back to indexDavid Chalmers: The Hard Problem of Consciousness | Lex Fridman Podcast #69
link |
The following is a conversation with David Chalmers.
link |
He's a philosopher and cognitive scientist
link |
specializing in the areas of philosophy of mind,
link |
philosophy of language, and consciousness.
link |
He's perhaps best known for formulating
link |
the hard problem of consciousness,
link |
which could be stated as why does the feeling
link |
which accompanies awareness of sensory information
link |
Consciousness is almost entirely a mystery.
link |
Many people who worry about AI safety and ethics
link |
believe that, in some form, consciousness can
link |
and should be engineered into AI systems of the future.
link |
So while there's much mystery, disagreement,
link |
discoveries yet to be made about consciousness,
link |
these conversations, while fundamentally philosophical
link |
in nature, may nevertheless be very important
link |
for engineers of modern AI systems to engage in.
link |
This is the Artificial Intelligence Podcast.
link |
If you enjoy it, subscribe on YouTube,
link |
give it five stars on Apple Podcast,
link |
support it on Patreon, or simply connect with me
link |
on Twitter at Lex Friedman, spelled F R I D M A N.
link |
As usual, I'll do one or two minutes of ads now
link |
and never any ads in the middle
link |
that can break the flow of the conversation.
link |
I hope that works for you
link |
and doesn't hurt the listening experience.
link |
This show is presented by Cash App,
link |
the number one finance app in the App Store.
link |
When you get it, use code LEXBODCAST.
link |
Cash App lets you send money to friends,
link |
buy Bitcoin, and invest in the stock market
link |
with as little as one dollar.
link |
Brokerage services are provided by Cash App Investing,
link |
subsidiary of Square, and member SIPC.
link |
Since Cash App does fractional share trading,
link |
let me mention that the order execution algorithm
link |
that works behind the scenes to create the abstraction
link |
of fractional orders is an algorithmic marvel.
link |
So big props to the Cash App engineers
link |
for solving a hard problem that, in the end,
link |
provides an easy interface that takes a step up
link |
to the next layer of abstraction over the stock market,
link |
making trading more accessible for new investors
link |
and diversification much easier.
link |
If you get Cash App from the App Store or Google Play
link |
and use the code LEXBODCAST, you'll get $10,
link |
and Cash App will also donate $10 to FIRST,
link |
one of my favorite organizations
link |
that is helping to advance robotics and STEM education
link |
for young people around the world.
link |
And now, here's my conversation with David Chalmers.
link |
Do you think we're living in a simulation?
link |
I don't rule it out.
link |
There's probably gonna be a lot of simulations
link |
in the history of the cosmos.
link |
If the simulation is designed well enough,
link |
it'll be indistinguishable from a non simulated reality.
link |
And although we could keep searching for evidence
link |
that we're not in a simulation,
link |
any of that evidence in principle could be simulated.
link |
So I think it's a possibility.
link |
But do you think the thought experiment is interesting
link |
or useful to calibrate how we think
link |
about the nature of reality?
link |
Yeah, I definitely think it's interesting and useful.
link |
In fact, I'm actually writing a book about this right now,
link |
all about the simulation idea,
link |
using it to shed light
link |
on a whole bunch of philosophical questions.
link |
So the big one is how do we know anything
link |
about the external world?
link |
Descartes said, maybe you're being fooled by an evil demon
link |
who's stimulating your brain into thinking,
link |
all this stuff is real when in fact, it's all made up.
link |
Well, the modern version of that is,
link |
how do you know you're not in a simulation?
link |
Then the thought is, if you're in a simulation,
link |
none of this is real.
link |
So that's teaching you something about knowledge.
link |
How do you know about the external world?
link |
I think there's also really interesting questions
link |
about the nature of reality right here.
link |
If we are in a simulation, is all this real?
link |
Is there really a table here?
link |
Is it really a microphone?
link |
Do I really have a body?
link |
The standard view would be, no, we don't.
link |
None of this would be real.
link |
My view is actually that's wrong.
link |
And even if we are in a simulation, all of this is real.
link |
That's why I called this reality 2.0.
link |
New version of reality, different version of reality,
link |
So what's the difference between quote unquote,
link |
real world and the world that we perceive?
link |
So we interact with the world by perceiving it.
link |
It only really exists through the window
link |
of our perception system and in our mind.
link |
So what's the difference between something
link |
that's quote unquote real, that exists perhaps
link |
without us being there, and the world as you perceive it?
link |
Well the world as we perceive it is a very simplified
link |
and distorted version of what's going on underneath.
link |
We already know that from just thinking about science.
link |
You don't see too many obviously quantum mechanical effects
link |
in what we perceive, but we still know quantum mechanics
link |
is going on under all things.
link |
So I like to think the world we perceive
link |
is this very kind of simplified picture of colors
link |
and shapes existing in space and so on.
link |
We know there's a, that's what the philosopher
link |
Wilfred Sellers called the manifest image.
link |
The world as it seems to us, we already know
link |
underneath all that is a very different scientific image
link |
with atoms or quantum wave functions or super strings
link |
or whatever the latest thing is.
link |
And that's the ultimate scientific reality.
link |
So I think of the simulation idea as basically
link |
another hypothesis about what the ultimate
link |
say quasi scientific or metaphysical reality
link |
is going on underneath the world of the manifest image.
link |
The world of the manifest image is this very simple thing
link |
that we interact with that's neutral
link |
on the underlying stuff of reality.
link |
Science can help tell us about that.
link |
Maybe philosophy can help tell us about that too.
link |
And if we eventually take the red pill
link |
and find out we're in a simulation,
link |
my view is that's just another view
link |
about what reality is made of.
link |
The philosopher Immanuel Kant said,
link |
what is the nature of the thing in itself?
link |
I've got a glass here and it's got all these,
link |
it appears to me a certain way, a certain shape,
link |
it's liquid, it's clear.
link |
And he said, what is the nature of the thing
link |
Well, I think of the simulation idea,
link |
it's a hypothesis about the nature of the thing in itself.
link |
It turns out if we're in a simulation,
link |
the thing in itself nature of this glass,
link |
it's okay, it's actually a bunch of data structures
link |
running on a computer in the next universe up.
link |
Yeah, that's what people tend to do
link |
when they think about simulation.
link |
They think about our modern computers
link |
and somehow trivially crudely just scaled up in some sense.
link |
But do you think the simulation,
link |
I mean, in order to actually simulate
link |
something as complicated as our universe
link |
that's made up of molecules and atoms
link |
and particles and quarks and maybe even strings,
link |
all of that would require something
link |
just infinitely many orders of magnitude more
link |
of scale and complexity.
link |
Do you think we're even able to even like conceptualize
link |
what it would take to simulate our universe?
link |
Or does it just slip into this idea
link |
that you basically have to build a universe,
link |
something so big to simulate it?
link |
Does it get this into this fuzzy area
link |
that's not useful at all?
link |
Yeah, well, I mean, our universe
link |
is obviously incredibly complicated.
link |
And for us within our universe to build a simulation
link |
of a universe as complicated as ours
link |
is gonna have obvious problems here.
link |
If the universe is finite,
link |
there's just no way that's gonna work.
link |
Maybe there's some cute way to make it work
link |
if the universe is infinite,
link |
maybe an infinite universe could somehow simulate
link |
a copy of itself, but that's gonna be hard.
link |
Nonetheless, just that we are in a simulation,
link |
I think there's no particular reason
link |
why we have to think the simulating universe
link |
has to be anything like ours.
link |
You've said before that it might be,
link |
so you could think of it in turtles all the way down.
link |
You could think of the simulating universe
link |
different than ours, but we ourselves
link |
could also create another simulating universe.
link |
So you said that there could be these
link |
kind of levels of universes.
link |
And you've also mentioned this hilarious idea,
link |
maybe tongue in cheek, maybe not,
link |
that there may be simulations within simulations,
link |
arbitrarily stacked levels,
link |
and that there may be, that we may be in level 42.
link |
Along those stacks, referencing Hitchhiker's Guide
link |
If we're indeed in a simulation within a simulation
link |
at level 42, what do you think level zero looks like?
link |
The originating universe.
link |
I would expect that level zero is truly enormous.
link |
I mean, not just, if it's finite,
link |
at some extraordinarily large finite capacity,
link |
much more likely it's infinite.
link |
Maybe it's got some very high cardinality
link |
that enables it to support just any number of simulations.
link |
So high degree of infinity at level zero,
link |
slightly smaller degree of infinity at level one.
link |
So by the time you get down to us at level 42,
link |
maybe there's plenty of room for lots of simulations
link |
of finite capacity.
link |
If the top universe is only a small finite capacity,
link |
then obviously that's gonna put very, very serious limits
link |
on how many simulations you're gonna be able to get running.
link |
So I think we can certainly confidently say
link |
that if we're at level 42,
link |
then the top level's pretty damn big.
link |
So it gets more and more constrained
link |
as we get down levels, more and more simplified
link |
and constrained and limited in resources.
link |
Yeah, we still have plenty of capacity here.
link |
What was it Feynman said?
link |
He said there's plenty of room at the bottom.
link |
We're still a number of levels above the degree
link |
where there's room for fundamental computing,
link |
physical computing capacity,
link |
quantum computing capacity at the bottom level.
link |
So we've got plenty of room to play with
link |
and we probably have plenty of room
link |
for simulations of pretty sophisticated universes,
link |
perhaps none as complicated as our universe,
link |
unless our universe is infinite,
link |
but still at the very least
link |
for pretty serious finite universes,
link |
but maybe universes somewhat simpler than ours,
link |
unless of course we're prepared to take certain shortcuts
link |
in the simulation,
link |
which might then increase the capacity significantly.
link |
Do you think the human mind, us people,
link |
in terms of the complexity of simulation
link |
is at the height of what the simulation
link |
might be able to achieve?
link |
Like if you look at incredible entities
link |
that could be created in this universe of ours,
link |
do you have an intuition about
link |
how incredible human beings are on that scale?
link |
I think we're pretty impressive,
link |
but we're not that impressive.
link |
Are we above average?
link |
I mean, I think human beings are at a certain point
link |
in the scale of intelligence,
link |
which made many things possible.
link |
You get through evolution, through single cell organisms,
link |
through fish and mammals and primates,
link |
and something happens.
link |
Once you get to human beings,
link |
we've just reached that level
link |
where we get to develop language,
link |
we get to develop certain kinds of culture,
link |
and we get to develop certain kinds of collective thinking
link |
that has enabled all this amazing stuff to happen,
link |
science and literature and engineering
link |
and culture and so on.
link |
So we had just at the beginning of that
link |
on the evolutionary threshold,
link |
it's kind of like we just got there,
link |
who knows, a few thousand or tens of thousands of years ago.
link |
So we're probably just at the very beginning
link |
for what's possible there.
link |
So I'm inclined to think among the scale
link |
of intelligent beings,
link |
we're somewhere very near the bottom.
