back to indexJeff Hawkins: Thousand Brains Theory of Intelligence | Lex Fridman Podcast #25
link |
The following is a conversation with Jeff Hawkins.
link |
He's the founder of the Redwood Center
link |
for Theoretical and Neuroscience in 2002
link |
and New Menta in 2005.
link |
In his 2004 book titled On Intelligence
link |
and in the research before and after,
link |
he and his team have worked to reverse engineer
link |
the New York Cortex and propose artificial intelligence
link |
architectures approaches and ideas
link |
that are inspired by the human brain.
link |
These ideas include hierarchical temporal memory,
link |
HTM from 2004, and New Work, The Thousand's Brain's Theory
link |
of Intelligence from 2017, 18, and 19.
link |
Jeff's ideas have been an inspiration
link |
to many who have looked for progress beyond the current
link |
machine learning approaches, but they have also received
link |
criticism for lacking a body of empirical evidence
link |
supporting the models.
link |
This is always a challenge when seeking more than small
link |
incremental steps forward in AI.
link |
Jeff is a brilliant mind and many of the ideas
link |
he has developed and aggregated from neuroscience
link |
are worth understanding and thinking about.
link |
There are limits to deep learning
link |
as it is currently defined.
link |
Forward progress in AI is shrouded in mystery.
link |
My hope is that conversations like this
link |
can help provide an inspiring spark for new ideas.
link |
This is the Artificial Intelligence Podcast.
link |
If you enjoy it, subscribe on YouTube,
link |
iTunes, or simply connect with me on Twitter
link |
at Lex Freedman spelled F R I D.
link |
And now here's my conversation with Jeff Hawkins.
link |
Are you more interested in understanding the human brain
link |
or in creating artificial systems
link |
that have many of the same qualities,
link |
but don't necessarily require that you actually understand
link |
the underpinning workings of our mind?
link |
So there's a clear answer to that question.
link |
My primary interest is understanding the human brain.
link |
No question about it, but I also firmly believe
link |
that we will not be able to create
link |
fully intelligent machines
link |
until we understand how the human brain works.
link |
So I don't see those as separate problems.
link |
I think there's limits to what can be done
link |
with machine intelligence
link |
if you don't understand the principles
link |
by which the brain works.
link |
And so I actually believe that studying the brain
link |
is actually the fastest way to get to machine intelligence.
link |
And within that, let me ask the impossible question.
link |
How do you not define, but at least think
link |
about what it means to be intelligent?
link |
So I didn't try to answer that question first.
link |
We said, let's just talk about how the brain works.
link |
And let's figure out how certain parts of the brain,
link |
mostly the neocortex, but some other parts too,
link |
the parts of the brain most associated with intelligence.
link |
And let's discover the principles by how they work.
link |
Because intelligence isn't just like some mechanism
link |
and it's not just some capabilities.
link |
It's like, okay, we don't even know where to begin
link |
And so now that we've made a lot of progress on this,
link |
after we've made a lot of progress
link |
on how the neocortex works, and we can talk about that,
link |
I now have a very good idea
link |
what's gonna be required to make intelligent machines.
link |
I can tell you today, some of the things
link |
are gonna be necessary, I believe,
link |
to create intelligent machines.
link |
Well, so we'll get there.
link |
We'll get to the neocortex and some of the theories
link |
of how the whole thing works.
link |
And you're saying, as we understand more and more
link |
about the neocortex, about our own human mind,
link |
we'll be able to start to more specifically define
link |
what it means to be intelligent.
link |
It's not useful to really talk about that until...
link |
I don't know if it's not useful.
link |
Look, there's a long history of AI, as you know.
link |
And there's been different approaches taken to it.
link |
Maybe they're all useful.
link |
So the good old fashioned AI, the expert systems,
link |
the current convolutional neural networks,
link |
they all have their utility.
link |
They all have a value in the world.
link |
But I would think almost everyone agreed
link |
that none of them are really intelligent
link |
in a sort of a deep way that humans are.
link |
And so it's just the question of how do you get
link |
from where those systems were or are today
link |
to where a lot of people think we're gonna go.
link |
And there's a big, big gap there, a huge gap.
link |
And I think the quickest way of bridging that gap
link |
is to figure out how the brain does that.
link |
And then we can sit back and look and say,
link |
oh, what are these principles that the brain works on
link |
are necessary and which ones are not?
link |
Clearly, we don't have to build this in,
link |
and tellage machines aren't gonna be built
link |
out of organic living cells.
link |
But there's a lot of stuff that goes on the brain
link |
that's gonna be necessary.
link |
So let me ask maybe, before we get into the fun details,
link |
let me ask maybe a depressing or a difficult question.
link |
Do you think it's possible that we will never
link |
be able to understand how our brain works,
link |
that maybe there's aspects to the human mind
link |
like we ourselves cannot introspectively get to the core,
link |
that there's a wall you eventually hit?
link |
Yeah, I don't believe that's the case.
link |
I have never believed that's the case.
link |
There's not been a single thing we've ever,
link |
humans have ever put their minds to that we've said,
link |
oh, we reached the wall, we can't go any further.
link |
People keep saying that.
link |
People used to believe that about life, you know,
link |
Alain Vitao, right?
link |
There's like, what's the difference in living matter
link |
and nonliving matter?
link |
Something special you never understand.
link |
We no longer think that.
link |
So there's no historical evidence that suggests this is the case
link |
and I just never even consider that's a possibility.
link |
I would also say today, we understand so much
link |
about the neocortex.
link |
We've made tremendous progress in the last few years
link |
that I no longer think of as an open question.
link |
The answers are very clear to me and the pieces
link |
that we don't know are clear to me,
link |
but the framework is all there and it's like, oh, okay,
link |
we're gonna be able to do this.
link |
This is not a problem anymore.
link |
It just takes time and effort, but there's no mystery,
link |
a big mystery anymore.
link |
So then let's get into it for people like myself
link |
who are not very well versed in the human brain,
link |
Can you describe to me at the highest level,
link |
what are the different parts of the human brain
link |
and then zooming in on the neocortex,
link |
the parts of the neocortex and so on, a quick overview.
link |
The human brain, we can divide it roughly into two parts.
link |
There's the old parts, lots of pieces,
link |
and then there's the new part.
link |
The new part is the neocortex.
link |
It's new because it didn't exist before mammals.
link |
The only mammals have a neocortex and in humans
link |
and primates is very large.
link |
In the human brain, the neocortex occupies about 70 to 75%
link |
of the volume of the brain.
link |
And the old parts of the brain are,
link |
there's lots of pieces there.
link |
There's a spinal cord and there's the brainstem
link |
and the cerebellum and the different parts
link |
of the basal ganglion and so on.
link |
In the old parts of the brain,
link |
you have the autonomic regulation,
link |
like breathing and heart rate.
link |
You have basic behaviors.
link |
So like walking and running are controlled
link |
by the old parts of the brain.
link |
All the emotional centers of the brain
link |
are in the old part of the brain.
link |
So when you feel anger or hungry,
link |
lust or things like that,
link |
those are all in the old parts of the brain.
link |
And we associate with the neocortex
link |
all the things we think about
link |
as sort of high level perception.
link |
And cognitive functions, anything from seeing and hearing
link |
and touching things to language, to mathematics
link |
and engineering and science and so on.
link |
Those are all associated with the neocortex.
link |
And they're certainly correlated.
link |
Our abilities in those regards are correlated
link |
with the relative size of our neocortex
link |
compared to other mammals.
link |
So that's like the rough division.
link |
And you obviously can't understand
link |
the neocortex completely isolated,
link |
but you can understand a lot of it
link |
with just a few interfaces
link |
to the old parts of the brain.
link |
And so it gives you a system to study.
link |
The other remarkable thing about the neocortex
link |
compared to the old parts of the brain
link |
is the neocortex is extremely uniform.
link |
It's not visually or anatomically,
link |
or it's very, it's like a,
link |
I always like to say it's like the size
link |
of a dinner napkin, about two and a half millimeters thick.
link |
And it looks remarkably the same everywhere.
link |
Everywhere you look in that two and a half millimeters
link |
is this detailed architecture.
link |
And it looks remarkably the same everywhere.
link |
And that's a cross species,
link |
a mouse versus a cat and a dog and a human.
link |
Where if you look at the old parts of the brain,
link |
there's lots of little pieces do specific things.
link |
So it's like the old parts of a brain evolved,
link |
like this is the part that controls heart rate
link |
and this is the part that controls this
link |
and this is this kind of thing.
link |
And that's this kind of thing.
link |
And these evolve for eons of a long, long time
link |
and they have those specific functions.
link |
And all of a sudden mammals come along
link |
and they got this thing called the neocortex
link |
and it got large by just replicating the same thing
link |
over and over and over again.
link |
This is like, wow, this is incredible.
link |
So all the evidence we have,
link |
and this is an idea that was first articulated
link |
in a very cogent and beautiful argument
link |
by a guy named Vernon Malkassel in 1978, I think it was,
link |
that the neocortex all works on the same principle.
link |
So language, hearing, touch, vision, engineering,
link |
all these things are basically underlying
link |
or all built in the same computational substrate.
link |
They're really all the same problem.
link |
So the low level of the building blocks all look similar.
link |
Yeah, and they're not even that low level.
link |
We're not talking about like neurons.
link |
We're talking about this very complex circuit
link |
that exists throughout the neocortex is remarkably similar.
link |
It is, it's like, yes, you see variations of it here
link |
and they're more of the cell, that's not old and so on.
link |
But what Malkassel argued was,
link |
it says, you know, if you take a section on neocortex,
link |
why is one a visual area and one is a auditory area?
link |
Or why is, and his answer was,
link |
it's because one is connected to eyes
link |
and one is connected to ears.
link |
Literally, you mean just as most closest
link |
in terms of the number of connections to the sensor?
link |
Literally, if you took the optic nerve
link |
and attached it to a different part of the neocortex,
link |
that part would become a visual region.
link |
This actually, this experiment was actually done
link |
by Murgankasur in developing, I think it was lemurs,
link |
I can't remember what it was, it's some animal.
link |
And there's a lot of evidence to this.
link |
You know, if you take a blind person,
link |
a person is born blind at birth,
link |
they're born with a visual neocortex.
link |
It doesn't, may not get any input from the eyes
link |
because of some congenital defect or something.
link |
And that region becomes, does something else.
link |
It picks up another task.
link |
So, and it's, so it's this very complex thing.
link |
It's not like, oh, they're all built on neurons.
link |
No, they're all built in this very complex circuit.
link |
And somehow that circuit underlies everything.
link |
And so this is, it's called the common cortical algorithm,
link |
if you will, some scientists just find it hard to believe.
link |
And they just say, I can't believe that's true.
link |
But the evidence is overwhelming in this case.
link |
And so a large part of what it means
link |
to figure out how the brain creates intelligence
link |
and what is intelligence in the brain
link |
is to understand what that circuit does.
link |
If you can figure out what that circuit does,
link |
as amazing as it is, then you can,
link |
then you understand what all these other cognitive functions are.
link |
So if you were to sort of put neocortex
link |
outside of your book on intelligence,
link |
you look, if you wrote a giant tome,
link |
a textbook on the neocortex,
link |
and you look maybe a couple of centuries from now,
link |
how much of what we know now would still be accurate
link |
two centuries from now.
link |
So how close are we in terms of understanding?
link |
I have to speak from my own particular experience here.
link |
So I run a small research lab here.
link |
It's like any other research lab.
link |
I'm sort of the principal investigator.
link |
There's actually two of us,
link |
and there's a bunch of other people.
link |
And this is what we do.
link |
We started the neocortex,
link |
and we publish our results and so on.
link |
So about three years ago,
link |
we had a real breakthrough in this field.
link |
Just tremendous breakthrough.
link |
We started, we now publish, I think, three papers on it.
link |
And so I have a pretty good understanding
link |
of all the pieces and what we're missing.
link |
I would say that almost all the empirical data
link |
we've collected about the brain, which is enormous.
link |
If you don't know the neuroscience literature,
link |
it's just incredibly big.
link |
And it's, for the most part, all correct.
link |
It's facts and experimental results
link |
and measurements and all kinds of stuff.
link |
But none of that has been really assimilated
link |
into a theoretical framework.
link |
It's data without, in the language of Thomas Kuhn,
link |
the historian, it would be sort of a preparadigm science.
link |
Lots of data, but no way to fit it in together.
link |
I think almost all of that's correct.
link |
There's gonna be some mistakes in there.
link |
And for the most part,
link |
there aren't really good cogent theories
link |
about how to put it together.
link |
It's not like we have two or three competing good theories,
link |
which ones are right and which ones are wrong.
link |
It's like, yeah, people just scratching their heads
link |
throwing things, you know, some people giving up
link |
on trying to figure out what the whole thing does.
link |
In fact, there's very, very few labs that we do
link |
that focus really on theory
link |
and all this unassimilated data and trying to explain it.
link |
So it's not like we've got it wrong.
link |
It's just that we haven't got it at all.
link |
So it's really, I would say, pretty early days
link |
in terms of understanding the fundamental theories,
link |
forces of the way our mind works.
link |
I would have said that's true five years ago.
link |
So as I said, we had some really big breakthroughs
link |
on this recently and we started publishing papers on this.
link |
So you can get to that.
link |
But so I don't think it's, you know, I'm an optimist
link |
and from where I sit today,
link |
most people would disagree with this,
link |
but from where I sit today, from what I know,
link |
it's not super early days anymore.
link |
We are, you know, the way these things go
link |
is it's not a linear path, right?
link |
You don't just start accumulating
link |
and get better and better and better.
link |
No, you got all the stuff you've collected.
link |
None of it makes sense.
link |
All these different things are just sort of around.
link |
And then you're going to have some breaking points
link |
all of a sudden, oh my God, now we got it right.
link |
That's how it goes in science.
link |
And I personally feel like we passed that little thing
link |
about a couple of years ago.
link |
All that big thing a couple of years ago.
link |
So we can talk about that.
link |
Time will tell if I'm right,
link |
but I feel very confident about it.
link |
That's when we'll just say it on tape like this.
link |
At least very optimistic.
link |
So let's, before those few years ago,
link |
let's take a step back to HTM,
link |
the hierarchical temporal memory theory,
link |
which you first proposed on intelligence
link |
and went through a few different generations.
link |
Can you describe what it is,
link |
how it would evolve through the three generations
link |
since you first put it on paper?
link |
Yeah, so one of the things that neuroscientists
link |
just sort of missed for many, many years.
link |
And especially people were thinking about theory
link |
was the nature of time in the brain.
link |
Brain's process, information through time,
link |
the information coming into the brain is constantly changing.
link |
The patterns from my speech right now,
link |
if you're listening to it at normal speed,
link |
would be changing on your ears
link |
about every 10 milliseconds or so, you'd have a change.
link |
This constant flow, when you look at the world,
link |
your eyes are moving constantly,
link |
three to five times a second,
link |
and the input's completely, completely.
link |
If I were to touch something like a coffee cup
link |
as I move my fingers, the input changes.
link |
So this idea that the brain works on time
link |
changing patterns is almost completely,
link |
or was almost completely missing
link |
from a lot of the basic theories like fears of vision
link |
It's like, oh no, we're gonna put this image in front of you
link |
and flash it and say, what is it?
