Latest / Daniel and Kelly’s Extraordinary Universe / How does consciousness emerge?

Transcript

How does consciousness emerge? · Daniel and Kelly’s Extraordinary Universe

• 0:05
  Speaker 1: On the podcast, we love to ask deep questions and
• 0:09
  Speaker 1: then explore the edge of our knowledge. Today we're going
• 0:12
  Speaker 1: to go a little bit meta and try to understand
• 0:14
  Speaker 1: the workings of our own minds, the us that is
• 0:18
  Speaker 1: asking those questions and yearning to understand the universe, from
• 0:22
  Speaker 1: biology to physics, and yes, sometimes even a little bit
• 0:25
  Speaker 1: of chemistry. Science gives us this incredible tool to consistently
• 0:29
  Speaker 1: build knowledge about the universe, but it's not clear that
• 0:33
  Speaker 1: it can tell us about ourselves. That it can help
• 0:36
  Speaker 1: us understand why we have a first person subjective experience
• 0:41
  Speaker 1: in the first place, why we have an experience at all,
• 0:45
  Speaker 1: why chocolate has a taste and pain has a negative valance,
• 0:48
  Speaker 1: Why there's something it's like to be me. How's that
• 0:52
  Speaker 1: different from what it's like to be a bat? Is
• 0:54
  Speaker 1: it like anything to be a rock or a puddle
• 0:57
  Speaker 1: or a star? Science, like all you and endeavors, is
• 1:00
  Speaker 1: limited by the senses that we use to interact with
• 1:03
  Speaker 1: the universe. It can help us understand what we sense
• 1:07
  Speaker 1: and find patterns and build models, but it can't ever
• 1:10
  Speaker 1: really tell us about the hard reality of the physical
• 1:13
  Speaker 1: universe that exists out there beyond our minds and our senses,
• 1:18
  Speaker 1: and that same limitation might apply to the other end
• 1:20
  Speaker 1: of that sensory pipe. Can a scientist use science to
• 1:24
  Speaker 1: probe the nature of the mind while trapped inside their
• 1:28
  Speaker 1: own mind? That might just be the province of philosophy.
• 1:32
  Speaker 1: Today on the podcast, we're getting philosophical and asking what
• 1:35
  Speaker 1: we know and how we can know about how consciousness emerges.
• 1:40
  Speaker 1: Welcome to Daniel and Kelly's Extraordinarily self Conscious Universe.
• 1:58
  Speaker 2: Hello, I'm Kelly Winnersmith. I study parasites and space, and
• 2:01
  Speaker 2: today I'm wondering if we live in a simulation.
• 2:05
  Speaker 1: Hi, I'm Daniel. I'm a particle physicist, and I don't
• 2:09
  Speaker 1: care if we live in a simulation. I want to
• 2:11
  Speaker 1: eat simulated dark chocolate.
• 2:13
  Speaker 2: So that's my question. Is it possible that we live
• 2:15
  Speaker 2: in a simulation? We're getting philosophical today, let's go all
• 2:17
  Speaker 2: the way.
• 2:18
  Speaker 1: Yeah, absolutely, it's possible. But it also seems like a
• 2:22
  Speaker 1: modern reflection on recent concepts about how computers work. You know,
• 2:27
  Speaker 1: one hundred years ago, people thought the whole universe was
• 2:30
  Speaker 1: like pipes and cogs, and these days we think it's
• 2:33
  Speaker 1: all computers. So it's possible, but we also have zero
• 2:37
  Speaker 1: evidence that it's a simulation, and dark chocolate tastes really
• 2:41
  Speaker 1: really good anyway.
• 2:42
  Speaker 2: That's true, and so does pad see You. I'm hungry today.
• 2:47
  Speaker 1: This is why we shouldn't report the podcast just before lunch.
• 2:50
  Speaker 2: That's right. I'm so glad that I have simulated taste
• 2:52
  Speaker 2: buds and smell receptors so that I can enjoy all
• 2:54
  Speaker 2: of those things very much.
• 2:56
  Speaker 1: But it is fascinating to think about where your consciousness were,
• 3:00
  Speaker 1: where your experience comes from, why it is that you
• 3:03
  Speaker 1: even have one while you can like or hate pad
• 3:06
  Speaker 1: see You, whether or not that's something that comes from
• 3:09
  Speaker 1: the meat in your mind, or whether it's just the
• 3:11
  Speaker 1: information circling through those neurons which could be uploaded to
• 3:15
  Speaker 1: the cloud.
• 3:16
  Speaker 2: Yeah. I mean, I think you'd have to be a
• 3:18
  Speaker 2: machine to not like TI food in general personally, but
• 3:20
  Speaker 2: I could be wrong. But I'm excited that so you
• 3:22
  Speaker 2: and I. We get a lot of emails from listeners.
• 3:26
  Speaker 2: You can send those emails to questions at Daniel and
• 3:29
  Speaker 2: Kelly dot org. And lately we've been getting a lot
• 3:32
  Speaker 2: of questions from folks who have been asking about consciousness.
• 3:35
  Speaker 2: I think we've been getting like one a week for
• 3:37
  Speaker 2: a while, and we decided it was time to jump
• 3:40
  Speaker 2: into the consciousness question, and you lucked out, and you
• 3:43
  Speaker 2: happen to know the perfect guest I do.
• 3:45
  Speaker 1: There's lots of really smart folks here at you See Irvine,
• 3:47
  Speaker 1: and I'm good at googling and finding them and persuading
• 3:50
  Speaker 1: them to come and talk to me on the podcast.
• 3:53
  Speaker 1: I have a lot of fun, and this is a
• 3:54
  Speaker 1: topic that's really close to my heart. I've been thinking
• 3:57
  Speaker 1: about it for a long time and also thinking about
• 3:59
  Speaker 1: whether it's something you really can ask scientifically. You know,
• 4:03
  Speaker 1: science is wonderful and powerful, and of course we are
• 4:06
  Speaker 1: advocates for science here on the podcast, But just because
• 4:08
  Speaker 1: it's a tool that does let us build knowledge about
• 4:10
  Speaker 1: the universe doesn't mean it's a tool that can answer
• 4:13
  Speaker 1: every question. Right. You can't answer the question like should
• 4:16
  Speaker 1: I get out of bed in the morning? Why do
• 4:18
  Speaker 1: people eat white chocolate? Anyway? Right, Not every question is scientific,
• 4:23
  Speaker 1: and so there are things about the universe that we
• 4:25
  Speaker 1: can probe with science, and there are limits to what
• 4:28
  Speaker 1: science can answer. And I've always wondered how much we
• 4:31
  Speaker 1: can understand about consciousness itself using science. So I was
• 4:35
  Speaker 1: really excited to talk to a cognitive scientist who thinks
• 4:38
  Speaker 1: about this all day in a very serious way.
• 4:41
  Speaker 2: Yeah, I'm excited that consciousness is sort of at the
• 4:43
  Speaker 2: interface of science and philosophy, which makes it a really
• 4:47
  Speaker 2: fun topic to think about. And so I had a
• 4:49
  Speaker 2: lot of fun today with Megan Peter's chatting with her
• 4:51
  Speaker 2: about this.
• 4:52
  Speaker 1: Stopping before we dive into this wonderful conversation, we were
• 4:55
  Speaker 1: curious what folks out there thought about their own consciousness,
• 5:00
  Speaker 1: what ideas people have for why they even have an experience,
• 5:04
  Speaker 1: how these few pounds of goo in your brain somehow
• 5:07
  Speaker 1: generate all of this joy and frustration that we experience.
• 5:11
  Speaker 1: So I asked our cadre of volunteers to chime in
• 5:14
  Speaker 1: on the question how does consciousness emerge? If you'd like
• 5:19
  Speaker 1: to provide answers for a future episodes, please don't be
• 5:21
  Speaker 1: shy to right to questions at danieland Kelly dot org.
• 5:25
  Speaker 1: So think about it for a minute. How do you
• 5:26
  Speaker 1: think consciousness emerges? Here's what our listeners had to say.
• 5:31
  Speaker 3: I think we have to really think about what consciousness
• 5:34
  Speaker 3: might mean before we can figure out where it comes from.
• 5:39
  Speaker 3: But if I had to make a guess, I'd say
• 5:41
  Speaker 3: consciousness is just energy sliding into different forms as it
• 5:46
  Speaker 3: rolls along from our brain.
• 5:48
  Speaker 1: Jews exponential, exponential extension of the emergence of life. I
• 5:52
  Speaker 1: don't think anyone knows human consciousness is an intersection of
• 5:55
  Speaker 1: the more mechanistic functions of the human brain and the
• 6:00
  Speaker 1: created soul. I think consciousness needs you to have both
• 6:04
  Speaker 1: theory of mind and the complexity to apply it to yourself.
• 6:08
  Speaker 4: Very slowly on a Saturday morning after a night out
• 6:11
  Speaker 4: down the town with several coffees.
• 6:14
  Speaker 1: Consciousness is fundamental and that we are as humans somehow
• 6:18
  Speaker 1: tapped into it. Wow, consciousness, that's a bigie. Well, I
• 6:23
  Speaker 1: think it's magic.
• 6:25
  Speaker 3: As the parent of twins who think I'm pretty much useless,
• 6:29
  Speaker 3: I'll say it that somehow it involves interaction with others.
• 6:32
  Speaker 1: Consciousness emerges when the penguin models in. I've always leaned
• 6:36
  Speaker 1: towards the idea that consciousness is an illusion.
• 6:40
  Speaker 4: I believe consciousness emerges when the brain's elegical florid of
• 6:42
  Speaker 4: plasma interacts with the noise and dangers in the brain.
• 6:45
  Speaker 2: Thanks the brain itself is a sensory organ that can
• 6:48
  Speaker 2: sense its own processing.
• 6:50
  Speaker 3: A pint of strong fresh coffee is a necessary but
• 6:53
  Speaker 3: not sufficient condition.
• 6:55
  Speaker 1: It's a very complicated dance in the brain. It's a
• 6:58
  Speaker 1: very slow process. I think consciousness takes many, many forms,
• 7:03
  Speaker 1: things that the brain is doing, neural inspiring, reaching out
• 7:06
  Speaker 1: a critical consistency, and an introduction number.
• 7:12
  Speaker 3: I don't think we can know how consciousness emerges until
• 7:15
  Speaker 3: we know what consciousness is and if we have it all.
• 7:18
  Speaker 2: Right, We've got coffee, penguins, magic, all kinds of amazing
• 7:23
  Speaker 2: things in there, and a large diversity of answers.
• 7:26
  Speaker 1: Mmmmm, And there's a thread here. I've noticed a lot
• 7:30
  Speaker 1: of people suggesting that somehow it arises from the complexity
• 7:34
  Speaker 1: of the brain neurons, reaching a critical number or you know,
• 7:38
  Speaker 1: all these interactions. To me, that sort of frames the
• 7:41
  Speaker 1: question but doesn't really answer it, right, Like, why is
• 7:44
  Speaker 1: it the complex network of cells can have a consciousness
• 7:48
  Speaker 1: emerge regardless of the complexity, Right, It's not obvious to
• 7:51
  Speaker 1: me that complexity is enough. And that's really I think
• 7:54
  Speaker 1: the focus of the question for today's episode.
• 7:57
  Speaker 2: I would personally love to know the evolutionary reason why
• 8:00
  Speaker 2: we have consciousness, how does it benefit us? And would
• 8:02
  Speaker 2: we expect to see other animals to have it as well? Yeah,
• 8:05
  Speaker 2: and anyway, there were so many exciting things we talked
• 8:07
  Speaker 2: about with Megan today.
• 8:08
  Speaker 1: Yeah, there's lots of really fun questions there, like where
• 8:11
  Speaker 1: are there other species of humans which actually were more
• 8:14
  Speaker 1: intelligent but didn't survive because our ancestors were more warlike
• 8:18
  Speaker 1: or more cannibalistic? Right? Or if you ran the Earth
• 8:22
  Speaker 1: experiment a thousand times, how often would you even get
• 8:25
  Speaker 1: intelligence or more or less intelligence? All right? Is this
• 8:28
  Speaker 1: a fluke or is it a common outcome? Man? I'd
• 8:31
  Speaker 1: love to know the answer to those questions.
