automatically generated captions
00:00:00 I'm Don Hoffman
00:00:01 and I'm at the University of California at Irvine in the department of cognitive sciences with joint appointments in
00:00:07 philosophy in computer science as well and I try to integrate all of those disciplines in the work that I do and when.
00:00:17 I did. You will in in the field of cognitive science I'm studying visual perception. How do we see a three D. World.
00:00:27 How do we see the colors of objects the shapes of objects the motions of objects.
00:00:32 What are the processes that are going on inside of our brains when we do that
00:00:36 and it turns out about a third of the brain's cortex is engaged just in vision.
00:00:41 So when you simply open your eyes look around the room.
00:00:44 Billions of neurons and trillions of synapses are springing into action
00:00:49 and so one of the things we try to do in the cognitive neuroscience is just understand what's going on why all of this
00:00:55 horse power.
00:00:57 A third of your highest processing horse power is involved in visual perception that's a bit surprising So the question
00:01:04 is why do we need to have that much horse power in something that seemed so simple.
00:01:08 Just open your eyes and see the world
00:01:11 and it turns out that we have to do a lot of computation because in some sense we're creating the worlds that we see.
00:01:19 And so computer science comes into it because we need to understand from a computational point of view what's going on
00:01:26 when you open your eyes and see the shapes of objects the colors in the motions of objects.
00:01:30 You're not just seeing them.
00:01:31 You're actually creating a virtual reality in in some sense for yourself you're a reality engine.
00:01:42 What I'm seeing now is pretty widely accepted in cognitive neuroscience is they they almost every cognitive
00:01:49 neuroscientist will say that we are constructing what we see. In real time.
00:01:54 So you close your eyes you just see a gray field you open your eyes and it looks like you just see. Seeing a three D.
00:02:01 World with objects and colors emotions as it is just like taking a snapshot.
00:02:07 But most cognitive neuroscientist will say
00:02:09 but that's not what's going on was happening really is that within about one hundred milliseconds about one tenth of a
00:02:15 second you're creating the three D.
00:02:18 World around you you're creating the objects in the colors in the motions you're doing it so quickly
00:02:23 and apparently so effortlessly that you're you know taken
00:02:27 and you think you're just seeing the world as it is that would say that's the majority the vast majority view in
00:02:34 cognitive neuroscience right now that we construct what we see. And so computer science comes into it in part because.
00:02:42 To make sure that we understand what we're doing that our theories about the construction process are accurate.
00:02:48 It's good for us to try to build them to actually build robotic vision systems that work.
00:02:52 So if you have a theory about how you see the three D. Structure of objects.
00:02:56 Well then build it and if you can actually have a computer that has video cameras giving inputs into the computer
00:03:03 and then software in the computer that creates inside the computer and the three D.
00:03:08 Model of the object that is what you think it should have done then maybe you've got a good theory.
00:03:13 So the computer science comes in in the following way. We're trying to reverse engineer.
00:03:19 What's going on in the human brain.
00:03:21 Well enough that we can then implement it in a computer vision system and build robotic visions
00:03:27 and if you can do that in that that's an existence proof that you might actually have a good theory
00:03:33 and then the philosophy.
00:03:38 That's actually the person who is most famous for sort of pushing this point of view is a guy named David Marr who was
00:03:48 a professor at MIT the late seventy's and early one nine hundred eighty S.
00:03:54 And David Marr he was in the Artificial Intelligence Laboratory at MIT and in what's now.
00:04:00 The brain and cognitive sciences department.
00:04:02 And he was the one that he had a background in mathematics and neuroscience
00:04:06 and worked in the AI lab Marvin Minsky invited him to come there because of this this new point of view that we should
00:04:13 you know vision is a constructive process
00:04:15 and we need to understand the mathematics of the construction well enough that we can actually build it.
00:04:21 So if you can build working systems robotic systems then it would which is in your contribution to artificial
00:04:27 intelligence then that will really show that you understand what's going on when we open our eyes and see. Well.
00:04:38 In the construction of vision. Yes Yes that's right. So well what Maher got us into what David Marr.
00:04:55 Got to feel to think about is trying to get our understanding of human vision human visual processing so rigorous that
00:05:03 we can actually build robotic vision systems so that means that that what seems to us like an immediate perception of
00:05:09 shapes and colors and motions is in fact a sophisticated computation.
00:05:13 And so that's where the computer science comes in we actually want to take RAW images coming in from video from like
00:05:20 video cameras which is just a bunch of numbers if you look at it it's just an interpretive ball a ray of numbers.