link |
A convolutional neural network's worked that way today, right?
link |
Classified this picture.
link |
But that's not what vision is like.
link |
Vision is this sort of crazy time based pattern
link |
that's going all over the place,
link |
and so is touch and so is hearing.
link |
So the first part of a hierarchical temporal memory
link |
was the temporal part.
link |
It's to say, you won't understand the brain,
link |
nor will you understand intelligent machines
link |
unless you're dealing with time based patterns.
link |
The second thing was, the memory component of it was,
link |
is to say that we aren't just processing input,
link |
we learn a model of the world.
link |
And the memory stands for that model.
link |
The point of the brain, part of the neocortex,
link |
it learns a model of the world.
link |
We have to store things that are experiences
link |
in a form that leads to a model of the world.
link |
So we can move around the world,
link |
we can pick things up and do things
link |
and navigate and know how it's going on.
link |
So that's what the memory referred to.
link |
And many people just, they were thinking about like,
link |
certain processes without memory at all.
link |
They're just like processing things.
link |
And then finally, the hierarchical component
link |
was a reflection to that the neocortex,
link |
although it's just a uniform sheet of cells,
link |
different parts of it project to other parts,
link |
which project to other parts.
link |
And there is a sort of rough hierarchy in terms of that.
link |
So the hierarchical temporal memory is just saying,
link |
look, we should be thinking about the brain
link |
as time based, model memory based and hierarchical processing.
link |
And that was a placeholder for a bunch of components
link |
that we would then plug into that.
link |
We still believe all those things I just said,
link |
but we now know so much more that I'm stopping to use
link |
the word hierarchical temporal memory yet
link |
because it's insufficient to capture the stuff we know.
link |
So again, it's not incorrect,
link |
but I now know more and I would rather describe it
link |
Yeah, so you're basically, we could think of HTM
link |
as emphasizing that there's three aspects of intelligence
link |
that are important to think about
link |
whatever the eventual theory converges to.
link |
So in terms of time, how do you think of nature of time
link |
across different time scales?
link |
So you mentioned things changing,
link |
sensory inputs changing every 10, 20 minutes.
link |
What about every few minutes?
link |
Every few months and years?
link |
Well, if you think about a neuroscience problem,
link |
the brain problem, neurons themselves can stay active
link |
for certain periods of time.
link |
They're parts of the brain where they stay active
link |
for minutes, so you could hold a certain perception
link |
or an activity for a certain period of time,
link |
but not most of them don't last that long.
link |
And so if you think about your thoughts
link |
or the activity neurons,
link |
if you're gonna wanna involve something
link |
that happened a long time ago,
link |
even just this morning, for example,
link |
the neurons haven't been active throughout that time.
link |
So you have to store that.
link |
So by I ask you, what did you have for breakfast today?
link |
That is, you've built it into your model of the world now.
link |
You remember that and that memory is in the synapses,
link |
it's basically in the formation of synapses.
link |
And so you're sliding into what used to different time scales.
link |
There's time scales of which we are
link |
like understanding my language and moving about
link |
and seeing things rapidly and over time.
link |
That's the time scales of activities of neurons.
link |
But if you wanna get in longer time scales,
link |
then it's more memory and we have to invoke those memories
link |
to say, oh, yes, well, now I can remember
link |
what I had for breakfast because I stored that someplace.
link |
I may forget it tomorrow, but I'd store it for now.
link |
So does memory also need to have,
link |
so the hierarchical aspect of reality
link |
is not just about concepts,
link |
it's also about time.
link |
Do you think of it that way?
link |
Yeah, time is infused in everything.
link |
It's like, you really can't separate it out.
link |
If I ask you, what is your, how's the brain
link |
learn a model of this coffee cup here?
link |
I have a coffee cup, then I met the coffee cup.
link |
I said, well, time is not an inherent property
link |
of the model I have of this cup,
link |
whether it's a visual model or tactile model.
link |
I can sense it through time,
link |
but the model itself doesn't really have much time.
link |
If I asked you, if I say, well,
link |
what is the model of my cell phone?
link |
My brain has learned a model of the cell phones.
link |
If you have a smartphone like this,
link |
and I said, well, this has time aspects to it.
link |
I have expectations when I turn it on,
link |
what's gonna happen, what water,
link |
how long it's gonna take to do certain things,
link |
if I bring up an app, what sequences,
link |
and so I have instant, it's like melodies in the world,
link |
you know, melody has a sense of time.
link |
So many things in the world move and act,
link |
and there's a sense of time related to them.
link |
Some don't, but most things do actually.
link |
So it's sort of infused throughout the models of the world.
link |
You build a model of the world,
link |
you're learning the structure of the objects in the world,
link |
and you're also learning
link |
how those things change through time.
link |
Okay, so it really is just a fourth dimension
link |
that's infused deeply,
link |
and you have to make sure
link |
that your models of intelligence incorporate it.
link |
So, like you mentioned, the state of neuroscience
link |
is deeply empirical.
link |
A lot of data collection, it's, you know,
link |
that's where it is, you mentioned Thomas Kuhn, right?
link |
And then you're proposing a theory of intelligence,
link |
and which is really the next step,
link |
the really important step to take,
link |
but why is HTM, or what we'll talk about soon,
link |
So is it more in this, is it backed by intuition,
link |
is it backed by evidence, is it backed by a mixture of both?
link |
Is it kind of closer to where string theory is in physics,
link |
where there's mathematical components
link |
which show that, you know what,
link |
it seems that this,
link |
it fits together too well for it not to be true,
link |
which is where string theory is.
link |
Is that where you're kind of thinking?
link |
It's a mixture of all those things,
link |
although definitely where we are right now,
link |
it's definitely much more on the empirical side
link |
than, let's say, string theory.
link |
The way this goes about, we're theorists, right?
link |
So we look at all this data,
link |
and we're trying to come up with some sort of model
link |
that explains it, basically,
link |
and there's, unlike string theory,
link |
there's vast more amounts of empirical data here
link |
than I think that most physicists deal with.
link |
And so our challenge is to sort through that
link |
and figure out what kind of constructs would explain this.
link |
And when we have an idea,
link |
you come up with a theory of some sort,
link |
you have lots of ways of testing it.
link |
First of all, I am, you know,
link |
there are 100 years of assimilated,
link |
und assimilated empirical data from neuroscience.
link |
So we go back and repapers, and we say,
link |
oh, did someone find this already?
link |
We can predict X, Y, and Z,
link |
and maybe no one's even talked about it
link |
since 1972 or something,
link |
but we go back and find that, and we say,
link |
oh, either it can support the theory
link |
or it can invalidate the theory.
link |
And then we say, okay, we have to start over again.
link |
Oh, no, it's support. Let's keep going with that one.
link |
So the way I kind of view it,
link |
when we do our work, we come up,
link |
we look at all this empirical data,
link |
and it's what I call it is a set of constraints.
link |
We're not interested in something that's biologically inspired.
link |
We're trying to figure out how the actual brain works.
link |
So every piece of empirical data is a constraint on a theory.
link |
If you have the correct theory,
link |
it needs to explain every pin, right?
link |
So we have this huge number of constraints on the problem,
link |
which initially makes it very, very difficult.
link |
If you don't have many constraints,
link |
you can make up stuff all the day.
link |
You can say, oh, here's an answer to how you can do this,
link |
you can do that, you can do this.
link |
But if you consider all biology as a set of constraints,
link |
all neuroscience as a set of constraints,
link |
and even if you're working in one little part of the Neocortex,
link |
for example, there are hundreds and hundreds of constraints.
link |
There are a lot of empirical constraints
link |
that it's very, very difficult initially
link |
to come up with a theoretical framework for that.
link |
But when you do, and it solves all those constraints at once,
link |
you have a high confidence
link |
that you got something close to correct.
link |
It's just mathematically almost impossible not to be.
link |
So that's the curse and the advantage of what we have.
link |
The curse is we have to meet all these constraints,
link |
which is really hard.
link |
But when you do meet them,
link |
then you have a great confidence
link |
that you've discovered something.
link |
In addition, then we work with scientific labs.
link |
So we'll say, oh, there's something we can't find,
link |
we can predict something,
link |
but we can't find it anywhere in the literature.
link |
So we will then, we have people we collaborated with,
link |
we'll say, sometimes they'll say, you know what,
link |
I have some collected data, which I didn't publish,
link |
but we can go back and look at it
link |
and see if we can find that,
link |
which is much easier than designing a new experiment.
link |
You know, neuroscience experiments take a long time, years.
link |
So although some people are doing that now too.
link |
So, but between all of these things,
link |
I think it's a reasonable,
link |
it's actually a very, very good approach.
link |
We are blessed with the fact that we can test our theories
link |
out to yin and yang here,
link |
because there's so much on a similar data,
link |
and we can also falsify our theories very easily,
link |
which we do often.
link |
So it's kind of reminiscent to whenever that was with Copernicus,
link |
you know, when you figure out that the sun is at the center,
link |
the solar system as opposed to Earth,
link |
the pieces just fall into place.
link |
Yeah, I think that's the general nature of the Ha moments,
link |
is in Copernicus, it could be,
link |
you could say the same thing about Darwin,
link |
you could say the same thing about, you know,
link |
about the double helix,
link |
that people have been working on a problem for so long,
link |
and have all this data,
link |
and they can't make sense of it, they can't make sense of it.
link |
But when the answer comes to you,
link |
and everything falls into place,
link |
it's like, oh my gosh, that's it.
link |
That's got to be right.
link |
I asked both Jim Watson and Francis Crick about this.
link |
I asked them, you know,
link |
when you were working on trying to discover the structure
link |
of the double helix,
link |
and when you came up with the sort of,
link |
the structure that ended up being correct,
link |
but it was sort of a guess, you know,
link |
it wasn't really verified yet.
link |
I said, did you know that it was right?
link |
And they both said, absolutely.
link |
We absolutely knew it was right.
link |
And it doesn't matter if other people didn't believe it or not,
link |
we knew it was right, they'd get around to thinking it
link |
and agree with it eventually anyway.
link |
And that's the kind of thing you hear a lot with scientists
link |
who really are studying a difficult problem,
link |
and I feel that way too, about our work.
link |
Have you talked to Crick or Watson about the problem
link |
you're trying to solve, the, of finding the DNA of the brain?
link |
In fact, Francis Crick was very interested in this,
link |
in the latter part of his life.
link |
And in fact, I got interested in brains
link |
by reading an essay he wrote in 1979
link |
called Thinking About the Brain.
link |
And that was when I decided
link |
I'm going to leave my profession of computers and engineering
link |
and become a neuroscientist.
link |
Just reading that one essay from Francis Crick.
link |
I got to meet him later in life.
link |
I got to, I spoke at the Salk Institute
link |
and he was in the audience
link |
and then I had a tea with him afterwards.
link |
You know, he was interested in a different problem.
link |
He was focused on consciousness.
link |
The easy problem, right?
link |
Well, I think it's the red herring
link |
and so we weren't really overlapping a lot there.
link |
Jim Watson, who's still alive,
link |
is also interested in this problem
link |
and when he was director of the Coltsman Harbor Laboratories,
link |
he was really sort of behind
link |
moving in the direction of neuroscience there.
link |
And so he had a personal interest in this field
link |
and I have met with him numerous times.
link |
And in fact, the last time,
link |
a little bit over a year ago,
link |
I gave a talk at Coltsman Harbor Labs
link |
about the progress we were making in our work.
link |
And it was a lot of fun because he said,
link |
well, you wouldn't be coming here
link |
unless you had something important to say,
link |
so I'm going to go attend your talk.
link |
So he sat in the very front row.
link |
Next to him was the director of the lab, Bruce Stillman.
link |
So these guys were in the front row of this auditorium, right?
link |
So nobody else in the auditorium wants to sit in the front row
link |
because there's Jim Watson there as the director.
link |
And I gave a talk and then I had dinner with Jim afterwards.
link |
But there's a great picture of my colleague,
link |
Subitai Amantik, where I'm up there
link |
sort of expiring the basics of this new framework we have.
link |
And Jim Watson's on the edge of his chair.
link |
He's literally on the edge of his chair,
link |
like, internally staring up at the screen.
link |
And when he discovered the structure of DNA,
link |
the first public talk he gave was at Coltsman Harbor Labs.
link |
And there's a picture, there's a famous picture
link |
of Jim Watson standing at the whiteboard
link |
with an overhead thing pointing at something,
link |
pointing at the double helix at this pointer.
link |
And it actually looks a lot like the picture of me.
link |
So there was a sort of funny, there's an area talking about the brain
link |
and there's Jim Watson staring up at the tent.
link |
And of course, there was, you know, whatever,
link |
60 years earlier he was standing pointing at the double helix.
link |
It's one of the great discoveries in all of, you know,
link |
whatever, by all the science, all science and DNA.
link |
So it's the funny that there's echoes of that in your presentation.
link |
Do you think in terms of evolutionary timeline and history,
link |
the development of the neocortex was a big leap?
link |
Or is it just a small step?
link |
So, like, if we ran the whole thing over again,
link |
from the birth of life on Earth,
link |
how likely would we develop the mechanism of the neocortex?
link |
Okay, well, those are two separate questions.
link |
One, was it a big leap?
link |
And one was how likely it is, okay?
link |
They're not necessarily related.
link |
And we don't really have enough data to make a judgment about that.
link |
I would say definitely it was a big leap.
link |
And I can tell you why.
link |
I don't think it was just another incremental step.
link |
I'll get that in a moment.
link |
I don't really have any idea how likely it is.
link |
If we look at evolution, we have one data point,
link |
which is Earth, right?
link |
Life formed on Earth billions of years ago,
link |
whether it was introduced here or it created it here
link |
or someone introduced it we don't really know,
link |
but it was here early.
link |
It took a long, long time to get to multicellular life.
link |
And then from multicellular life,
link |
it took a long, long time to get the neocortex.
link |
And we've only had the neocortex for a few hundred thousand years.
link |
So that's like nothing.
link |
Okay, so is it likely?
link |
Well, certainly it isn't something that happened right away on Earth.
link |
And there were multiple steps to get there.
link |
So I would say it's probably not going to something that would happen
link |
instantaneously on other planets that might have life.
link |
It might take several billion years on average.
link |
But you'd have to survive for several billion years to find out.
link |
Yeah, I think it is a qualitative difference
link |
in all other evolutionary steps.
link |
I can try to describe that if you'd like.
link |
Sure, in which way?
link |
Yeah, I can tell you how.
link |
Pretty much, let's start with a little preface.
link |
Maybe the things that humans are able to do do not have obvious
link |
survival advantages precedent.
link |
Is there a really survival advantage to that?
link |
What about mathematics?
link |
Is there a real survival advantage to mathematics?
link |
You can stretch it.
link |
You can try to figure these things out, right?
link |
But most of evolutionary history, everything had immediate survival
link |
I'll tell you a story, which I like.
link |
It may not be true.
link |
But the story goes as follows.
link |
Organisms have been evolving since the beginning of life here on Earth.
link |
Adding this sort of complexity onto that and this sort of complexity onto that.
