Your Brain Lies to You (And That's a Good Thing): Anil Seth on Consciousness

What if everything you see, hear, and feel isn't real? What if it's just your brain's best prediction of what's real? Dr. Heather Berlin and Dr. Christof Koch sat down with neuroscientist and consciousness researcher Anil Seth to explore the profound question of: What is consciousness? Seth's answer might surprise (and even unsettle) you. Consciousness isn't a window onto reality. It's a controlled hallucination constructed by your brain to make sense of a world it can never directly access. These questions of consciousness illuminate central ideas of the Perception Box, a framework developed by Elizabeth R. Koch for understanding the forces within us that shape how we see ourselves, relate to others, and move through the world. Anil Seth is Professor of Cognitive and Computational Neuroscience and Director of the Centre for Consciousness Science at the University of Sussex. He is the author of Being You: A New Science of Consciousness and more than 200 scientific papers. Dr. Heather Berlin is a neuroscientist, clinical psychologist, and Professor of Psychiatry and Neuroscience at the Icahn School of Medicine at Mount Sinai in New York City. Dr. Christof Koch is Chief Scientist for the Tiny Blue Dot Foundation and the current Meritorious Investigator and former President of the Allen Institute for Brain Science. Join us for new episodes of Science of Perception Box every Thursday. Follow the show on YouTube, Spotify, Apple Podcasts, Amazon Music, or wherever you get your podcasts. Love Science of Perception Box? Write us a review on your podcast app, or tell a friend! Visit unlikelycollaborators.com for more information. Chapters: 00:00 Escaping the echo chamber 01:43 Defining consciousness 09:05 Perception as "controlled hallucination" 11:26 Why perception feels like reality (an adaptive illusion) 17:06 When perception shapes social and historical "truth" 19:03 Hallucination as uncontrolled perception 21:57 What is the self? 33:38 Intelligence vs. consciousness 35:42 Anthropomorphizing AI and "bio-chauvinism" 36:10 Biological naturalism: can AI be conscious? 39:44 Should AI be designed to seem conscious? 40:52 Perception Box question

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  We need to realize that we're in one to start with. It's like any kind of echo chamber. You can't get out of an echo chamber unless, to start with, you know that you're inside such a chamber.

• 0:10Heather Berlin?

  Hi, I'm Elizabeth Koch. What's a perception box? It's this invisible mental box that each one of us lives inside, built from the material of our beliefs, life experiences, and the meaning we've made out of those experiences. You could think of it as a kind of reality filter that's constantly shaping how we see ourselves, other people, and the world around us. This podcast explores how, although the walls of this box are always present, they can expand in moments of curiosity, awe, and wonder, or contract in response to fear, anxiety, and anger. In each episode, we invite experts to share how to create lasting shifts in our perception.

• 0:55Heather Berlin?

  I'd like to introduce our esteemed host, Heather Berlin, professor of psychiatry and neuroscience at the Icahn School of Medicine at Mount Sinai, and Kristof Koch, chief scientist for the Tiny Blue Dot Foundation and former president of the Allen Institute for Brain Science. Welcome to the Science of Perception Box. This week, our guest is Dr. Anil Seth. As Anil explores, perception isn't passive. It's closer to a controlled hallucination, shaped by prediction and interpretation. What feels real and certain isn't reality itself, but our brain's best guess. Which raises a deeper question. If we're always living inside this construction,

• 1:38Heather Berlin?

  how much of what we take as truth is something we've learned to see? To start, we want to interrogate how we define consciousness. So Kristof, how do you define consciousness?

• 1:50Speaker 02

  Any subjective feeling, pain, pleasure, dreaming, imagining, treading, hoping, those are all conscious experiences.

• 1:58Heather Berlin?

  So those are just some examples, but how would you define it?

• 2:02Speaker 02

  You would have to describe the universe of all possible conscious experiences.

• 2:06Heather Berlin?

  Okay. Again, not a definition, but how about this? How about subjective first-person experience? Any experience? The feeling of pain?

• 2:14Speaker 02

  Yes.

• 2:15Heather Berlin?

  Okay. You don't need language for it?

• 2:17Speaker 02

  Correct.

• 2:17Heather Berlin?

  You don't need a sense of self for it?

• 2:20Speaker 02

  Correct.

• 2:20Heather Berlin?

  Meta-awareness, awareness of just pure feeling, pure sensation?

• 2:24Speaker 02

  You don't need to be intelligent about it?

• 2:26Heather Berlin?

  Nope, no intelligence. Okay. So pure, subjective experience, but it's first person. Only you know you have it.

• 2:32Speaker 02

  It's only you, yes.

• 2:34Heather Berlin?

  So...

• 2:35Speaker 02

  You can lie, of course.

• 2:36Heather Berlin?

  Well, yeah, I could say I'm conscious. How would you know?

• 2:39Speaker 02

  Well, more relevant, you can say, "I love you, but you don't love me." Or you can say, "I'm in pain," while you're not in pain.

• 2:44Heather Berlin?

  Right, right. Only you know.

• 2:46Speaker 02

  Only I know. Only I have access.

• 2:48Heather Berlin?

  Right. You can report to other people about it, but only you actually have access to the feeling. It's interesting, because to differentiate things that are happening at a very complex level unconsciously within us that are guiding our behavior versus things that we have subjective awareness of that's happening in our brain.

