Support the show to get full episodes and join the Discord community.
Anil and I discuss a range of topics from his book, BEING YOU A New Science of Consciousness. Anil lays out his framework for explaining consciousness, which is embedded in what he calls the “real problem” of consciousness. You know the “hard problem”, which was David Chalmers term for our eternal difficulties to explain why we have subjective awareness at all instead of being unfeeling, unexperiencing machine-like organisms. Anil’s “real problem” aims to explain, predict, and control the phenomenal properties of consciousness, and his hope is that, by doing so, the hard problem of consciousness will dissolve much like the mystery of explaining life dissolved with lots of good science.
Anil’s account of perceptual consciousness, like seeing red, is that it’s rooted in predicting our incoming sensory data. His account of our sense of self, is that it’s rooted in predicting our bodily states to control them.
We talk about that and a lot of other topics from the book, like consciousness as “controlled hallucinations”, free will, psychedelics, complexity and emergence, and the relation between life, intelligence, and consciousness. Plus, Anil answers a handful of questions from Megan Peters and Steve Fleming, both previous brain inspired guests.
Thanks for being here. A huge congratulations on the book. I know I sent this to you in email, but, uh, it was very refreshing, refreshing, really clear, and, uh, the writing style was just easy and fun to read. So nice job.
Anil 00:03:49 Thanks for that. That means a lot. And thanks for having me.
Paul 00:03:52 We’re going to, I don’t want to bury the lead here, but I want to ask a couple of questions just about writing the book and the book itself before we get into its contents. Um, so on a recent episode I had, uh, Steve Grossberg on and by the way, I have you, um, heard his or read his, uh, his recent tome or any of it conscious mind resonant brain. I have,
Anil 00:04:15 I haven’t yet. I have ordered it, uh, but I have not yet read it. Okay.
Paul 00:04:19 Well, it is, it is a massive, massive book and it’s just kind of a collection of his work and it took him like 30 years to write. Uh, and, um, this is the same, uh, like mark, I had mark Bickert on and he’s not done writing his 30, 40 year book. Uh, but you’re like 25 years old. Right. So yours couldn’t have taken that long to write. How long, how long have you been working on this?
Anil 00:04:40 Hold on, did you say I’m 25 years old? I wish I was 25 years old. I’ve been, I guess, in the business for about 25 years. So it might be my, uh, you know, I started out in the mid nineties. Um, it took me a bout, was five years that I signed an agreement to write the book. And then of course I think like many people I just said, wow, that’s great. Now I’ll put that away. And, uh, and forget about it for a bit. And I think I started seriously writing it, uh, about three and a half years ago. I think it was took about three years all in all to write.
Paul 00:05:18 Have you always been such a clear writer?
Anil 00:05:21 Of course not. No, it’s it’s, I think it’s, it’s really a skill that’s to be learned. And of course with this book, what clarity there is, is also great credit to the editors that I had to, and the contribution that a good editor can make to a book is just really hard to overestimate. And then they can do amazing things with otherwise garbled and completely incoherent text.
Paul 00:05:45 Yeah. I was harping on this fact, uh, many scientists and researchers don’t seem to use editors and the work suffers because of it. So, uh, I just appreciate that in general. So I told you, I have, uh, I have a handful of questions, uh, from some folks that, you know, and some that you don’t, uh, and I thought I’d start off by playing one of those questions here, as it pertains to communicating, uh, consciousness. Right? It does. It seems like there was a slew of consciousness books right now. It’s like the golden era of consciousness books. Um, do you agree with that?
Anil 00:06:21 I don’t dunno. There always seems to be a slew of consciousness book, so that’s one of the, one of the beautiful, but also slightly challenging things about the area everybody’s interested, but there’s a lot of stuff out there too. A lot of competition.
Paul 00:06:32 Maybe, maybe it’s my bias since I do a podcast. Okay. So I’m going to play this for you, and then I’ll let you answer here.
Megan Peters 00:06:40 Hi, and Neil and Paul. This is Megan Peters at UC Irvine. Thanks so much for giving me the chance to ask some questions. What do you see are the greatest challenges in public outreach about consciousness science right now? And relatedly, what do you think are some of the most promising approaches to solving those challenges?
Anil 00:07:02 Hello, Megan, how nice to hear for me? And, um, yeah, I, well, that’s a, it’s a good question. Of course. Um, Megan is a, a colleague and friend of mine. Who’s doing brilliant work in metacognition and consciousness. And I think she’s been on your podcast too. I think I remember hearing it. It was a great episode. The main challenge from my perspective is I think there are two challenges. The first is probably common to a lot of public communication of science in general, which is how do you express things, accessibly, and clearly without oversimplifying, that’s just a hard balance to find, but it’s a possible balance to find. And the approach to that is just continually trying to refine the way you put things, the examples that you might use, the metaphors that you might use, I’m always worried that I’ve oversimplified, that I’ve misrepresented others’ opinions or misrepresented the literature in general.
Anil 00:08:03 It’s hard to get away from that worry. Uh, but you can’t cover everything. You can’t caveat absolutely everything either. You have to have the message you want to convey, but make sure that the evidence, the arguments that you’re resting, that message on they stand up. I think that’s one general challenge for public communication of science. The other one for consciousness is an amazing thing about consciousness working on consciousness is really that it’s not hard to get people motivated by the topic itself. People will just come to the table with a strong April, very interest in consciousness usually,
Paul 00:08:44 And a priority view on it. Right.
Anil 00:08:46 Well, that’s right. Yeah. They also come with very strong opinions about certain things about consciousness, what it is, what the definition is, what something like freewill might mean. And sure, we’ll talk about that too. So here, the challenge is how to actually engage with preexisting views rather than just try to dump my own views into somebody else’s mind, how to turn it from election to it dialogue.
Paul 00:09:13 Hmm. All right. Very good. Well, I don’t want to bury the lead anymore. So I’m going to start actually, by reading a quote from your book that is kind of the crux of the, of your message, and then I’ll let you unpack it and, and, and, you know, kind of give an overview of, of your views. Uh, before we move on, like I said, I have, um, I have a host of my own questions of course, way more than we’ll get to, but then, uh, I got a lot of, um, questions from, from people like Megan. So it will be hearing from her again. So I want to make sure that we get to those. Okay. So, uh, this is from the book. This is about two thirds of the way through. Maybe this for me is the true ground state of conscious selfhood, a formless shapeless control, oriented perceptual prediction about the present and future physiological condition of the body itself. This is where being you begins. And it is here that we find the most profound connections between life and mind between our beast machine nature and our conscious self. So, uh, I don’t know if you feel that that kind of EnCap encapsulates the whole message, but there’s a lot to unpack there as well. So I’m going to let you have the floor
Anil 00:10:23 Well, great. There is a lot to unpack there because that’s in a way, that’s the culminating statement for the book’s main argument. So I don’t want to try and explain the whole book in, in, in unpacking that, that summary, but I think what’s important about it is, is the first thing I say about it is this wasn’t really the idea that I set out with it at the beginning, this idea that this ground state of self is, is in a predictive control oriented perception about the physiological condition of the body and its trajectory. That’s something that, that came as I followed the thread of ideas over the years. And I think that’s one of the rewarding things also about writing the book. It wasn’t that it was just putting down all the ideas already had. The writing of the book was an extremely for me by turns frustrating and challenging, but also rewarding way of putting all the different threads and weaving them together into something that was new to me as well, by the time I’d finished writing it.
Anil 00:11:26 And this is particularly true in this connection between life and consciousness. So the idea here is that I started to think many, many years ago about perception as this form of prediction about the causes of sensory signals. This is not a new idea. This goes back to him and Von Helmholtz in the 19th century. People talk about it in terms of predictive coding, predictive processing, active inference, all these related ideas that see perception as an inside out top down construction, rather than an outside in bottom up reading out of the world around us. The key part of that story for me was that the contents of what we perceive the nature of it, the, the character of the different perceptual experiences that we have should relate to the kinds of predictions that the brain is making about the sensory signals to visual, that experiences have a particular character because the brain is making predictions about how visual signals relate to behavior.