link |
And the brain itself is evolved this way.
link |
There's an old part, an older part, an older, older part to the brain that kind of just
link |
keeps calming on new things and we keep adding capabilities.
link |
When we got to the neocortex, initially it had a very clear survival advantage
link |
in that it produced better vision and better hearing and better touch and maybe
link |
a new place and so on.
link |
But what I think happens is that evolution took a mechanism, and this is in our
link |
recent theory, but it took a mechanism that evolved a long time ago for
link |
navigating in the world, for knowing where you are.
link |
These are the so called grid cells and place cells of that old part of the brain.
link |
And it took that mechanism for building maps of the world and knowing where you are
link |
on those maps and how to navigate those maps and turns it into a sort of a slim
link |
down idealized version of it.
link |
And that idealized version could now apply to building maps of other things,
link |
maps of coffee cups and maps of phones, maps of mathematics.
link |
Concepts, yes, and not just almost, exactly.
link |
And it just started replicating this stuff.
link |
You just think more and more and more.
link |
So we went from being sort of dedicated purpose neural hardware to solve certain
link |
problems that are important to survival to a general purpose neural hardware
link |
that could be applied to all problems and now it's escaped the orbit of survival.
link |
It's, we are now able to apply it to things which we find enjoyment, you know,
link |
but aren't really clearly survival characteristics.
link |
And that it seems to only have happened in humans to the large extent.
link |
And so that's what's going on where we sort of have, we've sort of escaped the
link |
gravity of evolutionary pressure in some sense in the near cortex.
link |
And it now does things which are not, that are really interesting,
link |
discovering models of the universe, which may not really help us.
link |
It doesn't matter.
link |
How does it help us surviving knowing that there might be multiple verses or that
link |
there might be, you know, the age of the universe or how do, you know,
link |
various stellar things occur?
link |
It doesn't really help us survive at all.
link |
But we enjoy it and that's what happened.
link |
Or at least not in the obvious way, perhaps.
link |
It is required, if you look at the entire universe in an evolutionary way,
link |
it's required for us to do interplanetary travel and therefore survive past our own fun.
link |
But you know, let's not get too quick.
link |
Yeah, but, you know, evolution works at one time frame.
link |
It's survival, if you think of survival of the phenotype,
link |
survival of the individual.
link |
What you're talking about there is spans well beyond that.
link |
So there's no genetic, I'm not transferring any genetic traits to my children.
link |
That are going to help them survive better on Mars.
link |
Totally different mechanism.
link |
So let's get into the new, as you've mentioned, this idea,
link |
I don't know if you have a nice name, thousand.
link |
We call it the thousand brain theory of intelligence.
link |
So can you talk about this idea of spatial view of concepts and so on?
link |
So can I just describe sort of the, there's an underlying core discovery,
link |
which then everything comes from that.
link |
That's a very simple, this is really what happened.
link |
We were deep into problems about understanding how we build models of stuff in the world
link |
and how we make predictions about things.
link |
And I was holding a coffee cup just like this in my hand.
link |
And I had my finger was touching the side, my index finger.
link |
And then I moved it to the top and I was going to feel the rim at the top of the cup.
link |
And I asked myself a very simple question.
link |
I said, well, first of all, let's say I know that my brain predicts what it's going to feel
link |
before it touches it.
link |
You can just think about it and imagine it.
link |
And so we know that the brain's making predictions all the time.
link |
So the question is, what does it take to predict that?
link |
And there's a very interesting answer.
link |
First of all, it says the brain has to know it's touching a coffee cup.
link |
It has to have a model of a coffee cup.
link |
It needs to know where the finger currently is on the cup, relative to the cup.
link |
Because when I make a movement, it needs to know where it's going to be on the cup
link |
after the movement is completed, relative to the cup.
link |
And then it can make a prediction about what it's going to sense.
link |
So this told me that the neocortex, which is making this prediction,
link |
needs to know that it's sensing it's touching a cup.
link |
And it needs to know the location of my finger relative to that cup
link |
in a reference frame of the cup.
link |
It doesn't matter where the cup is relative to my body.
link |
It doesn't matter its orientation.
link |
None of that matters.
link |
It's where my finger is relative to the cup, which tells me then that the neocortex
link |
has a reference frame that's anchored to the cup.
link |
Because otherwise, I wouldn't be able to say the location
link |
and I wouldn't be able to predict my new location.
link |
And then we quickly, very instantly, you can say,
link |
well, every part of my skin could touch this cup
link |
and therefore every part of my skin is making predictions
link |
and every part of my skin must have a reference frame
link |
that it's using to make predictions.
link |
So the big idea is that throughout the neocortex,
link |
there are, everything is being stored and referenced in reference frames.
link |
You can think of them like XYZ reference frames,
link |
but they're not like that.
link |
We know a lot about the neural mechanisms for this.
link |
But the brain thinks in reference frames.
link |
And as an engineer, if you're an engineer, this is not surprising.
link |
You'd say, if I were to build a CAD model of the coffee cup,
link |
well, I would bring it up in some CAD software
link |
and I would assign some reference frame and say,
link |
this features at this location and so on.
link |
But the fact that this, the idea that this is occurring
link |
throughout the neocortex everywhere, it was a novel idea.
link |
And then a zillion things fell into place after that, a zillion.
link |
So now we think about the neocortex as processing information
link |
quite differently than we used to do it.
link |
We used to think about the neocortex as processing sensory data
link |
and extracting features from that sensory data
link |
and then extracting features from the features
link |
very much like a deep learning network does today.
link |
But that's not how the brain works at all.
link |
The brain works by assigning everything,
link |
every input, everything to reference frames,
link |
and there are thousands, hundreds of thousands of them
link |
active at once in your neocortex.
link |
It's a surprising thing to think about,
link |
but once you sort of internalize this,
link |
you understand that it explains almost every,
link |
almost all the mysteries we've had about this structure.
link |
So one of the consequences of that is that
link |
every small part of the neocortex, say a millimeter square,
link |
and there's 150,000 of those.
link |
So it's about 150,000 square millimeters.
link |
If you take every little square millimeter of the cortex,
link |
it's got some input coming into it,
link |
and it's going to have reference frames
link |
where it's assigning that input to.
link |
And each square millimeter can learn complete models of objects.
link |
So what do I mean by that?
link |
If I'm touching the coffee cup,
link |
well, if I just touch it in one place,
link |
I can't learn what this coffee cup is
link |
because I'm just feeling one part.
link |
But if I move it around the cup and touch it in different areas,
link |
I can build up a complete model of the cup
link |
because I'm now filling in that three dimensional map,
link |
which is the coffee cup.
link |
I can say, oh, what am I feeling in all these different locations?
link |
That's the basic idea.
link |
It's more complicated than that.
link |
But so through time, and we talked about time earlier,
link |
through time, even a single column,
link |
which is only looking at, or a single part of the cortex,
link |
which is only looking at a small part of the world,
link |
can build up a complete model of an object.
link |
And so if you think about the part of the brain,
link |
which is getting input from all my fingers,
link |
so they're spread across the top of your head here.
link |
This is the somatosensory cortex.
link |
There's columns associated with all the different areas of my skin.
link |
And what we believe is happening is that
link |
all of them are building models of this cup,
link |
every one of them, or things.
link |
Not every column or every part of the cortex
link |
builds models of everything,
link |
but they're all building models of something.
link |
And so when I touch this cup with my hand,
link |
there are multiple models of the cup being invoked.
link |
If I look at it with my eyes,
link |
there are again many models of the cup being invoked,
link |
because each part of the visual system,
link |
the brain doesn't process an image.
link |
That's a misleading idea.
link |
It's just like your fingers touching the cup,
link |
so different parts of my retina are looking at different parts of the cup.
link |
And thousands and thousands of models of the cup
link |
are being invoked at once.
link |
And they're all voting with each other,
link |
trying to figure out what's going on.
link |
So that's why we call it the thousand brains theory of intelligence,
link |
because there isn't one model of a cup.
link |
There are thousands of models of this cup.
link |
There are thousands of models of your cell phone,
link |
and about cameras and microphones and so on.
link |
It's a distributed modeling system,
link |
which is very different than what people have thought about it.
link |
So that's a really compelling and interesting idea.
link |
I have two first questions.
link |
So one, on the ensemble part of everything coming together,
link |
you have these thousand brains.
link |
How do you know which one has done the best job of forming the cup?
link |
Great question. Let me try to explain.
link |
There's a problem that's known in neuroscience
link |
called the sensor fusion problem.
link |
And so the idea is something like,
link |
oh, the image comes from the eye.
link |
There's a picture on the retina.
link |
And it gets projected to the neocortex.
link |
Oh, by now it's all sped out all over the place,
link |
and it's kind of squirrely and distorted,
link |
and pieces are all over the, you know,
link |
it doesn't look like a picture anymore.
link |
When does it all come back together again?
link |
Or you might say, well, yes, but I also,
link |
I also have sounds or touches associated with the cup.
link |
So I'm seeing the cup and touching the cup.
link |
How do they get combined together again?
link |
So this is called the sensor fusion problem.
link |
As if all these disparate parts have to be brought together
link |
into one model someplace.
link |
That's the wrong idea.
link |
The right idea is that you get all these guys voting.
link |
There's auditory models of the cup,
link |
there's visual models of the cup,
link |
there's tactile models of the cup.
link |
In the vision system, there might be ones
link |
that are more focused on black and white,
link |
ones versioned on color.
link |
It doesn't really matter.
link |
There's just thousands and thousands of models of this cup.
link |
They don't actually come together in one spot.
link |
Just literally think of it this way.
link |
Imagine you have, each columns are like about the size
link |
of a little piece of spaghetti.
link |
Like a two and a half millimeters tall
link |
and about a millimeter in white.
link |
They're not physical like, but you can think of them that way.
link |
And each one's trying to guess what this thing is or touching.
link |
Now they can, they can do a pretty good job
link |
if they're allowed to move over time.
link |
So I can reach my hand into a black box and move my finger
link |
around an object and if I touch enough space,
link |
it's like, okay, I know what it is.
link |
But often we don't do that.
link |
Often I can just reach and grab something with my hand
link |
all at once and I get it.
link |
Or if I had to look through the world through a straw,
link |
so I'm only invoking one little column,
link |
I can only see part of something because I have to move
link |
But if I open my eyes to see the whole thing at once.
link |
So what we think is going on is all these little pieces
link |
of spaghetti, all these little columns in the cortex
link |
are all trying to guess what it is that they're sensing.
link |
They'll do a better guess if they have time
link |
and can move over time.
link |
So if I move my eyes and move my fingers.
link |
But if they don't, they have a, they have a poor guess.
link |
It's a, it's a probabilistic guess of what they might be touching.
link |
Now imagine they can post their probability
link |
at the top of little piece of spaghetti.
link |
Each one of them says, I think,
link |
and it's not really a probability distribution.
link |
It's more like a set of possibilities in the brain.
link |
It doesn't work as a probability distribution.
link |
It works as more like what we call a union.
link |
You could say, and one column says,
link |
I think it could be a coffee cup, a soda can or a water bottle.
link |
And another column says, I think it could be a coffee cup
link |
or a, you know, telephone or camera or whatever.
link |
And all these guys are saying what they think it might be.
link |
And there's these long range connections
link |
in certain layers in the cortex.
link |
So there's some layers in some cell types in each column
link |
send the projections across the brain.
link |
And that's the voting occurs.
link |
And so there's a simple associative memory mechanism.
link |
We've described this in a recent paper and we've modeled this
link |
that says they can all quickly settle on the only
link |
or the one best answer for all of them.
link |
If there is a single best answer, they all vote and say,
link |
yep, it's got to be the coffee cup.
link |
And at that point, they all know it's a coffee cup.
link |
And at that point, everyone acts as if it's a coffee cup.
link |
Yeah, we know it's a coffee.
link |
Even though I've only seen one little piece of this world,
link |
I know it's a coffee cup I'm touching or I'm seeing or whatever.
link |
And so you can think of all these columns are looking
link |
at different parts and different places,
link |
different sensory input, different locations.
link |
They're all different.
link |
But this layer that's doing the voting, it solidifies.
link |
It crystallizes and says, oh, we all know what we're doing.
link |
And so you don't bring these models together in one model,
link |
you just vote and there's a crystallization of the vote.
link |
That's at least a compelling way to think about the way you
link |
form a model of the world.
link |
Now, you talk about a coffee cup.
link |
Do you see this as far as I understand that you were proposing
link |
this as well, that this extends to much more than coffee cups?
link |
Or at least the physical world.
link |
It expands to the world of concepts.
link |
And well, the first, the primary phase of evidence for that
link |
is that the regions of the neocortex that are associated
link |
with language or high level thought or mathematics or things
link |
like that, they look like the regions of the neocortex
link |
that process vision and hearing and touch.
link |
They don't look any different or they look only marginally
link |
And so one would say, well, if Vernon Mountcastle,
link |
who proposed that all the parts of the neocortex
link |
are the same thing, if he's right, then the parts
link |
that are doing language or mathematics or physics
link |
are working on the same principle.
link |
They must be working on the principle of reference frames.
link |
So that's a little odd thought.
link |
But of course, we had no prior idea how these things happen.
link |
So let's go with that.
link |
And in our recent paper, we talked a little bit about that.
link |
I've been working on it more since.
link |
I have better ideas about it now.
link |
I'm sitting here very confident that that's what's happening.
link |
And I can give you some examples to help you think about that.
link |
It's not that we understand it completely,
link |
but I understand it better than I've described it in any paper
link |
But we did put that idea out there.
link |
It's a good place to start.
link |
And the evidence would suggest it's how it's happening.
link |
And then we can start tackling that problem one piece at a time.
link |
What does it mean to do high level thought?
link |
What does it mean to do language?
link |
How would that fit into a reference framework?
link |
I don't know if you could tell me if there's a connection,
link |
but there's an app called Anki that helps you remember different concepts.
link |
And they talk about like a memory palace that helps you remember
link |
completely random concepts by trying to put them in a physical space
link |
in your mind and putting them next to each other.
link |
It's called the method of loci.
link |
For some reason, that seems to work really well.
link |
Now that's a very narrow kind of application of just remembering some facts.
link |
But that's a very, very telling one.
link |
So this seems like you're describing a mechanism why this seems to work.
link |
So basically the way what we think is going on is all things you know,
link |
all concepts, all ideas, words, everything, you know,
link |
are stored in reference frames.
link |
And so if you want to remember something,
link |
you have to basically navigate through a reference frame the same way
link |
a rat navigates to a man.
link |
Even the same way my finger rat navigates to this coffee cup.
link |
You are moving through some space.
link |
And so if you have a random list of things you would ask to remember
link |
by assigning them to a reference frame,
link |
you've already know very well to see your house, right?
link |
And the idea of the method of loci is you can say,
link |
okay, in my lobby, I'm going to put this thing.
link |
And then the bedroom, I put this one.
link |
I go down the hall, I put this thing.
link |
And then you want to recall those facts.
link |
So recall those things.
link |
You just walk mentally.
link |
You walk through your house.
link |
You're mentally moving through a reference frame that you already had.
link |
And that tells you there's two things that are really important about that.
link |
It tells us the brain prefers to store things in reference frames.
