Can we measure consciousness?

[Music] Welcome to Many Mines, a podcast about our world's spectacular varieties of mind and what we can learn from them. We feature conversations that shed light on the diverse ways of thinking, sensing, and being that are out there. On your host, Kenzie Cooper-Rider, thanks for joining us. [Music] Cluster of brain cells in a dish pulsing with electrical activity, a bee buzzing its way through a garden in bloom, a newborn baby staring up into his mother's eyes. What all these entities have in common is that we don't quite know what it's like to be them, or really whether it's like anything at all. We don't really know, in other words, whether they're conscious, but maybe we could know if only we developed the right test. My guest today is Dr. Tim Bane. Tim is professor of philosophy at Monash University in Melbourne, Australia. He's a philosopher of mind and cognitive science, with a particular interest in the nature of consciousness. Along with a large team of co-authors, Tim recently published an article titled "Tests for Consciousness in Humans and Beyond." In it, they review the current landscape of consciousness tests, or C-tests as they call them, and outline strategies for building more and better tests down the road. Here, Tim and I discuss what consciousness is and why theories of it seem to be proliferating. We consider several of the boundary cases that are most hotly debated right now in the field, cases like brain organoids, neonates, and split brain patients. We sketch a few of the most prominent current consciousness tests, the command following test, the SIFT test, the unlimited associate of learning test, and the test for AI consciousness. We talk about how we might be able to inch our way slowly towards something like a thermometer for consciousness, a universal test that tells us whether an entity is conscious or to what degree, or even what kind of conscious it is. Along the way, Tim and I talk about zombies, chatbots, brains and vats, and islands of awareness. And we meet us about how, in certain respects, consciousness is like temperature, or perhaps more like happiness, or wealth, or intelligence, and maybe even a bit like fire. I think you'll enjoy this one, friends. It's a thought-provoking conversation on a foundational topic, and one that takes us far and wide. So without further ado, here's my interview with Dr. Tim Bain. Enjoy. Tim Bain. Welcome to many minds. Thanks very much, Kenji. Consciousness is obviously a fundamental topic in the study of the mind. It's evergreen topic. There are a ton of active researchers in this area. There's kind of a constant churn of new books and articles, both popular and academic. But at the same time, in my experience anyway, there's kind of a sizable number of folks who are sort of allergic to the topic, like they just kind of want to put it to one side and not really engage with the issue of consciousness. What do you think about this kind of ambivalence? Why do you think there's this kind of ambivalence in the field around consciousness? Well, I fully understand it. I feel the ambivalence myself. You know, it's my day job, but sometimes I wonder whether I wouldn't be happier being a plumber or cutting down trees or the like. It's like, I crash you. You can't stop looking at it, but it's not particularly edifying sometimes. I mean, I think the ambivalence comes from, look, it's fundamental to who we are. And if you've ever thought about the mind as a kid, you've probably wondered about consciousness. You know, I remember being at a zoo when I was 9 or 10. You know, you see the crocodile open its eye and stare at you and you wonder whether it's thinking about you and the way that you're thinking about it. You wonder whether it's experiencing anything. So the fascination, I think, is clear and it's a perennial topic. Any culture that's done philosophy is thought about experience, a culture that's thought about mind, but it's hard to study scientifically. The subjectivity of it is what's challenging. And there's a reason I think why there was this heyday of scientific interest in consciousness at the end of the 19th century. And then the behaviorist said to help with it, it's not a scientific, it's not the proper object of scientific study. You can't, you can't pin it down and you can't really subject it to proper scientific rigor. And you know, people now think you can, but exactly how to go about studying it is something that the field's constantly wrestling with. I think this was very much present in the work of the late Dandana trying to say, well, we can study consciousness scientifically, but you can't do it that way and you can't do it that way. And then of course, people respond by saying, you know, he's changed the topic. So, you know, this, this is part of the problem. You don't go very far without hitting your head against really deep methodological issues. And, and mature sciences, you know, philosophy is still involved in them, but insofar as they mature, you can get on and do the science without having a hotline to a philosopher or without getting yourself into philosophical issues. Consciousness is a bit different because, you know, it's rare that the science proceeds without any philosophical engagement because it's coming up against definitional and methodological and conceptual issues. You get conceptual issues in physics at fundamental physics when you're looking at the nature of space and time and how measurement works. And that's where the philosophers of physics talk to physicists in this real dialogue. The science, the philosophers, they're just not helpful. They're not going to help you, but with consciousness, I think part of the ambivalence comes from it's in this, another world between philosophy and science where you're constantly negotiating, is it conceptual or is it scientific? And maybe this also relates to this point you made about this sort of maturity or lack of maturity of the science of consciousness, but you note in this paper that we're going to focus on today that there are a lot of theories of consciousness that are currently still kind of on the table that are taken seriously. I think you give the number 22 in the paper. There are 22 theories out there. And you also observe that theories of consciousness in fact seem to be proliferating. Is that proliferation also just kind of a hallmark of the immaturity of the study of consciousness, the scientific study of consciousness? Yeah, I think it is. Science goes and waves, it was an early wave. I mean, conscious science became respectable in the 90s. If you've got a noble prize, you can start saying stuff that you couldn't say before you got the prize. And so Francis Crick teamed up with Christophe Koch with this very influential paper saying, "Look, let's talk about consciousness." I think scientists, especially vision scientists, were talking about consciousness prior to that. They were just weren't keen on using the word. And Rick and Koch said, "Yeah, let's study visual experience, let's study consciousness." And then the first, I would say 30 years, that resurgence of interest in scientific and interesting consciousness was dominated by data collection. People are looking for the neural correlates of consciousness. What kind of brain activity correlates with consciousness? And what kind of brain activity do you see when the brain is doing something? But there's no consciousness involved. And then there was a kind of, I think, a general sense that we need theories. And so there was a rushed a theory development. And now there's a proliferation of theories. And it's sometimes difficult to count theories, you know, theories. Fission, and do you have, you know, is the theory really different from that theory? Or everyone wants to put their name. No one wants to say, "I like my friend's theory," or "I like the guy down the road's theory." Right? Here's my theory. You always need to give your theory a new name, there's a lot of branding going on. And I think that's fine. It's good to have a proliferation of new ideas. But it doesn't mean that from the outside, also from the inside, there's certainly a lack of consensus, a lack of theoretical consensus, as to what we're dealing with. And so it appeals to theory problematic. Much in the way if someone wants to say, "Look, I want to ground ethics in religion," you might say, "Well, that's okay. Go ahead." But which religion? Right? There's too many religions. The religious marketplace is a busy one. Similarly, if you want to ground a science of consciousness in theory, which is a perfectly reasonable thing to do, you can't just walk into the shop and say, "Well, here's the theory of consciousness." You can't even say, "Here's the 203 theories." And we know that we don't know which of these theories are right, but we know that one of them is right. I think we're not at that position yet. So yeah, I think maturity is exactly the right word. I think it's a good place to be. And I think it's a great to have lots of theoretical ideas out there. But it wouldn't be appropriate to say, "We now have a mature theory of consciousness," or anything, like an emerging consensus about how to