See all Lex Fridman transcripts on Youtube

Speaker 1

00:00

The following is a conversation with John Hopfield, professor at Princeton, whose life's work weaved beautifully through biology, chemistry, neuroscience, and physics. Most crucially, he saw the messy world of biology through the piercing eyes of a physicist. He's perhaps best known for his work on associative neural networks, now known as Hopfield networks, that were 1 of the early ideas that catalyzed the development of the modern field of deep learning. As his 2019 Franklin Medal in Physics Award states, he applied concepts of theoretical physics to provide new insights on important biological questions in a variety of areas, including genetics and neuroscience with significant impact on machine learning.

Speaker 1

00:47

And as John says in his 2018 article titled, Now What? His accomplishments have often come about by asking that very question, now what? And often responding by a major change of direction. This is the Artificial Intelligence Podcast.

Speaker 1

01:06

If you enjoy it, subscribe on YouTube, give it 5 stars on Apple Podcast, support it on Patreon, or simply connect with me on Twitter, at Lex Friedman, spelled F-R-I-D-M-A-N. As usual, I'll do 1 or 2 minutes of ads now and never any ads in the middle that can break the flow of the conversation. I hope that works for you and doesn't hurt the listening experience. This show is presented by Cash App, the number 1 finance app in the App Store.

Speaker 1

01:34

When you get it, use code LEXPODCAST. Cash App lets you send money to friends, buy Bitcoin, and invest in the stock market with as little as $1. Since Cash App does fractional share trading, Let me mention that the order execution algorithm that works behind the scenes to create the abstraction of fractional orders is to me an algorithmic marvel. So big props to the Cash App engineers for solving a hard problem that in the end provides an easy interface that takes a step up the next layer of abstraction over the stock market, making trading more accessible for new investors and diversification much easier.

Speaker 1

02:12

So again, if you get Cash App from the App Store or Google Play and use code LEXPODCAST, you'll get $10, and Cash App will also donate $10 to FIRST, 1 of my favorite organizations that is helping advance robotics and STEM education for young people around the world. And Now, here's my conversation with John Hopfield. What difference between biological neural networks and artificial neural networks is most captivating and profound to you?

Speaker 2

02:44

At the higher philosophical level, Let's not get technical just yet.

Speaker 3

02:50

1 of the things that very much intrigues me is the fact that neurons have all kinds of components, properties to them. And evolutionary biology, if you have some little quirk in how a molecule works or how a cell works, and it can be made use of, evolution will sharpen it up and make it into a useful feature rather than a glitch. And so you expect in neurobiology for evolution to have captured all kinds of possibilities of getting neurons, of how you get neurons to do things for you.

Speaker 3

03:33

And that aspect has been completely suppressed in artificial neural networks.

Speaker 2

03:40

Do the glitches become features in the biological neural network?

Speaker 3

03:47

They can. Look, let me take 1 of the things that I used to do research on. If you take things which oscillate, have rhythms which are sort of close to each other.

Speaker 3

04:02

Under some circumstances, these things will have a phase transition, and suddenly the rhythm will... Everybody will fall into step. There was a marvelous physical example of that in the Millennium Bridge across the Thames River about... Built about

Speaker 1

04:18

2001.

Speaker 3

04:21

And pedestrians walking across, pedestrians don't walk synchronized, they don't walk in lockstep. But they're all walking about the same frequency. And the bridge could sway at that frequency, and the slight sway made pedestrians tend a little bit to lock into step.

Speaker 3

04:37

And after a while, the bridge was oscillating back and forth, and the pedestrians were walking in step to it. And you could see it in the movies made out of the bridge. And the engineers made a simple-minded mistake. They assumed when you walk, it's step, step, step, and it's back and forth motion.

Speaker 3

04:56

But when you walk, it's also right foot, left foot, side to side motion. And it's the side to side motion for which the bridge was strong enough, but it wasn't stiff enough. And as a result, you would feel the motion, and you'd fall into step with it. And people were very uncomfortable with it.

Speaker 3

05:15

They closed the bridge for 2 years while they built stiffening for it. Now, nerves, look, nerve cells produce action potentials. You have a bunch of cells which are loosely coupled together producing action potentials of the same rate. There'll be some circumstances under which these things can lock together.

