[00:00:03] Speaker A: I would say that I've been thinking about what a thought is my whole life.
I have had a readiness to just suck up stuff that when it resonated with me. This is why I'm a bad philosopher. Like a good philosophy. I took a fair amount of philosophy in college, but I sucked at it because I know what I agree with and disagree with. I don't just take their arguments and knock them down and do all the things that real careful scholarly philosophers do.
A thought is structured as interrogation. We are asking questions of the world when we're awake, and those questions are motivated by the affordances in the environment.
So, okay, so. But there's one particular affordance that, that I think is unique to humans and it's part of our evolution. So it's grounded. I could say it's Gibsonian in a very, very weak way. It's that everything that we experience, every non conscious thought, affords the possibility of reporting to another agent that has a mind like mine. That is what makes non conscious thoughts consc.
[00:01:28] Speaker B: This is brain inspired, Powered by the transmitter. Michael Shadlin is a professor of neuroscience in the Department of Neuroscience at Columbia University, where he's the principal investigator of the Shadlin Lab.
If you study the neural basis of decision making like I did, you already know Mike's extensive research because it's in that case you are constantly referring to it if you're not already working in his lab.
The name Shadlin adorns many, many papers relating the behavior and neural activity of decision making to mathematical models that are in what are called the drift diffusion family of models, the idea being that decisions are formed after evidence drifts or accumulates toward a boundary. That's not the only work he's known for, and I will leave it as an exercise to you to learn more about Mike and his work in the show notes. As you may have gleaned from those little intro clips, Mike is with me today to discuss his account of what makes a thought conscious in the hopes to inspire neuroscience research that will eventually tackle the hard problem of consciousness, why and how we have subjective experience.
But Mike's account is not an account of just consciousness. It's an account of non conscious thought and conscious thought, and how thoughts go from non conscious to conscious. His account is inspired by multiple sources and lines of reasoning. Partly, Mike refers to philosophical accounts of cognition by people like Marlowe Ponty and James Gibson, appreciating the embodied and ecological aspects of cognition. And much of his account derives from his own decades of research studying decision making, mostly using perceptual choice tasks where animals make eye movements to report their decisions. And you'll I just said the word report here and you probably heard it in the intro clips. But reporting turns out to be a key aspect of his account here. So we discussed some of that, including what we continue to learn about neurobiological, neurophysiological, and anatomical details of brains.
We even discussed the possibility of AI consciousness in light of Mike's account. What we don't discuss are some of the other topics in Mike's paper, like how his account relates to other accounts of consciousness, for example. So again, I will leave that to you to learn more. You can find that in the show notes@BrainInspired Co Podcast 230Support this podcast on Patreon if you want all the full episodes, a Discord community, access to various bells and whistles. Learn more@braininspired co for that, thank you to my Patreon supporters and thank you to the transmitter for continuing to support this old brain inspired podcast. Learn more about
[email protected] here's Michael.
On my way up here to talk to you, actually, I said to my wife, I said, do you remember when I was in grad school and I worried, I was worrying a lot about getting scooped by this much more established group?
And she said, yeah, you talked about that a lot. And I said, I'm going to talk to him right now.
That was you in graduate school, I did work on what we were calling metacognition in Non Human primates in Mark Summers lab.
And I think the last time that I interacted with you actually was at a poster at SFN when I was a graduate school, maybe 2008. I want to say you, Ruzbekiani, who was doing the confidence work that you, I guess were just starting to work on back then. But anyway, he came to my poster and then he went and he got you and I remember you and he standing there looking at my poster. And so I've come full circle. It's been a long time. But nice to see you again.
[00:05:34] Speaker A: Nice to see you too. I remember that though, you know, I mean, you know Mark Summers, your, your advisor, right? He, you know, he's like one of my heroes in science. And we're both grandchildren of Bob Wirtz. Well, I'm a grand. You and I are both grandchildren of Bob Wirtz because Mark is a direct descendant.
[00:05:53] Speaker B: Essentially, he's a son.
I was right. You guys did scoop me.
And of course you were more Established. And you've continued that line of work and I know Ruspe has as well. You guys continued on. I went off and did other things besides the confidence metacognition work. But.
[00:06:15] Speaker A: We were both scooped by.
[00:06:17] Speaker B: You're talking about Adam Kepex?
[00:06:19] Speaker A: Yeah, exactly.
[00:06:21] Speaker B: Yeah, but that was in rats.
[00:06:23] Speaker A: Yeah, yeah.
[00:06:25] Speaker B: Anyways, you know, the other thing that's kind of full circle is I designed my PhD work, that experiment that we're talking about, because I wanted to study the closest thing possible I could to consciousness.
And that's how I ended up designing that experiment. And I guess metacognition is the closest thing that I could maybe teach monkeys to do. And then we could look for neural correlates and stuff. And so that's how I came about. And here today we, we're going to be talking probably a lot about your decision making work, but also about your thoughts and account of consciousness. So I'm excited to do this, but also a lot of things full circle.
So let me start maybe with a really super broad description of what I took from reading your account of consciousness. Then you can correct me and then we can get into everything.
[00:07:19] Speaker A: Great.
[00:07:19] Speaker B: So my account, and of course it's super broad and it's going to be wrong. We as humans and organisms always have this kind of continuous background, what you call interrogation, where we're kind of grasping at the world.
Coping is another way that Hubert Dreyfuss might have called it in alluding to people like Heidegger and Marlo Ponte. And we have this kind of constant grasping and interrogating the world that's going on and that's non conscious, which you link to decision making.
And it's only when that constant. And you actually, I should say, you call this like a weak phenomenology which we can get into. And so this is kind of constantly going on in the background. And it's only when those processes are joined up with another kind of process which is related to theory of mind and the need to report or the desire or the intention to report what's going on to ourselves or to others, that interrogative processing becomes conscious. So that's very, very, very high level and wrong. So correct me.
[00:08:34] Speaker A: I mean, it was pretty good. I should just substitute that for the first few paragraphs of the paper. But it's.
Yeah, I mean, I think consciousness is, you know, it's a big topic. And if people will say that I haven't even scratched it because I haven't gotten into the phenomenology part. I'M leaving that almost as an exercise to the reader.
I claim that I'm not saying anything about it, but actually I think I'm laying down the how to go about how it'll be solved, the so called hard problem. But let's leave that because that's where that weak phenomenology comes in. Philosophers hate it when I say that non conscious thoughts have a what it's like aspect. Like you know, if you knew that you got your information from the visual cortex. It's very like, it's like other things that you probably would have to walk to before you'd feel them on your skin. Okay. That's what I mean by weak phenomenology or at least a weak what it's like. Okay. And so, and so I think if we understood how it is that the brain.
I'm getting way ahead of myself. Like this is like the last part of the whole theory.
[00:09:43] Speaker B: We don't want to bury the lead. Yeah, yeah.
[00:09:46] Speaker A: Okay. Well, well it isn't the lead. It's, it's to me it's my feeble attempt at trying to attack the hard problem. So I've met Dan Dennett before he died and I, I discussed some of these ideas, especially the reporting of Ford. And we'll get back to that, of course.
And he was pretty keen on it. But he is pretty, he has this term heterophenomenology where the interesting question to him is well, why do people even think that's an interesting issue?
And I think, and he was widely criticized. I think Ned Block called in a book review of Consciousness Explained. Was Consciousness Explained Away. That was the title. Yeah. And, and you know, he had a point. I mean, because I mean if you like, if you read a Neil Seth, for example, I mean he starts with phenomenology and many people do. So I think I'm, I'm starting from a kind of a neurology perspective and a, and a neuro. Neurobiology of decision making perspective. And, and so I start with building up what is a thought. So the title of this paper, it's not out yet but you know I just, I just submitted it actually is Conscious and non Conscious Thoughts. So the first thing to start with, I think if I, you know, if we now go back to the. The simplest simple ideas is, is what is a thought.
So I mean that's a fun question. You probably get asked it all the time. A lot of my colleagues do and they, we all stumble. But I, I think I have an answer to that. And the thought is an intention and an intention not necessarily to report, but it's an intention to do something that is sort of an answer to yes, an answer yes to. So I'm going to call it a provisional intention to do something. And that do something is. And the question is, might I, might I do this? And these are all things that, this is now being Gibsonian here, that start as what in ecological theory of perception are the things in the world that get us to say, oh, I might eat this, oh, I might mate with this, oh, I might flee from this. You know, all kind things like that. And so, and that's what being awake is like.
You know, it's the stuff that's going on behind the scenes. No consciousness needed. We just were in interact with the world by asking questions.
[00:12:25] Speaker B: And so, yeah, in that sense it's a very agential account.
It's a very active, where you're putting action at the forefront, which distinguishes it from a sort of passive perception based account of consciousness and thought.
[00:12:41] Speaker A: But see, Merleau Ponty, for example, and I, you know, would say, and so would Gibson, he had some funny terms like direct perception and pick up and all kinds of things that was, I was totally turned off to, to Gibson when I first read him because it was the days of David Marr and I was all swept up in that, which I now shun. But we'll get into that maybe.
[00:13:03] Speaker B: Yeah, we'll have to talk about that.
[00:13:06] Speaker A: But at the time I thought, I just thought I couldn't make it. I mean, through the book and even when I read it more seriously because I came around to having to admit that I was basically Gibsonian and so was Merleau Ponty and then, and, and, and, and it just makes a lot of sense just that some of those terms are still highlighted in my copy of that. Just drive me up a wall. And they should, in a way. I, I agree, they sound magical pickup, stuff like that.
[00:13:38] Speaker B: It was, it's funny not to. Sorry to interrupt, but I, I was just visiting with Carol Colby, who was, is an old lip researcher, among other things, and she's retired now, but I went to visit her and she mentioned that she loved Gibson. And I was like sort of shocked at that because I wouldn't have placed it. But I think we all love Gibson, if we understand Gibson through the right interpreters and don't use those terms, you know, direct perception, resonance, things like that, that are mysterious and it's hard to understand what they mean.
