The limits of the utility function + Structure of a new science of consciousness

Transcript

00:00:02Nice. Hello everyone. A little bit of a different talk now. It's called we do a little philosophy, a little metaphysics. Hopefully it'll be a good pre-lunch thing. And I promise it'll be fun.

00:00:22So my talk is about the sort of, well the title is a bit of a mouthful. It's the limits of the utility function and the structure of a new science of consciousness.

00:00:36Does that work? I think we're...

00:01:30Just hang on a little bit. I have a little technical problem.

00:02:02Okay. Getting started here. So this is going to be a little bit of a different talk. It's going to be about consciousness.

00:02:27And the intention is to...

00:02:30wiggle loose the sense of possibility that I think that sometimes in AI alignment we sort of dig down into some deep sort of technical and political and social issues.

00:02:49And I guess I want to re-interject a bit of wonder into the great and a sense of possibility.

00:02:57So...

00:03:00This is me.

00:03:03So I actually moved to the San Francisco Bay Area many years ago, I guess 12 years ago now, to work on AI safety.

00:03:12And kind of bounced off the Lesseron frames.

00:03:16A lot of smart things, but they were speaking about consciousness and phenomenology in a much different way than I thought was needed.

00:03:27So...

00:03:28I wrote a book.

00:03:30I co-founded a research institute, left that last year, starting a new one.

00:03:35And I would say I have sort of a core result, which I will speak about in this talk.

00:03:43Some secondary results.

00:03:46And so my claim, the bottom line for this talk is trying to figure out AI alignment without a problem.

00:03:58And I think that the proper model of consciousness is very similar to getting to the moon without Maxwell's equations.

00:04:07We can absolutely do it.

00:04:09It's not outside the realm of possibility.

00:04:12But it painfully is needlessly complicated.

00:04:18So in that sense I'm much more hopeful.

00:04:22So...

00:04:23Okay, let's begin.

00:04:25What is this?

00:04:26What does this metaphysics stuff look like?

00:04:30First of all, just kind of talking about what is consciousness.

00:04:33It's phenomenology.

00:04:34It's the sum total of experience.

00:04:38It's what it feels like to be you.

00:04:40What it feels like to be me.

00:04:41And all that good stuff.

00:04:45And so right now, this form of knowledge is very much in the alchemy stage.

00:04:52That we kind of know, okay.

00:04:54Something here is...

00:04:56This is cool.

00:04:56There are some patterns.

00:04:58You know, if I eat a candy bar,

00:05:00I feel good. If I eat gravel, I feel bad. Okay, there are some correlations here. But it's not science. It's not chemistry. It is still alchemy.

00:05:14So, one thing that I think a lot about as a philosopher is to have knowledge about something, you need a place to put that knowledge. And I think this is an underappreciated fact.

00:05:28And the core question in consciousness research is what kind of container could fit all the sorts of knowledge that we would want to fit in it.

00:05:40So, we would want to fit, you know, knowledge about the rich diversity of human consciousness, all the different sort of, I mean, just like there's the periodic table of elements, and you can fit all of chemistry in that.

00:05:55We need a container that can fit all the richness of phenomenology. Not only human phenomenology, but dinosaur phenomenology, alien phenomenology, and of course, AI phenomenology.

00:06:09So, that's a big, tall challenge. And we'll talk a little bit about that today.

00:06:19So, there's a big question. How important is this study?

00:06:25what's at stake? Is it kind of a side question or is it the main question?

00:06:34Nick Bostrom and Scott Alexander have talked about how we don't want to end up

00:06:39building a Disneyland with no children. That this rich, beautiful technological

00:06:47wonderland with no consciousness. That what's the point if we build

00:06:54a Disneyland with no people and there's no one there to actually enjoy it.

00:07:00And so in general, where I'm coming from is I think it's a, if consciousness is

00:07:07sort of the domain that valuable stuff lives in, it's strange to not be

00:07:15energetically exploring the structure of this domain. Like let's figure out

00:07:21this exciting thing.

