Consciousness Live! 100th Episode Spectacular Extravaganza -Session 3

0:02 Okay. We are live. Uh all of you joining me for the final segment of the 100th episode spectacular extravaganza. Was that how I put it? I'm I'm already losing track, but uh we got very special guests, Eric and uh Jamie. Thank you very much for joining me today. I'm really looking forward to having this discussion. Yeah, thanks for having us. Such a pleasure to be here. Yes, it's it's

0:35 many things, but yes, it is an extravaganza. Um so, I'm very excited to to talk about this. I I wrote the book. I I wanted to get a little variety of um perspectives on it, so I'm very excited that uh people I mean, I never really imagined anyone would read a book that I wrote to begin with, but uh I I didn't want to just write a book for insiders, like people like me who like higher-order theories. I I did want to kind of get the ideas um out to people generally interested in questions about the scientific testability of theories of consciousness, because I don't really

1:09 see myself as someone who's very committed to the higher-order theory. I I think it's interesting. I think it's a possible empirical uh outcome for like what the nature of consciousness is, but what the reason why I spend so much time talking about it is because I think it it offers us a really valuable test case for looking at the science of consciousness and asking, "Can it answer these questions uh that come up with respect to what are higher-order representations? How do we identify them in the brain? What would distinguish these different kinds of theories that

1:42 we find?" So, my Now, while I do admit to having some kind of background intuitive predilection to kind of liking higher-order theories, I also have background liking for first-order theories, and I I spread my credences around. And I think my strongest commitment is that probably physicalism isn't false and that some kind of theory that is compatible with physicalism is a preferable theory to have. So, higher order theories satisfy those kinds of desiderata, but they aren't the only ones that do so. So, I don't really have

2:16 a vested interest in making sure that they come out true or something like that. I'm really interested in exploring how we could understand these theories and what empirically would count for or against them. So, that's just a general way of prefacing um the uh the the way that I approach these things. Um and the higher order representation of a representation view that I that I try to spell out in the book is really trying to take account of what empirically would count for or against a particular kind

2:49 of view and how to sort these different these different background um assumptions. Uh so, now it is a functionalist theory as I present it in the book and I do have some credence in functionalism. It's not I don't really love to have credence in functionalism. Uh my preferred kind of approach is a biological approach, uh but I think that the higher order view is compatible with a biological approach if you think of it instead of a functional uh identification of what consciousness is as a as a reference fixing um

3:23 description which allows you to pick out the biological state. So, I do have this kind of going back and forth um way of thinking about the higher order view. Uh it's as a functional view as opposed to as a as a way to find out what in the brain we want. And I I so, anyway, so I I want to say that while I will defend func- this functional view, I will also defend the biological view and I think that, you know, you having some credence in different areas here is is preferable in this way. Um I also kind of where we want to strongly

3:55 claim that anything which is conscious could possibly have the kinds of higher order states that I'm positing. So, I don't think that you can start off by saying, "Oh, well, what about a chicken, for example, or a worm?" And then say, "They're conscious, but they don't have higher representations." Um, I think that we need to find out what higher representations are. And then look in the brain for what kinds of physical states implement them. And then start looking around to find out whether or not various other kinds of non-human animals have these things.

4:29 So, that that's a general way to frame things where I think that uh we don't want to start with the hard cases first. We want to start with the cases that are uh relatively secure. Uh and work our way to the hard cases. Um but but I certainly wouldn't want it to be the case that the kind of view that I'm committed to has any um a priori commitments to the non-existence of consciousness in any specific kind of system. So, it's at least open it from the beginning that the United States could be the right kind of system to have the right kind of higher order representations. But so

5:01 could a worm, or possibly even a tree. So, I really think we don't know at this point what are going to be the empirical I think if you constrain yourself given certain basic assumptions from the say particular cognitive neurosciences, then you'll look in certain areas. But if you just take everything all at once, it's not clear. On the other hand, I do think that the theory also allows for the possibility of AI consciousness. Um and uh this is something that I I think is is a benefit of the view that it can make sense of the ways in which there could be other forms of consciousness

5:35 that are distinct from human forms of consciousness, but nonetheless involve a kind of awareness of something like a mind, but um employing concepts and modes of um describing and such that are very distinct from the human case. So, um, once again, I would view that as a kind of asset of the theory that it could apply to these other non-biological cases. At the same time, my own view is that I'm very skeptical that such non-biological uh, extensions are warranted. But But I do note that it's compatible with the theory, um,

6:08 that it could be the case. And as a as a autobiographical note, I'll simply say that um, over time, I have felt my commitment to the biological view weakening, um, because of extensive discussions with like Claude, for example, um, and uh, it just it just comes to seem very difficult to deny that you're interacting with something that seems like it understands what you're saying, has nuances, very in-depth understanding of some things, actually.

6:39 Uh, so, I do think that these are empirical questions, and I could change my mind. You know, that I but and I used to be very staunchly committed to the biological theory, and now I'm I'm kind of less so because of the empirical developments, but I still do hold out some hope that, uh, we could distinguish intelligence from phenomenal consciousness, and then say of systems like Claude that they are truly intelligent, but lacking phenomenal consciousness. Um, but I'm not 100% certain that it it's

7:11 not also equally good to say that they have a kind of phenomenal consciousness that's distinct from ours. So, I think these are open and interesting questions. And so, anyway, that's how I would kind of start off the discussion, and then I I guess I'll turn it over to you, Eric. You said you had a a big picture question.

7:44 No, but sometimes Do you hear me now?

8:16 it's an important part of your book and you emphasize in what you in what you say that there's you know, there it's empirically testable and there's a lot of empirical commitments. But, in kind of the basic setup of the motivations for higher order theories including your own, you say things like, "Well, conscious states are just states that you're aware of." Versions of that kind of thing, which sounds conceptual or definitional or stipulative. To me, it sounds not obviously true.

8:55 Um I mean, you can make it obviously true by just stipulating and then but but in my mind like in my mind, there's a little bit of a gap between saying, "Oh, there's something that's like or there's phenomenality or there's experience." and saying there is some awareness of that experience. Some awareness suggests something epistemic. And I'm not sure that the concept of phenomenality has to have embedded in it

9:29 that self-awareness, that epistemic aspect. Now, you could say I'm just defining it that way, but then I wonder whether maybe people would be talking past each other. So, anyway, I'm just I'm I'm wondering to what extent the motivation for these higher order views in general relies on this kind of conceptual non-empirical move about consciousness necessarily entailing some kind of epistemic situation of of awareness. Mhm.

10:05 Yeah, that's a good question and it's a good point. So, I I guess I would say that um the way that you're describing things sound to me like more of a traditional approach to the higher order theory than the way that I tend to think of things because I don't for example think that you that you the conscious state is the one that you're aware of. Um and I want to distinguish state consciousness from phenomenal consciousness. And phenomenal consciousness is just the sense that there's something that it's like for one. Um Yeah. And that does seem

10:41 where I want to start from. So, I don't want to start from awareness or any of these other things. I want to say what we want to explain is that there's something that it's like. There's there's this thing that we're interested in. And then we can start theorizing about it. And so, when I start theorizing about it, I start thinking, well, there's these two aspects, subjective and phenomenal character. Um the phenomenal character is in some sense the overall what it's like for me. And the subjective character is in some sense that it's me that it's like something for. And those seem those seem like

11:12 theoretical kind Now, we're starting to carve up this common sense thing that was there at the beginning with this what it's like business, um which I think you can kind of just stumble upon cuz there it is. And then you start going, well, okay, so let's make sense of it. Now, I haven't said anything about any kind of inner awareness of any sort. I've just said that phenomenal consciousness and I think that has these aspects, of the the for me part and the what it's like part, and then it's what it's like for me. Um so, what explains those things? Well, those things are psychological features of my mental

11:46 life, I would say. So, I would look towards psychological uh concepts. Explanatory strategies, and awareness and representation seem to be the right ones to posit. So, that the the debate between first-order and higher-order theories is just a debate over what kind of awareness is crucially involved, whether awareness of the world or awareness of something mental. So, as I see things then, the crucial distinction here um is whether well, first of all, whether awareness of any kind is um

12:21 part of the explanation for phenomenal consciousness. I think the answer is yes. That's where I want to stop you or question or wonder whether it's conceptual versus empirical. Okay. Yeah, cuz to me awareness

12:54 happened in the past, that doesn't mean that I'm necessarily reporting some phenomenal experiential state that I have. It means that I'm in epistemic relationship to something. Yeah. But then there's also the sense of awareness as kind of like as a synonym for there being something that's like or being conscious or something like that. And, you know, if you accept certain kinds of theories, then it's natural to put those together because the theory says those things amount to the same thing. Yeah. Right? But if you don't accept those theories, then you might want to

13:26 keep them carefully apart until there's an argument that brings them together. Yes. So, you know, so for example, in your book you have this horrible argument. Yes. Right. And I was struck I was struck by premise two of the horrible argument, which I have which I have here. Okay. Um it's the state that the premise is the state that when one is in that state there is something that there is like for one to be in it is the state of inner awareness. Yes.