Speaker 3

05:35

Other circumstances in which they won't. Well, if they fire together, you can be sure that other cells are gonna notice it. So you can make a computational feature out of this in an evolving brain. Most artificial neural networks don't even have action potentials, let alone have the possibility for synchronizing them.

Speaker 2

05:59

And you mentioned the evolutionary process. So the evolutionary process that builds on top of biological systems leverages that, the weird mess of it somehow. So how do you make sense of that ability to leverage all the different kinds of complexities in the biological brain?

Speaker 3

06:24

Well, look, at the biological molecule level, you have a piece of DNA which encode for a particular protein. You could duplicate that piece of DNA and now 1 part of it can code for that protein but the other 1 could itself change a little bit and thus start coding for a molecule which is slightly different. Now that molecule which is slightly different had a function which helped any old chemical reaction which was important to the cell.

Speaker 3

07:00

It would go ahead and let that try, and evolution would slowly improve that function. And so you have the possibility of duplicating, and then having things drift apart. 1 of them retain the old function, the other 1 do something new for you. And there's evolutionary pressure to improve.

Speaker 3

07:23

Look, there is in computers too, but it's improvement has to do with closing some companies and opening some others. The evolutionary process looks a little different.

Speaker 2

07:34

Yeah, similar time scale, perhaps.

Speaker 3

07:37

Much shorter in time scale.

Speaker 2

07:39

Companies close, yeah, go bankrupt and are born, yeah, shorter, but not much shorter. Some companies last a century,

Speaker 1

07:49

but yeah, you're right.

Speaker 2

07:50

I mean, if you think of companies as a single organism that builds and you all know, yeah, it's a fascinating dual correspondence there between biological and...

Speaker 3

08:02

And companies have difficulty having a new product competing with an old product.

Speaker 2

08:07

Yeah.

Speaker 3

08:08

And when IBM built its first PC, you probably read the book, They made a little isolated internal unit to make the PC. And for the first time in IBM's history, they didn't insist that you build it out of IBM components. But they understood that they could get into this market, which is a very different thing by completely changing their culture.

Speaker 3

08:38

And biology finds other markets in a more adaptive way.

Speaker 2

08:45

Yeah, it's better at it. It's better at that kind of integration. So maybe you've already said it, but what to use the most beautiful aspect or mechanism of the human mind?

Speaker 2

09:01

Is it the adaptive, the ability to adapt as you've described, or is there some other little quirk that you particularly like?

Speaker 3

09:11

Adaptation is everything when you get down to it. But the difference, there are differences between adaptation, where your learning goes on only over generations, over evolutionary time, where your learning goes on at the timescale of 1 individual who must learn from the environment during that individual's lifetime. And biology has both kinds of learning in it.

Speaker 3

09:43

And the thing which makes neurobiology hard is that it's a mathematical system as it were built on this other kind of evolutionary system.

Speaker 2

09:58

What do you mean by mathematical system? Where is the math in the biology?

Speaker 3

10:03

Well when you talk to a computer scientist about neural networks, it's all math. The fact that biology actually came about from evolution and the fact that Biology is about a system which you can build in 3 dimensions. If you look at computer chips, computer chips are basically two-dimensional structures, maybe 2.1 dimensions, but they really have difficulty doing three-dimensional wiring.

Speaker 3

10:40

Biology is, neocortex is actually also sheet-like, and it sits on top of the white matter, which is about 10 times the volume of the gray matter and contains all of what you might call the wires. But there's a huge, the effect of computer structure on what is easy and what is hard is immense. And biology does, it makes some things easy that are very difficult to understand how to do computationally. On the other hand, you can't do simple floating point arithmetic, so it's awfully stupid.

Speaker 2

11:23

Yeah, and you're saying this kind of 3 dimensional complicated structure makes, it's still math. It's still doing math. The kind of math it's doing enables you to solve problems of a very different kind.

Speaker 3

11:38

That's right, that's right.

Speaker 2

11:40

So you mentioned 2 kinds of adaptation. The evolutionary adaptation and the adaptation or learning at the scale of a single human life, which do you, which is particularly beautiful to you and interesting from a research and from just a human perspective? And which is more powerful?