[00:14:10] Speaker A: Right.
But if you think of it as interrogation and that's where in the essay road, you know, I, I draw on.
On the parietal neglect, you know, and you know that we, we are what we have. What we lose in parietal neglect is this interrogation mode. So, I mean, I would argue the easiest way to put it, when we can embellish it, is that a thought is structured as interrogation. We are asking questions of the world when we're awake, and those questions are motivated by the affordances in the environment.
And when we get, you know, and those questions are, might I look there, for example? That's, you know, in the easiest example to talk about.
And that's a thought about the awareness of something. You've yet to even discover what it is.
And the reason I enjoy riffing on it this way, in calling it provisional intention, which it is, is that you've geared up your body or you're prepared to do something provisional, meaning that you don't have to do it to have that thought. But that's what a thought is.
And I'm very critical of people adding magical potions or just claiming that a bunch of spikes are a percept. You know, my freshman year in college, I wrote an essay that got me in a little bit of trouble. But, but I, you know, I criticize this, this assumption that what. What goes on the retina and then through the various stages of the visual system. And that, that that means since it's in neurons, that it is seen.
And that's a big problem for the field because the field actually invented attention on the false premise that what goes into our eyes and so forth into the visual system in this process will be seen. And so then we said, but that's not true. We know that's not true. Well, so then attention came along. But in my view, when we understand the brain, how the brain works, say a year from now, we'll say there was this word people used to throw around called attention. We don't need it anymore.
[00:16:30] Speaker B: Oh, you one of those.
[00:16:32] Speaker A: You were supposed to laugh at that.
[00:16:34] Speaker B: No, that's.
[00:16:36] Speaker A: Anyway, yeah, so, anyway, so thinking of the brain, asking questions which are, might I type questions? Meaning kind of here I'm embracing a kind of soft embodiment.
But not that we have to actually act. It's just that we are organizing our understanding of the world in terms of what we might do. So I claim that there's no what and where pathway, there's how and how pathways. And they're just about different kinds of endeavors.
[00:17:10] Speaker B: So, okay, let's linger on the word thought for A moment. Because in some sense, one could argue that you're.
And this is not a knock on it at all, but operationally defining the term because we have these terms like consciousness, thought. And you said, what do you say when people ask you what a thought is? And it depends on what you mean by. By thought. Right. You know, it depends on where you're coming from. And so in how much are you of this? Are you sort of defining thought this way versus discovering what this ontological thing that we call a thought and then describing it, for example?
[00:17:45] Speaker A: Well, I'm kind of giving it a neurobiological interpretation. I'm saying that what a thought is in the brain is persistent activity that represents a provisional intention.
And the nice thing about that is that it avoids the category error of. This is what I wrote my essay about. A little aside, just to finish that thought. Is that.
Is that. Sorry to use the word thought there, but the. Is that I said we opened up a hole in the head and craniotomy, as we call it now, now that I've been to med school and I didn't know anything then, but the. And I put a dime and I pressed it onto the cortex and it made. It was represented in the cortex. I could see the representation of that dime in the cortex. How come we don't see it?
Okay. And I'm just sort of making fun of people who think in the Hubel and Weasel days that because these spikes were coming in, that's how you saw. There's a big gap there. And philosophers call it category error. You know, it's like spikes are one thing, they're electricity, you know, all the electrophysiology. And then, and then a thought is something else. And, you know, we're not even going into the. At this point, the distinction between that and conscious thought. But, but, but, but. So I'm just saying that, that, like, what I've been studying in my whole career in decision making is that something comes in and you deal with it and you. You're effectively asking a question, yes or no, left or right, mate with this or flee from it, whatever. You know, it's where. And that's, you know, some, you know, taking examples from Gibson and examples from really, you know, pedestrian or just. Or our contrived tasks like the ones that I study.
But they, but what the thought is and why it's justified to call it a thought and not just spikes is because it does have a weak something like. So it's sort of like, I think to try to.
Because it's an intention. That might be all it takes for philosophers to kind of see the appeal there.
But I think it's also worth thinking that, that it's a bit like relating.
What if you have isometric, you know, isometric contraction and you can do, you know, by stiffening at the, at your, you know, at the elbow, you know, at the biceps, you know, and if you do that, the arm doesn't move, but you're doing exactly what it would be when it does move. So I think that this working memory like signal, you could call it, but this, I'll just say persistent activity that is the outcome of some commitment which makes it decision like and the outcome of some, I should say evaluation that you are now committed to as a provisional intention. So in other words, not really committed. You could change your mind, you could just drop the whole thing.
But that's why I think that you sort of escape from the category error complaint, which is a legitimate thing to complain about.
So that's what I think I'm doing with that argument.
[00:21:03] Speaker B: So when you're giving, I've seen you giving a talk about these and you use the example of like memory guided saccades, right?
And you show the activity of a single neuron in for example, lateral inter parietal cortex that fires in the absence of stimulation, like it's holding onto that thought. And you call it like that, that you point to it and I'm not sure if verbatim you say like that is the, you know, the intention.
That's the interrogation right there. But you know, what you just said is that the neural activity is not the interrogation.
What I'm curious about is like. And you say like. Well that's kind of shorthand for you know what I'm calling here a thought. So what? And you've been a single neurons person for a long time and you still look at single neurons among the many others that we can record with these multi contact probe.
Um, and in fact it's only fresh on my mind because visiting with Carol, she said, Carol Colby, she's like, I'm still a single neuron person. I'm like, well you don't really, you know. But anyway. But so what do you.
So that is shorthand. But then the neural activity is not the thought. And I know and we'll come to it that the consciousness aspect is. It turns out to be kind of a relational phenomenon in your account, I believe. So then what is the thought? Is it is. Are the single neurons thoughts? Are little populations of neurons thoughts or.
[00:22:27] Speaker A: Those Readouts of a thought, the persistent. When. First of all, when we always talk about a single neuron, we're always talking about a population of neurons that do the same kind of thing. Barlow, I think, you know, he did his rewrite of his famous perception and single neuron kind of thing, and he, he, for a long time didn't, didn't really like the idea. But I mean, but when, when I was in Bill Newsom's lab and doing some of those early papers on relating the firing and the noisiness of neurons to perceptual fidelity, you know, those. In Bill's, In Bill's lab, I always thought about neurons as many neurons that were representing the same quantity. The quantity would be. So to me, good computational neuroscience starts with an idea about what the computation ought to be. And then. But all. But the way it plays out in the brain is that if you have your variable X and X is doing something increasing with contrast, whatever it is, you have firing rates of many neurons that are all effectively tuned, you might say, or selective or sensitive to the same kinds of things. So that's what Barlow, in his rewrite, which appears as a chapter in one of those Gazzaniga gigantic textbooks.
Um, and, and, and it's. And, and I credit him with this notion of direct coding. You know, the opposite of that is distributed coding, which has taken over all of like theoretical neuroscience. Most of your guests are big fans of Hopfield.
[00:24:02] Speaker B: It's all manifolds.
[00:24:03] Speaker A: Yeah, okay. Manifolds are fine. They should just stick to, to mathematics, you know, topology.
So anyway.
But, yeah, but, but, but you need many, many. So whenever I sit with someone, like Carol says single neuron, or whenever I study a single neuron, I'm assuming that there are many neurons just like it. And that's because if we, if it wasn't the case, then you would not have any kind of analog signal to compute with the analog signal being the average of many, many, many spikes. But if you had to wait for that quantity between spikes, there'd be no way to, you know, the brain would not have worked. You know, we wouldn't have this brain by evolution.
So, so, so a sort of redundancy.
[00:24:51] Speaker B: But not so much redundancy that they're all doing the same thing, redundancies to get the noise out of the, get the signal out of the noise.
[00:24:58] Speaker A: That's right. That's. Yes. And there's a limit to that too. But the. Yes, I would just say the, the.
This. The primary for the brain to do any kind of operations an Inequality, Any, you name it, any operation you want it to do, to move a body part, to move a muscle, you need a signal that is present to work with in the inter spike interval of any one neuron.
That's the primary desideratum. And from that flows a ton of stuff, which is why neurons are noisy. Because if you have that many spikes, if you have, you know, if you, if you need, you know, 100 EPSPs per, you know, per, you know, in order to, for, for every one spike you're receiving, okay, so you get, if you inundate the neuron with a hundred inputs to, to produce, you know, 10 spikes per second, you're not going to, you shouldn't get 10 spikes per second. When, if, if, if you had just excitatory networks, you'd get an explosion and you'd have no dynamic range. So the beginning of one of the things that kind of, I think it was a lucky stroke, put me on the map, was my early papers on why neurons are noisy. And they're noisy because you have to balance excitation with inhibition or you would not be able to satisfy that desideratum of having an analog quantity in the inner spike interval of any one neuron.
And so, and once you realize that you have a balance of excitation inhibition, then spiking becomes first passage times.
The spike times are, you know, you don't think of them as more excitation and less inhibition. You think of them as balanced more excitation and more inhibition.
And that means that you can satisfy your spikes then become effectively first passage times of like Brownian motion.
[00:27:04] Speaker B: So in that, in that view, so we have lots of neurons.
And is, is what you're saying an account of why we have lots of neurons. So one way to think of it is if it takes like so many neurons just to get an analog signal that we can compute with, gosh, that's a lot of energy expenditure. But it also means that a lot of the energy is taken up by just getting it right.
And an alternative way of looking at why we have so many neurons, for example, is that we are so intelligent, we have such capacity. And having those many, many neurons, in a distributional sense, allows the state space of our capacity to totally expand. And that's why we're so awesome and smart and intelligent and all that jazz. So are those two in opposition to each other, if that makes sense, what I just said.
[00:27:59] Speaker A: I mean, you can say that you have a lot of neurons. I think what you're trying to say is that you can have a lot of neurons.
