back to indexJohn Hopfield: Physics View of the Mind and Neurobiology | Lex Fridman Podcast #76
link |
The following is a conversation with John Hopfield,
link |
Professor Princeton, whose life's work weaved beautifully
link |
through biology, chemistry, neuroscience, and physics.
link |
Most crucially, he saw the messy world of biology
link |
through the piercing eyes of a physicist.
link |
He's perhaps best known for his work
link |
on associative neural networks,
link |
now known as Hopfield Networks.
link |
They were one of the early ideas that catalyzed
link |
the development of the modern field of deep learning.
link |
As his 2019 Franklin Medal in Physics Award states,
link |
he applied concepts of theoretical physics
link |
to provide new insights on important biological questions
link |
in a variety of areas, including genetics and neuroscience,
link |
with significant impact on machine learning.
link |
And as John says in his 2018 article titled,
link |
Now What, his accomplishments have often come about
link |
by asking that very question, Now What,
link |
and often responding by a major change of direction.
link |
This is the Artificial Intelligence Podcast.
link |
If you enjoy it, subscribe on YouTube,
link |
give it five stars on Apple Podcasts,
link |
support it on Patreon, or simply connect with me on Twitter,
link |
and Lex Friedman, spelled F R I D M A N.
link |
As usual, I'll do one or two minutes of ads now
link |
and never any ads in the middle
link |
that can break the flow of the conversation.
link |
I hope that works for you
link |
and doesn't hurt the listening experience.
link |
This show is presented by Cash App,
link |
the number one finance app in the App Store.
link |
When you get it, use code Lex Podcast.
link |
Cash App lets you send money to friends, buy Bitcoin,
link |
and invest in the stock market with as little as $1.
link |
Since Cash App does fractional share trading,
link |
let me mention that the order execution algorithm
link |
that works behind the scenes
link |
to create the abstraction of fractional orders
link |
is to me an algorithmic marvel.
link |
So big props to the Cash App engineers
link |
for solving a hard problem
link |
that in the end provides an easy interface
link |
that takes a step up the next layer of abstraction
link |
over the stock market,
link |
making trading more accessible for new investors
link |
and diversification much easier.
link |
So again, if you get Cash App from the App Store,
link |
Google Play, and use code Lex Podcast,
link |
you'll get $10, and Cash App will also donate $10 to first,
link |
one of my favorite organizations
link |
that is helping advance robotics and STEM education
link |
for young people around the world.
link |
And now here's my conversation with John Hopfield.
link |
What difference between biological neural networks
link |
and artificial neural networks
link |
is most captivating and profound to you?
link |
At the higher philosophical level,
link |
let's not get technical just yet.
link |
One of the things that very much intrigues me
link |
is the fact that neurons have all kinds of components,
link |
properties to them.
link |
And evolutionary biology,
link |
if you have some little quirk
link |
into how a molecule works or how a cell works,
link |
and it can make use of evolution will sharpen it up
link |
and make it into a useful feature rather than a glitch.
link |
And so you expect in neurobiology
link |
for evolution to have captured all kinds of possibilities
link |
of getting neurons, of how you get neurons
link |
to do things for you.
link |
And that aspect has been completely suppressed
link |
in artificial neural networks.
link |
Do the glitches become features in the biological neural network?
link |
Look, let me take one of the things
link |
that I used to do research on.
link |
If you take things which oscillate,
link |
they have rhythms which are sort of close to each other.
link |
Under some circumstances,
link |
these things will have a phase transition
link |
and suddenly the rhythm will, everybody will fall into step.
link |
There was a marvelous physical example of that
link |
in the Millennium Bridge across the Thames River
link |
about, build about 2001.
link |
And pedestrians walking across,
link |
pedestrians don't walk, synchronize,
link |
they don't walk in lock step.
link |
But they're all walking about the same frequency.
link |
And the bridge could sway at that frequency
link |
and the slight sway made pedestrians
link |
tend a little bit to lock in the step.
link |
And after a while, the bridge was oscillating back and forth
link |
on the pedestrians for walking in step to it.
link |
And you could see it in the movies made out of the bridge.
link |
And the engineers made a simple mind at a mistake.
link |
They had assumed when you walk,
link |
it's step, step, step, and it's back and forth motion.
link |
But when you walk, it's also right foot left
link |
with side to side motion.
link |
It's the side to side motion
link |
for which the bridge was strong enough,
link |
but it wasn't stiff enough.
link |
And as a result, you would feel the motion
link |
and you'd fall into step with it.
link |
And people were very uncomfortable with it.
link |
They closed the bridge for two years
link |
while they built stiffening for it.
link |
Now, nerve cells produce action potentials.
link |
You have a bunch of cells which are loosely coupled together
link |
producing action potentials of the same rate.
link |
There'll be some circumstances
link |
under which these things can lock together.
link |
Other circumstances in which they won't.
link |
Well, if they fire together,
link |
you can be sure that the other cells are gonna notice it.
link |
So you can make a computational feature out of this
link |
in an evolving brain.
link |
Most artificial neural networks
link |
don't even have action potentials,
link |
let alone have the possibility for synchronizing them.
link |
And you mentioned the evolutionary process.
link |
So there, the evolutionary process
link |
that builds on top of biological systems
link |
leverages that the weird mess of it somehow.
link |
So how do you make sense of that ability
link |
to leverage all the different kinds of complexities
link |
in the biological brain?
link |
Well, look, in the biological molecule level,
link |
you have a piece of DNA
link |
which encodes for a particular protein.
link |
You could duplicate that piece of DNA
link |
and now one part of it encodes for that protein,
link |
but the other one could itself change a little bit
link |
and thus start coding for a molecule
link |
which is slightly different.
link |
Now, that molecule was just slightly different.
link |
Had a function which helped any old chemical reaction
link |
was important to the cell.
link |
You would go ahead and let that evolution
link |
slowly improve that function.
link |
And so you have the possibility of duplicating
link |
and then having things drift apart,
link |
one of them retain the old function,
link |
the other one do something new for you.
link |
And there's evolutionary pressure to improve.
link |
Look, there isn't in computers too,
link |
but it's improvement has to do with closing
link |
some companies and openings of others.
link |
The evolutionary process looks a little different.
link |
Similar timescale, perhaps.
link |
What's shorter in timescale?