Speaker 3

12:05

I find things most interesting that I begin to see how to get into the edges of them and tease them apart a little bit to see how they work. And since I can't see the evolutionary process going on, I'm in awe of it, but I find it just a black hole as far as trying to understand what to do. And so in a certain sense, I'm in awe of it, but I couldn't be interested in working on it.

Speaker 2

12:39

The human life's time scale is however thing you can tease apart and study.

Speaker 3

12:47

Yeah, you can do, there's developmental neurobiology which understands how the connections and how the structure evolves from a combination of what the genetics is like and the real, the fact that you're building a system in 3 dimensions.

Speaker 2

13:10

In just days and months, those early, early days of a human life are really interesting.

Speaker 3

13:17

They are and of course, there are times of immense cell multiplication. There are also times of the greatest cell death in the brain is during infancy.

Speaker 2

13:32

It's turnover.

Speaker 3

13:34

So what is not effective, what is not wired well enough to use at the moment, throw it out.

Speaker 2

13:42

It's a mysterious process. From, let me ask, from what field do you think the biggest breakthroughs in understanding the mind will come in the next decades? Is it neuroscience, computer science, neurobiology, psychology, physics, maybe math, maybe literature?

Speaker 2

14:06

Ha ha ha.

Speaker 3

14:09

Well, of course, I see the world always through a lens of physics. I grew up in physics. And the way I pick problems is very characteristic of physics and of an intellectual background which is not psychology, which is not chemistry and so on and so on.

Speaker 2

14:28

Both of your parents are physicists.

Speaker 3

14:30

Both of my parents were physicists and the real thing I got out of that was a feeling that the world is an understandable place. And if you do enough experiments and think about what they mean and structure things that you can do the mathematics of the relevant to the experiments. You also be able to understand how things work.

Speaker 2

14:55

But that was a few years ago. Did you change your mind at all through many decades of trying to understand the mind, of studying it different kinds of ways, not even the mind, just biological systems. You still have hope that physics, that you can understand.

Speaker 3

15:16

There's a question of what do you mean by understand?

Speaker 2

15:20

Of course.

Speaker 3

15:21

When I taught freshman physics, I used to say I wanted to give physics to understand the subject, to understand Newton's laws. I didn't want them simply to memorize a set of examples to which they knew the equations to write down to generate the answers, I had this nebulous idea of understanding. So that if you looked at a situation, you could say, Oh, I expect the ball to make that trajectory.

Speaker 3

15:48

I expect some intuitive notion of understanding. And I don't know how to express that very well. I've never known how to express it well. And you run smack up against it when you look at these simple neural nets, feed forward neural nets, which do amazing things and yet you know contain nothing of the essence of what I would have felt was understanding.

Speaker 3

16:20

Understanding is more than just an enormous lookup table.

Speaker 2

16:24

Let's linger on that. How sure you are of that? What if the table gets really big?

Speaker 2

16:31

So, I mean, ask another way, these feed forward neural networks, do you think they'll ever understand?

Speaker 3

16:40

Could answer that in 2 ways. I think if you look at real systems, feedback is an essential aspect of how these real systems compute. On the other hand, if I have a mathematical system with feedback, I know I can unlayer this and do it.

Speaker 3

17:00

But I have an exponential expansion in the amount of stuff I have to build if I'm gonna solve the problem that way.

Speaker 2

17:08

So feedback is essential. So we can talk even about recurrent neural networks, so recurrence. But do you think all the pieces are there to achieve understanding through these simple mechanisms?

Speaker 2

17:23

Like, back to our original question, what is the fundamental, is there a fundamental difference between artificial neural networks and biological? Or is it just a bunch of surface stuff?

Speaker 3

17:34

Suppose you ask a neurosurgeon, when is somebody dead? Yeah. They'll probably go back to saying, well, I can look at the brain rhythms and tell you this is a brain which is never going to function again.

Speaker 3

17:51

This 1 is, this other 1 is 1 which if we treat it well is still recoverable. And then just do that by some electrodes looking at simple electrical patterns which don't look in any detail at all at what individual neurons are doing. These rhythms are utterly absent from anything which goes on at Google.

Speaker 2

18:23

Yeah, but the rhythms.

Speaker 3

18:26

But the rhythms what?