• 3:02Speaker 02

  Like, most of the things in our brain are probably unconscious.

• 3:05Heather Berlin?

  Right.

• 3:05Speaker 02

  Right?

• 3:06Heather Berlin?

  I agree. Yeah. Well, on that note, I'd like to welcome our guest, Anil Seth. Welcome to the show.

• 3:11Speaker 00

  Hello, Heather. Hello, Kristof.

• 3:13Heather Berlin?

  Hello.

• 3:13Speaker 02

  Hello, Anil. Thanks for joining.

• 3:15Speaker 00

  It's a pleasure.

• 3:15Heather Berlin?

  So great to have you here. Not only are you an esteemed colleague, but you're a friend to both of us for many years. So what is your definition of consciousness?

• 3:24Speaker 00

  It's much the same. I think we tend to overstress definitions a bit. I mean, they usually come at the end of our scientific journey of understanding of a phenomenon, not at the start. And the utility is more to make sure we're not talking about completely different things. So I kind of agree with Kristof that the minimal definition that we're after should be just this notion of experience. There was a good example I was reading yesterday in a paper by Jonathan Birch, a philosopher, LSE. And he said, "A child throws a ball across a field and then runs after it." Now, the ball just flies through the air. There is nothing it is like for the ball to fly through the air. It's just an object that follows the laws of physics.

• 4:10Speaker 00

  The child runs after the ball, and is also, there is something it is like for the child to be the child running after the ball. That's the difference. It's any kind of experience whatsoever. And the key thing, as Kristof says, is to avoid, at the beginning, building in confounds with other things, whether it's language, whether it's intelligence, whether it's metacognition, and even actually whether it's first person. I think first person is dangerous here, because, yes, it's true that a conscious experience exists for the thing that has it, but it doesn't necessarily have to have a first-person perspective, like the world as seen from behind my eyes.

• 4:53Speaker 00

  That is also maybe a contingent feature of consciousness in humans and other animals. It's really any kind of experience whatsoever.

• 5:02Heather Berlin?

  So it could happen without any sort of self-awareness that you're having this conscious experience? Or one is having, or an organism is having this experience? Do you think that you don't need self-awareness, like knowing that I am the organism having this experience?

• 5:17Speaker 00

  That's right. I think conceptually, that makes sense. We can conceive of experience without the sense of a self reflecting on that experience. But I think there's also empirical evidence that this can happen. You can talk to monks who have thousands of hours of meditative experience, and they can report states of what is sometimes called pure awareness. So complete ego dissolution, experience without an experiencer. And if you don't want to spend 10,000 hours in a cave in the Himalayas, you can also access these states, to some extent, through pharmacology, through certain psychedelics.

• 5:54Heather Berlin?

  But there's still a physical being who's having the experience.

• 5:58Speaker 00

  That's right. There's a physical system, which is the basis of that experience, the locus of that experience.

• 6:03Speaker 02

  In fact, Thomas Metzinger calls these type of pure awareness experience also zero-person experience. Because to make, he makes a point that you make, there may not be anyone behind there that's looking out at awareness. It's just a state of being aware.

• 6:18Heather Berlin?

  So you can have pure perception, but just because, subjectively, it feels like it's sort of disembodied, doesn't mean that you don't need a physical organism, let's call it, creating that experience?

• 6:32Speaker 02

  You need a substrate. Is that what you're saying?

• 6:34Heather Berlin?

  Right.

• 6:34Speaker 02

  A brain?

• 6:34Speaker 00

  No, I think you do. I think there has to be a physical system, and this is betraying my metaphysical cards, put them on the table, of broadly speaking, materialism. So I like to try to understand. I don't know if materialism is true, but materialism is the perspective that conscious experiences are a property of physical stuff, you know, doing something or arranged in a particular way. So the basic starting point is that, yes, there has to be a physical system, a brain, maybe something else, but we know embodied, embedded brains are sufficient for this, because we are all conscious. That has to be in place. But then the nature of the experience that entailed by this physical system,

• 7:18Speaker 00

  that's where I think we have to be very careful not to build too much in. I really like the zero-person perspective, because it's far too easy to assume that features of consciousness, as they unfold for most of us, most of the time, just have to be there by definition, but they really don't.

• 7:38Speaker 02

  Of course, they are so prevalent in us, and particularly in people who write about, right? Who are highly literary, who think about themselves obsessively, who write books about consciousness. They reify the self then.

• 7:49Speaker 00

  Yeah. And of course, a lot of the early ideas about what consciousness fundamentally is emphasize things like that, emphasize language, emphasize rational thought. And this takes us back to Descartes, what Alison Gottlieb might call professor consciousness. But there's a more basic commonality of conscious experiences. I mean, we all experience the world in some way, not exactly the same way, as I'm sure we'll talk about. But if we're trying to understand the phenomenon itself, the trick is, we have to start somewhere. And human experience is a stable starting point, because we know we are conscious. But we must be careful of assuming that contingent features of consciousness for humans

• 8:33Speaker 00

  and other animals are there by definition. This is kind of easy to see for something like colors. You know, we sort of know that people see colors differently, and we can maybe appreciate that colors are not intrinsically part of the universe. But when we have something like first-person perspective, because we're so familiar with it, it can be hard to realize that that too is part of what needs to be explained, not something that can be taken for granted.