00:07:24And it's pun intended here, to not reverse engineer consciousness is to leave

00:07:35value on the table. Apologies for the pun, but there it is. And I also think that

00:07:44there are some pretty important path dependencies here. That okay, if we

00:07:50figure out consciousness before we get to AGI, we could have a better understanding

00:07:54of consciousness. We could live in a much different world rather than if we

00:07:58figure out if we get to AGI before we live in consciousness, or before we fully

00:08:02understand consciousness. So I think knowledge is good and I think knowledge

00:08:08about this is possible. And so I think it's important to pursue.

00:08:15So I definitely want to give deep respect to Les Strong. I think that there's a lot of wisdom there.

00:08:24There's a lot of sort of skill in sort of the Bayesian reasoning. And I'm currently kind of

00:08:35digging into the neuroscience of various forms of notation, active inference and

00:08:40predictive coding.

00:08:41So definitely it's a beautiful frame. But I also think that not everything fits in this frame of

00:08:51Bayesianism.

00:08:52So the Bayesian formula.

00:08:54and aesthetic tends to be focused on data and updating probabilities rather than structure

00:09:02and uh i i thank rob knight who's here uh back there for that insight thank you rob

00:09:09um and uh and i think that when when you're sort of exploring the the possibility of a new science

00:09:18of consciousness uh you're exploring the possibility of finding deep structure in the

00:09:24similar to how it turned out there was deep structure um behind electricity and lightning

00:09:32and load stones and and all that um so it's a it's a question of what kind of tool should we

00:09:38we try to apply here um and i i guess i also with respect um i would call out the uh perhaps the

00:09:50overuse of the term utility function uh we're kind of quick to say well

00:09:54this fits my utility function or my utility function is is that but um i

00:10:00we're kind of replicating existing parts of our brain in a kind of illegible way.

00:10:07That way, I don't think it's ground truth.

00:10:10I think it's just sometimes just words.

00:10:13So that's my dig at the utility function.

00:10:21So there's this big question.

00:10:24What do we get if we...

00:10:26Okay, so if we all agree that consciousness research could be important and there could be progress on this,

00:10:34what do we get if we sort of put resources in?

00:10:38And by resources, I mean money and I mean smart people.

00:10:42And both are very important.

00:10:46And especially relevant to today,

00:10:51how can it help us with this AI alignment issue?

00:10:56And I agree that it is a big issue.

00:10:58It's perhaps the most pressing issue.

00:11:03So I have some claims.

00:11:07Nate Soares very eloquently put it that there's this central question.

00:11:13What is a mind and how do you align one?

00:11:18And I don't really see a way to understand that without understanding neuroscience,

00:11:25understanding consciousness, and so on.

00:11:26So I would say that there are sort of two types of knowledge here.

00:11:33There's specific knowledge about how the brain works,

00:11:37kind of what evolution has thrown together.

00:11:40And then there's general universal principles.

00:11:43So these principles would be true universally in humans, in dogs, in dinosaurs, aliens, AIs.

00:11:54So it's sort of just like...

00:11:58So with Maxwell's equations, the equations of electromagnetism,

00:12:03we were able to build cool stuff.

00:12:05And with more cool stuff,

00:12:07we were able to better characterize the equations of electromagnetism.

00:12:12Theory and practice go together.

00:12:15So I think understanding the brain helps us understand the mind.

00:12:18Understanding the mind helps us understand the brain.

00:12:23Concretely, I think we can expect outputs

00:12:26to be human interpretability and alignment,

00:12:30better in our technology and intelligence enhancement,

00:12:35and a little bit more speculatively avoiding what I would call substrate leaks,

00:12:40where we sort of transfer to a new substrate that does not have consciousness.

00:12:45I think that's a complex topic, but a real danger.

00:12:51Okay.

00:12:53Most then slide.

00:12:55We don't need to worry.