00:05:27 Millions of numbers so cild the idea is that we want to build robotic vision systems that will take for example video
00:05:37 from cameras into the computer and that video.
00:05:40 If you look at it is just an array of numbers millions of numbers you look at it it's who knows what it means is just a
00:05:47 bunch of numbers and you want to take those numbers and then create a world. What.
00:05:53 That those numbers are trying to describe to you like a boy riding a bicycle getting hot
00:05:58 or whatever might be going on there.
00:06:00 So so to do that you can see going from those numbers to a three dimensional world with colors and objects
00:06:05 and boys riding bicycles is not a trivial thing.
00:06:08 And so that's what robotic vision
00:06:11 and computer vision has been doing ever since David Marr really got us going in this direction in the late mid to late
00:06:17 one nine hundred seventy S. and It was his work that got me into this field.
00:06:21 I read his work when I was an undergraduate and said Where is this guy I'd like to work with him
00:06:25 and I ended up going
00:06:26 and being a student at MIT so so I were I got the pleasure of working with David Marr for a couple years the last two
00:06:34 years of his life until he died of leukemia.
00:06:37 While I was still a graduate student but he said he set the field on this on this new path
00:06:44 and ever since then there have been the majority of the field has been trying to build rigorous mathematical models
00:06:51 that you could in principle
00:06:53 or in practice actually build robotic vision systems from so now I've got an intelligent vision system on my car partly
00:07:00 as a result of this kind of work they can see if I'm on the on the road. Going over the lines and so forth.
00:07:06 It will beep me so we're starting to get intelligent vision systems coming out of this where the the car itself can see.
00:07:12 in three D.
00:07:13 So so this is the result of this long decades long effort of trying to understand mathematically
00:07:19 and precisely how what why we have to spend a third of our brains cortex all that horsepower in seeing the three D.
00:07:27 World and objects and colors. Now where we're really understanding that. So that's the standard view.
00:07:33 I mean there are some dissenters
00:07:34 but I would say it's by far the the majority view that we create what we see we construct the world that we see.
00:07:51 You could a lot of people in order.
00:07:53 Fishel intelligence would say exactly that that we are machines we're just carbon based machines. So the.
00:08:00 Would be the standard view an artificial intelligence we we are complex machines
00:08:03 and they would say that we should not take that as an indication that we're not valuable
00:08:07 but we are just carbon based machines and we can reverse engineer the algorithms that are going on in the brain.
00:08:17 Once we reverse engineer those algorithms from the neural networks of the brain then we can implement them in
00:08:23 artificial neural networks for example in in silicon and so we transfer from CARBON SILICON. But the algorithms.
00:08:30 If we've done it right or still the same or close to the our algorithms that are that are in the brain.
00:08:36 So that's sort of the idea
00:08:37 and then the idea then is that those algorithms do a good job of Truthfully reconstructing the fruit shapes the true
00:08:48 colors the true motions of objects in the world. That's the standard view.
00:08:52 So it's not my view now but that is the standard view.
00:09:05 OK Well my my view is that we do construct what we see so I agree with the field with the majority of the field that we
00:09:15 construct in real time all the shapes colors and motions not objects that we see.
00:09:20 But I don't think that we're reconstructing the truth.
00:09:25 So everybody in my field pretty much believes that we're reconstructing a good faithful reproduction of the true shapes
00:09:34 and colors and I don't think we are. Instead I think that what's going on is.
00:09:41 It's more like a desktop interface on your computer.
00:09:45 So if you have an icon for it and say oh you're writing a file
00:09:49 or editing a photograph you're writing an e-mail to a friend.
00:09:52 And the icon for that file is blue and rectangular and in the middle of your screen.
00:10:00 That doesn't mean that the file itself in your computer is blue and rectangular
00:10:05 and in the middle of the computer that that's a silly notion anybody without that doesn't understand what the desktop
00:10:11 interface is for.
00:10:13 And I think that that's what we have our perceptions are like that desktop interface so space and time
00:10:20 or the desktop three dimensional space as you perceive it in time or the desktop
00:10:24 and then physical objects like a glass or a car a spoon. These are all just icons in that three dimensional desktop.
00:10:33 And the point of the interface the desktop and the on the icons is not to show you the truth.
00:10:39 The point is really to hide the truth right. If you had to know in the case of the computer.
00:10:44 If you had know all the diodes and resistors and voltages and you know magnetic fields.
00:10:50 If you don't know all that in deal with that you'd never finish reading your e-mail to your friend.
00:10:55 So you don't need to know the reality you want it in fact the desktop interface is there to hide that reality reality
00:11:02 gets in the way of what you really need to do. And so I think. In so. That's right.