• 8:33
  Speaker 2: And you're also getting into some of the complexity there
• 8:35
  Speaker 2: because you were calling it intelligence and where does intelligence
• 8:38
  Speaker 2: end in consciousness begin? And what is the difference in
• 8:41
  Speaker 2: those things? And it gets complicated pretty fast.
• 8:44
  Speaker 1: It does. And fortunately we have an expert to help
• 8:46
  Speaker 1: us dig through this. So let's jump right into our
• 8:49
  Speaker 1: conversation with Professor Megan Peters from the Cognitive Science department
• 8:52
  Speaker 1: here at UC Irvine. All right, so it's my pleasure
• 8:58
  Speaker 1: to have on the podcast Professor Megan Peters. She's an
• 9:01
  Speaker 1: associate professor here at you see Irvine in the Department
• 9:04
  Speaker 1: of Cognitive Science. She's a PhD from UCLA in cognitive
• 9:07
  Speaker 1: science and her research aims to reveal how the brain
• 9:11
  Speaker 1: represents and uses uncertainty. She uses neuroimaging, computational modeling, machine learning,
• 9:16
  Speaker 1: and neural stimulation techniques to study all of this, and
• 9:20
  Speaker 1: most importantly, she agreed to come on the podcast and
• 9:22
  Speaker 1: talk to us about what is consciousness and answer all
• 9:25
  Speaker 1: of our questions and fend off Kelly's poop references.
• 9:29
  Speaker 4: Ah, good luck, Thank you both for having me. It's
• 9:34
  Speaker 4: really a pleasure to be here. As an aside, my
• 9:38
  Speaker 4: PhD is in psychology with a focus on cognitive and
• 9:41
  Speaker 4: computational neuroscience. If that matters, So you can cut this
• 9:44
  Speaker 4: part out if you want to redo that intro.
• 9:46
  Speaker 1: No, we'll find out if that matters, we'll see Yes,
• 9:49
  Speaker 1: absolutely great, So let's dig right in. I want to
• 9:52
  Speaker 1: know And this is maybe the hardest question we're going
• 9:55
  Speaker 1: to ask you all day. How do we define consciousness?
• 9:58
  Speaker 1: We're going to be talking about it, We're going to
• 9:59
  Speaker 1: be arguing about We're going to be saying do rocks habit?
• 10:03
  Speaker 1: Do dogs have it? But it's kind of slippery if
• 10:05
  Speaker 1: we don't even know what it is we're talking about.
• 10:07
  Speaker 1: So what is consciousness? How do we define this thing
• 10:11
  Speaker 1: that we all sort of know intuitively but have a
• 10:13
  Speaker 1: hard time describing.
• 10:15
  Speaker 4: Super important question and leads me into one of the
• 10:18
  Speaker 4: first kind of technical things that we'll talk about today,
• 10:21
  Speaker 4: which is the distinction between are the lights on? But
• 10:25
  Speaker 4: is anyone home? Essentially, So we talk in the fields
• 10:29
  Speaker 4: of psychology and in the philosophy of mind about the
• 10:32
  Speaker 4: distinction between what's called access consciousness and phenomenal consciousness. So
• 10:38
  Speaker 4: these are terms that the philosopher and ned Block introduced
• 10:40
  Speaker 4: a while back, and the idea is that access consciousness
• 10:44
  Speaker 4: is about something like the global availability of information in
• 10:48
  Speaker 4: a complex system to allow that system to behave usefully
• 10:53
  Speaker 4: in its environment, to survive, to seek goals, to seek rewards,
• 10:57
  Speaker 4: to get food, to not be eaten by something. And
• 11:01
  Speaker 4: the distinction then is between you know, access consciousness, which
• 11:04
  Speaker 4: is information goes all over in the brain or in
• 11:07
  Speaker 4: the mind if there isn't a brain present. But the
• 11:10
  Speaker 4: difference then between that and phenomenal consciousness is that the
• 11:14
  Speaker 4: phenomenal consciousness has to do with the phenomenology of the
• 11:19
  Speaker 4: experience that the observer is having. So I'm going to
• 11:22
  Speaker 4: break that down that really just comes down to this
• 11:25
  Speaker 4: something that it is like to be you the qualitative
• 11:29
  Speaker 4: experiences that you have of the world. The fact that
• 11:32
  Speaker 4: pain hurts. It isn't just a signal like your tesla
• 11:36
  Speaker 4: would send to itself, like damage in my right front
• 11:38
  Speaker 4: tire or something like that. It's more like it has
• 11:41
  Speaker 4: qualitative character to it. There's something that it's like to
• 11:45
  Speaker 4: be in pain. If I whack you in the leg,
• 11:47
  Speaker 4: it's not just a signal that I've damaged the tissue
• 11:51
  Speaker 4: in your leg. There's more to it than that.
• 11:53
  Speaker 1: So are you saying that a tesla has access consciousness
• 11:56
  Speaker 1: because it notices damage in its tire, but it doesn't
• 11:59
  Speaker 1: feel pain, doesn't have anomenological consciousness, or if I miss.
• 12:02
  Speaker 4: Ansers, I wouldn't even go that far to say that
• 12:05
  Speaker 4: a tesla has any sort of access consciousness whatsoever. I
• 12:08
  Speaker 4: think that that's a higher bar there, But I think
• 12:10
  Speaker 4: that the distinction is important.
• 12:11
  Speaker 2: Right.
• 12:11
  Speaker 4: We can imagine a system that has all of the
• 12:14
  Speaker 4: hallmarks of access consciousness, like a fancy future tesla. Okay,
• 12:17
  Speaker 4: you know that has all of this global availability of information.
• 12:21
  Speaker 4: It enables flexible adaptive behavior. It can change based on
• 12:25
  Speaker 4: its context. It's not just going according to its programming.
• 12:28
  Speaker 4: You know that kind of thing. And yet there's nothing
• 12:30
  Speaker 4: that it's like to be that fancy future tesla. It's
• 12:33
  Speaker 4: just a zombie. It's just engaging in behaviors that are
• 12:37
  Speaker 4: useful for that organism, but there's no one home. So
• 12:42
  Speaker 4: someone being home aspect is also really important because we
• 12:45
  Speaker 4: can imagine in like, you know, not a future robot scenario,
• 12:48
  Speaker 4: but just a medical situation where you've got a person
• 12:52
  Speaker 4: who's in a coma or they're in a persistent vegetative state,
• 12:56
  Speaker 4: and they might wake up seemingly and have of sleep
• 13:00
  Speaker 4: wake cycles and maybe respond to external stimuli. But the
• 13:04
  Speaker 4: critical factor in deciding whether you know someone's home whether
• 13:07
  Speaker 4: to keep them on life support is whether there's anything
• 13:11
  Speaker 4: is like to be them inside. And the flip side
• 13:15
  Speaker 4: is also really important in the context of medical science,
• 13:17
  Speaker 4: because even if they don't exhibit any of the outward
• 13:21
  Speaker 4: signs or symptoms of being conscious, of having access to
• 13:26
  Speaker 4: that available information, of being able to behave in their environments,
• 13:29
  Speaker 4: someone still might be in there. They might be locked in.
• 13:33
  Speaker 4: And so it's this presence of phenomenal experience of someone
• 13:36
  Speaker 4: being home, so to speak, that I think is the
• 13:40
  Speaker 4: important thing to remember when we're talking about consciousness in
• 13:43
  Speaker 4: the context of biological systems or artificial ones.
• 13:47
  Speaker 2: So I'm still trying to wrap my head around the
• 13:49
  Speaker 2: two different kinds of consciousness. Is there a kind of
• 13:52
  Speaker 2: organism or a situation a person can be in where
• 13:54
  Speaker 2: they would have one but not the other to help
• 13:56
  Speaker 2: me sort of differentiate.
• 13:58
  Speaker 4: Well, For humans, the idea is that, presuming that you
• 14:01
  Speaker 4: don't subscribe to, you know, philosophical zombiesm that like, you
• 14:05
  Speaker 4: have no idea that I'm conscious, right, I exhibit all
• 14:07
  Speaker 4: the hallmarks, but you have no idea. But presuming that
• 14:10
  Speaker 4: you don't want to go down that rabbit hole, I.
• 14:12
  Speaker 1: Do actually in a minute.
• 14:15
  Speaker 4: In humans, it seems like they go hand in hand, right,
• 14:18
  Speaker 4: that if you have access consciousness, you have phenomenal consciousness
• 14:21
  Speaker 4: in general as an awake, behaving organism. But there might
• 14:25
  Speaker 4: be cases in specific scientific experiments where you can reveal
• 14:31
  Speaker 4: symptoms or cases or evidence of access consciousness in the
• 14:35
  Speaker 4: absence of phenomenal consciousness. So there's some very specific psychological
• 14:40
  Speaker 4: experiments that suggests that these are separable entities. One classic
• 14:45
  Speaker 4: example is a task that was actually attributed to and
• 14:49
  Speaker 4: was developed by a cognitive scientist here in my department,
• 14:52
  Speaker 4: George Spurling. So this is the classic spurling task where
• 14:55
  Speaker 4: you show someone an array of letters. There's like five
• 14:58
  Speaker 4: letters in a row, and there's rows of letters, and
• 15:01
  Speaker 4: you flash it very fast, and you ask people to
• 15:05
  Speaker 4: kind of give their impression of the overall array, like
• 15:08
  Speaker 4: did you see it? Did you feel like you got
• 15:10
  Speaker 4: the whole thing? And people say, yeah, I feel like
• 15:11
  Speaker 4: I got the whole array. That's fine. But then when
• 15:14
  Speaker 4: you ask them to report the whole array they can't,
• 15:17
  Speaker 4: But when you ask them to report a specific row
• 15:19
  Speaker 4: they can't. It feels like there's a distinction then between
• 15:22
  Speaker 4: this feeling that you've got phenomenal experience of the whole array,
• 15:27
  Speaker 4: but your access might be limited, and so ned Block
• 15:29
  Speaker 4: has famously called this phenomenal consciousness overflowing access.
• 15:34
  Speaker 1: It seems to me sort of the crucial distinction, right,
• 15:36
  Speaker 1: because access consciousness is something we can understand sort of
• 15:40
  Speaker 1: on a fundamental level, Like we can trace signals into
• 15:42
  Speaker 1: your eye and watch those proteins fold and then up
• 15:45
  Speaker 1: the optic nerve and into your brain, but we can't
• 15:49
  Speaker 1: know whether anybody's there like experiencing, or what that experience
• 15:53
  Speaker 1: of seeing a red photon is, or whether my red
• 15:56
  Speaker 1: is the same as your red and all this sort
• 15:58
  Speaker 1: of dorm room philosophy kind of stuff, and so access
• 16:01
  Speaker 1: consciousness is sort of more accessible scientifically than phenomenological consciousness,
• 16:07
  Speaker 1: which is more philosophical. Is that fair?