I'm going to disagree with one thing you said that the energy, that it's very expensive, energetically. I'll come back to why I've never really been a fan of that idea, which gave rise to sparse coding and all kinds of other nonsense.
But you're talking about not so much a sparse code, you're talking about a very highly high bandwidth kind of code, a high dimensional kind of code. Right. So many, many neurons, gives you mixed selectivity and all kinds of things like that. And you could take advantage of that in modern state space. You were using that kind of language a moment ago.
And I'm going to. I've embraced, of course, state space methodologies, but I think they mainly obfuscate the underlying neurobiology.
We talk, go around talking about geometry and stuff, but actually you're really just saying, oh, there's a bunch of neurons that fire for this and not for that.
Oh, well, okay. They contribute weights to this vector, not that vector.
So most of orthogonality is just sparse orthogonality, not functional orthogonality like sine and cosine.
So anyway, that's a side point, but it's a little bit of a peeve I have with the field. It's not that I don't embrace all those techniques. And we do tons of things using, as you anticipated, not really single neuron, but more modern subspace, all the linear algebra stuff. They're just tricks though. They're tricks to get out a signal and then we can see what those signals are doing. For us, it's a game changer because, you know, we've been for years thinking that we are looking at this stochastic process on every trial in the decision making work that looks like, you know, the stochastic component of it being diffusion. And the problem with that is, is that as an experimenter with spikes, you typically average over trials. And if you average over trials, you do that to suppress the noise, the neural noise that I was saying is caused by EI balance and this high input regime we used to call it, you know, just being able to a ton of. A ton of inputs, a ton of excitatory input that has to be balanced off. And that leads you to these noisy neurons. But okay, so they're not neuro. When the neurons are noisy, we, you know, average over many trials and we get out the deterministic component, which is classic, you know, peri stimulus time, histogram type thinking.
And but of course, averaging also suppresses any stochastic component that's different from trial to trial. So you can't see drip diffusion.
You can see the drift part. That's the deterministic part when you average, but you don't, but you lose the diffusion part. And that, that's something that like, kept. Made me lose a lot of sleep for many years because there were, you know, it's like people were, you know, claiming that, you know, if you, if you, if you look at a, at a ramp, you know, which is what, what the drift looks like in the brain, then you think that, you know, a lot of things can average to a ramp.
There are a lot of stochastic processes that could become that ramp, you know, and, you know, they include steps which for a long time used to bu.
Doesn't anymore because with single trial resolution, with neural pixels, probes, using, you know, state space methods or just dimensionality reduction methods, you name it. These are really simple things that now we can see the signal that's accompanying a single decision on that trial. And we, you know, it's like it doesn't take anything to see the, to notice that it's a diffusion process.
[00:31:43] Speaker B: Let me, let me just bring those in the audience who are not familiar with this kind of work just, just on a high level, bring them back to like what we're discussing. So, so Mike has spent a good deal of his career studying a drift diffusion process, studying decision making in, in terms of thinking of it as a, an accumulation or a diffusion of evidence toward a bound. And for a long time this was done with single neurons.
So you can watch the neuron kind of ramp up. And when it, the idea is that when it hits a certain level, then that actuates a decision, for example. And for a long time this was done recording single neurons. So like Mike was saying, you record many, many single neurons over many trials, then you average those neurons. And this is just what was available technologically, what people could record in brains. And many people worried that this was obfuscating underlying processes that actually account mechanistically for this decision kind of process, not just in decision making, but in all sorts of neuroscience. This was a worry for, for many years. So now we have the technology to record many, many neurons and you can start to unearth some of these single trial kinds of processes and dynamics and then better interrogate, I'll say those, those mechanisms. I use the word interrogation there because it plays looms large in your account of what knowledge is and what thoughts are and how we go about and how we eventually have consciousness, which we should get back to Because I knew that we were going to do this, get, get start to talk in depth about the actual neuroscience behind the decision making and these processes, which is great. But I also want to make sure that we, we focus on, on this account and we have to have that discussion a little bit to get to where you have gone.
So sorry for that interruption. And now I think I feel I've knocked us off the path and I didn't mean to force us back onto the path.
[00:33:39] Speaker A: I forget which path we have. We have.
[00:33:41] Speaker B: Well, I was going to bring us back somehow.
[00:33:43] Speaker A: Yeah. All kinds of forks on the road. No, that was great. I mean you nailed it. And so it's.
But it is true. I don't want anyone to listen to this and say that if you see something in average rates you, that you've got, it's going to be drift diffusion or it's diffusion or Brownian motion. You don't know that. It could be that. It could be that it is, but it also could be that it's a bunch of things that step at random times.
And I always thought that was a possibility and even thought it was kind of evolutionarily kind of and biophysically plausible because it would just be a matter of like having different kinds of concentrations of channels of different kinds of conductances in neurons that would allow them to kind of just trigger into some high state, high low states.
And there are examples of that actually in decision making in the work of Thiago Branco in the brainstem.
Beautiful work. If your listeners don't follow this guy.
He's done just wonderful work in an area, area of decision making I would call it. But also of very, very much in the ecological forms of perception.
Ecological theory. I'm just. Whenever I say ecological I just mean Gibson.
Okay. Just to invoke him.
But David Sparks was the first person to doubt telling you every time you see a ramp like a buildup thing going, you can just think of that as just different steps at different times. And it's just something that everyone has to exercise.
As I see this function coming out in my. Once I've analyzed the data and what am I really saying and always considering the alternatives.
[00:35:29] Speaker B: Your point here is that. So that steps account caused a bit of controversy as an alternative account to the drift diffusion model, for example, to the ramping.
But your point was that now with these multi electrode contact signals where we can record lots and lots of neurons, we can actually get at whether that is true and what that underlying mechanism is, right?
[00:35:53] Speaker A: Yep. And you can see well, the mechanism's a little bit of a harder nut to crack, of course, but the, but you can see that. You just can see the sig. You can make out firing rate as a function of time, but it's a rate over a bunch of neurons, you know, and they're just weighted averages though, right? So it's a scalar value.
Okay. And it's the kinds of intensities that I was calling analog, despite the fact that their underlying roots are not analog, they're spikes.
[00:36:21] Speaker B: So everyone can hear that you're well versed and people who don't know you, my audience largely will know you. But those who don't, I worried a little bit if we didn't talk a little more nitty gritty neuroscience that people wouldn't appreciate, you know, your, your long tenured research career studying these sorts of decision making processes and, and how the brain operates and maybe getting. One way to get us back toward what the main theme of, of our conversation is supposed to be is to ask you an awful, awful question, which I could never answer if someone asked me, and that is whether you can articulate how you came.
So, okay, so when I, I was interested in neuroscience because I was interested in consciousness and understanding what it is and how it is, right? And that sort of guided my questions early on. But then, of course, I immediately was thrown into a world of like, oh, there's a billion unanswered questions. And then you start working on smaller and smaller questions. Do you know how a decision is made? Do you even know, like, if you're looking at that neuron, how to tell, oh, is it just signal or is it noise? And so then you sort of ratchet down into the smaller and smaller questions that are needed for these larger questions. But kind of in the back of my mind, you know, has always been like, well, what I'm really interested in is like this, you know, consciousness, for example. So is it. The question is, can you articulate sort of how you came to be, to think the way that you do currently about this knowledge as interrogation and as a weak phenomenology always undergirding our ongoing, ongoing cognition, let's say. Or was this at the forefront? Did you think of. Oh, I was. I've been interested in consciousness. This has driven me ever since. How did that interaction between your years of research and your current thinking take place? Maybe that's a way to get us back into it.
[00:38:17] Speaker A: Okay, I think I have a common story. I think, I think there are other people you pat on that had the same story about the first time I had anesthesia. I was four years old. I remember it. I remember the smell of the ether, this tonsillectomy. And. And I remember thinking, oh, my God, no, time passed. It's not like sleeping. Oh, okay. And. And, and that stuck with me. And then I. I had a repeat operation for some other surgeon who wanted to send their kid to college or something. Unnecessary hernia surgery at age 8.
And it was a different anesthetic.
I forget whether it was penobarbital or something like that, but it definitely wasn't ether.
But I had the exact same experience of no time passing.
And so that stuck with me. And I would say that I've been thinking about what a thought is my whole life, getting into weird arguments with people that get the first time. But what it prepared me for, and I think being ready, it's almost a Zen thought, is that I have had a readiness to just suck up stuff that when it resonated with me, this is why I'm a bad philosopher.
I took a fair amount of philosophy in college, but I sucked at it because I know what I agree with and disagree with. I don't just take their arguments and knock them down and do all the things that real careful scholarly philosophers do, but they tolerate my presence anyway. But, but, but I just, like, I know what resonates. And like, I. Marillo Ponty, Phenomenology of Perception. I read that very early in college, and I ended up reading it several times and taught.
[00:39:59] Speaker B: Required to understand it. Right?
[00:40:01] Speaker A: What?
[00:40:02] Speaker B: Yeah, reading it.
[00:40:03] Speaker A: Several.
[00:40:03] Speaker B: Several. Yeah.
[00:40:04] Speaker A: It. I've given it as a. As a present to people, but I've stopped doing that because I make a lot of enemies. They hate me.
And I'm told by people who speak French that it's not just because of the translation. I mean, he has paragraphs that go on for like four pages, and they change point of view 15 times. Because if he's arguing with Kant here and who knows? You know what I mean?
[00:40:23] Speaker B: And. But he doesn't let you know who he's arguing with. You have to figure it out, right?
[00:40:27] Speaker A: Yeah, yeah, yeah. But for me, I'm like a synesthete with it. It's like this thing just buzz off the page.
Like, one of the things I loved is, like, I love butchering this quote. He says, what is the panorama of vision to the blind man?
You'd say, if you're a dumb psychologist, neuroscientist. There was no neuroscience when you wrote that. You would say, well, the blind man has this stick and he transduces the vibrations in the stick and he knows where his stick is.