00:11:18 What you see the whole three dimensional world that we perceive and all the physical objects that we perceive
00:11:25 and all their properties are just like the the colored icons on your desktop. Right. Right. That's right.
00:11:40 And that's that's a very useful belief was very useful for me to know if you know to think that the icon is the file
00:11:49 and I can just double click on the file and it will open what I drag the file to the trash can. I can delete it.
00:11:55 So it's very it's a very nice and useful fiction allows me to do what I need to do. But it is just I use perfection.
00:12:04 To see that that's right. Right. So. Right.
00:12:21 You're fooling yourself if you believe that what you're seeing is a true replication of what is there
00:12:28 when you don't look at so I do think that there is an objective world
00:12:32 and I think there is a reality that exists whether or not I.
00:12:36 I'm here and whether or not I'm looking so there is some objective reality.
00:12:40 But the chances that my perceptions are reconstructing part of that reality. So I'm seeing the truth.
00:12:49 The chances are actually you can prove the are zero and the argument comes from evolution.
00:12:56 So I started looking at evolution by natural selection and what it has to say about perception. OK sure. You are right.
00:13:20 After the playing right or wrong. Yes right. Yes maybe that's right that's right that's right.
00:13:52 Yeah that's it's a very alarming point of view. Yes. Well the.
00:14:01 Yes So this this point of view is quite surprising the idea that what you're seeing
00:14:07 when you look around you see a fireplace you see a cop you see people. Well what you.
00:14:17 Yes What you're seeing is a representation that you construct. For one purpose.
00:14:25 The representation is there to keep you alive and to guide your behaviors.
00:14:30 So that you can stay alive long enough to reproduce and evolutionary hardware that we can go into
00:14:35 but the whole point is just like the desktop interface on your compel Yeah you can and with but OK and then then
00:14:49 and then I'll continue my sentence to the end and then stop OK very very well absolutely. Right but not the truth.
00:15:29 Yes well it. So there is so exactly.
00:15:33 So what you see from this point of view what you're seeing when you see shapes of objects and their colors
00:15:39 and their motions in a three D. World around you is simply your interface your representation it's not the truth.
00:15:49 It's it's there to keep you alive long enough to reproduce.
00:15:53 There is a question Do other people see the same thing that you do and the answer is probably their perception.
00:16:00 Are very very similar. So if I am seeing a red rose.
00:16:03 And you look at it
00:16:04 and you say that there is a red rose there's a good chance that your experiences of the Red Rose are very very similar
00:16:09 to to my experiences in the reason is not because there's an objective red rose in the world.
00:16:16 But because whatever the objective world is when I interact with it. I construct an interface.
00:16:23 That's very similar to the interface that you construct
00:16:26 when you interact with that world because we're members of the same species.
00:16:30 We're not exactly the same your genes are slightly different from mine. There are mutations.
00:16:35 And so we're not going to see exactly the same thing.
00:16:39 And there are cases where we actually know this for facts in the case of color perception we know that roughly one
00:16:44 third of men have one illegal for the red photoreceptor and the other two thirds have a different a little
00:16:51 and they actually see the red orange yellow end of the spectrum a little bit differently from each other.
00:16:56 It's a measurable difference so we know that there are differences in the D.N.A.
00:17:00 Which lead to measurable differences in the interface.
00:17:04 But nevertheless we could know we could say that our perceptions are substantially similar. But.
00:17:25 You know how do we know if we see the truth. Oh we're right. So what if I see a snake.
00:17:35 For example you might say well you know if you think that the snake is just an icon on your interface.
00:17:43 Why don't you touch it. Why don't you play with it because it will bite you.
00:17:46 And after you're dead you know what will know that snake was more than just an icon in your desktop interface it's real
00:17:53 it's a real part of objective reality. And I wouldn't touch that snake. For the same reason.
00:18:00 I wouldn't take my blue rectangular icon on my screen and drag it carelessly to the trash can on my screen.
00:18:09 I don't drag it carelessly to the trash can.
00:18:11 Not because I take the icon literally the file is not literally blue and rectangular but I do take it seriously.
00:18:18 If I drag that icon to the trash can. I could lose.
00:18:21 Who knows how much you know your of work if it's a long paper I'm writing her book or something.
00:18:25 Could be a lot of work that I lose. So the interface.
00:18:29 I don't take it literally like the files are blue and rectangular but I do take it seriously.
00:18:35 In the same thing is true but our perceptions in our everyday life. So if I see a snake.
00:18:39 That's an icon I better take quite seriously if you see a snake. Don't touch it.