understand human consciousness. So I don't think we're going to be able to go into the different theories of consciousness in any detail, but I will flag in the Shunutsu flag this other paper that you co-authored where you review many of the leading theories in terms of branding. And as you said, you should involve some branding. My favorite theory that I thought sounded the coolest was the beast machine theory. I just mean in terms of the label. Hold that, we'll circle back to that at the end, I suspect. All right, I look forward to hearing more about that. So what was your first entry point into thinking about consciousness? How did you come to this, subfield? I was lucky. I was at the right place at the right time. I was a graduate student in Tucson, Arizona, University of Arizona. I think we just hit peak consciousness science. The first wave was breaking. A lot of great philosophy was being published on consciousness. There was a wonderful conversation between philosophers and the scientists about this thing that everyone had recognized was important, but it'd been too scared to touch. And a young Australian with long hair, but an ever David Chambers was just recruited to the department. And Dave was looking for PhD students. And I was looking for a PhD supervisor. He took me under his wing and I guess in some sense, that was that was the start of my consciousness career. And he's, you know, obviously a Titan in the field, coined the phrase, the hard problem. So if you've heard anything about, you know, if you listen to a single consciousness podcast, you've heard about Chambers talking about the hard problem. So that, yeah, that was that was my entry in. So you didn't have some burning desire to go into the subfield of consciousness. You just were in the, as you said, right place right now. Well, yeah, you know, I think it was both. I mean, I had, I wouldn't have gone had I not found the problems fascinating. Equally, if had I found the problems fascinating, but not being welcomed into the community in the appropriate way, you know, a lot of science and a lot of philosophy is about being in a community of nurturing scholars and being willing to, to learn from them and feeling supported, feeling that what you're doing, doing matters. So I mean, I've always been interested in consciousness, but it took a community to kind of give me the confidence that I had might have something with saying or, you know, there were people to collaborate with. I think that's an important part of my story is feeling, feeling part of a community. Yeah, that's super important feeling like you're talking to people, people want to hear what you have to say. I feel like I think that someone's going to listen, you know, you write these things and they go out into the void. I mean, now with the internet, you get a bit of response, but back in the day, you put a paper in an envelope, but you could go years without hearing from anyone. Right, right. Yeah. So maybe we could start with a little bit of definitional stuff, you know, obviously we could have an entire conversation around definitions, but I just want to kind of get a rough sense of how you define some of these terms. So first of all, how do you define consciousness in this context? So short answer and a long answer. I mean, the short answer is the one that Tom Nagel gave us in the famous paper that he wrote that many of your listeners would have heard of, you know, what is it like to be about? Nagel coined this phrase, well, I actually didn't coin it, but he gave it prominence, you know, consciousness is there being something it's like to be you. So we're recording this, this discussion through a whole lot of technology, you know, my phones involved your computer, my laptop. These were all incredibly complicated devices, very sophisticated. I don't think there's anything that's like to be my laptop. There's no subjective perspective. There's nothing for the laptop that it's like to be in a laptop. That's a definition, if you like, it's not a very useful definition. I mean, it points to something, it's not a definition of scientists can do anything with it. If by a definition, everything depends on what you mean by a definition here. You ask a philosopher to give me a definition of X and they say, would define definition. It's a horrible trick, but I think I think one has to attend to the texture of that term. What are we asking for? Because people say it all the time, you know, define consciousness, especially in the area that we're going to talk about with the test for consciousness. I think typically what people want is something that you can operationalize in the context of a lab. Say, if it performs well on this test, it's conscious by definition. That desire I fully get, I absolutely fully get, but I think it should be resisted because you don't want to start with definitions. Definitions fall out of theoretical understanding at the end of the day, you know, way. I mean, I mentioned Denet before. One of my favorite sayings of Denet's was he talked about the dangers of premature definition. So if you'd asked someone in the 16th century, would define water, they would have said, our water is the wet stuff that falls from the sky. It's, you know, in the lakes and rivers around here. No, that's okay. It's rough and ready that enabled a science of water to get up and running. That's not a really good definition of water. Water is not always wet and what's wet is not always water. And, you know, there can be stuff that falls from the sky that looks like water and tastes like water and it comes out of your shower, but it's not water if it's not H2O. We get to a scientific definition once we've done the science. What we need in order to do the science is something that gives us some kind of security that we're latching on to the same thing. That's what we want. That's the hard thing, yeah. So then you use Negel's definition, the idea that to be conscious is to have some kind of subjective experience to have something it is like to be that entity. So you use that as kind of, almost a sort of placeholder. That's right. So that we can get started on the work of flushing out theories and flushing out tests and that kind of stuff. And that's right. And I think the crucial thing, this is a kind of hard idea to articulate, but you kind of need something between the two. You know, that the Negel thinking of the something that's like to be doesn't give you much to go on when you're in the lab and you're trying to figure out whether a dolphin's conscious or a baby or, you know, someone with brain damage, that just that's what you want to find out. But what do I do? You know, how do I study this thing? And the thing that you get at at the end of the science is great, but you haven't done the science. You're trying to do the science. Right. So you need something in the middle that will serve in the context as a kind of operationalization of the construct. It's a bit like, well, for this thing, I'm going to treat water as the stuff that's falling out of the sky. Or for this project, I'm going to treat water as the stuff that's in my bathtub. And let's see what that's made of. But you want to hold that loosely because you want to be open to the possibility that the thing that you were studying as water in your bathtub isn't actually water. You want to be open to the possibility that the thing that you're studying in the lab that you think is consciousness might not turn out to be consciousness. So in the field more broadly, do people conceptualize consciousness as something that comes in degrees or something that's a little bit more all or nothing? Is it a great question? I mean, I think if I had to guess, there's a traditional assumption that it's that it's kind of all or nothing. And that I think goes along with Naples definition. Is there something it's like to be? And you asked yourself, well, could it be the case that there's kind of something it's like to be, but kind of not? Right. It's kind of hard to wrap your head around that, right? The light might be dim. The light of consciousness might be dim, but the lights either are off or on. I think more recently, there's a bit of a groundswell of people trying to push back and say, look, it's hard to get your head around. Yes, we get that, but we need to take seriously the idea that consciousness might not be all or nothing that there may be in determining it cases. And people will often point to life intuitively. You're either dead or alive, but when you take the microscope at someone's final seconds on life, metabolic processes are breaking down. Is this person that the they're at the threshold of death, they're at that border between life and death, but is absolutely determined at every microsecond that they're either alive or dead, you know, at a certain level of granularity that the concept I could be breaks down. We've got to reckon with indeterminacy, neither fully alive nor fully dead. We could stipulate, but that's just a matter of stipulation. And I suspect that we want to say something similar about consciousness. So how would you position consciousness relative to another fraught term that gets talked about in a number