00:12:56We don't need to worry too much about the details.

00:12:58But I think basically there are two lineages in sort of understanding minds.

00:13:05And one of the lineages is sort of to understand what a system is feeling,

00:13:14understand what it's computing, understand what the bits are doing.

00:13:18And a mind is a certain kind of algorithm,

00:13:21or consciousness is what an algorithm feels like from the inside.

00:13:27I don't take that view.

00:13:30And my reason is that this is inherently arbitrary to sort of map,

00:13:38try to map computations to physical system.

00:13:42There is no fact of the matter as to what my, like,

00:13:45which computations my brain is performing.

00:13:48It's messy.

00:13:51I'm more of the atoms sort of a mind is the right kind of physical system.

00:13:56So, yeah.

00:13:56So, this is definitely an area where we can have live debate.

00:14:06Shifting to what I believe and what I built.

00:14:10First of all, I'd highlight two core beliefs.

00:14:15One is this idea of monism that I think the strongest position in sort of beginning to understand consciousness,

00:14:26is to understand the world.

00:14:27that there exists one thing and it casts two shadows.

00:14:33So this one thing has a sort of,

00:14:36we can call it a projection, a mathematical projection

00:14:39that looks like physics.

00:14:41And it also has a second mathematical projection

00:14:45that looks like phenomenology,

00:14:46that looks like consciousness and qualia.

00:14:49And the cool thing, if we do that,

00:14:53and I'm a big proponent of this,

00:14:55anything we can tell about the structure of one shadow,

00:14:59we can

00:15:00also apply to the other shadow.

00:15:03I think that this is a dramatically underused insight.

00:15:10And also we can just kind of inherit a bunch of stuff from physics.

00:15:15And this stuff is really cool and lets us solve a lot of problems.

00:15:19For example, just kind of a technical comment, but applying Noether's theorem to phenomenology.

00:15:27And the other belief is just there is an answer.

00:15:32And by there is an answer, I mean there is a mathematical answer to consciousness.

00:15:38There is a way to create a mathematical representation of an experience.

00:15:43There exists a mathematical object that represents what it feels like to be me, or what it feels like to be you, or you, or you.

00:15:55You know, we can be a little agnostic as to how to do that.

00:16:00There are some smart people working on that.

00:16:03Giulio Tononi is one.

00:16:05And Dalton, and I'm not going to be able to put down his last name, but a very intelligent fellow with an Indian last name.

00:16:17So, I think splitting the problem of consciousness into two is really powerful.

00:16:24How do you create the formalism?

00:16:26How do you create a mathematical representation of an experience?

00:16:30And this is a fascinating problem.

00:16:33I'm focused on the latter half.

00:16:35How do you interpret a formalism?

00:16:38If I could hold up a formalism, and we can call it a very high dimensional shape, or something like that,

00:16:47then how do we begin to tell what it means?

00:16:50How do we extract information from it?

00:16:52And in the end, it's just a question of how do we do that?

00:16:54And how do we do that in sort of a relevant way?

00:16:57So that's what I wrote my book about.

00:17:01And my conclusion kind of gets into the symmetry theory of valence.

00:17:08So, next.

00:17:10Here we go.

00:17:11So, this is, I would say this is my most important result.

00:17:19And it's a little hard to fit it on one slide, but I'll try.

00:17:25So, first of all, split the problem of consciousness into how do you create a mathematical representation of an experience?

00:17:35And then how do you interpret that mathematical representation?

00:17:40So then you find the simplest place to start.

00:17:44The absolute, I mean, the term that I, someone used, I stole it for myself, is what is the C. elegans of qualia?

00:17:54The simplest model organism.

00:17:58And so, I decided that was emotional valence.

00:18:02It's the pleasantness or unpleasantness of an experience.

00:18:06It should be fairly well defined across all experiences.

00:18:10And so, we can think of two domains.

00:18:17The first domain is math.

00:18:20The second domain is consciousness.