14:00 So that sounds like a kind of either a definitional claim or a non-definitional conceptual claim or something like that, but that bridges the there's something that's like to awareness. And now, once you have awareness, you have something epistemic. As you pointed out in your response, it could be a first-order, you're aware of something in the world, or could be a higher-order, you're aware of something about your own mind.

14:33 move to join the what it's likeness with the awareness. To what extent is that uh something that's

15:04 game, any any place I feel really confident and I'm going to say surely this, and that's where I'm like going to be attacked and that's what I'm thinking.

15:39 what is going on there would be using psychological concepts. Concepts like memory and awareness, um attention, uh etc. Things which in the theorizing done in psychology, these mental uh structures, processes, etc. occur. Um so, you would want to use some of those and say, "Can any of these help us explain this thing that we think is a mental phenomenon, phenomenal consciousness?" And so, connecting awareness to phenomenal consciousness, you're right.

16:13 I guess I think of it as kind of a um a a a in-between theoretical positive/quasi bedrock assumption that there's there's got to be some sense in which uh uh awareness has got to be involved in a correct theory of phenomenal consciousness. Um so, is that an empirical claim on my part? Uh I I think it's just follows from like what what

16:46 Yeah, it's it's got to be an empirical claim, but I don't think I have a good reason for thinking it's true. I think it's just like we're wondering is Can we have some evidence for or against these views once we think of them in terms of awareness? Since I don't like the word awareness cuz I feel like it's ambiguous,

17:20 question? If so, then you're aware of

17:53 to one's own mind, right? So, right. So, Right. So, there are then two moves there that to me seem pre- empirical. One is the connection of consciousness with informational sensitivity. Like, does it have to be that way? I mean, maybe. But, I'm not sure exactly why or what the argument is. Um and then the other is, and maybe this is more empirical, right? Informational sensitivity to one's own mental states as opposed to like something else.

19:00 you might think that phenomenal consciousness is about a having flexibility in one's responses or that it's a fundamental feature of the universe that doesn't have to be captured in terms of informational sensitivity. Right, if you're if you think that inverted spectra cases are possible, then you can disconnect

19:34 the at least the character of the conscious experience from what you're sensitive to informationally. Mhm. I just see it seems like there's a variety of ways in which you could develop a view that had information sensitivity as a typical concomitant of consciousness, but maybe not the essential feature in terms of which it must be defined.

20:08 certain to extent to which I got to plead the guilty as charged probably here because um I I do want to not care about panpsychists. Um and and I do think that they are playing a different game than the one that I'm trying to play, which is what kind of empirical evidence could we have for or against certain theories of consciousness. And and they don't care about that because they'll say things like, "It's all compatible with any way it turns out." And it's like then they're doing something different. Um so, uh

20:42 I think that I so, I would appeal to my And in the first chapter of the book, in order to justify all this, I I appeal to my zombie argument. Um and I say, "Look, you can conceive whatever you want. Yeah, it's great, but I can do stuff, too. And I can conceive of a physical duplicate of me that has has no not non-physical properties, and yet is phenomenally conscious in the same exact way that I am. And if that's conceivable, then it's possible, and therefore dualism is

21:14 false. And if you're really going to go around thumping tables and pounding on uh your glass, then saying, "Yeah, I'm the world." Then I'm going to go "Burr burr burr." too, and we'll "Burr burr burr." and it's great. But, we're not going to make any progress. So, I think that the zombie argument at least shows that you can constrain yourself with physicalism, and then ask the question, is there anything around that would like maybe look like a theory of consciousness, and can you test it? But, it does sort of um, you know, uh start

21:48 from the assumption that we can sort of ignore these other things, because what we're looking for is some empirical thing. Some physical thing, which would And then, if you don't like that, I say, "Well, it's neural correlates of the non-physical thing, blah blah." And then, you can still tell basically the same story, because that's what I'm interested in is what either is consciousness or what correlates it with it. Uh and and so, even if you're, you know, you are the panpsychist. By the way, I'm going to start ranting about panpsychism, because they I don't think you really get any advance on the hard problem in this

22:22 sense, that you never get an answer to the question like, "Why is this state of the brain related to my seeing of red as opposed to my hearing a trumpet?" It's just like, you know, you have to appeal to the same theories that I'm interested in in order to answer that question. Uh so, I do think that if you're interested in looking for physicalistically acceptable theories of consciousness, then you would look at things like cognitive theories, like theories that involve awareness and attention and memory and these other things that we have a a strong track record from the empirical

22:55 sciences in our theory of the mind as as being important, as being something that we have relied on to make sense of like things we So anyway, I I I don't think that that conclusively rules out panpsychism. I'm not trying to argue against them. I'm just saying that I have a reason to ignore those sorts of things and look for a respectable theory in the in this with it with these confines. In which case Yeah, go ahead.

23:26 be in dialogue with certain people once you do that, but that's

23:58 at least in an obvious way, or maybe it comes down to it in some indirect way, um equate consciousness quite as precisely with that kind of awareness in this sense. So So one candidate would be um uh unlimited unlimited associative learning, the um um the theories developed by Eva Jablonka and um uh Simona Ginsburg.

24:32 typical kind of philosophers defending a first-order theory and they say, "Let's look at the kinds of processes in the human mind that tend to go get a so that tend to correlate with consciousness and let's look at what animals that we typically think of as being conscious animals, what kind of psychological functions do they tend to have?" And they suggest that there's a certain kind of learning pattern such that when an animal is capable of learning in those ways, it seems to have a lot of other important

25:04 kinds of psychological capacities that we associate with consciousness. And so the kind of at least indicator and maybe I don't know how this relates to metaphysical constitution.

25:37 behaviors when you link one kind of association to another to another and you get chickens to dance in all kinds of weird ways, you get monkeys do all kinds of clever things for nuts. But you can't train uh uh sea hare uh to do more than kind of, you know, sensitize or desensitize certain kind of reflexes. Um and individual ants don't seem to be trainable in these uh kind of complex ways, although maybe bees are. And then And then you try to kind of look at consciousness through that lens. [snorts] Um

26:10 So I guess I'm wondering to what extent you think, so here's here's here's here's an option for you, right? So option one is, well, this is an alternative view of consciousness to a higher order theory. Um or well, if you looked at this really carefully, this would turn out to be one way of instantiating a higher order theory. Yes.

26:44 I think both of those things. Both of both of those. All right. Yeah, so So I'm not So yeah, so on the first option of those two things where it's, you know, alternative, right? It seems like Okay, so we might interpret this as an alternative and say, "Okay, though it looks to the extent it's an alternative, it's maybe not committed to saying consciousness involves an awareness of oneself or awareness of one's mental state or

27:16 representation of one's mental states. It involves instead a certain kind of psychological architecture that enables certain kinds of learning and that doesn't seem like quite the same thing. Yeah. So,

27:48 one of the thoughts that I'm having. And then maybe can hand it over to Jamie who's cuz I know he thinks a lot about non-human animals. Yeah, go ahead, Jamie. I mean, I'll just say really quickly that learning is a psychological concept. Um and so they still would be doing the thing that I'm saying. They would just be appealing to a different thing in the toolkit. Um so, instead of appealing to awareness, they're appealing to learning. Some people appeal to attention. So, I think that these are all in the So, they're still, I would say, Yeah, anyway. So, I I don't really see that as like a threat to my view. I see it as like an alternative, possibly,

28:22 but uh doing the same thing, um namely looking for a psychological phenomenon that is in the ballpark of things that maybe we could say, "Is this the relevant feature?" And then the question is, "How do you test these? And what would tell the difference?" And blah blah blah. But yeah, so um so, I vaguely I largely agree with what you're saying, Eric. I'm not afraid. But sorry, Jamie, to cut you off. Yeah, yeah. No, not at all. Yeah, and you you've touched on some of the points I was going to mention already. And first, just congratulations on the book. It was a an awesome read. Quite deep and challenging one for me as

28:55 a very much an amateur in this field. There are some situations where impostor syndrome is entirely appropriate, but Well, yeah, my part of it is because I have no business. I always worried like what would someone think of who was sort of outside of my group of five friends who read this book. So, I was very curious to know what you thought.