link |
Companies close, yeah, go bankrupt and are born,
link |
yeah, shorter, but not much shorter.
link |
Some company lasts the century, but yeah, you're right.
link |
I mean, if you think of companies as a single organism
link |
that builds and you all know, yeah,
link |
it's a fascinating dual correspondence there
link |
between biological.
link |
And companies have difficulty having a new product,
link |
competing with an old product.
link |
And when IBM built its first PC,
link |
you probably read the book,
link |
they made a little isolated internal unit to make the PC.
link |
And for the first time in IBM's history,
link |
they didn't insist that you build it out of IBM components.
link |
But they understood that they could get into this market,
link |
which is a very different thing
link |
by completely changing their culture.
link |
And biology finds other markets in a more adaptive way.
link |
In a more adaptive way.
link |
Yeah, it's better at it.
link |
It's better at that kind of integration.
link |
So maybe you've already said it,
link |
but what to use the most beautiful aspect
link |
or mechanism of the human mind?
link |
Is it the adaptive,
link |
the ability to adapt as you've described?
link |
Or is there some other little quirk
link |
that you particularly like?
link |
Adaptation is everything when you get down to it.
link |
But the differences between adaptation
link |
where you're learning goes on only over a generation
link |
that over evolutionary time,
link |
where your learning goes on at the time scale
link |
of one individual who must learn from the environment
link |
during that individual's lifetime.
link |
And biology has both kinds of learning in it.
link |
And the thing which makes neurobiology hard
link |
is that it's a mathematical system as it were built
link |
on this other kind of evolutionary system.
link |
What do you mean by mathematical system?
link |
Where's the math and the biology?
link |
Well, when you talk to a computer scientist
link |
about neural networks, it's all math.
link |
The fact that biology actually came about
link |
from evolution and the fact that
link |
biology is about a system which you can build
link |
in three dimensions.
link |
If you look at computer chips,
link |
computer chips are basically two dimensional structures.
link |
Maybe 2.1 dimensions, but they really have difficulty
link |
doing three dimensional wiring.
link |
Biology is neocortex is actually also sheet like
link |
and it sits on top of the white matter,
link |
which is about 10 times the volume of the gray matter
link |
and contains all what you might call the wires.
link |
But there's a huge, the effect of computer structure
link |
on what is easy and what is hard is immense.
link |
And biology does, it makes some things easy
link |
that are very difficult to understand
link |
how to do computationally.
link |
On the other hand, it can't do simple
link |
floating point arithmetic, so it's awfully stupid.
link |
Yeah, and you're saying this kind of three dimensional
link |
complicated structure makes, it's still math.
link |
It's still doing math.
link |
The kind of math is doing enables you to solve problems
link |
of a very different kind.
link |
That's right, that's right.
link |
So you mentioned two kinds of adaptation.
link |
The evolutionary adaptation and the adaptation
link |
are learning at the scale of a single human life.
link |
Which is particularly beautiful to you
link |
and interesting from a research
link |
and from just a human perspective
link |
and which is more powerful.
link |
I find things most interesting that I begin to see
link |
how to get into the edges of them
link |
and tease them apart a little bit to see how they work.
link |
And since I can't see the evolutionary process going on,
link |
I am in awe of it, but I find it just a black hole
link |
as far as trying to understand what to do.
link |
And so in a certain sense, I'm in awe of it,
link |
but I couldn't be interested in working on it.
link |
The human life's time scale is however a thing
link |
you can tease apart and study.
link |
Yeah, you can do it.
link |
There's developmental neurobiology
link |
which understands how the connections
link |
and how the structure evolves
link |
from a combination of what the genetics is like
link |
and the real, the fact that you're building a system
link |
in three dimensions.
link |
In just days and months, those early days
link |
of a human life are really interesting.
link |
They are, and of course, there are times
link |
of immense cell multiplication.
link |
There are also times of the greatest cell death
link |
in the brain is during infancy.
link |
So what is not effective, what is not wired well enough
link |
to use at the moment, throw it out.
link |
It's a mysterious process.
link |
From, let me ask, from what field do you think
link |
the biggest breakthroughs in understanding
link |
the mind will come in the next decades?
link |
Is it neuroscience, computer science, neurobiology,
link |
psychology, physics, maybe math, maybe literature?
link |
Well, of course, I see the world
link |
always through a lens of physics.
link |
I grew up in physics and the way I pick problems
link |
is very characteristic of physics
link |
and of an intellectual background,
link |
which is not psychology, which is not chemistry
link |
and so on and so on.
link |
At both of your parents are physicists.
link |
Both of my parents were physicists
link |
and the real thing I got out of that
link |
was a feeling that the world is an understandable place.
link |
And if you do enough experiments
link |
and think about what they mean
link |
and structure things that you can do,
link |
the mathematics of the relevant to the experiments,
link |
you ought to be able to understand how things work.
link |
But that was, that was a few years ago.
link |
Did you change your mind at all?
link |
Through many decades of trying to understand the mind,
link |
of studying in different kinds of way,
link |
not even the mind, just biological systems.
link |
You still have hope that physics that you can understand.
link |
There's a question of what do you mean by understand?
link |
When I taught freshman physics, I used to say,
link |
I wanted to get physics to understand the subject,
link |
to understand Newton's laws.
link |
I didn't want them simply to memorize a set of examples
link |
to which they knew the equations to write down,
link |
to generate the answers.
link |
I had this nebulous idea of understanding.
link |
So that if you looked at a situation, you could say,
link |
oh, I expect the ball to make that trajectory.
link |
Or I expect some intuitive notion of understanding.
link |
And I don't know how to express that very well.
link |
And I've never known how to express it well.
link |
And you run smack up against it.
link |
When you do these, look at these simple neural nets,
link |
feed forward neural nets, which do amazing things.
link |
And yet, you know, contain nothing of the essence
link |
of what I would have felt was understanding.
link |
Understanding is more than just an enormous lookup table.
link |
Let's linger on that.
link |
How sure you are of that?
link |
What if the table gets really big?
link |
So, I mean, asks another way,
link |
these feed forward neural networks,
link |
do you think they'll ever understand?
link |
Could answer that in two ways.
link |
I think if you look at real systems,
link |
feedback is an essential aspect
link |
of how these real systems compute.