00:18:44 If there's a train coming down the tracks don't step in front of it. So take my my perceptions quite seriously.
00:18:51 But it's a logical error. To then say We must therefore take them literally.
00:18:58 That doesn't follow the fact we take them seriously
00:19:01 and must take them seriously does not entail that we have to take them literally that's a logical error
00:19:06 but it's one that we seem to be inclined to as a species we know that we have to be very very if you see a cliff don't
00:19:12 step on if you see a car goes to her front of it so we know that we have to take our perceptions quite seriously
00:19:17 and it's just natural for us. Somehow as human beings to say that means that we're seeing the truth.
00:19:22 Well no it doesn't it doesn't mean that all psychologically it does but logically it doesn't.
00:19:27 And so that's the the error that we all fall into. It's.
00:19:32 It's very interesting very human one that I feel mean this is very very unnatural for me as well I mean I know I assume
00:19:39 that because the car could hurt me. It's real but no the car is just my interface to something out there.
00:19:46 I don't know what it is and my interface is telling me certain behaviors I better do
00:19:50 or not do so that that whatever that real world is out there doesn't have impacts on me that I don't like.
00:20:01 Yes you are right. You know that's that's right.
00:20:10 So the first step then is that if you buy this interface idea space
00:20:14 and time as we see them are not the nature of reality physical objects so matter momentum mass position all this stuff
00:20:23 that's not the nature of reality either. So the first often is to just say what we thought was reality isn't reality.
00:20:32 There is some reality that's out there and the first step.
00:20:35 And we actually don't know what it is right as scientists we don't know what it is
00:20:39 and is best for us to just recognize that we don't know what it is that even the very language of our perceptions the
00:20:46 language of space and time the language of matter position momentum spin
00:20:52 and so forth is in fact the wrong language to describe reality. You can't possibly describe reality in that language.
00:21:01 In in the same way that for example suppose I had a class A computer science class
00:21:07 and I gave the students the following assignment.
00:21:09 You can only use the language of the pixels on the desktop screen of your interface that's the only language you can
00:21:16 use the pixels and I want you to use that language to describe exactly how a computer works.
00:21:22 What's really going on inside a computer that's the only language. Well good luck. Everybody's going to fail.
00:21:26 You can't use the language of pixels.
00:21:28 It's the wrong language to describe the voltage is a magnetic fields and so forth. That's that's inside the computer.
00:21:35 Similarly if we try to use the language of space and time and matter and motion particles
00:21:42 and so forth to describe ultimate reality.
00:21:45 We're guaranteed to fail because the that language does not have the possibility to describe the truth of the world
00:21:52 around us as scientists then we have to step back and ask what language might work. Why. Do we.
00:22:00 It's also possible by the way that we don't have the concepts necessary to describe reality right.
00:22:11 We don't expect that monkeys have the language and the knowledge the concepts needed to understand quantum mechanics.
00:22:22 No one would ever try to teach quantum mechanics to a monkey.
00:22:24 They simply lack the concepts that are needed even address the subject and it's quite possible that homo sapiens
00:22:30 or species has not evolved the concepts that are needed to understand the true nature of reality.
00:22:37 Now I can't dismiss the possibility. I mean we're just another species like the monkeys. I don't want to give up.
00:22:45 I mean as a scientist. I'm not going to say. Therefore I'm just you know let's let's have a drink and not worry.
00:22:50 We're going to say let's let's try
00:22:52 but we need to be very very aware of the possibility of limitations in our conceptual system.
00:22:58 And limitations from our perceptual system.
00:23:00 What's very clear to me is that the the perceptual language that we've evolved has no chance of being the right
00:23:08 language to describe reality.
00:23:10 No chance the we can actually show that the probability that our language of perception is the right language to
00:23:16 describe reality is zero is precisely zero so that might be precisely zero. That's right. So. To write. You. When you.
00:23:51 Well most of my colleagues believe that our perceptual systems evolved to tell us. Truth about reality around us.
00:24:02 Not all of the truth. No one thinks that we see all of the truth.
00:24:05 We can only see light in a narrow band we can't see cosmic rays or X. Rays or you know radio waves.
00:24:11 So no one believes that we see all of reality but most of vision scientists
00:24:16 and cognitive scientists think that our perceptual systems have evolved to report the truth because they feel that.
00:24:24 Sensory systems that report the truth give you a competitive advantage against you
00:24:29 when you're competing with other organisms so the organisms that see reality as it is are more fit
00:24:35 and more likely to pass on their genes to the next generation.
00:24:39 So that's the standard view in every generation the organisms that saw a reality.
