out and some people navigate that space explicitly. So I think. [01:04:19] Speaker B: Yeah, well, I just, I'm sorry to interject, but I mean, there's a career element here at play with like an organization like you. You can view an organization like the International Brain Lab as being like, more powerful because they're this combined conglomerate of forces. And then if you have Betsy in her lab wanting to embrace the variability, it's a lot harder for her to get. To increase her H index or to make herself more visible in the literature. When you have these thousand author papers that are, you know, setting down in nature, neuroscience and science and it looks really powerful like, anyway, like it sort of dominates like that perspective. So the, the field is not weighted equally. When these Walmarts come into existence, the mom and pop shop maybe can't compete as much. So that's just a sort of social commentary. That's besides the point. [01:05:14] Speaker A: Yeah. I mean, but also what's to their advantage is that we clearly value standardization. Right. When we talk about the kinds of, you know, the markers of good scientific practice, standardizing and generalizing are part of it. And I think that that's sort of, again, if I, I don't mean to kind of paint this in a very, you know, broad stroke way, but those standards historically arrive from a particular place. Right. And then if you look at practice, you see that that's not exactly what everyone's trying to do. So again, in the natural behavior piece, I actually talk about, like, one of the tensions that arises from this desire to retain experimental control under these naturalistic conditions when you're also trying to understand the behavioral flexibility of a system. Right. And so you have to kind of put a cap on that to some degree in order to come up with generalizable principles for what's going on. But maybe the reason that you got into this in the first place is to understand, like, how far can I push this system? Right. And so there you're constantly dancing with those tensions and you're trying to figure it out. But in terms of, like, you know, again, what's codified, we say, oh, standardization is really important. You know, that's. That's what we should aim to do. And the reason that those papers maybe are popular, I don't. I don't know how much of it has to do. I mean, this is an empirical question too, right? It's like, how much of it has to do actually with, like, it being powerful names behind it versus it just also feeding into this, like, intuition that that's what our science should be aiming for as well. [01:06:59] Speaker B: Sure, yeah. Probably a little bit of both. [01:07:02] Speaker A: Yeah. [01:07:04] Speaker B: I am sensitive to the popularity aspect, though, but it's just a social thing in science, so we don't need to go down that road. But. Because otherwise I'll just complain the whole time if we drop naturalistic. Do we call it nothing? Do we have a different term that we use? [01:07:23] Speaker A: I don't think replacing the term is going to, like, with one other term is going to be the solution here. [01:07:30] Speaker B: You call it unnatural. Everything is in this unnatural behavior that we're studying. [01:07:35] Speaker A: I just really don't like purchasing into this idea of natural behavior. I think, yeah, again, because it has [01:07:41] Speaker B: a value, but again, because it has a value associated with it. [01:07:44] Speaker A: Yeah. It can mask a bunch of things that I think are important in terms of what people are trying to accomplish. And. And again, I think it's, you know, I don't want to go too much in the weeds of this because it can take us off. But, you know, even with the behavioral tracking tools, I think one thing that was surprising to me as I entered different environments where these tools are getting developed is that they also come from very different kinds of scientific traditions and the people who are also Kind of, you know, steering These projects, the PIs of these groups, or, you know, the kind of the people sitting at the top, so to speak. They may have certain kinds of vision for, you know, a framework for thinking about behavior that is not always explicit, but you can really dig in and see, like, oh, how is it serving that kind of explanatory goal there? You know, this, like, kind of larger picture that they have in mind. And sometimes that can be about translation. Something like, oh, you want to develop something that thinks about behavior and predicts it for these kind of translational ends. And that may come apart from somebody who's also thinking about this differently, which is like, is there a language of the body that we can try to parse out that these tools can give us and we haven't been able to do before, kind of going back to this traditional ethological view. And so those are. Again, they can have overlapping agendas, but not always because your markers of success are going to look different for what you're trying to explain again. [01:09:33] Speaker B: Yeah, yeah. Depending on your goals. Okay, so there isn't like a replacement term necessarily, but it might be just depending on your specific goals. And I mean, you have to call it something. Right. And. And the nice thing about maybe not the nice thing, the powerful thing, the marketing aspect of calling it. Yeah. I mean, and there's a lot of marketing in science, and it's necessary. You can't avoid it. Right. So you have to call it something. And then as soon as there's a shiny, awesome term like naturalistic neuroscience, then everyone