link |
And the method of recalling things or thinking, if you will,
link |
is to move mentally through those reference frames.
link |
You could move physically through some reference frames,
link |
like I could physically move through the reference frame of this coffee cup.
link |
I can also mentally move through the reference frame of the coffee cup,
link |
imagining me touching it.
link |
But I can also mentally move my house.
link |
And so now we can ask ourselves, are all concepts stored this way?
link |
There was some recent research using human subjects in fMRI.
link |
And I'm going to apologize for not knowing the name of the scientists who did this.
link |
But what they did is they put humans in this fMRI machine,
link |
which was one of these imaging machines.
link |
And they gave the humans tasks to think about birds.
link |
So they had different types of birds, and birds that looked big and small
link |
and long necks and long legs, things like that.
link |
And what they could tell from the fMRI was a very clever experiment.
link |
You get to tell when humans were thinking about the birds,
link |
that the birds, the knowledge of birds was arranged in a reference frame
link |
similar to the ones that are used when you navigate in a room.
link |
These are called grid cells.
link |
And there are grid cell like patterns of activity in the neocortex when they do this.
link |
So that, it's a very clever experiment.
link |
And what it basically says is that even when you're thinking about something abstract
link |
and you're not really thinking about it as a reference frame,
link |
it tells us the brain is actually using a reference frame.
link |
And it's using the same neural mechanisms.
link |
These grid cells are the basic same neural mechanisms that we propose
link |
that grid cells, which exist in the old part of the brain, the entomonic cortex,
link |
that that mechanism is now similar mechanism, is used throughout the neocortex.
link |
It's the same nature to preserve this interesting way of creating reference frames.
link |
And so now they have empirical evidence that when you think about concepts like birds
link |
that you're using reference frames that are built on grid cells.
link |
So that's similar to the method of loci.
link |
But in this case, the birds are related so that makes,
link |
they create their own reference frame, which is consistent with bird space.
link |
And when you think about something, you go through that.
link |
You can make the same example.
link |
Let's take a math mathematics.
link |
Let's say you want to prove a conjecture.
link |
What is a conjecture?
link |
A conjecture is a statement you believe to be true,
link |
but you haven't proven it.
link |
And so it might be an equation.
link |
I want to show that this is equal to that.
link |
And you have some places you start with.
link |
You say, well, I know this is true and I know this is true.
link |
And I think that maybe to get to the final proof,
link |
I need to go through some intermediate results.
link |
What I believe is happening is literally these equations
link |
or these points are assigned to a reference frame,
link |
a mathematical reference frame.
link |
And when you do mathematical operations,
link |
a simple one might be multiply or divide,
link |
maybe a little plus transform or something else.
link |
That is like a movement in the reference frame of the math.
link |
And so you're literally trying to discover a path
link |
from one location to another location in a space of mathematics.
link |
And if you can get to these intermediate results,
link |
then you know your map is pretty good
link |
and you know you're using the right operations.
link |
Much of what we think about is solving hard problems
link |
is designing the correct reference frame for that problem,
link |
how to organize the information, and what behaviors
link |
I want to use in that space to get me there.
link |
Yeah, so if you dig in on an idea of this reference frame,
link |
whether it's the math, you start a set of axioms
link |
to try to get to proving the conjecture.
link |
Can you try to describe, maybe take a step back,
link |
how you think of the reference frame in that context?
link |
Is it the reference frame that the axioms are happy in?
link |
Is it the reference frame that might contain everything?
link |
Is it a changing thing as you...
link |
You have many, many reference frames.
link |
In fact, the way the thousand brain theories of intelligence
link |
says that every single thing in the world has its own reference frame.
link |
So every word has its own reference frames.
link |
And we can talk about this.
link |
The mathematics work out this is no problem for neurons to do this.
link |
But how many reference frames does the coffee cup have?
link |
Well, let's say you ask how many reference frames
link |
could the column in my finger that's touching the coffee cup have
link |
because there are many, many models of the coffee cup.
link |
So there is no model of the coffee cup.
link |
There are many models of the coffee cup.
link |
And you can say, well, how many different things can my finger learn?
link |
Is this the question you want to ask?
link |
Imagine I say every concept, every idea,
link |
everything you've ever know about that you can say,
link |
I know that thing has a reference frame associated with it.
link |
And what we do when we build composite objects,
link |
we assign reference frames to point another reference frame.
link |
So my coffee cup has multiple components to it.
link |
It's got a cylinder.
link |
It's got a handle.
link |
And those things have their own reference frames.
link |
And they're assigned to a master reference frame,
link |
which is called this cup.
link |
And now I have this mental logo on it.
link |
Well, that's something that exists elsewhere in the world.
link |
It's its own thing.
link |
So it has its own reference frame.
link |
So we now have to say, how can I assign the mental logo reference frame
link |
onto the cylinder or onto the coffee cup?
link |
So we talked about this in the paper that came out in December
link |
of this last year.
link |
The idea of how you can assign reference frames to reference frames,
link |
how neurons could do this.
link |
So my question is, even though you mentioned reference frames a lot,
link |
I almost feel it's really useful to dig into how you think
link |
of what a reference frame is.
link |
It was already helpful for me to understand that you think
link |
of reference frames as something there is a lot of.
link |
OK, so let's just say that we're going to have some neurons
link |
in the brain, not many actually, 10,000, 20,000,
link |
are going to create a whole bunch of reference frames.
link |
What does it mean?
link |
What is a reference frame?
link |
First of all, these reference frames are different than the ones
link |
you might be used to.
link |
We know lots of reference frames.
link |
For example, we know the Cartesian coordinates,
link |
XYZ, that's a type of reference frame.
link |
We know longitude and latitude.
link |
That's a different type of reference frame.
link |
If I look at a printed map, it might have columns,
link |
A through M and rows, 1 through 20,
link |
that's a different type of reference frame.
link |
It's kind of a Cartesian reference frame.
link |
The interesting thing about the reference frames in the brain,
link |
and we know this because these have been established
link |
through neuroscience studying the entorhinal cortex.
link |
So I'm not speculating here.
link |
This is known neuroscience in an old part of the brain.
link |
The way these cells create reference frames,
link |
they have no origin.
link |
So what it's more like, you have a point,
link |
a point in some space,
link |
and you, given a particular movement,
link |
you can then tell what the next point should be.
link |
And you can then tell what the next point would be.
link |
You can use this to calculate how to get from one point to another.
link |
So how do I get from my house to my home,
link |
or how do I get my finger from the side of my cup
link |
to the top of the cup?
link |
How do I get from the axioms to the conjecture?
link |
So it's a different type of reference frame.
link |
And I can, if you want, I can describe in more detail.
link |
I can paint a picture how you might want to think about that.
link |
It's really helpful to think.
link |
It's something you can move through.
link |
But is it helpful to think of it as spatial in some sense,
link |
or is there something?
link |
No, it's definitely spatial.
link |
It's spatial in a mathematical sense.
link |
How many dimensions?
link |
Can it be a crazy number of dimensions?
link |
Well, that's an interesting question.
link |
In the old part of the brain, the entorhinal cortex,
link |
they studied rats.
link |
And initially, it looks like, oh, this is just two dimensional.
link |
It's like the rat is in some box in a maze or whatever,
link |
and they know whether the rat is using these two dimensional
link |
reference frames and know where it is in the maze.
link |
We say, OK, well, what about bats?
link |
That's a mammal, and they fly in three dimensional space.
link |
How do they do that?
link |
They seem to know where they are, right?
link |
So this is a current area of active research,
link |
and it seems like somehow the neurons in the entorhinal cortex
link |
can learn three dimensional space.
link |
We just, two members of our team, along with Ilefet from MIT,
link |
just released a paper this literally last week,
link |
it's on bioarchive, where they show that you can,
link |
the way these things work, and unless you want to,
link |
I won't get into the detail, but grid cells
link |
can represent any n dimensional space.
link |
It's not inherently limited.
link |
You can think of it this way, if you had two dimensional,
link |
the way it works is you had a bunch of two dimensional slices.
link |
That's the way these things work.
link |
There's a whole bunch of two dimensional models,
link |
and you can slice up any n dimensional space
link |
with two dimensional projections.
link |
And you could have one dimensional models.
link |
So there's nothing inherent about the mathematics
link |
about the way the neurons do this,
link |
which constrained the dimensionality of the space,
link |
which I think was important.
link |
So obviously, I have a three dimensional map of this cup.
link |
Maybe it's even more than that, I don't know.
link |
But it's a clearly three dimensional map of the cup.
link |
I don't just have a projection of the cup.
link |
But when I think about birds,
link |
or when I think about mathematics,
link |
perhaps it's more than three dimensions.
link |
So in terms of each individual column
link |
building up more and more information over time,
link |
do you think that mechanism is well understood?
link |
In your mind, you've proposed a lot of architectures there.
link |
Is that a key piece, or is it, is the big piece,
link |
the thousand brain theory of intelligence,
link |
the ensemble of it all?
link |
Well, I think they're both big.
link |
I mean, clearly the concept, as a theorist,
link |
the concept is most exciting, right?
link |
A high level concept.
link |
A high level concept.
link |
This is a totally new way of thinking about
link |
how the near characteristics work.
link |
So that is appealing.
link |
It has all these ramifications.
link |
And with that, as a framework for how the brain works,
link |
you can make all kinds of predictions
link |
and solve all kinds of problems.
link |
Now we're trying to work through many of these details right now.
link |
Okay, how do the neurons actually do this?
link |
Well, it turns out, if you think about grid cells
link |
and place cells in the old parts of the brain,
link |
there's a lot that's known about them,
link |
but there's still some mysteries.
link |
There's a lot of debate about exactly the details,
link |
how these work, and what are the signs.
link |
And we have that same level of detail,
link |
that same level of concern.
link |
What we spend here, most of our time doing,
link |
is trying to make a very good list
link |
of the things we don't understand yet.
link |
That's the key part here.
link |
What are the constraints?
link |
It's not like, oh, this seems to work, we're done.
link |
It's like, okay, it kind of works,
link |
but these are other things we know it has to do,
link |
and it's not doing those yet.
link |
I would say we're well on the way here.
link |
We're not done yet.
link |
There's a lot of trickiness to this system,
link |
but the basic principles about how different layers
link |
in the neocortex are doing much of this, we understand.
link |
But there's some fundamental parts
link |
that we don't understand as well.
link |
So what would you say is one of the harder open problems,
link |
or one of the ones that have been bothering you,
link |
keeping you up at night the most?
link |
Well, right now, this is a detailed thing
link |
that wouldn't apply to most people, okay?
link |
But you want me to answer that question?
link |
We've talked about, as if, oh, to predict
link |
what you're going to sense on this coffee cup,
link |
I need to know where my finger's going to be on the coffee cup.
link |
That is true, but it's insufficient.
link |
Think about my finger touching the edge of the coffee cup.
link |
My finger can touch it at different orientations.
link |
I can touch it at my finger around here.
link |
And that doesn't change.
link |
I can make that prediction, and somehow,
link |
so it's not just the location.
link |
There's an orientation component of this as well.
link |
This is known in the old part of the brain, too.
link |
There's things called head direction cells,
link |
which way the rat is facing.
link |
It's the same kind of basic idea.
link |
So if my finger were a rat, you know, in three dimensions,
link |
I have a three dimensional orientation,
link |
and I have a three dimensional location.
link |
If I was a rat, I would have a,
link |
I think it was a two dimensional location,
link |
or one dimensional orientation, like this,
link |
which way is it facing?
link |
So how the two components work together,
link |
how does it, I combine orientation,
link |
the orientation of my sensor,
link |
as well as the location,
link |
is a tricky problem.
link |
And I think I've made progress on it.
link |
So at a bigger version of that,
link |
the perspective is super interesting,
link |
but super specific.
link |
Yeah, I warned you.
link |
No, no, no, it's really good,
link |
but there's a more general version of that.
link |
Do you think context matters?
link |
The fact that we are in a building in North America,
link |
that we, in the day and age where we have mugs,
link |
I mean, there's all this extra information
link |
that you bring to the table about everything else in the room
link |
that's outside of just the coffee cup.
link |
How does it get connected, do you think?
link |
Yeah, and that is another really interesting question.
link |
I'm going to throw that under the rubric
link |
or the name of attentional problems.
link |
First of all, we have this model.
link |
I have many, many models.
link |
And also the question, does it matter?
link |
Well, it matters for certain things.
link |
Of course it does.
link |
Maybe what we think about as a coffee cup
link |
in another part of the world is viewed as something completely different.
link |
Or maybe our logo, which is very benign in this part of the world,
link |
it means something very different in another part of the world.
link |
So those things do matter.
link |
I think the way to think about it as the following,
link |
one way to think about it,
link |
is we have all these models of the world.
link |
And we model everything.
link |
And as I said earlier, I kind of snuck it in there.
link |
Our models are actually, we build composite structures.
link |
So every object is composed of other objects,
link |
which are composed of other objects,
link |
and they become members of other objects.
link |
So this room is chairs and a table and a room and walls and so on.
link |
Now we can just arrange these things in a certain way.
link |
And you go, oh, that's in the Nementa conference room.
link |
So, and what we do is when we go around the world,
link |
when we experience the world,
link |
by walking to a room, for example,
link |
the first thing I do is like, oh, I'm in this room.
link |
Do I recognize the room?
link |
Then I can say, oh, look, there's a table here.
link |
And by attending to the table,
link |
I'm then assigning this table in a context of the room.
link |
Then I say, oh, on the table, there's a coffee cup.
link |
Oh, and on the table, there's a logo.
link |
And in the logo, there's the word Nementa.
link |
So if you look in the logo, there's the letter E.
link |
And look, it has an unusual surf.
link |
It doesn't actually, but I pretend it does.
link |
So the point is your attention is kind of drilling deep in and out
link |
of these nested structures.
link |
And I can pop back up and I can pop back down.
link |
I can pop back up and I can pop back down.
link |
So when I attend to the coffee cup,
link |
I haven't lost the context of everything else,
link |
but it's sort of, there's this sort of nested structure.
link |
The attention filters the reference frame formation
link |
for that particular period of time.
link |
It basically, a moment to moment, you attend the subcomponents
link |
and then you can attend the subcomponents to subcomponents.
link |
You can move up and down.
link |
You can move up and down.
link |
We do that all the time.
link |
You're not even, now that I'm aware of it,
link |
I'm very conscious of it.
link |
But most people don't even think about this.
link |
You know, you just walk in a room and you don't say,
link |
oh, I looked at the chair and I looked at the board
link |
and looked at that word on the board
link |
and I looked over here.
link |
So what percentage of your day are you deeply aware of this?
link |
In what part can you actually relax and just be Jeff?
link |
Me personally, like my personal day.
link |
Unfortunately, I'm afflicted with too much of the former.
link |
Well, unfortunately or unfortunately.
link |
You don't think it's useful?
link |
I think about this stuff almost all the time.
link |
And one of my primary ways of thinking is
link |
when I'm asleep at night,
link |
I always wake up in the middle of the night
link |
and I stay awake for at least an hour with my eyes shut
link |
in sort of a half sleep state thinking about these things.