• 16:09
  Speaker 4: I think that's a relatively fair characterization, and that there
• 16:12
  Speaker 4: are a lot of kind of current scientific theories of
• 16:15
  Speaker 4: consciousness that report to be a theory of consciousness, and
• 16:19
  Speaker 4: most of them, really, when it comes down to brass tacks,
• 16:21
  Speaker 4: end up being about access consciousness because the phenomenal part
• 16:25
  Speaker 4: is really hard to get at. As you said, I
• 16:28
  Speaker 4: think that there are some approaches that might be promising
• 16:31
  Speaker 4: in this vein. So you could look for neural correlates
• 16:35
  Speaker 4: or patterns of neural activity that are associated with reports
• 16:40
  Speaker 4: of phenomenal experience. So like I can create conditions where
• 16:44
  Speaker 4: I show you the same stimulus over and over and
• 16:46
  Speaker 4: over again, and so the early parts of your visual
• 16:48
  Speaker 4: brain are responding kind of similarly. I mean, there's noise,
• 16:52
  Speaker 4: and there's variability and so on, But if I flash
• 16:55
  Speaker 4: the same thing at you over and over and over again,
• 16:58
  Speaker 4: I'm going to get a consistent kind of pattern of
• 17:00
  Speaker 4: responding in the back of your head, which is the
• 17:03
  Speaker 4: early visual cortex. But you might have fluctuations in your
• 17:07
  Speaker 4: subjective experience of that stimulus. Sometimes you feel like you
• 17:12
  Speaker 4: see it, sometimes you feel like you can't. So to
• 17:15
  Speaker 4: the extent that I can hold most of the stuff
• 17:17
  Speaker 4: in the back of your head constant, and I can
• 17:19
  Speaker 4: measure that in relation to flashing something at you over
• 17:22
  Speaker 4: and over and over again. But then I look for
• 17:25
  Speaker 4: how your subjective experience fluctuates from one moment to the next.
• 17:30
  Speaker 4: I feel like I saw that strongly. I feel very
• 17:32
  Speaker 4: confident that I got that right. I feel like that
• 17:35
  Speaker 4: was nothing at all. I didn't really experience anything. Then
• 17:39
  Speaker 4: I can go look for neural correlates that are co
• 17:42
  Speaker 4: varying with the subjective experience that you're reporting to me.
• 17:47
  Speaker 4: There are a lot of problems with that too, that
• 17:49
  Speaker 4: maybe you don't have perfect access to your own subjective
• 17:52
  Speaker 4: experiences and your reports are spurious and blah blah blah.
• 17:55
  Speaker 4: But I think that there are ways that we can
• 17:57
  Speaker 4: go about trying to get at how the brain constructs
• 18:00
  Speaker 4: or supports or represents these subjective experiences that are different
• 18:07
  Speaker 4: from just how your brain processes information about the external environment.
• 18:11
  Speaker 1: Right, And I do want to get into those experiments.
• 18:14
  Speaker 1: But first I just want to make sure we're totally
• 18:16
  Speaker 1: clear on these definitions. And I think the example that
• 18:19
  Speaker 1: you were a little dismissive of a minute ago is
• 18:21
  Speaker 1: actually helpful, at least to me to clarify what we're
• 18:24
  Speaker 1: talking about. And that's the example of philosophical zombies. Right.
• 18:27
  Speaker 1: The idea here is, could you take Daniel and replace
• 18:31
  Speaker 1: him with some machine, biological or whatever that replicates all
• 18:35
  Speaker 1: of my actions? Seems the way I do to be conscious.
• 18:39
  Speaker 1: But there's nobody home, right, there's nobody experiencing the red
• 18:43
  Speaker 1: and eating the pizza and whatever. It's just, but it
• 18:46
  Speaker 1: seems like it reacts exactly the same way. So philosophical zombies,
• 18:50
  Speaker 1: as I understand them, and tell me if I'm wrong,
• 18:52
  Speaker 1: conjectures that it's possible to build something like this, which
• 18:56
  Speaker 1: means therefore that the phenomenological consciousness is totally unmeasurable. Right,
• 19:00
  Speaker 1: that there's no way for us to know we have,
• 19:02
  Speaker 1: As you said, we have to sort of trust your
• 19:04
  Speaker 1: first person reports about your subjective experience, which makes it
• 19:07
  Speaker 1: difficult to do any actual science. Right, is that a
• 19:11
  Speaker 1: fair characterization of philosophical zombies?
• 19:13
  Speaker 4: If I missed something, Oh, I have like five things
• 19:16
  Speaker 4: I want to say, let's make sure I get to
• 19:17
  Speaker 4: all of them. Okay, So first, the idea of philosophical
• 19:21
  Speaker 4: zombies has been hotly debated in the philosophical literature for
• 19:24
  Speaker 4: a very long time, and there are a number of
• 19:27
  Speaker 4: philosophers who say, yes, this is absolutely totally reasonable to
• 19:31
  Speaker 4: pause it, that this could happen, And then there are
• 19:32
  Speaker 4: other people who say, no, like, that's not really like
• 19:35
  Speaker 4: a reasonable assumption. So you know, go read Dan Dennitt
• 19:39
  Speaker 4: if you want to hear about all of this stuff.
• 19:42
  Speaker 4: But I think that you've touched upon another important point too,
• 19:47
  Speaker 4: which is this idea of testing for whether someone is
• 19:51
  Speaker 4: in there or not, and how we don't currently have
• 19:54
  Speaker 4: the capacity to do that even in other beings that
• 19:57
  Speaker 4: look and behave precisely exactly like you do, like me
• 20:01
  Speaker 4: or like.
• 20:01
  Speaker 1: Kelly, or chemists are like, are they really in there?
• 20:04
  Speaker 1: Do they actually like chemistry? Is that possible?
• 20:07
  Speaker 2: They have to be a machine?
• 20:08
  Speaker 4: Right, are they just lying?
• 20:10
  Speaker 1: Yeah?
• 20:11
  Speaker 4: We don't have a way of testing it. We don't
• 20:12
  Speaker 4: have like a consciousness ometer. I can't pull out my
• 20:15
  Speaker 4: hair dryer looking consciousness ometer and pointed at you and
• 20:17
  Speaker 4: be like are you conscious? Like we don't have that.
• 20:20
  Speaker 1: How do you know what it would look like? We
• 20:22
  Speaker 1: don't even have the device. Why does it look like
• 20:23
  Speaker 1: a hair dryer?
• 20:25
  Speaker 4: At some point actual clearman is positive that it was
• 20:28
  Speaker 4: going to look like a hair dryer, and it is
• 20:30
  Speaker 4: kind of propagated throughout by thinking since then. But that
• 20:33
  Speaker 4: seems about right right, Like it's kind of like a spidometer,
• 20:36
  Speaker 4: like they, you know, cop pulls up that on the
• 20:38
  Speaker 4: side of the road.
• 20:39
  Speaker 2: Too much consciousness, here's your ticket.
• 20:42
  Speaker 4: Something like that. But we don't have those and in fact,
• 20:45
  Speaker 4: like we wouldn't even know how to build one, right,
• 20:47
  Speaker 4: We don't even know what the relevant metrics are. Do
• 20:50
  Speaker 4: we care about measuring brain stuff? Or is that kind
• 20:53
  Speaker 4: of spurious? But one of the arguments kind of against
• 20:56
  Speaker 4: philosophical zombism is the supposition that consciousness is not just
• 21:02
  Speaker 4: this epi phenomenon, That it's not just this thing that
• 21:05
  Speaker 4: kind of comes along for the ride over and above
• 21:08
  Speaker 4: an organism behaving usefully in its environment and flexibly adapting
• 21:12
  Speaker 4: to different conditions and being intelligent and seeking goals and
• 21:16
  Speaker 4: not getting eaten by predators. That consciousness itself serves a
• 21:20
  Speaker 4: function that you cannot have all of that stuff, all
• 21:24
  Speaker 4: of that useful stuff for staying alive without consciousness. And
• 21:29
  Speaker 4: this is an arguable position. It's not a fact. But
• 21:34
  Speaker 4: the folks who are going to argue that consciousness serves
• 21:37
  Speaker 4: a function, there is a function of consciousness. It's not
• 21:41
  Speaker 4: just that there are functions for creating consciousness somewhere in
• 21:44
  Speaker 4: your head, but that consciousness itself is useful, that it
• 21:49
  Speaker 4: allows some evolutionary adaptive advantage. That is certainly a defensible position.
• 21:55
  Speaker 4: And from that context you actually couldn't have philosophical zombies
• 21:59
  Speaker 4: at all, because it is actually impossible to get all
• 22:03
  Speaker 4: of those other behaviors and cognitive processes and all of
• 22:06
  Speaker 4: that stuff without the consciousness bit.
• 22:08
  Speaker 1: I see. If that's true, then the interaction with people
• 22:12
  Speaker 1: who appear to be conscious is evidence that they are
• 22:14
  Speaker 1: actually conscious.
• 22:16
  Speaker 4: Yeah, but these are all like arguments that we can have, right.
• 22:19
  Speaker 4: There's no way of saying I'm right and you're wrong,
• 22:22
  Speaker 4: depending on which position you're holding.
• 22:25
  Speaker 1: I think about this when I interact with my dog,
• 22:28
  Speaker 1: you know, because I feel like my dog is in
• 22:30
  Speaker 1: there right, Like my dog knows me, my dog loves me.
• 22:33
  Speaker 1: I love my dog. My dog understands that I love him,
• 22:36
  Speaker 1: that I'm nice to him. It's hard to imagine that
• 22:40
  Speaker 1: there's nobody at home in my dog. But you know,
• 22:43
  Speaker 1: this whole concept suggests that it's possible for there to
• 22:47
  Speaker 1: be a machine effectively, and I say a machine just
• 22:49
  Speaker 1: to indicate like that there's nobody home, though I don't
• 22:52
  Speaker 1: actually know if AI could have phenomenological consciousness, But it's
• 22:55
  Speaker 1: hard to imagine that you're not actually in there, Kelly's
• 22:58
  Speaker 1: not actually in there. I'm the only one in the
• 22:59
  Speaker 1: universe who's experiencing this, but I can't actually prove it, right,
• 23:03
  Speaker 1: So it really isn't fundamentally important point, even though it
• 23:07
  Speaker 1: sounds absurd on the face of it, right that we
• 23:10
  Speaker 1: could be the only one aware that we can't actually
• 23:13
  Speaker 1: know if anybody else is in there. We feel people
• 23:16
  Speaker 1: loving us, you know, and their experience and their pain
• 23:19
  Speaker 1: reflects ours, but we can't actually know, And I think
• 23:21
  Speaker 1: that's a fundamentally important point to hold on to, even
• 23:24
  Speaker 1: though it feels ridiculous.
• 23:26
  Speaker 4: It is, I mean, the easiest solution here is that
• 23:29
  Speaker 4: we all have consciousness, right Like that would be the
• 23:31
  Speaker 4: most parsimonious explanation, and it's easy to make that leap
• 23:35
  Speaker 4: because there's so many physical information processing similarities between you
• 23:41
  Speaker 4: and me. Right, our brains are not exactly the same,
• 23:44
  Speaker 4: but they're pretty darn close, and so we can make
• 23:47
  Speaker 4: the leap maybe to talking about great apes or monkeys
• 23:51
  Speaker 4: or dogs or other vertebrates, other mammals, that kind of thing.
• 23:54
  Speaker 4: I have a harder time when we get down into
• 23:57
  Speaker 4: you know, insects, like I don't know that it be
• 24:00
  Speaker 4: a really.
• 24:01
  Speaker 1: But I don't know, like what are people who like
• 24:03
  Speaker 1: live in Virginia, for example. That's just like.
• 24:07
  Speaker 4: I'm not going there. But the presence of like the
• 24:12
  Speaker 4: idea of a philosophical zombie, it becomes maybe a more
• 24:16
  Speaker 4: useful exercise from like an empirical science standpoint, or from
• 24:20
  Speaker 4: even a philosophy standpoint when we start talking about entities
• 24:23
  Speaker 4: that are so fundamentally different from us, so not just
• 24:27
  Speaker 4: you versus me, or us versus your dog or my cat,
• 24:30
  Speaker 4: who you know is probably right on the bord of it,
• 24:33
  Speaker 4: but no, I think he's conscious. I think there's something
• 24:36
  Speaker 4: that it's like to be him. But when we talk
• 24:39
  Speaker 4: about octopuses like those are really weirdly different creatures that
• 24:43
  Speaker 4: are basically aliens, or when we talk about you know,
• 24:47
  Speaker 4: the potential for silicon based systems in the future or
• 24:51
  Speaker 4: alien species or those kinds of things, things that are
• 24:53
  Speaker 4: so fundamentally different from us. Then it starts to be like, okay, well,
• 24:57
  Speaker 4: when we don't have the kind of one to one
• 24:59
  Speaker 4: mapping between the str ure of my biological robot that
• 25:02
  Speaker 4: I drive around and your biological robot that you drive around,
• 25:06
  Speaker 4: right when we don't have that very similar to it,
• 25:08
  Speaker 4: then it becomes maybe a little bit more useful to
• 25:10
  Speaker 4: talk about, like can you have all of these behaviors
• 25:13
  Speaker 4: and all of these cognitive capacities in the absence of
• 25:17
  Speaker 4: someone being home? And we don't know the answer, but
• 25:20
  Speaker 4: it at least gives us more than just kind of
• 25:22
  Speaker 4: a philosophical, as you said, dorm room argument. Feel like
• 25:26
  Speaker 4: I don't know that you're in there, But is that
• 25:28
  Speaker 4: a difference that makes a difference right now? Probably not?