Anil 00:12:33 What happens when we move our eyes, for instance, these sorts of things, but the brain is also dealing with perception and control of the body. And the body is as remote to the brain as the outside world in, in, in the fundamental sense that it has no direct access to what’s going on in the body. It still has to infer what’s going on in the body on the basis of noisy and ambiguous sensory signals. So there’s this process or the argument is there’s this process of predictive perception about the interior of the body, but experiences of the interior of the body. They aren’t like visual experiences. They don’t have colors and shapes and locations in space. They have valence, things are good or bad or likely to be good or bad in the future. And so the idea to connect the two kinds of experiences is that, well, the predictions that are involved in perceiving the interior of the body have a different function, instead of figuring out where things are to a first approximation, they’re about controlling and regulating things.
Anil 00:13:38 That’s why we have brains in the first place to keep the body alive. And so if a prediction is being used for control, and there’s a whole literature from cybernetics and control engineering, and now to free energy principle and active inference that tells us that to control something, you need to be able to have a good predictive model of it. Well, then the character of the resulting experience can be understood as, as emerging from a relating to that control oriented function. So this is a good for me handle on why self related experiences feel differently from let’s say visual experiences of the outside world, and it underwrites this close link between life and mind, because now these very basic experiences of just being a living organism. And here’s where I do talk a little bit vaguely about the phenomenology, the experiential character of this. What, what really is the base level of being a self it’s not the identity that you have with the name and the set of memories.
Anil 00:14:43 It’s not really even the experience of what objects in the world is my body. There is, I think for me, this very deep lying sense of just being a living organism and that, and this is the proposal. It’s not something that I can justify on the basis of clear experiments with data that, that, that experience, that Bazell experience emerges from the, the role of the brains predictive mechanisms in regulating the body, and then everything else flows from that. So all the perceptual mechanisms that are, are, and now dealing with the outside world or with the body as an object in the world, all have that evolutionary developmental and kind of moment to moment functions being grounded in this basic imperative to stay alive. I mean, this is not a completely new idea either, right? I mean, it’s got lots of resonances with people like Antonio Damasio, uh, with mark Solms with Lisa Barrett, um, with Evan Thompson, there’s lots of rich territory, rich literature exploring these life, mind connections, but my way of doing it is to emphasize predictive perception and predictive processing as the common thread.
Paul 00:15:57 Yeah. So the, um, I mean, there’s a lot of different ways to go here. One of the things that, you know, so you talk about, uh, the possibility you end up talking in the book about the possibility of AI having consciousness, and maybe we’ll come onto that later. But one of the things that I have as my own thinking about consciousness has very minimally developed over the years. It seems because I have, I’ve not thought nearly as deeply about these sorts of things is that I do appreciate the richness of the, you know, the feeling of identity of having a self that you’re saying, you know, and the narrative self and personal identity self, uh, there’s a certain richness there that you’re saying doesn’t need to be accompanied in the, um, uh, predicting our life control setting processes to stay alive. And so then that, of course, uh, makes you wonder about all the different animals and organisms and their level of, you know, I suppose, minimally, they would need to have a predictive mechanism to, you would say that, um, the conscious experience is somehow situated in that predictive, uh, cognitive mechanism. Uh, but, and, and you write about, of course you write about animals in the book, but, um, maybe you can just describe what it means you think for the, the experiences of other animals, uh, and then we’ll come onto the experiences of other people as well.
Anil 00:17:20 Yeah, I think that that’s a very rich topic. It’s very important topic as well, because one of the main implications of a, of a well-grounded science of consciousness is to make informed judgments about the potential for suffering and the potential space of experiences of non-human animals. And there isn’t inevitable tension here because so far we just lack a consensus view on the sufficient mechanisms for consciousness. We, there are competing theories, we have different ideas. Um, so at the moment inferences about other animals is still using humans or mammals as a benchmark. You know, we, we take what we know from humans. We extend that to other animals, which, um, will mammals for instance, have pretty much the shared neural mechanisms that we know are important in humans. And then how far out can we go? This is a strategy that’s hard to get around, but of course there’s also the recognition that the way we are conscious, the way we experience having a self is not the only way it seems to be all bound together.
Anil 00:18:27 So this is the thing at the experience of being a human self is that we have all of these different aspects of it, a name and identity over time experiences of agency, of free world of seeing the world from a first person perspective of having a body of being a body of being the seat of emotions. All these things seem unified, but of course they aren’t neurology and psychiatry tells us that they aren’t and various experiments tell us that they aren’t as well. And if they aren’t necessarily unified, then there are different ways they might come together in different people, but also in different species. So the space of other minds is, is very large. And what can we say about that space? Well, we can’t have the experiences of another species or indeed of another person. And this is a very old point in philosophy, Thomas Nagel, what is it like to be about?
Anil 00:19:17 But that doesn’t mean we can’t understand from a third person perspective, something about what those experiences might be like, if I can characterize, for instance, what the differences between a visual experience and an emotional experience in terms of different kinds of predictions, then that provides a language for thinking about other kinds of experience to how they may relate to the experiences with familiar with, even though we can’t instantiate those experiences ourselves. The harder question to answer is how far does the magic circle act stands? Know when does sentience ground into nothingness in the animal kingdom? That’s, that’s really, really difficult. I don’t think there’s a sensible way to answer that there is something about level of complexity of the nervous system that, that seems to me has to be important. It seems unlikely to me that C elegans and had this, this tiny worm with, um, 300, two neurons is, is conscious, but maybe that’s just a species specific bias on my part. I mean, we know that number of neurons per se, doesn’t matter, the cerebellum has three quarters of the neurons in the brain, and doesn’t seem much involved in consciousness if at all, to humans. So all these, these intuitions that we have, have to be very careful about the extent to which they’re based on a sense of anthropocentrism of human exceptionalism. Uh, and that’s the tension at the heart of thinking about animal consciousness. For me,
Paul 00:20:47 Poor poor C elegans always, always gets shafted with regard to, uh, admitting consciousness in the NC elegance. All right. So, uh, I’m in danger of just going down the rabbit hole on my own question. So here’s what we’re going to do. We’re going to, we’re going to go through the rest of these questions and because you already mentioned some things that related to some of the questions, and we can just use that as a jumping off point to talk about, you know, things that you’ve, that you write about in the book. So, uh, here’s Meghan’s next question.
Megan Peters 00:21:15 Uh, Neil you’ve described the real problem of consciousness as being separate from the hard problem of consciousness. So related to this idea, do you believe that discovering and characterizing something like a quality space, like Rosenthal’s quality space and similar kinds of, uh, writings and efforts by hoc one loud and now to chia and others, do you think this is truly going to be enough? Do you think that this is going to make the hard problem truly just disappear. If we reach a relatively full description of this quality space and relatedly, um, what do you see as the biggest challenges in making the leap from a causal description, like a quality space, a full quality space description to a true explanation of consciousness. Do you think that this distinction is also going to disappear as we get closer to such a full description?
Paul 00:22:12 I told you Megan really went after it.
Anil 00:22:14 Yeah, I know. And I like the way she, her questions always have a and relatedly halfway through that’s when I start to worry. So, but it is a very good question. Of course. Thank you again, Megan. Um, yeah, I talk about the real problems quite informally, really, because it’s very related to other approaches that I’m sure Megan knows, but, but more generally for, for the listeners, these are approaches like neuro phenomenology that traces back to Francisco Varella. Uh, this general idea of, instead of trying to explain how it comes about that any physical system could be identical to, or give rise to a conscious experience. This is the hard problem of consciousness broadly speaking from, from David Chalmers, how consciousness happens to be part of this physical list material, this picture of the universe, um, or what’s the relationship between conscious experiences and stuff in the universe.
Anil 00:23:13 And the real problem is saying, well, let’s not address that directly. Let’s go after it in directly. Let’s try to explain, predict and control the properties of conscious experiences in terms of things happening inside brains and bodies. So these quality spaces that Megan mentioned, this is one aspect of doing that. It’s a way of trying to organize different kinds of experiences, according to metrics, you know, how similar they are, how different they are, how they relate to each other. And you can think of that as, as very related to this idea where you were talking about a bit earlier, that different kinds of predictions can go along with different kinds of experiences. It’s trying to, they’re both different ways of talking about how you organize a space of experiences and relate it to mechanisms. Um, this, by the way is not the same as David Charla’s easy problems, which are questions about how the brain works when you just basically take consciousness entirely out of the picture and just talk about function behavior and so on.