link |
I come up with answers to problems very often
link |
in that sort of half sleeping state.
link |
I think about on my bike ride, I think about on walks.
link |
I'm just constantly thinking about this.
link |
I have to almost schedule time to not think about this stuff
link |
because it's very, it's mentally taxing.
link |
Are you, when you're thinking about this stuff,
link |
are you thinking introspectively,
link |
like almost taking a step outside of yourself
link |
and trying to figure out what is your mind doing right now?
link |
I do that all the time, but that's not all I do.
link |
I'm constantly observing myself.
link |
So as soon as I started thinking about grid cells,
link |
for example, and getting into that,
link |
I started saying, oh, well, grid cells can have my place of sense
link |
That's where you know where you are.
link |
And it's interesting, we always have a sense of where we are
link |
unless we're lost.
link |
And so I started at night when I got up to go to the bathroom,
link |
I would start trying to do it completely with my eyes closed
link |
all the time and I would test my sense of grid cells.
link |
I would walk five feet and say, okay, I think I'm here.
link |
Am I really there?
link |
And then I would calculate my error again and see how the errors
link |
So even something as simple as getting up in the middle of the
link |
night to go to the bathroom, I'm testing these theories out.
link |
I mean, the coffee cup is an example of that too.
link |
So I think I find that these sort of everyday introspections
link |
are actually quite helpful.
link |
It doesn't mean you can ignore the science.
link |
I mean, I spend hours every day reading ridiculously complex
link |
That's not nearly as much fun,
link |
but you have to sort of build up those constraints and the knowledge
link |
about the field and who's doing what and what exactly they think
link |
is happening here.
link |
And then you can sit back and say, okay, let's try to have pieces
link |
Let's come up with some, you know, I'm very in this group here
link |
and people, they know they do this.
link |
I do this all the time.
link |
I come in with these introspective ideas and say, well,
link |
there we ever thought about this.
link |
Now watch, well, let's all do this together.
link |
It's not, as long as you don't, if all you did was that,
link |
then you're just making up stuff, right?
link |
But if you're constraining it by the reality of the neuroscience,
link |
then it's really helpful.
link |
So let's talk a little bit about deep learning and the successes
link |
in the applied space of neural networks, ideas of training model
link |
on data and these simple computational units,
link |
artificial neurons that with back propagation have statistical
link |
ways of being able to generalize from the training set on to
link |
data that similar to that training set.
link |
So where do you think are the limitations of those approaches?
link |
What do you think are strengths relative to your major efforts
link |
of constructing a theory of human intelligence?
link |
Well, I'm not an expert in this field.
link |
I'm somewhat knowledgeable.
link |
A little bit in just your intuition.
link |
Well, I have a little bit more than intuition,
link |
but I just want to say like, you know, one of the things that you asked me,
link |
do I spend all my time thinking about neuroscience?
link |
That's to the exclusion of thinking about things like convolutional neural
link |
But I try to stay current.
link |
So look, I think it's great the progress they've made.
link |
And as I mentioned earlier, it's very highly useful for many things.
link |
The models that we have today are actually derived from a lot of
link |
neuroscience principles.
link |
They are distributed processing systems and distributed memory systems,
link |
and that's how the brain works.
link |
And they use things that we might call them neurons,
link |
but they're really not neurons at all.
link |
So we can just, they're not really neurons.
link |
So they're distributed processing systems.
link |
And nature of hierarchy that came also from neuroscience.
link |
And so there's a lot of things, the learning rules, basically,
link |
not backprop, but other, you know, sort of heavy entire learning.
link |
I'll be curious to say they're not neurons at all.
link |
Can you describe in which way?
link |
I mean, some of it is obvious, but I'd be curious if you have specific
link |
ways in which you think are the biggest differences.
link |
Yeah, we had a paper in 2016 called Why Neurons of Thousands of Synapses.
link |
And if you read that paper, you'll know what I'm talking about here.
link |
A real neuron in the brain is a complex thing.
link |
Let's just start with the synapses on it, which is a connection between neurons.
link |
Real neurons can everywhere from five to 30,000 synapses on them.
link |
The ones near the cell body, the ones that are close to the soma, the cell body,
link |
those are like the ones that people model in artificial neurons.
link |
There's a few hundred of those.
link |
Maybe they can affect the cell.
link |
They can make the cell become active.
link |
95% of the synapses can't do that.
link |
They're too far away.
link |
So if you activate one of those synapses, it just doesn't affect the cell body
link |
enough to make any difference.
link |
Any one of them individually.
link |
Any one of them individually, or even if you do a mass of them.
link |
What real neurons do is the following.
link |
If you activate, or you get 10 to 20 of them active at the same time,
link |
meaning they're all receiving an input at the same time,
link |
and those 10 to 20 synapses or 40 synapses are within a very short distance
link |
on the dendrite, like 40 microns, a very small area.
link |
So if you activate a bunch of these right next to each other at some distant place,
link |
what happens is it creates what's called the dendritic spike.
link |
And dendritic spike travels through the dendrites
link |
and can reach the soma or the cell body.
link |
Now, when it gets there, it changes the voltage,
link |
which is sort of like going to make the cell fire,
link |
but never enough to make the cell fire.
link |
It's sort of what we call, it says we depolarize the cell.
link |
You raise the voltage a little bit, but not enough to do anything.
link |
It's like, well, what good is that?
link |
And then it goes back down again.
link |
So we proposed a theory, which I'm very confident in basics are,
link |
is that what's happening there is those 95% of the synapses
link |
are recognizing dozens to hundreds of unique patterns.
link |
They can write, you know, about 10, 20 synapses at a time,
link |
and they're acting like predictions.
link |
So the neuron actually is a predictive engine on its own.
link |
It can fire when it gets enough, what they call proximal input from those ones
link |
near the cell fire, but it can get ready to fire from dozens to hundreds
link |
of patterns that it recognizes from the other guys.
link |
And the advantage of this to the neuron is that when it actually does produce
link |
a spike in action potential, it does so slightly sooner than it would have otherwise.
link |
And so what could just slightly sooner?
link |
Well, the slightly sooner part is it, there's all the neurons in the,
link |
the excited throwing neurons in the brain are surrounded by these inhibitory neurons,
link |
and they're very fast, the inhibitory neurons, these baskets all.
link |
And if I get my spike out a little bit sooner than someone else,
link |
I inhibit all my neighbors around me, right?
link |
And what you end up with is a different representation.
link |
You end up with a representation that matches your prediction.
link |
It's a sparser representation, meaning fewer neurons are active,
link |
but it's much more specific.
link |
And so we showed how networks of these neurons can do very sophisticated temporal prediction, basically.
link |
So this summarizes real neurons in the brain are time based prediction engines,
link |
and there's no concept of this at all in artificial, what we call point neurons.
link |
I don't think you can mail the brain without them.
link |
I don't think you can build intelligence without them because it's where a large part of the time comes from.
link |
These are predictive models and the time is, there's a prior prediction and an action,
link |
and it's inherent through every neuron in the neocortex.
link |
So I would say that point neurons sort of model a piece of that,
link |
and not very well at that either, but, you know, like, for example, synapses are very unreliable,
link |
and you cannot assign any precision to them.
link |
So even one digit of precision is not possible.
link |
So the way real neurons work is they don't add these, they don't change these weights accurately,
link |
like artificial neural networks do.
link |
They basically form new synapses, and so what you're trying to always do is
link |
detect the presence of some 10 to 20 active synapses at the same time as opposed,
link |
and they're almost binary.
link |
It's like, because you can't really represent anything much finer than that.
link |
So these are the kind of, and I think that's actually another essential component
link |
because the brain works on sparse patterns, and all that mechanism is based on sparse patterns,
link |
and I don't actually think you could build real brains or machine intelligence
link |
without incorporating some of those ideas.
link |
It's hard to even think about the complexity that emerges from the fact that
link |
the timing of the firing matters in the brain, the fact that you form new synapses,
link |
and everything you just mentioned in the past couple minutes.
link |
Trust me, if you spend time on it, you can get your mind around it.
link |
It's not like it's no longer a mystery to me.
link |
No, but sorry, as a function in a mathematical way,
link |
can you start getting an intuition about what gets it excited, what not,
link |
and what kind of representation?
link |
Yeah, it's not as easy as there are many other types of neural networks
link |
that are more amenable to pure analysis, especially very simple networks.
link |
You know, oh, I have four neurons, and they're doing this.
link |
Can we describe them mathematically what they're doing type of thing?
link |
Even the complexity of convolutional neural networks today,
link |
it's sort of a mystery. They can't really describe the whole system.
link |
And so it's different.
link |
My colleague, Subitain Ahmad, he did a nice paper on this.
link |
You can get all the stuff on our website if you're interested.
link |
Talking about sort of mathematical properties of sparse representations,
link |
and so what we can do is we can show mathematically, for example,
link |
why 10 to 20 synapses to recognize a pattern is the correct number,
link |
is the right number you'd want to use.
link |
And by the way, that matches biology.
link |
We can show mathematically some of these concepts about the show
link |
why the brain is so robust to noise and error and fallout and so on.
link |
We can show that mathematically as well as empirically in simulations.
link |
But the system can't be analyzed completely.
link |
Any complex system can, and so that's out of the realm.
link |
But there is mathematical benefits and intuitions that can be derived from mathematics.
link |
And we try to do that as well.
link |
Most of our papers have a section about that.
link |
So I think it's refreshing and useful for me to be talking to you about deep neural networks,
link |
because your intuition basically says that we can't achieve anything like intelligence with artificial neural networks.
link |
Well, not in the current form.
link |
Not in the current form.
link |
I'm sure we can do it in the ultimate form, sure.
link |
So let me dig into it and see what your thoughts are there a little bit.
link |
So I'm not sure if you read this little blog post called Bitter Lesson by Rich Sutton recently.
link |
He's a reinforcement learning pioneer.
link |
I'm not sure if you're familiar with him.
link |
His basic idea is that all the stuff we've done in AI in the past 70 years, he's one of the old school guys.
link |
The biggest lesson learned is that all the tricky things we've done don't, you know, they benefit in the short term.
link |
But in the long term, what wins out is a simple general method that just relies on Moore's law on computation getting faster and faster.
link |
This is what he's saying.
link |
This is what has worked up to now.
link |
This is what has worked up to now.
link |
If you're trying to build a system, if we're talking about, he's not concerned about intelligence.
link |
He's concerned about a system that works in terms of making predictions on applied, narrow AI problems.
link |
That's what the discussion is about.
link |
That you just try to go as general as possible and wait years or decades for the computation to make it actually possible.
link |
Is he saying that as a criticism or is he saying this is a prescription of what we ought to be doing?
link |
Well, it's very difficult.
link |
He's saying this is what has worked.
link |
And yes, a prescription, but it's a difficult prescription because it says all the fun things you guys are trying to do.
link |
We are trying to do.
link |
He's part of the community.
link |
He's saying it's only going to be short term gains.
link |
This all leads up to a question, I guess, on artificial neural networks and maybe our own biological neural networks.
link |
Do you think if we just scale things up significantly?
link |
Take these dumb artificial neurons, the point neurons.
link |
If we just have a lot more of them, do you think some of the elements that we see in the brain
link |
may start emerging?
link |
No, I don't think so.
link |
We can do bigger problems of the same type.
link |
I mean, it's been pointed out by many people that today's convolutional neural networks aren't really much different than the ones we had quite a while ago.
link |
We just, they're bigger and train more and we have more labeled data and so on.
link |
But I don't think you can get to the kind of things I know the brain can do and that we think about as intelligence by just scaling it up.
link |
So that may be, it's a good description of what's happened in the past, what's happened recently with the reemergence of artificial neural networks.
link |
It may be a good prescription for what's going to happen in the short term.
link |
But I don't think that's the path.
link |
I've said that earlier.
link |
There's an alternate path.
link |
I should mention to you, by the way, that we've made sufficient progress on our, the whole cortical theory in the last few years.
link |
But last year, we decided to start actively pursuing how we get these ideas embedded into machine learning.
link |
That's, again, being led by my colleague, and he's more of a machine learning guy.
link |
I'm more of an neuroscience guy.
link |
So this is now our, I wouldn't say our focus, but it is now an equal focus here because we need to proselytize what we've learned.
link |
And we need to show how it's beneficial to the machine learning.
link |
So we're putting, we have a plan in place right now.
link |
In fact, we just did our first paper on this.
link |
I can tell you about that.
link |
But, you know, one of the reasons I want to talk to you is because I'm trying to get more people in the machine learning community to say,
link |
I need to learn about this stuff.
link |
And maybe we should just think about this a bit more about what we've learned about the brain.
link |
And what are those team, what have they done?
link |
Is that useful for us?
link |
Yeah, so is there elements of all the, the cortical theory that things we've been talking about that may be useful in the short term?
link |
Yes, in the short term, yes.
link |
This is the, sorry to interrupt, but the, the open question is it, it certainly feels from my perspective that in the long term,
link |
some of the ideas we've been talking about will be extremely useful.
link |
The question is whether in the short term.
link |
Well, this is a, always what we, I would call the entrepreneur's dilemma.
link |
You have this long term vision, oh, we're going to all be driving electric cars or we're all going to have computers or we're all going to whatever.
link |
And, and you're at some point in time and you say, I can see that long term vision.
link |
I'm sure it's going to happen.
link |
How do I get there without killing myself, you know, without going out of business?
link |
That's the challenge.
link |
That's the dilemma.
link |
That's the really difficult thing to do.
link |
So we're facing that right now.
link |
So ideally what you'd want to do is find some steps along the way that you can get there incrementally.
link |
You don't have to like throw it all out and start over again.
link |
The first thing that we've done is we focus on these sparse representations.
link |
So just in case you don't know what that means or some of the listeners don't know what that means.
link |
In the brain, if I have like 10,000 neurons, what you would see is maybe 2% of them active at a time.
link |
You don't see 50%, you don't see 30%, you might see 2%.
link |
And it's always like that.
link |
For any set of sensory inputs.
link |
It doesn't matter anything.
link |
It doesn't matter any part of the brain.
link |
But which neurons differs?
link |
Which neurons are active?
link |
So let me put this.
link |
Let's say I take 10,000 neurons that are representing something.
link |
They're sitting there in a little block together.
link |
It's a teeny little block of neurons, 10,000 neurons.
link |
And they're representing a location.
link |
They're representing a cop.
link |
They're representing the input from my sensors.
link |
It doesn't matter.
link |
It's representing something.
link |
The way the representations occur, it's always a sparse representation.
link |
Meaning it's a population code.
link |
So which 200 cells are active tells me what's going on.
link |
It's not individual cells aren't that important at all.
link |
It's the population code that matters.
link |
And when you have sparse population codes,
link |
then all kinds of beautiful properties come out of them.
link |
So the brain uses sparse population codes that we've written
link |
and described these benefits in some of our papers.
link |
So they give this tremendous robustness to the systems.
link |
You know, brains are incredibly robust.
link |
Neurons are dying all the time and spasming and synapses falling apart.
link |
And, you know, all the time and it keeps working.
link |
So what Subitai and Louise, one of our other engineers here have done,
link |
have shown that they're introducing sparseness into convolutional neural networks.