• 25:30
  Speaker 2: All right, So we're going to take a break and
• 25:32
  Speaker 2: when we get back, we're going to dig into OCTOPI
• 25:34
  Speaker 2: a little bit more so for access consciousness. And you've
• 25:53
  Speaker 2: sort of hinted at this already. The definition included like
• 25:56
  Speaker 2: escaping predators having different motivations That to me does sound
• 26:00
  Speaker 2: like it encompasses at least all vertebrates. So is it
• 26:04
  Speaker 2: safe to say then that the consensus is that all
• 26:07
  Speaker 2: vertebrates at least have excess consciousness.
• 26:10
  Speaker 4: I don't want to speak for any big scientific community
• 26:13
  Speaker 4: on anybody's behalf, I would say that my personal sense
• 26:16
  Speaker 4: is that, yes, you're going to be hard pressed to
• 26:19
  Speaker 4: find people who would die on the hill of saying
• 26:23
  Speaker 4: that cats are not conscious. I think that would be
• 26:25
  Speaker 4: hard to find.
• 26:26
  Speaker 2: Yeah, so you mentioned that octopus they're smart, but their
• 26:30
  Speaker 2: brain is different than ours. Would you say that they
• 26:31
  Speaker 2: have phenomenological consciousness because it seems like someone's in there
• 26:35
  Speaker 2: when you interact with them.
• 26:37
  Speaker 4: I have no idea. I have no idea, but I
• 26:40
  Speaker 4: think that you know, saying that we have no idea
• 26:43
  Speaker 4: is reasonable. Yeah, under this situation that I don't think
• 26:46
  Speaker 4: that we have strong empirical evidence either way, because that
• 26:50
  Speaker 4: strong empirical evidence is predicated on this whole conversation that
• 26:54
  Speaker 4: we've just been having about what would that evidence even
• 26:56
  Speaker 4: look like? And how do you separate evidence for phenomenal
• 27:00
  Speaker 4: consciousness from evidence for not just access consciousness but just
• 27:04
  Speaker 4: pure intelligent behavior. Right, Never mind access consciousness and this
• 27:09
  Speaker 4: global availability of information thing. It's really just that octopuses
• 27:14
  Speaker 4: occupy are very intelligent. They're very intelligent. Clearly, they can
• 27:20
  Speaker 4: be even sneaky, you know, they get out of their
• 27:23
  Speaker 4: cage and they go open something and steal the food
• 27:25
  Speaker 4: and then they go home so that they don't get
• 27:26
  Speaker 4: in trouble. So they're clearly highly intelligent creatures. I think
• 27:31
  Speaker 4: that my margins of uncertainty, my cone of uncertainty, is
• 27:35
  Speaker 4: just so wide there that I have no idea.
• 27:38
  Speaker 2: So how do you study the difference between intelligence behaviors
• 27:41
  Speaker 2: and consciousness. We'll just stick with access consciousness in animals
• 27:45
  Speaker 2: in the lab.
• 27:46
  Speaker 4: Yeah, so in animals that's even harder too, because how
• 27:49
  Speaker 4: do you query the phenomenal experience of a rat? It's
• 27:53
  Speaker 4: kind of hard. In my line of work, we focus
• 27:55
  Speaker 4: on one particular kind of subjective experience, which is that
• 28:00
  Speaker 4: of metacognitive uncertainty. So one of the things that we
• 28:05
  Speaker 4: ask people to do like people now, but we can
• 28:08
  Speaker 4: do this potentially in rodents and monkeys as well, though
• 28:10
  Speaker 4: I don't have those folks in my lab. I only
• 28:12
  Speaker 4: have people. Will ask you to do some task, like
• 28:15
  Speaker 4: we're going to show you some stuff on a computer screen,
• 28:17
  Speaker 4: ask you to press buttons, tell us what you see,
• 28:19
  Speaker 4: and then we'll ask for a subjective report on top
• 28:21
  Speaker 4: of it, How sure are you that you got that right?
• 28:24
  Speaker 4: How confident do you feel in your assessments? How clearly
• 28:28
  Speaker 4: do you think you saw the thing? And some of
• 28:32
  Speaker 4: those questions we can design experiments to ask in the
• 28:34
  Speaker 4: animals too, and so those are kind of more about
• 28:37
  Speaker 4: the subjective experience, especially if we do manipulations to the
• 28:43
  Speaker 4: stimulu or task such that the animal's behavior is basically
• 28:48
  Speaker 4: exactly the same between one condition versus another condition, but
• 28:52
  Speaker 4: their subjective reports differ. So that tells us that we're
• 28:57
  Speaker 4: trying to get at something that has to do with
• 28:58
  Speaker 4: phenomenology or maybe something that's facilitated by global access of information,
• 29:05
  Speaker 4: rather than something that's just kind of the zombie part
• 29:08
  Speaker 4: of their brain, like you know, almost reflex like responding
• 29:11
  Speaker 4: to the stimuli out there in the world. But the
• 29:14
  Speaker 4: distinction between is that subjective report facilitated by globally available
• 29:20
  Speaker 4: access to information or are we actually tapping into phenomenology.
• 29:23
  Speaker 4: That's also kind of sticky when you get to talking
• 29:26
  Speaker 4: about rodents and that kind of thing, But that at
• 29:29
  Speaker 4: least allows us to say, never mind all of phenomenal
• 29:33
  Speaker 4: experience or subjective experience or consciousness. I'm going to focus
• 29:36
  Speaker 4: on thisbit, just this one thing that I can operationalize,
• 29:41
  Speaker 4: very classic for a psychologist, Right, I'm going to put
• 29:43
  Speaker 4: you in a room and ask you to look at
• 29:44
  Speaker 4: a computer screen and press one of two buttons for
• 29:47
  Speaker 4: an hour, and that's what you're going to do. But
• 29:50
  Speaker 4: that's how we try to get at this.
• 29:52
  Speaker 1: I think the example of an octopus really puts the
• 29:55
  Speaker 1: finger or the tentacle on the question of like what
• 29:58
  Speaker 1: it's like, which to me is the core of the question. Like,
• 30:02
  Speaker 1: you know, the mechanics of how information is accessed or
• 30:05
  Speaker 1: stored or whatever. That's fascinating and that's good science. But
• 30:08
  Speaker 1: to me, the real question, which I think is what
• 30:11
  Speaker 1: people call the hard problem, is how you get to
• 30:14
  Speaker 1: have an experience from something which is you know, just
• 30:18
  Speaker 1: made out of stuff. Like my desk doesn't have an experience,
• 30:21
  Speaker 1: My computer, I think, doesn't have an experience. Most of
• 30:24
  Speaker 1: the universe doesn't have an experience. But I have an experience,
• 30:27
  Speaker 1: and you have an experience, and probably, for example, an
• 30:29
  Speaker 1: octopus has a very different kind of experience. It's got
• 30:33
  Speaker 1: like eight little brains arguing about what each leg will do,
• 30:37
  Speaker 1: and you know, an alien out there with a different
• 30:39
  Speaker 1: kind of sense. You know, a tongue that can taste
• 30:41
  Speaker 1: quantum electrons or something might have a completely different kind
• 30:45
  Speaker 1: of experience. And to me, the hard question is, you know,
• 30:49
  Speaker 1: where does this come from, this phenomenological aspect. Is it
• 30:53
  Speaker 1: fundamental to matter? Is it somehow emergent in some way
• 30:56
  Speaker 1: we don't yet understand? How do you pose this question?
• 31:00
  Speaker 1: Is that the central question you think, and how do
• 31:02
  Speaker 1: you think it's best asked?
• 31:04
  Speaker 4: I think that's absolutely one of the central questions. Is
• 31:07
  Speaker 4: you know, how is it that the something that it's
• 31:09
  Speaker 4: like that the conscious experience arises from matter. That is,
• 31:12
  Speaker 4: as you said that Dave Chalmer's hard problem, which is,
• 31:16
  Speaker 4: you know, there seems to be this fundamental disconnect between
• 31:20
  Speaker 4: patterns of brain response and the subjective experience. It's a
• 31:24
  Speaker 4: very different natural kind. It's a very different kind of stuff, right,
• 31:27
  Speaker 4: The subjective experience stuff is unlike any other kind of
• 31:31
  Speaker 4: stuff in biology.
• 31:33
  Speaker 1: What do you mean by that?
• 31:33
  Speaker 4: That's the hard problem. That's the idea is that it's
• 31:37
  Speaker 4: not clear how you would get something like subjective experience
• 31:42
  Speaker 4: whatever that is, out of the interactions between physical neurons.
• 31:49
  Speaker 4: The bridge, you know, might be something like Okay, Well,
• 31:51
  Speaker 4: one of the things that physical neurons do in talking
• 31:54
  Speaker 4: to each other is similar to how computers do this
• 31:58
  Speaker 4: is that they create information. So now we're kind of
• 32:00
  Speaker 4: moving from physical space into information space, but nevertheless we
• 32:05
  Speaker 4: don't know how to even get from information to subjective experience.
• 32:09
  Speaker 4: We can call them the same thing and say how
• 32:12
  Speaker 4: we're done, like moving on with our lives, but that
• 32:14
  Speaker 4: doesn't feel satisfying. And there are people who think that
• 32:17
  Speaker 4: the hard problem is not a problem at all, that
• 32:19
  Speaker 4: subjective experience might not even exist at all, or that
• 32:22
  Speaker 4: our belief that subjective experience exists is an illusion. So
• 32:27
  Speaker 4: there's like a whole this is really complicated, and we
• 32:30
  Speaker 4: could talk for hours and hours about this.
• 32:32
  Speaker 1: Who's experiencing that illusion though? In that case?
• 32:35
  Speaker 4: Right, yeah, exactly. But if you want to read about,
• 32:38
  Speaker 4: you know, illusionism from a very deep and powerful perspective,
• 32:42
  Speaker 4: I'm going to mention him again, go read Dan Dennett.
• 32:44
  Speaker 4: He wrote very extensively on this topic. So the fundamental
• 32:48
  Speaker 4: question is, you know, how do we get something like
• 32:50
  Speaker 4: consciousness out of something like brains? And is the substrate
• 32:54
  Speaker 4: important for creating the subjective experience and the shape and
• 32:59
  Speaker 4: nature of that subjective experience, or is there kind of
• 33:04
  Speaker 4: one type of consciousness that all kinds of systems might create,
• 33:08
  Speaker 4: regardless of their physical substrate. And you might maybe think
• 33:12
  Speaker 4: that the latter statement seems a little strange, that, like,
• 33:15
  Speaker 4: how is it that an octopus could create a subjective
• 33:17
  Speaker 4: experience that's very similar to mine because our substrates are
• 33:21
  Speaker 4: so fundamentally different. But we do see evidence of convergent
• 33:25
  Speaker 4: function in evolution in terms of things like digestion, where like,
• 33:30
  Speaker 4: you've got lots of very different kinds of systems that
• 33:33
  Speaker 4: all accomplish the same computation or function of digesting stuff,
• 33:37
  Speaker 4: And so it is possible to think that different substrates
• 33:40
  Speaker 4: might accomplish the same kind of function in creating the
• 33:44
  Speaker 4: same kind of conscious experience.