link |
On the other hand, if I have a mathematical system
link |
with feedback, I know I can unlayer this and do it.
link |
But I have an exponential expansion
link |
in the amount of stuff I have to build
link |
if I can solve the problem that way.
link |
So feedback is essential.
link |
So we can talk even about recurrent neural recurrence.
link |
But do you think all the pieces are there
link |
to achieve understanding through these simple mechanisms?
link |
Like, back to our original question,
link |
what is the fundamental,
link |
is there a fundamental difference
link |
between artificial neural networks and biological?
link |
Or is it just a bunch of surface stuff?
link |
Suppose you ask a neurosurgeon,
link |
when has somebody dead?
link |
They'll probably go back to saying,
link |
well, I can look at the brain rhythms
link |
and tell you if this is a brain
link |
which has never got a function again.
link |
This one is, this other one is one
link |
is if we treat it well, it's still recoverable.
link |
And then just do that by some electrodes
link |
looking at simple electrical patterns.
link |
Just don't look in any detail at all
link |
at what individual neurons are doing.
link |
These rhythms are already absent
link |
from anything that goes on at Google.
link |
Yeah, but the rhythms,
link |
but the rhythms what?
link |
So, well, that's like comparing,
link |
okay, I'll tell you, it's like you're comparing
link |
the greatest classical musician in the world
link |
to a child first learning to play.
link |
The question, but they're still both playing the piano.
link |
I'm asking, is there, will it ever go on at Google?
link |
Do you have a hope?
link |
Because you're one of the seminal figures
link |
in both launching both disciplines,
link |
both sides of the river.
link |
I think it's going to go on generation after generation
link |
the way it has where what you might call
link |
the AI computer science community says,
link |
let's take the following.
link |
This is our model of neurobiology at the moment.
link |
Let's pretend it's good enough
link |
and do everything we can with it.
link |
And it does interesting things.
link |
And after the while, it sort of grinds into the sand
link |
and you say, oh, something else is needed
link |
for neurobiology and some other grand thing comes in
link |
and enables you to go a lot further.
link |
What will go into the sand again?
link |
And I think there couldn't be generations
link |
of this evolution.
link |
I don't know how many of them
link |
and each one is going to get you further
link |
into what a brain does and in some sense
link |
past the Turing test longer and more broad aspects.
link |
And how many of these are there
link |
are going to have to be before you say,
link |
I've made something, I've made a human.
link |
But your senses, it might be a couple.
link |
My senses might be a couple more.
link |
And going back to my brainwaves, as it were.
link |
From the AI point of view,
link |
if they would say, ah, maybe these are an debut phenomenon
link |
and not important at all.
link |
The first car I had, a real wreck of a 1936 Dodge,
link |
go above about 45 miles an hour and the wheels would shimmy.
link |
Good, good speedometer that.
link |
Now, don't be designed at the car that way.
link |
The car is malfunctioning to have that.
link |
But in biology, if it were useful to know,
link |
when are you going more than 45 miles an hour?
link |
You just capture that
link |
and you wouldn't worry about where it came from.
link |
It's going to be a long time before that kind of thing,
link |
which can take place in large complex networks of things,
link |
is actually used in the computation.
link |
Look, the, how many transistors are there
link |
at your laptop these days?
link |
Actually, I don't know the number.
link |
It's on the scale of 10 to the 10.
link |
I can't remember the number either.
link |
And all the transistors are somewhat similar
link |
and most physical systems with that many parts,
link |
all of which are similar, have collective properties.
link |
Sound waves and air, earthquakes,
link |
what have you have collective properties, weather.
link |
There are no collective properties used
link |
in artificial neural networks, in AI.
link |
Yeah, it's very...
link |
If biology uses them,
link |
it's going to take us to more generations of things
link |
to further people to actually dig in
link |
and see how they are used and what they mean.
link |
See, you're very right.
link |
You might have to return several times to neurobiology
link |
and try to make our transistors more messy.
link |
At the same time, the simple ones
link |
will conquer big aspects.
link |
And I think one of the most biggest surprises to me was
link |
how well learning systems are manifestly nonbiological,
link |
how important they can be actually
link |
and how important and how useful they can be in AI.
link |
So, if we can just take a stroll to some of your work
link |
that is incredibly surprising,
link |
that it works as well as it does,
link |
that launched a lot of the recent work with neural networks.
link |
If we go to what are now called Hopfield Networks,
link |
can you tell me what is associative memory in the mind
link |
for the human side?
link |
Let's explore memory for a bit.
link |
Okay, what do you mean by associative memory is,
link |
you have a memory of each of your friends.
link |
Your friend has all kinds of properties
link |
from what they look like,
link |
what their voice sounds like,
link |
to where they went to college,
link |
where you met them, go on and on,
link |
what science papers they've written.
link |
And if I start talking about
link |
a five foot 10 wire rated cognitive scientist
link |
that's got a very bad back,
link |
it doesn't take very long for you to say,
link |
oh, he's talking about Jeff Hinton.
link |
I never mentioned the name or anything very particular,
link |
but somehow a few facts that are associated
link |
with a particular person enables you to get
link |
hold of the rest of the facts, or not the rest of them,
link |
another subset of them.
link |
And it's this ability to link things together,
link |
link experiences together,
link |
which goes on to the general name of associative memory.
link |
And a large part of intelligent behavior
link |
is actually just large associative memories at work,
link |
as far as I can see.
link |
What do you think is the mechanism
link |
of how it works in the mind?
link |
Is it a mystery to you still?
link |
Do you have inklings of how this essential thing
link |
for cognition works?
link |
What I made 35 years ago was, of course,
link |
a crude physics model to show the kind of
link |
to actually enable you to understand
link |
my old sense of understanding as a physicist,
link |
because you could say, ah, I understand
link |
why this goes to stable states.
link |
It's like things going downhill.
link |
And that gives you something with which to think
link |
in physical terms, rather than only in mathematical terms.
link |
So you've created these associative artificial,
link |
you know, that works.
link |
And now if you look at what I did,
link |
I didn't at all describe a system which gracefully learns.
link |
I described a system in which you could understand
link |
how learning could link things together,
link |
how very crudely it might learn.
link |
One of the things which intrigues me is I reinvestigate
link |
that this now, to some extent, is look, I see you,
link |
I'll see you every second for the next hour or what have you.