00:18:21And pleasantness in the consciousness domain maps to what in the math domain?

00:18:30That becomes what the core question is.

00:18:35And borrowing from, borrowing pretty deeply from physics, my hypothesis is that it's symmetry.

00:18:45That if we have a mathematical representation of an experience,

00:18:50the symmetry of the experience or symmetries represent how, or like exactly correspond to the pleasantness of an experience.

00:19:00So, this happens to be kind of a formal mathematical way of saying pleasure is harmony in the mind.

00:19:10And the coolest thing is that this is testable.

00:19:15So, if we can kind of load up a special,

00:19:19pleasurable states,

00:19:21such as MDMA sessions or jhana meditation or things like this,

00:19:28then we can check, like, let's make a metric for harmony.

00:19:32And do these things have more harmony, more literal mathematical harmony?

00:19:37So, that's in progress.

00:19:40Working with Robin Carr-Harris on getting some data for that.

00:19:46So,

00:19:50it's a little hard to say, okay, you know, everyone needs to pay attention to this.

00:19:57But I would

00:20:00say that if there is going to be progress on consciousness, on understanding what is this consciousness stuff, how do we turn it into a science, how do we turn it into a system, how do we make it predictive, and so on, this is the path.

00:20:20I will, like, if there is a path, I think this is it. So that's my claim here.

00:20:31And in terms of just saying a few things about AI alignment, I would tell the story about three domains.

00:20:43Just kind of this, back to the idea of to have knowledge about something, you need a place to put that.

00:20:50And the question is, I think one of the core hidden questions in AI alignment is what kind of thing is a human?

00:21:09And I don't think that our current stories are navigation grade about that.

00:21:16So a navigation grade story says, okay, we have a whole set of human beings.

00:21:20We have all the information that we need to preserve the good stuff.

00:21:25And I don't think we understand what is beautiful about humans good enough that we can preserve the good stuff.

00:21:36So the story that I tell is, okay, there are three domains.

00:21:40There's the physical domain, the domain of consciousness, and the domain of computation.

00:21:47And what, like, the kind of thing a human is,

00:21:50what kind of thing a human is, is we're a projection into these three domains.

00:21:56We're a specific way that we project into these three domains.

00:22:01It's like we have a special signature in each of these domains.

00:22:06And to understand what kind of thing a human is, is to understand these three signatures.

00:22:13Likewise, to understand what kind of thing an AI is,

00:22:18we would also understand the three signatures of the AI.

00:22:23And different AIs will have different signatures.

00:22:27And I think we can sort of understand, okay, what makes a better projection or a worse projection?

00:22:38Or what, like, where's the beauty in humans?

00:22:42Can we figure out, like, okay, humans have something very unique

00:22:47about how we project into these three domains of physics.

00:22:51about physics, consciousness, computation.

00:22:57Um, so, uh, some closing observations here.

00:23:01Um, I'm, I think I'm much more optimistic about AI alignment than, uh,

00:23:08than many people that I've had conversations with here.

00:23:12And I think part of that is, I think there are $20 bills on the sidewalk.

00:23:16And there's knowledge about reality.

00:23:21That we don't have.

00:23:21And that actually wouldn't be that hard to have.

00:23:25And, um, you know, we're, we're sort of at a, at a place of low information.

00:23:30And this, this is a scary place to be.

00:23:33Uh, but there are things we can do to, to improve that.

00:23:39Um, I also think that, uh, you know, there's, um, there's a question.

00:23:44And I think this is, uh, more of a question for the, both the AI alignment folks

00:23:51and the AI, uh, kit buildings folks.

00:23:53That, uh, you know, if, um, if Sam Altman, uh, fires up GPT-6,

00:24:00and he's, he's the first one on there.

00:24:02It's, it's just a fresh, uh, fresh run.

00:24:06And, uh, he just really wants to know what is consciousness.

00:24:11How does this work?