link |
I would expect that, for example,
link |
if we're in a simulation,
link |
then the simulators who created us
link |
have got the capacity to be far more sophisticated.
link |
If we're at level 42,
link |
who knows what the ones at level zero are like.
link |
It's also possible that this is the epitome
link |
of what is possible to achieve.
link |
So we as human beings see ourselves maybe as flawed,
link |
see all the constraints, all the limitations,
link |
but maybe that's the magical, the beautiful thing.
link |
Maybe those limitations are the essential elements
link |
for an interesting sort of that edge of chaos,
link |
that interesting existence,
link |
that if you make us much more intelligent,
link |
if you make us much more powerful
link |
in any kind of dimension of performance,
link |
maybe you lose something fundamental
link |
that makes life worth living.
link |
So you kind of have this optimistic view
link |
that we're this little baby,
link |
that then there's so much growth and potential,
link |
but this could also be it.
link |
This is the most amazing thing is us.
link |
Maybe what you're saying is consistent
link |
with what I'm saying.
link |
I mean, we could still have levels of intelligence
link |
but maybe those levels of intelligence on your view
link |
would be kind of boring.
link |
And we kind of get so good at everything
link |
that life suddenly becomes uni dimensional.
link |
So we're just inhabiting this one spot
link |
of like maximal romanticism in the history of evolution.
link |
You get to humans and it's like, yeah,
link |
and then years to come, our super intelligent descendants
link |
are gonna look back at us and say,
link |
those were the days when they just hit
link |
the point of inflection and life was interesting.
link |
So I'd like to think that if there is super intelligent
link |
somewhere in the future,
link |
they'll figure out how to make life super interesting
link |
and super romantic.
link |
Well, you know what they're gonna do.
link |
So what they're gonna do is they realize
link |
how boring life is when you're super intelligent.
link |
So they create a new level of assimilation
link |
and sort of live through the things they've created
link |
by watching them stumble about
link |
in their flawed ways.
link |
So maybe that's, so you create a new level of assimilation
link |
every time you get really bored with how smart and.
link |
This would be kind of sad though,
link |
because if we showed the peak of their existence
link |
would be like watching simulations for entertainment.
link |
Not like saying the peak of our existence now is Netflix.
link |
No, it's all right.
link |
A flip side of that could be the peak of our existence
link |
for many people having children and watching them grow.
link |
That becomes very meaningful.
link |
Okay, you create a simulation that's like creating a family.
link |
Creating like, well, any kind of creation
link |
is kind of a powerful act.
link |
Do you think it's easier to simulate the mind
link |
So I've heard several people, including Nick Bostrom,
link |
think about ideas of maybe you don't need
link |
to simulate the universe,
link |
you can just simulate the human mind.
link |
Or in general, just the distinction
link |
between simulating the entirety of it,
link |
the entirety of the physical world,
link |
or just simulating the mind.
link |
Which one do you see as more challenging?
link |
Well, I think in some sense, the answer is obvious.
link |
It has to be simpler to simulate the mind
link |
than to simulate the universe,
link |
because the mind is part of the universe.
link |
And in order to fully simulate the universe,
link |
you're gonna have to simulate the mind.
link |
So unless we're talking about partial simulations.
link |
And I guess the question is which comes first?
link |
Does the mind come before the universe
link |
or does the universe come before the mind?
link |
So the mind could just be an emergent phenomena
link |
So simulation is an interesting thing
link |
that it's not like creating a simulation perhaps
link |
requires you to program every single thing
link |
that happens in it.
link |
It's just defining a set of initial conditions
link |
and rules based on which it behaves.
link |
Simulating the mind requires you
link |
to have a little bit more,
link |
we're now in a little bit of a crazy land,
link |
but it requires you to understand
link |
the fundamentals of cognition,
link |
perhaps of consciousness,
link |
of perception of everything like that,
link |
that's not created through some kind of emergence
link |
from basic physics laws,
link |
but more requires you to actually understand
link |
the fundamentals of the mind.
link |
How about if we said to simulate the brain?
link |
Rather than the mind.
link |
So the brain is just a big physical system.
link |
The universe is a giant physical system.
link |
To simulate the universe at the very least,
link |
you're gonna have to simulate the brains
link |
as well as all the other physical systems within it.
link |
And it's not obvious that the problems are any worse
link |
for the brain than for,
link |
it's a particularly complex physical system.
link |
But if we can simulate arbitrary physical systems,
link |
we can simulate brains.
link |
There is this further question of whether,
link |
when you simulate a brain,
link |
will that bring along all the features of the mind with it?
link |
Like will you get consciousness?
link |
Will you get thinking?
link |
Will you get free will?
link |
And that's something philosophers have argued over
link |
My own view is if you simulate the brain well enough,
link |
that will also simulate the mind.
link |
But yeah, there's plenty of people who would say no.
link |
You'd merely get like a zombie system,
link |
a simulation of a brain without any true consciousness.
link |
But for you, you put together a brain,
link |
the consciousness comes with it, arise.
link |
Yeah, I don't think it's obvious.
link |
That's your intuition.
link |
My view is roughly that yeah,
link |
what is responsible for consciousness,
link |
it's in the patterns of information processing and so on
link |
rather than say the biology that it's made of.
link |
There's certainly plenty of people out there
link |
who think consciousness has to be say biological.
link |
So if you merely replicate the patterns of information
link |
processing in a nonbiological substrate,
link |
you'll miss what's crucial for consciousness.
link |
I mean, I just don't think there's any particular reason
link |
to think that biology is special here.
link |
You can imagine substituting the biology
link |
for nonbiological systems, say silicon circuits
link |
that play the same role.
link |
The behavior will continue to be the same.
link |
And I think just thinking about what is the true,
link |
when I think about the connection,
link |
the isomorphisms between consciousness and the brain,
link |
the deepest connections to me seem to connect consciousness
link |
to patterns of information processing,
link |
not to specific biology.
link |
So I at least adopted as my working hypothesis
link |
that basically it's the computation and the information
link |
that matters for consciousness.
link |
Same time, we don't understand consciousness,
link |
so all this could be wrong.
link |
So the computation, the flow, the processing,
link |
manipulation of information,
link |
the process is where the consciousness,
link |
the software is where the consciousness comes from,
link |
Roughly the software, yeah.
link |
The patterns of information processing at least
link |
in the hardware, which we could view as software.
link |
It may not be something you can just like program
link |
and load and erase and so on in the way we can
link |
with ordinary software, but it's something at the level
link |
of information processing rather than at the level
link |
of implementation.
link |
So on that, what do you think of the experience of self,
link |
just the experience of the world in a virtual world,
link |
in virtual reality?
link |
Is it possible that we can create sort of
link |
offsprings of our consciousness by existing
link |
in a virtual world long enough?
link |
So yeah, can we be conscious in the same kind
link |
of deep way that we are in this real world
link |
by hanging out in a virtual world?
link |
Yeah, well, the kind of virtual worlds we have now
link |
are interesting but limited in certain ways.
link |
In particular, they rely on us having a brain and so on,
link |
which is outside the virtual world.
link |
Maybe I'll strap on my VR headset or just hang out
link |
in a virtual world on a screen, but my brain
link |
and then my physical environment might be simulated
link |
if I'm in a virtual world, but right now,
link |
there's no attempt to simulate my brain.
link |
There might be some non player characters
link |
in these virtual worlds that have simulated
link |
cognitive systems of certain kinds
link |
that dictate their behavior, but mostly,
link |
they're pretty simple right now.
link |
I mean, some people are trying to combine,
link |
put a bit of AI in their non player characters
link |
to make them smarter, but for now,
link |
inside virtual world, the actual thinking
link |
is interestingly distinct from the physics
link |
of those virtual worlds.
link |
In a way, actually, I like to think this is kind of
link |
reminiscent of the way that Descartes
link |
thought our physical world was.
link |
There's physics, and there's the mind,
link |
and they're separate.
link |
Now we think the mind is somehow connected
link |
to physics pretty deeply, but in these virtual worlds,
link |
there's a physics of a virtual world,
link |
and then there's this brain which is totally
link |
outside the virtual world that controls it
link |
and interacts it when anyone exercises agency
link |
in a video game, that's actually somebody
link |
outside the virtual world moving a controller,
link |
controlling the interaction of things
link |
inside the virtual world.
link |
So right now, in virtual worlds,
link |
the mind is somehow outside the world,
link |
but you could imagine in the future,
link |
once we have developed serious AI,
link |
artificial general intelligence, and so on,
link |
then we could come to virtual worlds
link |
which have enough sophistication,
link |
you could actually simulate a brain
link |
or have a genuine AGI, which would then presumably
link |
be able to act in equally sophisticated ways,
link |
maybe even more sophisticated ways,
link |
inside the virtual world to how it might
link |
in the physical world, and then the question's
link |
gonna come along, that would be kind of a VR,
link |
virtual world internal intelligence,
link |
and then the question is could they have consciousness,
link |
experience, intelligence, free will,
link |
all the things that we have, and again,
link |
my view is I don't see why not.
link |
To linger on it a little bit, I find virtual reality really
link |
incredibly powerful, just even the crude virtual reality
link |
we have now of perhaps there's psychological effects
link |
that make some people more amenable
link |
to virtual worlds than others, but I find myself
link |
wanting to stay in virtual worlds for the most part.
link |
With a headset or on a desktop?
link |
No, with a headset.
link |
Really interesting, because I am totally addicted
link |
to using the internet and things on a desktop,
link |
but when it comes to VR, with a headset,
link |
I don't typically use it for more than 10 or 20 minutes.
link |
There's something just slightly aversive about it, I find,
link |
so I don't, right now, even though I have Oculus Rift
link |
and Oculus Quest and HTC Vive and Samsung, this and that.
link |
You just don't wanna stay in that world for long.
link |
Not for extended periods.
link |
You actually find yourself hanging out in that.
link |
Something about, it's both a combination
link |
of just imagination and considering the possibilities
link |
of where this goes in the future.
link |
It feels like I want to almost prepare my brain for it.
link |
I wanna explore sort of Disneyland
link |
when it's first being built in the early days,
link |
and it feels like I'm walking around
link |
almost imagining the possibilities,
link |
and something through that process allows my mind
link |
to really enter into that world,
link |
but you say that the brain is external to that virtual world.
link |
It is, strictly speaking, true, but...
link |
If you're in VR and you do brain surgery on an avatar,
link |
and you're gonna open up that skull,
link |
what are you gonna find?
link |
Sorry, nothing there.
link |
The brain is elsewhere.
link |
You don't think it's possible to kind of separate them,
link |
and I don't mean in a sense like Descartes,
link |
like a hard separation, but basically,
link |
do you think it's possible with the brain outside
link |
of the virtual rhythm, when you're wearing a headset,
link |
create a new consciousness for prolonged periods of time?