link |
Each look at you is a little bit different.
link |
I don't store all those second by second images.
link |
I don't store 3,000 images.
link |
I somehow compact this information.
link |
So I now have a view of you which I can use.
link |
It doesn't slavishly remember anything in particular,
link |
but it compacts the information into useful chunks
link |
which are somehow, it's these chunks
link |
which are not just activities of neurons,
link |
bigger things than that,
link |
which are the real entities which are useful to you.
link |
Useful to you to describe, to compress this information.
link |
And you have to compress it in such a way
link |
that if the information comes in just like this again,
link |
I don't bother to rewrite it or efforts to rewrite it.
link |
Simply do not yield anything
link |
because those things are already written.
link |
And that needs to be, not look this up,
link |
as if I started somewhere already.
link |
There has to be something which is much more automatic
link |
in the machine hardware.
link |
Right, so in the human mind,
link |
how complicated is that process, do you think?
link |
So you've created, feels weird to be sitting
link |
with John Hopfield calling him Hopfield Networks, but...
link |
Yeah, but nevertheless, that's what everyone calls him.
link |
So that's a simplification.
link |
That's what a physicist would do.
link |
You and Richard Feynman sat down
link |
and talked about associative memory.
link |
Now, as a, if you look at the mind
link |
where you can't quite simplify it so perfectly, do you...
link |
Well, let me back track just a little bit.
link |
Biology is about dynamical systems.
link |
Computers are dynamical systems.
link |
You can ask, if you're about to math,
link |
the model biology, if you want to model neurobiology,
link |
what is the time scale?
link |
There's a dynamical system in which of a fairly fast time
link |
scale in which you could say,
link |
the synaptes don't change much during this computation.
link |
So think of the synaptes as fixed
link |
and just do the dynamics of the activity.
link |
Or you can say, the synaptes are changing fast enough
link |
that I have to have the synaptic dynamics
link |
working at the same time as the system dynamics
link |
in order to understand the biology.
link |
Most are, if you look at the feedforward artificial neural
link |
nets, they're all done as learnings.
link |
First of all, I spent some time learning,
link |
not performing, and I turned off learning and I performed.
link |
That's not biology.
link |
And so as I look more deeply at neurobiology,
link |
even as an associative memory, I've
link |
got to face the fact that the dynamics of a synapse change
link |
is going on all the time.
link |
And I can't just get by by saying,
link |
I'll do the dynamics of activity with fixed synapses.
link |
So the synaptic, the dynamics of the synapses
link |
is actually fundamental to the whole system.
link |
And there's nothing necessarily separating the time scales.
link |
When the time scales can be separated,
link |
it's neat from the physicists of the mathematicians
link |
But it's not necessarily true in neurobiology.
link |
See, you're kind of dancing beautifully
link |
between showing a lot of respect to physics
link |
and then also saying that physics cannot quite
link |
reach the complexity of biology.
link |
So where do you land?
link |
Or do you continuously dance between the two points?
link |
I continuously dance between them
link |
because my whole notion of understanding
link |
is that you can describe to somebody else
link |
how something works in ways which are honest and believable
link |
and still not describe all the nuts and bolts in detail.
link |
I can describe weather as 10 to the 32 molecules
link |
colliding in the atmosphere.
link |
I can simulate weather that way or have a big enough machine.
link |
I'll simulate it accurately.
link |
It's no good for understanding.
link |
If I want to understand things, I want to understand things
link |
in terms of wind patterns, hurricanes, pressure
link |
differentials, and so on.
link |
All things as they're collective.
link |
And the physicists in me always hopes
link |
that biology will have some things which
link |
can be said about it which are both true and for which you
link |
don't need all the molecular details of the molecules
link |
That's what I mean from the roots of physics
link |
So what did, again, sorry, but Hopfield Networks
link |
help you understand?
link |
What insight did it give us about memory, about learning?
link |
They didn't give insights about learning.
link |
They gave insights about how things having learned
link |
could be expressed.
link |
How having learned a picture of you
link |
reminds me of your name, but it didn't describe
link |
a reasonable way of actually doing the learning.
link |
They only said if you had previously
link |
learned the connections of this kind of pattern
link |
would now be able to behave in a physical way with the day
link |
off with part of the pattern in here,
link |
the other part of the pattern will complete over here.
link |
I could understand that physics if the right learning
link |
stuff had already been put in.
link |
And it could understand why then putting in a picture
link |
of somebody else would generate something else over here.
link |
But it did not have a reasonable description
link |
of the learning process.
link |
But even if we get learning, that's
link |
just a powerful concept that forming representations that
link |
are useful to be robust for error correction kind of thing.
link |
So this is kind of what the biology does
link |
that we're talking about.
link |
What my paper did was simply enable you.
link |
There are lots of ways of being robust.
link |
If you think of a dynamical system,
link |
you think of a system where a path is going on in time.
link |
And if you think of a computer, there's
link |
a computational path which is going on in a huge dimensional
link |
space of 1s and 0s.
link |
And an error correction system is a system
link |
which, if you get a little bit off that trajectory,
link |
will push you back onto that trajectory again.
link |
So you get to the same answer in spite of the fact
link |
that there were things that the computation wasn't being
link |
ideally done all the way along the line.
link |
And there are lots of models for error correction.
link |
But one of the models for error correction
link |
is to say there's a valley that you're following
link |
And if you push a little bit off the valley,
link |
just like water being pushed a little bit by a rock,
link |
it gets back and follows the course of the river.
link |
And that, basically, the analog in the physical system
link |
which enables you to say, oh, yes, error free computation
link |
and an associative memory are very much like things
link |
that I can understand from point of view
link |
of a physical system.
link |
The physical system can be, under some circumstances,
link |
an accurate metaphor.
link |
It's not the only metaphor.
link |
There are error correction schemes
link |
which don't have a valley and energy behind them.
link |
But those are error correction schemes
link |
which a mathematician may be able to understand,
link |
So there's the physical metaphor that seems to work here.
link |
So these kinds of networks actually
link |
led to a lot of the work that is going on now
link |
in neural networks, artificial neural networks.
link |
So the follow on work with the restricted Boltzmann machines
link |
and deep belief nets followed on from these ideas
link |
of the Hopfield network.