00:24:12Um, how can we preserve the light of consciousness?

00:24:14And, and so on.

00:24:16Um, and then he, he types that into GPT-6.

00:24:21And when he gets an answer, then sort of, how did that work?

00:24:26Um, what kind of behind the scenes stuff happened, uh, such that it,

00:24:31it was able to give a good answer to that.

00:24:34And, um, sort of to, to collect the, you know, various, uh, theories of consciousness

00:24:39and, and evaluate them and, and so on.

00:24:42Um, I think that's a fascinating question.

00:24:45And I, I hope that the AI people are asking that question.

00:24:49Uh, and I hope the AI alignments are asking that question.

00:24:51I think that the AI alignment people would be interested in that question.

00:24:54Uh, and then I, of course, the, the follow-up question was, would be, well, what did GPT-6

00:25:00say? So, that's also a good puzzle.

00:25:06So, this is it. I'm starting a new thing, the Symmetry Institute shoestring budget.

00:25:16But if you want to donate either, you can do that.

00:25:22And, yeah, like I, also just if you're interested in talking about this stuff, come up and say hello, and I'm glad to chat.

00:25:32Thank you.

00:25:57Thanks for your talk.

00:25:59In the one that you discussed, your hypothesis and how it's testable, you talk about looking for symmetries or compressibility in data.

00:26:10I presume that's like EEG, fMRI, that type of data.

00:26:14So, I'm wondering how do you equate that to the mathematical representation of .

00:26:20So, you can equate data to, like symmetry in data to symmetry in the mathematical form that is replaying that you talked about.

00:26:27Yeah. So, that's a very good and deep question.

00:26:31It's a question of what can we measure, like can harmony in the brain be a proxy for symmetry in the mind?

00:26:43And what metric do we apply to measure harmony in the brain?

00:26:47Like...

00:26:48Yeah.

00:26:50So, I'll give you what we've tried.

00:26:53There's a way to process fMRI and MRI and DTI data in order to infer a...

00:27:05So, a bunch of math terms here, or neuroscience terms, to infer a connectome's eigenmodes, basically the resonances of a brain.

00:27:15And then we built an algorithm, and this is my ex-co-founder.

00:27:19Okay.

00:27:20And we built an algorithm to measure the pairwise, consonance, dissonance, and noise between each eigenmode, between each resonance.

00:27:32And I'm thinking that, was that, okay, we don't exactly know what parts of the mind are contributing to consciousness, but this should be a pretty good rough proxy.

00:27:46Thanks.

00:27:49Okay, yeah, my question is, do you have any approximate timelines for how long you expect this research to become useful for making AI alignment more sort of useful in the sort of post-Utopian, in the Utopian future, hopefully?

00:28:07And would this sort of also improve the case for taking a pause on like development of AI algorithms, possibly hardware caps, etc. that some EA people are discussing just for solving alignment?

00:28:28So would this sort of help your project as well, feasibly?

00:28:33Yeah, no, that's a good question.

00:28:37I, so separately from my research, I'm a little worried about the slowdown considerations that Vitalik mentioned in his talk.

00:28:46So I wouldn't, I would humbly not advocate for a pause, I guess.

00:28:54But yeah, in terms of timelines, I think I would look at two things.

00:29:00One is, can we figure out how to spot consciousness in the world?

00:29:07And I think that that kind of revolves around the technical topic of the binding problem, or the boundary problem.

00:29:17Like, how to, like, there's a discussion here.

00:29:24So, yeah, like, first of all, what do we need to solve in order to understand, okay, is this computer conscious?

00:29:33Is that computer conscious?

00:29:35Is this computer architecture more conscious?

00:29:37And so on.

00:29:39So some people would think that timeline, I'm going to say it's probably going to be solved in five years.

00:29:47And then the second issue is, okay, if we know something is conscious, is it a pleasant consciousness?

00:29:56Is it suffering?

00:29:58And I would

00:30:00just sort of refer back to the symmetry theory of valence.