Anil 00:24:15 So I do think it’s a useful middle ground. It’s a pragmatic way to approach the science of consciousness. The question is, will it be enough? And this is a tricky question to answer from where we are now with the tools and with the concepts that we have now, I think very hubristic to say it definitely will be enough, but I also think it will be disappointingly or an unwarranted, really pessimistic to say it definitely won’t be enough. I think that is a case that can be made for a healthy agnostic optimism about it. And the reason I say that is because the history of science just gives us plenty of examples, where things that have seen mysterious no longer seem that mysterious because of insights that we have, uh, because of a sort of real problem like approach. The classic example of course, is the study of life that we, instead of looking for one Eureka solution, a spark of life and our own Vitel biologist characterize the different properties of life and explain them as a related set of problems.
Anil 00:25:23 And the hard problem of life wasn’t solved. It was dissolved, but life is not the same thing as consciousness. So that’s why I can’t be fully optimistic about it because you can still agree objectively on the data about life. Whereas the data about consciousness are intrinsically private and subjective. It makes it harder, but in my view, it doesn’t make it impossible. And so what will the trajectory of this approach look like? The thing that I actually think is most likely to happen and will be most explanatory in the end to Megan’s. The second part of Megan’s question is when our whole idea of explaining consciousness, uh, should achieve actually changes. Now, we, we, when we set out the problem that we think we’re setting out to solve at the beginning turns into a slightly different problem, and we see consciousness is continuous with the rest of nature. And we worry a bit less about how to explain this apparent division between the mental and the physical, the final point on this is that we sometimes ask too much of a science of consciousness. And this gets at Megan’s second point. When I think she asked what would count as a true explanation of consciousness, right. And there’s a lot, there’s a lot under that. What does she really mean? What does, what, what, what should a true explanation actually mean? And in philosophy of science, this is a complicated question.
Paul 00:26:45 Yeah. I was going to ask you, um, part of my own questions, whereas, you know, when we get there too enough, I’m using air quotes. Is it going to feel intuitively, uh, satisfying or, you know, because of, uh, theories like integrated information theory that you kind of just have to accept that there’s a complexity and there’s a number and it doesn’t feel quite intuitively satisfying, but I don’t know if that’s what she means by enough, but I was going to ask you anyway, if getting there wherever there is, if that will feel, but, you know, because I’m not even sure we understand life yet. I agree that the, um, the mystery, the mystery of it, Ella and Vitale has dissolved because we started asking different questions. And I agree with you that we will need to shift our approach and shift the conceptual, um, uh, approach to, to understand what it is. But, um, and yet, I don’t know if it’ll feel intuitively, uh, satisfying,
Anil 00:27:44 Right? So when it comes to life, that’s a really important point you made because indeed not everything is understood, but the sense of mystery about things being ex explicable has dissolved at least for most people. I think that’s a good indication of, of a mature science of X doesn’t have to explain absolutely every detail, but the sense of deep mystery about details being explicable should have dissolved. And when it comes to consciousness, it’s sort of often put the other way around that people might say things like, well, let’s assume we can explain every single detail wouldn’t there still be a sense of big mystery. Um, this is an odd way to put it because even the premise is not necessarily something that, that we should take as a, as a necessary criteria. The important thing is, does the overall sense of mystery dissolve? And then I think in terms of the standard criteria that we apply to scientific explanations, can we explain a phenomenon now, this it’s tricky, but here I might say, this is back to the realm of different kinds of predictions and Qualia spaces, whether they’re the sorts that David Rosen, Tyler and Megan, and hot corn loud talk about, or the Qualia spaces and integrated information theory.
Anil 00:28:59 It’s another way of thinking about that. Um, explanation prediction. Can you predict when a particular kind of experience will occur and control? Can you intervene in a mechanism to bring about particular kinds of experiences in, in systematic ways? If you can do that, then you’re doing pretty well. Will this be intuitively satisfying? We’d like to think so, but I don’t think there’s any guarantee that it, that it will be. And of course, whether something like integrated information theory is intuitively satisfying, really depends on who you are. It’s firstly, it’s a very complicated, uh, theory. And when you do delve into the math, there are some beautifully intuitively appealing points about it, but there are things that challenge intuition as well. The key point though, is that we tend to require or smuggle in this criteria and for intuitive satisfaction when it comes to consciousness in a way that we don’t for other areas of science, like we, we know that quantum mechanics is just makes no sense. It makes no sense, whichever interpretation of quantum mechanics, you tend to favor. None of them make any sense at all. Uh, but it’s a beautifully successful site. It’s not a complete science, it’s a beautifully successful science. Does it have to be intuitively satisfying? I think we smuggle that into consciousness and we feel that that has to play out partly because we ourselves are conscious. We’re trying to explain us. And that I think leads us to ask different things from a science of consciousness where the scientific method may not justify as asking those things.
Paul 00:30:40 Well, I think it means that we need to use quantum mechanics to, um, situate consciousness in microtubules right.
Anil 00:30:45 Do not go there.
Paul 00:30:49 Okay. Where we are going is, is the next question. You might recognize this chap,
Steve Fleming 00:30:56 Steve Fleming hair, congratulations to the book. It’s a fantastic achievement. So my question for you is about how we should think about the contents of consciousness within the beast machine framework in biological agents, such as ourselves, there are some things we’re aware of and other processes or newer representations that we’re not aware of. And for me, modeling this kind of distinction pushes you towards a more cognitive or higher order model of how consciousness works, even when we’re thinking about embodied living systems. I suspect you disagree though, and would love to hear your thoughts, enjoy your chat with Paul, and I’ll look forward to listening.
Paul 00:31:40 Another friend of yours,
Anil 00:31:42 Another friend of mine. Hello, Steve, thank you for the question and thank you for your book as well and reverse congratulations to you on know thyself. It’s a, it’s a brilliant book, um, and very much enjoyed it. It’s another good question. Of course. And the answers I think I, I, I quite agree actually with, with Steve probably more than he was expecting, if you take as one of the core methods and it is one of the core methods in consciousness science to contrast of conscious versus unconscious perceptions, um, then your you’re maybe drawn to the cognitive processes that mark that distinction. I mean they can, but they can play out in different ways. They could be on some theories of consciousness, like the favorite higher order type theories of, of Steve and Megan and how allow the differences in the kind of higher order representation that in some sense, it looks down at other processes going on in lower order perceptual circuits, whether they’re to do with the, the world or the body, but these differences could also be in these lower order circuits as well.
Anil 00:32:53 That explain the difference between conscious and unconscious perception. I think this is a very valid approach. I think that is, as Steve will know, there’s still a surprising degree of controversy about whether unconscious perception really exists at all. As, as a phenomenon, the closer you look, sometimes it seems to go away entirely. And it’s also a methodology that is, it works better for some kinds of experiences than others. It works to the extent that it does work. It works really well for extra receptive perceptions, vision auditions, things like that. You can have all these rapid swabbed, his toolbox of masking techniques that we can use to at least approximate this conscious versus unconscious content divine. These toolboxes, just that don’t apply or they’re certainly not as readily available for studying processes about the perceptual regulation and experience of these deep levels of self, like emotion.
Anil 00:34:01 A lot of discussion goes into questions about, are there such things as unconscious emotions? What would that mean? Is there an unconscious mood? It strikes me as completely plausible that there are some aspects of the brain regulating of the body that do not arise into our conscious experiences. That they’re very basic homeostatic reflexes that don’t seem to shape conscious contents in terms of moods and motions or anything else. And there are others that do. So the question for me is, is more, it’s a very open question. Like what, what, what level of perceptual regulation or perceptual inference are there corresponding aspects of conscious content and at some level there aren’t, and at some level that are, I think this is, this is likely to supply, although, you know, maybe less so to the body, but it’s just a much harder question to get at experimentally. And it doesn’t mean that we just give up, it just means that what are the other methods, the thinking about consciousness, uh, without using these contrasts between con conscious and unconscious perceptions. And this is where I do get drawn to these ideas, more of explaining the phenomenological properties of a conscious experience that is there rather than worrying about when it is or when it isn’t. And what marks that specific difference
Paul 00:35:31 Do you think that there, um, since the majority, the vast majority of consciousness neuroscience has focused on perception and specifically visual perception that that has biased our intuitions about what might be needed, uh, in the perception predict in, you know, higher order toward higher order thought type of, um, uh, approaches, uh, because, and the other side of that coin is do you think that processes like self self-maintenance and, uh, life processes that you’re focused on have been under appreciated?