link |
So what do you think about this continued progress
link |
of that work towards now revigorated exploration
link |
of feedforward neural networks and recurrent neural networks
link |
and convolutional neural networks
link |
and kinds of networks that are helping solve image recognition,
link |
natural language processing, all that kind of stuff?
link |
It always intrigued me that one of the most long lived
link |
of the learning systems is the Boltzmann machine, which
link |
is intrinsically a feedback network.
link |
And with the brilliance of Hinden and Sinovsky
link |
to understand how to do learning in that,
link |
and it's still a useful way to understand learning
link |
and understand, and the learning that you understand in that
link |
has something to do with the way that feedforward systems
link |
work, but it's not always exactly simple
link |
to express that intuition.
link |
But it always amuses me to see Hinden going back
link |
to the will yet again on a form of the Boltzmann machine,
link |
because really that which has feedback and interesting
link |
probabilities in it is a lovely encapsulation
link |
of something computational.
link |
Something computational?
link |
Something both computational and physical.
link |
Computational, and it's very much related
link |
to feedforward networks.
link |
Physical in that Boltzmann machine learning
link |
is really learning a set of parameters
link |
for a physics, Hamiltonian, or energy function.
link |
What do you think about learning in this whole domain?
link |
Do you think the F4 mentioned guy, Jeff Hinton,
link |
all the work there with back propagation,
link |
all the kind of learning that goes on in these networks,
link |
if we compare it to learning in the brain, for example,
link |
is there echoes of the same kind of power
link |
that back propagation reveals about these kinds
link |
of recurrent networks, or is it something fundamentally
link |
different going on in the brain?
link |
I don't think the brain is as deep
link |
as the deepest networks go, the deepest computer science
link |
And I do wonder whether part of that depth of the computer
link |
science networks is necessitated by the fact
link |
that the only learning that's easily done on a machine
link |
And so there is the question of, to what extent
link |
is the biology, which has some feedforward and some feedback,
link |
been captured by something which has got many more neurons,
link |
but much more depth than the neurons in it?
link |
So part of you wonders if the feedback is actually
link |
more essential than the number of neurons or the depth,
link |
the dynamics of the feedback.
link |
The dynamics of the feedback, look,
link |
if you don't have feedback, it's a little bit like a building,
link |
a big computer, and running it through one clock cycle.
link |
And then you can't do anything.
link |
Do you reload something coming in?
link |
How do you use the fact that there
link |
are multiple clocks like that?
link |
How do I use the fact that you can close your eyes,
link |
stop listening to me, and think about a chess board
link |
for two minutes without any input whatsoever?
link |
Yeah, that memory thing.
link |
That's fundamentally a feedback kind of mechanism.
link |
You're going back to something.
link |
Yes, it's hard to understand.
link |
It's hard to introspect.
link |
Let alone consciousness.
link |
Oh, let alone consciousness, yes, yes.
link |
Because that's tied up in there, too.
link |
You can't just put that on another shelf.
link |
Every once in a while, I get interested in consciousness,
link |
and then I go and I've done that for years
link |
and ask one of my betters, as it were,
link |
their view on consciousness.
link |
It's been interesting collecting them.
link |
What is consciousness?
link |
Let's try to take a brief step into that room.
link |
Well, I asked Marvin Minsky, as you go on consciousness.
link |
And Marvin said, consciousness is basically overrated.
link |
It may be an epiphenomenon.
link |
After all, all the things your brain does,
link |
these are actually hard computations you do unconsciously.
link |
And there's so much evidence that even the simple things you do,
link |
you can make committed decisions about them.
link |
The neurobiologist can say, he's now committed.
link |
He's going to move the hand left before you know it.
link |
So his view that consciousness is not,
link |
that's just like little icing on the cake.
link |
The real cake is in the subconscious.
link |
Subconscious, nonconscious.
link |
Nonconscious, that's the better word, sir.
link |
It's only that Freud captured the other word.
link |
Yeah, it's a confusing word, subconscious.
link |
Nicholas Chater wrote an interesting book.
link |
I think the title of it is The Mind is Flat.
link |
In an neural net sense, it might be
link |
flat as something which is a very broad neural net
link |
without really any layers in depth,
link |
or as a deep brain would be many layers and not so broad.
link |
In the same sense that if you pushed Minsky hard enough,
link |
he would probably have said, consciousness
link |
is your effort to explain to yourself
link |
that which you have already done.
link |
Yeah, it's the weaving of the narrative
link |
around the things that already been computed for you.
link |
And so much of what we do for our memories of events,
link |
for example, if there's some traumatic event, you witness,
link |
you will have a few facts about it correctly done.
link |
If somebody asks you about it, you
link |
will weave a narrative which is actually
link |
much more rich in detail than that.
link |
Based on some anchor points you have of correct things
link |
and pulling together general knowledge on the other,
link |
but you will have a narrative.
link |
And once you generate that narrative,
link |
you are very likely to repeat that narrative
link |
and claim that all the things you have in it
link |
are actually the correct things.
link |
There was a marvelous example of that in the Watergate
link |
slash impeachment era of John Dean.
link |
John Dean, you're too young to know,
link |
had been the personal lawyer of Nixon.
link |
And so John Dean was involved in the cover up.
link |
And John Dean ultimately realized
link |
the only way to keep himself out of jail for a long time
link |
was actually to tell some of the truths about Nixon.
link |
And John Dean was a tremendous witness.
link |
He would remember these conversations
link |
in great detail and very convincing detail.
link |
And long afterward, some of the techniques
link |
of some of the tapes, the secret tapes
link |
from which John Dean was recalling these conversations
link |
And one found out that John Dean had a good,
link |
but not exceptional memory.
link |
What he had was an ability to paint vividly and in some sense
link |
accurately the tone of what was going on.
link |
By the way, that's a beautiful description
link |
Do you, like, where do you stand in your today?
link |
So perhaps it changes day to day,
link |
but where do you stand on the importance of consciousness
link |
in our whole big mess of cognition?
link |
Is it just a little narrative maker?
link |
Or is it actually fundamental to intelligence?
link |
That's a very hard one.
link |
When I asked Francis Crick about consciousness,
link |
he launched forward a long monologue about Mendel
link |
and the peas and how Mendel knew that there was something.
link |
And how biologists understood that there was something
link |
in inheritance, which was just very, very different.