link |
Really feel, like really, like forget
link |
that your brain is outside.
link |
So this is, okay, this is gonna be the case
link |
where the brain is still outside.
link |
It's still outside.
link |
But could living in the VR, I mean,
link |
we already find this, right, with video games.
link |
They're completely immersive, and you get taken up
link |
by living in those worlds,
link |
and it becomes your reality for a while.
link |
So they're not completely immersive,
link |
they're just very immersive.
link |
Completely immersive.
link |
You don't forget the external world, no.
link |
Exactly, so that's what I'm asking.
link |
Do you think it's almost possible
link |
to really forget the external world?
link |
Really, really immerse yourself.
link |
To forget completely?
link |
Why would we forget?
link |
We got pretty good memories.
link |
Maybe you can stop paying attention to the external world,
link |
but this already happens a lot.
link |
I go to work, and maybe I'm not paying attention
link |
I go to a movie, and I'm immersed in that.
link |
So that degree of immersion, absolutely.
link |
But we still have the capacity to remember it,
link |
to completely forget the external world.
link |
I'm thinking that would probably take some,
link |
I don't know, some pretty serious drugs or something
link |
to make your brain do that.
link |
So, I mean, I guess what I'm getting at
link |
is consciousness truly a property
link |
that's tied to the physical brain?
link |
Or can you create sort of different offspring,
link |
copies of consciousnesses based on the worlds
link |
Well, the way we're doing it now,
link |
at least with a standard VR, there's just one brain.
link |
Interacts with the physical world.
link |
Plays a video game, puts on a video headset,
link |
interacts with this virtual world.
link |
And I think we'd typically say there's one consciousness here
link |
that nonetheless undergoes different environments,
link |
takes on different characters in different environments.
link |
This is already something that happens
link |
in the nonvirtual world.
link |
I might interact one way in my home life,
link |
my work life, my social life, and so on.
link |
So at the very least, that will happen
link |
in a virtual world very naturally.
link |
People sometimes adopt the character of avatars
link |
very different from themselves,
link |
maybe even a different gender, different race,
link |
different social background.
link |
So that much is certainly possible.
link |
I would see that as a single consciousness
link |
is taking on different personas.
link |
If you want literal splitting of consciousness
link |
into multiple copies,
link |
I think it's gonna take something more radical than that.
link |
Like maybe you can run different simulations of your brain
link |
in different realities
link |
and then expose them to different histories.
link |
And then you'd split yourself
link |
into 10 different simulated copies,
link |
which then undergo different environments
link |
and then ultimately do become 10
link |
very different consciousnesses.
link |
Maybe that could happen,
link |
but now we're not talking about something
link |
that's possible in the near term.
link |
We're gonna have to have brain simulations
link |
and AGI for that to happen.
link |
So before any of that happens,
link |
it's fundamentally you see it as a singular consciousness,
link |
even though it's experiencing different environments,
link |
it's still connected to same set of memories,
link |
same set of experiences and therefore,
link |
one sort of joint conscious system.
link |
Yeah, or at least no more multiple
link |
than the kind of multiple consciousness
link |
that we get from inhabiting different environments
link |
in a non virtual world.
link |
So you said as a child,
link |
you were a music color synesthete.
link |
So where songs had colors for you.
link |
So what songs had what colors?
link |
You know, this is funny.
link |
I didn't pay much attention to this at the time,
link |
but I'd listen to a piece of music
link |
and I'd get some kind of imagery
link |
of a kind of color.
link |
The weird thing is mostly they were kind of murky,
link |
dark greens and olive browns
link |
and the colors weren't all that interesting.
link |
I don't know what the reason is.
link |
I mean, my theory is that maybe it's like different chords
link |
and tones provided different colors
link |
and they all tended to get mixed together
link |
into these somewhat uninteresting browns and greens.
link |
But every now and then there'd be something
link |
that had a really pure color.
link |
So there's just a few that I remember.
link |
There was a Here, There and Everywhere by the Beatles
link |
was bright red and has this very distinctive tonality
link |
and it's called structure at the beginning.
link |
So that was bright red.
link |
There was this song by the Alan Parsons Project
link |
called Ammonia Avenue that was kind of a pure, a pure blue.
link |
Anyway, I've got no idea how this happened.
link |
I didn't even pay that much attention
link |
until it went away when I was about 20.
link |
This synesthesia often goes away.
link |
So is it purely just the perception of a particular color
link |
or was there a positive or negative experience?
link |
Like was blue associated with a positive
link |
and red with a negative?
link |
Or is it simply the perception of color
link |
associated with some characteristic of the song?
link |
For me, I don't remember a lot of association
link |
with emotion or with value.
link |
It was just this kind of weird and interesting fact.
link |
I mean, at the beginning, I thought this was something
link |
that happened to everyone, songs of colors.
link |
Maybe I mentioned it once or twice and people said, nope.
link |
I thought it was kind of cool when there was one
link |
that had one of these especially pure colors,
link |
but only much later once I became a grad student
link |
thinking about the mind that I read about this phenomenon
link |
called synesthesia and I was like, hey, that's what I had.
link |
And now I occasionally talk about it in my classes,
link |
in intro class and it still happens sometimes.
link |
A student comes up and says, hey, I have that.
link |
I never knew about that.
link |
I never knew it had a name.
link |
You said that it went away at age 20 or so.
link |
And that you have a journal entry from around then saying,
link |
songs don't have colors anymore.
link |
Yeah, it was definitely sad that it was gone.
link |
In retrospect, it was like, hey, that's cool.
link |
The colors have gone.
link |
Yeah, can you think about that for a little bit?
link |
Do you miss those experiences?
link |
Because it's a fundamentally different set of experiences
link |
that you no longer have.
link |
Or is it just a nice thing to have had?
link |
You don't see them as that fundamentally different
link |
than you visiting a new country and experiencing
link |
I guess for me, when I had these experiences,
link |
they were somewhat marginal.
link |
They were like a little bonus kind of experience.
link |
I know there are people who have much more serious forms
link |
of synesthesia than this for whom it's absolutely central
link |
I know people who, when they experience new people,
link |
they have colors, maybe they have tastes and so on.
link |
Every time they see writing, it has colors.
link |
Some people, whenever they hear music,
link |
it's got a certain really rich color pattern.
link |
For some synesthetes, it's absolutely central.
link |
I think if they lost it, they'd be devastated.
link |
Again, for me, it was a very, very mild form
link |
of synesthesia, and it's like, yeah,
link |
it's like those interesting experiences
link |
you might get under different altered states
link |
of consciousness and so on.
link |
It's kind of cool, but not necessarily
link |
the single most important experiences in your life.
link |
So let's try to go to the very simplest question
link |
that you've answered many a time,
link |
but perhaps the simplest things can help us reveal,
link |
even in time, some new ideas.
link |
So what, in your view, is consciousness?
link |
What is the hard problem of consciousness?
link |
Consciousness, I mean, the word is used many ways,
link |
but the kind of consciousness that I'm interested in
link |
is basically subjective experience,
link |
what it feels like from the inside to be a human being
link |
or any other conscious being.
link |
I mean, there's something it's like to be me right now.
link |
I have visual images that I'm experiencing.
link |
I'm hearing my voice.
link |
I've got maybe some emotional tone.
link |
I've got a stream of thoughts running through my head.
link |
These are all things that I experience
link |
from the first person point of view.
link |
I've sometimes called this the inner movie in the mind.
link |
It's not a perfect metaphor.
link |
It's not like a movie in every way,
link |
and it's very rich.
link |
But yeah, it's just direct, subjective experience.
link |
And I call that consciousness,
link |
or sometimes philosophers use the word qualia,
link |
which you suggested.
link |
People tend to use the word qualia
link |
for things like the qualities of things like colors,
link |
redness, the experience of redness
link |
versus the experience of greenness,
link |
the experience of one taste or one smell versus another,
link |
the experience of the quality of pain.
link |
And yeah, a lot of consciousness
link |
is the experience of those qualities.
link |
Well, consciousness is bigger,
link |
the entirety of any kinds of experiences.
link |
Consciousness of thinking is not obviously qualia.
link |
It's not like specific qualities like redness or greenness,
link |
but still I'm thinking about my hometown.
link |
I'm thinking about what I'm gonna do later on.
link |
Maybe there's still something running through my head,
link |
which is subjective experience.
link |
Maybe it goes beyond those qualities or qualia.
link |
Philosophers sometimes use the word phenomenal consciousness
link |
for consciousness in this sense.
link |
I mean, people also talk about access consciousness,
link |
being able to access information in your mind,
link |
reflective consciousness,
link |
being able to think about yourself.
link |
But it looks like the really mysterious one,
link |
the one that really gets people going
link |
is phenomenal consciousness.
link |
The fact that there's subjective experience
link |
and all this feels like something at all.
link |
And then the hard problem is how is it that,
link |
why is it that there is phenomenal consciousness at all?
link |
And how is it that physical processes in a brain
link |
could give you subjective experience?
link |
It looks like on the face of it,
link |
you'd have all this big complicated physical system
link |
in a brain running without a given
link |
subjective experience at all.
link |
And yet we do have subjective experience.
link |
So the hard problem is just explain that.
link |
Explain how that comes about.
link |
We haven't been able to build machines
link |
where a red light goes on that says it's not conscious.
link |
So how do we actually create that?
link |
Or how do humans do it?
link |
And how do we ourselves do it?
link |
We do every now and then create machines that can do this.
link |
We create babies that are conscious.
link |
They've got these brains.
link |
That brain does produce consciousness.
link |
But even though we can create it,
link |
we still don't understand why it happens.
link |
Maybe eventually we'll be able to create machines,
link |
which as a matter of fact, AI machines,
link |
which as a matter of fact are conscious.
link |
But that won't necessarily make the hard problem go away
link |
any more than it does with babies.
link |
Cause we still wanna know how and why is it
link |
that these processes give you consciousness?
link |
You just made me realize for a second,
link |
maybe it's a totally dumb realization, but nevertheless,
link |
that as a useful way to think about
link |
the creation of consciousness is looking at a baby.
link |
So that there's a certain point
link |
at which that baby is not conscious.
link |
The baby starts from maybe, I don't know,
link |
from a few cells, right?
link |
There's a certain point at which it becomes consciousness,
link |
arrives, it's conscious.
link |
Of course, we can't know exactly that line,
link |
but that's a useful idea that we do create consciousness.
link |
Again, a really dumb thing for me to say,
link |
but not until now did I realize
link |
we do engineer consciousness.
link |
We get to watch the process happen.
link |
We don't know which point it happens or where it is,
link |
but we do see the birth of consciousness.
link |
Yeah, I mean, there's a question, of course,
link |
is whether babies are conscious when they're born.
link |
And it used to be, it seems,
link |
at least some people thought they weren't,
link |
which is why they didn't give anesthetics
link |
to newborn babies when they circumcised them.