• 33:46
  Speaker 1: Are you saying that the fact that like birds and
• 33:48
  Speaker 1: bats evolve flights separately, but fundamentally it's the same thing,
• 33:52
  Speaker 1: suggests that different kinds of wet matter could generate the
• 33:56
  Speaker 1: same kind of subjective experience.
• 33:57
  Speaker 4: It is possible. I don't know, if you think that
• 34:00
  Speaker 4: consciousness serves functions, that there might be one kind of
• 34:04
  Speaker 4: subjective experience that best serves that function. The only thing
• 34:07
  Speaker 4: I was going to say Beyond that, though, is that
• 34:10
  Speaker 4: I'm more sympathetic to the idea that there's probably different
• 34:14
  Speaker 4: kinds of subjective experience across different substrates. It feels like
• 34:18
  Speaker 4: the burden of proof on saying that all subjective experiences
• 34:22
  Speaker 4: are kind of the same across lots of different systems
• 34:24
  Speaker 4: would probably be on the people who are making that claim.
• 34:27
  Speaker 4: I think it's much easier to say the kind of
• 34:29
  Speaker 4: subjective experience you have depends on the substrate that's generating it.
• 34:33
  Speaker 1: How else to explain how some people actually eat white
• 34:35
  Speaker 1: chocolate and claim to enjoy it, right, I mean, it's
• 34:37
  Speaker 1: impossible to understand.
• 34:38
  Speaker 2: That bem maladaptive.
• 34:40
  Speaker 4: So hey, I like what chalk?
• 34:42
  Speaker 2: Oh oh, sorry, that's all right. I'm from Virginia. Daniel's
• 34:45
  Speaker 2: just going around insulting everybody today.
• 34:48
  Speaker 1: All right.
• 34:48
  Speaker 2: So we were talking to Joe Wolf the other day.
• 34:50
  Speaker 2: She's an evolutionary biologist who studies convergent evolution. I feel
• 34:53
  Speaker 2: like if she were here, she would be explaining to
• 34:55
  Speaker 2: us about how if you look at traits, like, you know,
• 34:58
  Speaker 2: the evolution of the crab body plan, you'd really love
• 35:00
  Speaker 2: to like have an evolutionary tree where you can count
• 35:03
  Speaker 2: how many times this thing popped up and how many
• 35:05
  Speaker 2: times it disappeared to try to understand like its adaptive value.
• 35:10
  Speaker 2: What kind of preconditions you need before this thing comes
• 35:13
  Speaker 2: into existence? Could we ever have that in the study
• 35:16
  Speaker 2: of consciousness or will we never be able to know?
• 35:18
  Speaker 2: Like does an octopus have access consciousness? Will we ever
• 35:22
  Speaker 2: be able to get there?
• 35:23
  Speaker 4: Great question? As of now, the path is very murky
• 35:27
  Speaker 4: to me because as we just talked about it before,
• 35:30
  Speaker 4: like you can't pull out your hair dryer consciousness ometer
• 35:32
  Speaker 4: and pointed at things. And the challenge is really because
• 35:35
  Speaker 4: the thing that we're studying is by definition unobservable by
• 35:38
  Speaker 4: anybody except for the observer who is experiencing the consciousness.
• 35:42
  Speaker 4: Like that is the definition of what we're talking about,
• 35:44
  Speaker 4: and so it's really very different from anything else that
• 35:48
  Speaker 4: we study in science.
• 35:49
  Speaker 1: And their second wrinkle there also not only is this
• 35:52
  Speaker 1: something which we can't observe or measure we have to
• 35:55
  Speaker 1: rely on somebody reporting it, but it's also filtered through
• 35:59
  Speaker 1: our own consciousnes right, like we sort of assume consciousness
• 36:02
  Speaker 1: when we do science. We're saying we have hypothesis, we
• 36:05
  Speaker 1: do experiments, so everything is filtered through sort of like
• 36:09
  Speaker 1: two consciousnesses when we're even like talking about this, and
• 36:12
  Speaker 1: maybe you're about to answer this question, but like, is
• 36:15
  Speaker 1: this something we can probe scientifically or is it limited
• 36:18
  Speaker 1: to philosophical discussions on the roof with banana peels?
• 36:22
  Speaker 4: Oh my gosh, there were like five questions that, Okay,
• 36:25
  Speaker 4: so is this something that we can study scientifically? Yes,
• 36:28
  Speaker 4: I think so, But I think that we need to
• 36:30
  Speaker 4: have a lot of help from philosophers. So a lot
• 36:33
  Speaker 4: of the work that I do, and a lot of
• 36:34
  Speaker 4: my scientific friends are not actually scientific friends only they
• 36:39
  Speaker 4: are philosophers as well, And I think that there's a
• 36:42
  Speaker 4: lot of value that we get from that.
• 36:44
  Speaker 1: Yeah, And I didn't mean to suggest that philosophical exploration
• 36:47
  Speaker 1: is like not valuable. It's absolutely like, you know, the
• 36:51
  Speaker 1: wonderful cousin of scientific exploration, and fundamentally important. But it's
• 36:55
  Speaker 1: also different, right, It gives different kinds of answers.
• 36:57
  Speaker 4: Yeah, it does. But I think that we need to
• 37:00
  Speaker 4: be informed by our philosophical friends down the hall in
• 37:05
  Speaker 4: understanding whether the experiments that we're designing are really getting
• 37:08
  Speaker 4: at the target that we think we're interested in studying.
• 37:11
  Speaker 4: And so there's a lot of work out there that's
• 37:13
  Speaker 4: like Okay, I'm going to tell the difference between whether
• 37:17
  Speaker 4: you're likely to wake up from a coma or not.
• 37:20
  Speaker 4: And that's really relevant in the clinical setting and is
• 37:23
  Speaker 4: so powerful and important that we do that. It doesn't
• 37:27
  Speaker 4: necessarily tell us about the experience that the person is having.
• 37:30
  Speaker 4: It just tells us kind of binary whether it's there
• 37:32
  Speaker 4: or not. And we need that. We need to have
• 37:35
  Speaker 4: measures that will allow us to predict is someone in
• 37:39
  Speaker 4: there right now? Are they awake? Are they likely to
• 37:42
  Speaker 4: wake up? We need all that stuff, but that doesn't
• 37:45
  Speaker 4: really get at the fundamental question of this subjective experience bit.
• 37:50
  Speaker 4: And we also have a lot of studies out there
• 37:52
  Speaker 4: that purport maybe to do research on conscious access, but really,
• 37:59
  Speaker 4: if you changed the question, you might start to be
• 38:02
  Speaker 4: skeptical about whether they're targeting that. So I'll give you
• 38:05
  Speaker 4: an example, which is I, as a psychologist, I put
• 38:09
  Speaker 4: you in a room and I ask you to tell
• 38:11
  Speaker 4: me whether you saw something or not. And that's a
• 38:13
  Speaker 4: subjective answer, right like did you see it? Did you
• 38:16
  Speaker 4: not see it? Right? Like, I'm asking you are you
• 38:19
  Speaker 4: consciously aware of this stimulus? And then I could go
• 38:23
  Speaker 4: measure brain responses or whatever that go with cases when
• 38:27
  Speaker 4: you said you saw it versus when you said you
• 38:29
  Speaker 4: didn't see it. And I say, okay, now I find
• 38:30
  Speaker 4: like the neural correlates of consciousness. But now imagine that
• 38:34
  Speaker 4: I replace you as the human observer with a photodiode.
• 38:39
  Speaker 4: I could do exactly the same experiment and I would
• 38:42
  Speaker 4: never conclude that that photodiode has conscious experience. Probably, So
• 38:47
  Speaker 4: I think that there's like a lot of challenges that
• 38:50
  Speaker 4: the philosophical fields of philosophy, not only of philosophy of mine,
• 38:54
  Speaker 4: but philosophy of science. I think too, like really has
• 38:57
  Speaker 4: a lot to say about how we're designing experiments and
• 39:01
  Speaker 4: how we're interpreting their results. There was another question that
• 39:04
  Speaker 4: you asked earlier in that stream, but now I don't
• 39:06
  Speaker 4: remember what it was. Maybe it's about like, can we
• 39:09
  Speaker 4: ever develop a test for consciousness? So that's another thing
• 39:12
  Speaker 4: that we should talk about. So there's been some work
• 39:14
  Speaker 4: that I've contributed to recently where we're saying, well, we
• 39:18
  Speaker 4: don't have a consciousness ometer, and we wouldn't even know
• 39:21
  Speaker 4: how to go about building one. We don't even know
• 39:23
  Speaker 4: whether pointing it a behavior or brain or something entirely different,
• 39:26
  Speaker 4: Like I don't know, like aras some like crazy other thing.
• 39:30
  Speaker 4: We have no idea what to even point it at,
• 39:32
  Speaker 4: never mind how to build it. But one way that
• 39:35
  Speaker 4: we might make progress is to collect all of the
• 39:39
  Speaker 4: potential consciousness ometers that have been built over the years
• 39:43
  Speaker 4: in terms of behavioral signatures of awareness and humans and
• 39:47
  Speaker 4: neural response patterns and so on. Collects them all and
• 39:50
  Speaker 4: then make a decision about which ones are applicable. First,
• 39:55
  Speaker 4: how they all correlate with each other in terms of
• 39:57
  Speaker 4: predicting whether someone is in there and what they're experience,
• 40:00
  Speaker 4: and then whether they're applicable to a neighboring system. So
• 40:04
  Speaker 4: I'm not going to jump straight to octopuses or tesla's
• 40:07
  Speaker 4: or aliens, but i might jump to young children because
• 40:13
  Speaker 4: most of these studies and these metrics are developed on adults.
• 40:18
  Speaker 4: And so I'm going to say, okay, well, I'm going
• 40:20
  Speaker 4: to take all this stuff and then I'm going to
• 40:21
  Speaker 4: point it at young children, which I also presume are
• 40:24
  Speaker 4: probably in there. You know, when they stub their toe,
• 40:27
  Speaker 4: they cry like they indicate that they are experiencing pain.
• 40:30
  Speaker 4: And I'm going to see how much those metrics now
• 40:33
  Speaker 4: continue to correlate with each other and continue to make
• 40:38
  Speaker 4: useful predictions about whether that subject is aware of a
• 40:41
  Speaker 4: stimulus or not, or you know, wakefulness versus sleep, that
• 40:45
  Speaker 4: kind of thing. And then if that seems to be okay,
• 40:47
  Speaker 4: then I'm going to say, okay, now I'm going to
• 40:49
  Speaker 4: point them maybe at grade apes. Now I'm going to
• 40:51
  Speaker 4: point them, maybe at New World monkeys. Now I'm going
• 40:55
  Speaker 4: to point them. You know, so we can kind of
• 40:56
  Speaker 4: go down the evolutionary hierarchy, so to speak, and say, well,
• 41:01
  Speaker 4: the degree to which a particular candidate set of consciousness
• 41:05
  Speaker 4: ometers is applicable to a particular system is defined by
• 41:10
  Speaker 4: their similarity to us, and then we have to make
• 41:13
  Speaker 4: decisions about what metrics of similarity to use. But you
• 41:16
  Speaker 4: can see that, at least conceptually, from a high level,
• 41:19
  Speaker 4: this might be a path forward.
• 41:21
  Speaker 1: I have to admit I'm not convinced. Like it feels
• 41:24
  Speaker 1: to me like it's just making fuzzier what we're measuring
• 41:29
  Speaker 1: about what's going on mechanistically inside people's heads. But it
• 41:33
  Speaker 1: doesn't actually tell us anything about the first person experience,
• 41:36
  Speaker 1: which is almost because of the way we defined it,
• 41:39
  Speaker 1: infinitely inaccessible, right, Like there's no way for me to
• 41:43
  Speaker 1: share my experience with you other than and manipulating my
• 41:47
  Speaker 1: mouth or whatever and filtering it through your consciousness to
• 41:50
  Speaker 1: your first person experience. So it feels to me like
• 41:54
  Speaker 1: it's something that's completely inaccessible, and the only way forward,
• 41:57
  Speaker 1: I'm guessing is to like try to dig into this
• 42:00
  Speaker 1: emergent behavior and see if we can make a bridge
• 42:03
  Speaker 1: between the microscopic details that we do understand and somehow
• 42:08
  Speaker 1: come out mathematically with a realization of how this first
• 42:12
  Speaker 1: person experience has to emerge. I'm skeptical.