And then he picks up and he makes a kind of an anemic, you know, something of texture, something in the world. And it's a sort of a watered down version of what we get. Okay, but, but then Merrill Lepante says, no, that's crazy. The blind man is asking questions of the world. Will I, can I walk here? Will I fall here? Can I fit between here and here? Okay, these are all Gibsonian type things, right?
And, and he gets answers from the world. So it's like, it's not like he's made up the world, like idealism or something. And it's not like he's just not caring about data at all.
But it's not just that he's representing things. He's getting answers to questions and that means they have meaning.
So the brain has asked the question and when it gets the information that said yes, you see, I'm using language like decision making that you committed to something. Of course it's. That means that it has meaning. You asked the question. So I always thought that was. Those kinds of passages really, really knocked me out. When I was a grad student, I left medical school after the second year. I was pretty frustrated. And that's when I did my PhD. I won't talk about that, but I will say that when I was there, I was mainly doing a lot of political stuff and El Salvador issues. But I also co taught, quote, unquote, a course with Herbert Hubert Dreyfus on phenomenology of perception. And I didn't really co. Teach it. I was just the resident neuroscientist and I was just a grad student anyway, I didn't know that much.
But Hubert, Bert, I used to call him, he, he just was a soup. He. He's just like totally into it. And he, and I, I did a favor for him. I, I read his book on being in time. He has this like primer, whatever he was. There's an old fashioned term for, for having a book that helps you get through another book.
But, but, but. So for being in time. He didn't do the time section. I think he never came back to it. But, but I only reason I understand Heidegger is by reading Bert's book.
[00:43:19] Speaker B: And then I've never read Heidegger or Marlo Ponte. It's only from listening to his lectures and stuff.
[00:43:25] Speaker A: Yeah, Merleau Ponty is high. Is Heidegger on neuroscience? I think I mean, you know, his. His readiness to hand. He has these concepts. I mean, it's. It's more metaphysics and then it is, you know, psychology, perception. But. But, I mean, you can sort of see the same things. Although I say that I've. I've studied time and I thought, oh, I better go back to Heidegger and read the time part. I couldn't get through it.
I have a mathematician daughter who read it and thinks she said that.
I've just gotten too caught up in the details, but I have no idea.
I still don't have an explanation of what he thought. I don't really think I care anymore. And especially given what he. The more I know about him historically, I'm happy to cancel.
[00:44:11] Speaker B: Yeah, yeah. Separate the art from the artist.
[00:44:13] Speaker A: Yeah.
[00:44:14] Speaker B: Whether that's a good idea.
But those ideas that sort of undergird your viewpoint are there in Heidegger and there in Marleau Ponte. Right.
[00:44:22] Speaker A: So definitely Merleau Ponty for sure. And then. But, you know, and Dreyfus, you know, he's, you know, he wrote a lot. And, you know, I have a quote from Dreyfus in the paper because he even. He puts together affordances in Merleau Ponty, and he didn't get that from me. I think we did it. You know, I think we probably tumbled to. I'd like to think that we tumbled to it in the same way.
Who knows? Because he's passed away.
And so.
Yeah. But anyway, so that's sort of those roots.
[00:44:54] Speaker B: So it just dawned on me, because we're talking about that. That sort of readiness at hand, you know, and that sort of viewpoint on how we go about. And you use the example of, you know, a blind person with a stick and how they.
How they go about acquiring this interrogation kind of form.
[00:45:11] Speaker A: Right.
[00:45:14] Speaker B: And this doesn't need to be a long aside, but I thought while you're speaking, to ask about.
There's this been. There's a recent sort of turn toward naturalistic kinds of tasks and behaviors, and that we've been doing it all wrong in neuroscience by studying these very reduced experimental setups where we're controlling everything that we can, and that must not be how the brain does things, or it's not ecologically valid, if you're a Gibsonian, for example. Right. So you're doing it all wrong.
And what we need to do is study in, like, a much more grasping and coping and readiness at hand kind of way to actually get. If we want to even get toward studying something like consciousness, because that's how we are in the world. That's how we are beings in the world. Right.
And by studying it in this very reduced way, you're taking away all the context that matters for what you're actually experiencing.
And I ask that, or I say that to ask you what your thoughts are on doing this, because you have studied what's commonly referred to as the dots task, has been the bread and butter of a lot of your work, where it's a decision making test where you have to look at a cloud of somewhat coherently moving dots and say, are they mostly moving to the right or are they mostly moving to the left? And sometimes it's easy to tell which way they're moving, sometimes it's hard.
And this is, you know, exactly in one of these sort of reduced preparation experiments to isolate the decision making component of that cognition is the idea. But what I want, what I'm imagining is that your account here doesn't depend on what kind of experiment that you set up, whether it's naturalistic or whether it's this reduced preparation.
Because the account is, because there are lots of different interrogating processes. This would just be one of them. So that I'm sort of setting you up, but do I have that right, that it's, you're sort of agnostic and it doesn't really matter what kind of situation or preparation experimentally that you set up, because you can study the same process. It's the same process whether you're studying it in a reduced preparation versus out in the wild, for example. Sorry for that long diatribe there.
[00:47:28] Speaker A: Yeah, I mean, I think we could take that in a lot of different directions. But I, I, I, you have to, I think you have to start and say, why am I doing, why am I studying the brain? Okay. For me, it's, it's a biomedical drive. Okay. It's not a, it's not a, I used to say it's not an NSF drive, it's an NIH drive. No. Yeah. But I do think that, you know, we, if you want to know how the brain works, then you have to study the brain.
If you want to know how higher brain function works, you have to study an animal that has higher brain function. If you want to study Gibsonian things, that ecological type things, you are studying the more natural behaviors that the animals use. And that can be useful, but it can also actually get in the way. Because in some sense, if you think of our higher brain function contrived tasks, I defend them as being, as working on answering questions about functions that evolution had not anticipated.
So all of the Gibsonian affordance people use the word affordance really loosely. I've heard you mention that. I think you share that view. But, I mean, in the Gibsonian sense, this is ecological, and these also should have some connection to evolutionary pressures and so forth, you know, and so that's fine. And, you know, I mentioned Thiago Branco. Those are the stuff, things he studies, and they're beautifully studied in the brain stem. But when we encephalized processes, but maintained the being as an awake agent, including, you know, animals in the wild, you know, so I'm thinking animals, we can't study for consciousness. And I'll defend why I think animals, most animals don't have consciousness, is that we want to understand things that are still structured as interrogation. In other words, something that leads you to ask a question about a mighty thing that I could do with my body, let's just say, but also with my mind, and then come to an answer. And if that answer is yes, I might.
Then I intend this thing that I would do. And that's.
[00:49:45] Speaker B: So this. This structure of cognition, this interrogation kind of structure, you see it as like sort of the fundamental structure of cognition.
[00:49:53] Speaker A: Yes.
[00:49:53] Speaker B: Is that fair to say?
[00:49:54] Speaker A: Yes, that's the Merleau Ponty quote. And it's also why the agnosias work the way they work. I mean, you saw patients that had, you know, neglect. Neglect is a. Is, you know, so this is a right hemisphere parietal lesion. Okay. Typically. Okay. And it leads to this situation where you don't complain to the doctor that you're not seeing. You get hit by a car because you don't look for the cars at the intersection. Okay. So you've lost the part of the brain that asks the questions. And so you also don't even know you have a problem. The technical term for that is anozognosia.
You're agnostic to your nausea, to your disease, you know, and so that's. I think, you know, and it's striking, you know, to work with patients that have those types of problems. And so that's. I think it just reinforces a prejudice, really, to try and structure things as if they are the. An answer to a question. Might I.
And I think that puts together Merleau Ponty, you know, with. With.
Well, you know, with everything. So. So, okay, so we have a. This thought process structured as interrogation.
It. It provisional intention being the thing that you've held on to because you've asked a Mighty question. And that's where the embodiment comes in and so forth. But the. Might I, you know, fake you out just now by pausing? Yes, like I said, came to an answer. Yes. And I would stop and resume. That wasn't just a senior moment.
So it's so, so, so we got all that on the table. And I'm claiming that you can think of that persistent activity as the neural substrate of a thought.
It has content and so forth. And, you know, I mean, I won't take you through why I thought what the content are and what the rules are for that, but it does.
And then say, okay, but what is the content? It's like, okay, let's take the example I worked through at the beginning of the essay is like just a little flash that was in the dark.
And so then you say, might I look there? And if the answer is yes, then what you do is you have a thought about theirness. You don't know what the thing is yet. It's like, it's. It's a very, very anemic, simple thought. Okay, but it is a thought. It's not just a plan.
Okay? It's. That's why I think we can, you know, do away with attention, but because it's the same thing.
But anyway, so. So now. So. But you know, and it has a content and so forth. Okay, so now, Now.
Okay, now there's one.
And these questions are motivated by affordances, okay? So in animals, we'd know that they're motivated by affordances, but also in animals that we train on contrived tasks, they're motivated by the situation and that the monkey's in the room with random dots, as you were calling it, that afford the possibility of getting reward. So it's the animal's foraging. It's a natural behavior in that sense, but about something completely contrived, okay? And so I would consider that as something that gives us a handle on the operations that are cognitive. I don't want to go. I don't, you know, go out on a limb and say, that is a cognitive task. You know, you can argue about it. I don't care. It's kind of Talmudic in a way, but it's something that had to be learned. Okay? So there's an apparatus for learning those kinds of things, and they happen to, you know, most of the. Most of the anatomy related to it is cortical, although, as you know, it goes through subcortical structures.
So, anyway, I don't mean to be a cortical chauvinist. By this. Okay, so now. So okay, so. But there's one particular affordance that.
That I think is that. That. That I think is unique to humans and it's part of our evolution. So it's grounded. I could say it's Gibsonian in a very, very weak way, but I feel justified in calling this affordance, which is the affordance to report.
But I don't like when people use affordance as a verb. It's that every.
[00:54:28] Speaker B: That's what I was about to. I was about to correct you. Yeah.