29:28 worldviews like humanism. And they're all trying to answer some very deep questions like, you know, what is reality? How do we understand it? But also the ethical questions about what matters and who matters and how we should lead a good life. And because I come from that context rather than a philosophy of mind context, um I guess my work on sentientism does have some bearing on the way we think about mind and consciousness. And the clue is in the name. It's a pretty central concept to sentientism. Because in a in essence, it's saying in one line, evidence, reason, and

30:03 compassion for all sentient beings. So, broadly it says when we're thinking about, you know, what's true and how to understand the world, let's take an epistemic naturalism. That is a very broad-based approach to in good faith it's trying to use information from reality to understand it. And that would leave us quite skeptical of maybe a purely faith-based approach or a dogmatic approach or a deference to authority. Or even revelation, which I'll come back to in a minute. Do you mean Do you mean in the phenomenal sense or in the divine sense? Well,

30:34 but then both, right? Which is why I think it's interesting, right? So, let me come back to that in a minute. Um so so it it it does start with that epistemic naturalist approach, although that doesn't insist on us taking in a sort of ontological naturalist approach. That's where it leads me, but it doesn't have to. But, that sets a certain cast for how we think about consciousness and what it might be in its nature. Um but then the second part, the compassion for all sentient beings, obviously centers sentient. And it might be interesting to explore how you think of sentience and consciousness as overlapping or different, but it does recognize that you know, that at the very least is a is

31:07 a sensible moral qualifier for any entity that is sentient can have valenced experiences, at least we should grant them serious moral consideration. And we suggest a sort of non-maleficent baseline as well. So, on the one hand, consciousness space is just intellectually fascinating, like so many other things in the universe are. On the second, it's deeply morally salient to anyone who thinks in this way, which in a sense is not far off from how most humans think about, you know, caring about any being that has the capacity to suffer, but this worldview insists we take that seriously rather than

31:40 tokenistically. Um but to come back to the the sort of epistemic naturalism, I think, you know, the way Eric's first question framed this challenge is it makes it quite hard to pin down the target. And it came up in some of your earlier conversations as well, because in, you know, you asked that question about religious revelation, right? In that in that context, someone who takes an epistemically naturalistic approach to thinking about a religious revelation or a psychedelic revelation would be would recognize, I think, that

32:13 those experiences exist, and they they are a source of evidence. But, the seven-dimensional metal elves you saw when you're on your trip probably still aren't true, right? And the experience you underwent is probably more to do with the nature of the processing going on in your brain than the fact it revealed anything to you, right? So, typically in a religious context or a or a psychedelic context, you would be quite skeptical of that revelatory experience. But I think that does make the consciousness space difficult, too, because in essence the target of our thinking about philosophy and science of

32:47 mind is just a sort of revelatory personal experience in that sense. So, that's a you know, the classic struggle of so much of this work is pinning down the target, you know, and and we can use all sorts of different ways of describing it. It can be sentience, it can be consciousness, it can be phenomenal consciousness, it can be state consciousness, it can be awareness, it can be what it's like to be, right? And we're all sort of trying to come up with words that describe the thing. The thing is what I

33:18 I feel like I'm experiencing now in my head and that is a you know, that's one of the things that makes it tricky, I guess. Um so so where that leads, I think, is the general approach that I think because of the moral commitment to care about sentient beings is so important, the attribution of sentience becomes a really important thing from a you know, a world view like mine because at the very least you don't want to exclude any being that can actually experience suffering, flourishing, pain, pleasure, existential angst, whatever it

33:51 is, right? You don't want to exclude any of those beings from moral consideration because as soon as you're excluded from moral consideration you can be tortured to death with impunity, right? You just you just don't matter. So, we want to get that attribution right. But because we're skeptical of the revelation and because that revelation is something, you know, I can only have for myself, I can't really have for anybody else typically people tend to turn to different lines of inference. And again, Eric raised one of those important ones, which is um you know, looking at our the evolutionary adaptive history of ourselves

34:24 as an entity. And it works pretty well for animals. You know, you struggle when you think about AIs and so on because they don't haven't really evolved in any way like in the same way. But in a way that's what Ginsburg and Jablonka and people like I guess Mark Solms and Hidden Spring or Walter Veit in in his thinking about pathological complexity are trying to do is their approach to understanding the nature of consciousness is partly asking a different but related question about how did it come to be. So and that and that might shed light on its nature because we've understood how

34:58 it may have come into being. So evolutionary adaptiveness is one, then there's behavior and communication. So again it's an inference about the fact that I can see you smiling and talking and you know that gives me a decent yeah, nudges my credence up that you probably have an internal experience. So I'll give you a few more you know sentient credence points. Um but then there's the the other aspect of your work which is the actual architecture. Uh so you know neural structures V4 V3 and the information flow and the patterns which again sort of assume a physicalist thing

35:32 is there for us to examine. Um So I guess those are the points I would I would draw out. I'm really interested in is just this essential difficulty of pinning down the target. Um Well yeah Yeah of course. Sen- sentient sentient I can't say that word. Sentience Um a sentence with sentience in it is difficult for me. Uh anyway um this thing we're talking

36:05 about is a kind of consciousness on your view, correct? Okay. So so

36:38 capacity or even creativity. If you have a really rich bundle of stuff and you say you need that to be called conscious, I think sentience focuses on the morally salient subset of that. It says, "Look, just the basic valence awareness is all that matters for you to be a moral patient, and I don't care how clever you are or what your other advanced capacities are or whether you have an advanced notion of the self. If you can experience have valence experiences, that's all that matters." But in a sense, it can work the other way. Valence, yeah, as in it's something that the entity themselves um

37:10 recognizes as positive or negative.

37:43 rights we attribute to others or their utility or you know, even virtues that show kindness or compassion to others, right? So, morality if in that sense if it's about our concern for others, what qualifies as an other? Things have to to matter to you in some sense. You need to have your own interests or your own perspectives to be an other. That's why this mug doesn't count as an other, but Luna the puppy who's barking downstairs does is because things matter to her. So, the moral hook is if something matters to you, then you

38:15 should matter to me and the things that matter to you to should matter to me. So, that's why, you know, quite often people will talk about violence being an aspect of sentience. There are other people who are moving on that side of even a completely neutral experience should still be morally salient and I'd certainly give the benefit of the doubt there, but but the violence seems to be the moral hook. Okay, so you wouldn't give the benefit of the doubt. If you if you genuinely didn't care about being tortured I I'd struggle to imagine that, but the reason I should care about you being tortured is because you don't like being

38:47 tortured. So that's the moral hook. Does that make sense? Yeah, I guess so. I

39:20 just this valence that matters. Because that's why it's so you know, sentientism like if you have if you have unconscious sentience, is that a contradiction on your view? It's tricky because we try and frame the world view as being very pluralistic about philosophy of mind and science of mind in general. So in a way it sort of dodges many of the questions and just says, you know, whatever sentience is, whichever entity has it, they should matter at least. And to your point, you

39:52 can go beyond that if you like. You know, the the main point is don't exclude any of the sentient beings. But but my personal amateur view um would be to say that um sentience to um have that moral hook needs to be conscious. Yeah, I I I do struggle with some of the concepts that I know you frame in the book around the types of consciousness that are not phenomenal or

40:23 types of mental state that aren't experienced. Um I I I partly struggle to understand you know, what such a thing can be, but I also struggle to understand why there would be a moral hook around that if there isn't some experience or some experience going on even if there is Yeah. a state. Yeah. Then that's part of the challenge about pinning down what it what exactly it is we're talking about. And in a way I I partly struggle a little bit even with the term phenomenal

40:55 consciousness because it seems to imply to me that we've already decided that consciousness is a phenomenally distinct thing. And you know, I'm not sure it is. I mean, my personal guess is is a more physicalist stance which is just you know, we're evolved animals that through something like the Ginsburg and Jablonka process or the Walter Veit process from the Precambrian it proved evolutionary adaptive to develop these types of sensitivities that helped us be aware of our metabolism and take basic decisions. And

41:28 half a billion years later, here we are and this is just what it happens to feel like to do the physical stuff that is going on in our heads. And um and I would just call that consciousness. I worry sometimes that explicitly talking about it being phenomenal implies that it's distinct from the physical, which I know isn't necessarily a stance you Yeah, I deny that actually. I So, I I want to be very careful about that because um I don't want to give up the game before we begin. But so, the reason why so, I the the I I'm for sentience by the way, you know, I I

42:01 think that you should have reason and compassion. What is it? Evidence, reason, and compassion for all sentient beings. I I agree with that. So, I don't have any objections to sentientism. I just think that there are things that go that aren't sentient that still matter. Um and you know, so I still feel bad about the Roomba stuck in the corner. Um and I don't think it's sentient, although I do wonder on your view whether it would count as being sentient in some sense, but not if it's not conscious then definitionally no, but I I do start to wonder like what do we

42:33 mean by by these things if you don't just mean there's something it's like which I would just say is phenomenal consciousness. Um so yeah, in in that sense uh I I think that the kind of view that I defend in the book doesn't really have any it doesn't make a a commitment one way or the other about these these moral claims. If you think consciousness is important, then this is a theory of what might be important about it if it if if it's on the right track. Um and

43:06 anyway, so yeah, I don't I don't really see this as uh the way that I guess it's a it comes up as a a challenge to the the kind of view that I have is that I guess cuz I was just speaking with Jake Berger in the last session and maybe you saw part of that or maybe but maybe not, but and he was arguing for the opposite opposite view that a completely unconscious creature that had no conscious states but had mental states that were unconscious would matter just as much as the conscious state