of similar contexts, which is sentience? How do we distinguish those two? Sentience is a term I tend to avoid. I think some people in the field use sentience just as a synonym for consciousness. And awareness would be another term experienced, you know, is this cat sentient? Is this cat aware? Is that kind of conscious? These are just synonyms, and you might, you know, mix it up for startlistic reasons, but you mean the same thing. The problem is that many people use sentience for effectively valanced consciousness, stuff that's got an emotional tonality in particular pain and pleasure. So debates about animal sentience are often either exclusively or at least focused on debates about animal pain, animal suffering, animal pleasure. And there's just half the field goes one way, half the field goes the other, there's subfields, you know. And so because of that, I tend to try just to stick with consciousness. But the whole field is a bit of a terminological nightmare. You've just, there's landmines everywhere. Yeah, but again, it keeps philosophers employed. So, let's, so I thought we could, we're gonna obviously move to talking about the meat of this paper, which is really this idea of consciousness tests and developing them and evaluating them and so on. But just to kind of get people's gears going, get people's intuitions going about why we even need consciousness tests. I thought we could linger a little bit on some boundary cases. So cases where there's argument, there's disagreement about whether or not an entity should be considered conscious. So you've written about a number of these and we'll, we'll link to your papers on these cases. But so I had, I had three picked out. The first one that caught my eye is this work on brain organoids or cerebral organoids, I guess they're sometimes called. You're gonna have to tell us what, what those even are because folks may not be familiar, but what are those and then what is the debate there? So there organisms that are grown in the lab from stem cells and there's lots of different kinds. But they resemble parks or small parts or large parts of brains. And so the organoid, you know, field, there's organoids of lots of different body organs, kidneys, livers, consciousness, sciences, become interested in cerebral organoids, brain organoids. They don't have all the properties of brains. I mean, for one thing, they're just, they're tiny and they're not vascularized so they don't, you know, I don't get the blood flow, but they do have some of the properties of the developing human brain. And there's enough there in terms of spontaneous, you know, rhythms that in one study resembled some of the brain activity that you see in preterm human infants to at least, you know, raise in some kind of serious way. The question of, well, might there be consciousness here? And I think the pressing issue is not so much with a current generation brain organoids might be conscious, but whether some of the technical developments where they become more complex when they become larger where the future developments might potentially lead to consciousness in brain organoid. And then of course, the question is what how would you tell? And one of the crucial things is many of the organoids are isolated from the environment. So they don't take information, they don't have perceptual systems. They're a bit like a brain in the lab thought experiment, except the brain in the lab has never been in anything other than the that, you know, it was created, it came into existence in the that the classic brain and the that kind of scenario is the brain had been in a body and bodies aren't vats. Right? Bodies are taking information in from the environment, they're acting in environments, they're embedded in environments. What's fascinating about organoids is they haven't had this history of embodiment of environmental embeddedness. And yet they're showing some of the signs, you know, spontaneous electrical activity in the neurons. So, you know, my own feeling is I suspect that the first synthetic consciousness that we're going to create isn't going to be AI, you know, it's not going to be the success of chat GPT or anything like that or deep blue. I suspect that the first genuinely synthetic conscious agent we're going to create is going to be something like an organoid that's hooked up to a robot. So it's going to be involved in taking information in and out of the environment. My own feeling is that disembodied brain organoids are not likely candidates for consciousness, but once you've hooked them up to things that they can control and some tentative steps have been taken in this direction, that's probably the first time human beings will have created conscious agents without breeding animals, which of course we do every day of the week. But are there folks in this area who think that no organoids can certainly reach the consciousness threshold, even if they remain disembodied? I struggled to recall anyone who would say that any currently created organoid is likely to be conscious. There are certainly folks who say we should be extremely cautious about this technology and a kind of a precautionary principle ought to kick in maybe even now because for all we know these systems will soon become conscious. Could you say a word about what you meant by a precautionary principle? The idea is when you're not sure about whether a kind of system is conscious and the consequences of creating it, if it were conscious would be serious, then you want to err on the side of caution and avoid creating it. It's very tricky in these issues because you might say well maybe some of these organoids might soon develop a kind of rudimentary consciousness, but there's a reason people scientists want to develop this technology, it enables them to investigate all sorts of neurological conditions that they wouldn't otherwise be able to investigate. You can intervene. I mean that's why all organoids get created in the first place. They're extremely useful from a lab perspective. And so a whole bunch of research that we might want to do is off the table if you impose a blanket down on the development and use of neural organoids. The ethical issues are complicated. Okay, so maybe we can move to another one that you've written about, which is questions around the emergence of consciousness in early infancy. Disconsciousness started in utero, does it start during birth, does it start sometime perinataly or whatever? I'm curious, what is the spectrum of opinion on that question right now or is there a consensus that has emerged on that question? There's a bit of a mini-consensus emerging, I guess I would say, maybe it's just people on reading, maybe it's something to describe it as a consensus. When people have written on this topic in the past, they've typically pushed for a relatively late view of the onset of consciousness. Peter Karellas has argued that consciousness might not be in place until the age of about three. Oh wow. Yeah, I mean he's appealing to a particular view of what consciousness requires and I think that's a very late view. I told my partner about that and that was her response to, oh wow. Most people find that view very counterintuitive, but armed with a certain kind of theoretical perspective. Some of those who have written on the topic previously have gone for a certainly after an early infancy, the age of one, something like that. When thinking that consciousness requires a certain amount of cognitive machinery and the ability not to be aware of the world, but to be aware of your own representations of the world as such. And I think that's very much a minority of you now, but you can find people who are still committed to it. They might not be so keen to articulate it in public, but they're tempted by that thought. And one of the things that's fascinating about that is it runs counter to our ordinary intuitions. You know, I mean anyone who's had a baby, I think, finds it pretty irresistible to describe consciousness to a baby. Although having said that, this is an astonishing fact to many people, it wasn't so long ago that it was routine for neonates a few days, a few weeks old to be operated on without any set it in the UK, in Europe, in America, right? And physicians would say, well, what's the evidence of infant pain? That seems like a good place where the precautionary principal could have been. That's right. That's right. So sorry, I think you said there's a mini consensus building. Well, I think, you know, maybe it's just me and my friends, but I think most people who have contributed to the topic recently would argue for something like three or four months of age. Okay, interesting. But, you know, there is some very interesting evidence to suggest some of the forms of neural activity that is likely responsible or implicated in visual consciousness. Isn't really online till much later, like six or seven or eight months of age. That's kind of interesting. So yeah, there's significant voices would say consciousness is likely in place from birth, maybe a little bit earlier. I wouldn't describe that as a consensus. I would think that consensus is likely to be there in fairly early infancy. Consensus