starts using it. Right. And it kind of loses its meaning over time. I mean, I rant on about terms like mechanistic and computation and all these terms that just get thrown in there without any justification for using the term or whatever. Right. And those don't go away. So I'm not sure. I don't know. I don't know if, despite the accuracy of what you're saying, I don't see. I wonder if naturalistic will get dropped. I don't know. What's your outlook on this? Do you think that it's going to. [01:10:43] Speaker A: Whether or not it gets dropped is not something I can control. I think the only thing is to say that we're not talking about the same things every time we use this term. Let's pay attention to how we're talking about them differently. One thing to also say is when Nachlim's book was published, it was just about to come out. I was having a conversation with this brilliant scientist who works more in the neuroethology tradition. And I'm not going to name them, but they were really pissed off that this book was getting published. And they were like, we've been doing natural behavior for so long, why on earth is this person getting contract through MIT Press to publish this book on natural va? And they were just so angry about it. And I was like, yeah, I understand the frustration because of the coinage of they're trying to say that they're reinventing a wheel or whatever. But actually if you look at this book, if you look at what they're saying, it's not at all what's happening. And I mean it's not that it's like light and day different from what's happening in neuroethology, but again, it's like the tools that are driving the questions in neuroscience are so than these other traditions in which people have been working in. And they're setting a new benchmark on what natural means that I think is getting missed. So it's like, okay, you can keep using the term or whatever, you're gonna maybe piss off some people in the scientific community by laying claim to it. But I think if you look, I mean, my only point is to say if you look at the details, you'll see these are not the same practices. And we wanna kind of maybe flesh that out because I actually am super sympathetic towards coming up with, you know, unifying theories and we're not getting closer to that by doing this. So yeah, I should be more closeted maybe about thinking that way. But you know, I am really sympathetic towards that kind of ultimate goal for the neurosciences. But. [01:12:41] Speaker B: Well, I know you were not talking about me because I was not like upset about it or anything, but when we were talking about it over that dinner months ago, because I, you had mentioned that book and how you really enjoy the section at the end of the book where there's like these sort of ongoing questions. But my sort of knee jerk reaction was like, I don't understand why. Well, am I gonna have to edit this out anyway? I don't understand. I don't understand like what the big deal is about this book because it's like it's been going on for some time now and I don't really know what it adds. Like it seems like late to the party. But then I also recognize that almost any book is going to come way after it. Like things have been progressing in that direction for a long time. And even though the book seems to market it as something new and big which it's not new, you know, but that sort of is inevitable with the way that publishing works, I guess. But, yeah, I don't know. [01:13:38] Speaker A: I do think. I mean, again, if you buy what I'm saying about these kind of new technologies being central to what this notion of naturalistic is today, so, you know, doing calcium imaging and freely moving animals or wireless opto or like these behavioral tracking algorithms we're talking about, like, all of that stuff is relatively new, you know, and it's really opened how we're able to study behavior in different ways. And I do think that, again, it's like, if you see these tools as, like, central to this narrative of what this person is doing and what other people are saying, too, when they talk about the study of natural behavior and neuroscience, like, it's not. It's not the same thing as what was there before. Yeah. [01:14:21] Speaker B: So in. In here's what I want to ask you is what you think about the. About the tools changing the theory, essentially, right? So in something like, I don't know, quantum. In something like physics, right, you get, like, bigger, bigger particle accelerators or just a bigger microscope, and you can see smaller and smaller. Then you're like, sort of discovering new organizations of matter, for example. And so that is fine because you're really exploring and discovering. But in cognitive sciences, many people think, well, you really need to begin with theory. Like, why would. Here's a question is like, why would better tools reinvent your theory? Like, that must mean that you have poor theory if you're letting what your tools show you dictate what you actually think about the cognitive function. I'm not explaining this in a very clear, coherent manner, but this goes back to the idea of, yes, there should be a virtuous circle of. You get better tools to ask better questions, and those inform your theory and change your theory, and you build the tools. But one could see this as well. It seems like the tools are just dictating how we think about these things. Like, we just have no idea what we're talking about. We must not. Right. [01:15:49] Speaker A: Yeah. So, yeah, yeah. I think there are so many. And there have been papers written on this tool versus