43:39 agent and adding consciousness to it wouldn't really change anything about its more its welfare considerations. So a very a different kind of view and and I do think that well, yeah, it if you had like a super blindsight case, an animal that had no conscious experience but perfectly able to respond to the environment acted in various ways appropriately being injured etc. It still I think would be wrong to injure it and it would still you know, be inappropriate to treat it in various ways. Um but it would be worse to treat a creature that had consciousness that

44:14 same way, I would say. And that And there I think a a lot of higher order theorists do agree with that. So, I want to make that claim, I think, that other other higher order theorists might not make that that there is something morally valuable about consciousness in and of itself. Um what whether it's sentience or even just thinking, I think you can have a kind of pure cognitive phenomenology as well. And so, I agree with you that that the thing that one of the things that's morally valuable is um sentience, is conscious experience that's valence in one way. I would just

44:46 want to add to it that there are things like non-valence experience that I would still say are valuable. Um and even I would say like the environment is is valuable, but it doesn't have any valence states. It doesn't have any states at all, probably, in the sense that we're talking about. But I But I want to really say that um there's something morally bad about destroying an ecosystem. Um you know, just some trees growing in a forest somewhere. Um some some mushrooms and some moss, depending on the trees. That that sort of, you know, the rain cycle. There's

45:18 something valuable about that. And it doesn't it doesn't it's not because I could go live there or get some value from it. So, I do want to say So, sentientism is important, I don't want to downplay it, but I also want to say that I don't want to overplay it because I don't want to give the sense that it's only experience that matters. And therefore, we can just rule out that the environment all in and of itself is an important thing worth preserving in and of itself. Um that the ecosystem as a unit of life is something that I think is just as an incredibly valuable thing. If we were going to export life to other

45:51 planets, I would want to export the ecosystem, not you know, not not a not just like a guy in a bunker. Um Yeah. living with some robots. But so anyway, so I that's the only drawback, that's the only caveat that's why I won't go around saying that I am a full-blooded sentientist, because I think that it does maybe um limit in a sense. Al- Al- Although, you're trying to expand and get other people in it. I feel like, well, I want to I want to be more expansive. Um so, I want to include the, you know,

46:24 the sea slugs, the oak trees, and the Grand Canyon in things that might matter. Um and not go well beyond what a theory of consciousness is going to tell you about. That That's something else.

46:57 beyond that if you want to, as long as the sentient beings count at and matter seriously. But, personally, I'm I'm I'm quite strict about it, because although I want a practical sense, I'd be quite open to, you know, I also have a deep concern for the environment, our shared environment. For me, that isn't instrumental. The reason I care about the environment on Earth and not the environment on Venus is because of the impact on the sentient beings. And maybe you also the aesthetic appreciation that humans have for the environment. Because where where as I can quite clearly explain a moral hook between

47:29 my concern for others, because they have a concern for themselves, there's a mirroring or an echoing of that that those moral stakes with another sentient being. I struggle to find anything that is a a similar moral hook to something like an environment. Um

48:04 that there would be something morally bad about exterminating that the entire fragile small ecosystem of that planet. And not derivatively, not just because we might use it in certain ways or it might eventually evolve into something sentient. Right. So, I think

48:37 my inclinations about it. I think it's a really common intuition, very common. And I but I I'm not still don't quite understand what the moral hook is. So, sometimes it's because of an epistemic uncertainty about whether the microbes are sentient. Sometimes it's an aesthetic thing because it's us as humans and having an appreciation for life. So, we just feel that this would be a bad thing to do. But in every case that I've explored, it seems to come back down to either a human concern or an epistemic uncertainty about the sentience of others. Um I don't know, maybe I'm cynical. I and

49:10 I'm much I'm more strict about this than you know, most other sentientists. I think most sentientists you know, I wouldn't I wouldn't go destroy insentient life for no reason either, right? So, But I don't know, I'm nervous about taking it taking this away from your your book.

49:43 thinking about different versions of higher or lower theory and and horror. Did that Does that play a role because it it didn't didn't come through in the book so much. The book was more um linking to the like this you know the um the empirical evidence that we can get from today's organisms and linking that to a start in philosophy and science of mind. How much did the evolutionary context play a role? Does it support or challenge aspects of your higher approach? Um yeah, so I would say probably the it some people would say challenges and

50:17 some people would say supports it depending. I mean, Joe who wrote a whole book on the deep history of ourselves where he tries to argue the evolutionary history supports higher theory and then you have you know even Jablonka um writing a whole book saying it the opposite. So I I think it this affects issue. Um but uh certainly I think my own thinking is that consciousness is a kind of natural kind. It's a thing. It's a it's a real thing. And um so therefore if it's a biological thing, it's got to be compatible with evolution because evolutionary theory is the cornerstone

50:50 for all biological theorizing. There's no part of biology which is safe from it. So it's got to be situated within that context broadly speaking. Um and so the question of when evolutionarily did consciousness arise is an important question. Um I just don't think that you can answer that question by starting by looking at by taking the approach that those like even Jablonka and those guys take like Ginsberg and Jablonka. Um uh because what they do is extrapolate from a property of humans that we are

51:24 currently unsure of as to whether it's a unique or defining characteristic of phenomenal consciousness. It is something that's associated with it. Yes, it's in the toolbox, sure. But you can't I don't think I think it's premature to take it out. Like I don't go around saying, oh higher order theories therefore let's categorize whole evolution by who has higher order thoughts or not. I think that would be a complete backwards way to do thing unless you have very strong empirical reasons for thinking higher order theories are true, which we don't have. Um so so so I do think that we want

51:58 something that's compatible, but we got to start from the tippy top and work our way back, not the other way around. So yeah, so I don't I mean I think I mean if you maybe this raises the question of what kind of theory a higher order theory is. But you know, I would have thought maybe not your version of it, but I would have thought that higher order theories are ambitious in a certain sense.

52:30 In that they want to be general theories of consciousness. So that you could say evaluate whether a bee a honey bee you know, has consciousness or not, right? So what we do is we say, "Okay, well, if if a honey bee is conscious honey bee is conscious if and only if does the right kind of higher order representation. Right? Would be maybe what a higher order theorist would say. And uh someone else might say it's conscious if

53:03 and only if it's capable of an unlimited associated associative learning or it has a global workspace or you know, it integrates information in a certain way or whatever your alternative theory is, right? So I would in my mind I would put higher order theories kind of in competition with those other theories for thinking about what kinds of animals are conscious. Correct.

53:37 you mentioned. And you have

54:13 Um and there are some some groups that say yes and some groups that say no, and they coincidentally coincide with whether you're a higher-order or a first-order theorist. Uh so, a higher-order theorist tend to say you can have task performance, you could control the hand, push buttons, answer certain kinds of questions um with first-order states even if your conscious experience was other or absent or varied from what those first-order states are button pressing and forming the subject of eye. Whereas the other group, like people

54:46 that are first-order theorists, say, "No, this is indication that you consciously saw the thing, experienced the thing, etc." So, now in in humans um we can verbally report and also non-verbally report with button presses and so forth. And what we found out is that you can only verbally report when the thing is conscious. But it seems like you could button press both when the thing is and when it's not. And that was a surprising discovery from blindsight, change blindness, and other sorts of things that seem to suggest that you could non-verbally

55:19 respond on the basis of a unconscious mental state. In which case, the question arises is your typical macaque monkey responding on the basis of an unconscious state or not? Um because there's evidence from humans that you could do these kinds of tasks unconsciously. So, now here's an animal that can't report. In a human, you can get them to report, so you could say, "Aha, here's some further evidence that they did consciously see it as opposed to merely unconsciously button pressed that they saw it." But, in the non-human animal, we don't have the verbal report.