with some pretty large caliets and qualifiers, you know, it's a fascinating topic. Yeah, it's interesting. And we will certainly point to your review article from just in the last couple of years. So another one you've written about is the case of split brain patients. What is being debated there? So this is a debate that goes back in the 70s. When a surgery was conducted earlier than that in the 50s and 60s. So these are patients with really intractable epilepsy, refractory to drugs. And an effort to keep the epileptic seizure from spreading surgeons would cut significant chunks, sometimes pretty much all of the corpus colosum, the bundle of fibres that connects the two hemispheres to parts of the cortex together. So, you know, you get an epileptic seizure in one hemisphere, but because the fibres are being cut, it can't spread. You know, it wasn't huge numbers of patients, but there were a certainly a handful of patients that were studied quite intensively. And there's a sort of a paradox that these folks generated, which is on the one hand, everyday life was not significantly impacted. I mean, there were impacts, but generally speaking, friends, family would still engage with them as a single person, as a single agent that was continuous with the pre-operation individual. People could still hold down jobs. They could still drive. But folks, most famously Michael Gzzanigar, Jerry Levy, Colin Trevathan, a bunch of folks studied some of these patients in the lab. And we're able to show that large amounts of information that was received by the right hemisphere never made it to the left hemisphere and vice versa. So, there's a kind of encapsulation. And there's some reason to think it's controversial, and the debates recently been given a new lease of life from a bunch of folks. But there's some reason to believe that each hemisphere houses its own center of consciousness. That's the upshot of all this. That there's a clearly, there's this little debate that the left hemisphere supports consciousness. Most of the debate is around whether the right hemisphere might be a kind of zombie system that's able to initiate action, but does so unconsciously. That's one view and other view as it's kind of an autonomous, isolated, conscious agent. A third view is that it's actually unified with the consciousness that's supported in the left hemisphere. It just seems as though the two hemispheres are independently conscious, but they're actually not there in some way, a single consciousness. They just seem to be distinct consciousnesses. I see. So, the idea is that in people that have their, you know, their corpus callosa intact, the two hemispheres are integrated in such a way that it forms a unitary single consciousness. But in these cases, there might be a way to think about a kind of splitting of one consciousness into two. That's beautifully put. I mean, you and I seem to be a unified single subject of experience. I mean, we seem to ourselves and to each other. I don't think there's two conscious agents in you, one in your left hemisphere and one in your right hemisphere. So, the unity of consciousness, the singularity of it, the fact that there's a single subject of thought and action, a single center of awareness seems to be pretty deeply part of our ordinary thinking of consciousness. And the split brain syndrome is interesting because it's suggest that that unity might be disrupted by certain types of surgery. One of the reasons this has become a hot topic again is that there's another kind of surgery, often performed on children, with epilepsy, where they will cut the afferent and efferent connections, the input in the output to one hemisphere. Now, sometimes the hot that hemisphere is removed, but sometimes it's just, there were compte medical complications with actually just removing the hemisphere. So, so what's typically done now is surgeons will cut the input and output to the hemisphere, but leave it there. What's fascinating here is we don't have a clear picture of what kind of neural activity is going on in this isolated hemisphere. Some very recent studies suggest that maybe there's a potential for consciousness, other studies have cast out on that. It's all up in the air, but it's fascinating in a paper with Marcello, Massimini and Neil Seth. We talked about these kind of cases as potentially islands of awareness. They're islands of awareness because they're you know, they're like the ordinary case. There's no contact with the outside world. There's no capacity for action. If there's some kind of consciousness there, you can't get at it directly. You know, it's not going to communicate with you. It's not going to behave. Your tools for identifying it are going to be have to be very indirect. Interesting. Yeah, I love that phrase, islands of awareness. But yeah, so that would be another conversation we could do. We could do again, right? Yeah, indeed. Because there's, I imagine there's some other really cool edge cases there too. But so before we leave this question of intriguing boundary cases, I'm just curious if you could just rattle off, you know, just other boundary cases that that have been attracting attention among consciousness researchers recently. So there's there's three classes, I think. We've talked about examples of certainly two of the three classes. So there's these edge cases with respect to our species, the infants, for example, or individuals with the city of brain damage. There's edge cases with respect to non-human animals. We're most of us probably think that many may be most, maybe old mammals are conscious, but we're not sure about the insects or the snakes or the birds. And there's edge cases when it comes to technology. I don't think you're going to find many serious people who think that chat GP2 is conscious, but it's pretty damn smart. You know, it's success is going to be telling us it's conscious pretty soon. They're going to be using the language. And I think we're going to have AI systems that are going to be triggering all the consciousness, ascribing tendencies that kids have. We're going to be engaging with these systems in the way that we engage with conscious creatures. And so the artificial cases like AI systems like humanoids, that's going to be an increasing important edge case class. Yeah, for sure. So let's move now to the notion of consciousness test or C test as you call it. So what are C tests and why do we need them? What do they do for us? Well, what they do for us is they enable us to make justified claims about whether or not a system is conscious. And ideally, they would tell us what kind of conscious capacities the system had. So, you know, in the best of all possible worlds, I kind of have this picture of a universal C test, something that you could point at a system and it would tell you whether it was conscious, what kind of conscious experiences it was having, whether it was awake or dreaming, you know, what kind of global state of consciousness it was in. That would be the holy grail of consciousness test research would be to have something be like a thermometer that you could point at any substance and it would tell you how hot it was. So a good C test wouldn't necessarily just return like a yes or no answer. Yes, this is conscious. No, this is not. It would actually give you maybe some more textured information about both degree or maybe kind of consciousness, that kind of thing. Yeah, it's just a construct and that's that's one way to think of it. What we focus on in the paper is a more kind of stripped down thing that would kind of give you a yes, no, or if it's neither a yes, no, it would say, well, it's indeterminately conscious. It's a little bit conscious, so it's a borderline case. The important thing in my mind is it goes beyond ordinary folk judgment intuition, you know, it grounds descriptions of consciousness. You know, policymakers, people thinking about animal ethics and animal welfare, they get philosophers or scientists to tell, you know, they say, you know, how should we treat lobsters? And we're not that far away, I think, from having a court asked for expert opinion on consciousness in organoids or AI systems, you know, it's already an issue when it comes to fetuses and infants, you've already famously got people worrying about a particular type of conscious state in the fetus pain. So, you know, if we had the means, if we had something, the C test terminology is just a kind of, it's branding, it's just a fancy label for the kind of procedures that we might want to rely on to answer those kind of questions. So you sort of alluded to the fact that part of why we need C tests is because it's not like we're just engaged in this academic taxonomic exercise of yes, this is conscious, no, this is not conscious. There's always this kind of moral layer to our descriptions of consciousness, to different beings. And so if moral questions arise, then it'd be nice to be able to appeal to external tests, right, that give us some sense of whether or not an description of moral status might be appropriate. That's right. People care about it. It matters, it matters to the science, but it matters ethically, it matters politically. So