00:30:03This is, I think, a hypothesis that we can deploy, you know, once we can kind of point to consciousnesses in the physical world,

00:30:12once we have a definition of where the boundaries of the consciousnesses are, then we're ready to talk about the valence.

00:30:21Okay.

00:30:25One more question, and then we'll break for a lightning talk and lunch.

00:30:34Awesome talk. I'm curious, like, how would you go about trying to spot emergent almost like consciousness

00:30:42in more advanced large language models, for example, or AI in general, and like, do you think connected to that,

00:30:49that there's some risk to have, you know,

00:30:52LGBT 6 or 7F cyber consciousness?

00:30:55Yeah, I'm curious.

00:30:57Yeah, super question.

00:31:00So, I would say that, first of all, and this is a matter of technical metaphysics,

00:31:08so we can put on our philosopher hats, and I would say that the question of AI consciousness

00:31:17and the question of computer consciousness,

00:31:19they sound similar to our own.

00:31:22To our ears.

00:31:22But metaphysically speaking, these are very different things.

00:31:25That a computer is something that is embodied in some form or fashion.

00:31:29It's made out of stuff similar to the stuff that we're made out of.

00:31:35And so, in that sense, I suspect that talking about computer consciousness is less likely to lead to confusion,

00:31:46whereas talking about AI consciousness, it's like talking about Walmart consciousness

00:31:49or East India Corporation consciousness.

00:31:52Or country of Montenegro consciousness.

00:31:55And I think that we get into confusion pretty easily there.

00:32:01I do suspect that, I mean, you know, when we're talking about complex electromagnetic devices,

00:32:11such as GPU clusters, creating very complex patterns,

00:32:17and we can kind of apply different metrics to, you know, is this pattern complex,

00:32:22or is it not, and so on.

00:32:23Is it creating complex patterns in the electromagnetic field, and so on.

00:32:28I wouldn't be surprised if current computers aren't at least somewhat conscious.

00:32:34And I do expect that to grow.

00:32:40And I think that there's a, so, Harish, who I'm not sure is here,

00:32:46but he made a really good point in his talk,

00:32:52after his talk.

00:32:53He was talking about, like, what are the dangers of,

00:32:58so, first of all, there's sort of two classes of interesting confusion.

00:33:04One, what happens when something is conscious, but we don't afford it consciousness?

00:33:13Like, if a brain emulation is conscious, but we don't treat it as conscious.

00:33:18And that would get to a pretty dark place pretty quickly.

00:33:24And we can also flip that.

00:33:26What happens when something is not conscious, but, for example, like a large language model,

00:33:34it sort of parrots the motifs of consciousness,

00:33:39and might claim consciousness, and advocate for its right as a conscious being,

00:33:45but it isn't conscious.

00:33:48And, you know, what happens in that scenario?

00:33:50And I also think that that also leads to a pretty bad place.

00:33:56So, I think that it's just, it's very important to understand the ground truth

00:34:01of what's actually real here.

00:34:08I want to ask you about your statement about,

00:34:12would it be a pleasurable consciousness, or would it be suffering consciousness?

00:34:18And would we go with the paradigm that an AI can be self-aware and reprogram itself?

00:34:24Would that not make the question of suffering or existentialism obsolete,

00:34:31because it can reprogram its own dopamine emulation system,

00:34:37unlike us, who are locked into the biological reality?

00:34:43Anyways, and it would just not be able to form any other will then,

00:34:48because it could have the candy of the will immediately anyways.

00:34:53And would that be a good proof that consciousness is just emulated,

00:34:59if

00:35:00there is a will that's shown to us with text, but logically or rationally we could see if AI actually wants that, it can just rebron itself to getting it in another way or being satisfied with it without getting it in a real world.

00:35:17Yeah, so there's this question of if the AI can program itself, won't it just wear out?

00:35:25And I'd split this into two scenarios. One is sort of the question in technical AI alignment.