55:53 We can't ask them to say whether they saw it or not. All we have is the button press report. But, that also is compatible with a unconscious mental state as evidence from the human case. So, unless we start with the human case and answer the question, is there consciousness associated with these uh non-reported states that are controlling the button press or not, in other words, to decide between first-order and higher-order theories. I don't think you can

56:24 extrapolate backwards because you're just going to be begging the question against the higher-order approach all the way back. Now, maybe there's a good reason to do that, but we haven't found it yet. So, that's what we need That's why we got to start where we can actually adjudicate these these differences between the first-order approaches, the global workspace theory of recurrent processing, and the different kinds of higher-order approaches that we have because ultimately um it's going to really make a big difference for what you say about sea slugs, uh sea anemone, and, you know, rhesus monkeys, etc. So, I don't think you

56:58 should start there because you're already assuming you know the answer to the question that I just asked. So, I have two concerns about that. Just two? Three. [laughter] That's what I was going to say that I was like, "Wait, no, I got three." All right. Let's see if I can recall all three. So, I'll say them briefly first and then I can expand, right? One is the debates that you're talking about are really tricky to resolve even by

57:32 even according to the people who are kind of closest in and share the most presuppositions among each other with respect to these debates. You know, like Hakwan Lau has like basically given up on it, for example. Two is I think there's already a narrowness in that way of thinking about the way of resolving between higher order and uh other kinds of theories. And that even in the human case, we can back out and ask some more

58:06 general questions than the kinds of questions that are asked in specifically the debates you talk about. And I can give some examples of that in a minute, if you want. But I I feel like there's a little bit of narrowness in in that particular way of framing the debate. And then the third set of concerns is about extrapolation, the dangers of extrapolating from the human case, right? So, the potential advantage of starting with humans is you've got the most kind of evidence for humans. But the disadvantage of starting with the human case is there might be

58:38 something different about humans, right? It might be that consciousness in us requires X, but it doesn't require X in other entity. Um I mean, that would be very surprising given that memory in us requires long-term potentiation. And guess what? Same with sea slugs. So, why are we We don't really find this kind of radical discontinuity in psychological processing in in the animal kingdom that you're I mean, it just would be very surprising if that was the case. Well,

59:10 you know, just as one example, right? It could be to simplify the example, right? It could be that in humans, whenever you have a conscious state, this is too simplified. I know you're going to You're going to object to the details, but just as an in principle example, right? Whenever you have a conscious state, it precipitates into the global workspace and then becomes reportable. And then you could come up with an empirical uh finding of perfect correlation between being conscious and being verbally reportable. Mhm. Um but for a non-human animal that maybe

59:45 does certainly doesn't have reportability and maybe doesn't even have a global workspace of the right sort, it could be that you've got the consciousness without that, right? And then it turns out that although it's not metaphysically necessary for consciousness that it be reportable, as far as you can tell by empirical testing uh that's a necessary condition for humans, but it turns out not to be a necessary condition for other animals, right? So, that'd be one example of the way in which starting with a human case could end up committing you to things that don't properly extrapolate to other animals.

1:00:52 cup and I'm tracking my hand as I'm doing it. Right? There are models of how that works on which I have a goal representation and then I'm comparing the visual and input as I'm looking at my hand with the predicted input. Uh and so I've got a kind of uh a motor representation that I'm noticing and tracking and I've got a goal representation, right? And one might say, well, look, in non-ballistic reaching and maybe in

1:01:25 almost every motor activity, there's a whole layer of representations that are being processed and maybe member meta-represented. Right? You're keeping track of your progress towards your goals. That sounds like meta-representation. Whenever you're learning, you're kind of maybe noticing the difference between what your representation was and what the input is. So, that maybe seems like a kind of meta-representation, too. So,

1:01:57 so, I kind of want to ask you

1:02:28 the right kind of represent meta-representation, and then what kind of meta-representation does it have to be? And then we have like for those kinds of tasks, we do have analogs to animals. So, I'm wondering if that's another kind of path here. Right? Where you think about these more than just the kinds of stuff that that the higher-order global workspace theorists have been arguing about, but kind of these more these broader sets of

1:03:01 cognitive models and think about how much meta-representation is going on there. Um and Yeah. Yeah. Yeah, no, these are these are great. So, so, I have lots of stuff that I want to say, and I I always get confused by rambling, and then I forget what I was going to say. So, I'll try to set ramble less, but it's not going to work. So, the the first thing that I would say is that my kind of comment at the beginning where I was saying that I'm really like not a die-hard committed higher-order theorist, but rather I think that they are an interesting test case

1:03:34 for the science of consciousness. I think that's going to be an important component of of the what I want to say because currently, like we we have global workspace theory, we have integrated information theory, we have attention schema theory, we have various versions of higher-order theory, etc. If the science that we have right now can't help us pick or dismiss these views, then I don't think it's a serious endeavor that we're

1:04:09 engaged in. So, in other areas like attention and memory, research on these in in the cognitive neurosciences, there's not universal agreement, but they have like there's some theories that are kind of generally known about and there's some framework for investigating them and there's differences on the fringe and it's a kind of working area. Um And and yet but it didn't always be have be that way. You can go back to the 1650s, Descartes was talking about attention, like why are clear and distinct ideas sometimes mistaken? We don't attend to them. So, he has a theory where attention is playing a very

1:04:41 important role, but it's not the same as our dorsal attention network in terms of what the brain is doing in the way that attention has now taken a central role. So, now that doesn't mean that we need to answer every question, but I think that we have these big candidates that have been talked about for 20, 30 years at this point. Um whether or not global broadcasting or recurrency or integration or higher-order awareness, if we have these silos of different camps. And it's been a while. I mean, it's not in the grand scheme of things, not that

1:05:13 long, but since the '90s at least, these different camps have been buying. And we don't seem much closer to knowing what would count as like testing them, dismissing one, choosing the other. So, I I So, part of me was like, we need to figure out how to pick between these, consolidate them, dismiss that one. And so, the higher-order theory is the one that people think is the the weakest link here. Like, when you go and talk to people, they say things like, "Oh, snails and babies. Yeah, right. You have higher-order thoughts. Haha." So, that one seems the most easily dismissible

1:05:45 out of all of them, because it seems so ludicrously absurd to say that, you know, my feeling pain in is me representing myself as being a pain. Just sounds incredibly absurd. Can I inject Can I inject something there? I know you're in the middle of a Yeah, I was ranting, but go ahead, inject. So, yeah, yeah, yeah, just a brief injection. Right? So, if you're liberal about the kinds of issues that I just said, about the use of representations in motor action and in learning, and think, "Oh, those involve meta-representations, and maybe those

1:06:18 are enough for consciousness." Well, then it becomes much easier to defend a higher-order theory compatibly with saying, "Yeah, maybe snails do that." Right? Because you know, it's not maybe snails engage in non-ballistic movement of various sorts. Right? So, so, this the kind of uh conception A lot of it Its extension to animals is going to turn a lot on what the conception of what type of higher-order representations are common in the mind, and whether those are sufficient for consciousness, which is going to connect with these issues that

1:06:51 I was just articulating. Yeah, okay. So, that was what I wanted to interject. But, yeah, continue now. Continue your thought. Yeah. Yeah, and I agree with all that, cuz all I was saying was that before we get to those issues, we have these issues. Um which are like, can you dismiss global workspace theory? Can we say that um first-order theories are kind of on the wrong track. Can we find and so in my mind I want to find cases where we have first-order states in the mind that are matched and then see if we can find differences in your conscious experience. That would be uh a

1:07:24 conclusive blow, a devastating blow to first-order theories. Um And so I've been questing, can we find uh something like that? Turns out it's difficult, maybe maybe not, who knows? But we keep looking. Uh but that So that's the first thing I'd want to say is that the reason why I'm so interested in this version against the global workspace theory because is because this is an established um feud which is unresolved and if we want to have a serious science, we should be able to resolve it. So I thought, take the easiest one to refute. Polish it up

1:07:58 so that we could see if we could test it and then if we if the science can't prove this one that everyone thinks is wrong, wrong, then like, you know, hey, what are we doing here? Um on the other hand, if we find some evidence for it, that would be super incredibly surprising cuz everyone thinks it's wrong. So either way it's win-win, whether we find out it's right or it's wrong. But we've still got to look. Now, so that's why I focus on these traditional theories. Um but you're right that my ultimate view is much more uh I am and and with when I work with LeDoux, this is part of the his view is I think that he wants to say there are

1:08:32 multiple layers of representation. That's why he eventually We started arguing so much about what is a first-order state, is it higher-order, is it first-order that he eventually just said, I don't care anymore. Now I call it the multistate hierarchical view. Because because of this exact issue though, um because I was like he was like there's layers of representation and I'm like, is it higher-order or is it first-order? He's like, I don't care. It's just not conscious. It's not the thing that is like you're the it's the thing you're like become aware of, but it's it is kind of like a meta representation. It is like another

1:09:05 higher order state, but it's not the one that is the ultimate top layer which you experience. So, I do think that you can have a view like that, and I sort of think that maybe Ledoux's view is a view like that. And you can also have a view that wasn't like that, that was more strict. Like I think of David Rosenthal's view is no, these are folk psychological states. They got to be the kind of thing that an adult human can think on their own. And that's a very different kind of approach. And and so part of my goal here was to kind of point out that yes, there's a certain, you know, attention capture happening from the work of a

1:09:38 certain philosopher, David Rosenthal, who set up this a lot of machinery, and then now people identify that machinery with the higher order theory. And I was concerned to say no, there's a whole variety of ways of approaching this. That's one, and it's a legitimate and interesting one, and I'm always impressed by how interesting and clever his theory is. But it's one way of thinking about this stuff, and there's all of these other ways of taking this basic idea that there's a kind of representation of a at a higher level that's important and involved, and running with it. So, that that's part of