00:24:46 More closely the way it is had a competitive advantage
00:24:48 and were more likely to pass on their genes that could for their sensory systems to the next generation.
00:24:54 So after thousands of generations we can be pretty confident that we're the offspring of those who saw more truly in
00:25:00 each generation. So we can be confident that we see reality as it is not exhaustive fully but truly.
00:25:09 I'm not completely on my own. I would say.
00:25:12 My estimate is that maybe five percent maybe five percent of my colleagues might be you know game they might believe
00:25:22 that we're not seeing reality as it is
00:25:26 and some very very very good colleagues so yawn can drink for example from the Netherlands.
00:25:32 Arguably the brightest vision scientist alive today.
00:25:36 Does agree with me he thinks that evolution by natural selection does not favor true perceptions
00:25:42 and that we just have interface perceptions in a number of my colleagues that I've talked with initially are quite
00:25:51 and they even have the you know the reaction I thought you were smart guy until you said that is really a stupid idea
00:25:57 that we don't see reality as it is. But I.
00:26:00 After talking with them for an hour or two about evolution and the mathematics.
00:26:03 Then a lot of my colleagues do these come around to say well OK maybe I mean at least I can't reject this idea out of
00:26:12 But I think it's catching on my hope is actually not so much to get my generation of scientists to to follow this idea
00:26:21 but to get the graduate students the next generation that's what I'm really after so much
00:26:25 and they seem to be quite open to it the next generation of scientists the young scientists have seen the Matrix.
00:26:31 They've seen movies like that their minds are more open to the possibility that we don't see reality as it is
00:26:38 and I think that the next generation is going to really catch this catch on to this and
00:26:42 when we have right now our desktops on our computers are flat but very very soon.
00:26:48 We're going to have holographic dustups you'll be interacting with a three D.
00:26:51 Duster you pull open your laptop and they'll be a three D.
00:26:53 Virtual world that opens up in front of you and you'll be moving icons around in three D.
00:26:58 and Suddenly the idea that a three D.
00:27:02 World could just be a virtual world just an interface and not the truth will become not a strange weird idea
00:27:10 but part of your everyday experience every time you open your laptop so I think that you know what I'm saying right now.
00:27:16 I mean.
00:27:17 In fifteen twenty years people will ask why was that so hard for people to even understand back then is so silly.
00:27:27 So what happens to the truth then is first we have to be very very careful in our claims about truth.
00:27:36 Any of the normal language we use of space and time and matter
00:27:40 and particles is almost surely the wrong language into the mention it's possible that we don't have the right language
00:27:48 that none of our concepts are adequate. But I don't want to be a solid assist.
00:27:54 So a solid system from the philosophical point of view is someone who claims that nothing exists.
00:28:00 Except me in my perceptions. I'm not a softest I think that there does exist something besides me in my perceptions.
00:28:10 And I have to first say I don't know what it is so as a scientist right off the confession is now that I've given up
00:28:19 space and time and Mattern in physical objects as the nature of reality by I honestly don't know
00:28:25 but the nature of objective reality is but as a scientist. It's my job to theorize I can make proposals.
00:28:33 I'm probably going to be wrong
00:28:34 but the idea in science is to make specific precise mathematically precise if you can mathematically precise proposals
00:28:43 about the nature of reality.
00:28:45 Knowing full well that you're probably wrong
00:28:47 but being so precise that you can then do experiments to prove that you're wrong
00:28:53 and then figure out how you might change your theory so that you can get something that's not quite as wrong.
00:28:58 So that's what I've been working on and the direction I'm pursuing is motivated in the following way.
00:29:10 Perhaps I know nothing. There's a good chance that everything I believe is false but if I know anything at all.
00:29:20 I know that I'm experiencing headaches smells sounds visual perceptions
00:29:26 and so forth as experiences not as a truth about an external world just as my experiences.
00:29:33 So a headache is a good example because a headache is you know something that no one else can see
00:29:38 and you can't see my headache I can't see your headache I can't experience it.
00:29:42 It's my own personal experience and it's real as an experience.
00:29:49 It's not real as a claim about the external world
00:29:51 but it is real as an experience to me if you said all your headache isn't real. I'd be very angry with you.
00:29:56 It is real headache and you are my mom I need aspirin for. So so my idea is to say I could be wrong about everything.
00:30:06 But if I'm not wrong. If a film wrong about experiences but having experiences then it's really game over.
00:30:13 There's not really any place I can go.
00:30:16 So I'm going to start with that I'm going to start with there are experiences
00:30:22 and so have a mathematical model of what I call a conscious agent something like me they can have experiences conscious
00:30:31 experiences of smells and tastes and colors and sound.