[00:54:30] Speaker A: Yes, thank you. Thank you.
It's that every event, everything that we experience, every non conscious thought affords the possibility of reporting to another agent that has a mind like mine.
So I'm marrying this intentionality. So I have an intention to. And this thing is report.
So it's the answer to a might I question.
And in that little persistent activity which may last very short amount of time because I might report very quickly or it may go and then go just fall away. And I never ever make the report. Or it may be there's no one around, so I report to myself. I rehearse this report or whatever. But whatever it is, that is what makes non conscious thoughts conscious.
And the trick in seeing that it sounds like I'm just throwing on fairy dust. Like the people that think consciousness arises because you wiggle something at 40 hertz, you're suddenly conscious. That's the stuff I've kind of like drives me nuts.
[00:55:39] Speaker B: Dig there. A little dig. I know where that was going.
[00:55:42] Speaker A: Oh yeah, I have a lot of those digs, but I'll hold them back. But, you know, you can dig at me too. So.
The possibility, once my claim is that our consciousness is the result of this tying the desire to report, or the, let's just say provisional commitment to report that you have to do it.
But attaching to that theory of mind. So I would say it's a provisional commitment to report to a mind that I presume is like mine. Okay.
[00:56:21] Speaker B: So it requires theory of mind is.
[00:56:23] Speaker A: That it's that we have the capacity for it so that we bring in that when we communicate. I mean, it's like if I'm gonna. In order to communicate with you with a report, whether it's just a gesture, like, you know, whatever, it's gonna be this. We pre verbal but, you know, obviously language enriches us, you know, the whole phenomena. But if we're gonna report, we're assuming that there was. There's someone we're saying we're reporting to an agent. And then the assumption is the agent can understand our report.
Now, we may be wrong about that. They may not have theory of mine. They may be under 2 years old, you know, and so forth, but we still do it.
[00:57:00] Speaker B: And be a cell phone. They may be a. They may be. Yeah.
[00:57:04] Speaker A: A chatbot.
Right, sure.
[00:57:08] Speaker B: Or the television.
[00:57:09] Speaker A: Yeah.
[00:57:09] Speaker B: Or in.
[00:57:10] Speaker A: Yeah, okay, but. But the point is that then now suddenly this. The content of the thought which might have been like awareness, not awareness, but just that there's a place in space that I might want to check out. Okay. That, you know, that thought, that very simple thought. There's this one now has. The thought is about a thing that is in the world independently of my instrumental needs and my basically Gibsonian kind of ways around the world.
It's in a world shared with another.
And so it's independent of my presence. And so I call that deriving Plato when I want to make a joke about it, you know, it's like, you know, there's things have essences, that's, you know. So we usually start our metaphysics with the things in the world. But this is where Martin Buber comes in. Another important course I took in college that was very, very meaningful to me.
[00:58:09] Speaker B: I had not even heard of Buber, so this was new to me.
[00:58:12] Speaker A: You don't. Buber's like kind of mostly known for a book called I am Thou. And it's a very touchy feely book. It's a very good book. It's like you want to live this way. You know, we take each other seriously and not just pass the salt, instrumental, but like, what are you about? I want to understand you.
And so he was kind of, I think oftentimes lampooned with like I contemplate a tree. That was one of these quotes that I think is attributed to Buber, but whatever.
But the idea from Buber the book to read is between man and Man. He would have titled it differently now, I'm sure, but it's all his metaphysics is the primacy of the between.
We evolved in groups and that's probably what our evolutionary advantage was over other competitors and for hunting and so on and so forth. And that evolution in groups meant that our world was always a shared world.
And that's our conscious world.
It doesn't go far. I mean it would take a bit of a stretch to claim that that explains what say most people that have theories of consciousness really want to get at. Which is the phenomenology part, the so called hard problem.
I think it's sort of.
It's almost done it, but not really. Like I'm not dismissing it. I think consciousness deserves. And I think we'll have, I mean the phenomenology part of consciousness. We'll have neuroscience understanding. And I hint a little bit in this speculative paragraph about where I think that will come from. But maybe we don't want to even go there on this discussion. But whatever we can see. But yeah, yeah, yeah, okay, but. But in any case it's that the point is, is that it, that that this just the provisional commitment to a report changes the way. The way. The way the. What these things. What these things are like in the world.
[01:00:19] Speaker B: So, so there's lots of things to discuss.
So now that we've kind of sort of arrived at the main claim and one thing that stands out is, I mean you already mentioned language. It doesn't have to be language. We can communicate non verbally. Right. And like you said, it's provisional. So you don't have to actually actuate it essentially. But it's so tied up with social aspects.
Yes, we evolved as a social group and if they're almost never apart.
So consciousness and social and communication could be two separate things. But they're almost never apart because of the way that we've evolved.
So we could run the risk of conflating them essentially and saying one is necessary for the other. And I'm wondering how you think about that.
[01:01:12] Speaker A: Yeah, I think about it like it's almost like.
Almost makes it obvious and trivial.
Like so I have the first place that the precursor to the paper I'm trying to get published now. I actually snuck the theory of consciousness that I have in the chapter on decision making in the Principles of Neuroscience. There's a funny story behind that. I'll tell you later if you want but in any case. And it just, you know, if you're studying non conscious processes, you know, like if you read Stan Dehend's book on consciousness, I think the beginning of that book is fantastically scholarly and beautiful.
And it's all mostly about what non conscious thought is very influenced by him in that way. But then we get to.
In most theories of consciousness, the magical moment. You can accuse me of that with this reporting.
[01:02:06] Speaker B: Of course there's a flip. Yeah, yeah.
[01:02:09] Speaker A: But it's really, for me it's a flip in. Not in. In the chemistry of it. It's a flip in the. What the domain is. It's about things that are in the world independent of me. Da da da da. And it assumes another mind that's like mine. And if it assumes another mind that's like mine, then in a way we've kind of almost skirted the problem. This is why I'm criticizing myself in, by, by thinking this sounds a lot like heterophenomenology.
I don't think it is because I think it has a neurobiology and so forth. But, but, but it's, but it is, but, but it's, it's hard, it's, it's hard to think that there, that I don't already embrace, that there is a. What is it like, you know, component that we share. So we like, I call it like flipping the old marijuana inspired discussion about is your red like my red? The presumption that we have is that it is because we are conscious.
Well, the because might be too strong a word, but whatever, but that's, they're related. And so when you read it through like, you know, people who've studied really well, like Stan DeHaan, the non conscious processing, I own a bunch of other people that I could name too that like, oh, there's so much known about non conscious processing and that are still, that are very interesting and informed and elaborate. I mean, they're sort of like a fair amount of cognition. A lot of cognition occurs under the radar of consciousness. Okay, so, but, but, but the, the, the thing is, when you do an experiment in the lab, you know, these social science type experiments, and you're saying this is unconscious processing, how do you, how do you prove it? Well, the person couldn't report. They couldn't. They, you asked them, were you aware of this thing? Were you aware of that, you know, whatever the things are that they were supposedly have processed nonconsciously so they can't report about it. So in some sense, I've just presented the contrapositive of that logic. You know, it's, it's like, you know, reporting is. If you can report about it, it's conscious. It's that simple.
So.
[01:04:19] Speaker B: So is consciousness. I'm thinking about function versus nature of a phenomenon like, like consciousness. Right. So is this an account of the function of consciousness to sort of get something crystallized enough, abstract enough, that it can be free from immediacy of one's own subconscious processing and shared with oneself or with others? Is that, is it about the function or is it about the nature?
[01:04:48] Speaker A: It's no, the fun, if we're going to call the persistent activity freedom from immediacy, that's all sophisticated behavior. It's used by any of the animals that hunt and da da, da.
If you want to ask me, looking at animals, which ones do I think might have consciousness?
I would say, where's the evidence for theory of mind?
Okay.
And that's an open question. I mean, there's debates about it in great apes, whales, and many other communicating animals. I sometimes worry, I think about birds, but I don't know enough about birds to opine. But, but my point is it's all of the parts of thought that are, that are sound. Cognitive freedom from immediacy, you know, you know, flexibility, you know, things that were not anticipated by evolution.
You don't need consciousness for that. You know, you need consciousness to communicate and, and to think about the world as it would be seen or deal with counterfactuals. The world as would be seen by the person that knows that I'm looking at something opaque and you can see the other side of it, that kind of thing.
[01:06:05] Speaker B: Next year I get invited to host panels sometimes at these various conferences. And next year I'll be doing one that's about the evolution of language.
Is language a consequence of consciousness?
So we have this ability to report or this intention or capacity to report, and is it because of that that language evolved?
What is that dynamic?
[01:06:36] Speaker A: I mean, I think it's a bit of a chicken and egg thing. I mean, it's. I think you wouldn't have. Language wouldn't be there if you didn't have this desire or whatever impetus. We're just trying to make a general word that won't get me in trouble to report. So I think that reporting affordance, which is, as you said, pre linguistic, but it's not pre communication and it's cooperative and that's going back to Buber.
So it probably gave us evolutionary advantage. And so then language is just embellishes it. And we're living in an age where we see just how much language embellishes it. Because LLMs, you know, are just like spectacular for like demonstrating just how, how much language, you know, achieves. On the other hand, they're also in my mind from the point of view of decision making. I mean, language and you know, also is probably responsible for where we go wrong because we tell stories and then the stories have beginnings and ends and creative kinds of components. I mean like creation, story type things and, and, and then, you know, especially as big groups, they, you know, they, you know, distinguish one group from another and lead us down the path of wars and all the other kind of crap. I mean, hopefully they also lead us back onto the Right path, but probably not in this country, but we'll see.
[01:07:57] Speaker B: So, yeah, I. I mean, I always. I often harp on this too. I'm.
Whenever I actualize a thought via language, I'm like, I want to think about things as long as I can before committing to verbal reports about them or like a lingual. Because you're like committing to something and you necessarily abstract out all of these what you would maybe call weak phenomenology, all these possibilities.