Anil 00:36:07 That’s a bit of a judgment call, isn’t it? I mean, the, the, the focus on vision is, is I think very sensible in many ways. It’s if you, if you think back to the early 1990s, when Francis Crick and Christophe cock were talking for the first time about the pragmatic strategy of looking for neural correlates of consciousness, they were focusing on vision. And I mean, this is a bit of historical reconstruction, but I imagine one good reason for doing so was just that you could do experiments that way you could build on a whole literature of psychophysics to actually do those experiments. And it could just be a more compelling argument to the rest of the community in psychology and neuroscience, that there was a reasonable way to study consciousness. So I think if Kurt and Cox had started off by talking about the Deepak’s deep embodied experiences, it would have been much, much harder to sell because you can’t go and do your experiment the next day.
Anil 00:37:02 You can’t do binocular rivalry and go, oh, look at this. Uh, so there are good reasons to focus on vision and visual experiences are well characterized, the level of phenomenology. We have all these super interesting aspects of it. What’s, what’s the relationship between central vision and peripheral vision. Uh, there’s a lot known about the visual cortex too. It’s, uh, it’s a relatively well understood part of the brain in terms of its organization. So these are all the good reasons. I don’t know if a study, if th this, this bias towards division also bias towards higher order theories. I think, I don’t think that’s true, actually, because there, there are people like in that block also deeply rooted in the visual tradition who put the opposite perspective and say, look, you know, visual visual experience gives us a compelling case to think that are experienced in the nature of our experiences.
Anil 00:37:56 In fact, more than we have higher order access to at any given time. And this debate rumbled on, I think, between phenomenal consciousness on the one side and access consciousness in very productive ways. Actually, I think it’s a really good debate, but yes, this focus does bias against, uh, recognizing I think these deep roots of all consciousness in the regulation of the body. And as, as we were discussing earlier, it’s not that these ideas weren’t also there from the very beginning. It Demasio one of the early pioneers of consciousness science to, uh, said this very, very, very, uh, my old mentor, Gerald Edelman also talks about the role of the body in consciousness. So it’s been there all the time. Adam Thompson, Francisco, rather the embodied mind, there are, there are very deep traditions that make this point, but they don’t give you the experimental availability that the focus on vision did.
Paul 00:38:55 Yeah. I mean, the vast majority of the predictive processing framework is, um, focused on vision as well because everything in neuroscience, the vast majority is focused on vision, but one of your main moves was to point that predictive processing, um, inference process to, uh, bodily processes. So, um, that was D do you, do you think that that was a key connection that you made that allowed you to start thinking about these things?
Anil 00:39:21 Yeah, actually I do. And I think, I mean, the idea was around at the time. So I think Lisa Barrett in particular had a very similar idea more or less at the same time. It’s not a, it’s not a massive leap at all. They’ve been going a long history of, of thought and emotion that thought about emotions as cognitive appraisals of changes in physiological state. So there was already this sort of interpretive framework applied to thinking about emotions, uh, but it was still in this, this framework of a cognitive part of the brain and a non-cognitive part of the brain and one appraising the other. And the simple idea was just a staircase look in predictive processing, where you have this general principle that perceptions depend on the whole ensemble of top-down predictions about causes of sensory signals at multiple levels, without any clear, bright line between the cognitive and the non-cognitive.
Anil 00:40:15 Well, that that just provides a, a modern gloss, a predictive processing gloss on these older ideas of how emotions are formed. They now become multilevel multimodal predictions about the causes of sensory signals, but now the sensory signals are the sensory signals that come from within the body interoceptive processes. And just making that connection is a start to then, for me, the really important part of that was thinking about these predictions is having different functions in the two cases. So we already mentioned this, that vision to a first approximation perceptual predictions, trying to figure out what’s their interoceptive predictions, try to control and regulate.
Paul 00:41:01 Uh, we’re going to continue on with Meghan number three, and then we’ll have one more Megan. Okay.
Anil 00:41:06 Let’s see if there’s an and relatedly halfway through.
Megan Peters 00:41:09 Do you think that there is a conceptual distinction that can be drawn between the qualitative character of an experience and the content of that experience? So I’m not talking about level here, but more like the qualitative nature itself. So, uh, in other words is phenomenal character like a substrate or foundation on which the content will rest or is the qualitative character fully inextricable from the content.
Paul 00:41:42 Does that make sense? And do you want to answer,
Anil 00:41:46 I’ll try to answer though. I think this is more of an extended discussion than a simple question, because it turns on what precisely is meant by phenomenal character and, and content. So I think I’ll make up an example. It might not be the kind of thing that Megan has in mind, but vision has a phenomenal character. It has a spatial organization, objects have the property of objects that they seem to have locations in space in three-dimensional, um, volume, metric extensions. And then within that, against that background, we have specific experiences. I see a cup, I see a laptop computer. I see, uh, another house across the street. So are these the same thing? No, they’re not the same thing. Um, but they are to use Megan’s words. I would think inextricably intertwined, the reason I can experience an object as having the property of object, goodness, as having volumetric extension as having a back, even though I can’t directly see the bang that to me does rest on the phenomenal character of visual experiences in general, that they have this spatiality and volume metric extension in a way that let’s say emotional experiences do not.
Anil 00:43:04 So this is really off the top of my head. Now I have to say, but thinking about it, it seems that the two aspects of conscious experiences co determined each other. Uh, and I wonder whether there’s any, if you subtract out all the, all the possible contents, do you have anything left? Is there some sort of rule, phenomenal character to a modality? I think actually probably, probably not probably there’s that there’s some, even the experience of nothing is a sort of specific content against that kind of phenomenal background. But, but here I am, I am speculating. And, um, this is something, yeah. Megan, we should definitely talk about more. It’s very interesting question.
Paul 00:43:44 Yeah. All right. More in public. Okay. Do you want to do the drum roll or should I, here comes here comes
Anil 00:43:50 The last question before from Megan pieces
Megan Peters 00:43:53 Regarding the measure of levels of complexity, would you say it is a measure of conscious level only, or is it a measure of the complexity or richness of a conscious experience that can be held in awareness if that awareness exists to begin with?
Anil 00:44:13 That’s another very good question on the Meghan piece is show. So we have this measure of levels of complexity. This is just to summarize it for, for people. This is, it turns out to be quite a robust measure of brain dynamics that can be used to distinguish between different global conscious states like sleep anesthesia, the vegetative states, and a paper that we did with in collaboration with Imperial college, the psychedelic state too. And what levels of complexity measures very broadly is the diversity of different patterns in a signal. So if you apply it to, let’s say EEG in the brain, the brain’s electrical activity, the higher, the levels of complexity, the more diverse the activity patterns are. And the way it works is it measures how compressible the data are. So it’s the same algorithm roughly that’s used to compress digital photos into JPEG files.
Anil 00:45:12 So if you just have a photo of a featureless blue sky, it’s very easy to compress it because all the pixels are more or less the same. But if you have actually just white noise, visual, snow, you can’t compress it at all because you have to specify every pixel in every place. So the levels of complexity of a visual snow is really high of a featureless blue skies is very low. And the finding is that as you index through these global states of consciousness, the level of, or the levels of complexity goes down when consciousness fades away. Uh, so hold on. If I got that the right way round, yes, it goes down. The brains is less diverse. It becomes more predictable as you, as you lose consciousness, as consciousness is lost, the perhaps surprising thing. And it was surprising to us. This was an exploratory study, was that in the psychedelic state, it goes the other way around.
Anil 00:46:07 So, uh, the brain becomes more diverse, less predictable in the psychedelic state than in the baseline of, of normal waking. Well, certainly when we did this study, this was the first time we’d seen, uh, this metric go above the level of the baseline waking state usually goes below when you lose consciousness in one way or the other. Now having explained that I’ll try and remember what Megan’s good question was, which was something it was like, is it really a metric of the level of consciousness in the sense of this distinction between wakeful awareness and anesthesia and drowsiness and all these things? Or is it indexing the richness of experience, uh, that’s possible in each of these states? I can’t, I’m not sure if I’m recollecting the question quite accurately.