in the paper, you focus on seven, I think it is of the most prominent current consciousness tests, C tests. I thought we could just very quickly go through these because each of them is kind of interesting in its own way and they come out of different literatures and they're sort of fun. So why don't we just start with the command following tests. So what is the command following test? So what we focus on is the command following test, and I guess you could probably more strictly speaking covert command following is a test developed by a British neuroscientist Adrian Owen. When he was at Cambridge, he was a patient who had been in a serious car accident and was in the vegetative state. What was then called the vegetative state. Now it's been rebranded as the unresponsive waitfulness syndrome, but she was sort of non-communicative. So she wasn't in the coma exactly. She was breathing and there were moments of a rouse of where she would open her eyes, perhaps momentarily move her eyes around, but there were no clinical, no overt, clinical signs of consciousness. And what they did was they put her in a brain scanner and they gave her some imagery instructions in one trial that I asked her to imagine playing tennis. So a motor imagery task. Now they said a trial that I asked her to imagine walking around the rooms of her house with spatial navigation imagery task. And they use those tasks because they generate predictable activity in certain parts of the brain. And in fact, they had 12 neurotypical controls and they gave them the same task. And you know, it was astonishing that the patient generated brain activity in response to those commands that were sustained for about 30 seconds and then they said stop and it was time locked to, you know, it began in response to the instruction when they said stop it finished. And in fact, I think Adrian's work is relatively well known. What people tend not to focus on is the time locked element, which I think is one of the most interesting parts of it. It's not just where in the brain it happened, but when it happened. And so, you know, the key there is arguably the patient was following commands following the instruction to produce visual imagery. And that was identified by brain activity, not via movement of her arms or movement of her eyes, but brain activity. So it's I think of that test as piggybacking on the kind of test that clinicians use to tell whether someone might have regained consciousness, they'll typically say, you know, lift your arm or follow my finger with your eyes, something like that. Yeah, or blink twice, blink twice, right. And in this case, she's arguably following commands is just that the command is being picked up by brain imaging. Yeah, it's interesting because I agree that it's it's compelling that this person is probably conscious if they're able to follow these commands. On the other hand, it seems like a pretty stringent test in the sense that presumably you can be conscious even as a human without actually being able to understand language, right? I mean, we were talking about the case of three month old infants earlier. They're not going to be able to imagine playing tennis, right? Right. Exactly. So this is why it's crucial, I think, to distinguish between whether, you know, failing the test is evidence that you're not conscious. I see. So in this case, we wouldn't want to say it is three month olds fail the test or in some sense, it's not even appropriate to even put the test to them, but a negative result doesn't give you evidence that consciousness is absent. Right. The logic of these tests is really interesting and tricky. I think your point is an excellent one that we need to attend to. The conditions under which failing a test tells us something interesting and Owen in his collaborative is a very keen to say failing the test doesn't tell you anything passing it does. Let's move on to another one of these. I was intrigued by this sniff test. That's not one I'd heard of. So what is the sniff test? Yeah. So this has been less discussed. The primary author here is Ana Arzi, an Israeli neuroscientist. She looked at the way in which patients, and again, this has been primarily applied exclusively applied as far as I know to patients in these disorders of consciousness states, you know, unresponsive weightfulness syndrome, vegetative state patients, and showed that their response to an odorant seems to be very strongly diagnostic of recovery of consciousness down the track. And then so they present positive odorance or negative odorance, and they contrast that to the way in which the patient responds to just plain air. And the response to the amount of sniffing as it were, the inhalation volume looks like it's, those who go on to recover seem to respond to the odorance very differently from those who don't. Interesting. So they must be using pretty strong odors for this, as it like gasoline or something. I'm not sure I'm not across the details of what exactly the odorance were. And I should say actually there were about 12 co-authors. I forget how many of us were on this paper. And this is one of the tests that my co-authors would much more across the details of that than I am. Part of what intrigued me about it is, you know, in popular representations of coming back to consciousness after being knocked out or something, you often have people giving like a strong odor as a way of kind of jolting someone to consciousness. So there's that idea kind of out there. I don't know if there's any evidence, any scientific evidence to support that's actually affected, but it sort of interested me that, yeah. Well, you know, smell is very interesting. It's the circuitry is different. It's got a subcortical root as opposed to the other perceptual modalities. And of course, it's bound up with memory in a way. Smells evoke times and places. I mean, proofs talks about this, doesn't it? You know, so I think smell is fascinating. Okay, so let's talk about a couple others of these. So another one that caught my eyes, the unlimited associative learning test. So what is that? The idea is that the capacity to associate a stimulus with a response in a very plastic, easily reversible kind of way. And to learn in this open-ended fashion is a crucial transition marker for the evolution of consciousness. This idea has been developed and you, you know, you mentioned that some of these tests come out of different literatures. So this test comes out of the animal consciousness literature. It was pioneered by Eva De Blanca and Simone Ginsburg. More recently, the philosopher Jonathan Birch has collaborated with them in developing some of this approach to thinking about unlimited associative learning as a test for consciousness. So the idea is we can use this capacity as a marker when consciousness is likely to have first emerged in evolution. It hasn't, to the best of my knowledge, been applied to to some of the other difficult cases that we've been talking about, infancy to sort of a consciousness in the like, whether it could be applied, I think is a really interesting question. Is it an actual test or is it sort of like a hypothetical test? Like if you could devise a test for unlimited associative learning, then it would be useful in sort of looking at the evolution of consciousness. It's not going to be, it's more of a framework for a test in the sense that there's not going to be a single test that tells you whether a creature has that capacity. It's a capacity to figure out whether a creature has this capacity for unlimited associative learning. You've got to do a whole bunch of tests and rule out all sorts of other possibilities. You've got to basically figure out how flexible its capacity for learning is. Can it associate dissimulus with this response and can it easily reverse those two? The logic of it is very different from say a command for a willing kind of test. Right. That's right. Interesting. Part of why I asked is that it does seem like, well, how are you going to demonstrate an unlimited capacity? It seems like a hard, a hard bar to clear, right? For sure. The last one I thought we could touch on is the AI consciousness test. So what is that one? So this is something that the philosophers Susan Schneider has talked about in thinking about AI consciousness. So the proposal here is that you engage an AI in conversation and it's crucial that the AI is kind of boxed in in the sense that it doesn't have unlimited access to what's on the internet or the like. And you see if it spontaneously engages in the same types of thoughts and discussions about consciousness that we do. Does it find thought experiments involving consciousness equally gripping? So the zombie thought experiment that there might be something that's got all your physical and functional properties, but there's something that's like, or inverted spectra, right? That what looks yellow to you might look red to me or what looks red to me like more like green to you, and vice versa. Does it spontaneously come up with these kind of thoughts and puzzles and does it find the problem of consciousness as gripping as as intractable as we do? Interesting. I thought there is that suggests that it's