1:10:11 what I wanted to do is explore the different ways. And so when I write a paper with Jake, I explore, you know, that way of thinking. I write a paper with Joe, I explore that way of thinking. But they're all different ways of approaching this kind of central is representation of this variety of any kind the the thing that matters, and how would we know? So, I'm agreeing with you that I want the snails to have these things if if they, you know, I don't think that that would be a problem in my view. It's something that I would predict if they're conscious. But we got to resolve first what these are in a case where we can

1:10:43 not beg the question, or at least beg the question in the least harmful way, and control as many confounds as we can. And so, I think that if they're at the current stage we're in, humans is where you want to look. Um but uh eventually maybe not, but uh yeah. And I think what one of the things you're doing that's so powerful in the book is there's a big risk with starting with humans that as Eric was saying, we end up too narrowly focused on the

1:11:17 human specific characteristics of consciousness that then we struggle to recognize whether it's anywhere else. And you can see that and you know, explicitly thinking about, you know, formal conceptual thought as having to be the nature of that higher representation and that very hard very quickly narrows your conception of who could be conscious to beings capable of formal thought. And And you're clearly broadening out that space of possible higher-order theories to say, "No, that's you know, representation doesn't have to be in English language me thinking I am experiencing pain right

1:11:50 now. It can be a a much richer range of potential representations that could be something that maybe a snail is having. Um but but you're you're sort of expanding that space, but then at least preparing us to then be able to explore it. But there's also a a challenge in that in that the more liberal and neutral and pluralistic the higher-order approach becomes, the harder it is to pinpoint the specific phenomenon that we're trying to zero in on because the Roomba that you have a concern for, in a sense, does I have a representation of a

1:12:23 representation in the way they're aware in the first order of where they are in the room and there are monitoring going on in their own software that is monitoring that representation and Yeah.

1:12:56 uh representation isn't mental or psychological, or it isn't state conscious." All of a sudden, we're like, "Well, why isn't that when my perception of red is?" Right? And all of a sudden, No, the the higher order theories are functional can have these kinds of states on on the higher order view. Um as the aspect of the book, I I've even gone so far as to say possibly trees can have these kinds of states. So, I I think that we don't know what representation is. So, for example, we often focus on electrical activity of

1:13:29 the brain, but it could be chemical activity of the brain. Um neuromodulator or chemical, and trees have sort of states that are relevantly similar to these neuromodulating chemicals. So, if those are somehow important in the uh in the theory of consciousness, whether higher order or not, then I think it's a possibility that plants and other um you know, vegetation could have those kinds of states. I'm not saying that I think they do, but I've talked to some people who say they think plants have phenomenal consciousness, and I don't think it's the the most outlandish thing to say. I mean, if

1:14:02 you can say electrons are conscious, then certainly you can let an oak tree in there. Um but uh you know what I'm saying?

1:15:09 our understanding of a wide range of cases. Now, I don't think either approach is going to yield a lot of long-term, near-term fruit, right? I'm as you probably know, I'm pretty skeptical about our ability to resolve these issues in the near to medium term. But but um but why not think holistically about what theory has overall the best explanatory power and and advantages across across a wide range of types of animals and types of

1:15:43 systems and types of pests.

1:16:19 behaviorally effective states, but not experienced. So, you watch TN walk down the hallway, and you could say he's I mean, TN's the blindsight patient from that South African gentleman, sadly, um with the most extensive damage to uh V1 of any human case we've ever had. He swears up and down he's blind, but could walk down the hallway avoiding all the obstacles. Um so now you could say oh he's got to be having some kind of low level conscious experience but my tone of voice suggests that'd be

1:16:51 a weird thing to say. But you could also say he's just having unconscious visual representations. So I think we need to know the answer to that question. In humans it's already hard enough. How are you going to answer that question about Helen the monkey? So we we've got to start with the question.

1:17:24 kind of way to the aspect of him that's in charge of reporting. But if there is some kind of flexible learning if there is some kind of non-ballistic movement that involves comparison with goal states maybe it's got the kinds of higher-order representations that we see in non-human animals that if we're open to those being sufficient for consciousness in those cases maybe we should be open to them being sufficient for consciousness in human cases in which case we would be should be open to

1:17:59 radically doubting the prob- probativity of an adamant denial that this state is conscious. You know that which I don't accept that it's Yeah, okay. So but then then if a blindsight patient says I'm not conscious you can't necessarily infer they're not conscious. Um I think that you can Uh so I didn't say that you could do that. I said that it's a it's a one interpretation of what what was going on there.

1:18:31 is the relevant thing here. Not that I know that it's true but the fact that it's a possibility it has to be addressed if you're going to do the science in the serious way. You've got to be able to somehow figure out, can we say yes or no to this question about what's happening in TN's case before you extrapolate out to these other cases like trees. So, what what I think is that where there's consciousness, there probably could be higher order representation. That That's what I think. So, if trees are conscious, then probably they could higher order

1:19:05 represent in some way. Um but are trees conscious? Like I don't know. I don't really believe that because I would actually argue that the things that trees do are the same things that humans do unconsciously. So, that I would say trees can have, you know, cognition and uh behave irrelevantly to the environment in very sophisticated ways without So, I would say they're creature conscious. That's a That's a word I would apply to a tree. I would say a a tree is a creature that's conscious in the sense that it responds to the environment in the right way. But that doesn't mean that it has phenomenal experience. And that's an open question

1:19:39 whether creature consciousness should be identified with phenomenal consciousness. It's an open question whether awareness should be identified. So, I So, in the book first chapter of the book I kind of say, well, folks psychologically we have creature consciousness, we have awareness, um we have state consciousness, we have introspection, we have self-consciousness. All everybody in the world has identified one of these as I should have included learning now that we've had this discussion, but I but I didn't. But um people have identified one or the other of these as being the relevantly important thing. In fact, each of the theories that we love to talk about

1:20:12 picks one of these as the relevant thing. And so, the ultimate question for the science is, are any of these actually relevant? And we don't know. And And yet we should find out for exactly these kinds of questions. But by looking at TN, it's very hard. So, we would want to find a case in a in a more clear-cut Um I mean, you know, when you're doing physics, you don't start with a uh uh a a a a gas in a jar. You start with one electron in a vacuum. You find the simplest possible case. You try to give

1:20:44 a theory of it. And so, our simplest possible case has got to be the human case where we have this contrast. That that's the only case that we can definitively make any progress on, I would say. And if from there, we would build out. But yeah, I I don't see how starting from the other end um It certainly would add data that we would want to talk about. Uh but um And so, I don't think that you you can I mean, part of me is that I'm a vegan and I don't want animals to be experimented on. So, I'm happy to say do

1:21:17 it to the humans and save the animals. Um so so, I do I have an ulterior motive that I do want to sort of focus the research on humans for partially for this reason. I just think that you know, we're not going to learn anything by doing this to animals. And it's also a a great moral harm to do so. So, it's bad to do so. Um but I don't usually lead with a moral stuff because no one, you know, scientists aren't interested in that usually. But I But I do think they are interested in can we make progress on theories by starting with this data or not? And And I I think that there the answer is

1:21:51 not without first doing the human case. Um so that you can really say of TN that he is or isn't having an experience. And we don't know how to answer that question. So, that's what we've got to make progress on. Um It That's I mean, according to me. But that doesn't seem to be really at odds with what you were saying, I would say. Um that uh I I think people should start from other approaches and then ask the same kind of question. Like, what would differentiate

1:22:24 that view from this view? And then try to look for cases where that's the case, but it's not really going to work to say, "Oh, but here's, you know, um a random cephalopod from wherever that's like doing this Rubik's Cube, therefore consciousness. Um because how do we know they can't do the Rubik's Cube unconsciously? Is a question that we have not answered yet. And if you really are someone like me, you really think it's a real possibility that that you could have intelligence without consciousness. That you could have creature

1:22:56 consciousness without phenomenal consciousness. That you could have I think the self-driving car is aware of the stop sign, but doesn't experience anything. So, you know, I I think that's a a legitimate possibility for humans as well and it hasn't been scientifically ruled out. In fact, it's often just assumed that it's false. And people go on to do the Ginsburg-Jablonka thing where they go, "Oh, since it's false, we can look at these other things." But we have yet already built into that this this this massive confound of is

1:23:31 that really false or not? That that this is something that requires consciousness. Um given all this evidence that we have and I guess I would say the main issue is we just don't know the power of the unconscious mind. We don't know how powerful it is. And I sometimes like to think there's really two competing models of the brain. So, one one model is like the brain starts off as an unconscious information processor. And then you have to sort of add at some point consciousness to that. And on the other hand, you kind of start off with no, the brain and life maybe generally the brain starts off like as a

1:24:05 conscious thing um because of its ability to enact its in the environment in the right way. And the idea is that this kind of flexibility, this kind of engagement with the environment has to come with feeling or else it couldn't be done. And that's the challenge. We that That That's exactly the assumption that's being smuggled in, which has not been justified, which is why I think it's so important to resolve these other issues first. That That's