theory framework, right? Like, for thinking about these questions. And actually, this is so relevant to the context in which we met, because we met at this neural populations workshop. And technically, you can think of that as like a theoretical overturn due to different tools that have come about. Right. At least that was the case for Rafa's talk and some of the other things that were going on there. But I think that I like to think about these things not as tool versus theory, but maybe more in that the way that you frame it in terms of them being mutually informing. And I think that depending on how you frame it, you could always say, oh, the tool came first or the theory came first. I mean, I think one breath, one time in a talk, I heard someone make the argument both ways. So I think how I think about this is more from. And we haven't talked so much about this, but this framework of lived experience that I draw from to me is really productive because here I see experience as something kind of under theorized in the experimental process. It's playing a role at kind of every stage of experimentation. [01:17:09] Speaker B: Sorry, just to. I'm sorry to stop you, but can you just define lived experience? [01:17:13] Speaker A: Yeah, totally. Actually, this is really important because it's a term of art that comes from phenomenology. And it's not something that I think people are always confused. They're like, well, why call it lived experience and not just experience? You know, because it comes from this tradition. And phenomenology, of course, is not a single tradition either. So the term that I'm using it from is mainly rooted in this kind of Husserlian usage of this term. The idea here is that the sciences kind of are just unaware of the conditions for its own possibility through experience. And so here you're not just talking about something that's like that first person subjective stuff, but also the way in which your ideas can be historically situated, the ways in which possibilities are afforded to us. Notions of embodiment enter here. And so I think that when I talk about lived experience, it's not just the way that we were talking about it before. When we say, hey, we operationalize sleep and we draw from these everyday ideas of what sleep is in order to help us set these definitions in place. It's also things like knowing when, kind of going back to that behavioral tracking example, how much time is actually the necessary amount for thinking about the transition between this behavior and that behavior. These are the ways in which experience plays a role throughout our procedures of judging things in science. I think there you would see that that really blurs that line between tool and dairy if you think about it from this frame. Frame. One thing that's also kind of problematic here is that the tool theory stuff comes from this tradition of thinking about sciences where the theories were really specified. And I'm not really sure that in neuroscience we're navigating theories in that way. Right? We have these ideas about how things work, but they're not as solid. And sometimes they come also from the local research communities in which we're situated. And so I can, right now I'm embedded in a lab that does more evolutionary neurobiology. And I can tell you about these theoretical debates that exist on brain homology or something like that. It doesn't really mean anything in terms of a theory and a robot robust sense. Right? Like, I mean, it means something to these communities that are using particular tools and techniques for their questions. And yeah, it is theoretical. It has this theoretical component to it, but it's not like a robust theory in that sense. And so I think when you kind of shift this to this like, kind of framework of experience and you look at the ways in which experience is getting embedded throughout this process, it really blurs that line. And that to me is where this stuff becomes interesting. Because again, when I say tools, I think it comes with this kind of idea that, oh, nut is just talking about this boring technical stuff that we're using in order to do experiments. But you dig in and you see that, no, these, these things are really extensions of how we're, you know, seeing. And that's, that's where it's much more interesting to me. So, I don't know, I'm making it complicated. [01:21:01] Speaker B: Well, no, well, I was going to ask if you can give like an example of like where lived experience has, you know, come into play or where it maybe could have come into play and if, if it were considered, for example, I don't know, just like a concrete example. Do you have any that are. [01:21:16] Speaker A: Yeah, yeah. So I love the sleep case because that's what I thought about most with behavior. But I think. Yeah, the sleep case in the paper I show is that you have. Okay, so you have these kind of first person things. You know, when we look at people sleeping, for example, we know that they don't move around most of the time. I mean, they move around a little bit, but they're not getting up and walking, you know, and so you have those, those kinds of things. You also have lived experience in the sense of like, why. Well, why am I even investigating this topic to begin with? Right? Like, was there something about, you know, did my mother suffer from insomnia? And that is what drove me to think about sleep in the first place. So there, there are all these things that have to do with your personal, like the way in which you came about in the world. Right. But I think there is, there's a host of