link |
Now other people are thinking along these lines,
link |
but we're going about it in a more principled way, I think.
link |
And we're showing that if you enforce sparseness throughout these convolutional neural networks,
link |
in both the sort of which neurons are active and the connections between them,
link |
that you get some very desirable properties.
link |
So one of the current hot topics in deep learning right now are these adversarial examples.
link |
So, you know, you give me any deep learning network
link |
and I can give you a picture that looks perfect and you're going to call it, you know,
link |
you're going to say the monkey is, you know, an airplane.
link |
So that's a problem.
link |
And DARPA just announced some big thing and we're trying to, you know, have some contests for this.
link |
But if you enforce sparse representations here,
link |
many of these problems go away.
link |
They're much more robust and they're not easy to fool.
link |
So we've already shown some of those results,
link |
just literally in January or February, just like last month we did that.
link |
And you can, I think it's on bioarchive right now or on iCry, you can read about it.
link |
But so that's like a baby step.
link |
Okay. That's a take something from the brain.
link |
We know, we know about sparseness.
link |
We know why it's important.
link |
We know what it gives the brain.
link |
So let's try to enforce that onto this.
link |
What's your intuition why sparsity leads to robustness?
link |
Because it feels like it would be less robust.
link |
Why would you feel the rest robust to you?
link |
So it, it just feels like if the fewer neurons are involved,
link |
the more fragile the representation.
link |
Yeah, but I didn't say there was lots of few.
link |
I said, let's say 200.
link |
There's still a lot.
link |
So here's an intuition for it.
link |
This is a bit technical.
link |
So for, you know, for engineers, machine learning people this be easy,
link |
but God's listeners, maybe not.
link |
If you're trying to classify something,
link |
you're trying to divide some very high dimensional space into different pieces, A and B.
link |
And you're trying to create some point where you say all these points in this high dimensional space are A
link |
and all these points in this high dimensional space are B.
link |
And if you have points that are close to that line, it's not very robust.
link |
It works for all the points you know about, but it's, it's not very robust
link |
because you can just move a little bit and you've crossed over the line.
link |
When you have sparse representations, imagine I pick, I have,
link |
I'm going to pick 200 cells active out of, out of 10,000.
link |
So I have 200 cells active.
link |
Now let's say I pick randomly another, a different representation, 200.
link |
The overlap between those is going to be very small, just a few.
link |
I can pick millions of samples randomly of 200 neurons and not one of them will overlap more than just a few.
link |
So one way to think about it is if I want to fool one of these representations to look like one of those other representations,
link |
I can't move just one cell or two cells or three cells or four cells.
link |
I have to move 100 cells.
link |
And that makes them robust.
link |
In terms of further, so you mentioned sparsity.
link |
Will we be the next thing?
link |
So we have, we picked one.
link |
We don't know if it's going to work well yet.
link |
So again, we're trying to come up incremental ways of moving from brain theory to add pieces to machine learning,
link |
current machine learning world in one step at a time.
link |
So the next thing we're going to try to do is, is sort of incorporate some of the ideas of the, the thousand brains theory that you have many,
link |
many models and that are voting.
link |
Now that idea is not new.
link |
There's a mixture of models that's been around for a long time.
link |
But the way the brain does it is a little different.
link |
And, and the way it votes is different and the kind of way it represents uncertainty is different.
link |
So we're just starting this work, but we're going to try to see if we can sort of incorporate some of the principles of voting
link |
or principles of a thousand brain theory, like lots of simple models that talk to each other in a, in a very certain way.
link |
And can we build more machines and systems that learn faster and, and also, well, mostly are multimodal and robust to multimodal type of issues.
link |
So one of the challenges there is, you know, the machine learning computer vision community has certain sets of benchmarks.
link |
So it's a test based on which they compete.
link |
And I would argue, especially from your perspective, that those benchmarks aren't that useful for testing the aspects that the brain is good at or intelligent.
link |
They're not really testing intelligence.
link |
They're very fine and has been extremely useful for developing specific mathematical models, but it's not useful in the long term for creating intelligence.
link |
So do you think you also have a role in proposing better tests?
link |
Yeah, this is a very, you've identified a very serious problem.
link |
First of all, the tests that they have are the tests that they want, not the tests of the other things that we're trying to do.
link |
Right. You know, what are the, so on.
link |
The second thing is sometimes these to be competitive in these tests, you have to have huge data sets and huge computing power.
link |
And so, you know, and we don't have that here.
link |
We don't have it as well as other big teams that big companies do.
link |
So there's numerous issues there.
link |
You know, we come at it, you know, we're our approach to this is all based on, in some sense, you might argue elegance.
link |
You know, we're coming at it from like a theoretical base that we think, oh my God, this is so clearly elegant.
link |
This is how brains work.
link |
This is what intelligence is.
link |
But the machine learning world has gotten in this phase where they think it doesn't matter.
link |
Doesn't matter what you think, as long as you do, you know, 0.1% better on this benchmark.
link |
That's what that's all that matters.
link |
And that's a problem.
link |
You know, we have to figure out how to get around that.
link |
That's a challenge for us.
link |
That's one of the challenges we have to deal with.
link |
So I agree you've identified a big issue.
link |
It's difficult for those reasons.
link |
But, you know, part of the reasons I'm talking to you here today is I hope I'm going to get some machine learning people to say,
link |
I'm going to read those papers.
link |
Those might be some interesting ideas.
link |
I'm tired of doing this 0.1% improvement stuff, you know.
link |
Well, that's why I'm here as well, because I think machine learning now as a community is at a place where the next step is needs to be orthogonal to what has received success in the past.
link |
You see other leaders saying this, machine learning leaders, you know, Jeff Hinton with his capsules idea.
link |
Many people have gotten up saying, you know, we're going to hit road, maybe we should look at the brain, you know, things like that.
link |
So hopefully that thinking will occur organically.
link |
And then we're in a nice position for people to come and look at our work and say, well, what can we learn from these guys?
link |
Yeah, MIT is just launching a billion dollar computing college that's centered around this idea.
link |
On this idea of what?
link |
Well, the idea that, you know, the humanities, psychology, neuroscience have to work all together to get to build the S.
link |
Yeah, I mean, Stanford just did this human center today, I think.
link |
I'm a little disappointed in these initiatives because, you know, they're focusing on sort of the human side of it,
link |
and it can very easily slip into how humans interact with intelligent machines, which is nothing wrong with that.
link |
But that's not, that is orthogonal to what we're trying to do.
link |
We're trying to say, like, what is the essence of intelligence?
link |
In fact, I want to build intelligent machines that aren't emotional, that don't smile at you, that, you know, that aren't trying to tuck you in at night.
link |
Yeah, there is that pattern that you, when you talk about understanding humans is important for understanding intelligence.
link |
You start slipping into topics of ethics or, yeah, like you said, the interactive elements as opposed to, no, no, no, let's zoom in on the brain,
link |
study what the human brain, the baby, the...
link |
Let's study what a brain does.
link |
And then we can decide which parts of that we want to recreate in some system.
link |
But until you have that theory about what the brain does, what's the point?
link |
You know, it's just, you're going to be wasting time, I think.
link |
Just to break it down on the artificial neural network side, maybe you can speak to this on the, on the biologic neural network side,
link |
the process of learning versus the process of inference.
link |
Maybe you can explain to me, what, is there a difference between, you know, in artificial neural networks, there's a difference between the learning stage and the inference stage?
link |
Do you see the brain as something different?
link |
One of the big distinctions that people often say, I don't know how correct it is, is artificial neural networks need a lot of data.
link |
They're very inefficient learning.
link |
Do you see that as a correct distinction from the biology of the human brain, that the human brain is very efficient?
link |
Or is that just something we deceive ourselves with?
link |
No, it is efficient, obviously.
link |
We can learn new things almost instantly.
link |
And so what elements do you think...
link |
Yeah, I can talk about that.
link |
You brought up two issues there.
link |
So remember I talked early about the constraints, we always feel, well, one of those constraints is the fact that brains are continually learning.
link |
That's not something we said, oh, we can add that later.
link |
That's something that was upfront, had to be there from the start, made our problems harder.
link |
But we showed, going back to the 2016 paper on sequence memory, we showed how that happens, how the brains infer and learn at the same time.
link |
And our models do that.
link |
They're not two separate phases or two separate sets of time.
link |
I think that's a big, big problem in AI, at least for many applications, not for all.
link |
So I can talk about that.
link |
There are some parts of the neocortex in the brain where actually what's going on, there's these cycles of activity in the brain.
link |
And there's very strong evidence that you're doing more of inference on one part of the phase and more of learning on the other part of the phase.
link |
So the brain can actually sort of separate different populations of cells that are going back and forth like this.
link |
But in general, I would say that's an important problem.
link |
We have all of our networks that we've come up with do both.
link |
They're continuous learning networks.
link |
And you mentioned benchmarks earlier.
link |
Well, there are no benchmarks about that.
link |
So we have to like, we get in our little soapbox and hey, by the way, this is important.
link |
And here's the mechanism for doing that.
link |
But until you can prove it to someone in some commercial system or something, it's a little harder.
link |
So one of the things I had to linger on that is in some ways to learn the concept of a coffee cup.
link |
You only need this one coffee cup and maybe some time alone in a room with it.
link |
So the first thing is I imagine I reach my hand into a black box and I'm reaching, I'm trying to touch something.
link |
I don't know up front if it's something I already know or if it's a new thing.
link |
And I have to, I'm doing both at the same time.
link |
I don't say, oh, let's see if it's a new thing.
link |
Oh, let's see if it's an old thing.
link |
As I go, my brain says, oh, it's new or it's not new.
link |
And if it's new, I start learning what it is.
link |
And by the way, it starts learning from the get go, even if it's going to recognize it.
link |
So they're not separate problems.
link |
And so that's the thing there.
link |
The other thing you mentioned was the fast learning.
link |
So I was just talking about continuous learning, but there's also fast learning.
link |
Literally, I can show you this coffee cup and I say, here's a new coffee cup.
link |
It's got the logo on it.
link |
Take a look at it.
link |
You can predict what it's going to look like, you know, in different positions.
link |
So I can talk about that too.
link |
In the brain, the way learning occurs.
link |
I mentioned this earlier, but I mentioned it again.
link |
The way learning occurs, I imagine I have a section of a dendrite of a neuron.
link |
And I want to learn, I'm going to learn something new.
link |
It just doesn't matter what it is.
link |
I'm just going to learn something new.
link |
I need to recognize a new pattern.
link |
So what I'm going to do is I'm going to form new synapses.
link |
New synapses, we're going to rewire the brain onto that section of the dendrite.
link |
Once I've done that, everything else that neuron has learned is not affected by it.
link |
Now, it's because it's isolated to that small section of the dendrite.
link |
They're not all being added together, like a point neuron.
link |
So if I learn something new on this segment here, it doesn't change any of the learning
link |
that occur anywhere else in that neuron.
link |
So I can add something without affecting previous learning.
link |
And I can do it quickly.
link |
Now, let's talk, we can talk about the quickness, how it's done in real neurons.
link |
You might say, well, doesn't it take time to form synapses?
link |
Yes, it can take maybe an hour to form a new synapse.
link |
We can form memories quicker than that.
link |
And I can explain that happens too, if you want.
link |
But it's getting a bit neurosciencey.
link |
But is there an understanding of these mechanisms at every level?
link |
So from the short term memories and the forming new connections.
link |
So this idea of synaptogenesis, the growth of new synapses,
link |
that's well described, as well understood.
link |
And that's an essential part of learning.
link |
You know, back, you know, going back many, many years, people, you know,
link |
was, what's his name, the psychologist proposed,
link |
Hebb, Donald Hebb.
link |
He proposed that learning was the modification of the strength
link |
of a connection between two neurons.
link |
People interpreted that as the modification of the strength of a synapse.
link |
He didn't say that.
link |
He just said there's a modification between the effect of one neuron and another.
link |
So synaptogenesis is totally consistent with Donald Hebb said.
link |
But anyway, there's these mechanisms, the growth of new synapse.
link |
You can go online, you can watch a video of a synapse growing in real time.
link |
It's literally, you can see this little thing going.
link |
It's pretty impressive.
link |
So that those mechanisms are known.
link |
Now, there's another thing that we've speculated and we've written about,
link |
which is consistent with no neuroscience, but it's less proven.
link |
And this is the idea.
link |
How do I form a memory really, really quickly?
link |
Like instantaneous.
link |
If it takes an hour to grow a synapse, like that's not instantaneous.
link |
So there are types of synapses called silent synapses.
link |
They look like a synapse, but they don't do anything.
link |
They're just sitting there.
link |
It's like if an action potential comes in, it doesn't release any neurotransmitter.
link |
Some parts of the brain have more of these than others.
link |
For example, the hippocampus has a lot of them,
link |
which is where we associate most short term memory with.
link |
So what we speculated, again, in that 2016 paper,
link |
we proposed that the way we form very quick memories,
link |
very short term memories, or quick memories,
link |
is that we convert silent synapses into active synapses.
link |
It's like saying a synapse has a zero weight and a one weight.
link |
But the long term memory has to be formed by synaptogenesis.
link |
So you can remember something really quickly
link |
by just flipping a bunch of these guys from silent to active.
link |
It's not from 0.1 to 0.15.
link |
It doesn't do anything until it releases transmitter.
link |
If I do that over a bunch of these, I've got a very quick short term memory.
link |
So I guess the lesson behind this is that most neural networks today are fully connected.
link |
Every neuron connects every other neuron from layer to layer.
link |
That's not correct in the brain.
link |
We don't want that.
link |
We actually don't want that.
link |
You want a very sparse connectivity so that any neuron connects
link |
to some subset of the neurons in the other layer,
link |
and it does so on a dendrite by dendrite segment basis.
link |
So it's a very parcelated out type of thing.
link |
And that then learning is not adjusting all these weights,
link |
but learning is just saying, OK, connect to these 10 cells here right now.
link |
In that process, you know, with artificial neural networks,
link |
it's a very simple process of back propagation that adjusts the weights.
link |
The process of synaptogenesis.
link |
Back propagation requires something that really can't happen in brains.
link |
This back propagation of this error signal.
link |
They really can't happen.
link |
People are trying to make it happen in brains, but it doesn't happen in brain.
link |
This is pure Hebbian learning.
link |
Well, synaptogenesis is pure Hebbian learning.
link |
It's basically saying there's a population of cells over here that are active right now,
link |
and there's a population of cells over here active right now.
link |
How do I form connections between those active cells?
link |
And it's literally saying this guy became active.
link |
These 100 neurons here became active before this neuron became active.
link |
So form connections to those ones.
link |
There's no propagation of error, nothing.
link |
All the networks we do, all models we have work on almost completely on Hebbian learning,
link |
but on dendritic segments and multiple synaptoses at the same time.
link |
So now let's turn the question that you already answered,
link |
and maybe you can answer it again.
link |
If you look at the history of artificial intelligence, where do you think we stand?
link |
How far are we from solving intelligence?
link |
You said you were very optimistic.
link |
Can you elaborate on that?
link |
Yeah, it's always the crazy question to ask, because no one can predict the future.
link |
So I'll tell you a story.