link |
And the fact that inherited traits
link |
didn't just wash out into a gray, but were this or this
link |
But that was absolutely fundamental to biology.
link |
And it took generations of biologists
link |
to understand that there was genetics.
link |
And it took another generation or two
link |
to understand that genetics came from DNA.
link |
But very shortly after Mendel, thinking biologists did realize
link |
that there was a deep problem about inheritance.
link |
And Francis would have liked to have said,
link |
and that's why I'm working on consciousness.
link |
But of course, he didn't have any smoking
link |
gun in the sense of Mendel.
link |
And that's the weakness of his position.
link |
If you read his book, which he wrote with Koch, I think.
link |
Yeah, Christoph Koch, yeah.
link |
I find it unconvincing for the smoking gun reason.
link |
So I go on collecting views without actually having taken
link |
a very strong one myself.
link |
Because I haven't seen the entry point.
link |
Not seeing the smoking gun from the point of view of physics.
link |
I don't see the entry point.
link |
Whereas in neurobiology, once I understood
link |
the idea of a collective evolution of dynamics, which
link |
could be described as a collective phenomenon,
link |
I thought, ah, there's a point where
link |
what I know about physics is so different
link |
from any neurobiologist that I have something
link |
that I might be able to contribute.
link |
And right now, there's no way to grasp
link |
a consciousness from a physics perspective.
link |
From my point of view, that's correct.
link |
And of course, people, physicists, like everybody else,
link |
they think very muddly about things.
link |
You have the closely related question about free will.
link |
Do you believe you have free will?
link |
Physicists will give an offhand answer,
link |
and then backtrack, backtrack, backtrack,
link |
where they realize that the answer they gave
link |
must fundamentally contradict the laws of physics.
link |
Naturally answering questions of free will and consciousness
link |
naturally lead to contradictions
link |
from a physics perspective.
link |
Because it eventually ends up with quantum mechanics,
link |
and then you get into that whole mess
link |
of trying to understand how much, from a physics perspective,
link |
how much is determined, already predetermined,
link |
how much is already deterministic about our universe.
link |
And there's lots of different.
link |
And if you don't push quite that far,
link |
you can say essentially all of neurobiology,
link |
which is relevant, can be captured
link |
by classical equations of motion.
link |
Because in my view, the mysteries of the brain
link |
are not the mysteries of quantum mechanics,
link |
but the mysteries of what can happen
link |
when you have a dynamical system, driven system,
link |
with 10 of the 14 parts.
link |
That complexity is something which
link |
is that the physics of complex systems
link |
is at least as badly understood as the physics of phase
link |
coherence in quantum mechanics.
link |
Can we go there for a second?
link |
You've talked about attractor networks.
link |
And just maybe you could say, what are attractor networks?
link |
And more broadly, what are interesting network dynamics
link |
that emerge in these or other complex systems?
link |
You have to be willing to think in a huge number of dimensions,
link |
because in a huge number of dimensions,
link |
the behavior of a system can be thought
link |
as just the motion of the point over time
link |
in this huge number of dimensions.
link |
And an attractor network is simply
link |
a network where there is a line and other lines
link |
converge on it in time.
link |
That's the essence of an attractor network.
link |
In a highly dimensional space.
link |
And the easiest way to get that is
link |
to do it in a high dimensional space, where
link |
some of the dimensions provide the dissipation, which
link |
I have a physical system.
link |
Projectories can't contract everywhere.
link |
They have to contract in some places and expand in others.
link |
There's a fundamental classical theorem
link |
of statistical mechanics, which goes
link |
under the name of Leoville's theorem, which
link |
says you can't contract everywhere.
link |
If you contract somewhere, you expand somewhere else.
link |
It is an interesting physical systems.
link |
You get driven systems where you have a small subsystem, which
link |
is the interesting part.
link |
And the rest of the contraction and expansion,
link |
the physicists would say, is entropy
link |
flow in this other part of the system.
link |
But basically, attractor networks
link |
are dynamics funneling down so that you can't be any.
link |
So that if you start somewhere in the dynamical system,
link |
you will soon find yourself on a pretty well determined
link |
pathway, which goes somewhere.
link |
If you start somewhere else, you'll
link |
wind up on a different pathway.
link |
But I don't have just all possible things.
link |
You have some defined pathways, which are allowed,
link |
and under which you will converge.
link |
And that's the way you make a stable computer,
link |
and that's the way you make a stable behavior.
link |
So in general, looking at the physics
link |
of the emergent stability in these networks,
link |
what are some interesting characteristics
link |
that what are some interesting insights
link |
from studying the dynamics of such high dimensional systems?
link |
Most dynamical systems, most driven dynamical systems,
link |
by driven they're coupled somehow to an energy source.
link |
And so if their dynamics keeps going
link |
because it's coupling to the energy source,
link |
most of them it's very difficult to understand at all
link |
with the dynamical behavior is going to be.
link |
You have to run it.
link |
You have to run it.
link |
There's a subset of systems, which
link |
is what is actually known to the mathematicians
link |
as a Lyapunov function.
link |
And those systems, you can understand convergent dynamics
link |
by saying you're going downhill on something or other.
link |
And that's what I found with ever knowing
link |
what the Lyapunov functions were in the simple model
link |
I made in the early 80s, was that energy functions
link |
you could understand how you could get this channeling
link |
on the pathways without having to follow the dynamics
link |
in infinite detail.
link |
You started rolling a ball off of a mountain
link |
that's going to wind up at the bottom of a valley.
link |
You know that it's true without actually watching
link |
the ball roll down.
link |
There's certain properties of the system
link |
that when you can know that.
link |
And not all systems behave that way.
link |
Most don't, probably.
link |
But it provides you with a metaphor
link |
for thinking about systems which are stable
link |
and who do have these attractors behave,
link |
even if you can't find the Lyapunov function behind them
link |
or an energy function behind them.
link |
It gives you a metaphor for thought.
link |
Speaking of thought, if I had a glint in my eye
link |
with excitement and said, you know,
link |
I'm really excited about this something called
link |
deep learning and neural networks
link |
and I would like to create an intelligent system
link |
and came to you as an advisor, what would you recommend?
link |
And is it a hopeless pursuit to use neural networks
link |
to achieve thought?