Paul 00:46:57 Would you want, do you want me to play it again,
Megan Peters 00:47:02 Or is it a measure of the complexity or richness of a conscious experience that can be held in awareness? If that awareness exists to begin with? The
Paul 00:47:13 Last part gets me to,
Anil 00:47:16 So getting back to, so with that background about levels of complexity in mind, let’s, let’s return to Megan’s question, which was, is it a measure of the level of consciousness that just what it is this measure of the difference between sleep and wakefulness, or is it a measure of the richness of experiences that can be held in mind when consciousness is already there? I don’t think it’s either of those things really. I mean, th th pools of complexity just is what it is. It’s a measure of the diversity of the neural dynamics now quite what that relates to it’s, it’s a bit of both, it’s, it’s an interesting and open question. Um, it’s clearly not simply a measure of conscious level for the simple point that there probably isn’t just one scale of conscious level. There’s a nice paper by, by Adrian I think just looking at multidimensional conceptions of conscious level, it’s not, it’s probably not as simple as it just being one point on a, on a linear scale.
Anil 00:48:26 Um, there is certainly something about signal diversity levels of complexity that tracks conscious level across many of these dimensions of conscious level. Uh, but I wouldn’t say it’s identical to that thing also because when you lose consciousness, you still, you know, your brain still has some degree of, of dynamic collectivity. There’s still some values of levels of complexity going on. I think the more interesting question is indeed does it index something about the richness of experiences that are possible in indifferent states and here the psychedelic application is informative because I think it’s misleading to say that the psychedelic state is a higher state of consciousness. Although, you know, newspapers were tempted to report the finding that way. If you took levels of complexity as just a measure of conscious level, that’s the conclusion you would draw, it goes up. So psychic psychedelic state is a highest state of consciousness and sleep is a lower one, but I think it’s spot on speed or something.
Anil 00:49:27 Well, I think very good questions. And I think, um, I think Daniel bore is beginning to do some work on, on looking at levels of complexity across a range of different conditions now. Uh, but it seems to me a more honest description of the psychedelic state. Not that it’s a highest state on some single scale, but that it’s an experientially, more diverse state, less constraints state. There are other theories that suggest something like this, the, uh, the Rebus theory of Robin Carhart, Harrison, Cal forest, and relaxed beliefs, um, sort of relaxed priors and a predictive processing framework that, so I think there’s at least a, a, a fairly informal way where you can say that the increased diversity or lower predictability of brain dynamics in the psychedelic state goes along with the somewhat freewheeling nature of perceptual experience. And you go to mixture of experiences of self and the world that often characterize it’s a psychedelic state.
Anil 00:50:33 That’s probably a better way to think of it, but the levels of complexity measure itself, I think, has to be recognized that it’s a pretty brute force measure. I mean, it’s not that sophisticated, and this is a real challenge in, in developing measures like this. It tends to be that deliberately oversimplified metrics like levels of complexity have empirical traction, but as you make a measure, more sort of theoretically sophisticated, like one of the earliest things that I remember doing was this a measure of causal density, which is supposed to be a more principled measure of brain complexity than the samples of complexity. So a little bit in common with measures of integrated information in the sense that it’s, it’s low, both for completely random and for completely welded processes. Uh, but these more sophisticated measures just tend to not perform very reliably on empirical data. You know, you get numbers, but they’re just sort of all over the place often. And they’re very sensitive to the small differences in the data. And so they’re not yet that useful. I think that’s the real challenge is developing measures. That’s have the empirical traction of simple metrics like , but half, uh, the rest on more interesting, deeper theoretical principles that that will allow us to answer Megan’s question and then much more satisfactory way.
Paul 00:52:01 And this goes back to the, you know, the brain is a complex system and, uh, you know, whether it’s going to feel intuitively satisfying when we have a good enough explanation of consciousness, because, um, I mean, you know, modern science is still grappling with complexity, even though you’re talking about measures of complexity, obviously, but, uh, this is within systems where we still don’t know what’s actually important. We don’t know all of the pieces potentially that are important to even test for empirically, right? So, uh, I mean, you’re measuring these processes, but we might not have the whole story of what is important in, in that complex, uh, realm. Now I’m just kind of drooling out and
Anil 00:52:44 You’re absolutely right. And I mean, there’s something completely, right? So that’s, that’s one thing, what’s the, what’s the appropriate granularity. So we have one problem, which is, we don’t yet have the neuro imaging methods to give us both high time and space resolution and global coverage. Even if we did, we don’t really know what’s the appropriate granularity to, to look at brain dynamics or which levels of granularity are useful to look at brain dynamics, we assume sort of, is it neurons or collections of neurons, whatever. Um, it’s, you know, it’s not just like voxels, that’s pretty arbitrary, right? That’s just what an FMRI scanner can can resolve. Um, and also the there’s just, I think, a need for exploring different kinds of measures that can characterize the behavior of complex systems in general. And this goes back to that, I think a theme that we’ve had throughout this conversation, which is, is mapping between mechanism and experience going to be enough, is it going to be intuitivity satisfying?
Anil 00:53:48 And one thing that speaks to that is, well, how do we construct these mappings? If we just develop correlations, like say there’s a correlation between you being conscious of a house and these brain areas being active, this is never going to be that either intuitively satisfying or that explanatorily powerful, right? It’s a correlation, but we, although this identifying of neural correlates is important, it’s just the starting point. And there are much more sophisticated ways to build bridges between mechanism and experience, quality spaces, different kinds of prediction, uh, complexity measures. And one of the things that, uh, with colleagues, Sussex and London and Cambridge now that, that I’m particularly interested in is this whole bugbear of emergence. How do we characterize emergents in complex systems and what might, uh, and as a well-grounded, but an applicable measure of emergence, tell us about consciousness. People often want to stay well clear this cause there’s a lot of rubbish spoken when people bring the concept of emergence together with consciousness.
Anil 00:55:00 Generally again, thinking that the two things are quite spooky, so they probably have to be related. Uh, but there are sensible ways. Think about emergence too. There are sensible ways in which the brain acts as a whole, yet it contains many distributed parts and there are sensible ways that a flock of birds looks to be like a flock of birds, even though it’s not identical to the behavior of any individual bird coming up with a sort of right set of mathematical tools to identify and characterize emergence. It’s not going to solve the problem with consciousness, whether it’s the hard problem or the real problem, but it will give us a different perspective, a different way of relating mechanism to experience. And I don’t know what that will provide ultimately, but it might again get to Megan’s question about what are these measures telling us about phenomenology,
Paul 00:55:54 Uh, nearly eight. You, uh, actually, we’ve talked about psychedelics a little bit. You actually write in the book about, of course your research on psychedelics, but then your own personal experience. I don’t know if you’re revisiting or if this was a first time experience for you. I remember trying to explain to my father driving through the Backwoods of Texas, what an, uh, an acid trip felt like. Uh, of course I was 17 or so. Um, and you know, I haven’t done psychedelics in a long time. Do I need to, uh, revisit this as an adult, as a, as an old man? Like I am,
Anil 00:56:26 Well, it’s almost, Metzinger once said on another podcast, I think with Sam Harris, that there, there are serious consciousness scientists, and there are non-serious consciousness, scientists, some, something like that. And so you make it that what you will, my description in the book was, was from a relatively recent one. I grew up in south Oxfordshire in the UK. And I dunno, I mean, my, my, with hindsight, fairly sheltered upbringing that didn’t at least to provide any opportunities that I found for, uh, experimenting with psychedelics at the time. So yeah, when I, uh, decided to try them, it was from the perspective of someone who was already very interested in consciousness it’s brain basis. And what psychedelics might, might tell us about that?
Paul 00:57:17 Yeah. Okay. I’ll leave it to the reader to read your descriptions in the book, but it sounded like a, um, a good experience. I had, uh, a couple, a couple unexplained, uh, unpleasant experiences, um, very vivid, vividly unpleasant, but, um, that might speak to my age then and insecurities and who knows?
Anil 00:57:35 Well, I think this is, I do want to be careful about this because I think there’s quite a lot of boosterism around psychedelics at the moment as well, but it’s for it’s, uh, you know, th there, there is a huge amount of clinical potential, and I’m, I’m fully behind efforts to conduct clinical research into that, into the merits of the potential for, for clinical treatments of, especially in the domain of psychedelic assisted psychotherapy. Uh, but at the same time, they’re not a panacea. They’re not a magic bullet. People do and can have, uh, adverse experiences. And that’s not simply a matter of age that’s sometimes they just happen. It’s a matter of sentence setting as well. So I’m, I’m slightly concerned that the pendulum is swinging too far in the other direction, but there’s no doubt in my mind that they are not only potentially clinically very, very useful, but as a tool for consciousness science, they have extraordinary value because you can go in, you can make a very simple pharmacological manipulation that we know, uh, pretty well at the low level.