link |
And so now people think, oh, that would be incredibly cruel.
link |
Of course, babies feel pain.
link |
And now the dominant view is that the babies can feel pain.
link |
Actually, my partner Claudia works on this whole issue
link |
of whether there's consciousness in babies
link |
And she certainly thinks that newborn babies
link |
come into the world with some degree of consciousness.
link |
Of course, then you can just extend the question backwards
link |
to fetuses and suddenly you're into
link |
politically controversial territory.
link |
But the question also arises in the animal kingdom.
link |
Where does consciousness start or stop?
link |
Is there a line in the animal kingdom
link |
where the first conscious organisms are?
link |
It's interesting, over time,
link |
people are becoming more and more liberal
link |
about ascribing consciousness to animals.
link |
People used to think maybe only mammals could be conscious.
link |
Now most people seem to think, sure, fish are conscious.
link |
They can feel pain.
link |
And now we're arguing over insects.
link |
You'll find people out there who say plants
link |
have some degree of consciousness.
link |
So, you know, who knows where it's gonna end.
link |
The far end of this chain is the view
link |
that every physical system has some degree of consciousness.
link |
Philosophers call that panpsychism.
link |
You know, I take that view.
link |
I mean, that's a fascinating way to view reality.
link |
So if you could talk about,
link |
if you can linger on panpsychism for a little bit,
link |
what does it mean?
link |
So it's not just plants are conscious.
link |
I mean, it's that consciousness
link |
is a fundamental fabric of reality.
link |
What does that mean to you?
link |
How are we supposed to think about that?
link |
Well, we're used to the idea that some things in the world
link |
are fundamental, right, in physics.
link |
We take things like space or time or space time,
link |
mass, charges, fundamental properties of the universe.
link |
You don't reduce them to something simpler.
link |
You take those for granted.
link |
You've got some laws that connect them.
link |
Here is how mass and space and time evolve.
link |
Theories like relativity or quantum mechanics
link |
or some future theory that will unify them both.
link |
But everyone says you gotta take some things as fundamental.
link |
And if you can't explain one thing,
link |
in terms of the previous fundamental things,
link |
you have to expand.
link |
Maybe something like this happened with Maxwell.
link |
He ended up with fundamental principles
link |
of electromagnetism and took charge as fundamental
link |
because it turned out that was the best way to explain it.
link |
So I at least take seriously the possibility
link |
something like that could happen with consciousness.
link |
Take it as a fundamental property,
link |
like space, time, and mass.
link |
And instead of trying to explain consciousness wholly
link |
in terms of the evolution of space, time, and mass,
link |
and so on, take it as a primitive
link |
and then connect it to everything else
link |
by some fundamental laws.
link |
Because there's this basic problem
link |
that the physics we have now looks great
link |
for solving the easy problems of consciousness,
link |
which are all about behavior.
link |
They give us a complicated structure and dynamics.
link |
They tell us how things are gonna behave,
link |
what kind of observable behavior they'll produce,
link |
which is great for the problems of explaining how we walk
link |
and how we talk and so on.
link |
Those are the easy problems of consciousness.
link |
But the hard problem was this problem
link |
about subjective experience just doesn't look
link |
like that kind of problem about structure,
link |
dynamics, how things behave.
link |
So it's hard to see how existing physics
link |
is gonna give you a full explanation of that.
link |
Certainly trying to get a physics view of consciousness,
link |
yes, there has to be a connecting point
link |
and it could be at the very axiomatic
link |
at the very beginning level.
link |
But first of all, there's a crazy idea
link |
that sort of everything has properties of consciousness.
link |
At that point, the word consciousness
link |
is already beyond the reach of our current understanding.
link |
Like far, because it's so far from,
link |
at least for me, maybe you can correct me,
link |
as far from the experiences that I have as a human being.
link |
To say that everything is conscious,
link |
that means that basically another way to put that,
link |
if that's true, then we understand almost nothing
link |
about that fundamental aspect of the world.
link |
How do you feel about saying an ant is conscious?
link |
Do you get the same reaction to that
link |
or is that something you can understand?
link |
I can understand ant,
link |
I can understand an atom, a particle.
link |
Plant, so I'm comfortable with living things on Earth
link |
being conscious because there's some kind of agency
link |
where they're similar size to me
link |
and they can be born and they can die.
link |
And that is understandable intuitively.
link |
Of course, you anthropomorphize,
link |
you put yourself in the place of the plant,
link |
but I can understand it.
link |
I mean, I'm not like, I don't believe actually
link |
that plants are conscious or that plants suffer,
link |
but I can understand that kind of belief, that kind of idea.
link |
How do you feel about robots?
link |
Like the kind of robots we have now?
link |
If I told you like that a Roomba
link |
had some degree of consciousness
link |
or some deep neural network.
link |
I could understand that a Roomba has consciousness.
link |
I just had spent all day at I, robot.
link |
And I mean, I personally love robots
link |
and I have a deep connection with robots.
link |
So I can, I also probably anthropomorphize them.
link |
There's something about the physical object.
link |
So there's a difference than a neural network,
link |
a neural network running a software.
link |
To me, the physical object,
link |
something about the human experience
link |
allows me to really see that physical object as an entity.
link |
And if it moves and moves in a way that it,
link |
there's a, like I didn't program it,
link |
where it feels that it's acting based on its own perception.
link |
And yes, self awareness and consciousness,
link |
even if it's a Roomba,
link |
then you start to assign it some agency, some consciousness.
link |
So, but to say that panpsychism,
link |
that consciousness is a fundamental property of reality
link |
is a much bigger statement.
link |
That it's like turtles all the way.
link |
It's like every, it's, it doesn't end.
link |
The whole thing is, so like how,
link |
I know it's full of mystery,
link |
but if you can linger on it,
link |
like how would it, how do you think about reality
link |
if consciousness is a fundamental part of its fabric?
link |
The way you get there is from thinking,
link |
can we explain consciousness given the existing fundamentals?
link |
And then if you can't, as at least right now, it looks like,
link |
then you've got to add something.
link |
It doesn't follow that you have to add consciousness.
link |
Here's another interesting possibility is,
link |
well, we'll add something else.
link |
Let's call it proto consciousness or X.
link |
And then it turns out space, time, mass plus X
link |
will somehow collectively give you the possibility
link |
for consciousness.
link |
Why don't rule out that view?
link |
Either I call that pan proto psychism,
link |
because maybe there's some other property,
link |
proto consciousness at the bottom level.
link |
And if you can't imagine there's actually
link |
genuine consciousness at the bottom level,
link |
I think we should be open to the idea
link |
there's this other thing X.
link |
Maybe we can't imagine that somehow gives you consciousness.
link |
But if we are playing along with the idea
link |
that there really is genuine consciousness
link |
at the bottom level, of course,
link |
this is going to be way out and speculative,
link |
but at least in, say, if it was classical physics,
link |
then we'd have to, you'd end up saying,
link |
well, every little atom, every little,
link |
with a bunch of particles in space time,
link |
each of these particles has some kind of consciousness
link |
whose structure mirrors maybe their physical properties,
link |
like its mass, its charge, its velocity, and so on.
link |
The structure of its consciousness
link |
would roughly correspond to that.
link |
And the physical interactions between particles,
link |
I mean, there's this old worry about physics.
link |
I mentioned this before in this issue
link |
about the manifest image.
link |
We don't really find out
link |
about the intrinsic nature of things.
link |
Physics tells us about how a particle relates
link |
to other particles and interacts.
link |
It doesn't tell us about what the particle is in itself.
link |
That was Kant's thing in itself.
link |
The nature in itself of a particle is something mental.
link |
A particle is actually a conscious,
link |
a little conscious subject
link |
with properties of its consciousness
link |
that correspond to its physical properties.
link |
The laws of physics are actually ultimately relating
link |
these properties of conscious subjects.
link |
So in this view, a Newtonian world
link |
actually would be a vast collection
link |
of little conscious subjects at the bottom level,
link |
way, way simpler than we are without free will
link |
or rationality or anything like that.
link |
But that's what the universe would be like.
link |
Now, of course, that's a vastly speculative view.
link |
No particular reason to think it's correct.
link |
Furthermore, non Newtonian physics,
link |
say quantum mechanical wave function,
link |
suddenly it starts to look different.
link |
It's not a vast collection of conscious subjects.
link |
Maybe there's ultimately one big wave function
link |
for the whole universe.
link |
Corresponding to that might be something more
link |
like a single conscious mind
link |
whose structure corresponds
link |
to the structure of the wave function.
link |
People sometimes call this cosmo psychism.
link |
And now, of course, we're in the realm
link |
of extremely speculative philosophy.
link |
There's no direct evidence for this,
link |
but yeah, but if you want a picture
link |
of what that universe would be like,
link |
think, yeah, giant cosmic mind
link |
with enough richness and structure among it
link |
to replicate all the structure of physics.
link |
I think therefore I am at the level of particles
link |
and with quantum mechanics
link |
at the level of the wave function.
link |
It's kind of an exciting, beautiful possibility,
link |
of course, way out of reach of physics currently.
link |
It is interesting that some neuroscientists
link |
are beginning to take panpsychism seriously,
link |
that you find consciousness even in very simple systems.
link |
So for example, the integrated information theory
link |
of consciousness, a lot of neuroscientists
link |
are taking seriously.
link |
Actually, I just got this new book
link |
by Christoph Koch just came in,
link |
The Feeling of Life Itself,
link |
why consciousness is widespread, but can't be computed.
link |
He likes, he basically endorses a panpsychist view
link |
where you get consciousness
link |
with the degree of information processing
link |
or integrated information processing in a simple,
link |
in a system and even very, very simple systems,
link |
like a couple of particles will have some degree of this.
link |
So he ends up with some degree of consciousness
link |
And the claim is that this theory
link |
can actually explain a bunch of stuff
link |
about the connection between the brain and consciousness.
link |
Now, that's very controversial.
link |
I think it's very, very early days
link |
in the science of consciousness.
link |
It's interesting that it's not just philosophy
link |
that might lead you in this direction,
link |
but there are ways of thinking quasi scientifically
link |
that lead you there too.
link |
But maybe it's different than panpsychism.
link |
What do you think?
link |
So Alan Watts has this quote that I'd like to ask you about.
link |
The quote is, through our eyes,
link |
the universe is perceiving itself.
link |
Through our ears, the universe is listening
link |
We are the witnesses through which the universe
link |
becomes conscious of its glory, of its magnificence.
link |
So that's not panpsychism.
link |
Do you think that we are essentially the tools,
link |
the senses the universe created to be conscious of itself?
link |
It's an interesting idea.
link |
Of course, if you went for the giant cosmic mind view,
link |
then the universe was conscious all along.
link |
It didn't need us.
link |
We're just little components of the universal consciousness.
link |
Likewise, if you believe in panpsychism,
link |
then there was some little degree of consciousness
link |
at the bottom level all along.