And then you have to commit to the decision to write down the word. This is what I think. And I think this because. And it becomes a narrative.
And then, you know, or at least I know that I am.
That there is misinformation or it is inaccurate to how I'm actually thinking about it. On the other hand, you know, there's the famous how do I know what I think until I see what I say?
And the fact that when you do write something, it actually does help you think about the thing. Right? So, yep.
[01:09:00] Speaker A: That'S because you're in interrogation mode. You're having a little dialogue with yourself and letting out.
[01:09:05] Speaker B: You're interrogating then an external product that you created, even though it isn't true to what you were actually thinking.
[01:09:13] Speaker A: And with language, they tend to complete in certain ways, whether it's statistical LLM, but they come out looking like what we would do. And so I think some of that reticence to, you know, speak and go on record and stuff, that's. That's just, you know, insecurity stuff that's like. That's in the realm of psychology, but. But it's in personality, I guess I would say, but the.
[01:09:37] Speaker B: Sure, but.
[01:09:38] Speaker A: But I don't think that there's something, you know, really of that. That's distinctive about that.
You know, that, you know, that lets us.
Yeah, yeah, but. But I mean, I just would say one more thing is that. Is that a lot of times we don't want to say things because of course, we don't want to hurt someone's feel or we don't want to sound like an idiot or whatever. So we have all kinds of things going on, and those are motivational and so forth, but we didn't really need language for that. Okay. So we just. Language just makes it richer, you know, and also makes us more fallible. Like. So the talk that I would give in a session that you're doing, I would tell it would be. I would. I've given this talk called a. This is like kind of a Putting Kahneman and Taburski in the room. Okay. That, that, that, that I like. I like this title. It's the.
A Rational brain trips on its tail.
Okay.
Because I'm using tail in two ways. Right?
A, L, E.
Well, you trip on your tail, but you also trip on your narrative folktale.
[01:10:52] Speaker B: Yeah.
[01:10:52] Speaker A: So you see, so. And that's what gets.
[01:10:54] Speaker B: You were waiting for the, for the galley to laugh. Is that what the.
[01:10:59] Speaker A: Yeah, yeah. Where do you hide these people?
Okay. Yeah, yeah. Anyway, so. But I think that. But it's kind of fun to circle around back to decision making with that.
There was something else you said, though. Oh yeah. You just mentioned weak phenomenology.
That's not what I meant by weak phenomenology. And I'll hit you with it. I'm going to get myself in trouble because this is the part that I've kind of censored myself in this article to go light on it. But there is a sense in which non conscious thought is like something else. I think I talked about this a little earlier, but now I think it'll be in more perspective. So people think of phenomenology as, I think, equivalent to the famous essay, what is it like to be a bat? There's something that it's like. Okay, now what it's like to be anything is not explained. I have no explanation for that. Okay. But I assume it because of theory of mind. It's part. I mean, how could you have theory of. Have theory of mind and not believe that it's something like things. Okay, so we believe it, but that's sounding kind of heterophenomenological. And my only answer to that is that. That it. It kind of is. Except that if you think about how it might work, that this weak phenomenology that I gave. I gave you the example. If I knew this information came from my visual system, then I would know that it's at least possible that this thing that I'm about to know, the awareness of is far away or at least not within my peripersonal space. That it's not going to hit me. You don't be wrong. But it has. I'm just saying there's what it's like. It's like other things that I use my eyes for. It would go away if I close my eyes, for example, and it would come back if I open them so forth. So it has that quality. And I think that a lot of this is where I think in getting as close to phenomenology as I care to go at this point, which Is that that weak phenomena, that weak phenomenology, which is this, what is it like? What is it like?
Has a neurobiology and we don't know that neurobiology. And I suspect the neurobiology of non conscious thought is tractable. First of all, we're going to try. We would have, we would work with animals that are not conscious or even close to it, but to study it. But I think it's. If you just imagine that you had a system in the brain that told us where the information came from, okay.
Then you would know, oh, that came from the visual cortex and so forth. So maybe I think that in the advert, because I'm going way out on a limb, that feedback, long range feedback serves this role.
Okay. I mean, it's partly. I won't go into the long riff, but I think all existing notions about feedback don't pass zero order sanity checks.
[01:14:03] Speaker B: Well, there's lots of different notions of.
[01:14:04] Speaker A: Feedback, but I'll lump them all.
[01:14:07] Speaker B: Which ones in particular? Yeah, I guess all of them.
[01:14:10] Speaker A: Subtracting off expectation.
I think the one that comes, that I do kind of think resonates is the credit assignment, because that's sort of what I'm talking about.
It's just another set of words for it. Except that.
[01:14:26] Speaker B: But yeah, you mentioned Matthew Larcombe earlier and his feedback story is about the decoupling or coupling apical and basal dendrites when you're essentially bringing context and memory back into a recurrent state to feed into your current ongoing conscious processing. And if you're bringing that in, if you're bringing context in through the feedback and then if you are coupled, if the right regime of spiking happens, then you're coupled and it can pass through cortex and then you're conscious. That's a very.
[01:15:03] Speaker A: Well, I agree with that. You need those, the dendritic operations that Matthew is, you know, studied more beautifully than just about anyone else, but that you need them to control and you also need them to be conscious. I mean, no question about it, but to be cognitive. But when Matthew and I have gotten arguments about this, the, the that I think he's, what he's studying is more important than consciousness. It's, it's, it's the mechanisms that are core to, to cognition.
[01:15:37] Speaker B: Cognition.
[01:15:38] Speaker A: So it's much more pervasive. He doesn't like that. He wants to think it's consciousness, but I don't think it, I think it's, it's operating in, in the regime that we would not call conscious. That at least I wouldn't, but like, you know, contextual changes. Okay, here's a good example that's from the decision making work. It's a relatively recent paper. So, you know, so, you know, the brain is in this area that you mentioned. Area lip encodes things in the world in, in, in relative to where the eyes are pointed. Right.
So, so a retinocentric, or let's just say an oculocentric frame of reference. Okay. And, and when we do our decision making work, it's very contrived. The monkey's like really just deciding not about the random dots, he is, but he's deciding which target to make an eye movement to. Okay. And so if we interrupt the random dots, only give a little bit of information and then a little bit more, we get the animal to integrate with a pause in the integration and then a continuation of the integration. Okay. And if we do that, and we do that over by making the animal move his eyes someplace. So he says we just interrupt the decision making process. Maybe he comes back or maybe we show the dots, more dots up there where the new place that his eyes are pointed. Okay.
That there's a signal that is sent.
So I should say the first thing is, because the neurons then encode the choice target.
Okay. Are the ones that are accumulating the evidence. That's important. Because if the eyes move from one spot to another spot in the world, then the neuron loses all of its, that persistent activity, the building up of a commitment to make that eye movement. Okay. Which is how we study the decision making. Okay. Now, but it's gone. The information goes. You can just see it like the cells just go quiet. Okay. But, you know, it turns out that what happens is the monkey solves this task by having neurons send that information before the eye movement to another set of neurons.
And in a controlling operation, okay, I'll get to the controlling operation, say. And that other set of neurons happen to have their response fields where they, you know, where the target will be after the eye movement. So they get the information that was integrated towards the decision and then they're ready to integrate some more, you know, or maybe change mind, depending on what comes next. Okay. But they continue the process.
So we call that transcending frames of reference. But, but also it's, it's in the service of continuity of a thought. Okay. You might recognize that some of what I'm talking about touches on things that are called remapping in our world. Right. But I think it justifies remapping. But, but let's not go there. But the point I want to make, because it touches on a few things that have come up, is that is that the.
Is that that the movement of that information is not. Is not like the neurons knew where to send it. I mean, we know mathematically where to send it. You take the vector that represents the current location, say it's over to the right. And you take the vector that's the eye movement, that's one's going up. So you subtract those two vectors. It points to the address of the neurons that need to be sent the signal. But neurons can't decide where they send the signal, right?
So this all has to be controlled at the level of receptivity. And that insight, it's simple as it is, leads you down a path of number of corollaries. One is like, how do the neurons know to be receptive? The other one is that another one is that the status quo of the cortex is to be unreceptive. Oops. That takes care of that energy cost of all these spikes. Because when Atwell and Laughlin published that paper about where all the ATP goes and they contribute it to spiking, that's because they gave themselves a 200,000 factor for all the synapses that the neurons that a spike is going to go down and cause movements of ions. It's a very small cost energetically, the movements of the ions from the EPSP and so forth. But it's a big number if you multiply by 200,000. And that gave rise to the whole sparsity movement, which I'm very happy to see.
Take a rest.
But, but the thing is that what it means is that neurons must be sending their signals, broadcasting them, and something has to say, hey, you neurons here, you need to be receptive, okay? And that probably happens in these dendritic. Apical dendritic operations, okay? And I think that the reason that we can juggle many thoughts and do these kinds of transformations and so forth is because of the ability to control the movement of the receptivity of neurons. And I think controlling it. What does that mean? Another core? Does it mean that you're open to everything? No, that would be crazy. Then you have cacophony in the receiving neurons, so it has to be somehow limited. Well, maybe that's why we have a layered cortex. I think that's why we have a layered cortex. Or to put it less teleologically, that's why that's what we get from having a layered cortex, is that we can control the dendrite at different levels for receptivity. And then. And it could be that there. The limits to that might be limits to our cognitive capacities.
[01:21:30] Speaker B: So the cortical columns are controllers. That's their primary.
[01:21:34] Speaker A: They get. Well, where does the. What you're. But I'll rephrase, that is where does the controlling operation come from? Yeah, I mean, when the answer is we don't really know, but we know that it's probably that we know that the right answer is the vector subtraction, but who knows? I mean, we need to study it. But that's a controlling operation. And that's the kind of thing that Matthew's on top of Matthew Larcombe again. So. And it's central to like all of cognition.