link |
Is it, what kind of mechanisms should we explore?
link |
What kind of ideas should we explore?
link |
Well, you look at the simple networks,
link |
one past networks.
link |
They don't support multiple hypotheses very well.
link |
As I have tried to work with very simple systems
link |
which do something which you might consider to be thinking,
link |
thought has to do with the ability to do mental exploration
link |
before you take a physical action.
link |
Almost like we were mentioning playing chess,
link |
visualizing, simulating inside your head different outcomes.
link |
And now you could do that in a feed forward network
link |
because you've pre calculated all kinds of things.
link |
But I think the way neurobiology does it
link |
hasn't pre calculated everything.
link |
It actually has parts of a dynamical system
link |
in which you're doing exploration in a way which is...
link |
There's a creative element.
link |
Like there's an...
link |
There's a creative element.
link |
And in a simple minded neural net,
link |
you have a constellation of instances
link |
from which you've learned.
link |
And if you are within that space,
link |
if a new question is a question within this space,
link |
you can actually rely on that system pretty well
link |
to come up with a good suggestion for what to do.
link |
If on the other hand the query comes from outside the space,
link |
you have no way of knowing how the system is going to behave.
link |
There are no limitations on what could happen.
link |
And so the artificial neural net world
link |
is always very much...
link |
I have a population of examples.
link |
The test set must be drawn from this equivalent population.
link |
The test set has examples which are from a population
link |
which is completely different.
link |
There's no way that you could expect to get the answer right.
link |
What they call outside the distribution.
link |
And so if you see a ball rolling across the street in dusk,
link |
if that wasn't in your training set,
link |
the idea that a child may be coming close behind that
link |
is not going to occur to the neural net.
link |
And it is to our...
link |
There's something in your biology that allows that.
link |
There's something in the way of what it means
link |
to be outside of the population of the training set.
link |
The population of the training set
link |
isn't just sort of this set of examples.
link |
There's more to it than that.
link |
And it gets back to my question of,
link |
what is it to understand something?
link |
You know, in a small tangent,
link |
you've talked about the value of thinking
link |
of deductive reasoning in science versus large data collection.
link |
So sort of thinking about the problem.
link |
I suppose it's the physics side of you
link |
of going back to first principles and thinking.
link |
But what do you think is the value of deductive reasoning
link |
in the scientific process?
link |
Well, they're obviously scientific questions
link |
in which the root to the answer to it
link |
comes through the analysis of what hell of a lot of data.
link |
Cosmology, that kind of stuff.
link |
And that's never been the kind of problem
link |
in which I've had any particular insight.
link |
I must say, if you look at...
link |
Cosmology is one of those.
link |
If you look at the actual things that Jim Peebles,
link |
one of this year's Nobel Prize in physics,
link |
one's from the local physics department,
link |
the kinds of things he's done,
link |
he's never crunched large data.
link |
Never, never, never.
link |
He's used the encapsulation
link |
of the work of others in this regard.
link |
But ultimately boiled down to thinking through the problem.
link |
Like, what are the principles
link |
under which a particular phenomenon operates?
link |
And look, physics is always going to look for ways
link |
in which you can describe the system
link |
in a way which rises above the details.
link |
And to the hard dyed and the wool biologist,
link |
biology works because of the details.
link |
And physics, to the physicists,
link |
we want an explanation which is right
link |
in spite of the details.
link |
And there will be questions which we cannot answer as physicists
link |
because the answer cannot be found that way.
link |
There's not sure if you're familiar
link |
with the entire field of brain computer interfaces
link |
that's become more and more intensely researched
link |
and developed recently,
link |
especially with companies like Neuralink with Elon Musk.
link |
Yeah, I know there have always been the interest
link |
both in things like getting the eyes
link |
to be able to control things
link |
or getting the thought patterns
link |
to be able to move what had been a connected limb
link |
which is now connected through a computer.
link |
So in the case of Neuralink,
link |
they're doing a thousand plus connections
link |
where they're able to do two way,
link |
activate and read spikes, neural spikes.
link |
Do you have hope for that kind of computer brain interaction
link |
in the near or maybe even far future
link |
of being able to expand the ability of the mind of cognition
link |
or understand the mind?
link |
It's interesting watching things go
link |
when I first became interested in neurobiology.
link |
Most of the practitioners thought
link |
you would be able to understand neurobiology
link |
by techniques which allowed you to record
link |
only one cell at a time.
link |
People like David Hubel
link |
very strongly reflected that point of view.
link |
And that's been taken over by a generation,
link |
a couple of generations later,
link |
by a set of people who says
link |
not until we can record from 10 to the 4
link |
or 10 to the 5 at a time
link |
will we actually be able to understand
link |
how the brain actually works.
link |
And in a general sense,
link |
I think that's right.
link |
You have to begin to be able to look
link |
for the collective modes,
link |
the collective operation of things.
link |
It doesn't rely on this action
link |
and potential of that cell.
link |
It relies on the collective properties
link |
of this set of cells connected
link |
with this kind of patterns and so on.
link |
And you're not going to succeed
link |
in seeing what those collective activities are
link |
without recording many cells at once.
link |
The question is how many at once?
link |
What's the threshold?
link |
Look, it's being pursued hard
link |
in the motor cortex.
link |
The motor cortex does something
link |
and yet the problem you're trying to address
link |
Neurobiology does it in ways that are different
link |
from the way an engineer would do it.
link |
An engineer would put in six highly accurate
link |
stepping motors controlling a limb
link |
rather than 100,000 muscle fibers,
link |
each of which has to be individually controlled.
link |
And so understanding how to do things
link |
in a way which is much more forgiving
link |
and much more neural,
link |
I think, would benefit the engineering world.
link |
The engineering world touch.
link |
Let's put it in a pressure sensor or two,
link |
rather than an array of a gazillion pressure sensors,
link |
none of which are accurate,
link |
all of which are perpetually recalibrating themselves.
link |
You're saying your hope is your advice
link |
for the engineers of the future
link |
is to embrace the large chaos of a messy,
link |
error prone system like those of the biological systems.
link |
That's probably the way to solve some of these.
link |
I think you'll be able to make better computations
link |
slash robotics that way than by trying
link |
to force things into a robotics where joint motors
link |
are powerful and stepping motors are accurate.