Speaker 2

18:27

So, well that's like comparing, okay, I'll tell you. It's like you're comparing the greatest classical musician in the world to a child first learning to play. The question I'm at, but they're still both playing the piano.

Speaker 2

18:44

I'm asking, will it ever go on at Google? Do you have a hope? Because you're 1 of the seminal figures in both launching both disciplines, both sides of the river.

Speaker 3

19:00

I think it's going to go on generation after generation, the way it has, where what you might call the AI computer science community says, let's take the following. This is our model of neurobiology at the moment. Let's pretend it's good enough and do everything we can with it." And it does interesting things, and after a while, it sort of grinds into the sand, and you say, ah, something else is needed for neurobiology, and some other grand thing comes in and enables you to go a lot further.

Speaker 3

19:42

But we'll go into the sand again. I think there could be generations of this evolution. I don't know how many of them, and each 1 is going to get you further into what our brain does, and in some sense, pass the Turing test longer and for more broad aspects. And how many of these are there are going to have to be before you say, I've made something, I've made a human, I don't know.

Speaker 2

20:15

But your sense is it might be a couple. It might

Speaker 3

20:17

be a few. My sense is it might be a couple more. Yeah.

Speaker 3

20:20

And going back to my brain waves as it were. Yes. From the AI point of view, they would say, ah, maybe these are an heavy phenomenon and not important at all. The first car I had, a real wreck of a

Speaker 1

20:43

1936

Speaker 3

20:44

Dodge, go above 45 miles an hour and the wheels would shimmy.

Speaker 2

20:50

Yeah.

Speaker 3

20:53

Good speedometer, that. Now, nobody designed the car that way. The car is malfunctioning to have that.

Speaker 3

21:01

But in biology, if it were useful to know, when are you going more than 45 miles an hour, you just capture that, and you wouldn't worry about where it came from. Yeah. It's going to be a long time before that kind of thing, which can take place in large, complex networks of things, is actually used in the computation. Look, the, How many transistors are there in your laptop these days?

Speaker 2

21:34

Actually, I don't know the number. It's...

Speaker 3

21:37

On the scale of 10 to the

Speaker 1

21:38

10,

Speaker 3

21:39

I can't remember the number either. And all the transistors are somewhat similar. And most physical systems, with that many parts, all of which are similar, have collective properties.

Speaker 2

21:54

Yes.

Speaker 3

21:55

Sound waves in air, earthquakes, what have you, have collective properties. Weather. There are no collective properties used in artificial neural networks in AI.

Speaker 2

22:10

Yeah, it's very…

Speaker 3

22:12

If biology uses them, it's going to take us to more generations of things for people to actually dig in and see how they are used and what they mean.

Speaker 2

22:23

See, you're very right. We might have to return several times to neurobiology and try to make our transistors more messy.

Speaker 3

22:32

Yeah, yeah. At the same time, the simple ones will conquer big aspects. And I think 1 of the most biggest surprises to me was how well learning systems, which are manifestly non-biological, how important they can be actually, and how important and how useful they can be in AI.

Speaker 2

23:03

So if we can just take a stroll to some of your work that is incredibly surprising that it works as well as it does that launched a lot of the recent work with neural networks. If we go to what are now called Hopfield networks, can you tell me what is associative memory in the mind for the human side? Let's explore memory for a bit.

Speaker 3

23:34

Okay, what you mean by associative memory is, ah, you have a memory of each of your friends. Your friend has all kinds of properties, from what they look like, to what their voice sounds like, to where they went to college, where you met them, go on and on, what science papers they've written. If I start talking about a 5 foot 10, wire-aided cognitive scientist who's got a very bad back, it doesn't take very long for you to say, oh, he's talking about Jeff Hinton.

Speaker 3

24:12

I never mentioned the name or anything very particular. But somehow a few facts that are associated with this, with a particular person enables you to get a hold of the rest of the facts. Or not the rest of them, another subset of them. And it's this ability to link things together, link experiences together, which it goes under the general name of associative memory.

Speaker 3

24:44

And a large part of intelligent behavior is actually just large associative memories at work, as far as I can see.

Speaker 2

24:53

What do you think is the mechanism of how it works in the mind? Is it a mystery to you still? Do you have inklings of how this essential thing for cognition works?