Stuff that also has to do with, again, those local research environments in which you're familiar with and your background in training. So this is also what I look for when I'm occupying some role as a philosopher in a lab. Is one thing I'll always ask people is what were you trained in scientifically? Were you trained kind of in molecular biology? Were you trained in physics? Were you trained in straight neuroscience? Were you trained in some other kind of science? Were you an engineer? And I think that again, there's. That. That is playing a role as well as like kind of your. The communities in which you're interacting with. Okay, so you're actually studying sleep in. I mean, this is just totally made up, but like, imagine you're setting sleep in just Drosophila or you're setting sleep in Xenopost or something like that. Are you just going to Xenopost meetings? Are you just going to Drosophila meetings? Like, who are you actually interacting with? So I want to resist taxonomizing experience in these kinds of rigid ways. But the point here is to say that all of these influences then are going to affect the way that you not just think about the kind of experiment that you should be doing, but also then what is the meaningful line between the phenomena that you're observing and everything else. In that paper, I talk about the apparatus for studying inactivity in Drosophila and the way that we've standardized inactivity to like, kind of being around five minutes, representing a bout of sleep. And then the question is, like, why five minutes? You know, why that amount of time? And then once we get that answer, the next thing that we may do is also kind of tie this stuff to a genetic story. And then we want to take that and relay that information across different systems, right? We don't want to just come up with a molecular story that remains like, kind of local to the system that we're just trying to study. We want to understand how does this link up with other kinds of information that we know. And so there again also experiences playing a role. And I think that that stuff has much more of a fluid component to it than we may think. It's not just this thing that we can procedurally represent and say that, hey, go follow it. Right? It's a real. It takes a lot of experience to get really good at doing that. And I can also say, you know, one of the. Just going back to these stories that we tell about how we get into these things, when I did my neuroscience training, I had not been. I had Done lab rotations, but I had not been working in a lab full time, you know, and I went. When I got my. My first job, like, I was like, you know, full time going to be working on. There were two experiments I was doing. And I knew exactly, like, I knew the protocols, like, for how to do this stuff. I knew all this stuff that I had to do. Right. And I had. I had just gotten my degree. Like, I knew neuroscience now. And when I got in, I remember on my desk there were like this huge stack of papers. And. And that was it. I mean, my job was to just like, read through those paper, like to read through those papers and then to go do the experiment. And I remember asking, I was like, well, can you tell me what to do? Like what I'm supposed to be looking for? Like, what. He's like, no, just gonna figure it out. And so I think, like, that's the. That's what I'm trying to say is that it takes a lot of time to cultivate that. And you can't just put it down conceptually and then like follow it like a recipe. And you get better and better at it through just kind of doing Right. [01:26:06] Speaker B: But see that what you're describing. Okay, so. Could be construed as sort of antithetical to what science is supposed to be. An idealized version where lived experiences. Lived. So in this version, lived in experience gets. You could substitute the derogatory term bias for lived experience. And that's supposed to be bad. But that's not what you're saying. What you're saying is that the lived experience needs to be a part of the story, not to reduce the impact of the inference that you're making or the interpretation that you're making from the results, but as part of the actual story of the science. So it's not a negative thing here. But it is. I mean, what is. Does that mean it's a positive bias or a neutral bias? Or is bias the wrong term? [01:27:01] Speaker A: Bias? Yeah. I don't even think of it as bias. And I actually find the word bias kind of problematic sometimes too, when people invoke and say, like, oh, this is such a philosopher. Yeah, yeah. So I think it's a necessary condition for doing science. Science is to like, kind of invoke this. Invoke experience. And I think experience, again, has been under theorized. And that's why we kind of like, kind of relegate certain decisions that people may make that push ideas in certain directions over others as something like bi. You know, it's calling it something like bias. And so that gets us to say, oh, we need to strip all of that out, you know, so strip the human out of this process. And I think that actually this is becoming. You know, when I got really interested in this, it was always from that frame of where's the person in all of this? Right. Like that is actually essential to the explanation and story that we tell. But I think that it's becoming more pressing actually to think about this because we're now trying to automate science. And so there are all these efforts now to actually just have AI become the scientist and do the science. And so is that still science? You know, I think that that's something