link |
But then the physicists, the physicists in you
link |
will be lost forever in such systems
link |
because there's no simple fundamentals
link |
to explore in systems that are so large and messy.
link |
You say that, and yet there's a lot of physics
link |
in the Navier Stokes equations,
link |
the equations of nonlinear hydrodynamics,
link |
huge amount of physics in them.
link |
All the physics of atoms and molecules has been lost
link |
but has been replaced by this other set of equations
link |
which is just as true as the equations at the bottom.
link |
Those equations are going to be harder to find
link |
in general biology, but the physicist in me
link |
says there are probably some equations of that sort.
link |
They're out there.
link |
They're out there, and if physics is going to
link |
contribute anything, it may contribute to trying
link |
to find out what those equations are
link |
and how to capture them from the biology.
link |
Would you say that's one of the main open problems
link |
of our age is to discover those equations?
link |
If you look at, there's molecules
link |
and there's psychological behavior,
link |
and these two are somehow related.
link |
There are layers of detail,
link |
there are layers of collectiveness,
link |
and to capture that in some vague way,
link |
several stages on the way up to see how these things
link |
can actually be linked together.
link |
It seems in our universe there's a lot of
link |
elegant equations that can describe the fundamental way
link |
that things behave, which is a surprise.
link |
It's compressible into equations.
link |
It's simple and beautiful.
link |
It's still an open question whether that link
link |
is equally between molecules and the brain
link |
is equally compressible into elegant equations.
link |
But you're both a physicist and a dreamer.
link |
You have a sense that...
link |
Yeah, but I can only dream physics dreams.
link |
There was an interesting book called Einstein's Dreams,
link |
which alternates between chapters on his life
link |
and descriptions of the way time might have been,
link |
The linking between these being of course
link |
ideas that Einstein might have had
link |
to think about the essence of time
link |
as he was thinking about time.
link |
So speaking of the essence of time
link |
and your biology, you're one human,
link |
famous impactful human,
link |
but just one human with a brain
link |
living the human condition.
link |
But you're ultimately mortal, just like all of us.
link |
Has studying the mind as a mechanism
link |
changed the way you think about your own mortality?
link |
It has really, because particularly as you get older
link |
and the body comes apart in various ways,
link |
I became much more aware of the fact
link |
that what is somebody is contained in the brain
link |
and not in the body that you worry about burying.
link |
And it is to a certain extent true
link |
that for people who write things down, equations,
link |
dreams, note beds, diaries,
link |
fractions of their thought does continue to live
link |
after their dead and gone,
link |
after their body is dead and gone.
link |
And there's a sea change in that going on
link |
in my lifetime between
link |
when my father died, except for the things
link |
that were actually written by him,
link |
as there were very few facts about him
link |
that have been recorded,
link |
and a number of facts that are recorded
link |
about each and every one of us forever now,
link |
as far as I can see, in the digital world.
link |
And so the whole question of what is death,
link |
may be different for people a generation ago
link |
and a generation further ahead.
link |
We have become immortal under some definitions.
link |
Last easy question.
link |
What is the meaning of life?
link |
Looking back, you've studied the mind,
link |
us weird descendants of apes.
link |
What's the meaning of our existence on this little earth?
link |
Oh, that word meaning is as slippery as the word understand.
link |
Interconnected somehow, perhaps.
link |
Is there, it's slippery, but is there something
link |
that you, despite being slippery,
link |
can hold long enough to express?
link |
Well, I've been amazed at how hard it is
link |
to define the things in a living system
link |
in the sense that one hydrogen atom is pretty much like another.
link |
But one bacterium is not so much like another bacterium,
link |
even of the same nominal species.
link |
In fact, the whole notion of what is the species
link |
gets a little bit fuzzy.
link |
And do species exist in the absence of certain classes of environments?
link |
And pretty soon one winds up with the biology
link |
which the whole thing is living.
link |
But whether there's actually any element of it
link |
which by itself would be said to be living
link |
becomes a little bit vague in my mind.
link |
So in a sense, the idea of meaning
link |
is something that's possessed by an individual,
link |
like a conscious creature.
link |
And you're saying that it's all interconnected
link |
in some kind of way that there might not even be an individual,
link |
or all kind of this complicated mess of biological systems
link |
at all different levels,
link |
where the human starts and when the human ends, it's unclear.
link |
Yeah, and we're in neurobiology.
link |
We're the, oh, you say the neocortex does the thinking,
link |
but there's lots of things that are done in the spinal cord.
link |
And so we say, what is the essence of thought?
link |
Is it just going to be neocortex?
link |
Can't be. Can't be.
link |
Yeah, maybe to understand and to build thought,
link |
you have to build the universe along with the neocortex.
link |
It's all interlinked through the spinal cord.
link |
John, it's a huge honor talking today.
link |
Thank you so much for your time. I really appreciate it.
link |
Well, thank you for the challenge of talking with you,
link |
and it would be interesting to see whether you can win
link |
five minutes out of this and just go here in a sense to anyone.
link |
Thanks for listening to this conversation with John Hopfield,
link |
and thank you to our presenting sponsor, Cash App.
link |
Download it, use code LEX Podcast.
link |
You'll get $10, and $10 will go to first.
link |
An organization that inspires and educates young minds
link |
to become science and technology innovators of tomorrow.
link |
If you enjoy this podcast, subscribe on YouTube,
link |
give it five stars on Apple Podcasts, support on Patreon,
link |
or simply connect with me on Twitter at Lex Freedman.
link |
And now let me leave you with some words of wisdom
link |
from John Hopfield in his article titled, Now What.
link |
Choosing problems is the primary determinant
link |
of what one accomplishes in science.
link |
I have generally had a relatively short attention span
link |
in science problems. Thus, I have always been on the lookout
link |
for more interesting questions, either as my present ones
link |
get worked out, or as they get classified by me
link |
as intractable, given my particular talents.
link |
He then goes on to say,
link |
What I have done in science relies entirely
link |
on experimental and theoretical studies by experts.
link |
I have a great respect for them, especially for those
link |
who are willing to attempt communication with someone
link |
who is not an expert in the field.
link |
I would only add that experts are good at answering questions.
link |
If you're brash enough, ask your own.
link |
Don't worry too much about how you found them.
link |
Thank you for listening and hope to see you next time.