link |
And we were just a more complex form of consciousness.
link |
So I think maybe the quote you mentioned works better.
link |
If you're not a panpsychist, you're not a cosmo psychist,
link |
you think consciousness just exists
link |
at this intermediate level.
link |
And of course, that's the Orthodox view.
link |
That you would say is the common view?
link |
So is your own view with panpsychism a rare view?
link |
I think it's generally regarded certainly
link |
as a speculative view held by a fairly small minority
link |
of at least theorists, most philosophers
link |
and most scientists who think about consciousness
link |
are not panpsychists.
link |
There's been a bit of a movement in that direction
link |
the last 10 years or so.
link |
It seems to be quite popular,
link |
especially among the younger generation,
link |
but it's still very definitely a minority view.
link |
Many people think it's totally batshit crazy
link |
to use the technical term.
link |
But the philosophical term.
link |
So the Orthodox view, I think is still consciousness
link |
is something that humans have
link |
and some good number of nonhuman animals have,
link |
and maybe AIs might have one day, but it's restricted.
link |
On that view, then there was no consciousness
link |
at the start of the universe.
link |
There may be none at the end,
link |
but it is this thing which happened at some point
link |
in the history of the universe, consciousness developed.
link |
And yes, that's a very amazing event on this view
link |
because many people are inclined to think consciousness
link |
is what somehow gives meaning to our lives.
link |
Without consciousness, there'd be no meaning,
link |
no true value, no good versus bad and so on.
link |
So with the advent of consciousness,
link |
suddenly the universe went from meaningless
link |
to somehow meaningful.
link |
Why did this happen?
link |
I guess the quote you mentioned was somehow,
link |
this was somehow destined to happen
link |
because the universe needed to have consciousness
link |
within it to have value and have meaning.
link |
And maybe you could combine that with a theistic view
link |
or a teleological view.
link |
The universe was inexorably evolving towards consciousness.
link |
Actually, my colleague here at NYU, Tom Nagel,
link |
wrote a book called Mind and Cosmos a few years ago
link |
where he argued for this teleological view
link |
of evolution toward consciousness,
link |
saying this led the problems for Darwinism.
link |
It's got him on, this is very, very controversial.
link |
Most people didn't agree.
link |
I don't myself agree with this teleological view,
link |
but it is at least a beautiful speculative view
link |
What do you think people experience?
link |
What do they seek when they believe in God
link |
from this kind of perspective?
link |
I'm not an expert on thinking about God and religion.
link |
I'm not myself religious at all.
link |
When people sort of pray, communicate with God,
link |
which whatever form,
link |
I'm not speaking to sort of the practices
link |
and the rituals of religion.
link |
I mean the actual experience of that people
link |
really have a deep connection with God in some cases.
link |
What do you think that experience is?
link |
It's so common, at least throughout the history
link |
of civilization, that it seems like we seek that.
link |
At the very least, it is an interesting
link |
conscious experience that people have
link |
when they experience religious awe or prayer and so on.
link |
Neuroscientists have tried to examine
link |
what bits of the brain are active and so on.
link |
But yeah, there's this deeper question
link |
of what are people looking for when they're doing this?
link |
And like I said, I've got no real expertise on this,
link |
but it does seem that one thing people are after
link |
is a sense of meaning and value,
link |
a sense of connection to something greater than themselves
link |
that will give their lives meaning and value.
link |
And maybe the thought is if there is a God,
link |
then God somehow is a universal consciousness
link |
who has invested this universe with meaning
link |
and somehow connection to God might give your life meaning.
link |
I guess I can kind of see the attractions of that,
link |
but it still makes me wonder why is it exactly
link |
that a universal consciousness, God,
link |
would be needed to give the world meaning?
link |
If universal consciousness can give the world meaning,
link |
why can't local consciousness give the world meaning too?
link |
So I think my consciousness gives my world meaning.
link |
Is the origin of meaning for your world.
link |
Yeah, I experience things as good or bad,
link |
happy, sad, interesting, important.
link |
So my consciousness invests this world with meaning.
link |
Without any consciousness,
link |
maybe it would be a bleak, meaningless universe.
link |
But I don't see why I need someone else's consciousness
link |
or even God's consciousness to give this universe meaning.
link |
Here we are, local creatures
link |
with our own subjective experiences.
link |
I think we can give the universe meaning ourselves.
link |
I mean, maybe to some people that feels inadequate.
link |
Our own local consciousness is somehow too puny
link |
and insignificant to invest any of this
link |
with cosmic significance.
link |
And maybe God gives you a sense of cosmic significance,
link |
but I'm just speculating here.
link |
So it's a really interesting idea
link |
that consciousness is the thing that makes life meaningful.
link |
If you could maybe just briefly explore that for a second.
link |
So I suspect just from listening to you now,
link |
you mean in an almost trivial sense,
link |
just the day to day experiences of life have,
link |
because of you attach identity to it,
link |
they become, I guess I wanna ask something
link |
I would always wanted to ask
link |
a legit world renowned philosopher.
link |
What is the meaning of life?
link |
So I suspect you don't mean consciousness gives
link |
any kind of greater meaning to it all.
link |
And more to day to day.
link |
But is there a greater meaning to it all?
link |
I think life has meaning for us because we are conscious.
link |
So without consciousness, no meaning,
link |
consciousness invests our life with meaning.
link |
So consciousness is the source of the meaning of life,
link |
but I wouldn't say consciousness itself
link |
is the meaning of life.
link |
I'd say what's meaningful in life
link |
is basically what we find meaningful,
link |
what we experience as meaningful.
link |
So if you find meaning and fulfillment and value
link |
in say, intellectual work, like understanding,
link |
then that's a very significant part
link |
of the meaning of life for you.
link |
If you find that in social connections
link |
or in raising a family,
link |
then that's the meaning of life for you.
link |
The meaning kind of comes from what you value
link |
as a conscious creature.
link |
So I think there's no, on this view,
link |
there's no universal solution.
link |
No universal answer to the question,
link |
what is the meaning of life?
link |
The meaning of life is where you find it
link |
as a conscious creature,
link |
but it's consciousness that somehow makes value possible.
link |
Experiencing some things as good or as bad
link |
something comes from within consciousness.
link |
So you think consciousness is a crucial component,
link |
ingredient of assigning value to things?
link |
I mean, it's kind of a fairly strong intuition
link |
that without consciousness,
link |
there wouldn't really be any value
link |
if we just had a purely universe of unconscious creatures.
link |
Would anything be better or worse than anything else?
link |
Certainly when it comes to ethical dilemmas,
link |
you know about the old trolley problem.
link |
Do you kill one person
link |
or do you switch to the other track to kill five?
link |
Well, I've got a variant on this,
link |
the zombie trolley problem,
link |
where there's a one conscious being on one track
link |
and five humanoid zombies.
link |
Let's make them robots who are not conscious
link |
on the other track.
link |
Do you, given that choice,
link |
do you kill the one conscious being
link |
or the five unconscious robots?
link |
Most people have a fairly clear intuition here.
link |
Kill the unconscious beings
link |
because they basically, they don't have a meaningful life.
link |
They're not really persons, conscious beings at all.
link |
We don't have good intuition
link |
about something like an unconscious being.
link |
So in philosophical terms, you referred to as a zombie.
link |
It's a useful thought experiment construction
link |
in philosophical terms, but we don't yet have them.
link |
So that's kind of what we may be able to create with robots.
link |
And I don't necessarily know what that even means.
link |
Yeah, they're merely hypothetical.
link |
For now, they're just a thought experiment.
link |
They may never be possible.
link |
I mean, the extreme case of a zombie
link |
is a being which is physically, functionally,
link |
behaviorally identical to me, but not conscious.
link |
I don't think that could ever be built in this universe.
link |
The question is just could we,
link |
does that hypothetically make sense?
link |
That's kind of a useful contrast class
link |
to raise questions like, why aren't we zombies?
link |
How does it come about that we're conscious?
link |
And we're not like that.
link |
But there are less extreme versions of this like robots,
link |
which are maybe not physically identical to us,
link |
maybe not even functionally identical to us.
link |
Maybe they've got a different architecture,
link |
but they can do a lot of sophisticated things,
link |
maybe carry on a conversation, but they're not conscious.
link |
And that's not so far out.
link |
We've got simple computer systems,
link |
at least tending in that direction now.
link |
And presumably this is gonna get more and more sophisticated
link |
over years to come where we may have some pretty,
link |
it's at least quite straightforward to conceive
link |
of some pretty sophisticated robot systems
link |
that can use language and be fairly high functioning
link |
without consciousness at all.
link |
Then I stipulate that.
link |
I mean, we've caused, there's this tricky question
link |
of how you would know whether they're conscious.
link |
But let's say we've somehow solved that.
link |
And we know that these high functioning robots
link |
Then the question is, do they have moral status?
link |
Does it matter how we treat them?
link |
What does moral status mean, sir?
link |
Basically it's that question.
link |
Does it matter how we treat them?
link |
For example, if I mistreat this glass, this cup
link |
by shattering it, then that's bad.
link |
Why is it bad though?
link |
It's gonna make a mess.
link |
It's gonna be annoying for me and my partner.
link |
And so it's not bad for the cup.
link |
No one would say the cup itself has moral status.
link |
Hey, you hurt the cup and that's doing it a moral harm.
link |
Likewise, plants, well, again, if they're not conscious,
link |
most people think by uprooting a plant,
link |
you're not harming it.
link |
But if a being is conscious on the other hand,
link |
then you are harming it.
link |
So Siri, or I dare not say the name of Alexa.
link |
Anyway, so we don't think we're morally harming Alexa
link |
by turning her off or disconnecting her
link |
or even destroying her, whether it's the system
link |
or the underlying software system,
link |
because we don't really think she's conscious.
link |
On the other hand, you move to like the disembodied being
link |
in the movie, her, Samantha,
link |
I guess she was kind of presented as conscious.
link |
And then if you destroyed her,
link |
you'd certainly be committing a serious harm.
link |
So I think our strong sense is if a being is conscious
link |
and can undergo subjective experiences,
link |
then it matters morally how we treat them.
link |
So if a robot is conscious, it matters,
link |
but if a robot is not conscious,
link |
then they're basically just meat or a machine
link |
and it doesn't matter.
link |
So I think at least maybe how we think about this stuff
link |
is fundamentally wrong,
link |
but I think a lot of people
link |
who think about this stuff seriously,
link |
including people who think about,
link |
say the moral treatment of animals and so on,
link |
come to the view that consciousness
link |
is ultimately kind of the line between systems
link |
that where we have to take them into account
link |
and thinking morally about how we act
link |
and systems for which we don't.
link |
And I think I've seen you the writer talk about
link |
the demonstration of consciousness from a system like that,
link |
from a system like Alexa or a conversational agent
link |
that what you would be looking for
link |
is kind of at the very basic level
link |
for the system to have an awareness
link |
that I'm just a program
link |
and yet, why do I experience this?