got the concept of consciousness and that in turn suggests that it is conscious. So it seems like across these four tests that we've talked about, you can already get a sense that there's a lot of variation in these tests. So one this is just me noticing from the outside, but you know one kind of variation is just in like how much operationalization is in place, right? Like how much people have actually tried to turn this into a real test that you can go out and run today. And then another that just strikes me as like they seem to put the bar in different places, right? I mean like being gripped by thought experiments about consciousness, that's a pretty high bar. I mean no one's going to pass that three year old's not going to pass that. And maybe maybe some precocious three. That seems like a pretty high bar. So anyway, it just interests me that there's this heterogeneity. Yeah, look in one sense putting these different tests into conversation with each other and you know we put them on a single graphic is kind of weird because they do come out of different literatures. They're focused on different populations, right? The AI consciousness test is focused on AI. It's not focused on babies. It's not focused on human animals. You know, the command following test that we talked about is focused on brain damaged patients hasn't seriously been applied to AI and probably shouldn't be, you know, you tell the AI to raise its hand or imagine something. I mean, that's not really going to be particularly compelling evidence of consciousness. So in a one sense, it's sort of a weird thing to do. But in another sense, I think it's quite provocative. And it forced you to step back and say, okay, what's going on here? Do we want a single test that could be applied to all of these populations? Right. Which of these tests is, you know, passing it is positive evidence of consciousness, but failing it isn't is the bar here too high and the bar here too low or or or was the low bar in the right way. How do we set the bars and should we have the same kind of bars in different places? Should we demand the same of a three-year-old that we demand of an AI system? And what way should we demand? So you mentioned that I forget how you phrase it, but your dream essentially would be a universal CTAS, right? One that can be applied to any class of entity, one that would give you maybe not just a yes-no answer, but would return more granular information about kind of consciousness or degree or those kinds of things. I'm curious, though, like, is there a difference between your dream for these kinds of tests and you're like, what you actually think is a reasonable achievable goal for these tests? So in other words, is there some more realistic near-term goal that you have in mind? It's not a universal test, but is something beyond the kind of heterogeneous mix that we currently have? So what I think we should be doing is starting with the systems that are near and dear as it were. The systems that are closest to us. So, you know, we've talked about these edge cases, these hard cases, but I think it's crucial to keep in mind that they're not all equally difficult. Some of them, you know, the AI systems, maybe the Xenobots, maybe animals that are very, rather genetically distant from us, are really hard. We're to say about those kind of cases, but the place where I think we have made the most progress, potentially can make the most progress in the near-term is with the edge cases that are closest to home-base to the consensus cases. So the consensus cases, you or me, right? I know that you're conscious. Philosophers, you know, there's a problem of other minds in philosophy where the philosopher will say, "Well, how do I really know that you're conscious? Maybe you're just a zombie?" That's a problem that only a philosopher could love. Right. Right. Scientists aren't worried about that problem. They're worried about bees in AI systems and organoids. So let's take for granted that other human beings above the age of three who haven't suffered from immense brain damage, who can communicate, who can respond to contingencies in the environment they can learn. Let's take for granted that those folks are conscious and see how far we can go beyond the consensus cases. What can we say about infants, fetuses? What can we say about non-responsive individuals like the one that Owens' command following study was developed? And crucially, by thinking about those kind of cases, can we validate, can we as it were, provide scientific backing for some of these tests that aren't just folks' psychological intuitions? So the covert command following is great, I think. But it suffers from a kind of limitation in that the logic of it appeals to this link between being able to follow commands and being conscious. And if someone says, "Yeah," sure the person was able to imagine doing playing tennis. Yeah, that's cool, but I don't think that's evidence of consciousness. Because look, here's my AI system that, you know, I built this AI system that can do that. Or so, you want to have something just say to them, right? You want to be able to, as it were, ground that inference in something rather than just, well, seems obvious to me. This is how we do it in hospitals when it's overt command following. That's the challenge it seems to me. And the key, I think, is to try and validate some of these measures in populations that are most like us using us as a kind of base case and then trying to bootstrap our way, inch our way to the more alien kinds of systems, other kinds of mammals. And using what we learned about those other kinds of systems to revise our starting point, our intuitive judgments about consciousness. The model here is kind of thinking about the development of ways of measuring temperature. Back in the 10th century, you know, we had ways of measuring temperature. You know, you put one hand in icy water, it's colder than a hand and bottling water. Okay. But we weren't very good at it, right? And we couldn't make fine-grained judgments. We couldn't measure temperature a long way away from us. We didn't have thermometers. And so you have to start with intuitive common sense judgments and kind of construct a model that can then guide the development of thermometers that can be applied to more exotic substances. And that can go beyond intuitive judgment every day judgment and can correct every day judgment when it's found to be misleading. You know, if you've got one hand that's hot and one hand that's cold and you put them in the same bucket of water, they're going to deliver different judgments. So you know that every day judgment about temperature is fallible, is courageable and needs to be corrected. That's what thermometers do. So we need to take the same kind of approach to consciousness. So, but then, okay, so this is a really interesting analogy and I can fast now that the book that you may be alluding to about the development of temperature. I think it's called the invention of temperature. Yeah, that's a challenge. I still have not read. Yeah. Yeah. It's been on my list for like 10 years. But it sounds, I don't know why I haven't read it. It's like one of those ones that's on the list and I'm like, that sounds so interesting. And yet, but anyhow, could you just briefly fill in the analogy for us? Like, what was the between, oh, this water is icy and cold, this water is warm and thermometers. Like, what were some of the intermediate stages that you think could be analogous to like an intermediate stage in the development of how to measure consciousness? The key, I'll get to the intermediate stage just sort of in a minute, but the key I think is to think of consciousness as this real phenomenon in the world that you're trying to latch onto. That's what enabled us to develop ways of measuring temperature. It's a real phenomenon. We didn't invent it. I mean, the title was kind of sent me ironic, I think, the invention of temperature. Because what we're really doing is allowing the world, we're probing the world and we're trying to discover the underlying nature of temperature such that we can invent ways of latching onto that thing. And so we develop rudimentary thermometers and we try and check them against each other. And when we get outliers, we think, well, we don't have a direct way of telling that that's an outlier because we don't yet know that thing itself, right? You don't have the perfect thermometer that you would need in order to know whether this thermometer that you've invented is unreliable. You can only use these other thermometers so you construct different ways of getting at the underlying phenomenon and you try to understand what those different ways are measuring by a process of model construction and then you iterate your way to a better understanding of that underlying kind. So what I think we should be doing with consciousness is something similar. We think of it as this underlying phenomenon that we haven't yet fully latched onto. For any particular way of measuring it in verbal report or command following or sniffing, for any of those particular things, a pre-ethereetical faith in them might be unjustified. So we're trying