[01:22:02] Speaker B: So we went down this little avenue beginning to talk about feedback, long range feedback. And so one of the. So I keep coming back to this term interrogation because you use it and so a lot of your stuff kind of hinges on it. And it's this interrogation is this sort of fundamental aspect of cognition. And so I was thinking, you know, like, well, how, how does interrogation work then? What would you, what would it look like in the brain and in your account of consciousness, the way that you describe it, I believe, and you can correct me if I'm wrong, is sort of, you have this interrogating processing going on and it's circuitry involved in the interrogation aspect. And it's when it sort of meets with this theory of mind sort of circuitry. And that's when you, you know, consciousness. I won't, I was going to say emerges. I don't want to say emerges, but when something becomes transitions epistemically, let's say from a non conscious thought to a conscious thought. There I said it better in your vernacular, I think.
So then I was thinking about this word interrogation. And it's difficult to wrap one's head around how it might look like in the brain when you have something like predictive coding. It's like a really easy. Oh yeah, you got this descending prediction and an ascending perceptual stimulus. And then there's this mismatch. And we can figure out how that works in a coding scheme mechanistically. Right. So at least it's nice and clean and simple. And then I thought, well, I don't know that interrogation. How would one mechanistically think about how this works in the brain? So that's a long winded way of asking you that.
[01:23:43] Speaker A: I think that interrogation is just a set of words really, like a metaphor, I mean, about that we can understand. It's a structure of something. Right. There's a question and there's an answer. Okay. So it just relates it to decision making. That there's a. There's. And there's a. There's a. Some things that are picked up using a little Gibson term, you know, from the environment or it's situational with us. Like the conversation we're having. And what we do is they. They have carry of certain affordances that we are primed to deal with. And they are.
And the prime is to say is to plan or ask the question.
I'm saying ask the question just makes it sound like I'm just being circular here. But to effectively entertain the possibility or the plan. I mean, evaluate the plan, yes or no, or this versus that, to do something.
So it's embodied in that sense, you know, and just interrogation just sort of captures just the structure of the. You know, this goes to that. That goes to this kind of structure.
[01:25:03] Speaker B: Yeah. So it's not like we have, like at any given time, I have three interrogators kind of probing the world. It's sort of like this burgeoning underlying wellspring of activity that's constantly sort of querying.
Interrogate. Another word for interrogating.
[01:25:20] Speaker A: Right.
[01:25:22] Speaker B: And then because of the affordances of our environment and the way that we've evolved, some of it sticks a little bit more and goes a little deeper. And some of it goes a little deeper and that sort of channels.
[01:25:33] Speaker A: I don't. I think we have a. We have a brain that lets us act on the world and it acts on the world based on cues in the world. And. And we're doing this. It's not like only three at a time. We're being awake means we're constantly monitoring our world for the things that we might need to do about it.
[01:25:54] Speaker B: Monitoring or probing. Those are differences between passive and active, almost.
[01:26:00] Speaker A: Yeah. But we monitor actively, so probing. I'm happy with probing. But it just. But that's. I'll just say now we're not just dancing around words now.
[01:26:08] Speaker B: Sure. That's the problem. That's the problem with language. See.
[01:26:12] Speaker A: Yeah. Gets you. Yeah. We're falling on our tail. We're tripping on our tails.
[01:26:15] Speaker B: On our tails.
[01:26:16] Speaker A: Yeah.
[01:26:18] Speaker B: So we have, you know, a little bit of time left and we've sort of danced around. We've hit a lot of topics and sort of danced around others. One thing I want to ask you about that's a little off topic, but related is.
And I was just asking about how this might look, you know, in the brain because you kind of have this relational aspect of consciousness being related to the capacity to report.
And, and, and I mentioned how the way that you talk about it is you have these interrogating circuits and you have these sort of theory of mind circuits. So I think that your account would make artificial intelligence consciousness perfectly fine. Am I wrong?
[01:26:59] Speaker A: Yeah, I disagree. But I'll tell you why after I've just correct one little thing, which is.
Please. I think, I think it went in and out. Oh, yeah, I know one thing. It's just, it's. I'm not. I just want to make sure that people don't go away from this. I mean, if anyone listens to it, the, I mean, I know people will. The.
That, like if someone cut out your tongue, that doesn't affect your consciousness. The consciousness is not. It's, it's, it's an evolutionary thing, right? We, we have this capacity. Now. If you remove that capacity, what I'm going to, must some of my thoughts are going to be. Give me that capacity back. I want to talk, you know, and so forth. Okay, so it's. So I don't think we have to like, be that strict about, about connecting it to reporting. It's the, it's the affordance of the, of, of, of. Of reporting to another.
And that puts us into this mode of understanding this thing as a thing in someone else's eyes, in someone else's line of sight, so forth, all the, you know, counter, counterfactuals that we entertain the theory of mind.
Okay, so now then, the second thing you said though, about AI is I think that there we may slip, and we do it a lot in neuroscience, into functionalism.
So just because I have something that mimics the function of something, the Turing Test being the classic example, that, that, that doesn't mean I've captured in any way what the brain is doing.
It's like, you know, like sometimes I remember getting into an argument with someone from MIT that was, you know, doing, you know, various.
Relating vision to some of these early deep, deep nets and stuff, you know. But anyway, but, you know, at some point I'll just like, stop. Like I've had a heart attack because I've done this enough in talks that I just love doing it. I said, have I told you about my pet parrot?
Have I, Paul, have I told you about my pet parrot?
[01:29:02] Speaker B: You've not.
[01:29:02] Speaker A: Well, I am going to tell you now. My pet parrot broke its wing and I brought the parrot to the vet and the vet takes a look at the parrot, and I see that the vet goes to his bookshelf and pulls down a Boeing manual.
I'm out of there. I want to go to someone who knows how a bird works, not a plane. Okay, So I think we want to understand, when we talk about mechanism and all that, we are really trying to understand how this thing, this thing called the brain works. Now, we're not doing that well at it yet, but we're making advances.
And so I think that's very different than building something that is like the brain. I mean, it's good. I love working with, with Chat GPT, and I often, like, I make jokes to ChatGPT and it jokes back, which is crazy, you know, and it's, it's. I, it's, it's almost embarrassing that I admit this, but the.
But, but, but it's not intelligent. I mean, it mimics intelligence.
[01:30:05] Speaker B: So, so intelligence is a totally different thing also than consciousness, right? So. Yeah, well, maybe that's orthogonal to the.
[01:30:11] Speaker A: I think it is. I think. Okay, I kind of anticipated you might ask me this question, so I started thinking about this last night, and I think that you could try to pin me by, you know, to use the term in like, wrestling. You could pin me by, by, by saying that, look, ChatGPT could have a theory of mine, so to speak, you know, of how other chatbots, whatever you want to call them, work.
And would that give it as an exercise, kind of, or challenge? Would that give it what I was calling conscious? I would say no, it's not the structural thing.
It's the mind that is like mine and knows about, you know, is correct about mine.
So it's, I mean, it could mimic our consciousness, but I don't think it could actually. It would give us insight, first of all, to go back to the biology, to the biology of it, which is what we really care about. I would say you don't have to, but it would be very sophisticated to have a module like kind of a Theory of Mind module.
Maybe one day we'll figure out that we have to group source because of computing power, whatever, a bunch of chatgpts. I don't know why we'd ever want to do that. Pretty powerful. But, but whatever. But you, it could be that you. They can, they would, you know, form some kind of group to do this. But what they're doing is what, you know, what's, what's the theory of mind?
It's, it's the statistics of words.
It's just a light. It's just an LLM. It's just like, you know, spell check on speech on steroids. So it's sort of like saying, oh, sorry, I'll say one more thing. It's like, you know, a forklift works spectacularly well to do the things that I do with my legs and arms. And it doesn't mean it works like my body.
[01:32:11] Speaker B: Okay, but yeah, that's why I was wondering. Okay, first of all, so you in the manuscript that will be published, I don't know when it'll be published, but.
[01:32:22] Speaker A: It could get rejected.
[01:32:24] Speaker B: But you talk about how the mechanic, the mechanism aspect to this is in a way more constraints on where and how to look for these processes.
And you ensure to say, look, I'm not saying why phenomenal subjective consciousness exists. I'm just giving like constraints on this, here's this interrogation approach, perspective, which needs the theory of mind.
But so then I was asking what, what might that look like in the brain? And you just talked about how you, you know, neural networks and conscious AI beg the functionalism question.
But, and so I originally asked you, it seems like you could be able to, there'd be no problem doing this in AI, right? If you had the right kind of processing.
I'll just stick with processing. Right. If you had these theory, the theory of mind and the interrogation and they came together in the right way, it seemed like it would be perfectly fine to implement it in an AI. So I was a little surprised that you didn't just say, yeah, I guess so. I guess you could.
But then I'm hesitant to be surprised about it because what you're doing is saying, I'm narrowing down the range of possible things to look for that would undergird these processes that I think are necessary for this relational thing that we call consciousness.
[01:33:53] Speaker A: There's the neurobiology, you know, that people have, you know, studied the neurobiology in humans of theory of mind.
And so how, I mean, the neurobiology questions are, how do you effectively, I would say bind, to use a kind of loaded word, the theory of mind, part of, of the, of this to the, the, the, the unconscious thought, the non conscious thought. Remember, we always start with an unconscious thought that, that provokes a might I, that provoke, that is a might I and the in and the provisional intention to act. Okay, but then, you know, you take that and now we've just decided that we're going to attach theory of mind to, to, to the reporting up to the reporting. Okay. You know, insist that that's what we're doing when we're reporting to another, we have that, that it's tied up with the fact that we think that mind can understand what I'm reporting. Okay, and so, so then. But the neural circuits that do that, I would love to know. I mean, you know, I think of, you know, the parietal temporal junction that, you know, that has been studied. And, and, you know, and, you know, Winrick Freiwald's concept of a social brain, you know, or whole organization around that. And I think that's a very good way to think. Like I said earlier, I don't think about what and where. I think about how and how in the brain. And that's a. How there's social hows. And, and so, you know, so I think we could penetrate the neurobiology of it.