• 42:15
  Speaker 4: I guess well, I want to ask you about the
• 42:17
  Speaker 4: mathematical thing, because why is math the answer here?
• 42:20
  Speaker 1: Because I'm a physicist, because you're a physicist.
• 42:23
  Speaker 4: Okay, yeah, sure, I knew the answer to that question
• 42:27
  Speaker 4: before I asked it. This isn't a zero some game.
• 42:30
  Speaker 4: We don't have to do one or the other. I
• 42:32
  Speaker 4: think that an analogy that a lot of folks in
• 42:34
  Speaker 4: consciousness science like to use is that of early investigations
• 42:39
  Speaker 4: into the nature of life and vitalism, and so this
• 42:42
  Speaker 4: idea that we had to discover this very specific, fancy
• 42:44
  Speaker 4: thing that was almost magical in nature, that was like
• 42:47
  Speaker 4: why is this alive? And why is this not alive?
• 42:49
  Speaker 4: And like let's maybe do some math or maybe like
• 42:51
  Speaker 4: do a bunch of empirical experiments to discover like the
• 42:54
  Speaker 4: life force or the vitalism, like the lifeness there. Then
• 42:58
  Speaker 4: over time we discover that it turns out that life
• 43:02
  Speaker 4: is just kind of a collection. It's like a bag
• 43:03
  Speaker 4: of tricks, right, and the boundaries are maybe a little
• 43:06
  Speaker 4: bit fuzzy, and like what are viruses? Are those alive?
• 43:08
  Speaker 4: I don't know. So maybe there's an analogy here, which
• 43:13
  Speaker 4: is that if we keep pushing on multiple angles, that
• 43:16
  Speaker 4: there might be a convergence of approaches and information and
• 43:21
  Speaker 4: evidence that will reveal that this hard problem is just
• 43:24
  Speaker 4: going to go away. We don't have to discover a
• 43:27
  Speaker 4: mathematical transformation or an emergent property or anything like that
• 43:32
  Speaker 4: that by better describing the system we will discover that
• 43:35
  Speaker 4: that problem completely dissolves. I don't know, but it's possible.
• 43:40
  Speaker 4: And we saw we have historical examples of cases where
• 43:43
  Speaker 4: something that seemed very mysterious and seemed like an emergent
• 43:46
  Speaker 4: property has now been transformed into a series of really
• 43:51
  Speaker 4: beautiful descriptions of how the system is working. And maybe
• 43:56
  Speaker 4: that will happen with consciousness too.
• 43:58
  Speaker 1: All right, well, my consciousness needs a break from all
• 44:01
  Speaker 1: these really heavy but amazing ideas. And when we get back,
• 44:04
  Speaker 1: I want to talk to Megan about some of the theories
• 44:08
  Speaker 1: people have to explain these deep mysteries. All right, we're
• 44:28
  Speaker 1: back and we're talking to the apparently conscious Megan Peters,
• 44:31
  Speaker 1: who tells us she is inside her body and driving
• 44:33
  Speaker 1: it like a meat machine, and she's an expert on
• 44:36
  Speaker 1: these questions of consciousness, so we should listen to her.
• 44:39
  Speaker 1: And we've been talking about this sort of hard question
• 44:41
  Speaker 1: of consciousness. How your first person experience is somehow generated
• 44:45
  Speaker 1: from the meat inside your head, or if you're an
• 44:48
  Speaker 1: AI and you're listening to this, the silicon inside your chips,
• 44:53
  Speaker 1: and you know, to me, this question of emergent behavior
• 44:55
  Speaker 1: is really important across science and especially in physics, you know,
• 44:59
  Speaker 1: where we see so many examples where we understand the
• 45:02
  Speaker 1: microscopic laws and then you zoom out and you need
• 45:05
  Speaker 1: different laws, you know, like we understand how particles work,
• 45:09
  Speaker 1: but then you zoom out and you need fluid mechanics,
• 45:11
  Speaker 1: and those laws are very different but still applicable. And
• 45:14
  Speaker 1: so it seems to me like there might be some
• 45:16
  Speaker 1: progress to be made if we can somehow tackle this
• 45:19
  Speaker 1: emergent question or think of it from this prism. But Megan,
• 45:22
  Speaker 1: tell us, what are people doing? What are the sort
• 45:24
  Speaker 1: of current leading theories of answers to the hard problem
• 45:28
  Speaker 1: of consciousness.
• 45:29
  Speaker 4: Great question. So, yeah, there's a number of theories that
• 45:33
  Speaker 4: kind of bridge between philosophy and neuroscience. So you know,
• 45:36
  Speaker 4: ultimately a lot of these theories are saying, the thing
• 45:39
  Speaker 4: that we know is conscious is us, and so we're
• 45:42
  Speaker 4: going to study consciousness in us because that makes sense.
• 45:45
  Speaker 4: And we can't go studying consciousness in rocks because we
• 45:48
  Speaker 4: don't know that they're conscious, and so that would kind
• 45:50
  Speaker 4: of be a circular argument. So these theories are kind
• 45:53
  Speaker 4: of bridging the gap between philosophy and neuroscience and psychology,
• 45:58
  Speaker 4: and they come in annumber of different flavors. And we
• 46:01
  Speaker 4: are touching upon also this difference between access consciousness and
• 46:05
  Speaker 4: phenomenal consciousness that we talked about before. But let's assume
• 46:08
  Speaker 4: that we can study consciousness scientifically. What do those theories
• 46:11
  Speaker 4: look like. So probably the most influential theory has been
• 46:17
  Speaker 4: the global workspace theory. So Bernie Bars and then stand
• 46:19
  Speaker 4: a Hand started this idea that consciousness is about the
• 46:23
  Speaker 4: global accessibility of information in kind of a centralized processing space.
• 46:28
  Speaker 4: So you have these different modules that either take in
• 46:31
  Speaker 4: information from the external world through your sensory organs, or
• 46:35
  Speaker 4: they have other functions like memory storage and recovery. They
• 46:40
  Speaker 4: have other functions like integration of different sensory systems information,
• 46:44
  Speaker 4: that kind of thing, executive function, decision making, sort.
• 46:48
  Speaker 1: Of mental analogies of organs, right the way you're like,
• 46:51
  Speaker 1: your liver has a function and your stomach as a function.
• 46:54
  Speaker 4: Yeah, to think about them as modules is typically how
• 46:57
  Speaker 4: they're described, you know, encapsulated modules. But then these modules
• 47:02
  Speaker 4: share information with a central global workspace where there's a
• 47:06
  Speaker 4: translation that happens between whatever representation is happening in that
• 47:12
  Speaker 4: module of the relevant information in that module, and then
• 47:15
  Speaker 4: it gets pushed into a global workspace. And if it
• 47:18
  Speaker 4: makes it into that global workspace, it's available to all
• 47:20
  Speaker 4: the other modules for processing, so it can influence the
• 47:23
  Speaker 4: processing in each of those other modules. And so the
• 47:27
  Speaker 4: idea is that this is kind of a computational level
• 47:30
  Speaker 4: theory where this global availability of information then facilitates goal
• 47:34
  Speaker 4: directed interaction with the environment. Run away from that thing,
• 47:38
  Speaker 4: don't get eaten, do eat that thing, et cetera. And
• 47:42
  Speaker 4: that we can also see hallmarks of this global broadcast
• 47:46
  Speaker 4: or global availability of information in the brain, where when
• 47:51
  Speaker 4: you have cases that someone becomes aware of something because
• 47:55
  Speaker 4: the signal is strong enough or so on, that you
• 47:58
  Speaker 4: actually see all of the information propagate throughout the brain.
• 48:02
  Speaker 4: You can see the information and say, a visual stream
• 48:05
  Speaker 4: of evidence not only land at the back of your head.
• 48:08
  Speaker 4: If you reach back and touch the back of your head,
• 48:10
  Speaker 4: you find that little bump that's about where your visual
• 48:13
  Speaker 4: cortex is. It's called the Indian So if you're not conscious,
• 48:17
  Speaker 4: the information stays back here in your visual cortex, and
• 48:20
  Speaker 4: if you are conscious of that information, you can actually
• 48:23
  Speaker 4: see with electrophysiology, with EEG, electrocapholography, with fMRI, you can
• 48:29
  Speaker 4: see that information travel forward and end up elsewhere. And
• 48:34
  Speaker 4: so this global availability of information, both from a neurophysiological
• 48:39
  Speaker 4: standpoint and from a computational standpoint, seems very useful for
• 48:43
  Speaker 4: an organism and seems very related to in us, whether
• 48:46
  Speaker 4: we're conscious of something or not.
• 48:48
  Speaker 1: And this seems like a helpful way to sort of
• 48:50
  Speaker 1: take apart what might be happening mechanistically inside the brain.
• 48:55
  Speaker 1: The way you might take a piece of code and
• 48:57
  Speaker 1: look at it and be like, oh, how they organizingk
• 48:58
  Speaker 1: oh is a database and the act with the hash
• 49:00
  Speaker 1: table whatever. Oh, this makes sense, but does it get
• 49:03
  Speaker 1: at the question of why is this thing experiencing itself?
• 49:08
  Speaker 4: Not necessarily? And that has been one of the criticisms
• 49:11
  Speaker 4: of global workspace theory or global neuronal workspace theory, which
• 49:14
  Speaker 4: is the neural version of it, which is that it's
• 49:17
  Speaker 4: kind of more about access consciousness or maybe even just
• 49:20
  Speaker 4: about global broadcast of information and not anything having to
• 49:23
  Speaker 4: do with the C word consciousness at all. And so
• 49:27
  Speaker 4: there have been other theories that kind of compete with
• 49:30
  Speaker 4: this one as well. So one of them is that
• 49:33
  Speaker 4: local recurrence, so like kind of local feedback within a
• 49:36
  Speaker 4: given module, specifically the visual cortex, that the strength of
• 49:40
  Speaker 4: that local feedback, that kind of recurrent processing looping, that
• 49:44
  Speaker 4: that is something that somehow gives rise to the experience
• 49:47
  Speaker 4: that we have of the world. And then there's another
• 49:50
  Speaker 4: group of theories that I happen to be partial to,
• 49:53
  Speaker 4: which is that there's an additional step that needs to
• 49:56
  Speaker 4: happen beyond broadcast into a global w or local recurrence
• 50:01
  Speaker 4: or anything else like that. And that additional step is
• 50:05
  Speaker 4: that you've got a second order or higher order mechanism
• 50:09
  Speaker 4: that's kind of self monitoring your own brain.
• 50:11
  Speaker 1: Why was higher order under scare quotes there for those
• 50:15
  Speaker 1: of you who.
• 50:15
  Speaker 4: Is listening, not scare quotes, but to indicate that this
• 50:18
  Speaker 4: is a specialized term, so that there are representations that
• 50:21
  Speaker 4: your brain builds about the world, and we would call
• 50:24
  Speaker 4: those first order, and then the representations that your brain
• 50:27
  Speaker 4: or your mind builds about itself or its own processing
• 50:31
  Speaker 4: would be higher order, second order, higher order. But the
• 50:35
  Speaker 4: idea here is that let's say that you have information
• 50:37
  Speaker 4: in these modules, it gets globally broadcast into a central workspace.