link |
Or not to have that experience,
link |
but to communicate that to you.
link |
So that's what us humans would sound like.
link |
If you all of a sudden woke up one day,
link |
like Kafka, right, in a body of a bug or something,
link |
but in a computer, you all of a sudden realized
link |
you don't have a body
link |
and yet you were feeling what you were feeling,
link |
you would probably say those kinds of things.
link |
So do you think a system essentially becomes conscious
link |
by convincing us that it's conscious
link |
through the words that I just mentioned?
link |
So by being confused about the fact
link |
that why am I having these experiences?
link |
I don't think this is what makes you conscious,
link |
but I do think being puzzled about consciousness
link |
is a very good sign that a system is conscious.
link |
So if I encountered a robot
link |
that actually seemed to be genuinely puzzled
link |
by its own mental states
link |
and saying, yeah, I have all these weird experiences
link |
and I don't see how to explain them.
link |
I know I'm just a set of silicon circuits,
link |
but I don't see how that would give you my consciousness.
link |
I would at least take that as some evidence
link |
that there's some consciousness going on there.
link |
I don't think a system needs to be puzzled
link |
about consciousness to be conscious.
link |
Many people aren't puzzled by their consciousness.
link |
Animals don't seem to be puzzled at all.
link |
I still think they're conscious.
link |
So I don't think that's a requirement on consciousness,
link |
but I do think if we're looking for signs
link |
for consciousness, say in AI systems,
link |
one of the things that will help convince me
link |
that an AI system is conscious is if it shows signs of,
link |
if it shows signs of introspectively recognizing something
link |
like consciousness and finding this philosophically puzzling
link |
in the way that we do.
link |
It's such an interesting thought, though,
link |
because a lot of people sort of would,
link |
at the Shao level, criticize the Turing test for language.
link |
It's essentially what I heard Dan Dennett
link |
criticize it in this kind of way,
link |
which is it really puts a lot of emphasis on lying.
link |
Yeah, and being able to imitate
link |
human beings, yeah, there's this cartoon
link |
of the AI system studying for the Turing test.
link |
It's gotta read this book called Talk Like a Human.
link |
It's like, man, why do I have to waste my time
link |
learning how to imitate humans?
link |
Maybe the AI system is gonna be way beyond
link |
the hard problem of consciousness,
link |
and it's gonna be just like,
link |
why do I need to waste my time pretending
link |
that I recognize the hard problem of consciousness
link |
in order for people to recognize me as conscious?
link |
Yeah, it just feels like, I guess the question is,
link |
do you think we can ever really create
link |
a test for consciousness?
link |
Because it feels like we're very human centric,
link |
and so the only way we would be convinced
link |
that something is conscious is basically
link |
the thing demonstrates the illusion of consciousness,
link |
that we can never really know whether it's conscious or not,
link |
and in fact, that almost feels like it doesn't matter then,
link |
or does it still matter to you that something is conscious
link |
or it demonstrates consciousness?
link |
You still see that fundamental distinction.
link |
I think to a lot of people,
link |
whether a system is conscious or not
link |
matters hugely for many things,
link |
like how we treat it, can it suffer, and so on,
link |
but still, that leaves open the question,
link |
how can we ever know?
link |
And it's true that it's awfully hard
link |
to see how we can know for sure
link |
whether a system is conscious.
link |
I suspect that sociologically,
link |
the thing that's gonna convince us
link |
that a system is conscious is, in part,
link |
things like social interaction, conversation, and so on,
link |
where they seem to be conscious,
link |
they talk about their conscious states
link |
or just talk about being happy or sad
link |
or finding things meaningful or being in pain.
link |
That will tend to convince us if we don't,
link |
if a system genuinely seems to be conscious,
link |
we don't treat it as such,
link |
eventually it's gonna seem like a strange form
link |
of racism or speciesism or somehow,
link |
not to acknowledge them as conscious.
link |
I truly believe that, by the way.
link |
I believe that there is going to be
link |
something akin to the Civil Rights Movement,
link |
I think the moment you have a Roomba say,
link |
please don't kick me, that hurts, just say it.
link |
I think that will fundamentally change
link |
the fabric of our society.
link |
I think you're probably right,
link |
although it's gonna be very tricky
link |
because, just say we've got the technology
link |
where these conscious beings can just be created
link |
and multiplied by the thousands by flicking a switch.
link |
The legal status is gonna be different,
link |
but ultimately their moral status ought to be the same,
link |
and yeah, the civil rights issue is gonna be a huge mess.
link |
So if one day somebody clones you,
link |
another very real possibility.
link |
In fact, I find the conversation between
link |
two copies of David Chalmers quite interesting.
link |
Who is this idiot?
link |
He's not making any sense.
link |
So what, do you think he would be conscious?
link |
I do think he would be conscious.
link |
I do think in some sense,
link |
I'm not sure it would be me,
link |
there would be two different beings at this point.
link |
I think they'd both be conscious
link |
and they both have many of the same mental properties.
link |
I think they both in a way have the same moral status.
link |
It'd be wrong to hurt either of them
link |
or to kill them and so on.
link |
Still, there's some sense in which probably
link |
their legal status would have to be different.
link |
If I'm the original and that one's just a clone,
link |
then creating a clone of me,
link |
presumably the clone doesn't, for example,
link |
automatically own the stuff that I own
link |
or I've got a certain connect,
link |
the things that the people I interact with,
link |
my family, my partner and so on,
link |
I'm gonna somehow be connected to them
link |
in a way in which the clone isn't, so.
link |
Because you came slightly first?
link |
Because a clone would argue that they have
link |
really as much of a connection.
link |
They have all the memories of that connection.
link |
Then a way you might say it's kind of unfair
link |
to discriminate against them,
link |
but say you've got an apartment
link |
that only one person can live in
link |
or a partner who only one person can be with.
link |
But why should it be you, the original?
link |
It's an interesting philosophical question,
link |
but you might say because I actually have this history,
link |
if I am the same person as the one that came before
link |
and the clone is not,
link |
then I have this history that the clone doesn't.
link |
Of course, there's also the question,
link |
isn't the clone the same person too?
link |
This is a question about personal identity.
link |
If I continue and I create a clone over there,
link |
I wanna say this one is me and this one is someone else.
link |
But you could take the view that a clone is equally me.
link |
Of course, in a movie like Star Trek
link |
where they have a teletransporter
link |
basically creates clones all the time.
link |
They treat the clones as if they're the original person.
link |
Of course, they destroy the original body in Star Trek.
link |
So there's only one left around
link |
and only very occasionally do things go wrong
link |
and you get two copies of Captain Kirk.
link |
But somehow our legal system at the very least
link |
is gonna have to sort out some of these issues
link |
and that maybe that's what's moral
link |
and what's legally acceptable are gonna come apart.
link |
What question would you ask a clone of yourself?
link |
Is there something useful you can find out from him
link |
about the fundamentals of consciousness even?
link |
I mean, kind of in principle,
link |
I know that if it's a perfect clone,
link |
it's gonna behave just like me.
link |
So I'm not sure I'm gonna be able to,
link |
I can discover whether it's a perfect clone
link |
by seeing whether it answers like me.
link |
But otherwise I know what I'm gonna find is a being
link |
which is just like me,
link |
except that it's just undergone this great shock
link |
of discovering that it's a clone.
link |
So just say you woke me up tomorrow and said,
link |
hey Dave, sorry to tell you this,
link |
but you're actually the clone
link |
and you provided me really convincing evidence,
link |
showed me the film of my being cloned
link |
and then all wrapped in here being here and waking up.
link |
So you proved to me I'm a clone,
link |
well, yeah, okay, I would find that shocking
link |
and who knows how I would react to this.
link |
So maybe by talking to the clone,
link |
I'd find something about my own psychology
link |
that I can't find out so easily,
link |
like how I'd react upon discovering that I'm a clone.
link |
I could certainly ask the clone if it's conscious
link |
and what his consciousness is like and so on,
link |
but I guess I kind of know if it's a perfect clone,
link |
it's gonna behave roughly like me.
link |
Of course, at the beginning,
link |
there'll be a question
link |
about whether a perfect clone is possible.
link |
So I may wanna ask it lots of questions
link |
to see if it's consciousness
link |
and the way it talks about its consciousness
link |
and the way it reacts to things in general is likely.
link |
And that will occupy us for a while.
link |
So basic unit testing on the early models.
link |
So if it's a perfect clone,
link |
you say that it's gonna behave exactly like you.
link |
So that takes us to free will.
link |
Is there free will?
link |
Are we able to make decisions that are not predetermined
link |
from the initial conditions of the universe?
link |
You know, philosophers do this annoying thing
link |
of saying it depends what you mean.
link |
So in this case, yeah, it really depends on what you mean,
link |
If you mean something which was not determined in advance,
link |
could never have been determined,
link |
then I don't know we have free will.
link |
I mean, there's quantum mechanics
link |
and who's to say if that opens up some room,
link |
but I'm not sure we have free will in that sense.
link |
But I'm also not sure that's the kind of free will
link |
that really matters.
link |
You know, what matters to us
link |
is being able to do what we want
link |
and to create our own futures.
link |
We've got this distinction between having our lives
link |
be under our control and under someone else's control.
link |
We've got the sense of actions that we are responsible for
link |
versus ones that we're not.
link |
I think you can make those distinctions
link |
even in a deterministic universe.
link |
And this is what people call the compatibilist view
link |
of free will, where it's compatible with determinism.
link |
So I think for many purposes,
link |
the kind of free will that matters
link |
is something we can have in a deterministic universe.
link |
And I can't see any reason in principle
link |
why an AI system couldn't have free will of that kind.
link |
If you mean super duper free will,
link |
the ability to violate the laws of physics
link |
and doing things that in principle could not be predicted.
link |
I don't know, maybe no one has that kind of free will.
link |
What's the connection between the reality of free will
link |
and the experience of it,
link |
the subjective experience in your view?
link |
So how does consciousness connect
link |
to the reality and the experience of free will?
link |
It's certainly true that when we make decisions
link |
and when we choose and so on,
link |
we feel like we have an open future.
link |
Feel like I could do this, I could go into philosophy
link |
or I could go into math, I could go to a movie tonight,
link |
I could go to a restaurant.
link |
So we experience these things as if the future is open.
link |
And maybe we experience ourselves
link |
as exerting a kind of effect on the future
link |
that somehow picking out one path
link |
from many paths were previously open.
link |
And you might think that actually
link |
if we're in a deterministic universe,
link |
there's a sense of which objectively
link |
those paths weren't really open all along,
link |
but subjectively they were open.
link |
And that's, I think that's what really matters
link |
in making a decisions where our experience
link |
of making a decision is choosing a path for ourselves.