• 9:01Speaker 02

  As we know from babies, or little kids, right? Who don't have a lot of self.

• 9:05Heather Berlin?

  But even if, let's say, we take the self out of it and just think about this pure experience, right? As we understand it from neuroscience, we're creating this experience. We're creating our own reality. You know, you've called it controlled hallucination, right? Can you talk a little bit about that? It's sort of like we're creating from the inside out versus the outside in. And how do you bring in this kind of idea of like being a prediction machine?

• 9:32Speaker 00

  I think it's a very old idea, really, about how brains and bodies and living systems interact with the world. It goes right back, really, to Plato in Western philosophy, but we won't retrace the whole history. For me, a good starting point...

• 9:51Speaker 02

  You mean the cave?

• 9:52Speaker 00

  Yeah, the allegory of the cave. So, yeah, since we mentioned it. I mean, in the allegory of the cave, you have prisoners who are inside this cave. They're chained up in there. All they can see are shadows that are cast on the walls of the cave by a fire. And for the prisoners in the cave, these shadows are taken to be reality, because that's all they have access to. In fact, the real world is outside the cave. So one intuition towards the ideas of perception as a controlled hallucination is that instead of prisoners trapped inside a cave, think of your own brain trapped inside the cave, the bony vault of the skull, trying to figure out what's happening in the world. Instead of shadows cast by firelight,

• 10:35Speaker 00

  we have the electrical, chemical signals produced by our senses. And the brain interprets these in various ways. And a picture, a world appears. Is that the same thing as objective reality? No, there's a level of indirectness there. And I think it's a similar indirectness to prisoners within the cave. The only difference is... The prisoners are people with brains and perceptual systems themselves. When it comes to our conscious experience, there are no little people inside the skull trying to figure out what's going on.

• 11:12Speaker 02

  No homunculus.

• 11:13Speaker 00

  Yeah, no homunculus. The experience of being a self is part of the interpretation, part of the controlled hallucination itself. There's no mini-me inside my head that's trying to figure out what's going on.

• 11:26Heather Berlin?

  But that's the only access we have, given these kind of, in some ways, feeble pieces of matter that we are given to construct our sense of what's out there with. But we're so fooled by it. You know, most of the day, we're thinking what we are perceiving is reality.

• 11:42Speaker 00

  Well, for good reason. I mean, perception, like everything in bios biology, is shaped by evolution. And it wouldn't be very useful if evolution had designed our brains so that we perceived the world to be a kind of construction. I think perceptual systems generally work when we experience things as if they exist, because our perceptual experiences are not unrelated to what's there. They're very closely related to what's there.

• 12:10Heather Berlin?

  So it's almost like an adaptive illusion that we have, that our perception is reality.

• 12:15Speaker 00

  That's right. I'm always a little careful with illusion, because it sort of suggests that there would be a better way, a more veridical way of experiencing things. I think it's actually very good for us that we experience things as if they exist, because there is a very tight relationship. You know, if a car is coming across the road, then there is really something going on in physical reality that my brain is keeping track of. But the way I experience the car coming and the road and everything else going on, that's a construction. That's not a direct importing of objective reality into subjective experience.

• 12:55Speaker 02

  So that's what we call the perception box. And of course, we're never truly aware of the true nature of things.

• 13:03Speaker 00

  This is very clear when we do science, because we know that we don't have perceptual access to things. So we need all these complex instruments. We need telescopes, we need microscopes, and so on.

• 13:13Speaker 02

  I mean, even in vision, we only have a tiny access to the total spectrum, right? We don't see UV in above, we don't see IR in below. And some of us don't even see all the, with the normal colors.

• 13:26Speaker 00

  That's right. And I think vision is such a good place to start, although it's not everything, as some consciousness researchers pretend. It's quite a lot. But it's not that what we perceive is this restricted subset of objective reality. Right? So you're absolutely right to point out that the cells in our retinas are sensitive to, for most of us, three wavelengths. And that thin slice of reality, if we're only thinking about electromagnetic radiation, that thin slice of reality is where we live. But our visual experience isn't just sort of three things, three colors, right? We experience millions of distinctive colors.

• 14:08Speaker 00

  This, I think, makes the point nicely, that our conscious experience, our controlled hallucination, is simultaneously less than what's really there, less than the thing itself, because we only sample a thin slice of the electromagnetic spectrum. But it's also more than what's there, because we experience millions of colors rather than three wavelengths, which are not in themselves colored. They're just different wavelengths of radiation.

• 14:34Speaker 02

  Yeah, because we keep on having to emphasize that there is no color out there in the world. What there are are photons with a specific wavelength that get reflected into my eyes and then get processed in different bands that my brain then constructs. You call it white, right?

• 14:50Speaker 00

  Correct.

• 14:50Speaker 02

  It's a construction of our brain.

• 14:52Speaker 00

  Yeah. And this is why I think the language of illusion can sometimes mislead us, because if you take that book outside, and now we're inside, the ambient light has a particular color balance. The people videoing this notice very well, that's part of the job. If you go outside, it's a sunny day, there'll be a lot of bluish ambient light. Yet that cover of the book will still look white. So you could say that our visual system is doing a terrible job, because the light coming into our eyes has changed, from the book has changed, but yet we still see it to be the same color.

• 15:26Heather Berlin?