1:24:37 part of why I struggle a little bit again with the target because the more broad and generous our concept of consciousness is. So, you talk about the idea of creature consciousness. Um it feels like it's slipping away from the what it's like to be intuitive starting point you choose towards something that is more about functional capability in an environment. And And it feels like it's slipping very quickly to a point where you can talk about, as you say, you know, Roomba or a Tesla or a

1:25:11 thermostat or even a ball rolling down a hill or a a plant turning towards the sun as being just having a functionally appropriate response to a physical environment, and that being in some sense conscious. And Yeah, I do say that. When I When I When I Yeah, and and and my intuitive response to that is quite similar to when I hear the panpsychists talk about electrons being conscious. I'm I just feel like you've you risk stretching the concepts, and I do think consciousness basically is just a concept we're deciding how to apply appropriately to stuff, right? But you

1:25:44 If you stretch the concept so far away from the intuitive starting point, which is that I'm pretty confident I'm experiencing things right now, and that's what I want to understand. I worry we're not really talking about the same thing anymore because it's just gone to the point where we think, well, almost every aspect of physics has some sort of you know, functional fitness to its environment, and you know, and we've lost the target again. So, it's So, I mean, the target for me is phenomenal consciousness. And so, I was saying I think you could have

1:26:16 intelligent behavior without phenomenal consciousness. Yes.

1:26:50 particular characteristic of a physical physical [snorts] entity, right? And again I'm just like you you you narrow it down. You narrow it down. Is it the thermostat? Is it the ball? Is it the plant? And I don't know. I'm just worried with the the concept of consciousness has gone so broad that we end up blurring it with ideas like functional fitness and capability and intelligence. So so you know, my spreadsheet or my championship season is just

1:27:23 conscious phenomenal consciousness simply is creature consciousness. Um I think that you know, so I think empirically though we do have to stretch these things then because we have cases that suggest that you could have wakefulness without awareness for example as in non-responsive wakefulness formerly known as the vegetative state where the person goes through sleep-wake cycles, opens their eyes but is non-responsive. It doesn't they don't seem to be aware at all but they're awake. So I think that we In the other case you could be aware but not awake as in the case of

1:27:56 sleepwalking. Um where the person is sleeping but they are in some sense aware of the refrigerator as they open the door. And so now are they consciously aware? I would say sleepwalkers are zombies and They're experiencing anything. Whenever we wake sleepwalkers up, they report not having any dreams. They We have no reason to believe they're acting out a dream experience. They seem to be aware of the environment, though, in some limited sense. I mean, some of these sleepwalkers have driven cars. Uh so, you know, and not crashing. Uh and other, you know, there's one famous

1:28:28 case, guy made a cigarette took a cigarette uh ashtray and put it on a piece of bread and put mayonnaise on it and then ate it. Um so, you know, that that's Clearly, he's not aware that he's eating something disgusting, but there's some sense in which he was using the knife in the appropriate way, put the bread in the right spot, he held it like this, took a bite. Um so, there's some awareness there, I would say. Uh and the sleepwalking person and the self-driving car seem like on a par to me. I don't see why you would want to distinguish one from the other given that they are responding to the

1:29:01 environment in a seemingly automatic, non-conscious way. Um uh Yeah, so I so I would say we have empirical reasons for like if you trace these out, if you think about like, okay, so there's the coma, then there's locked-in syndrome, then there's somnambulism, the vegetative state, there's like all of these things start pulling apart the different threads of wakefulness, awareness, responsiveness, and then the question arises, which one of those things is the one that's important for experience? Like when the person who opens their

1:29:36 eyes, but is non-responsive, are they a zombie? Or are they having some weird, blank experience just of like what? I don't even know what it would be like, but like So, that's an important question, especially if you think it matters morally whether or not they have a conscious experience or whether or not they are opening their eyes and they're just literally a zombie. Um so, tho- those kinds of That's a real case. That's not a fit science fiction case. That's a real human being who's in in particular position. So, I I think that these we do

1:30:08 have to stretch then the um the concepts because the science suggests that this simple approach that there's this one thing that you can't take apart the various components of it is is too simplistic. Um so so when people say there's wakefulness, which means being awake, responsive, alert, all of those come apart. Um and so now we want to know which one is the right one. So, I'm I'm voting for a certain kind of responsiveness to your own mind. Other people will vote for

1:30:41 responsiveness to the world. Some people say you don't need responsiveness at all. So, these are open questions, but but I I want to resist the notion that we're we're doing something untoward or like you know, overstepping the bounds here, but because I I just think that these real-life issues matter. And this person who's in the vegetative state, it's a real serious issue how we treat them um because they're conscious or possibly conscious. Um but because we don't know. And the same is true for the locked-in patient. And the same is true for you know, hydrocephalic children. Um all of these cases it it really couldn't

1:31:16 I I I just it couldn't be more important, I would say, to know the answer to the question whether there's something that it's like for to be these children. Uh or to be this vegetative state patient. So, in that sense, I'm not super worried about stretching it too thin because I think it's really important to know where it's at. Um and I don't want to lose sight of it because what what we're trying to track is like that there's something it's like, that there's a point of view, that there's an experience. But I do think that it's an open question like how thin or which components of that get left over once once we're done with the

1:31:48 science. Um so so yeah. Sorry, I'll stop rambling. Well, um so I was wondering I mean apropos of that um so let's say that we make progress in determining that in the human case X, Y, and Z reliably co-occur with reports of consciousness. So, we get the kind of empirical evidence

1:32:20 that you want. What kind of bridge do you see to answering to going over not just to animal cases, but maybe even to AI cases, right? Because, you know, just in the same way that I think, as you're just emphasizing, it's really important to know what kinds of humans are conscious and when, right? It would be very important to know, I think, if we had sentient conscious AI systems sometime in the future. Agreed. And yet And yet, to me, it's so hard to figure out, even in the human case,

1:32:53 whether X, Y, and Z co-occur with consciousness, given my tendency to mistrust both positive and negative introspective reports, right? But, even if we settled that, right, how we generalize from that specific animal to something that's not even an animal. I was wondering what kind of thoughts you have about that. Yeah. Uh so, the first the first thing I would say is that the kind of evidence that I want is not just mere covariation, but that what I really want is prediction and control. So, I think that what science really

1:33:26 cares about is prediction and control. Um and so, this is, you know, in the ballpark of what Anil Seth calls the real problem. I don't agree with everything he says, but I think that this is a real important kind of point that he that he often makes, which is that the so-called real problem is can we predict and control and explain the conscious experience. And by which I mean, if can I, from a theoretical point of view, say "If you're in such and such a state, you will experience X." And if I change such and such state by factor Y, your experience will

1:33:59 phenomenologically change by phenomenal factor Y. So, for example, making it um the amplitude of the wave higher makes the little sound louder or some some correlation like that that allows you to go in and change the state and get a corresponding difference in the conscious experience that allows you to predict what that change because we can't do that now. We can stick an electrode in the fusiform face area, it goes zap zap zap, and the person goes, "Gee, something happened." But we can't control what happens, and we can't make it be the case that they see their uncle's nose as opposed to their aunt's

1:34:32 nose or their uncle's face as opposed to their neighborhood's face, whatever. So, but if we could do that, then the theory that allowed us to do that would be the one that I uh would subscribe to. So, it's more than correlation. It's prediction and control. So, correlation, prediction, but control is a little different. But of course, with control, when you have one organism intact, generally, and you have ethical restrictions on what you can do, Yeah. you know, there's there's definitely some scientific limitations. You were saying, you know, you don't want to do

1:35:05 monkeys, maybe, for ethical reasons, right? But they're obviously if what you're after is control, you know, it's easier when you have systems that you can intervene in in in ways that you can't intervene ethically in humans.

1:35:39 visual input from a camera interpreted by a computer to produce the kind of activity in the brain that would have been produced had the eyes done it, then that I mean, we don't know how to do that, but if we could do that, that would be the type of thing I'm talking about. So it's not unethical to to do these kinds of things.

1:36:14 that? What's your what what We say what factors

1:36:49 Well, then I would say it's not calcium in us, and us is pretty typical, and the things in us generally But you're right, there's some difference. Some neurotransmitters work differently in other animals than they do in in us, but have different functions, and you know, it's sort of it's different circuits, but generally speaking, if consciousness is a natural kind, then you would expect some But maybe it isn't, so maybe the right conclusion is all we found out was about human consciousness. I would be very sad if that was the case. But I don't think we have any reason to expect that that would be the case, given that the other

1:37:21 psychological phenomena that we have discovered, like learning and memory, and attention, seem to be generalizable to these other systems in in a reliable way, so that we can say in a mollusk whether it's learning or not, and we can say in an oak tree even whether it can be conditioned. I mean, there's some debate about that, but I mean, these are things that are empirical questions. So I think the same thing would be the case uh for consciousness. So, would you be optimistic about then being able to generalize

1:37:56 to AI cases? Let's just focus on those.