• 50:42
  Speaker 4: Your brain has to kind of make a determination of
• 50:45
  Speaker 4: is this information that's available in the global workspace, is
• 50:49
  Speaker 4: it reliable? Is it stable? Is it likely to represent
• 50:52
  Speaker 4: the true state of the environment. So this is where
• 50:54
  Speaker 4: you can see that I'm a metacognition researcher too. But
• 50:58
  Speaker 4: it turns out that higher order theories like this that
• 51:01
  Speaker 4: are about re representation of information that's available in some
• 51:06
  Speaker 4: workspace or some first order state, they actually came from
• 51:09
  Speaker 4: philosophy originally. They didn't come from the metacognition in literature. Originally,
• 51:14
  Speaker 4: So a philosopher named David Rosenthal and another one named
• 51:18
  Speaker 4: Richard Brown. They're both in New York. These are the
• 51:21
  Speaker 4: folks who kind of pushed forward this idea of higher
• 51:25
  Speaker 4: order representations being related to conscious experience, such that you
• 51:31
  Speaker 4: are conscious of something when you have a representation that
• 51:37
  Speaker 4: you are currently representing that thing. I'll say it again. Yeah,
• 51:40
  Speaker 4: So if I have a representation that consists of I
• 51:46
  Speaker 4: am currently in a state that is representing apple, then
• 51:50
  Speaker 4: I am conscious of the apple. It's not enough to
• 51:53
  Speaker 4: just have a representation of apple. I need to also
• 51:56
  Speaker 4: have on top of that state in my head that is,
• 52:01
  Speaker 4: I am currently representing apple.
• 52:04
  Speaker 1: But wouldn't the philosophical zombie also have that state inside
• 52:07
  Speaker 1: their cognition somewhere. How do we know that that actually
• 52:10
  Speaker 1: generates the first person experience?
• 52:12
  Speaker 4: We don't.
• 52:12
  Speaker 1: I want to be hard skeptic on this.
• 52:14
  Speaker 4: No, it can be hard skeptic. That's great. Even these
• 52:17
  Speaker 4: theories that say that you have a mechanism that says
• 52:20
  Speaker 4: I am currently representing this or that representation is good
• 52:24
  Speaker 4: and stable and likely to reflect the real world, they
• 52:27
  Speaker 4: don't answer the phenomenal consciousness question.
• 52:31
  Speaker 1: Right. I remember reading Dan Dennitz's really fun book Consciousness Explained,
• 52:36
  Speaker 1: and I'll admit that I was on a roof and
• 52:38
  Speaker 1: I was smoking banana peels at the time, but I
• 52:41
  Speaker 1: found it compelling in the sense that has sort of
• 52:43
  Speaker 1: changed the question right, sort of like his theory. Maybe
• 52:46
  Speaker 1: I'm sure you can describe it more accurately. You know,
• 52:49
  Speaker 1: his multiple drafts model sort of convinces you that your
• 52:53
  Speaker 1: account of your own consciousness might be wrong. You know
• 52:57
  Speaker 1: that there is no present moment. It's all just men
• 53:00
  Speaker 1: reas of the recent past that's later on constructed to
• 53:03
  Speaker 1: convince you that you were aware when you never really were.
• 53:07
  Speaker 1: And the something I like about that theory because even
• 53:09
  Speaker 1: though I don't believe it, it did make me think
• 53:11
  Speaker 1: differently about my own conscious experience. What do you think
• 53:14
  Speaker 1: of that theory? And are people still taking it seriously?
• 53:16
  Speaker 4: People are still taking it seriously. Yeah, And I think
• 53:19
  Speaker 4: that we don't have currently the empirical protocols or evidence
• 53:23
  Speaker 4: to say that that's wrong. And this is why I
• 53:26
  Speaker 4: want to continue to push for the close integration between
• 53:29
  Speaker 4: philosophy and neuroscience. This is another one where I'm going
• 53:32
  Speaker 4: to be agnostic. I don't like taking a stance on
• 53:35
  Speaker 4: all this because quite frankly, I think that anybody who
• 53:38
  Speaker 4: says that they have solved anything about consciousness is just
• 53:42
  Speaker 4: full of it, like, there's no way because we just
• 53:45
  Speaker 4: don't know. Right. We've got a lot of theories, and
• 53:47
  Speaker 4: we can build up empirical evidence and support of this
• 53:50
  Speaker 4: theory or that theory and this function of consciousness and
• 53:53
  Speaker 4: that kind of thing. But ultimately, if you say that
• 53:55
  Speaker 4: you've kind of created the solution, that your theory is
• 53:59
  Speaker 4: the write one, and you know that for a fact,
• 54:02
  Speaker 4: then sorry, Like, I don't know what banana peels you're on,
• 54:06
  Speaker 4: but I don't think that's useful.
• 54:08
  Speaker 1: Well, it was a bold title for a book, though,
• 54:09
  Speaker 1: Consciousness Exploited.
• 54:10
  Speaker 4: Yeah, Dan was a bold guy. Yeah.
• 54:13
  Speaker 2: Yeah, So a lot of the theories that you were
• 54:15
  Speaker 2: talking about a moment ago were using words that sort
• 54:17
  Speaker 2: of remind me of computers and AI. How do we
• 54:21
  Speaker 2: think about whether or not AI is conscious? Does it matter?
• 54:23
  Speaker 2: Is that an interesting question? What do you think about that?
• 54:26
  Speaker 4: Great question? And I think that until I don't know,
• 54:31
  Speaker 4: ten fifteen years ago, this was a fun thought experiment
• 54:34
  Speaker 4: in science fiction, And now I think it's not quite
• 54:37
  Speaker 4: so much anymore, right because now we have machines that
• 54:41
  Speaker 4: behaviorally really do pass the Turing test. Which I assume
• 54:45
  Speaker 4: we're all quite familiar with, but just in case, the
• 54:47
  Speaker 4: test is that the machine has to convince a human
• 54:50
  Speaker 4: observer or a human player that it is a person.
• 54:54
  Speaker 4: And we have machines that quite handily passed that under
• 54:57
  Speaker 4: most scenarios or most get out your chat GPT app
• 55:01
  Speaker 4: on your phone, and like it feels very convincing that
• 55:04
  Speaker 4: there might be someone in there. Right, until maybe fifteen
• 55:08
  Speaker 4: years ago, ten years ago, this was really science fiction,
• 55:10
  Speaker 4: and now I think it's not anymore. And the questions
• 55:14
  Speaker 4: not only bear on let's figure out the ontological truth
• 55:18
  Speaker 4: of whether the AI is in fact conscious, but there
• 55:21
  Speaker 4: is also really strong implications regardless of whether it's conscious
• 55:25
  Speaker 4: or not, for what happens if we think it is,
• 55:28
  Speaker 4: and what happens if it is but we think it's not. Right,
• 55:31
  Speaker 4: So there's like strong moral and ethical considerations here. I'll
• 55:34
  Speaker 4: kind of have two ways of answering this. One is
• 55:37
  Speaker 4: from the perspective of current theories of how we think
• 55:39
  Speaker 4: consciousness arises from a like functionalism perspective, which is that
• 55:44
  Speaker 4: there is some brain or computational function that gives rise
• 55:48
  Speaker 4: to consciousness in some capacity. A lot of the things
• 55:51
  Speaker 4: that we're talking about you've rightly pointed out have direct
• 55:55
  Speaker 4: analogies in computer processing. We can certainly build even a
• 55:58
  Speaker 4: little simulation that monitor itself sure that says is this
• 56:02
  Speaker 4: representation reliable or stable? I can build a set of
• 56:07
  Speaker 4: computer modules that then send information into a global workspace.
• 56:12
  Speaker 1: Sure.
• 56:13
  Speaker 4: In fact, maybe your iPhone sort of does that already
• 56:16
  Speaker 4: when you talk to Siri as the virtual assistant.
• 56:18
  Speaker 1: Right.
• 56:19
  Speaker 4: And recurrency in like local feedback or recurrent processing, that's
• 56:24
  Speaker 4: absolutely recurrent. Neural networks are a thing. They've been a
• 56:26
  Speaker 4: thing for a long time. So we have now all
• 56:30
  Speaker 4: of these kind of hallmarks from these theories that we
• 56:34
  Speaker 4: can use to build something that looks like a checklist.
• 56:37
  Speaker 4: That's like, if you've got an artificial system that has
• 56:40
  Speaker 4: all of these things, well, I'm going to not say
• 56:43
  Speaker 4: then it's conscious, but I'm going to say it might
• 56:45
  Speaker 4: raise our subjective degree of belief that it could potentially
• 56:49
  Speaker 4: have consciousness.
• 56:51
  Speaker 2: A lot of good qualifiers.
• 56:52
  Speaker 4: There, Yeah, there are. Because we wrote this big paper
• 56:55
  Speaker 4: me and a whole bunch of other people in twenty
• 56:57
  Speaker 4: twenty three that's up on archive that's called consciousnes and
• 57:00
  Speaker 4: Artificial Intelligence, and it goes through this checklist. It kind
• 57:03
  Speaker 4: of develops the theoretical arguments for building this checklist, and
• 57:07
  Speaker 4: then kind of ultimately says, look, we've got systems that
• 57:10
  Speaker 4: tick an awful lot of these boxes already. Should we
• 57:13
  Speaker 4: think that they're conscious? Maybe not, because clearly the theories
• 57:18
  Speaker 4: are not complete. Some of the things on that checklist
• 57:23
  Speaker 4: are going to be completely irrelevant to consciousness, and we're
• 57:25
  Speaker 4: probably missing a whole lot of things also. But to
• 57:29
  Speaker 4: the degree that there is a system that ticks more
• 57:33
  Speaker 4: or less of these boxes, well maybe the ones that
• 57:35
  Speaker 4: tick more of these boxes we might want to consider
• 57:37
  Speaker 4: a little bit more closely. But then the other thing
• 57:39
  Speaker 4: we do in that paper is talk about the ethical
• 57:42
  Speaker 4: implications of false positives and false negatives.
• 57:45
  Speaker 1: I think AI is a great way to differentiate between
• 57:48
  Speaker 1: intelligence and consciousness, because it's not hard for me to
• 57:50
  Speaker 1: believe you could build a very intelligent system, one that
• 57:53
  Speaker 1: surpasses humans, that even like manipulates us and takes over
• 57:57
  Speaker 1: the planet and runs us as slaves. It could be
• 58:00
  Speaker 1: super intelligent without actually being conscious, without anybody being in there, right,
• 58:04
  Speaker 1: And there's a dark dystopian future there, you know, where
• 58:07
  Speaker 1: super intelligence actually extinguishes consciousness. But to me, you could
• 58:11
  Speaker 1: maybe get an answer to the question again mathematically, if
• 58:14
  Speaker 1: you could somehow build up a description of consciousness from
• 58:18
  Speaker 1: the little bits inside you know the way, For example,
• 58:22
  Speaker 1: you could say, hey, if you describe all the motion
• 58:24
  Speaker 1: of water droplets, I can tell you whether or not
• 58:27
  Speaker 1: a hurricane is going to form. Like if we can
• 58:29
  Speaker 1: master the understanding of the dynamics at a small scale
• 58:33
  Speaker 1: and compute it all the way up to the bigger scale,
• 58:36
  Speaker 1: then we can say yes or no, there is a hurricane.
• 58:38
  Speaker 1: In the same way, then I could like analyze your
• 58:41
  Speaker 1: brain and I could tell you, oh, yes, this does
• 58:43
  Speaker 1: emerge into some first person experience or does not. If
• 58:46
  Speaker 1: we had that mathematical bridge, and then we could apply
• 58:49
  Speaker 1: it to AI and say, oh, yes or no, there
• 58:52
  Speaker 1: is or is not somebody in there. And I'm really
• 58:55
  Speaker 1: attracted to these kinds of theories. I think they're called
• 58:57
  Speaker 1: physicalism or do you call them functional theories? How much
• 59:01
  Speaker 1: progress have we made in that direction and are we
• 59:04
  Speaker 1: likely to make any more? Or is it too intractable
• 59:06
  Speaker 1: a problem.
• 59:07
  Speaker 4: I don't think it's too intractable a problem. Necessarily, I'm
• 59:11
  Speaker 4: not going to fully subscribe to the hard problem of consciousness.