that we need to ask because, I mean, there's so. There's so much room here to just to think about this because what the AI is is also open to discussion, you know, how much of it is. It represents of us and our interests. But the point here is just to say that science is a human activity, and I think of it as a humanistic activity, and the conditions that make it possible are us. So that if you want to call that bias, I would say also, good luck trying to do this without us kind of making those calls throughout the process. I just, I tried to show in that sleep paper that in certain moments when you. And I mean, historically, we have also seen this. When you kind of remove the human and in that process of interpretation, what you get is sometimes an explanation that. That makes no sense. And so. So there are time, like there are mechanistic characterizations of. That are mechanistic happenings, I would say, that correspond to sleep processes that are never invoked when we explain sleep. And that to me is an interesting thing. Like, why is certain. Like why are certain things ignored, completely ignored? I mean, the. Does a whole host of stuff anytime that you're kind of engaged in a new behavior, but not all of it's invoked. And so that's just like an illustration of something to say that we don't. Not everything matters. Not everything is part of the story. And there are systematic reasons for that that haven't been thought about so much because we place our focus elsewhere. And there are also historical reasons for why we want to think about neuroscience in this very fictionalized way [01:30:35] Speaker B: that. I like that term fiction. Fictionalized. Are you pointing to, like the quote unquote idealized version that I was. [01:30:42] Speaker A: Yeah, yeah, yeah, there. [01:30:45] Speaker B: Yeah, but it's fictionalized. But it's also sort of something that we're taught to strive toward. Right. I Guess I'm wondering what the. Maybe I'm just hung up on and eventually somehow need to get comfortable with the idea of the lived experience being an essential part of it. Because I'm so. Whether I'm implicitly taught this or not, I'm so used to the idea that anything that the human brings into it is sort of a degradation of the actual pure science. Right. So I can only contaminate. But what you're saying is, is. Which I know because whether it's biased or lived experience or the act of the human, the involvement of the human is a necessary part, is, is an inescapable part of at least human based science. We can talk about AI in a minute. But I'm wondering, like, how far away are we from being really comfortable with how to, how to be, how to tell that story in a way where it all fits together and we're still happy with this is with saying this is progress in science, not. This would have been progress in science had we not contaminated with our, contaminated it with our human endeavor. Right. So are we. Do you feel like, do you feel comfortable that we. You can tie that in together, the lived experience component of it in a way that will satisfy people like me who have been taught implicitly that that is antithetical to what real science is. [01:32:26] Speaker A: Yeah. So one thing I'd ask you is like, who sets the kind of standards on progress? Right. And so if you want to have progress, I would. Okay, so I'm, I don't care about progress. Okay, so cards on the table. I think that, you know, if there's this like, idea that, you know, we've got, there's a truth and then we're slowly working our way towards truth. I don't, I don't. That's not what I'm. What I think of when I think of progress. [01:32:57] Speaker B: What do you care about, Netta? Do you care about anything? [01:33:00] Speaker A: Yeah, So I would say like. But who sets the kind of standard on progress is really important here because it's us, right? We're doing that. And, and then the question is, wouldn't you want to think about the role of experience in science, kind of dive into that and think about it more robustly as a way of knowing whether or not you're actually meeting those benchmarks. So there. I think that, yeah, if you've got this idea of progress which is like, there's truth and we're just the. These things in the world that are so, you know, limited and we're biased, as you say, like the Way that we conceive of the world is already like, mediated through our nervous system, which only sees things partially and. Whatever. Whatever. I mean. Yeah, then I have nothing to say to that. I guess I think that, you know, I think you've gotten it wrong. But if the idea here is that, yeah, you, you can define progress differently than that, and yeah, then I think experience is something important to think about. One way of also saying this is just to go back to how we started the conversation is the thing that I'm trying to do by occupying a role in a lab as a philosopher kind of is a material expression of this idea. Just trying to understand, you know, how can we create a meaningful space for philosophy within science that helps draw attention to these dimensions of scientific practice that the scientists themselves are, you know, for many reasons, not incentivized to think about and to frame and don't have the time to do? I mean, in so many ways, like, I, I love science and I love doing neuroscience, like physically doing experiments, and I love the culture of a lab. And there were things that I really miss about it, but I just didn't feel like I had a space in it for the kinds of things that I was asking, the kind of space I needed to think about the questions