link |
I used to run a different neuroscience institute called the Redwood Neuroscience Institute,
link |
and we would hold these symposiums, and we'd get like 35 scientists from around the world to come together.
link |
And I used to ask them all the same question.
link |
I would say, well, how long do you think it'll be before we understand how the New York Cortex works?
link |
And everyone went around the room, and they had introduced the name, and they had to answer that question.
link |
So I got, the typical answer was 50 to 100 years.
link |
Some people would say 500 years.
link |
Some people said never.
link |
I said, why are you a neuroscience institute?
link |
But it doesn't work like that.
link |
As I mentioned earlier, these are step functions.
link |
Things happen, and then bingo, they happen.
link |
You can't predict that.
link |
I feel I've already passed a step function.
link |
So if I can do my job correctly over the next five years, then meaning I can proselytize these ideas.
link |
I can convince other people they're right.
link |
We can show that machine learning people should pay attention to these ideas.
link |
Then we're definitely in an under 20 year time frame.
link |
If I can do those things, if I'm not successful in that, and this is the last time anyone talks to me,
link |
and no one reads our papers, and I'm wrong or something like that, then I don't know.
link |
But it's not 50 years.
link |
It's the same thing about electric cars.
link |
How quickly are they going to populate the world?
link |
It probably takes about a 20 year span.
link |
It'll be something like that.
link |
But I think if I can do what I said, we're starting it.
link |
Of course, there could be other use of step functions.
link |
It could be everybody gives up on your ideas for 20 years, and then all of a sudden somebody picks it up again.
link |
Wait, that guy was onto something.
link |
That would be a failure on my part.
link |
Think about Charles Babbage.
link |
Charles Babbage used to invented the computer back in the 1800s.
link |
Everyone forgot about it until 100 years later.
link |
This guy figured this stuff out a long time ago, but he was ahead of his time.
link |
As I said, I recognize this is part of any entrepreneur's challenge.
link |
I use entrepreneur broadly in this case.
link |
I'm not meaning like I'm building a business trying to sell something.
link |
I'm trying to sell ideas.
link |
This is the challenge as to how you get people to pay attention to you.
link |
How do you get them to give you positive or negative feedback?
link |
How do you get the people to act differently based on your ideas?
link |
We'll see how well we do on that.
link |
There's a lot of hype behind artificial intelligence currently.
link |
Do you, as you look to spread the ideas that are in your cortical theory, the things you're working on,
link |
do you think there's some possibility we'll hit an AI winter once again?
link |
It's certainly a possibility.
link |
That's something you worry about?
link |
I guess, do I worry about it?
link |
I haven't decided yet if that's good or bad for my mission.
link |
That's very true because it's almost like you need the winter to refresh the pallet.
link |
Here's what you want to have it.
link |
To the extent that everyone is so thrilled about the current state of machine learning and AI,
link |
and they don't imagine they need anything else, it makes my job harder.
link |
If everything crashed completely and every student left the field,
link |
and there was no money for anybody to do anything,
link |
and it became an embarrassment to talk about machine intelligence and AI,
link |
that wouldn't be good for us either.
link |
You want the soft landing approach, right?
link |
You want enough people, the senior people in AI and machine learning to say,
link |
you know, we need other approaches.
link |
We really need other approaches.
link |
Damn, we need other approaches.
link |
Maybe we should look to the brain.
link |
Okay, let's look to the brain.
link |
Who's got some brain ideas?
link |
Okay, let's start a little project on the side here trying to do brain idea related stuff.
link |
That's the ideal outcome we would want.
link |
So I don't want a total winter,
link |
and yet I don't want it to be sunny all the time either.
link |
So what do you think it takes to build a system with human level intelligence
link |
where once demonstrated, you would be very impressed?
link |
So does it have to have a body?
link |
Does it have to have the C word we used before consciousness as an entirety in a holistic sense?
link |
First of all, I don't think the goal is to create a machine that is human level intelligence.
link |
I think it's a false goal.
link |
Back to Turing, I think it was a false statement.
link |
We want to understand what intelligence is,
link |
and then we can build intelligent machines of all different scales,
link |
all different capabilities.
link |
You know, a dog is intelligent.
link |
You know, that would be pretty good to have a dog, you know,
link |
but what about something that doesn't look like an animal at all in different spaces?
link |
So my thinking about this is that we want to define what intelligence is,
link |
agree upon what makes an intelligent system.
link |
We can then say, okay, we're now going to build systems that work on those principles,
link |
or some subset of them, and we can apply them to all different types of problems.
link |
And the kind, the idea, it's not computing.
link |
We don't ask, if I take a little, you know, little one chip computer,
link |
I don't say, well, that's not a computer because it's not as powerful as this, you know,
link |
big server over here.
link |
No, no, because we know that what the principles are computing on,
link |
and I can apply those principles to a small problem or into a big problem.
link |
And same, intelligence needs to get there.
link |
We have to say, these are the principles.
link |
I can make a small one, a big one.
link |
I can make them distributed.
link |
I can put them on different sensors.
link |
They don't have to be human like at all.
link |
Now, you did bring up a very interesting question about embodiment.
link |
Does it have to have a body?
link |
It has to have some concept of movement.
link |
It has to be able to move through these reference frames.
link |
I talked about earlier, whether it's physically moving,
link |
like I need, if I'm going to have an AI that understands coffee cups,
link |
it's going to have to pick up the coffee cup and touch it and look at it with its eyes
link |
and hands or something equivalent to that.
link |
If I have a mathematical AI, maybe it needs to move through mathematical spaces.
link |
I could have a virtual AI that lives in the Internet
link |
and its movements are traversing links and digging into files,
link |
but it's got a location that it's traveling through some space.
link |
You can't have an AI that just takes some flash thing input,
link |
you know, we call it flash inference.
link |
Here's a pattern done.
link |
No, it's movement pattern, movement pattern, movement pattern.
link |
Attention, digging, building, building structure,
link |
just trying to figure out the model of the world.
link |
So some sort of embodiment, whether it's physical or not, has to be part of it.
link |
So self awareness in the way to be able to answer where am I?
link |
You bring up self awareness, it's a different topic, self awareness.
link |
No, the very narrow definition, meaning knowing a sense of self enough to know
link |
where am I in the space where essentially.
link |
The system needs to know its location,
link |
where each component of the system needs to know where it is in the world at that point in time.
link |
So self awareness and consciousness.
link |
Do you think, one, from the perspective of neuroscience and neurocortex,
link |
these are interesting topics, solvable topics,
link |
do you have any ideas of why the heck it is that we have a subjective experience at all?
link |
Yeah, I have a lot of questions.
link |
And is it useful, or is it just a side effect of us?
link |
It's interesting to think about.
link |
I don't think it's useful as a means to figure out how to build intelligent machines.
link |
It's something that systems do, and we can talk about what it is,
link |
that are like, well, if I build a system like this, then it would be self aware.
link |
Or if I build it like this, it wouldn't be self aware.
link |
So that's a choice I can have.
link |
It's not like, oh my God, it's self aware.
link |
I heard an interview recently with this philosopher from Yale.
link |
I can't remember his name.
link |
I apologize for that.
link |
But he was talking about, well, if these computers were self aware,
link |
then it would be a crime to unplug them.
link |
And I'm like, oh, come on.
link |
I unplug myself every night.
link |
I plug myself in again in the morning.
link |
People get kind of bent out of shape about this.
link |
I have very detailed understanding or opinions about what it means to be conscious
link |
and what it means to be self aware.
link |
I don't think it's that interesting a problem.
link |
You talked about Christoph Koch.
link |
He thinks that's the only problem.
link |
I didn't actually listen to your interview with him.
link |
But I know him, and I know that's the thing.
link |
He also thinks intelligence and consciousness are disjoint.
link |
So I mean, it's not, you don't have to have one or the other.
link |
I just agree with that.
link |
I just totally disagree with that.
link |
So where's your thoughts and consciousness?
link |
Where does it emerge from?
link |
Then we have to break it down to the two parts.
link |
Because consciousness isn't one thing.
link |
That's part of the problem with that term.
link |
It means different things to different people.
link |
And there's different components of it.
link |
There is a concept of self awareness.
link |
That can be very easily explained.
link |
You have a model of your own body.
link |
The neocortex models things in the world.
link |
And it also models your own body.
link |
And then it has a memory.
link |
It can remember what you've done.
link |
So it can remember what you did this morning.
link |
It can remember what you had for breakfast and so on.
link |
And so I can say to you, okay, Lex,
link |
were you conscious this morning when you had your bagel?
link |
And you'd say, yes, I was conscious.
link |
Now, what if I could take your brain and revert all the synapses
link |
back to the state they were this morning?
link |
And then I said to you, Lex,
link |
were you conscious when you ate the bagel?
link |
And he said, no, I wasn't conscious.
link |
I said, here's a video of eating the bagel.
link |
And he said, I wasn't there.
link |
That's not possible because I must have been unconscious at that time.
link |
So we can just make this one to one correlation
link |
between memory of your body's trajectory through the world
link |
over some period of time.
link |
And the ability to recall that memory
link |
is what you would call conscious.
link |
I was conscious of that. It's self awareness.
link |
And any system that can recall,
link |
memorize what it's done recently
link |
and bring that back and invoke it again
link |
would say, yeah, I'm aware.
link |
I remember what I did. All right, I got it.
link |
That's an easy one. Although some people think that's a hard one.
link |
The more challenging part of consciousness
link |
is this is one that's sometimes used
link |
by the word Aqualia,
link |
which is, you know, why does an object seem red?
link |
And why does pain feel like something?
link |
Why do I feel redness?
link |
So why do I feel a little painless in a way?
link |
And then I could say, well, why does sight
link |
seems different than hearing? You know, it's the same problem.
link |
It's really, you know, these are all just neurons.
link |
And so how is it that why does looking at you
link |
feel different than, you know, hearing you?
link |
It feels different, but there's just neurons in my head.
link |
They're all doing the same thing.
link |
So that's an interesting question.
link |
The best treatise I've read about this
link |
is by a guy named Oregon.
link |
He wrote a book called Why Red Doesn't Sound Like a Bell.
link |
It's a little, it's not a trade book, easy to read,
link |
but it, and it's an interesting question.
link |
Take something like color.
link |
Color really doesn't exist in the world.
link |
It's not a property of the world.
link |
Property of the world that exists is light frequency,
link |
and that gets turned into we have certain cells
link |
in the retina that respond to different frequencies
link |
different than others.
link |
And so when they enter the brain, you just have a bunch
link |
of axons that are firing at different rates,
link |
and from that we perceive color.
link |
But there is no color in the brain.
link |
I mean, there's no color coming in on those synapses.
link |
It's just a correlation between some axons
link |
and some property of frequency.
link |
And that isn't even color itself.
link |
Frequency doesn't have a color.
link |
It's just what it is.
link |
So then the question is, well, why does it even
link |
appear to have a color at all?
link |
Just as you're describing it, there seems to be a connection
link |
to those ideas of reference frames.
link |
I mean, it just feels like consciousness having the subject,
link |
assigning the feeling of red to the actual color
link |
or to the wavelength is useful for intelligence.
link |
Yeah, I think that's a good way of putting it.
link |
It's useful as a predictive mechanism,
link |
or useful as a generalization idea.
link |
It's a way of grouping things together to say
link |
it's useful to have a model like this.
link |
Think about the well known syndrome that people
link |
who've lost a limb experience called phantom limbs.
link |
And what they claim is they can have their arms removed
link |
but they feel the arm.
link |
Not only feel it, they know it's there.
link |
I know it's there.
link |
They'll swear to you that it's there.
link |
And then they can feel pain in their arm.
link |
And they'll feel pain in their finger.
link |
And if they move their non existent arm behind their back,
link |
then they feel the pain behind their back.
link |
So this whole idea that your arm exists
link |
is a model of your brain.
link |
It may or may not really exist.
link |
And just like, but it's useful to have a model of something
link |
that sort of correlates to things in the world
link |
so you can make predictions about what would happen
link |
when those things occur.
link |
It's a little bit of a fuzzy,
link |
but I think you're getting quite towards the answer there.
link |
It's useful for the model to express things certain ways
link |
that we can then map them into these reference frames
link |
and make predictions about them.
link |
I need to spend more time on this topic.
link |
It doesn't bother me.
link |
Do you really need to spend more time on this?
link |
It does feel special that we have subjective experience,
link |
but I'm yet to know why.
link |
I'm just personally curious.
link |
It's not necessary for the work we're doing here.
link |
I don't think I need to solve that problem
link |
to build intelligent machines at all.
link |
But there is sort of the silly notion that you described briefly
link |
that doesn't seem so silly to us humans is,
link |
you know, if you're successful building intelligent machines,
link |
it feels wrong to then turn them off.
link |
Because if you're able to build a lot of them,
link |
it feels wrong to then be able to, you know,
link |
to turn off the...
link |
Let's break that down a bit.
link |
As humans, why do we fear death?
link |
There's two reasons we fear death.
link |
Well, first of all, I'll say when you're dead, it doesn't matter.
link |
So why do we fear death?
link |
We fear death for two reasons.
link |
One is because we are programmed genetically to fear death.
link |
That's a survival and propaganda in the genes thing.
link |
And we also are programmed to feel sad when people we know die.
link |
We don't feel sad for someone we don't know dies.
link |
There's people dying right now.
link |
They're only scared to say,
link |
I don't feel bad about them because I don't know them.
link |
But I knew they might feel really bad.
link |
So again, these are old brain genetically embedded things
link |
that we fear death.
link |
Outside of those uncomfortable feelings,
link |
there's nothing else to worry about.
link |
Well, wait, hold on a second.
link |
Do you know the denial of death by Beckard?
link |
You know, there's a thought that death is, you know,
link |
our whole conception of our world model kind of assumes immortality.
link |
And then death is this terror that underlies it all.
link |
So like, well, some people's world model, not mine.
link |
So what, what Becker would say is that you're just living in an illusion.
link |
You've constructed illusion for yourself because it's such a terrible terror.
link |
The fact that this illusion, the illusion that death doesn't matter.
link |
You're still not coming to grips with the illusion of what?
link |
That death is going to happen.
link |
Oh, it's not going to happen.
link |
You're, you're actually operating.
link |
You haven't, even though you said you've accepted it,
link |
you haven't really accepted the notion of death is what you say.
link |
So it sounds like it sounds like you disagree with that notion.
link |
Like, I literally every night, every night I go to bed, it's like dying.
link |
And if I didn't wake up, it wouldn't matter to me.
link |
Only if I knew that was going to happen would it be bothersome.
link |
If I didn't know it was going to happen, how would I know?
link |
Then I would worry about my wife.
link |
So imagine, imagine I was a loner and I lived in Alaska and I lived them out there and there
link |
Nobody knew I existed.
link |
I was just eating these roots all the time.
link |
And nobody knew I was there.
link |
And one day I didn't wake up.
link |
Where, what, what pain in the world would there exist?
link |
Well, so most people that think about this problem would say that you're just deeply enlightened
link |
or are completely delusional.
link |
But I would say, I would say that's a very enlightened way to see the world is that that's the rational one.
link |
Well, I think it's rational.