00:30:47 Yes yes the mathematical structure but that's was about to describe things. That's right exactly right.
00:30:59 So I'll try to describe them informally. And so.
00:31:15 So the idea of them is to have what I call a conscious agent that has conscious experiences sights smells sounds
00:31:23 and tastes that can then make choices based on what it experiences
00:31:30 and then once it's made a choice about what it wants to do it can then act on the world whatever that world is in the
00:31:39 net world will again affect our insistence so there's a loop between the world affecting my experiences my experience
00:31:46 is affecting the decisions I make about how to act in those actions down working on the world.
00:31:52 It's a loop
00:31:54 and then I also think about having a counter for every experience I have I can have my own little personal time which
00:32:00 is a counter of the experiences and in I've discussed it here informally but we've made this a mathematical model
00:32:07 and what we're trying to do is to develop this will we call theory of conscious agents of having networks so the so the
00:32:13 idea is there is a universe that exists independent of me whether or not I existed.
00:32:19 But it's a universe of consciousness of conscious agents agents that have experiences make decisions
00:32:25 and act interacting with each other so so I'm just one. I'm one participating in this.
00:32:35 And in fact I'm not just one i'm when we look at the whole theory.
00:32:38 I'm perhaps an infinite lattice of these conscious agents all interacting.
00:32:44 But and then so are you so as everybody is not just one can't you're one conscious agent
00:32:49 but you're also two roughly corresponding to the two hemispheres of your brain
00:32:54 and then within each of us for more conscious agents to perhaps an indefinite indefinitely large number. But. We. So.
00:33:21 That's right. Sue so that this theory.
00:33:23 I would so again I could be wrong
00:33:25 but what I'm proposing is that consciousness is fundamental is the fundamental nature of reality and
00:33:33 but I don't want to just have that be some kind of loose. You know semi spiritual kind of idea.
00:33:40 I'm trying to get it mathematically precise idea.
00:33:42 So what do I mean by consciousness I'm getting a mathematical model of what I mean by consciousness that absolute
00:33:47 precise mathematical precise and I call this mathematical model conscious agent and then it turns it.
00:33:55 It's all works out very very well actually the mathematics is all is quite clear
00:33:59 and we've published a paper with the mathematics has been out for a couple years.
00:34:07 Yes and as dynamical systems and we can write down the equations of the dynamics.
00:34:11 It's a very very rich mathematical area.
00:34:15 So you mean most of the time when you hear people say I think consciousness is fundamental.
00:34:19 It's more about well this meditating hold hands and
00:34:22 and things like that that what I'm trying to do is to take that idea and make it very very rigorous.
00:34:27 Here's a mathematical model of consciousness. These are the equations of that an Amex.
00:34:32 So so so the goal is to get a mathematically precise model of consciousness that we can then use.
00:34:41 To solve one of the biggest unsolved problems in science the so-called mind body problem.
00:34:46 This is a problem that has perplexed human beings for thousands of years.
00:34:52 And that is what is the relationship between our conscious experiences.
00:34:57 The taste of garlic the smell of an onion the sound of a trumpet and our physical bodies the physical world.
00:35:05 What is that relationship. How should we understand it. Most neuroscientists and philosophers of mind today are.
00:35:12 Trying to solve the problem by saying that neural activity in the brain is the foundation. That's the reality.
00:35:20 So neurons in space and time physical objects and their dynamics create or are there identical to consciousness.
00:35:29 So somehow when you get a complicated system of neurons somehow their dynamics
00:35:34 or their properties boot up consciousness
00:35:38 but the surprising thing is that we've never been able to get a theory of of how that could be there are ideas maybe
00:35:46 some how information theoretic properties of the dynamics of neural networks.
00:35:52 Maybe somehow those could bring up consciousness we do have correlations or we know that
00:35:57 when that your conscious your brain has certain information theoretic properties of its dynamics that's certainly true.
00:36:06 That's right. It's the other way so we do this right.
00:36:16 So instead of going from physics to the consciousness all start with cops doesn't get physics or go the other way.
00:36:27 Yup yup with that up. Yeah yeah yeah. So.
00:36:31 So most neuroscientists in philosophers of mind are trying to start with properties of neurons neural networks
00:36:38 and neural activity and to try to then get a theory of how consciousness could emerge from that
00:36:45 or somehow be identical to that neural activity.
00:36:48 And there are a lot of ideas about how we might get a scientific theory.
00:36:54 Information theoretic properties of of the I makes certain quantum properties of microtubules maybe certain you know
00:37:01 frequencies of firings of neurons and things like that.