Speaker 3

25:09

What I made

Speaker 1

25:12

35

Speaker 3

25:12

years ago was of course a crude physics model to show the kind, actually enable you to understand, my old sense of understanding as a physicist, because you could say, ah, I understand why this goes to stable states, it's like things going downhill.

Speaker 2

25:32

Right.

Speaker 3

25:33

And that gives you something with which to think in physical terms rather than only in mathematical terms.

Speaker 2

25:42

So you've created these associative artificial networks.

Speaker 3

25:47

That's right. And now if you look at what I did, I didn't at all describe a system which gracefully learns. I described a system in which you could understand how learning could link things together, how very crudely it might learn.

Speaker 3

26:09

1 of the things which intrigues me as I reinvestigate that system now to some extent is, look, I'll see you every second for the next hour or what have you. Each look at you is a little bit different. I don't store all those second-by-second images, I don't store 3,000 images, I somehow compact this information. So I now have a view of you, which I can use.

Speaker 3

26:46

It doesn't slavishly remember anything in particular, but it compacts the information into useful chunks, which are somehow, it's these chunks, which are not just activities of neurons, bigger things than that, which are the real entities which are useful to you.

Speaker 2

27:08

Useful to you to describe, to compress this information. Coming at you.

Speaker 3

27:13

And you have to compress it in such a way that if I get, if the information comes in just like this again, I don't bother to rewrite it. Or efforts to rewrite it simply do not yield anything because those things are already written. And that needs to be not, look this up, have I written this, have I started somewhere already.

Speaker 3

27:36

It's gotta be something which is much more automatic in the machine hardware.

Speaker 2

27:41

Right, so in the human mind, how complicated is that process, do you think? So you've created, it feels weird to be sitting with John Hopfield calling him Hopfield Networks, but. It is weird.

Speaker 2

27:57

Yeah, but nevertheless, that's what everyone calls him, so here we are. So that's a simplification. That's what a physicist would do. You and Richard Feynman sat down and talked about associative memory.

Speaker 2

28:09

Now, if you look at the mind, well, you can't quite simplify it so perfectly. Do you think that-

Speaker 3

28:18

Well, let me backtrack just a little bit. Yeah. Biology is about dynamical systems.

Speaker 3

28:27

Computers are dynamical systems. You can ask, if you want to model biology, if you want to model neurobiology, what is the time scale? There's a dynamical system in which, of a fairly fast time scale, in which you could say, the synapses don't change much during this computation, so I'll think of the synapses fixed and just do the dynamics of the activity. Or you can say, the synapses are changing fast enough that I have to have the synaptic dynamics working at the same time as the system dynamics in order to understand the biology.

Speaker 3

29:13

Most, if you look at the feedforward artificial neural nets, they're all done as learning is, First of all, I spend some time learning, not performing, and I turn off learning and I perform. Right. That's not biology. And so, as I look more deeply at neurobiology, even as an associate of memory, I've got to face the fact that the dynamics of a synapse change is going on all the time.

Speaker 3

29:44

And I can't just get by by saying, I'll do the dynamics of activity with fixed synapses.

Speaker 2

29:52

So the synaptic, the dynamics of the synapses is actually fundamental to the whole system.

Speaker 3

29:58

Yeah, yeah. And there's nothing necessarily separating the time scales. When the time scales can be separated, it's neat from the physicist's or the mathematician's point of view, but it's not necessarily true in neurobiology.

Speaker 2

30:14

So You're kind of dancing beautifully between showing a lot of respect to physics and then also saying that physics cannot quite reach the complexity of biology. So where do you land? Or do you continuously dance between the 2?

Speaker 3

30:33

I continuously dance between them, because my whole notion of understanding is that you can describe to somebody else how something works in ways which are honest and unbelievable, and still not describe all the nuts and bolts in detail. Weather. I can describe weather as 10 to the 32 molecules colliding in the atmosphere.

Speaker 3

31:05

I can simulate whether that way I have a big enough machine. I'll simulate it accurately. It's no good for understanding. If I want to understand things, I want to understand things in terms of wind patterns, hurricanes, pressure differentials, and so on.