Anil 00:58:39 What happens, we know which receptors are affected. We know where these receptors are roughly in the brain, this serotonin, uh, five HT, two, a receptors. And then we get these very reliable and very dramatic changes in the nature of conscious experience. And that does two things. It, firstly tells us that the space of possible experiences that we as humans can have is probably larger than we would realize without that. And it also just opens up this opportunity to say, okay, you change, you change one very low level thing about the brain experience changes in these dramatic ways. So what’s going on in the middle, how are the global patterns, brain dynamics changing that, explain why psychedelics have the effect that they have. And I think that’s a really important intervention and assistant is always a very valuable thing you can do when you try and explain its workings.
Paul 00:59:35 So as I was going to start writing notes, Dr. Neil, Seth suggests I get a new drug dealer, but no, that’s not what you’re. That is
Anil 00:59:43 Definitely not what I am saying. Okay.
Paul 00:59:46 So, um, before we move on, because I want to talk about freewill and hopefully we’ll have time for a few other topics, but, uh, another thing that you read about in the book in a S somewhat related, I think actually it’s physically near, um, the psychedelics are out of body experiences. Um, what you don’t write about. Uh, but I’m going to ask about, because at some point, YouTube, uh, decided that I really liked hearing people’s vivid and, um, to them extremely convincing near death experience accounts where there’s this commonality, right, where they’re, you know, in the tunnel, they meet the, usually the guy who brings them, you know, th th they, they feel a strong, strong, spiritual presence with them that will guide them through. And they feel like they’re there for a lifetimes. Uh, then they have to decide whether they’re going to stay, you know, there’s this very common narrative, uh, that seems to go along with near death experiences. And I’m just curious if you have thoughts, whether you know, any of those narratives or not, but, uh, of what might be going on there.
Anil 01:00:50 You’re right. I didn’t, I didn’t write about it. It’s a bit of a can of worms, but I just, actually, it brings up a common thread with a psychedelics, which is the danger of taking things as they seem as a reliable guide to how things are. So you, when I experienced psychedelics for me, this was so dramatic validation of a broadly materialist picture of the brain. Now you change the brain and your experience changes what could be more consonant with a materialist picture than that. Uh, but it turns out in a study by, by Chris Timmerman at, at, at, in London at Imperial, that when he did a large survey of people about how psychedelic experiences had affected their metaphysical beliefs about consciousness, for most people, it reinforced a more immaterial or dualistic perspective that, oh, consciousness is must, must, can’t be just what’s going on in my brain because it changed so much.
Anil 01:01:49 And it was a sort of filter to a wider open the filter to a wider universe. And I found this very interesting, but I must admit slightly deflating. I thought, okay. Th th th I thought it would go the other way, but that was me just generalizing my own take on it. Um, and a similar thing plays out with, out of body experiences and near death experiences as well. So if you take a sort of realist view of the experiences you’re having, then you reach quite metaphysically, radical conclusions. Like I experienced that my first person perspective is now somewhere other than in my head between my eyes is somewhere on the ceiling or it’s nowhere at all. Um, if you take how things seem as how they are conclusion, well, the soul or my conscious the essence of my consciousness can in fact, leave the body similar story with near death experiences, right?
Anil 01:02:48 Do you seem to be the point of, of ceasing to exist yet? All these, all these things are happening. Natural conclusion is that there’s something that persists beyond death of the body. There’s some entering into a different realm, something like that, but these explanations, we should take people’s descriptions of their experiences very seriously. And it’s very interesting that there are all these commonalities in things like near death experiences like this, this sort of tunneling of the vision, but that doesn’t mean that these experiences are direct reflections of what’s actually happening in the universe. So there are very good reasons that for instance, to think of the tunneling of vision as well, that’s different parts of the visual cortex, just shutting down according to pretty reliable patterns of blood flow. Now you can sort of see, well, as, as blood flow declines, you’re going to get peripheral vision falling away.
Anil 01:03:43 Um, now I don’t know if that’s demonstratively tree, but it’s certainly for me a more plausible experience than you are literally entering a ton of light to it, to another realm. And the same goes for out of body experiences. For me, instead of demonstrating the reality that the soul can leave the body. It tells us a much more interesting story, which is that the first person perspective is not to be taken for granted. It’s a construction it’s part of the act of perception, and this fits along with plenty of other evidence that okay, you can stimulate parts of the brain as will depend fuel did back in the 1940s or fifties, or whenever it was. And transit induce an ounce of body experience, uh, tie it to the brain. Now that’s would be strange to surmise that stimulating, um, part of the brain causes the soul to temporarily leave the body, but it makes much more sense if you think of what you’ve disrupted, the circuitry that is deciding that is inferring, where in space, the first person perspective is.
Anil 01:04:45 So I love all these, all these examples, but I think we just have to be careful about, uh, on the one hand respecting people’s descriptions and also respecting what it means for them. You know, it’s no good. Somebody had a near death experience is going to be one of the more meaningful experiences of their life. And it’s just not right to go and say, oh, by the way, no, that’s just your visual cortex shutting down means nothing. Now that’s not, that’s not helping anybody, but neither should we take them at face value. There’s no reason to take that explanations for what’s going on as really what is actually going on.
Paul 01:05:24 I mean, one of the compelling things about, um, a high number of these stories is that they have felt that it is didn’t, wasn’t, didn’t only feel real. It was the most real thing they’ve ever felt and that after they have lost all fear of death, because they have had this experience and it has really, you know, changed their life, I would say in a good way, I’d love to not fear death. Uh, it’s not really the afterlife. I fear it’s the suffocation while I’m drowning, you know, the, the panic of my last moments, right. That’s what I actually fear, but, um, but it seemed to have had, you know, a benefit for a lot of people. So I just, I find it interesting. Okay. I, Neil. So, uh, let’s talk about another interesting topic that has been a thorny issue in the history of philosophy and continues to be a thorny issue, uh, freewill to which you devote a chapter in the book. And instead of asking you a particular question about freewill, I’m going to just let you summarize, uh, your position and account of free will, if you will.
Anil 01:06:30 I will. And the first thing I want to say about it goes right back to the beginning of our conversation, because free will is not something I’ve written about in any scientific papers of mine. So, but I thought I couldn’t, I thought about it and I couldn’t have a book about the neuroscience of consciousness and self write about discussing free. Well, so it was writing this chapter was probably the most challenging chapter to write, but also again, the most rewarding cause I was really figuring out what I thought about it and how justifiable those, those thoughts were. So what are these thoughts? Well, discussions about will get derailed in so many ways. And whenever I give public talks about consciousness, when free will comes up in the Q and a it’s often the sign that, okay, we’re done, nothing else is going to come up now where we’re on the FreeWheel roller coaster and we’re not getting off.
Anil 01:07:23 Uh, and it could be a number of reasons for this. I think of all the aspects of selfhood that we cling to, and that we have some just deep sets resistance to that being explained in terms of science free, where there’s probably, uh, the most clingy. Like it’s okay. If I tell someone your experience of the world is visual experience is a construction like, oh, okay. But if you, if you make the claim that no, you’ve, there is a real sense in which you don’t have the freewill, you might think you have, that can be very personally disruptive, but that’s not a good reason for there not being a good scientific mechanistic explanation of free will. And so what is that? Well, I think very much in tune with the rest of the book, the right way to understand free will is as a kind of perception, right?
Anil 01:08:22 Free, well experiences of freewill arise in the field of consciousness, just as other experiences arise. Um, and the common lens that I applied to all of these experiences is a kind of perceptual inference. So what does that mean when it comes to free will? Well, firstly it means that it’s not consistent with this idea of, of a libertarian or spooky freewill, you know, this idea that there is in fact some way in which an experience of freewill in virtue of the kind of experience that it is swoops in and makes things happen that otherwise wouldn’t happen to me, that idea just makes no sense. Anyway, that disrupts the co causal closure of the universe. It requires conscious experiences to have this spooky, uh, causal power and some kind of very savvy causal power as well. That makes sure you do the thing that intervenes in the brain and exactly the right way to make stuff happen.