  It's amazing, though, because it adapts to keep consistency.

• 15:30Speaker 00

  That's right.

• 15:31Heather Berlin?

  The color of the book doesn't change, depending on the lighting. But then there's these sort of cracks in the matrix, like with the dress, you know, where it was ambiguous, and then some people saw it as black and blue, and others as white and gold, because some corrected for blue light, and other corrected for yellow light. So there are these subtleties where we find the crack in the matrix. But overall, our brains do an amazing job of keeping consistent. I would say, you know, our brains evolved for survival, not for truth. And but given that, you know, as a society, we run up against problems, like, can we ever actually get to truth and reality, given these cognitive biases that we have?

• 16:13Speaker 00

  There's just no way that our perceptual systems will access reality as it is. I mean, no matter how hard we try, we can't.

• 16:22Heather Berlin?

  Well, maybe we need things in place to work against our natural inclinations or biases to get to some other objective reality.

• 16:31Speaker 00

  Maybe the way towards that, and this is why I like the concept of perception box, is very consonant with the general view that I've been working with for years, is that we need to realize that we're in one to start with. It's like any kind of echo chamber. You can't get out of an echo chamber unless, to start with, you know that you're inside such a chamber. And just recognizing that the world that we experience, the world that appears for us, is not the world as it is, and that it might appear differently, but not arbitrarily differently for each of us.

• 17:06Speaker 02

  Okay, so when it comes to social events, which the perception box also includes, it's not just the reality, the physical reality as it were, what colors is, but also who caused what, who's responsible for what. Then it becomes incredibly important. It is a well-posed question, although the methods we have are very weak to try to get at. What actually happened to any one particular historical or social event?

• 17:31Speaker 00

  That's right. I think that in this case, the example of perception can be very useful, because you're completely right, that looking at things that happen in society, so these kinds of beliefs about social events or historical events that are very influenced by the media that we access and so on. But people can be as convinced that their version of events is the right one, in just the same way that those people who saw the dress as white and gold are convinced that they're right. And they can't imagine how it's possible for other people to experience something differently. This is why things like the dress and other studies of perception are so useful, because it's a relatively unthreatening context.

• 18:15Speaker 00

  And I think if you can have people realize that, oh, look, even for something as simple and as politically non-threatening as a dress can engender these two very different experiences. I think that can help cultivate a bit of humility for these larger questions, where the entrenched differences of opinion really do matter.

• 18:38Heather Berlin?

  But there's still a difference between differences at the perceptual level and differences at the sort of higher, let's say, cognitive level of belief systems, right? You know?

• 18:49Speaker 02

  I think it's a smooth transition. Yeah, clearly there's a difference, but it's still a perceptual. Who I believe is responsible and why and what feelings I have about that.

• 18:58Heather Berlin?

  It makes you question your own sort of beliefs at some level.

• 19:03Speaker 02

  We all know from our own experience, from clinical experience, that people who hallucinate. How do you think about when we all believe, well, he is totally hallucinating? What does it mean in your terms?

• 19:15Speaker 00

  Just as I like to describe perception as a controlled hallucination, we can think about hallucination as a kind of uncontrolled form of perception, uncontrolled by causes in the real world. So just as you said, there's a continuum from perception to belief. I think there's also a continuum of perception that's very tightly controlled by things in the environment, whether it's how a book reflects light or some other feature of objective reality. And to what extent our perceptual inferences become less controlled by these causes. And as you move along that scale, you get closer to what colloquially we call hallucination.

• 19:58Speaker 00

  People like Chris Frith long ago started to talk about hallucination this way as what happens when our perceptual best guesses lose their grip on reality. You can think of it in a way as a sort of extension of a phenomenon we're all familiar with, which is pareidolia. And we've all had this experience of walking out on a day when there's fluffy clouds. And if you look at clouds in the sky, sometimes you can see faces in these clouds.

• 20:24Speaker 02

  Dogs.

• 20:25Speaker 00

  We all see dogs, most of us. Some people see cats, other people see faces.

• 20:28Heather Berlin?

  They're usually informed by our biases and our expectations.

• 20:33Speaker 00

  And the phenomenology of this, I think, is very interesting, because we know there's no face there, but what we're experiencing is some kind of statistical shadow of a face, an echo of a perceptual prediction that doesn't quite grip. And you can imagine, if you turn a dial on that further and further, then you actually see faces in clouds. And if you turn it even further, then not only do people see faces, but they believe the faces are actually there. And I think a lot of cases of hallucination lie along the spectrum and can be usefully differentiated. And given actually mechanistic basis, we can start to think about what's happening in the brain that accounts for different kinds of hallucinations. And this is something we've been doing in my group.

• 21:15Speaker 00

  We've been trying to build computational models of visual hallucinations in different conditions, in Parkinson's disease, in psychedelics, in another condition, Charles Bonnet syndrome.

• 21:27Heather Berlin?

  Not speaking from experience, but if anybody has ever had a psychedelic experience, it really does look like that, what you've simulated in many ways, of what it is in some ways to be.

• 21:39Speaker 00

  In some ways, you know, I wouldn't over-egg the similarity. But I think it certainly captures a feature of some psychedelic experiences which have this pareidolic quality, that we see patterns in things. The face in the cloud becomes Oprah Winfrey marching across the sky, followed by a load of dinosaurs, that kind of thing.