I was interested in. And so I think that finding a way to meaningfully bring that in is almost like a. Like I said when I was saying this is all an experiment for me is like, can we actually get these ideas off the ground in an active practice way rather than just me writing papers about lived experience in science? Like, is there a way for me to physically bring this in? [01:35:51] Speaker B: Yeah, see, that's. That's too bad to hear because I was really leaning toward naming this episode Neta Nemati. I don't give a damn. But now I can't, right? [01:36:03] Speaker A: No, I give too much of a damn. [01:36:06] Speaker B: Yeah, that's the. Yeah, I mean, I like that you use the term dimension there, because going back to the idea of progress in science, I mean, you can define it differently and everyone does. And everyone has the, their own personal, subjective sense of what they consider progress. And for me, it's about explanation and what constitutes a satisfying explanation. And so that's where something like the lived experience comes in. Because if I think of explanation as kind of a high dimensional problem as well, where you have prediction, you have understanding, you have theory, and maybe lived experience is one of those dimensions along. Maybe not in terms of what satisfies an explanation, but something just is another dimension, which is why I like the way that you use the term dimension. So, lacking a ton of time left, I want to open the door to have a conversation, if you'd like to, about this new project that you're funded for. I don't know how new it is, but looking at the role of AI, since we brought that up, and how we want to. To take humans out of the equation and just make it automated science. So do you want to talk about that project a little bit and where you are and where it's going? [01:37:24] Speaker A: Yeah, yeah, sure. So we've talked about it, actually, a little bit already. So this project started out as a kind of comparative assessment of different behavioral tracking tools in neuroscience. Because I'm interested in behavior, I'm interested in the tools around studying behavior. And. And of course, because you have, like, you know, firsthand familiarity with this, and this is where behavioral studies are going now. Even if you, you know, didn't think you needed this stuff now, every paper that's getting published is kind of validating their observations with one of these tools. Right. So for me, the project kind of began with an interest because again, I'm in a lab that uses a model that is not famously used in neuroscience as much anymore, but it's an old neuroscience model. It's a salamander. And so they wanted to do behavioral tracking of this animal. And of course, it's always easier to do behavioral tracking of animals that these tools have been kind of optimized for. So they were just using a whole bunch of stuff. They were throwing different. I don't want to get into too many specifics to call anything out, but the point is just to say that they were using many different methods. To me, it became really interesting because I saw each of these methods encoded different ways of thinking about the behavior. And the scientists who were using them knew this, but outside of those users, people weren't really aware. And also, I think that it's just until you're really using these tools, you don't understand how it's kind of shifting your way of seeing what's happening. And so it started with just this question of, like, what are the different conceptual assumptions that are getting kind of like, built into these tools that I can sort of just like, document, you know, across the board, and especially when it comes to seemingly unsupervised methods, because I could tell that even though these things are supervised, there's a lot of supervision still happening. And I want to flesh that out as I've been spending time with engineers. I think that this project has kind of blown up in some sense because it speaks to. This is almost a case that speaks to a shifting moment in neuroscience where engineering and neuroscience are. Their relationship with one another is like changing to some degree. So maybe it's not even good to kind of pitch them as like separate in some sense. But what I see here is that the tool builders, that's maybe what I should say. It's like the people who are coding these things and trying to build them in ways that can be generalizable and the technical expertise it takes to do this is creating a meaningful space within these laboratories for engineers that I don't think existed before. Right. [01:40:29] Speaker B: Yeah. [01:40:30] Speaker A: So I started. Yeah, so go ahead. [01:40:33] Speaker B: Yeah, well, back in the. I'm just thinking like back in the old days, I mean, you sort of, the, the neuroscientist was partly an engineer and you still kind of have to be because you have to know how to, how the oscilloscope works and connect it to the stimulator. And so there are engineering aspects, but that's not what you're talking about. You're talking about these new highly sophisticated tools. [01:40:54] Speaker A: Yeah, totally. Yeah. So definitely, like you, you still have to have that. You have to kind of create the stuff you're working with to some degree. But no, this is, this is different. And I think that I found it really interesting just kind of going on these visits, like to different groups, seeing the model, the different models that exist right now. So there are now. And I had, I don't think that this really existed before. There are these kind of like third party