Speaker 3

31:23

All things as they're collective. And The physicist in me always hopes that biology will have some things which can be said about it which are both true and for which you don't need all the molecular details of the molecules colliding. That's what I mean from the roots of physics by understanding.

Speaker 2

31:49

So what did, again, sorry, but Hopfield Networks help you understand, what insight did it give us about memory, about learning?

Speaker 3

32:01

They didn't give insights about learning. They gave insights about how things having learned could be expressed. How having learned A picture of you reminds me of your name." That would, but it didn't describe a reasonable way of actually doing the learning.

Speaker 3

32:27

It only said if you had previously learned the connections of this kind of pattern would now be able to behave in a physical way, which is to say, ah, if I put part of the pattern in here, the other part of the pattern will complete over here. I could understand that physics if the right learning stuff had already been put in. And it could understand why then putting in a picture of somebody else would generate something else over here. But it did not have a reasonable description of the learning process.

Speaker 2

33:03

But even, so forget learning, I mean that's just a powerful concept that sort of forming representations that are useful to be robust, you know, for error correction kind of thing. So this is kind of what the biology does we're talking about.

Speaker 3

33:22

Is yeah. And what BiPAPER did was simply enable you. There are lots of ways of being robust.

Speaker 3

33:33

If you think of a dynamical system, you think of a system where a path is going on in time. And if you think of a computer, there's a computational path which is going on in a huge dimensional space of ones and zeros. And an error correcting system is a system which, if you get a little bit off that trajectory, will push you back onto that trajectory again, so you get to the same answer in spite of the fact that there were things, the computation wasn't being ideally done all the way along the line. And there are lots of models for error correction, but 1 of the models for error correction is to say, there's a valley that you're following, flowing down, and if you push it a little bit off the valley, it's just like water being pushed a little bit by a rock.

Speaker 3

34:26

It gets back and follows the course of the river. And that, basically, the analog in the physical system, which enables you to say, oh yes, error-free computation and an associative memory are very much like things that I can understand from the point of view of a physical system. The physical system can be, under some circumstances, an accurate metaphor. It's not the only metaphor.

Speaker 3

34:59

There are error correction schemes which don't have a value and energy behind them. But those are error correction schemes which a mathematician may be able to understand, but I don't.

Speaker 2

35:13

So there's the physical metaphor that seems to work here.

Speaker 3

35:18

That's right, that's right.

Speaker 2

35:20

So these kinds of networks actually led to a lot of the work that is going on now in neural networks, artificial neural networks. So the follow on work with restricted Boltzmann machines and deep belief nets followed on from these ideas of the Hopfield network. So what do you think about this continued progress of that work towards now re-revigorated exploration of feed-forward neural networks and recurrent neural networks and convolutional neural networks and kinds of networks that are helping solve image recognition, natural language processing, all that kind of stuff.

Speaker 3

36:05

It's always intrigued me that 1 of the most long-lived of the learning systems is the Boltzmann machine, which is intrinsically a feedback network. And with the brilliance of Hinton and Sanofsky to understand how to do learning in that. And it's still a useful way to understand learning and understand, and the learning that you understand in that has something to do with the way that feedforward systems work.

Speaker 3

36:38

But it's not always exactly simple to express that intuition. But it always amuses me to see Hinton going back to the will yet again on a form of the Boltzmann machine, because really, that which has feedback and interesting probabilities in it is a lovely encapsulation of something computational.

Speaker 2

37:07

Something computational?

Speaker 3

37:10

Something both computational and physical. Computational in the, it's very much related to feed forward networks. Physical in that, both from machine learning is really learning a set of parameters for a physics Hamiltonian or energy function.

Speaker 2

37:29

What do you think about learning in this whole domain? Do you think the aforementioned guy, Jeff Hinton, all the work there with back propagation, all the kind of learning that goes on in these networks, how do you, if we compare it to learning in the brain, for example, is there echoes of the same kind of power that back propagation reveals about these kinds of recurrent networks? Or is it something fundamentally different going on in the brain?

Speaker 3

38:10

I don't think the brain is as deep as the deepest networks go, the deepest computer science networks. And I do wonder whether part of that depth of the computer science networks is necessitated by the fact that the only learning that's easily done on a machine is feed forward. And so there's the question of to what extent has the biology, which has some feed forward and some feed back, been captured by something which has got many more neurons, a bunch more depth than the neurons.