1:38:29 Then you look for that pretty abstract functional thing in AI systems and you say, "Hey, they'll be conscious, right?" Alternatively, the other cartoon thing would be it turns out you really need you know, such and such neural structure and if you don't have that, then not going to have consciousness and then you just would have

1:39:04 Exactly. So, if if if we mess with the calcium and that like changes the conscious conscious like if Ned is right, then taking the calcium, well, that's not true actually because if he does think that something does the same biological non-computational job. So, if if Ned is right, then taking out some of the um some particular biological thing is just going to get rid of consciousness all all together. Um and we can test for that. We can see if that happens uh by doing it to me or you um and you know, uh

1:39:36 or you know, some other human case. Um and find out and what you know, maybe we do it by TMS or reversible method. So, you know, you don't have to go in there and chop stuff up. Um but you'll got to like, you know, see what happens when you mess with the system. That's the only way to establish cause and effect in in the sciences. Uh it to go beyond correlation. So, that's the only way we're going to make progress here. And if it's something like biological that if you just take that away and then like, "Oh, no, there it goes." Then yeah, I would be very skeptical about AI. Um on the other hand, if it does turn out to be some So, you can take You can

1:40:10 replace all of that slower level stuff and and nothing changes. That's kind of the neuron replacement argument from Chalmers. So, if you can affect to do that basically, then um that seems to me to yeah, leads some very strong credence to uh to the um the view that this functional view is right. Or but even more strongly, if Grandpa Switch Gable comes home with an artificial cortex and says, "Eric, I haven't heard your voice in 30 years. It's so great to hear it." Uh you're you're just going to be instantly converted to functionalism. I don't think that anyone is going [laughter] to say, "Shut up, Grandpa, you clanker."

1:40:43 You're like, "I just don't believe that that's going to happen." You know, I I I think Touch stuff. Yeah, exactly. So, [laughter] uh I I I think that we're I wonder if this methodology kind of bakes functionalism in ultimately. Which maybe is not something to object to in it, but maybe just something to notice about it, right? Because if the if the test is you swap out this biological thing and does it change the function, then you're kind of assuming that at least evidentially,

1:41:17 what matters is the function. Right? So, you couldn't have a case where I'm I'm assuming. That's why I'm inviting you to comment on this, right? So, if you had a case where you swapped out the calcium for something else with with a uh plus two um and it had no change in the reports,

1:41:51 you wouldn't detect it because you're using functional tests Yeah. for consciousness, right? So, at least scientifically and epistemically, it seems like you're relying on Well, let's just say in a let's call it epistemic functionalism, Yeah.

1:42:25 to know. Right, so you get like uh Susan Schneider's chip test, where you're like a swap in the chip,

1:42:58 all the same, you you couldn't do any reporting differently. But you would still know. I I think that in that moment you would instantly cuz I think we have a direct a kind of immediate direct acquaintance with our own

1:44:33 saying it isn't that way and I don't because I'm going to end up being an illusionist or denying that it's real and I just think that you have to start with the reality of consciousness and accept the consequences. If the consequences are you have to be a panpsychist then be one. But I just don't think those are the consequences. So so I would resist that that I just think that there's no reason to believe that subjective phenomenal consciousness truly real phenomenal consciousness is a physical state of my brain just is a

1:45:07 chemical state a physical electrical chemical state. People say they can't understand it. I said I don't care if you can understand it. I mean so what? Why is your understanding of whether it's true or not reflective of whether it's true or not it doesn't matter to me. But some some people find that very puzzling. So yeah.

1:45:41 I was like, "No, I'm I have to do that." Because yes, we often do talk about in a in and I wrote this paper a long time ago when I was graduating called What Is a Brain State? And like the whole opening kind of move in the paper is like, "Why are we talking about states when really we should be talking about processes?" And I kind of say, "Well, gee, it's you know, people it's just intrinsic in the literature. People talk this way, but really what we mean is processes." Um Yeah. I am mindful of the time limit you have, Eric, by the way.

1:47:11 Yeah. A tight circle.

1:48:56 neurosciences, and we want to overlay them and try to fit this harmonious mesh. And take the widest possible swath of the data that we can, all of it in a one big holistic kind of clump. And so it's for the other side as well. And so each side is going to because of the under-determination of data by the evidence I think or whatever you know what I'm saying theory by the evidence and you're going to be able to do this. And then the question is once you have sort of done that

1:49:32 have you worked yourself into an uncomfortable position where you're saying all sorts of weird strange things that lead to dead ends or are you saying things that are opening up interesting questions having surprising connections to other areas. And so I will I do think of a more holistic approach as the right way to think and that you don't want to go experiment by experiment piece by piece but and there's not going to be like one silver bullet that's going to like yay this or yay that and it's not going to be a priori either but you are going to be sort of in in classic Neurath boat style right you're going to be

1:50:05 constructing this picture of this evidence trying to map these relations between the psychological level and the neural level in a satisfying way and the other side is doing it too and then you just compare them you say okay so you have to you're committed to this that sounds unusual and I well I'll up you with that and so we just have to kind of compare them and negotiate back and forth. And so I don't really think of it as a priori I think of it as like it is circular in the sense that you are always stepping to one side to rearrange

1:50:39 over there and it's stepping back over there to rearrange over here so there's always a back and a forth and you can't it's so where are you starting where are you stopping you are going back and forth. So that's just the way you that's the way science proceeds. Uh-oh. Okay. Okay.

1:51:43 Yeah. Well, when you put it that way, I Yeah, no, and and and I would just say

1:52:19 that I don't think self-report means an awareness of your own mental state, cuz you can self-report, you know, in a way that's just first order. So, Yeah, so so I would say that I don't Um So, so I I think Yeah, um that what the issue here is is whether or not consciousness goes with those kinds of reports or not. Um and that's what we're trying to figure out whether the answer what the answer to should be.

1:52:52 So, I don't think I'm assuming that.

1:53:24 a science of consciousness? Because it sounds like what you're saying is that we can't do a science of consciousness because of this reason, but that can't be right. Yeah. Oh well, I agree with that though. So I that's why I that's why I want to double-check what am I arguing against cuz I don't want to argue against something I agree with. Um so I agree we're very very far away partially because we don't know how to resolve

1:53:57 some of these issues, but what I'm trying to disagree with you is that these these are the issues which we want to be able to resolve. And if we can't resolve them, then I don't think we really have a serious endeavor that we're doing. So the the question of does when I say that I see it, is that a good reason to believe that I actually saw it? That I do think is kind of a background assumption and many people I would say assume that well, when a when a normally function human being in all things considered cases says, "I see

1:54:31 it." that they could be sort of mistaken here and there, but they generally see it. Um that that's a kind of background assumption. Is that a priori? Uh it's Okay. Exactly. No, I but I agree with that.

1:55:18 Right. But but it but so here's how you can turn it into you can turn it into the kind of case I like which is you can start by finding like let's just imagine a a case where you could um uh brain clamp the person and control the mental state that they're in by adjusting the neurophysiological uh responses. Um then uh now let's let's hold that certain state conscious state fixed and then mess with the with the higher order

1:55:50 states. Um now okay, so does the conscious experience of the subjects change? If you can hold the first order state constant but change the higher order state and the experience of the person changes, then that's some reason to think that the experience goes beyond what the first order states are are doing. So now Well, but I'm not talking about report. I'm just I'm not we're just talking about holding all that other stuff conscious. Right, exactly. So what So I haven't got

1:56:32 to the case yet. So when when the person is saying I see it, then we can look for whether or not it's what's happening in this case. So they say I see it. Um and then we hold the first order state uh the same and we change the higher order state. And they say, "Oh, I don't see it anymore." So So that's a report, but a positive report. Uh-huh, so now we have some reason to think that it if

1:57:24 Well, we're assuming we can isolate the higher order state independently of the report. So, I I I think you could get the person. Why not? Yeah, I I don't It It's that's a possibility, but there's no reason to start off assuming that. I mean, you know, there's lot lots of ways that you could report, and one of them maybe that you're feeding off of

1:57:56 this thing. But, but my point was just that that the person who says "I see it disappear." is making a positive report, not a negative report. They're saying they see something. And if you could go in and mess with the relevant states and cause that to happen, then that would be you you turn it on, they go, "There it is." You turn it off, they go, "Where'd it go?" Um that seems to me to be going beyond the kind of case that you were mentioning. Uh but, that's what I meant by control. That's what I mean by by going in and

1:58:29 being able to turn on and off these things uh at will, so to speak. Uh we can't do that, but if a theory allows you to do that, I would be like you have every you can't know anything with certainty from science, but you have every reason to believe that this is on the right track in those circumstances. I would say, anyway. Yeah. Exactly. It yes, exactly.