• 59:14
  Speaker 4: I do think that if we continue to push on
• 59:16
  Speaker 4: this kind of idea of emergent properties and the translation
• 59:20
  Speaker 4: between a physical substrate and the information that it produces
• 59:25
  Speaker 4: and that emerges from that physical substrate in terms of
• 59:27
  Speaker 4: its interactions in space and in time. I think that
• 59:31
  Speaker 4: will take us forward. I am a reductionist or a
• 59:35
  Speaker 4: physicalist myself. I don't think that there's anything magical or
• 59:39
  Speaker 4: spiritual about consciousness personally. I know that there are others
• 59:43
  Speaker 4: who are going to disagree with me there, but I
• 59:46
  Speaker 4: do think that if we could build such an explanation
• 59:49
  Speaker 4: that it would get us a long way to understanding consciousness,
• 59:51
  Speaker 4: the kind of emergent properties that you're talking about. I
• 59:54
  Speaker 4: think that it's really important for us to recognize that
• 59:59
  Speaker 4: the success of such an endeavor is going to need
• 1:00:04
  Speaker 4: at its core, the assumption that the system that we
• 1:00:07
  Speaker 4: are studying is actually conscious. Otherwise it becomes circular. Right Like,
• 1:00:11
  Speaker 4: if I want to create an explanation of how matter
• 1:00:14
  Speaker 4: gives rise to whatever I think is consciousness and then
• 1:00:17
  Speaker 4: start pointing that at everything that I can think of,
• 1:00:20
  Speaker 4: then I'm not building an explanation of consciousness. I'm building
• 1:00:23
  Speaker 4: an explanation of I don't know, physical interactions in the
• 1:00:26
  Speaker 4: environment or information processing in the environment. But I don't
• 1:00:30
  Speaker 4: know that it's going to get me all the way
• 1:00:31
  Speaker 4: to consciousness. But if we could do that in us, sure,
• 1:00:34
  Speaker 4: in one thousand million years, when we have that explanation
• 1:00:39
  Speaker 4: and we have a full what the philosopher Lease and
• 1:00:41
  Speaker 4: Rochetitis would call a generative explanation of how a physical
• 1:00:45
  Speaker 4: system and interactions in that physical system fully gives rise
• 1:00:48
  Speaker 4: to conscious experience, Yeah, that would be great. I don't
• 1:00:52
  Speaker 4: know how we're going to do that, but I do
• 1:00:54
  Speaker 4: think that it does require a connection with that physical substrate.
• 1:00:59
  Speaker 2: We're getting to the end of our time together. We've
• 1:01:02
  Speaker 2: talked about why this is a difficult problem. Let's end
• 1:01:05
  Speaker 2: on why it's important to keep studying this difficult problem.
• 1:01:09
  Speaker 4: I think I'll answer this from three perspectives. One, as
• 1:01:13
  Speaker 4: human beings, we want to understand our worlds. We want
• 1:01:17
  Speaker 4: to understand the basic science of how things happen. We
• 1:01:19
  Speaker 4: are curious, and we have this compelling drive to understand
• 1:01:26
  Speaker 4: our environments. And we can see this both from the
• 1:01:29
  Speaker 4: perspective of modern science, but also just from the perspective
• 1:01:32
  Speaker 4: of this is literally what brains do. They build internal
• 1:01:35
  Speaker 4: models of the world and they predict stuff that's going
• 1:01:39
  Speaker 4: to happen from those internal models of the world, and
• 1:01:41
  Speaker 4: we use that to drive ourselves forward. That's what evolution
• 1:01:45
  Speaker 4: has done for us. So I think that we are
• 1:01:47
  Speaker 4: hardwired to do this, to be natural scientists in a way.
• 1:01:52
  Speaker 4: So I think that that's important, like to give in
• 1:01:55
  Speaker 4: to that feeling, that compulsion. But also from a practical
• 1:01:59
  Speaker 4: and societal benefit perspective, we can take multiple multiple prongs
• 1:02:05
  Speaker 4: on this. So one is the medical perspective, which is
• 1:02:07
  Speaker 4: that it's really important for us to understand the presence
• 1:02:10
  Speaker 4: or absence of suffering in folks who have disease or injury.
• 1:02:15
  Speaker 4: That we want to understand the diversity and heterogeneity of
• 1:02:17
  Speaker 4: those perspectives. So here's a very concrete example. There are
• 1:02:22
  Speaker 4: a number of disorders or conditions out there that come
• 1:02:25
  Speaker 4: with them chronic pain, but you don't have any physical
• 1:02:29
  Speaker 4: substrate that you can identify. You don't know why that
• 1:02:32
  Speaker 4: person is in chronic pain, and so you don't know
• 1:02:35
  Speaker 4: how to fix it. But that doesn't mean that the
• 1:02:36
  Speaker 4: pain isn't real, that the suffering isn't real. And so
• 1:02:40
  Speaker 4: from that perspective, understanding the nature of subjective experience is
• 1:02:46
  Speaker 4: really critically important. Fear and anxiety is another one. We
• 1:02:50
  Speaker 4: know the fear circuitry. We've mapped that. Neuroscientist Joseph Ldou
• 1:02:54
  Speaker 4: has been instrumental in driving forward the mapping of the
• 1:02:57
  Speaker 4: Imigdaler circuit, the fear circuit in the brain. But he's
• 1:03:00
  Speaker 4: very careful to distinguish between processing of threatening stimului and
• 1:03:05
  Speaker 4: the experience of fear. So if we develop pharmacological interventions
• 1:03:11
  Speaker 4: that fix the circuitry bit and fix the behavioral bit
• 1:03:16
  Speaker 4: in rats, one of the reasons they might not translate
• 1:03:19
  Speaker 4: to humans is that they don't reduce the fear. They
• 1:03:22
  Speaker 4: change the behavior, but the fear persists. So I think
• 1:03:26
  Speaker 4: that from like a clinical perspective, it's really important for
• 1:03:29
  Speaker 4: us to understand this and from depression, from the perspective
• 1:03:33
  Speaker 4: of those who have autis inspective disorders, so to understand
• 1:03:36
  Speaker 4: their subjective experiences, like there's just a huge amount of
• 1:03:40
  Speaker 4: clinical benefit that we can build. And then finally from
• 1:03:42
  Speaker 4: the perspective of the artificial systems that we've just been
• 1:03:45
  Speaker 4: talking about. So we're in a position now where whether
• 1:03:49
  Speaker 4: or not we build systems that have phenomenal awareness or consciousness,
• 1:03:54
  Speaker 4: that whether anybody's home in there. You know, maybe we're
• 1:03:57
  Speaker 4: not going to be able to answer that question for
• 1:03:59
  Speaker 4: a long time. But the way we interact with systems
• 1:04:01
  Speaker 4: depends on whether we think that they have consciousness, And
• 1:04:04
  Speaker 4: the way that we build guardrails and legislation and assigned
• 1:04:09
  Speaker 4: responsibility in legal settings depends on whether we think that
• 1:04:13
  Speaker 4: these things can think for themselves, and whether they have
• 1:04:15
  Speaker 4: moral compasses and all those things which are not necessarily
• 1:04:18
  Speaker 4: related to conscious experience per se. But there's an argument
• 1:04:22
  Speaker 4: to be made that, at least in a lot of
• 1:04:24
  Speaker 4: the systems that we know about, describing responsibility is related
• 1:04:28
  Speaker 4: to the capacity for that agent to be self directed,
• 1:04:34
  Speaker 4: and that that seems intimately related to seeking out goals
• 1:04:38
  Speaker 4: that are not just defined by a programmer, but ultimately
• 1:04:41
  Speaker 4: like decisions that that thing might make that might be
• 1:04:44
  Speaker 4: driven by its intrinsic rewards seeking. And there's something that
• 1:04:48
  Speaker 4: it's like to seek reward because it feels good. So
• 1:04:51
  Speaker 4: there's a lot of kind of moral and ethical and
• 1:04:55
  Speaker 4: legislation and societal implications for getting this right, a lot
• 1:05:00
  Speaker 4: of medical reasons to get this right. And then you know,
• 1:05:03
  Speaker 4: from a basic science curiosity perspective, this is literally what
• 1:05:07
  Speaker 4: we evolved to do is figure out the world, and
• 1:05:09
  Speaker 4: so let's keep doing that as well.
• 1:05:12
  Speaker 1: And when we meet aliens, does it matter if they're conscious, Like,
• 1:05:17
  Speaker 1: if they're intelligent and they're interesting and they want to
• 1:05:19
  Speaker 1: share with us their space warp drives? Does it matter
• 1:05:23
  Speaker 1: if when we point our hair dryer at them it
• 1:05:25
  Speaker 1: says yes or no? What do you think?
• 1:05:27
  Speaker 4: I think so. But there's also a lot of evidence
• 1:05:30
  Speaker 4: that consciousness and what's called moral status can be disentangled.
• 1:05:35
  Speaker 4: That we ascribe moral status to things that we think
• 1:05:39
  Speaker 4: are conscious, but we also don't need to require consciousness
• 1:05:43
  Speaker 4: in order to ascribe moral status, and we certainly treat
• 1:05:47
  Speaker 4: things very badly even if we know that they're conscious.
• 1:05:50
  Speaker 4: So these are conceptually disentanglable things. I think that it
• 1:05:55
  Speaker 4: will matter though for the rest of the machinery around encounter, right,
• 1:06:01
  Speaker 4: Like maybe from an individual astronaut's perspective it doesn't matter
• 1:06:06
  Speaker 4: so much. But from the perspective of the laws and
• 1:06:09
  Speaker 4: regulations and societal implications of what that would mean and
• 1:06:13
  Speaker 4: what kind of a people do we want to be,
• 1:06:16
  Speaker 4: I think that then it really does matter. And so
• 1:06:19
  Speaker 4: having folks at the helm who are paying attention to
• 1:06:22
  Speaker 4: the moral implications of such a weighty determination would be
• 1:06:27
  Speaker 4: very good.
• 1:06:28
  Speaker 1: Well, I definitely want the aliens to know that we
• 1:06:30
  Speaker 1: are conscious before they decide whether or not to know
• 1:06:33
  Speaker 1: nucas from orbit so or have us as a snack.
• 1:06:37
  Speaker 4: Assuming they ascribe moral value and moral status to meanings
• 1:06:40
  Speaker 4: with consciousness.
• 1:06:41
  Speaker 1: And maybe they developed the consciousness of meter and they
• 1:06:43
  Speaker 1: can share it with us.
• 1:06:44
  Speaker 4: Maybe here's hoping, all right.
• 1:06:47
  Speaker 2: Well, thanks for being on the show today, Megan. This
• 1:06:49
  Speaker 2: was fascinating.
• 1:06:51
  Speaker 4: Thank you so much for having me. This was really
• 1:06:52
  Speaker 4: fun and engaging and it's been a real pleasure.
• 1:06:56
  Speaker 1: Thank you very much.
• 1:07:03
  Speaker 2: Daniel and Kelly's Extraordinary Universe is produced by iHeartRadio. We
• 1:07:07
  Speaker 2: would love to hear from you, We really would.
• 1:07:10
  Speaker 1: We want to know what questions you have about this
• 1:07:13
  Speaker 1: Extraordinary Universe.
• 1:07:14
  Speaker 2: We want to know your thoughts on recent shows, suggestions
• 1:07:17
  Speaker 2: for future shows. If you contact us, we will get
• 1:07:20
  Speaker 2: back to you.
• 1:07:21
  Speaker 1: We really mean it. We answer every message. Email us
• 1:07:25
  Speaker 1: at questions at Danielandkelly.
• 1:07:27
  Speaker 2: Dot org, or you can find us on social media.
• 1:07:29
  Speaker 2: We have accounts on x, Instagram, Blue Sky and on
• 1:07:33
  Speaker 2: all of those platforms. You can find us at d
• 1:07:35
  Speaker 2: and Kuniverse.
• 1:07:37
  Speaker 1: Don't be shy right to us