00:37:07 But there's not yet been any scientific theory that's actually been proposed which says.
00:37:12 This neural activity with these say information theoretic properties has to be the taste of chocolate.
00:37:19 It could not be the taste of a strawberry it could not be a headache and these are the mathematical reasons why.
00:37:26 So we need laws that take us from neural activity.
00:37:30 Whatever the properties of neural activity are that we want to propose
00:37:32 or the foundation takes us from those properties of the neurons into the specific conscious experiences
00:37:38 and explain exactly why this neural activity lawfully must be that conscious experience that has never been done.
00:37:47 So there. So it when I say there are no scientific theories. That's what I'm saying.
00:37:51 No one has ever proposed laws to say this neural activity.
00:37:56 Based on this law must be the taste of chocolate it could not be the smell of garlic.
00:38:02 Nothing is is even close to trying to do that. So there are so put a very boldly there are no scientific theories.
00:38:10 That start with a physical description of the brain neural activity and give you consciousness.
00:38:15 There's nothing remotely plausible and there are no good ideas about how that might be done.
00:38:20 That's the state of play and we should be very very frank about it. There are no scientific theories.
00:38:24 There are no remotely plausible ideas about how to do that and that's what got me thinking about this. I mean I tried.
00:38:32 I'm a physicalist but at heart like everybody else but when everybody's failing deeply
00:38:38 and I have no good ideas no one has any good ideas about how to start with a brain and get consciousness.
00:38:44 I decided let's try the other direction.
00:38:46 So let's try to solve the mind body problem with the theory of consciousness on its own terms.
00:38:51 So first start with consciousness
00:38:53 and say propose as a scientific hypothesis that consciousness is fundamental get a mathematical model of it
00:39:01 and then solve the mind body problem. The other direction.
00:39:04 So instead of starting with physics
00:39:06 and getting consciousness start with consciousness mathematically described not a hand wave a mathematical model can.
00:39:12 Yes and get back all of quantum physics
00:39:15 and relativity theory that's the that would be solving the mind body problem in the other direction. That's right.
00:39:22 So ultimately we as a scientist.
00:39:24 We want one theoretical framework that covers everything we know right from the physicalist point of view would most
00:39:31 people or you want to start with what we know about physics you know string theory relativity theory and so forth
00:39:36 and then neural networks and their activity
00:39:38 and get consciousness out so we have one big picture that of the universe that gets it all
00:39:42 and we haven't been able to do that because we can't get consciousness in so I'm trying to start with a mathematical
00:39:49 model of consciousness and its dynamics and then see if I can't get out. You know string theory.
00:39:56 Quantum gravity
00:39:58 and maybe ideally make some new predictions that the physicists haven't made if I can do that then we're off to a real
00:40:04 scientific adventure.
00:40:08 But yeah it would be it would be of a scientific breakthrough if we could if we could do that because partly because it
00:40:13 would be unifying to things that we've never been able to unify namely consciousness our conscious experiences
00:40:19 and what we take to be the physical world. If this unification works.
00:40:24 What it would reveal as what we took to be an independent objective space time physical reality is simply a species
00:40:34 specific user interface and different species will have evolved different user interfaces.
00:40:42 Maybe they don't use space and time. Maybe they don't use color.
00:40:45 Maybe they used senses in formats of interfaces that we can't even imagine.
00:40:51 And it's very easy for us to blow out our imagination. Right.
00:41:00 So we have our interface in terms of space and time and objects snakes
00:41:05 and trains that we have to avoid to keep us alive.
00:41:08 That's it's evolved to keep us alive but there are many ways to stay alive. Right.
00:41:13 Evolution shapes different organisms for different niches with different gambits different strategies for staying alive.
00:41:21 Ours is just one of millions.
00:41:24 We we know there been many many millions of species that have lived even just on this one planet.
00:41:30 And we know that the nature of their perceptual experiences in general is very very different from ours.
00:41:39 There are snakes that see an infrared.
00:41:44 You know fish that see electric fields
00:41:49 and sense electrical things that we can't even imagine what it would be like in even birds for example that have four
00:41:56 color receptors. We only have three Try to imagine a specific color that you've never seen before.
00:42:08 Nothing happens right. You can't even imagine a specific concrete color that you've never seen before.
00:42:13 And yet apparently pigeons are in a a richer color world they're experiencing colors that perhaps no human can even
00:42:22 imagine and their animals that he even have more color receptors the the mantis shrimp has ten or more color receptors.
00:42:30 So try to imagine their color I can't even imagine in some sense. My my sensory systems are a window on the world.