• 21:57Heather Berlin?

  Right. Your book talks a lot about the self. You know, what is the self? Is there a self? What is it? How do you define it?

• 22:07Speaker 00

  This is another. Area where I think we we get misled by confusing how things seem with how things are. So in perception in general, it seems as though the world is just there and we open our eyes and have direct access to it, as we've just been discussing. How things are is likely very different. It seems as though the self is this essence of you or me that's somewhere, maybe in the head, that is receiving all these perceptions and then figuring out what to do next and then making some action and

• 22:39Speaker 02

  Here, it's exactly here.

• 22:40Speaker 00

  Yeah, it's there, it's there, and it's the locus of all your memories, all your all your plans, all your experiences,

• 22:45Speaker 02

  Your will.

• 22:46Speaker 00

  Your emotions, the source of your agency, the author of your thoughts. All of these things bundled together somewhere as an essence of you in there that maybe, depending on your religious beliefs, will survive the death of your brain and reappear in another body somewhere. There's that's the sort of the manifest image of the self, what the self seems like for most of us most of the time. But I think it's a bit like color. You know, there's a utility in the brain creating an experience of the self that has this feeling of of being real, of being an essence, being a stable quality. But it doesn't mean that that's actually what's going on at all. I think the self is a kind of perception. It's not the thing that does the perceiving.

• 23:31Speaker 00

  It's a bundle of perceptual inferences. But instead of being about how light bounces off a book, the self, for me anyway, starts with the body. So the the more basic levels of self are things like emotion, mood, which are well understood, and I think this is a tradition in psychology going hundreds of years back, that emotions are well understood as perceptions of changes in the physiological condition of the body. You say perception is for survival, and this is nowhere more clear in how we experience the state of our own living bodies. We perceive the body as an object in the world. You know, I I I know this is my hand, I experience it as my hand.

• 24:14Speaker 00

  That's also a perceptual inference.

• 24:17Speaker 02

  So I'm afraid because I see myself running and I sweat and my blood pressure goes up, right? And then I have the emotion of fear.

• 24:24Speaker 00

  Right, right. So an emotion is a response to the change in physiological condition. Actually, it can be a very closely coupled thing and also generate those those changes. I think the really challenging thing about the self is like you can, just like when you start with color and then you try and go to more challenging aspects of our experience of the world outside, like the flow of time or the structure of space around us, which seem harder to account for, but I think you still can. It's the same with the self, the self as an object in the world. You might say, okay, that's that I can get on board with as the brain making a best guess. But what about free will? What about mood? What about my sense of personal identity?

• 25:09Speaker 00

  You know, why am I Ann Ellen and not Heather or Kristoff? Those, I think, are more challenging, but I do also think they they can be accounted for by the same story.

• 25:19Speaker 02

  So in principle, it's another form of controlled hallucination.

• 25:24Speaker 00

  Yeah.

• 25:25Heather Berlin?

  It's another perception. And, you know, I

• 25:27Speaker 02

  It's not an illusion, though, right? I mean, or as much an illusion as seeing color is.

• 25:31Heather Berlin?

  Exactly.

• 25:32Speaker 00

  So some people, this is this is why, you know, I got a bit grumpy when you said illusion to start with, because illusions sort of suggest that it's it's it's we would do better if we saw things or experienced things more accurately. And people say, many people have said, the self is an illusion. And I probably said it, you'll probably point out that I'm contradicting myself now. But, you know, Anil says today, because the self is not one thing, so me now doesn't have to agree with the me that wrote the book.

• 25:59Heather Berlin?

  That's correct.

• 26:00Speaker 00

  I think illusion is the wrong is the wrong word for it. The the parallel with color is very useful. The experience of self is a very adaptive form of perception that helps the organism get along.

• 26:13Heather Berlin?

  It's a very creative, you know, perception that the brain can put together such a thing that is consistent. It's also built up on your memories. You know, when people lose their memories, they lose their sense of self and their identity.

• 26:27Speaker 00

  Well, not all of it. I mean, you know this, ah, I mean, you'll know this better than neither of us, working with patients all the time as you do. But there are famous examples of people who have complete loss of autobiographical memories, famous neurological patients.

• 26:39Speaker 02

  Clive Waring.

• 26:40Speaker 00

  Clive Waring, talk about him in the book, other ones, H M is another good example. Yes, they they lose an aspect of self, but what's interesting, at least to me, is that most aspects of their selfhood are preserved. You know, their first person perspective on the world is preserved, their emotions are preserved, they typically don't have an issue with knowing what their body is and is not. It's just this selective disruption of self as personal identity. But that's only one part of what a human self is.

• 27:14Speaker 02

  You discuss this remarkable case in this book of Clive Waring, this this music conductor in England with encephalitis and then has one of the deepest, most profound amnesia, doesn't remember anything going back and going forward. Yet what's so remarkable, and you point this out, when he meets his wife, who he married just a year before he had the encephalitis, he's overwhelmed by enthusiasm and and by love for her. So his love is somehow still has survived his loss of his person.

• 27:43Heather Berlin?

  The other thing is that it's it's an amazing case, because he it's like you keep pressing the refresh button. So he keeps saying, I he writes this journal, I am just now first conscious for the very first time. And then like, like a minute later, I'm now just first conscious of it, because he doesn't have that consistency of, you know, continuity, the continuity. The other thing that's really interesting with that case is that, because the subcortical parts of the brain, like the basal ganglia and parts that have to do with like sort of muscle memory, he if are still intact, so he could play, if he started to play a piece, he could suddenly get and play the entire piece.