Anil 01:09:21 Um, it’s, it’s just not the kind of will that we should be wanting to, to preserve. It’s got a lot of echoes of dualism of an immaterial mind pulling strings in a material brain and body. And once you get rid of wanting to preserve that, then one source of debate in the, in the whole area just falls away, which is this debate between determinism and in determinism in the universe. Like, does it matter if the world is completely deterministic or if there’s a bit of chance here and that, no, it doesn’t matter at all. It’s a complete red herring. Why should it matter? Because the only reason you might want a little bit of indeterminacy is so that’s where this spooky free will can come in and, and change the course of events, get a bit of elbow room for, for something spooky to him to make its play.
Anil 01:10:12 I don’t think we need that, right, but you don’t want to throw the baby out with the bath water. Uh, we have experiences of free will and they mean something. We as organisms also make voluntary behaviors. There are some things that we do the causes of which are relativity immediately found in our environment and the world around us. Typical example, you put your hand on a hot stove and you recoil, there’s no experience of freewill that goes along with that reflexive, uh, um, recoil of the heart of the arm. That’s involuntary though. You said, did you say, sorry, that’s involuntary. Now that would be an example of an involuntary reflex, right? Contrast that with a voluntary action, like picking up this cup of tea and having a sip from it. Um, I decided to do that and it felt like I decided to do that.
Anil 01:11:09 There was a feeling of intention for me picking up that cup and a feeling of agency, um, that accompany the cup arriving at my mouth and me taking a sip from it. These are the sorts of things that characterize the feet. The experience of free will. In fact, I think there are three things. There’s the experience that an action comes from within there’s the experience that the action is aligned with my beliefs and desires know I want, I wanted a sip of tea just then. And there’s a feeling that I could have done. Otherwise, this is the really tricky one. There’s, there’s the feeling that I might not have picked that cup up, or I might have picked something else that seems to me that those are the three characteristics of, of experiences of freewill. And I prefer to think of those characteristics in much the same way that I think of the perceptual experience of something like color.
Anil 01:12:00 So we know color seems to exist in the real world. If I look out of the window, now I can see a gray sky. It’s not really a color, but it is bright. And so, um, and it really seems to be that color, right? That the color seems to exist as an objective mind, independent property of the world, but we know that’s not true. We know, and don’t need neuroscience, Newton, Suzanne, all tellers that colors are constructed by the brain. Um, I think the same thing goes for our experiences of freewill, right? An experience of freewill has this metaphysically subversive content that’s it has causal power over events. So just as red things really seem to be read the experience of, of a 3d world action is that that experience somehow had causal agency in that action. Now, redness doesn’t really exist in the world, but it’s a very useful thing for the brain to construct in the same exact way.
Anil 01:13:05 These experiences of freewill don’t really have the causal power that they seem to have, but they also very useful for the brain in very specific ways. And this chapter tries to tell the story of why that, why that, so, and this is building on work by people like Patrick Haggard and Mike shadowing and, and others, um, that I think a good way to think about why we experienced voluntary actions as freely weld is so the brain can learn about what happened after then, uh, and learn about their consequences. So the brain, the organism might do things differently. The next time you can’t replay the same tape and get a different outcome. But the organism, I might sit down at the desk this time tomorrow and do another podcast. The universe will have changed. My brain would have changed from talking to you today. So if the experience of drinking tea went badly today, then I might have a glass of water tomorrow.
Anil 01:13:59 That’s a useful thing for the organism to have picked up on. And the way the organism can track those regularities in the world is by sort of labeling these voluntary actions with a particular kind of character. And that character is this character of counterfactual, reality of internal origin out of alignment with, with beliefs and desires. So there is free will in the sense that we have voluntary actions and there are lawfully and meaningfully associated with particular kinds of experiences, but we don’t have the spooky kind of freewill that just leaps into the brain from another dimension.
Paul 01:14:38 Do you worry though, or have you had any feedback from people who might reply that you, so even in the book you’ll use freewill in quotations because you know, like, like we were talking about earlier in the science of consciousness, we have to reframe how we think about these things conceptually, to actually get a grasp on them, but one could respond to that. What I actually do care about is this, uh, what you’re calling the spooky stuff. I, um, and, and by describing the phenomenology of it and our perception of the freewill, which might be satisfying causally, uh, it doesn’t satisfy my need and desire to feel at the helm of my own voluntary actions. When what your, um, what, you know, what you’re describing is my perception of a voluntary action, but I wouldn’t call that a voluntary action. I would say I wasn’t in control causally. Right.
Anil 01:15:35 Right. And I think this is the key point, right? Here’s a situation where I do think it’s a very good example of how the mystery that we started with productivity changes for me, this is a satisfying account of freewill. Uh, it does everything it should do. And it makes sense to me as well, your example of somebody who might respond well, that doesn’t feel right to me, that it doesn’t feel like I’m in that kind of field. Doesn’t explain the feeling that I am in control. Well, there’s another, um, another issue that slipped in there, which is this idea that there is an eye that is in control. And of course, part of our conversation has been that the self isn’t, this unique thing that, that sits behind the windows of the eyes pairing out, and that decides what to do. And then contracts various muscles.
Anil 01:16:23 The self is a kind of perception, whether it’s this perception of the body as a living organism, whether it’s emotion, mood, first person perspective, all of these things are aspects of perception too. So the experience of free will is not something that, uh, a self has on users in some way. It’s just part of what the experience of being a self is. And again, there are clinical examples which show us that this aspect of self, it can go away too. There’s this condition of a kinetic mutism where people specifically seem to lose the experience and ability to engage in voluntary behavior. But other aspects of their self might, might remain intact. For me, it’s a very satisfying way to think of it. And in fact, it also doesn’t, and here’s, here’s another really important point. It’s not, this just leads me into an apathetic life that I think, okay.
Anil 01:17:20 Right. If free will, is in fact, a perception of voluntary action, that’s mainly useful for the future, then yeah, screw it. I don’t actually have the ability to behave in the world as I, as I, as I thought I did. No, of course, I still see red when I look out the window and see surfaces with particular kinds of reflectance is I will still experience free, will in the same way. And that is intrinsically coupled to my voluntary behavior for the reasons we’ve just been talking about. So it changes everything, but it also leaves the essential things completely unchanged that I still go about my business in the same way as before
Paul 01:17:59 Looping back real quick to psychedelics does the, because it’s like a deluxe are often associated with the disillusion of the ego. Uh, do you think that there is a connection and I, I have no idea about this, that connection, um, between someone who has experienced a dissolution of the ego through something like psychedelics and the acceptance of this account of free will as, um, as satisfying
Anil 01:18:25 You’d like to thank. So, I mean, again, for me, it’s incredibly, it’s incredibly compatible. Like the, this sort of ego dissolution that goes along with, with psychedelics is completely in line with thinking about freewill as this kind of perception of, of voluntary action oriented to the future. Um, but I rather worry that, uh, just as Chris Tillman and the study found that people who’ve taken psychedelics generally move away from a materialistic belief, they may also move away from the way I think about freewill as well. It depends on your starting point. It depends on where you’re coming from, what the psychedelic experience is going to do to your beliefs about these things.
Paul 01:19:07 All right. Uh, I’m aware of our time here. And, uh, one of the things that you talk about in your book that I, I guess, because I’ve been thinking a lot about the relation between life and intelligence, and of course your book is about consciousness and life, and you write about how intelligence and consciousness, uh, are not necessarily orthogonal, but they aren’t along the same axes. So I wondered, um, I, I kind of want to throw all these three in the bag, uh, because I’ve come to appreciate intelligence. And I think the success of deep learning and computational neuroscience approaches, uh, to mechanistic, computational accounts of how minds and brains work has made me, um, ironically appreciate other life processes, which is interesting because you, this is, you know, the, the focus of your, uh, introspective inference, where there are PR there’s predictive processing exactly on these life processes and, uh, needing to control, uh, the processes to stay alive. So I’m sorry. That was a huge mouthful, but I’m wondering if you can explain your view on the relationship of life consciousness and intelligence.