link |
I mean, in general, our introspective models of the mind,
link |
I think are generally very distorted representations
link |
So it may well be that our experience of ourself
link |
in making a decision, our experience of what's going on
link |
doesn't terribly well mirror what's going on.
link |
I mean, maybe there are antecedents in the brain
link |
way before anything came into consciousness
link |
Those aren't represented in our introspective model.
link |
So in general, our experience of perception,
link |
so I experience a perceptual image of the external world.
link |
It's not a terribly good model of what's actually going on
link |
in my visual cortex and so on,
link |
which has all these layers and so on.
link |
It's just one little snapshot of one bit of that.
link |
So in general, introspective models
link |
are very over oversimplified.
link |
And it wouldn't be surprising
link |
if that was true of free will as well.
link |
This also incidentally can be applied to consciousness itself.
link |
There is this very interesting view
link |
that consciousness itself is an introspective illusion.
link |
In fact, we're not conscious,
link |
but the brain just has these introspective models of itself
link |
or oversimplifies everything and represents itself
link |
as having these special properties of consciousness.
link |
It's a really simple way to kind of keep track of itself
link |
And then on the illusionist view,
link |
yeah, that's just an illusion.
link |
I find this view, when I find it implausible,
link |
I do find it very attractive in some ways,
link |
because it's easy to tell some story
link |
about how the brain would create introspective models
link |
of its own consciousness, of its own free will
link |
as a way of simplifying itself.
link |
I mean, it's a similar way when we perceive
link |
the external world, we perceive it as having these colors
link |
that maybe it doesn't really have,
link |
but of course that's a really useful way
link |
of keeping tracks, of keeping track.
link |
Did you say that you find it not very plausible?
link |
Because I find it both plausible
link |
and attractive in some sense,
link |
because I mean, that kind of view
link |
is one that has the minimum amount of mystery around it.
link |
You can kind of understand that kind of view.
link |
Everything else says we don't understand
link |
so much of this picture.
link |
No, it is very attractive, I recently wrote an article
link |
about this kind of issue called
link |
the meta problem of consciousness.
link |
The hard problem is how does a brain
link |
give you consciousness?
link |
The meta problem is why are we puzzled
link |
by the hard problem of consciousness?
link |
Because being puzzled by it,
link |
that's ultimately a bit of behavior.
link |
We might be able to explain that bit of behavior
link |
as one of the easy problems, consciousness.
link |
So maybe there'll be some computational model
link |
that explains why we're puzzled by consciousness.
link |
The meta problem has come up with that model.
link |
And I've been thinking about that a lot lately.
link |
There's some interesting stories you can tell
link |
about why the right kind of computational system
link |
might develop these introspective models of itself
link |
that attributed itself, these special properties.
link |
So that meta problem is a research program for everyone.
link |
And then if you've got attraction
link |
to sort of simple views, desert landscapes and so on,
link |
then you can go all the way
link |
with what people call illusionism
link |
and say, in fact, consciousness itself is not real.
link |
What is real is just these introspective models
link |
we have that tell us that we're conscious.
link |
So the view is very simple, very attractive, very powerful.
link |
The trouble is, of course, it has to say
link |
that deep down, consciousness is not real.
link |
We're not actually experiencing right now.
link |
And it looks like it's just contradicting
link |
a fundamental datum of our existence.
link |
And this is why most people find this view crazy.
link |
Just as they find panpsychism crazy in one way,
link |
people find illusionism crazy in another way.
link |
But I mean, so yes, it has to deny
link |
this fundamental datum of our existence.
link |
Now, that makes the view sort of frankly unbelievable
link |
On the other hand, the view developed right
link |
might be able to explain why we find it unbelievable.
link |
Because these models are so deeply hardwired into our head.
link |
And they're all integrated.
link |
You can't escape the illusion.
link |
And it's a crazy possibility.
link |
Is it possible that the entirety of the universe,
link |
our planet, all the people in New York,
link |
all the organisms on our planet,
link |
including me here today, are not real in that sense?
link |
They're all part of an illusion inside of Dave Chalmers's head.
link |
I think all this could be a simulation.
link |
No, but not just a simulation.
link |
Because the simulation kind of is outside of you.
link |
What if it's all an illusion?
link |
Yes, a dream that you're experiencing.
link |
That's, it's all in your mind, right?
link |
Is that, can you take illusionism that far?
link |
Well, there's illusionism about the external world
link |
and illusionism about consciousness.
link |
And these might go in different.
link |
Illusionism about the external world
link |
kind of takes you back to Descartes.
link |
And yeah, could all this be produced by an evil demon?
link |
Descartes himself also had the dream argument.
link |
He said, how do you know you're not dreaming right now?
link |
How do you know this is not an amazing dream?
link |
And it's at least a possibility that yeah,
link |
this could be some super duper complex dream
link |
in the next universe up.
link |
I guess though, my attitude is that just as,
link |
when Descartes thought that if the evil demon was doing it,
link |
A lot of people these days say if a simulation is doing it,
link |
As I was saying before, I think even if it's a simulation,
link |
that doesn't stop this from being real.
link |
It just tells us what the world is made of.
link |
Likewise, if it's a dream,
link |
it could turn out that all this is like my dream
link |
created by my brain in the next universe up.
link |
My own view is that wouldn't stop this physical world
link |
It would turn out this cup at the most fundamental level
link |
was made of a bit of say my consciousness
link |
in the dreaming mind at the next level up.
link |
Maybe that would give you a kind of weird kind of panpsychism
link |
about reality, but it wouldn't show that the cup isn't real.
link |
It would just tell us it's ultimately made of processes
link |
in my dreaming mind.
link |
So I'd resist the idea that if the physical world is a dream,
link |
then it's an illusion.
link |
By the way, perhaps you have an interesting thought
link |
Why is Descartes demon or genius considered evil?
link |
Why couldn't have been a benevolent one
link |
that had the same powers?
link |
Yeah, I mean, Descartes called it the malign genie,
link |
the evil genie or evil genius.
link |
Malign, I guess was the word.
link |
But yeah, it's an interesting question.
link |
I mean, a later philosophy, Barclay said,
link |
no, in fact, all this is done by God.
link |
God actually supplies you all of these perceptions
link |
and ideas and that's how physical reality is sustained.
link |
And interestingly, Barclay's God is doing something
link |
that doesn't look so different
link |
from what Descartes evil demon was doing.
link |
It's just that Descartes thought it was deception
link |
and Barclay thought it was not.
link |
And I'm actually more sympathetic to Barclay here.
link |
Yeah, this evil demon may be trying to deceive you,
link |
but I think, okay, well, the evil demon
link |
may just be working under a false philosophical theory.
link |
It thinks it's deceiving you, it's wrong.
link |
It's like there's machines in the matrix.
link |
They thought they were deceiving you
link |
that all this stuff is real.
link |
I think, no, if we're in a matrix, it's all still real.
link |
Yeah, the philosopher O.K. Bousma had a nice story
link |
about this about 50 years ago, about Descartes evil demon,
link |
where he said this demon spends all its time
link |
trying to fool people, but fails
link |
because somehow all the demon ends up doing
link |
is constructing realities for people.
link |
So yeah, I think that maybe it's a very natural
link |
to take this view that if we're in a simulation
link |
or evil demon scenario or something,
link |
then none of this is real.
link |
But I think it may be ultimately a philosophical mistake,
link |
especially if you take on board sort of the view of reality
link |
where what matters to reality is really its structure,
link |
something like its mathematical structure and so on,
link |
which seems to be the view that a lot of people take
link |
from contemporary physics.
link |
And it looks like you can find
link |
all that mathematical structure in a simulation,
link |
maybe even in a dream and so on.
link |
So as long as that structure is real,
link |
I would say that's enough for the physical world to be real.
link |
Yeah, the physical world may turn out
link |
to be somewhat more intangible than we had thought
link |
and have a surprising nature of it.
link |
We're already gotten very used to that from modern science.
link |
See, you've kind of alluded
link |
that you don't have to have consciousness
link |
for high levels of intelligence,
link |
but to create truly general intelligence systems,
link |
AGI systems at human level intelligence
link |
and perhaps super human level intelligence,
link |
you've talked about that you feel like
link |
that kind of thing might be very far away,
link |
but nevertheless, when we reached that point,
link |
do you think consciousness
link |
from an engineering perspective is needed
link |
or at least highly beneficial for creating an AGI system?
link |
Yeah, no one knows what consciousness is for functionally.
link |
So right now there's no specific thing we can point to
link |
and say, you need consciousness for that.
link |
So my inclination is to believe
link |
that in principle AGI is possible.
link |
The very least I don't see why
link |
someone couldn't simulate a brain,
link |
ultimately have a computational system
link |
that produces all of our behavior.
link |
And if that's possible,
link |
I'm sure vastly many other computational systems
link |
of equal or greater sophistication are possible
link |
with all of our cognitive functions and more.
link |
My inclination is to think that
link |
once you've got all these cognitive functions,
link |
perception, attention, reasoning,
link |
introspection, language, emotion, and so on,
link |
it's very likely you'll have consciousness as well.
link |
So at least it's very hard for me to see
link |
how you'd have a system that had all those things
link |
while bypassing somehow conscious.
link |
So just naturally it's integrated quite naturally.
link |
There's a lot of overlap about the kind of function
link |
that required to achieve each of those things
link |
that's, so you can't disentangle them
link |
even when you're recreating.
link |
It seems to, at least in us,
link |
but we don't know what the causal role of consciousness
link |
in the physical world, what it does.
link |
I mean, just say it turns out
link |
consciousness does something very specific
link |
in the physical world like collapsing wave functions
link |
as on one common interpretation of quantum mechanics.
link |
Then ultimately we might find some place
link |
where it actually makes a difference
link |
and we could say, ah,
link |
here is where in collapsing wave functions
link |
it's driving the behavior of a system.
link |
And maybe it could even turn out that for AGI,
link |
you'd need something playing that.
link |
I mean, if you wanted to connect this to free will,
link |
some people think consciousness collapsing wave functions,
link |
that would be how the conscious mind exerts effect
link |
on the physical world and exerts its free will.
link |
And maybe it could turn out that any AGI
link |
that didn't utilize that mechanism would be limited
link |
in the kinds of functionality that it had.
link |
I don't myself find that plausible.
link |
I think probably that functionality could be simulated.
link |
But you can imagine once we had a very specific idea
link |
about the role of consciousness in the physical world,
link |
this would have some impact on the capacity of AGI's.
link |
And if it was a role that could not be duplicated elsewhere,
link |
then we'd have to find some way to either
link |
get consciousness in the system to play that role
link |
or to simulate it.
link |
If we can isolate a particular role to consciousness,
link |
of course, it seems like an incredibly difficult thing.
link |
Do you have worries about existential threats
link |
of conscious intelligent beings that are not us?
link |
So certainly, I'm sure you're worried about us
link |
from an existential threat perspective,
link |
but outside of us, AI systems.