Speaker 2

38:56

So part of you wonders if the feedback is actually more essential than the number of neurons or the depth, the dynamics of the feedback.

Speaker 3

39:06

The dynamics of the feedback. Look, if you don't have feedback, it's a little bit like building a big computer and running it through 1 clock cycle. And then you can't do anything till you put, you'd reload something coming in.

Speaker 3

39:24

How do you use the fact that there are multiple clocks like, how do I use the fact that you can close your eyes, stop listening to me, and think about a chessboard for a few minutes without any input whatsoever.

Speaker 2

39:38

Yeah, that memory thing, that's fundamentally a feedback kind of mechanism. You're going back to something. Yeah, it's hard to understand.

Speaker 2

39:51

It's hard to introspect, let alone consciousness.

Speaker 3

39:58

Oh, let alone consciousness, Yes, yes.

Speaker 2

40:00

Because that's tied up in there too. You can't just put that on another shelf.

Speaker 3

40:06

Every once in a while I get interested in consciousness and then I go and I've done that for years and ask 1 of my betters as it were their view on consciousness. It's been interesting collecting them.

Speaker 2

40:21

What is consciousness? Let's try to take a brief step into that room.

Speaker 3

40:30

Well, I asked Marvin Minsky, the G1 in consciousness, and Marvin said, consciousness is basically overrated. It may be an epiphenomenon. After all, all the things your brain does, which are actually hard computations, you do non-consciously.

Speaker 3

40:55

And there's so much evidence that even the simple things you do, you can make decisions, you can make committed decisions about them. The neurobiologist can say, he's now committed, he's going to move the hand left before you know it.

Speaker 2

41:14

So his view that consciousness is not, that's just like little icing on the cake. The real cake is in the subconscious.

Speaker 3

41:21

Yeah, yeah. Subconscious, non-conscious. Non-conscious, what's the better word, sir?

Speaker 3

41:27

It's only that Freud captured the other word.

Speaker 2

41:30

Yeah, it's a confusing word, subconscious.

Speaker 3

41:33

Nicholas Chater wrote an interesting book. I think the title of it is, The Mind is Flat. Flat in a neural net sense, might be flat as something which is a very broad neural net without really any layers in depth, or the deep brain would be many layers and not so broad.

Speaker 3

42:00

In the same sense that if you push Minsky hard enough, he would probably have said, consciousness is your effort to explain to yourself that which you have already done.

Speaker 2

42:16

Yeah, it's the weaving of the narrative around the things that have already been computed for you.

Speaker 3

42:23

That's right. And then so much of what we do for our memories of events, for example, If there's some traumatic event you witness, you will have a few facts about it correctly done. If somebody asks you about it, you will weave a narrative which is actually much more rich in detail than that, based on some anchor points you have of correct things, and pulling together general knowledge on the other, but you will have a narrative.

Speaker 3

42:56

And once you generate that narrative, you are very likely to repeat that narrative and claim that all the things you have in it are actually the correct things. There was a marvelous example of that in the Watergate slash impeachment era of John Dean. John Dean, you're too young to know, had been the personal lawyer of Nixon. And so John Dean was involved in the cover-up, and John Dean ultimately realized the only way to keep himself out of jail for a long time was actually to tell some of the truths about Nixon.

Speaker 3

43:38

And John Dean was a tremendous witness. He would remember these conversations in great detail, and very convincing detail. And long afterward, some of the tapes, the secret tapes as it were, from which Jane was recalling these conversations, were published. And 1 found out that John Dean had a good but not exceptional memory.

Speaker 3

44:06

What he had was an ability to paint vividly and in some sense accurately the tone of what was going on.

Speaker 2

44:16

By the way, that's a beautiful description of consciousness. Do you, like where do you stand in your, today? Perhaps it changes day to day, but where do you stand on the importance of consciousness in our whole big mess of cognition?

Speaker 2

44:41

Is it just a little narrative maker, Or is it actually fundamental to intelligence?

Speaker 3

44:54

That's a very hard 1. When I asked Francis Crick about consciousness, He launched forward in a long monologue about Mendel and the peas. And how Mendel knew that there was something, and how biologists understood that there was something in inheritance which was just very, very different.