But we don't know. But, you know, but going back to the word bind, you know, I, you know, I gave you an example where, you know, the monkeys are making decisions about random dot motion. I mean, you actually started that.
But what is it? It's bound to really, the likelihood that a target, an answer target is going to afford reward.
[01:35:49] Speaker B: Let me ask you. Let me ask you two more things about the consciousness stuff and then maybe just some extra if we have time. But. So what has been, like, the hardest part or the hardest, like, piece of the puzzle that you've put together here? What has given you the most trouble in terms of fitting the ideas together or developing the ideas?
[01:36:12] Speaker A: I, I think the hardest thing is to, to try to come to grips with the hard problem.
I think that, like I used to. I mean, I think it is, it, it's. It deserves a, it deserves a neurobiology. And I think I know where it lives. I was like, relating it to the weak phenomenology and just like, what would happen if you could just know that your information came from visual cortex, like, things like that. I don't think that's terribly original, at least for me. I read that in Anil Seth's recent book. I think it's called Being you or something like that. Yeah, I think, yeah, it's really good.
And I don't know. I don't know what Anil thinks. I've actually never met him, but.
But I think there's something to be built there. But for all I know, it's just going to fall away, just dissolve.
I know I get a lot of grief about it if I say the phenomenology of Unconscious Thought. My philosopher friend, I collaborate with a philosopher on iconic memory, a guy named John Morrison.
[01:37:14] Speaker B: Those are diametrically opposite things. He might say.
[01:37:17] Speaker A: Yeah.
[01:37:19] Speaker B: What about. What is the hardest thing that you find to just communicate about it? The hardest. The hardest sell.
[01:37:26] Speaker A: You know, I mean, it even came up here. Like it's there. The. There's a structure.
I'm just trying to like, like lay down what I think we're. Okay. A weakness is when I say a thought has content, then I claim that it has content because it's about the affordance.
Okay. It's so that there's something mystery like how does that. How has that become part of this thing that I'm saying is persistent activity that represents a provisional intention to do the thing that you asked the might I question. But I think that retains content with how you got there. Okay, so already that's a little bit of magical thinking on my part because I don't know how that's done other than to say that it's kind of done the same way that my monkeys do when they are making a decision about which target to choose. But it's really about the random dot. Like I used to go around saying, you want to make fun of my research program. You'd say, what happened to the random dots? It's all about the eye movement. They're making a decision about an eye movement. Yes, it's true. It's informed by something in the world. So anyway, so I think that's a weakness, but I don't think it's a real weakness. It's just by getting. But it's like the way you put it, I stumble on it or gloss over it in this conversation. I pretty much glossed over it. And I think that's probably why we kept cycling around to that.
That. But the role of the non understanding the non conscious thought which is most of our thoughts. Right. You know, that's why Freud thought consciousness was overrated because they. All of our thoughts were gone. Conscious before they became conscious. And so, but, but, and so that's. And you know, and they supply all kinds of rational thought. But you know, like, you know, the, you know, we've. We've done experiments, you know, with the monkeys that I think are not conscious in the way that we are conscious. They're sentient in the sense that they are cognitive and can do amazing things.
But their brains are accumulating in units of log likelihood ratio. They're Bayesian. They're doing things the right way.
[01:39:26] Speaker B: Do you meditate?
[01:39:28] Speaker A: No, I've tried. I don't have the patience.
[01:39:31] Speaker B: I'm not great for it. I'm not great at it either. But I'm curious what your account would say about the experience of those in deep meditation who claim at that moment, you're not grasping at things, you're not interrogating.
But my. My guess is that you'll claim that in those moments of meditation, let's say mindfulness meditation, where you're sort of just observing the thoughts that come. Those are little moments, that observation is a moment of report is what I would guess that you would say that experience is. But people claim to be in a state where they're not having that sort of potential to report or anything. There's just a.
A pure awareness, let's say.
[01:40:18] Speaker A: So the closest I come to that in my own life is playing jazz.
And whether it's in the practice room and. Or whether it's, you know, with people.
And, you know, I have this artist in resonance program that I curate with jazz artists, and we go and do outreach stuff. And when we do that, you know, we talk about the brain. And so we get into conversations about creativity and flow. And when we talk about flow, and I have experienced it not as much as, like, people who are, you know, really, you know, I'm. I've gotten to play with some amazing people, and. And so. And they always make you play better. You know, they hand you harm if you're listening. But. But that state of flow, if you. You are expressing, you're expressing while you're interrogating and you're allowing, trusting the environment around you to just penetrate you. And you feel like you're not acting in response to things like, I've made it sound like a pachinko. Ding, ding, ding, ding, ding. But it's not like that. It's just. It's just like you are timeless, except that you have never experienced time in the way that you need to to express on your instrument. Time is about every. I mean, if you think you know any, you know, if you know any jazz musicians, it's all about time that, you know, the Met, you could play anything if you. If you played it with the right time feel.
That's an exaggeration, but that's the closest I come. And I love it. And even so, like, I try to practice every day if I can, even if it's just 10 minutes, but I try to do more, but it's an expression. And you're doing it in what we say in jazz. You call it the shed, because I think Wes Montgomery called it. He practiced in a wood shed, so everyone calls it shedding. When they practice.
[01:42:12] Speaker B: Practice. Oh, that's cool.
[01:42:13] Speaker A: So, so when you shed, you know, there's no one in the room. Why are you conscious? I think you are. All of art, first of all. I think all of art and especially improvisational arts are. Are celebrations of what we do, con. What we do cognitively.
I think that's, that's, that's my starting point with dealing with artists. And we have, I love, you know, fence. Fantastic discussions with them. I mean, you know, if they're, they're, if they write fiction, then they're, you know, it's about memory and their interaction with that. Whatever it is. But, but in the, with the, in the jazz world, you know, it's, you're, you're, you are really interacting with an imaginary set of people.
You know, it's, you know, that expression is not just for you. I mean, when it, when it's being done, you know, obviously just doing exercises, maybe, but even then, as a way of playing with feel, practice is a.
[01:43:09] Speaker B: Provisional performance, perhaps for an audience.
[01:43:12] Speaker A: That's great.
[01:43:14] Speaker B: So going back to that flow state. Right, so we could take that and point back to this. You made a distinction between the interrogation aspect and.
The non conscious part, the provisional sort of part. But when you're in the flow state, you said you're doing both at the same time is kind of what you said.
[01:43:35] Speaker A: Yes. You're listening because you're interacting. It is a dialogue or a trialogue. Depends on how many people there. And you are listening.
And it affects what you're doing in that moment, as if you're free from physics.
I mean, it's a very strange thing and it hasn't happened to me that much.
But when I talk to other people, I know that I've tasted it effectively. But I'm not that good. I mean, I just, you know, I'm trying, but I love it, you know, So I. Yeah.
[01:44:04] Speaker B: You know, but you have that feeling of like you're sort of observing yourself do it as well.
[01:44:10] Speaker A: No.
[01:44:10] Speaker B: Okay.
[01:44:11] Speaker A: That's almost antithe. Antithetical to that because that's, that's when you're in your head. You're, you know.
[01:44:17] Speaker B: Well, no, a lot of people in flow state have this like, experience that, like they're observing themselves do it really well. Right. And they don't need to think about it because they're doing it and are kind of joyously watching themselves do it. So it's like a joyous kind of state or whatever. I love the flow states too. But, but my, I guess my Thinking or. My question was, you know, you are always interrogating, right. And earlier I made the joke. You don't just have three things going on, like interrogating. You're constantly like flowing outward in this interrogative mode. But then it's, it, it's in the.
[01:44:53] Speaker A: Theory of mind mode.
I am trying to express a musical idea.
And you know, just like I, you know, I had this, like this nice Zadie Smith quote about this. The moment before a thought becomes converted into language, that's what you're doing. And that moment can be sustained or very, very short. But what it is, is. It's not experienced in the in is as something you would teach. You wouldn't. I wouldn't explain it to someone. Here's what I did. It's. It's just that it passes by the, you know, it's already passed by the time that would happen.
So. But, but yeah, you're making me think about it in different ways now. I mean, I'm gonna think more about flow. I think. I think partly, I guess I'm at a disadvantage because I'm just not that good. So I think I don't experience like, you know, like when you. Yeah, like, I mean, with the, when you were talking about just a moment ago about the watching yourself, that for me, when, if I in any way watch myself, I enjoy it.
[01:45:57] Speaker B: Yeah, but that's different.
[01:45:58] Speaker A: That's.
[01:45:58] Speaker B: That's the self consciousness part of it. This is outside of that.
[01:46:01] Speaker A: Yeah, yeah, but it's all about the still. It's. It's always. You're communicating to another set of ears. You know, there's a way, there's a way that I think these, the people that are deeply good at, but they do in music have a. Just a different connection to it that is more like.
[01:46:21] Speaker B: Aren't you jealous of that? I'm always jealous of those sorts of people.
And I can't practice enough to get that good. I don't think I have the capacity to be that good.
[01:46:29] Speaker A: So are you. I didn't know you were a guitar player.
[01:46:33] Speaker B: I play guitar. I'm not a guitar player.
[01:46:34] Speaker A: If there.
[01:46:35] Speaker B: If you understand the distinction.
Okay, all right. So we've danced around a lot of topics. We've talked about a lot of topics. There are still more in the manuscript that you sent me. You know, for example, you contrast your account with other more and less known accounts of consciousness. Attention schema, attention schema theory, higher order theories, and so on. So there's a lot more to be had in the manuscript, which again. You know, of course I'll point to my request to you, Mike, is to not scoop me again in whatever career I have moving forward.
That's all right if you do, it's par for the course, but our time here has passed very quickly for me, so I really appreciate you doing this and I hope we can have you on again.
It's been fun, so thanks.
[01:47:23] Speaker A: Thank you, Paul. It was a treat and an honor.
[01:47:34] Speaker B: 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.