groups that are just engineers. These are software engineers or people with that kind of background who are consulted as part of labs. There are also labs that are employing software engineers within them. And it's interesting to see the variability too with respect to how they're treated. So some of them like, you know, attend every lab meeting and they, you know, even when they're gonna like hire some, a postdoc, like they're there on the ground with everyone else, you know, like part of the group. Whereas like, some of them are just like. No, they go off and do their own thing. They come to lab meetings to like present the progress they've made, but outside of that, they're not seen as like part of the science. And so, you know, there's, there's that aspect of this project that's become really like, of interest to me. Just like, how is the line between tool building and science getting shaped as a result of all of this? So there's that side of it. And I think there's also another side of it that I've become interested in. And that's sort of the opacity stuff that we were talking about earlier, which is, if it's not the case that these things are just technically opaque, what are other ways in which opacity is kind of entering this that maybe have to relate to the restructuring of the science with this engineering here? I found it really fascinating that even within the labs that these tools have been developed in, not everyone's using them. They don't use them because they have varying degrees of confidence in their ability to use it for their experimental question expertise. Right. And the thing is, like, that's not a knowledge acquisition problem. Like, these people are, first of all very smart. I mean, they could learn it if they wanted to. They second of all, have the physical resources at hand. They have like, you know, the engineers on site, or they have, you know, their scientist friend who could help them with this. So the reason for not adopting it and not feeling confident in adopting it has to do with a bunch of other stuff that isn't about like, you know, not not having technical knowledge. And that to me, interesting is like, okay, what started out as this question about like a conceptual comparison, you know, is now turning into this like, much more complicated thing where what the tool even is and how to understand it is shifting the way that we like, think about what understanding something within these, like, physical environments is. So, so that is where this project has gone. And so I think I've been in contact with other groups who are using now AI agents to help do science and just more directly thinking about AI as a replacement of the scientists. But at the moment, I'm kind of really focused on this case space because I think that it's an interesting one for thinking about behavior conceptually. Like, how is this. How is this understanding the space materially better going to help us understand how we're studying behavior and how that's changing through these AI tools? So that's where this is at the moment. Yeah. [01:44:58] Speaker B: Well, you certainly have a lot of work ahead of you. There's so many different directions that can go. So you're not. Not suffering for lack of interesting things to do. So that's awesome. So maybe just a couple quick questions to finish this up and then I will not demand any of your any more of your time. One is, it's just very simple. Like, do you. Should I hire a philosopher in my lab when I start running a lab? You recommend it? [01:45:28] Speaker A: Yeah. For sure. I mean, I think. Think you should hire someone you like. [01:45:32] Speaker B: What are you gonna say? [01:45:33] Speaker A: No, no, no, no, I think you should hire someone you like, someone you get along with. I think that's really important and absolutely, you should have a philosopher. I think that. Yeah, I've said this to you, like when we first met. I think of your podcast as a philosophical resource. I think that there's. Yeah, I think it's been. [01:45:53] Speaker B: That's a compliment. You don't realize that, but that's a compliment to me. [01:45:56] Speaker A: So. Yeah, yeah, no, I, I think, I think it's a common. I think it's a really interesting space for doing philosophy of science in a way, because you get to kind of picket the closeted things that people are thinking about when they're doing their studies and what motivates them. And you ask a whole bunch of interesting questions. I think in some sense you're already doing some kind of philosophy of science [01:46:24] Speaker B: here, some light bastardized version. Yeah, no, that's fine. [01:46:29] Speaker A: No, I wouldn't think of it that way. I mean, I just think it's just, it's important to be patient with it because I think that when I first start going into any environment, unless it's the way that I told you I was doing it now, which is, you know, I have these qualitative methods I'm exploiting at this moment. Yeah, I think it just takes time to really get an understanding of what the lab is also about. [01:47:02] Speaker B: Because from the philosopher's perspective. It takes time, you mean? [01:47:06] Speaker A: Yeah, and also I think even from the scientists, like, if you ask people, they can tell you what they're studying, they can tell you what interventions they're making and all of these things. But then we ask them, what's your kind of overarching program or what kind of orientations do you have to thinking about these issues? I mean, people come with varying degrees of self awareness in these things. Right. [01:47:30] Speaker