• 28:16Speaker 02

  Or conduct, yeah.

• 28:17Heather Berlin?

  And, yeah.

• 28:18Speaker 00

  The ability to think of ourselves in the future is an enormously adaptive feature, and to think of that future in terms of the continuity of self is is very, very adaptive. But it's also come with a cost that we can conceive of our own death.

• 28:32Heather Berlin?

  Right.

• 28:33Speaker 00

  And that, you know, that's not something that I doubt evolution selected for our ability to conceive our own death, but it did select for our ability to plan for the future organized around a self. So, yeah, it's a price we pay. There's a there's a trade off.

• 28:48Heather Berlin?

  I mean, the whole reason I became a neuroscientist to study consciousness is because I want to figure out how to continue my consciousness after I don't have a body.

• 28:54Speaker 02

  Uploading.

• 28:55Speaker 00

  Really?

• 28:56Heather Berlin?

  I would, yes, I'd be such, because of a fear of death.

• 28:58Speaker 00

  Okay, well, I think you just, I think that that's, this is, it's the oldest challenge in the book, right? Whether it's religion or, you know, the new religion of tech and mind uploading, ultimately, there's a lot of human motivation in trying to to deal with the prospect of death. And if I I'm not saying that I've got my head around it at all, and hopefully it's still a long way away, but I rather suspect the best strategy, instead of trying to live forever, whether whether it's in a biological form or in some silicon cloud, is actually to just get our heads around that not existing is okay.

• 29:35Heather Berlin?

  No, I'm not.

• 29:36Speaker 02

  And being in the here, so this is where the self can also be dysfunctional, right? Because as you said, we always project ourselves in the future. Either we live in the past, what I should have said, and why did she do this to me? Or we project ourselves in the future. And what Buddhism tells us, that by the way, of course, also denies the reality of self as as as usually constructed, that it's really important to be in the here and now, not in the past, not in the future, but now.

• 30:03Heather Berlin?

  All we have is this moment. I think my fear of death has ultimately led me to this concept of really appreciating and being aware of every moment, of this, the taste of food.

• 30:12Speaker 02

  In the here and now.

• 30:13Heather Berlin?

  In the here and now, like being present, it's actually made me much more present.

• 30:16Speaker 00

  While still planning for the future a little bit.

• 30:18Heather Berlin?

  A little bit, a little bit, yes.

• 30:19Speaker 00

  Like the present is a nice place to visit, but it's it's not a great place to hang out forever. You know, we do still need to worry about the future. I don't think it's, so there is a lot of resonance with with Buddhism, actually, in all of these views about perception as construction, but especially the nature of the self. There are so many resonances. I think the challenge is in accepting and and learning from those resonances. But in the real world, we we can't live completely in the present all the time.

• 30:44Heather Berlin?

  It's a balance.

• 30:45Speaker 00

  Appreciate it.

• 30:46Heather Berlin?

  Yes, you can, I mean, if you sit there meditating all day, you know, you'll go hungry, you won't have water, I mean.

• 30:50Speaker 02

  But there's still books, I can show you, I've invited many of these famous monks, you still have to plan a year ahead to get them to come to talk.

• 30:58Heather Berlin?

  So, speaking of the future, I did want to get into just finally, you know, A I and kind of the future. Now, you've mentioned this mind uploading, like I I don't think that, you know, that is a way to survive death or to keep conscious, because even if you can copy every connection of my brain and you put it in in in silicon, it might it might have consciousness, like a neuromorphic computer, it's based on my brain, but it wouldn't be me, it would be a different thing, or else I'd suddenly be uploaded into it, right? If if you created it right now, based on my brain, it would be something else, maybe sort of like me, but it wouldn't be this consciousness, right?

• 31:30Speaker 00

  What do you mean by this consciousness?

• 31:32Heather Berlin?

  The one that's in this physical, biological, well, let me ask you a question. Does it matter? Do you think that the substrate that consciousness is instantiated in matters?

• 31:44Speaker 00

  There's two, I I want to be careful to pull apart two things.

• 31:48Heather Berlin?

  Okay, two separate questions.

• 31:49Speaker 00

  The first one is really, it's the subject of a very famous thought experiment by Derek Parfit, the teletransporter thought experiment. So forget about uploading to a computer at the moment, just imagine that I have the technology to recreate the exact physical structure of your brain somewhere else, maybe in London.

• 32:07Speaker 02

  SpaceX, beat me up.

• 32:08Speaker 00

  Yeah, it's it's that. So now we now have to worry about how whether computers are sufficiently similar. I'm going to recreate you down to the molecular atomic quantum level.

• 32:17Heather Berlin?

  Okay.

• 32:18Speaker 00

  Now, it would be very hard to deny that if I do that, the the Heather that's recreated somewhere else, I mean, that Heather will will be you. It will, you know, there'll be something there that will experience being Heather Berlin, that will have the same memories, the same plans for the future, might be slightly surprised to be in London rather than New York. So you'll start to diverge. But at the moment of creation, there'll be another Heather that will feel exactly like you. And so the question is, you know, which is the which is the real Heather? And the answer is, well, they're both real Heathers, and neither are the real Heather. It's it's a process which has has diverged, and as time goes by, the the Heathers will will will diverge. So that's one, one of the topics.