link |
But the fact is we don't, I mean, we really don't have an understanding of why the heck
link |
it is we're born and why we die and what happens after we die.
link |
Well, maybe there isn't a reason.
link |
So I'm interested in those big problems too, right?
link |
You know, you, you interviewed Max Tagmark, you know, and there's people like that, right?
link |
I'm interested in those big problems as well.
link |
And in fact, when I was young, I made a list of the biggest problems I could think of.
link |
First, why does anything exist?
link |
Second, why did we have the laws of physics that we have?
link |
Third, is life inevitable?
link |
And why is it here?
link |
Fourth, is intelligence inevitable?
link |
And why is it here?
link |
I stopped there because I figured if you can make a truly intelligent system, we'll be,
link |
that'll be the quickest way to answer the first three questions.
link |
And so I said, my mission, you know, you asked me earlier, my first mission is to understand
link |
the brain, but I felt that is the shortest way to get to true machine intelligence.
link |
And I want to get to true machine intelligence because even if it doesn't occur in my lifetime,
link |
other people will benefit from it because I think it'll occur in my lifetime.
link |
But, you know, 20 years, you never know.
link |
And, but that will be the quickest way for us to, you know, we can make super mathematicians.
link |
We can make super space explorers.
link |
We can make super physicists brains that do these things and that can run experiments
link |
that we can't run.
link |
We don't have the abilities to manipulate things and so on.
link |
But we can build and tell the machines to do all those things.
link |
And with the ultimate goal of finding out the answers to the other questions.
link |
Let me ask, you know, the depressing and difficult question, which is once we achieve that goal,
link |
do you, of creating, no, of understanding intelligence, do you think we would be happier,
link |
more fulfilled as a species?
link |
The understanding intelligence or understanding the answers to the big questions?
link |
Understanding intelligence.
link |
It would be far more fun place to live.
link |
I just put aside this, you know, terminator nonsense and, and, and, and just think about,
link |
you can think about the, we can talk about the risk of AI if you want.
link |
So let's talk about.
link |
But I think the world is far better knowing things.
link |
We're always better than no things.
link |
Do you think it's better?
link |
Is it a better place to live in that I know that our planet is one of many in the solar system
link |
and the solar system is one of many in the galaxy?
link |
I think it's a more, I dread, I used to, I sometimes think like, God, what would be like
link |
the 300 years ago, I'd be looking up the sky, I can't understand anything.
link |
I'd be like going to bed every night going, what's going on here?
link |
Well, I mean, in some sense, I agree with you, but I'm not exactly sure.
link |
So I'm also a scientist.
link |
So I have, I share your views, but I'm not, we're, we're like rolling down the hill together.
link |
What's down the hill?
link |
I feel for climbing a hill.
link |
Whatever we're getting, we're getting closer to enlightenment.
link |
We're climbing, we're getting pulled up a hill.
link |
Pulled up by our curiosity.
link |
We're pulling ourselves up the hill by our curiosity.
link |
Yeah, sycophers are doing the same thing with the rock.
link |
But okay, our happiness aside, do you have concerns about, you know, you talk about Sam Harris, Elon Musk,
link |
of existential threats of intelligence systems?
link |
No, I'm not worried about existential threats at all.
link |
There are some things we really do need to worry about.
link |
Even today's AI, we have things we have to worry about.
link |
We have to worry about privacy and about how it impacts false beliefs in the world.
link |
And we have real problems that, and things to worry about with today's AI.
link |
And that will continue as we create more intelligent systems.
link |
There's no question, you know, the whole issue about, you know, making intelligent armament and weapons is something that really we have to think about carefully.
link |
I don't think of those as existential threats.
link |
I think those are the kind of threats we always face and we'll have to face them here and we'll have to deal with them.
link |
We can talk about what people think are the existential threats, but when I hear people talking about them, they all sound hollow to me.
link |
They're based on ideas, they're based on people who really have no idea what intelligence is.
link |
And if they knew what intelligence was, they wouldn't say those things.
link |
So those are not experts in the field, you know.
link |
So there's two, right?
link |
So one is like superintelligence.
link |
So a system that becomes far, far superior in reasoning ability than us humans.
link |
And how is that an existential threat?
link |
So there's a lot of ways in which it could be.
link |
One way is us humans are actually irrational, inefficient and get in the way of not happiness,
link |
but whatever the objective function is of maximizing that objective function and superintelligence.
link |
The paperclip problem and things like that.
link |
So the paperclip problem, but with the superintelligence.
link |
So we already faced this threat in some sense.
link |
They're called bacteria.
link |
These are organisms in the world that would like to turn everything into bacteria.
link |
And they're constantly morphing.
link |
They're constantly changing to evade our protections.
link |
And in the past, they have killed huge swaths of populations of humans on this planet.
link |
So if you want to worry about something that's going to multiply endlessly, we have it.
link |
And I'm far more worried in that regard, I'm far more worried that some scientists in the laboratory
link |
will create a super virus or a super bacteria that we cannot control.
link |
That is a more existential threat.
link |
Putting an intelligence thing on top of it actually seems to make it less existential to me.
link |
It's like, it limits its power.
link |
It limits where it can go.
link |
It limits the number of things it can do in many ways.
link |
A bacteria is something you can't even see.
link |
So that's only one of those problems.
link |
So the other one, just in your intuition about intelligence,
link |
when you think about intelligence of us humans,
link |
do you think of that as something,
link |
if you look at intelligence on a spectrum from zero to us humans,
link |
do you think you can scale that to something far superior to all the mechanisms we've been talking about?
link |
I want to make another point here, Alex, before I get there.
link |
Intelligence is the neocortex.
link |
It is not the entire brain.
link |
The goal is not to make a human.
link |
The goal is not to make an emotional system.
link |
The goal is not to make a system that wants to have sex and reproduce.
link |
Why would I build that?
link |
If I want to have a system that wants to reproduce and have sex,
link |
make bacteria, make computer viruses.
link |
Those are bad things.
link |
Those are really bad.
link |
Don't do those things.
link |
But if I just say, I want an intelligent system,
link |
why doesn't it have to have any human like emotions?
link |
Why does it even care if it lives?
link |
Why does it even care if it has food?
link |
It doesn't care about those things.
link |
It's just in a trance thinking about mathematics,
link |
or it's out there just trying to build the space for it on Mars.
link |
That's a choice we make.
link |
Don't make human like things.
link |
Don't make replicating things.
link |
Don't make things that have emotions.
link |
Just stick to the neocortex.
link |
That's a view, actually, that I share, but not everybody shares,
link |
in the sense that you have faith and optimism about us as engineers
link |
and systems, humans as builders of systems,
link |
to not put in different stupid things.
link |
This is why I mentioned the bacteria one,
link |
because you might say, well, some person's going to do that.
link |
Well, some person today could create a bacteria
link |
that's resistant to all the known antibacterial agents.
link |
So we already have that threat.
link |
We already know this is going on.
link |
It's not a new threat.
link |
So just accept that, and then we have to deal with it, right?
link |
Yeah, so my point is nothing to do with intelligence.
link |
Intelligence is a separate component that you might apply
link |
to a system that wants to reproduce and do stupid things.
link |
Let's not do that.
link |
Yeah, in fact, it is a mystery why people haven't done that yet.
link |
My dad as a physicist believes that the reason,
link |
for example, nuclear weapons haven't proliferated amongst evil people.
link |
So one belief that I share is that there's not that many evil people in the world
link |
that would use whether it's bacteria or nuclear weapons,
link |
or maybe the future AI systems to do bad.
link |
So the fraction is small.
link |
And the second is that it's actually really hard, technically.
link |
So the intersection between evil and competent is small.
link |
And by the way, to really annihilate humanity,
link |
you'd have to have sort of the nuclear winter phenomenon,
link |
which is not one person shooting or even 10 bombs.
link |
You'd have to have some automated system that detonates a million bombs,
link |
or whatever many thousands we have.
link |
So it's extreme evil combined with extreme competence.
link |
And despite building some stupid system that would automatically,
link |
you know, Dr. Strangelup type of thing, you know,
link |
I mean, look, we could have some nuclear bomb go off in some major city in the world.
link |
I think that's actually quite likely, even in my lifetime.
link |
I don't think that's an unlikely thing, and it would be a tragedy.
link |
But it won't be an existential threat.
link |
And it's the same as, you know, the virus of 1917 or whatever it was,
link |
you know, the influenza.
link |
These bad things can happen and the plague and so on.
link |
We can't always prevent it.
link |
We always try, but we can't.
link |
But they're not existential threats until we combine all those crazy things together.
link |
So on the spectrum of intelligence from zero to human,
link |
do you have a sense of whether it's possible to create several orders of magnitude
link |
or at least double that of human intelligence,
link |
to talk about neural cortex?
link |
I think it's the wrong thing to say, double the intelligence.
link |
Break it down into different components.
link |
Can I make something that's a million times faster than a human brain?
link |
Yes, I can do that.
link |
Could I make something that is, has a lot more storage than a human brain?
link |
Yes, I can do that. More copies come.
link |
Can I make something that attaches to different sensors than a human brain?
link |
Yes, I can do that.
link |
Could I make something that's distributed?
link |
We talked earlier about the departure of neural cortex voting.
link |
They don't have to be co located.
link |
They can be all around the places. I could do that too.
link |
Those are the levers I have, but is it more intelligent?
link |
What depends what I train in on? What is it doing?
link |
So here's the thing.
link |
Let's say larger neural cortex and or whatever size that allows for higher and higher hierarchies
link |
to form, we're talking about reference frames and concepts.
link |
So I could, could I have something that's a super physicist or a super mathematician? Yes.
link |
And the question is, once you have a super physicist, will they be able to understand something?
link |
Do you have a sense that it will be orders, like us compared to ants?
link |
Could we ever understand it?
link |
Most people cannot understand general relativity.
link |
It's a really hard thing to get.
link |
I mean, you can paint it in a fuzzy picture, stretchy space, you know?
link |
But the field equations to do that and the deep intuitions are really, really hard.
link |
And I've tried, I'm unable to do it.
link |
It's easy to get special relative, but general relative, man, that's too much.
link |
And so we already live with this to some extent.
link |
The vast majority of people can't understand actually what the vast majority of other people actually know.
link |
We're just either we don't have the effort to or we can't or we don't have time or just not smart enough, whatever.
link |
So, but we have ways of communicating.
link |
Einstein has spoken in a way that I can understand.
link |
He's given me analogies that are useful.
link |
I can use those analogies for my own work and think about, you know, concepts that are similar.
link |
It's not like he's existed in some other plane.
link |
There's no connection to my plane in the world here.
link |
So that will occur.
link |
It already has occurred.
link |
That's when my point at this story is it already has occurred.
link |
We live it every day.
link |
One could argue that with we create machine intelligence that think a million times faster than us that it'll be so far we can't make the connections.
link |
But, you know, at the moment, everything that seems really, really hard to figure out in the world when you actually figure it out is not that hard.
link |
You know, almost everyone can understand the multiverses.
link |
Almost everyone can understand quantum physics.
link |
Almost everyone can understand these basic things, even though hardly any people could figure those things out.
link |
Yeah, but really understand.
link |
But you don't need to really, only a few people really understand.
link |
You need to only understand the projections, the sprinkles of the useful insights from that.
link |
That was my example of Einstein, right?
link |
His general theory of relativity is one thing that very, very, very few people can get.
link |
And what if we just said those other few people are also artificial intelligences?
link |
In some sense they are, right?
link |
Yeah, they say already.
link |
I mean, Einstein wasn't a really normal person.
link |
He had a lot of weird quirks.
link |
And so the other people who work with him.
link |
So, you know, maybe they already were sort of this actual plane of intelligence that we live with it already.
link |
It's not a problem.
link |
It's still useful and, you know.
link |
So do you think we are the only intelligent life out there in the universe?
link |
I would say that intelligent life has and will exist elsewhere in the universe.
link |
There is a question about contemporaneous intelligence life, which is hard to even answer when we think about relativity and the nature of space time.
link |
We can't say what exactly is this time someplace else in the world.
link |
But I think it's, you know, I do worry a lot about the filter idea, which is that perhaps intelligent species don't last very long.
link |
And so we haven't been around very long.
link |
As a technological species, we've been around for almost nothing, you know, what, 200 years or something like that.
link |
And we don't have any data, a good data point on whether it's likely that we'll survive or not.
link |
So do I think that there have been intelligent life elsewhere in the universe?
link |
Almost certainly, of course.
link |
In the past, in the future, yes.
link |
Does it survive for a long time?
link |
This is another reason I'm excited about our work, is our work meaning the general world of AI.
link |
I think we can build intelligent machines that outlast us.
link |
You know, they don't have to be tied to Earth.
link |
They don't have to, you know, I'm not saying they're recreating, you know, you know, aliens.
link |
I'm just saying, if I asked myself, and this might be a good point to end on here.
link |
If I asked myself, you know, what's special about our species?
link |
We're not particularly interesting physically.
link |
We're not, we don't fly.
link |
We're not good swimmers.
link |
We're not very fast.
link |
We're not very strong, you know.
link |
That's the only thing.
link |
And we are the only species on this planet that's built the model of the world that extends beyond what we can actually sense.
link |
We're the only people who know about the far side of the moon and other universes and other galaxies and other stars and what happens in the atom.
link |
That knowledge doesn't exist anywhere else.
link |
It's only in our heads.
link |
Monkeys don't do it.
link |
That is what we've created that's unique.
link |
And if I ask me, what is the legacy of humanity?
link |
What should our legacy be?
link |
It should be knowledge.
link |
We should preserve our knowledge in a way that it can exist beyond us.
link |
And I think the best way of doing that, in fact, you have to do it, is to have to go along with intelligent machines to understand that knowledge.
link |
It's a very broad idea, but we should be thinking, I call it a state planning for humanity.
link |
We should be thinking about what we want to leave behind when as a species we're no longer here.
link |
And that will happen sometime.
link |
Sooner or later, it's going to happen.
link |
And understanding intelligence and creating intelligence gives us a better chance to prolong.
link |
It does give us a better chance to prolong life.
link |
It gives us a chance to live on other planets.
link |
But even beyond that, I mean, our solar system will disappear one day.
link |
It's given enough time.
link |
I doubt we will ever be able to travel to other things.
link |
But we could tell the stars, but we could send intelligent machines to do that.
link |
Do you have an optimistic, a hopeful view of our knowledge of the echoes of human civilization living through the intelligent systems we create?
link |
Well, I think the intelligent systems are greater.
link |
In some sense, the vessel for bringing them beyond Earth or making them last beyond humans themselves.
link |
So how do you feel about that?
link |
That they won't be human, quote unquote.
link |
Human, what is human? Our species are changing all the time.
link |
Human today is not the same as human just 50 years ago.
link |
What is human? Do we care about our genetics?
link |
Why is that important?
link |
As I point out, our genetics are no more interesting than a bacterium's genetics.
link |
It's no more interesting than a monkey's genetics.
link |
What we have, what's unique and what's valuable is our knowledge, what we've learned about the world.
link |
And that is the rare thing.
link |
That's the thing we want to preserve.
link |
Who cares about our genes?
link |
It's the knowledge.
link |
That's a really good place to end.
link |
Thank you so much for talking to me.