link |
There's a couple of different kinds
link |
of existential threats here.
link |
One is an existential threat to consciousness generally.
link |
I mean, yes, I care about humans
link |
and the survival of humans and so on,
link |
but just say it turns out that eventually we're replaced
link |
by some artificial beings that aren't humans,
link |
but are somehow our successors.
link |
They still have good lives.
link |
They still do interesting and wonderful things
link |
with the universe.
link |
I don't think that's not so bad.
link |
That's just our successors.
link |
We were one stage in evolution.
link |
Something different, maybe better came next.
link |
If on the other hand, all of consciousness was wiped out,
link |
that would be a very serious moral disaster.
link |
One way that could happen is by all intelligent life
link |
And many people think that, yeah,
link |
once you get to humans and AIs and amazing sophistication
link |
where everyone has got the ability to create weapons
link |
that can destroy the whole universe just by pressing a button,
link |
then maybe it's inevitable all intelligent life will die out.
link |
That would certainly be a disaster.
link |
And we've got to think very hard about how to avoid that.
link |
But yeah, another interesting kind of disaster
link |
is that maybe intelligent life is not wiped out,
link |
but all consciousness is wiped out.
link |
So just say your thought,
link |
unlike what I was saying a moment ago,
link |
that there are two different kinds of intelligent systems,
link |
some which are conscious and some which are not.
link |
And just say it turns out that we create AGI
link |
with a high degree of intelligence,
link |
meaning high degree of sophistication and its behavior,
link |
but with no consciousness at all.
link |
That AGI could take over the world maybe,
link |
but then there'd be no consciousness in this world.
link |
This would be a world of zombies.
link |
Some people have called this the zombie apocalypse
link |
because it's an apocalypse for consciousness.
link |
Consciousness is gone.
link |
You've merely got this super intelligent,
link |
nonconscious robots.
link |
And I would say that's a moral disaster in the same way,
link |
in almost the same way that the world
link |
with no intelligent life is a moral disaster.
link |
All value and meaning may be gone from that world.
link |
So these are both threats to watch out for.
link |
Now, my own view is if you get super intelligence,
link |
you're almost certainly gonna bring consciousness with it.
link |
So I hope that's not gonna happen.
link |
But of course, I don't understand consciousness.
link |
No one understands consciousness.
link |
This is one reason for,
link |
this is one reason at least among many
link |
for thinking very seriously about consciousness
link |
and thinking about the kind of future
link |
we want to create in a world with humans and or AIs.
link |
How do you feel about the possibility
link |
if consciousness so naturally does come with AGI systems
link |
that we are just a step in the evolution?
link |
That we will be just something, a blimp on the record
link |
that'll be studied in books
link |
by the AGI systems centuries from now?
link |
I mean, I think I'd probably be okay with that,
link |
especially if somehow humans are continuous with AGI.
link |
I mean, I think something like this is inevitable.
link |
The very least humans are gonna be transformed.
link |
We're gonna be augmented by technology.
link |
It's already happening in all kinds of ways.
link |
We're gonna be transformed by technology
link |
where our brains are gonna be uploaded
link |
and computationally enhanced.
link |
And eventually that line between what's a human
link |
and what's an AI may be kind of hard to draw.
link |
How much does it matter, for example,
link |
that some future being a thousand years from now
link |
that somehow descended from us actually still has biology?
link |
I think it would be nice if you kind of point
link |
to its cognitive system, point to some parts
link |
that had some roots in us and trace a continuous line there.
link |
That would be selfishly nice for me to think that,
link |
okay, I'm connected to this thread line
link |
through the future of the world,
link |
but if it turns out, okay, there's a jump there.
link |
They found a better way to design cognitive systems.
link |
They designed a wholly new kind of thing.
link |
And the only line is some causal chain of designing
link |
and systems that design better systems.
link |
Is that so much worse?
link |
We're still at least part of a causal chain of design.
link |
And yes, they're not humans,
link |
but still they're our successes.
link |
So, I mean, ultimately I think it's probably inevitable
link |
that something like that will happen.
link |
And at least we were part of the process.
link |
It'd be nice if they still cared enough about us
link |
to maybe to engage with our arguments.
link |
I'm really hoping that the AGI's are gonna solve
link |
all the problems of philosophy.
link |
They'll come back and read all this crappy work
link |
for the 20th and 21st century,
link |
hard problem of consciousness.
link |
And here is why they got it wrong.
link |
And so, and if that happened,
link |
then I'd really feel like I was part of
link |
at least an intellectual process over centuries.
link |
And that would be kind of cool.
link |
Well, I'm pretty sure they would clone
link |
or they would recreate David Chalmers
link |
and for the fun of it,
link |
sort of bring back other philosophers.
link |
Yeah, bring back Descartes.
link |
Descartes and just put them in a room and just watch.
link |
It'll be a Netflix of the future show
link |
where you bring philosophers from different human,
link |
100% human philosophers from previous generations,
link |
put them in a room and see them.
link |
I am totally up for that.
link |
Simulators, AGI's of the future,
link |
if you're watching this podcast, do that.
link |
I would like to be recreated and ending out with Descartes.
link |
Where Descartes would be the first,
link |
if you could hang out as part of such a TV show
link |
with a philosopher that's no longer with us from long ago,
link |
who would you choose?
link |
Descartes would have to be right up there.
link |
Oh, actually a couple of months ago,
link |
I got to have a conversation with Descartes,
link |
an actor who's actually a philosopher
link |
came out on stage playing Descartes.
link |
I didn't know this was gonna happen.
link |
And I just after I gave a talk
link |
and told me about how my ideas were crap
link |
and all derived from him.
link |
And so we had a long argument.
link |
I would love to see what Descartes would think about AI,
link |
for example, and the modern neuroscience.
link |
And so I suspect not too much would surprise him,
link |
but yeah, William James,
link |
for a psychologist of consciousness,
link |
I think James was probably the richest.
link |
But, oh, there are Immanuel Kant.
link |
I never really understood what he was up to
link |
if I got to actually talk to him about some of this.
link |
Hey, there was Princess Elizabeth who talked with Descartes
link |
and who really got at the problems
link |
of how Descartes ideas of a nonphysical mind
link |
interacting with the physical body couldn't really work.
link |
She's been kind of, most philosophers
link |
think she's been proved right.
link |
So maybe put me in a room with Descartes
link |
and Princess Elizabeth and we can all argue it out.
link |
What kind of future?
link |
So we talked about zombies, a concerning future,
link |
but what kind of future excites you?
link |
What do you think if we look forward sort of,
link |
we're at the very early stages
link |
of understanding consciousness.
link |
And we're now at the early stages
link |
of being able to engineer complex, interesting systems
link |
that have degrees of intelligence.
link |
And maybe one day we'll have degrees of consciousness,
link |
maybe be able to upload brains,
link |
all those possibilities, virtual reality.
link |
Is there a particular aspect to this future world
link |
that just excites you?
link |
Well, I think there are lots of different aspects.
link |
I mean, frankly, I want it to hurry up and happen.
link |
It's like, yeah, we've had some progress lately in AI and VR,
link |
but in the grand scheme of things, it's still kind of slow.
link |
The changes are not yet transformative.
link |
And I'm in my fifties, I've only got so long left.
link |
I'd like to see really serious AI in my lifetime
link |
and really serious virtual worlds.
link |
Cause yeah, once people,
link |
I would like to be able to hang out in a virtual reality,
link |
which is richer than this reality
link |
to really get to inhabit fundamentally different kinds
link |
Well, I would very much like to be able to upload
link |
my mind onto a computer.
link |
So maybe I don't have to die.
link |
If this is maybe gradually replaced my neurons
link |
with a Silicon chips and inhabit a computer.
link |
Selfishly, that would be wonderful.
link |
I suspect I'm not gonna quite get there in my lifetime,
link |
but once that's possible,
link |
then you've got the possibility of transforming
link |
your consciousness in remarkable ways,
link |
augmenting it, enhancing it.
link |
So let me ask then,
link |
if such a system is a possibility within your lifetime
link |
and you were given the opportunity to become immortal
link |
in this kind of way, would you choose to be immortal?
link |
Yes, I totally would.
link |
I know some people say they couldn't,
link |
it'd be awful to be immortal, be so boring or something.
link |
I don't see, I really don't see why this might be.
link |
I mean, even if it's just ordinary life that continues,
link |
ordinary life is not so bad.
link |
But furthermore, I kind of suspect that,
link |
if the universe is gonna go on forever or indefinitely,
link |
it's gonna continue to be interesting.
link |
I don't think your view was that we just have to get
link |
this one romantic point of interest now
link |
and afterwards it's all gonna be boring,
link |
super intelligent stasis.
link |
I guess my vision is more like,
link |
no, it's gonna continue to be infinitely interesting.
link |
Something like as you go up the set theoretic hierarchy,
link |
you go from the finite cardinals to Aleph zero
link |
and then through there to all the Aleph one and Aleph two
link |
and maybe the continuum and you keep taking power sets
link |
and in set theory, they've got these results
link |
that actually all this is fundamentally unpredictable.
link |
It doesn't follow any simple computational patterns.
link |
There's new levels of creativity
link |
as the set theoretic universe expands and expands.
link |
I guess that's my future.
link |
That's my vision of the future.
link |
That's my optimistic vision
link |
of the future of super intelligence.
link |
It will keep expanding and keep growing,
link |
but still being fundamentally unpredictable at many points.
link |
I mean, yes, this creates all kinds of worries
link |
like couldn't all be fragile and be destroyed at any point.
link |
So we're gonna need a solution to that problem.
link |
But if we get to stipulate that I'm immortal,
link |
well, I hope that I'm not just immortal and stuck
link |
in the single world forever,
link |
but I'm immortal and get to take part in this process
link |
of going through infinitely rich, created futures.
link |
Rich, unpredictable, exciting.
link |
Well, I think I speak for a lot of people in saying,
link |
I hope you do become immortal and there'll be
link |
that Netflix show, The Future,
link |
where you get to argue with Descartes,
link |
perhaps for all eternity.
link |
So David, it was an honor.
link |
Thank you so much for talking today.
link |
Thanks, it was a pleasure.
link |
Thanks for listening to this conversation
link |
and thank you to our presenting sponsor, Cash App.
link |
Download it, use code LexPodcast,
link |
you'll get $10 and $10 will go to FIRST,
link |
an organization that inspires and educates young minds
link |
to become science and technology innovators of tomorrow.
link |
If you enjoy this podcast, subscribe on YouTube,
link |
give it five stars on Apple Podcast,
link |
follow on Spotify, support it on Patreon,
link |
or simply connect with me on Twitter at Lex Friedman.
link |
And now let me leave you with some words
link |
from David Chalmers.
link |
Materialism is a beautiful and compelling view of the world,
link |
but to account for consciousness,
link |
we have to go beyond the resources it provides.
link |
Thank you for listening and hope to see you next time.