• 33:01Speaker 00

  The other one is, is I think, the question that's facing us with much more urgency now, because we can't build these teletransporters, but people are already wondering whether artificial intelligence systems, AIs, can be not only smart but also conscious.

• 33:19Speaker 00

  And if that's the case, if consciousness is the kind of thing that AI is made out of silicon can have, one of the consequences of that is the idea of mind uploading, even though it's still not practical now, becomes at least conceptually possible.

• 33:38Speaker 02

  Anil, can you distinguish for the audience between intelligence and consciousness?

• 33:43Speaker 00

  Very different things. I mean, I think this is this is a super important distinction. I think we probably all are on board with making this distinction.

• 33:52Speaker 00

  One of the problems in all these debates about AI is that we tend to conflate the two concepts. Intelligence is very broadly about doing the right thing at the right time, not the morally right thing, but the appropriate thing to fulfill a goal.

• 34:08Speaker 02

  It's doing, it's behaving.

• 34:09Speaker 00

  It's doing, it's behaving, it's external behavior manifest in some way, solving a problem. That's intelligence. Consciousness is fundamentally about being. Now, I think our colleague and friend Julio Tinani puts it like this. Consciousness, going back right to the beginning of our conversation, is any kind of experience at all. There are many ways to be conscious without doing anything especially smart at the time. We already distinguish between consciousness and language and so on. So they are fundamentally different things. Now, that doesn't mean that unrelated in living systems, evolved systems, it looks like there's a very close relationship. You have to have certain kinds of conscious experience,

• 34:54Speaker 00

  you need to have certain kinds of cognitive capability. And I can't experience regret unless I have the cognitive infrastructure to imagine.

• 35:04Speaker 02

  But they're also much more basic, like toothache.

• 35:06Speaker 00

  Yeah.

• 35:07Speaker 02

  When you have a god awful toothache, there's nothing smart or intelligent about a really molar pounding away, yet it fills your entire mental.

• 35:15Speaker 00

  Absolutely.

• 35:15Speaker 02

  Life.

• 35:16Speaker 00

  Yeah, so they're very different. So the one problem is if we if we think that consciousness and intelligence are somehow so intimately bound that you can't have one with the other, then it becomes natural to think that as AI systems get smarter, that at some point the inner lights come on and they become conscious as well. But I think that's an insight into our own psychological biases and not an insight into things as they are.

• 35:42Heather Berlin?

  We anthropomorphize, right? And we, because it acts and sounds as if it's conscious, we assume it is, but that doesn't necessarily mean that it is. I've used this term like I'm a biochauvinist. You don't like this term because it has a negative connotation. But I think we're on the same page of that there's something in particular about being a biological being embodied that is tied in with being conscious. Do you think that these AI systems could be conscious? What would it take to get them to be conscious?

• 36:10Speaker 00

  So I share with you, not the chauvinism, I think, I don't like chauvinism because it sort of connects like a belief that you have that you don't have a good reason for having.

• 36:20Heather Berlin?

  What did you call a biological naturalism?

• 36:22Speaker 00

  The term in philosophy that it's been around for a long time. I think John Searle was the person who first used it in the sense that we now think of it as biological naturalism. And this is, this is the claim, and it's only a claim, that properties of living systems are necessary, but may not be sufficient for consciousness. And I do think this is like, what, this really is kind of key question for for our time because the metaphor of the brain as a computer has been so powerful, so influential, and it's led to a lot of insights, like all models can be useful, but it's become somehow rarefied. So people, many people, think implicitly or explicitly

• 37:05Speaker 00

  that the computer is not only a good model for aspects of the brain, but that the brain literally is a computer.

• 37:12Speaker 02

  Particularly of young people have these very emotional intimate relationships, whether that was a robotic therapist, ChatGPT, or with romantic friends, right, they're all these services where you can also describe where in some sense you see it doesn't matter whether academics like us discuss are we really conscious or not. They behave increasingly, they interact with them increasingly. They say it's easier to interact with my ChatGPT therapist than with an actual person. I have more friends in my ChatGPT or Claude or whatever. A friend, he understands me better than any human does.

• 37:49Speaker 00

  Yeah, so Heather will be out of a job soon.

• 37:51Heather Berlin?

  Not really, because, I mean, I might even get into it, but but there's a lot of problems with relying on these AI systems for psychological health. They don't, they don't feel, they might give you.

• 38:00Speaker 00

  That's the question, right?

• 38:01Heather Berlin?

  Well.

• 38:02Speaker 02

  Do they or do they not?

• 38:03Heather Berlin?

  They can give you the most rational. I'll just give you one very quick example. Like I have a patient, he's a perfectionist and wants to get straight A's, whatever. So he said, I asked ChatGPT, what should be my schedule, my study schedule, to be most efficient to do well on my finals? And it gave this ridiculous schedule. What it didn't take into, and then he was so anxious if he diverted from that. But the problem, it didn't take into account that a human brain needs rest and we need for psychological well being. And you know what, some days you have really good days and some days you wake up, you don't feel so good, you're not as sharp, you can't do as much. You can't, so the most rational, logical schedule of studying doesn't make sense with the human feeling.