Speaker 3

45:16

And the fact that inherited traits didn't just wash out into a gray, but were this or this and propagated, that that was absolutely fundamental to biology. And it took generations of biologists to understand that there was genetics, and it took another generation or 2 to understand that genetics came from DNA. But very shortly after Mendel, thinking biologists did realize that there was a deep problem about inheritance. And Francis would have liked to have said, and that's why I'm working on consciousness.

Speaker 3

46:01

But of course, he didn't have any smoking gun in the sense of Mendel. And that's the weakness of his position. If you read his book, which he wrote with Koch, I think.

Speaker 2

46:15

Yeah, Christoph Koch, yeah.

Speaker 3

46:17

I find it unconvincing for the smoking gun reason. So I've gone on collecting views without actually having taken a very strong 1 myself, because I haven't seen the entry point. Not seeing the smoking gun from the point of view of physics, I don't see the entry point.

Speaker 3

46:41

Whereas in neurobiology, once I understood the idea of a collective evolution of dynamics, which could be described as a collective phenomenon, I thought, ah, there's a point where what I know about physics is so different from any neurobiologist that I have something I might be able to contribute.

Speaker 2

47:01

And right now there's no way to grasp that consciousness from a physics perspective.

Speaker 3

47:07

From my point of view, that's correct. And of course people, physicists like everybody else, think very muddily about things. You ask the closely related question about free will.

Speaker 3

47:23

Do you believe you have free will? Physicists will give an offhand answer and then backtrack, backtrack, backtrack where they realize that the answer they gave must fundamentally contradict the laws of physics.

Speaker 2

47:38

And that's your answering questions of free will and consciousness naturally lead to contradictions from a physics perspective. Because it eventually ends up with quantum mechanics and then you get into that whole mess of trying to understand how much, from a physics perspective, how much is determined, already predetermined, much is already deterministic about our universe. There's lots of different.

Speaker 3

48:03

And if you don't push quite that far, you can say essentially all of neurobiology, which is relevant, can be captured by classical equations of motion. Because in my view of the mysteries of the brain are not the mysteries of quantum mechanics, but the mysteries of what can happen when you have a dynamical system, driven system, with

Speaker 1

48:28

10

Speaker 3

48:29

to the 14 parts. That complexity is something which is, that the physics of complex systems is at least as badly understood as the physics of phase coherence in quantum mechanics.

Speaker 2

48:46

Can we go there for a second? You've talked about attractor networks and just maybe you could say what are attractor networks and more broadly, what are interesting network dynamics that emerge in these or other complex systems?

Speaker 3

49:05

You have to be willing to think in a huge number of dimensions, because in a huge number of dimensions, the behavior of a system can be thought of as just the motion of a point over time in this huge number of dimensions. And an attractor network is simply a network where there is a line and other lines converge on it in time. That's the essence of an attractor network.

Speaker 3

49:31

That's how you- In

Speaker 2

49:31

a highly dimensional space.

Speaker 3

49:35

And the easiest way to get that is to do it in a highly dimensional space, where some of the dimensions provide the dissipation, which means… Which, like if I have a physical system, trajectories can't contract everywhere, they have to contract in some places and expand in others. There's a fundamental classical theorem of statistical mechanics which goes under the name of Liouville's theorem, which says you can't contract everywhere. If you contract somewhere, you expand somewhere else.

Speaker 3

50:12

And it's an interesting physical systems. You get driven systems where you have a small subsystem, which is the interesting part, and the rest of the contraction and expansion, the physicists would say, is the entropy flow in this other part of the system. But basically, attractor networks are dynamics funneling down so that you can't be any... So that if you start somewhere in the dynamical system, you will soon find yourself on a pretty well-determined pathway which goes somewhere.

Speaker 3

50:46

If You start somewhere else, you'll wind up on a different pathway. But you don't have just all possible things. You have some defined pathways which are allowed and onto which you will converge. And that's the way you make a stable computer And that's the way you make a stable computer and that's the way you make a stable behavior.

Speaker 2

51:06

So in general, looking at the physics of the emergent stability in these networks, What are some interesting characteristics that, what are some interesting insights from studying the dynamics of such high dimensional systems?