Anil 01:20:25 Definitely. Just to clarify briefly, I’ve talk about interoceptive inferences rather than introspective inferences. They’re sort of, it’s often confused, but just to make the intraception being about perceptions of the body, introspection, thinking about your own thoughts very broadly. So life consciousness and intelligence, it is a, it is a big bag. It’s a mixed bag. And the theme of the book really has led me to recognize these deep connections between consciousness, especially conscious self and life and, and the, the claim, the primary claim being that we perceive ourselves and the world around us with through, and because of our living bodies, that all the predictive machinery that underpins all our experiences operates in, in light of this primary biological imperative to stay alive. Another reason the life thing is critical here is because unlike a computer where you have relatively sharp distinction between hardware and software, and if you use a computational metaphor for the, for the mind, you tend to think of the brain is the hardware and the mind is the software.
Anil 01:21:37 And maybe if you write software in the right way, does the right kind of information processing, which is a slippery return, uh, then consciousness will, would arise. Um, if you think about living systems there, isn’t such a sharp distinction between hardware and software or mind where in wetware, uh, it’s kind of just the hierarchical dependencies all the way down. I mean, this goes right back into literature that inspired me many years ago by, um, but in my Tirana about Alto pollicis and the, the sort of the way in which cells, uh, construct the components, their own components over time, they, they self generating processes. Um, and so this, just this recognition that there is no clear line between mind where, and wetware makes me very suspicious of the idea that consciousness is substrate independent, that it could be easily run on a different thing, because where does the substrate start and stop in a biological system?
Anil 01:22:44 And suspicious of the idea that consciousness is simply a matter of information processing, because that tends to go along with substrate independence, um, as well, but what goes with it. But if you do think about consciousness in terms of information processing or something that might potentially be run on a, on a, with the right kind of software that tends to get groups with intelligence and this, this sort of often unstated assumption that a sufficiently intelligent computer will become conscious that consciousness is a function, maybe a complicated function, but a function of intelligence in a substrate independent way. And I just think this is, this is based on also, it might be right. Like, I cannot say that it’s not right, but I just think it’s based on a lot of questionable assumptions. There is for me, no good reason to think that consciousness is substrate independent and at least one interesting reason to think that it might not be this lack of a clear, uh, label for where the division between mind, where and wetware, and then why would we even think that consciousness is related so intimately to intelligence?
Anil 01:23:58 And here I worry that, that we have a spec serve another kind of human exceptionalism that we think we’re intelligent, which is a bit questionable these days, but we’re certainly smart in some objective ways compared to other species. What we do with that smartness is the questionable part. Um, we think we’re smart. We know we’re conscious, so the two must go together. And this is a really dubious assumption because it leads us, uh, maybe to overestimate the possibility of building machines that are conscious by the way. I don’t think we should even be trying to do that. Um, and it may lead us to underestimate,
Paul 01:24:39 I don’t think we’re in any danger of, uh, creating conscious machines through building
Anil 01:24:44 AI. No, I don’t think we are time soon. That’s right. But, but I, I subscribed to the thing, the view here, that, for something that could be so ethically cataclysmic, um, even the tiny possibility that we might succeed is worth a little bit of worry. Part of our worry budget should be devoted that way. Not a massive amount, but part of it, but I think just more, more broadly, it seems to be often treated as this just, oh, that’d be cool. Let’s just do it. You know, even if we can’t do it. So the skepticism is often about like, oh, but you can’t know you’re not going to succeed, but I think there’s a deeper reason to question the motivations and the goals, even if they’re not achievable, because we don’t know what it would take to build an actually conscious machine, but we also don’t know what it would not take.
Anil 01:25:33 We might do it by accident without realizing here, I’m actually worried about things like brain organoids, um, brain organoids, or these brain like structures of increasing complexity that are grown at scale in lamps for good reasons, for good medical reasons. Um, but these are made out of neurons. So the whole question of substrate dependence or independence goes away. And so the possibility of, uh, organoid consciousness I think is much more concerning than consciousness, suddenly shimmering into existence in my next generation laptop. Uh, so we’ve, we’ve kind of rummaged around in this bad, quiet, quite quite a bit here, but I do think it’s worth with separating the tight bounds in my mind, at least between consciousness and life and the looser bounds between consciousnesses intelligence. Now you probably have to have a minimal degree of intelligence to have a conscious experience. And, um, quite what that minimal degree is.
Anil 01:26:32 It’s hard to say. I mean, intelligence is a wooly concepts, roughly doing the right thing at the right time. Uh, and being intelligent certainly gives an organism, a richer space of possible conscious experiences. As simple, relatively dumb organism might be able to experience the difference between sadness and happiness or suffering and Joyson, but, but we humans can, can experience sadness in all sorts of ways in regrets and anticipated. We regret in all these, all these things that depend on thinking about possible futures and possible counterfactual pasts. So the specific nature of conscious experience is very much tied to the kinds of thoughts and intelligent competences that we as humans have. But at root, I think it’s much more closely tied up with life,
Paul 01:27:25 Given our conversation today. So, you know, I didn’t know we’re going to talk about cerebral organoids at all, but as you were speaking, I was thinking, let’s go speculative to, to at the very last moment here, I was thinking that the kind of consciousness, the quality, the phenomenal experience of a cerebral organoid would be much more alien. I to my brain than let’s say a Chimp or some someone, you know, I would imagine that a Chimp would have a much more similar, phenomenal experience to me than a cerebral organoid made from my own neurons, right from let’s say we, we, uh, use a sample of my neurons and grow an organoid. It wouldn’t be like me at hall.
Anil 01:28:10 Yeah, that’s right. And firstly, I, I, I also think that the prospect of, of building a conscious organoid is, is very remote, but I think it’s much less remote than the prospect of building a conscious laptop. Um, one of the big questions about potentially conscious organoids, and again, I don’t think we should be setting out to build these things, um, is whether a history of, of interaction with an environment matters. Uh, so we organisms brains, non organoid brains, invariably have an evolutionary history that involved bodies and a developmental history that involves bodies, senses interaction with the environment as well. It’s pretty clear that we don’t need that interaction in the moment to have conscious experiences, can we can be dreaming and be basically cut off at least from the external environment, but we may need a history of that interaction in order to provide any determinant, uh, conscious content.
Anil 01:29:12 Uh, and so we, I think we wrote something along these lines in a paper, I had a paper with a tin Bain and my child had massive meany in trends in neurosciences in 2020 called islands of awareness. And it was discussing candidate situations where we might then might be consciousness completely cut off from a body in an environment. And an organoid was, was one of these cases. And I think we wrote there that we might be able in some, some future to tell, uh, whether an organoid is conscious, but have no idea what it is conscious of at all. And I think that that could be one way. It goes on the other hand, people building organoids, designing organoids these days are equipping them with sensors with, with actuators too. So you have now the possibility of, of organoids that can interact with the environment, but do they do so, do they have a body that they maintain in the state of being alive? And, and that’s, that’s a whole other question then you’re not really talking about an organoid you’re talking about in synthetic creature with a brain, with a synthetic brain, very different thing.
Paul 01:30:21 Uh, Neil, we w we went through a pretty good chunk of the book, but, you know, there was a ton more that we didn’t get to cover. And I hope that people read the book. Um, if not just for the pleasure of reading it, because it is easy on the eyes and mind as you read it. And of course, uh, has tons of good ideas and descriptions in there. So this has been a joy for me. Thank you for being here and good luck with the book.
Anil 01:30:43 Ah, thank you, Paul. It’s been a terrific ramble through various landscapes of the Birkin. Very much enjoyed the conversation. And thank you, Megan, for lots of questions to
0:00 – Intro
6:32 – Megan Peters Q: Communicating Consciousness
15:58 – Human vs. animal consciousness
19:12 – BEING YOU A New Science of Consciousness
20:55 – Megan Peters Q: Will the hard problem go away?
30:55 – Steve Fleming Q: Contents of consciousness
41:01 – Megan Peters Q: Phenomenal character vs. content
43:46 – Megan Peters Q: Lempels of complexity
52:00 – Complex systems and emergence
55:53 – Psychedelics
1:06:04 – Free will
1:19:10 – Consciousness vs. life vs. intelligence
Brian and I discuss a range of topics related to his latest book, The Alignment Problem: Machine Learning and Human Values. The alignment problem...
Businessportrait Francisco Webber Cortical.ioThe white paper we discuss: Semantic Folding Theory And its Application in Semantic Fingerprinting. A nice talk Francisco gave: Semantic fingerprinting:...
Support the show to get full episodes and join the Discord community. The Transmitter is an online publication that aims to deliver useful information,...