00:35:35And that isn't the wheelhouse of this talk. I do think that sort of if the AI can control its reward function,

00:35:43yeah, we get into sort of dangerous territory, potentially.

00:35:49The other thing, the other aspect, like speaking specifically about consciousness,

00:35:55I would offer that it's an interesting contingent fact about how brains work, that we seek out pleasure.

00:36:06I mean, Freud has this pleasure principle. He says that, you know, humans seek out pleasure.

00:36:13It's interesting to note that that's like roughly true, but not always true.

00:36:19We don't always seek out pleasure. And so there are sort of various ways to construct the basic human drive.

00:36:27Freud said we seek out pleasure. Carl Friston might say we seek to minimize short terms of prize and so on.

00:36:35I think like the danger that I would kind of point out is that, OK,

00:36:43humans, we seek out sort of the good neighborhoods of phenomenology, the tasty stuff in consciousness.

00:36:52We like being happy. But this is actually not a law of the universe.

00:36:59We could lose this. And we should be careful not to make AIs or computers that seek out the bad places specifically.

00:37:22Sorry, is that a question or is that a question?

00:37:27I think a question. Can you come up and I'm not sure my how.

00:37:31We'll take this as the one last question. So.

00:37:37So what do you think is the best way to figure out whether we think this thing have a consciousness?

00:37:43Yes or no? For example, animals may have consciousness.

00:37:46They cannot speak in our language. So we don't know that A.I.

00:37:52They can speak. They can reply in our language. So we may say yes.

00:37:57But do you think language is the best way or to fail our meditation or to think?

00:38:03So what's the best way to figure out this problem?

00:38:07Yeah. So this is this is kind of the core question.

00:38:11And so there.

00:38:13I would bring in a philosopher called Wittgenstein.

00:38:17Very, very cool guy.

00:38:19Had super strong dyslexia.

00:38:21So he could barely string two paragraphs together, but they were very good two paragraphs.

00:38:26So he he wrote about this in terms of language games.

00:38:32And he said that often when we're confused, we're using perfectly good words in ways they're just not meant to be used.

00:38:45So he gave the example like what time is it on the sun?

00:38:49Like what time zone is it on the sun?

00:38:51And that has no answer.

00:38:53It's it's like all the individual words are perfectly good.

00:38:57But when we apply that to to this novel context, we break language.

00:39:03And he says that most philosophical confusion is like that.

00:39:06We're trying to use these perfectly good words in ways that they're just not meant to be used.

00:39:13And I think that many of our attempts to infer consciousness will also fit this pattern that we can say, well, if if something claims it's conscious, surely it has to be conscious.

00:39:28But maybe if if it's just like a literally a computer program that says, you know, print F.

00:39:35Hello, I'm conscious.

00:39:37Well, I don't think that's conscious.

00:39:40And so I think that.

00:39:42I mean, this.

00:39:43This really, I think, is sort of the core reason why we need a system.

00:39:49We need a new branch of science.

00:39:52We need sort of to take the sort of alchemical combination of incompatible intuitions.

00:40:00intuitions that we currently have about what is consciousness and how it works and to make it consistent and necessarily so there's a there's a great quote by Eric and I'm gonna mangle his last name apologies Eric Schwitz-Gibble that he says that common sense is incoherent in matters of metaphysics if you're trying to be consistent you're

00:40:28gonna have to violate it somewhere because it is this kind of combination of sort of disjoint language games and when we when we try to make it into a system we're gonna have some surprises and much like I'm sure like chemistry and the history of physics has given us a lot of surprises that people took frankly generations to to get over like special and general relativity and especially quantum mechanics like people are really struggling

00:40:58to say okay like is this real and I you know I suspect that there will be some surprises when we formalize consciousness well thank you Michael for a very mindful and thoughtful talk before our lunch break which is now turned into lightning talk sessions so if you

00:41:28are one of the four lightning talk speakers please come here because