to find correlations between those things when we've got six markers of consciousness and five of them seem to hang out together pretty well and predict each other pretty well and the six one over here doesn't seem to be good. The thing to do is to downgrade our faith in that sixth one, upgrade our faith in those five ones and to try and figure out what's the underlying phenomenon that explains why these five things hang out together as closely as they do. So then in your vision for the future of C-tests, would it be something like we develop a battery of tests or maybe we pull from some existing tests, but we basically pull together a battery of tests that we apply to the same populations or like we apply all of those tests to a population of interests and see okay which of those tests return the same general result and then maybe we take a subset of tests that could actually be usefully applied across different populations. So we get are we getting the same result across these populations from something like that and then you develop more tests inspired by the existing battery and that sort of thing or yeah yeah that's exactly right. So let me come back to the baby's case because I think it's nice to have a worked example. Yeah kind of see what you might be trying to do. So in the review paper that you mentioned we looked at four markers of infant consciousness. So one is development of various neural networks, various networks that seem to be implicating consciousness as a default mode network and it connections between that and various potential networks in the front of the brain. Looks like the re-emergence of that is correlated with the recovery of consciousness following anacetic and following brain damage. Looks like that's got something to do with consciousness. So it's kind of interesting to know when that emerges in development. Looks like it's online pretty much from birth actually in the weeks before birth. So that's one. There's the recruitment of the front of the brain in attention in tracking various salient features in the environment that looks like on many accounts. It's closely connected with consciousness. There's various kinds of multi-century integration where what you hear and what you see are integrated in complex ways that seem to require consciousness. So we looked at when that emerges in infancy and then there's another thing that has got a lot of attention which is neural responses to second order oddballs. So if you get a bunch of auditory stimuli and you get a pattern set up with a pattern involves an outlier. So you get that pattern set up. The outlier is a first order oddball and then you get another sequence where you expect the oddball at the end of the set and you don't get the oddball. So there's a second order violation. So this is like if you set up AAB, AAB, but then you do AAB. Something really okay. Yeah. Yeah. So it looks like the neural responses to that second order oddball, the global effect as it's called. It looks like that might require consciousness. So we looked at see whether that emerges in infancy. And you know the interesting thing is the ages don't line up fully. Okay. Some of these things emerge earlier than others according to the data we have. Now the data we have is patchy sparse wasn't necessarily collected with exactly this issue in mind. But you know suppose it turns out that even with all the funding in the world and all the studies have been done, the data don't line up. Well then in so far as consciousness is a thing and these things are markers of that. There are tests if you like of that. Then you expect them to line up. And so now we've got an interesting challenge, interesting question from which we can hopefully learn quite a bit. So maybe one of these things actually isn't a really good C test because it doesn't line up with the others. If three of them line up really nicely, suppose the first three we talked about line up really nicely and you know come online at one week of age and the fourth one doesn't. Okay. Now tentatively we could downgrade that so actually you know that's not a very strong marker of consciousness. Or it's a marker but only if this other thing is in place relative to this other thing right, it gets masked by this other thing in it. That's why you see it only at four months of age rather than one week of age. Yeah it's interesting. I mean it does. This idea that all the tests would line up on the one hand. It makes intuitive sense but I think only if you buy into this model that we mentioned earlier in the conversation that consciousness is kind of an all or nothing thing right. If you think of consciousness as either graded or maybe even graded implies that like varies on a linear scale right but it could even be sort of more multi-dimensional right there could be different a different shape to consciousness in certain parts of the shape get filled out at different time. I guess I'm wondering to what extent we even need all of them to line up. Could it just be that the thing we call consciousness actually ends up being a little bit less uniform than we thought. I think it's all to play for a present and you know for what it's worth I do have some papers on where the consciousness is a multi-dimensional construct. Okay. Yeah I think it's an interesting idea and different things could be meant by that. Right. I think in the simplest case you've got a creature that's conscious and is determinedly conscious. So you know it's a bit like this bucket of water is under 40 degrees you know if you've got six thermometers and they're all accurate you want them all you want them each to say 40 degrees. But of course one of those thermometers might not work very well if it's used to measure a gas just doesn't work for gases. The other thermometers might be quite good at both gases and liquids. You know there are some contexts in which different kinds of thermometers will have a different range of applications and it's also possible and maybe this is where you're going that in certain respects heat is not a good model to thinking about consciousness. So maybe constructs that are much more nebulous are better models like or like wealth. You know think think of social scientists trying to measure wealth or happiness. Social scientists spend a lot of money trying to measure happiness. There are facts about happiness that some people are maybe happier than others but there's a lot of indeterminacy there and different measurements of happiness might disagree without one of them being wrong. Yeah I think that is kind of what I'm grappling with. I mean the temperature analogy which I like and is very helpful. It sort of smuggles in the idea that because we now know from the vantage of the present with our improved scientific understanding of physics. Like we now know that it makes sense to think about temperature as this kind of dimension of the physical world that cuts across things like gases and whatever. So in hindsight it's like oh this is how we went about measuring that thing that's external and pre-exists our measurements and it's just out there in unitary. But in the case of consciousness it's not settled yet that it is that kind of. I agree. Yeah I think that's right and I think that's important point. And I think this is the more you know we talked at the start of the conversation about you know what's the definition of consciousness how do you define it. In a way I think this is the pointy end of that issue. It's not so much the particular definition that is so crucial here although that doesn't hurt but it's the conception of the type of thing that it is. The type of way we ought to be doing it. Is he to good model for thinking about consciousness or as happiness a better model. You know should we treat consciousness like water this underlying phenomena that you can do a real science with or should I thinking about the distribution of consciousness in these edge cases you know should we be looking for stipulation and making decisions rather than discover is it more a matter of decision than discovery. That I think strikes many people as wrong-headed but I don't think that that approach should be dismissed. Yeah interesting. The other constructs that I I've been thinking about and we don't need to we don't need to take these up because it would send us off in all kinds of directions but everything about the problem of intelligence right just saying the other system is intelligent and that's I think there it's clear that like there isn't some objective underlying ground truth. It's more obvious to me that it's kind of multi-dimensional and it's more about description and judgment and then the other one that you know comes to mind is the question of like when does a system have language when is the system a language a language using system where again it doesn't have the same on off intuitively doesn't have the same on off kind of quality the same sort of singular dimension. Anyway so those just have me thinking about other ways we could imagine the kind of measurement the measurement problem. Yeah yeah I think that's really insightful I think there's a jump it's it's natural to kind of reify consciousness and think