B: But also, especially coming as a graduate student and you join the lab, you're actually learning your orientation as you're doing the science. Right. You don't cut. People think. You come in with like a total worldview, but you're actually learning that worldview as you go. [01:47:43] Speaker A: Totally, totally. But I would say, like the culture of the lab in general, like the. As an entity, like a whole thing, even that can be like really not, you know, something discussed. And it's nice to have a philosopher sometimes in a lab meeting asking a very obvious question, because these things are not. Sometimes even you don't know the answer you haven't asked. You know, no one else has asked the question. And so I think it depends on how much patience you have for this kind of thing. The people I work with right now, I wouldn't say that they're impatient or patient people, but they've at least been very open to me being there. I will say that one of the biggest advantages that I had was that my funding was not tied to them. So one of the really nice things about Columbia's program and unfortunately no longer exists, but it allowed me to have independence in these spaces that I think is very important. So if you kind of do invite people in whose job is not to just produce data for you and that kind of thing, it's also important to get clear on what their job is if you're employing them. Them. Oh, yeah, yeah. So having clarity around that, I think matters. [01:49:11] Speaker B: And for everyone else in the lab to also know what their job is. Right. Last, last thing. So you got into it via, got into all of this via interest in consciousness and the potential of neuroscience to, to make progress on consciousness. Do you think that you'll ever go back to studying aspects of consciousness or are you too far embedded in what you're doing now, now? [01:49:34] Speaker A: So I, I do think what I'm doing now is certainly still on the path. So I'm. Like I said, I think this shifted to an interest in behavior because a lot of neuroscience is studying non human animals and it's so hard to understand consciousness in these systems. So we test them behaviorally and then draw extrapolate from that, you know, and, and I think that one of the things I'm doing right now is really trying to bridge philosophy of science, like this kind of science and practice work with philosophy of mind and thinking about the kind of categories and phenomena that relate to these things that are relevant for consciousness. So that I think I'm still kind of, you know, I just, I think that you go in pretty naive about something and then you realize how complicated it is and you start studying a little piece of it just like, you know anyone else. And that's, that's the journey so far. [01:50:37] Speaker B: Yeah, well, in philosophy of mind and consciousness philosophy, it's so crowded too, that like, it's sort of. I mean, I don't. When you think in neuroscience, when you think of like, oh, okay, I'm going to study, I don't know what hippocampus, right? Like, I'm going to study cognitive maps. And it's like, oh, God, everybody's studying, like, I Don't know you. It. There's an advantage to studying something that is less crowded and more niche. Right. Because there's more room to like blossom and make, and make claims and make progress. I hate to use the word progress or whatever, but, but just in terms of a strategy. Yeah, it might, it might make sense. Like your path might end up making the most sense. Even if you, you start carving out the space from where you are now into philosophy of mind from the perspective that you've gained, you know, and the tools that you've gained philosophically from what you've done so far, that might be a beautiful way to come back to it. So. [01:51:30] Speaker A: Yeah, yeah, but also I'll say, like, you know, maybe from us, where we stand, things look crowded in that area. Like, generally speaking, though, all of this is not crowded enough. Like, if there were more people, you know, using their brain power to just do work at the intersection of these fields, I would be very happy. You know, I don't think that. [01:51:53] Speaker B: Well, but what you're, what you're doing is not crowded. That's what I'm saying is like you, you have like a. That's why there's like so much to do in it. Right. [01:52:01] Speaker A: Actually, I don't think it's crowded at all. Like even this other area, you know, of consciousness science or, you know, people thinking about these other questions. I mean, yeah, like I'm saying, you know, kind of local to us, maybe it looks that way, but I still, but yeah, we, we could use a lot more people thinking about these questions in general. So. Yeah. [01:52:22] Speaker B: All right, Netta put the call out, so. All right, well, thank you for you've gone over a little bit with me, so I'm sorry to keep you over, but thank you so much for joining me. Your work is cool and I wish you success in your new job. Congrats on the new job there. So. [01:52:37] Speaker A: All right, thank you for having me. This was so fun. Thank. [01:52:45] Speaker B: You. Brain Inspired is powered by the Transmitter, an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives written by journalists and scientists. If you value Brain Inspired, support it through Patreon to access full length episodes, join our Discord community and even influence who I invite to the podcast. Go to BrainInspired Co to learn more. The music you hear is a little slow jazzy blues performed by my friend Kyle Donovan. Thank you for your support. See you next time.