00:42:40 But they're also a prism.
00:42:42 I can't actually see outside of it
00:42:45 and I can't even concretely imagine perceptual experience outside of it I can do it abstractly I could imagine
00:42:52 abstractly a world that's not three dimensional I can imagine a four dimensional world mean Einstein did that.
00:42:59 And it's hard but you can you can imagine a four dimensional world
00:43:01 or you know mathematicians can go to any dimension you want. We can go there.
00:43:07 Conceptually but nobody not even the most brilliant mathematician can conquer.
00:43:12 Create Lee imagine in their mind a four dimensional world just like you can't imagine completely a specific new color
00:43:19 that you've never seen before. So our our desktop interface is a species specific interface.
00:43:28 It's our window on the world.
00:43:29 It's our way to stay alive it gives us the symbols we need in our particular nice homo sapiens has taken a particular
00:43:38 kind or set of niches. The parent paramecium equal I all these various organisms.
00:43:44 They have different issues they don't need the same user interface that we have so the interface is going to vary
00:43:49 widely from from from organism to organism. But if I'm correct about this conscious agent thesis.
00:43:56 It's consciousness all the way down different different user interfaces that are allowing different conscious agents to
00:44:03 do what they need to do. But in their own format. That's right that I see. But you know we.
00:44:22 You and I have very very similar interfaces is my assumption.
00:44:26 Again that as a scientist I can never say no for sure because. Exactly right.
00:44:34 We are I think so because you and I are members of the same species.
00:44:44 It's reasonable for me to assume are we guess that's right. Well that the yeah it's interesting.
00:44:50 My my perceptions of classified you as being similar to me
00:44:54 but again it's fallible as a scientist I never can say for sure.
00:44:58 Anything I can only give probabilities of but my statements but I think it's highly likely that
00:45:05 when I'm interacting with you I'm interacting with someone who's perceptions there who's interface is.
00:45:12 Very very similar to mine. There's no way for me to actually prove that your experiences are dentical to mine.
00:45:21 In fact I have a mathematical proof that I published about eight years ago that actually proves that we that we can't
00:45:26 do that. So it's actually a theorem that there's no way for us to verify that your experiences.
00:45:34 Are the same as mine even.
00:45:45 You know the fact that I well I wouldn't say that I would say that the fact that maybe what I'm saying is a little
00:45:51 difficult for you
00:45:52 or other people to understand is not so much a matter of a difference of the user interface that we have
00:45:57 or difference in intelligence. It's more just a matter of a difference in what we spend our time thinking about so.
00:46:05 So for example if if someone is crocheting I've never done any crocheting
00:46:12 and so things that are obvious to someone who spent their life crocheting are not obvious to me I could know I couldn't
00:46:16 pick up the darning needles and in the threads and do it.
00:46:23 I would say that the person is crocheting a smarter than me or dumber than me.
00:46:26 They just have had a different you know thing that they focused on
00:46:29 and so the same thing here with the user interface idea.
00:46:32 It's difficult to understand partly because our species seems to want to think that what we experience is the truth.
00:46:41 So we seem to have that inclination.
00:46:46 But I do think that you are very very similar to me in your in your perceptual experiences and but I could be wrong.
00:46:55 And I do know it's it's a theorem that even if in every experiment you behave exactly the same way as me.
00:47:03 So in every experiment.
00:47:05 Same color perception you give exactly the same answers as me
00:47:08 and we even do brain scans in your brain is behaving exactly the same way. As mine.
00:47:14 You might say well that must prove that your color experiences are identical to mine and it turns out.
00:47:19 No it doesn't prove that they're identical.
00:47:21 It makes it likely from my point of view as a scientist that the that they're the same but it isn't it's not a proof.
00:47:28 and in fact that might be one thing we could talk about is something called synesthesia So these are people who see
00:47:33 when when you see you just see but they also maybe hear something or if they see a letter they see a color
00:47:40 and so this mean these are Pete These are real people.
00:47:43 This is not real science fiction it's a synesthesia is real people that have what we call mixing of the senses
00:47:48 but that really shows that our user interface could be very very different. So we could talk about that. Yeah. Then.
00:48:06 That's right.
00:48:07 So it's one thing to say that our perceptions don't report the truth that they're just user interface
00:48:17 and anybody can say that in you might ask.
00:48:19 Well what's the logical argument on what grounds are you making such a wild claim. And.
00:48:25 The argument is based on evolution by natural selection most researchers in my field have the argument that I mentioned
00:48:33 earlier that. Those of us who saw reality as it is had a competitive advantage compared to those who don't.