it's going to be a thing and we don't make that mistake or at least those of us who are a little bit scientifically informed perhaps don't make the same assumption so quickly about intelligence or necessary language I mean we're happy to think of these as graded phenomena you can proto language is fine proto intelligence you know seven types of intelligence or however many there are right you know I think and maybe those are exactly the right ways to think about consciousness I'm not persuaded that they are but I'm also not inclined to dismiss them but I guess I would say that even once you make a lot of distinctions between types of consciousness and we you know we clearly can right I mean if a rabbits of consciousness they're clearly not conscious in the way that you and I are our consciousness is inflected by language by our sociality by culture we're not locked into the here and the now you know we engage in mind wondering how the time our thoughts are in the future of the past you know we've got this rich in a monologue that accompanies our experience lots of animals presumably don't have that but this brute basic kind of primitive awareness it's not so clear to me that that itself just sort of gets parceled out into lots of subcapacities that such that is there's no fact of the matter and you can call that consciousness if you like but you know it doesn't really matter where you draw the lines and it might be much more like fire you know I was burning off some stuff in my backyard the other day you've got the diesel and you've got a match or actually it was smoldering from some ashes and it was smoldering smoldering smoldering and then bang you get ignition right there's a step change where something new happens and it's not that a new immaterial substance comes into existence we don't need to be dualists about this we can still think that the world is at base a physical world but sometimes there are step changes and something radically different happens and maybe consciousness is more like that rather than incremental piecemeal changes but I think we don't know so there's this concept of the expanding moral circle right the idea that the kind of circle of beings that we have thought of as deserving our moral concern has been getting bigger it's been expanding over time and it seems like there's something similar going on with consciousness that our consciousness circle if you like is also expanding I mean just within the last couple weeks there was this New York Declaration on Animal Consciousness which I'm assuming you are familiar with and I guess I'm just curious to hear your thoughts on that expansion I guess I'm on board with the expansion as a general approach I think it's true sociologically that at least in the west I'm not sure it's true of all cultures I doubt that it is but at least in the west this sense of you know who's in the consciousness club has certainly been expanding I say that with some tentativeness because I'm not a historian of ideas so just to take the infant consciousness you know I do know that at least in the medical profession the idea of infant consciousness wasn't taken as seriously 30 40 50 years ago as it was now what did people think in the 17th century ago infant consciousness what did they think in the 10th century what did they think of the Stone Age I mean I think it's a genuine question and in some sense you might say animism is kind of the default view of human beings maybe this is an expansion that has followed a retraction as it were you know I mean we don't think trees most of us don't think trees are conscious or mountains are conscious or rivers are conscious many people have said things like that and some people still do whether they mean by consciousness what you and I might mean that I'm not so sure of yeah so what you know what counts as an expansion might depend a little bit on your historical time frame yeah that is interesting because generally across intellectual quarters you more often see these kind of pendulum swings right and so it was interesting to me that in this case in the case of the moral circle and in in the case of this apparent maybe expanding conscious circle you have more of just a one directional change so that's really interesting to think that actually know if you zoom out even further maybe it's not a unidirectional change maybe there is there is some pendulous behavior here but then I guess do you think that that expansion is because our ways of measuring consciousness have gotten better in other words we're better at tapping into some underlying reality or simply that our construct of consciousness has gotten looser and more permissive yeah that's a great question or a third option we just sort of more attentive right interesting nothing's changed we're just watching better and listening better and maybe not so scared of ascribing consciousness to other creatures right I think it's a great question what's behind it I agree that within the last 40-50 years the pendulum is definitely swung into an expansionist direction like if I had to hazard a guess I think some of it's got to do with a severing of the ties between consciousness and rationality and intelligence interesting so we're much less keen and we used to be on the thought that consciousness is all about reasoning and language and communication you still hear people keen on that idea but I think we're much more inclined to think the roots of consciousness either about tracking the contrast between what I'm up to and what the world's up to so it's important that I can distinguish movement in my visual field that's caused by something moving in front of me from movement in my visual field that's called by me moving my head in order to do that I need a model of myself and what I'm up to as an agent and I need a model of the world so I know which is changing and maybe consciousness is about that or maybe consciousness is about primitive bodily sensations and emotions hunger thirst the need for oxygen pain and those systems are presumably much more ancient and primitive than language communication so I think it's that and that thought that movement I think has received surprising support from the development of AI things that are kind of smart in one sense and linguistically inflected in one sense and yet we look at them and think they're not conscious they're not beast machines right they're not they don't care they don't care about their bodies they often don't have bodies they're not tracking pain they're not tracking hunger and thirst they're not looking for mother's milk so I think that a newborn animal cares about chat GPT does not care about those things you know if you think consciousness is connected to these basic bodily drives then you're going to be moved in an expansionist direction because they're everywhere they go a long way down as it were the phylogenetic tree okay well I will resist the temptation to ask you more about this phrase beast machine I know I know we said we'd come back to that but well I'll point you in the direction which is a Neil sets book this this is it's a Neil sets phrase oh I see okay his book being you this is his theory of consciousness so yeah for beast machines pick up a copy of the Neil's wonderful book it's a great read and full of great science but yeah you get you get the full rundown on the beast machine idea there all right awesome well that's where I was going to go next was to ask you about any reading recommendations for folks who want to dive deeper into the stuff we've been talking about so there's there's one a Neil's book so well a Neil's book which is pretty recent slightly older but I think is very much in the same tone which combines wonderful science with some thoughtful philosophical reflections and a lot of just gripping it's just a gripping readers Adrian Owens book into the gray zone so we talked about Owens where I can give you the backstory on on how we develop the command following test and different applications of that and I guess the third recommendation is Peter Godfrey Smith's wonderful book again it's it's a little older other minds the problem he's focusing on octopus and cuttlefish consciousness so you know these creatures that are fascinating a branch of evolution that's far far removed from us so alien in some sense and yet in Peter's hands also familiar in a certain way you know we can recognize playfulness and intelligence there's a kind of familiarity there and just a wonder fleeting philosophically deep and scientifically informed reflection on consciousness and octopus and it's octopus is not the pie apparently yeah that's important let's get that it's important get that right all right awesome well Tim this has been super fascinating we could obviously go on for much longer on all these really cool threads and provocative questions but for now we'll have to close the conversation but thank you so much for your time thank you Kenzie that was wonderful we enjoyed it the many minds podcast as a project of the diverse intelligences summer institute which is made possible by a generous grant from the Templeton World Charity Foundation to UCLA I'm Kenzie Cooper writer thanks so much for listening