• 38:37Speaker 02

  Oh, so it's a bad therapist.

• 38:38Heather Berlin?

  Maybe you hung over from the night before.

• 38:40Speaker 02

  So it's a bad therapist, but the fact is more and more people are using it.

• 38:43Heather Berlin?

  It can be dangerous.

• 38:44Speaker 00

  You're absolutely right, but I think what this highlights is the importance of separating two issues. So one issue is very hard to decide on, which is, is it even possible that AI systems could be conscious? Now, my suspicion is not, and the more we realize the differences between brains and computers, you pointed out that if you look at a brain, you don't see anything like the clean separation between hardware and software that's absolutely foundational to how computers work. So my suspicion is that, in fact, you won't find consciousness as a point in the space of possibility for silicon based AI. But we don't know for sure, and opinions can reasonably differ on this.

• 39:28Speaker 00

  But then that's very separate from this other problem that regardless of what academics natter on about, people will project consciousness into these things and they have a real social impact. There's a design question here. Regardless of the actual facts of the matter, should we be developing systems that encourage their users to believe that they are conscious? Now, it's like all such questions, it's not super simple. There may be some use cases where it's no bad thing. It encourages a fluency of interaction. But I think as an overall strategy, it's very, very dangerous.

• 40:07Speaker 02

  It may not matter, do you remember, well, you talk about the movie Ex Machina, right? Where the idea is, you know, this robot looks like a robot. She has a beautiful face, but her body is a robotic body, but it still doesn't matter, right?

• 40:19Heather Berlin?

  It does, well, just like the illusion. Listen, just like the illusion we have with, you know, there's two different colored gray squares on this checkerboard illusion and we see them as different, but they're exactly the same. Even though you know consciously that they are the same, you can't get over your illusion.

• 40:35Speaker 02

  And so you still fall in love.

• 40:35Heather Berlin?

  So you still fall in love.

• 40:42Heather Berlin?

  At the end of every episode, we ask a perception box question of our guests. This week, Kristoff is going to ask the question of us. So what is the question?

• 40:52Speaker 02

  Anil and Heather, so there's a skill that we all have that we learn early on called mind reading, right? So I infer that maybe because the way you look at me or smile at me when I say something you like me and you think I'm a reasonable person to converse with. Have there been noticeable situations where you just totally misread in some dramatic way what you assumed another person, for good or for bad, had assumed about you? Because this happens very often, right? That we assume she means, she probably thinks this about me, but in fact, she thinks the very opposite.

• 41:24Speaker 00

  It may be happening right now, such an example.

• 41:30Speaker 02

  Oh, that's very meta, that's very she, she.

• 41:34Heather Berlin?

  You thought I liked you, Kristoff, this whole time, but little do you know.

• 41:39Speaker 00

  It is a tough one, it's a tough one. I think it's very plausible. I'm trying to think of examples which which which remain.

• 41:47Speaker 02

  Like, I mean, examples I have, which we all must have. Someone doesn't, you text and they don't text for half an hour and then you go into these loops, oh, they don't like me anymore. Oh, there's something happened, they've cut me off, I mean, all sorts of things.

• 41:59Speaker 00

  I'll tell you an example, it's not a specific individual, because I wouldn't want to do that, but I had the kind of experience like this when I moved to California in the early 2000s. And I remember living there and going shopping and everyone wishes you a lovely day and, you know, asks you what plans you have for the weekend. And as a Brit, I found this extremely awkward. And like, I don't know, you know, how are you? I'm fine. And I was trying to figure out what they really meant. And it's not that I completely misread them, it's just that I didn't really know. In fact, I assumed they didn't really care and it was just something that Californians said.

• 42:34Speaker 02

  A habit.

• 42:35Speaker 00

  As a habit, just as a, yeah, yeah, just as a complete habit. But the longer I lived there, I realized that they kind of do care. It's not that they really care. You don't need to tell them all the details of what you have in mind, but it's not that they don't care. So I think that was a calibration towards a middle ground, which is one reason I'm very fond of California. I don't know about New York, though, I don't think they care so much, yeah.

• 42:58Heather Berlin?

  We are just more pragmatic, like if you ask for help, we'll give it to you. We're kind, we're just not going to waste our time with, you know, how are you doing? Like, we don't have enough time to spare, but but that doesn't mean you don't care.

• 43:09Speaker 02

  I'm always surprised how polite people are in New York. They all hold, you know, just walking around, they hold open the door and say, thank you and please. It surprises me because the image of New York is they're all hustle.

• 43:19Heather Berlin?

  We're just, we don't want to waste time, but that doesn't mean that we're not kind. You know, we're just efficient, it makes sense, trust me. So thank you, it's been a pleasure to have you, as always, great conversation.

• 43:31Speaker 02

  Thank you, Anil, it was enjoyable.

• 43:33Heather Berlin?

  I want to thank Anil Seth for joining us today. And if you'd like to learn more about your own perception box, spend some time this week answering the same perception box questions that we asked our guests and check out other questions on our website at unlikelycollaborators.com. You can also subscribe to our YouTube channel and watch the show or listen wherever you get your podcasts. This has been Science of Perception Box, created by Unlikely Collaborators in partnership with Pod People. I'm Dr. Heather Berlin.

• 44:05Speaker 02

  And I'm Dr. Kristoff Koch, thank you.