Kind: captions
Language: en
the following is a conversation with
risto michelinen a computer scientist at
the university of texas at austin
and associate vice president of
evolutionary artificial intelligence at
cognizant
he specializes in evolutionary
computation
but also many other topics in artificial
intelligence
cognitive science and neuroscience quick
mention of our sponsors
jordan harbinger show grammarly belcampo
and indeed check them out in the
description to support this podcast
as a side note let me say that nature
inspired algorithms
from ant colony optimization to generic
algorithms to
cellular automata to neural networks
have always captivated my imagination
not only for their surprising power in
the face of long odds
but because they always opened up doors
to new ways of thinking about
computation
it does seem that in the long arc of
computing history
running toward biology not running away
from it
is what leads to long-term progress this
is
the lex friedman podcast and here is my
conversation
with risto mcelinen if we ran the
earth experiment this fun little
experiment we're on
over and over and over and over a
million times and watch the evolution of
life
as it uh pans out how much variation in
the outcomes of that evolution do you
think we would see
now we should say that you are a
computer scientist
that's actually not such a bad question
for computer scientists because
we are building simulations of these
things and we are simulating evolution
and that's a difficult question to
answer in biology but we can build a
computational model
and run it million times and actually
answer that question how much variation
do we see when we when we simulate it
uh and um you know that's a little bit
beyond what we can do today but
but i think that we will see some
regularities and it took evolution also
a really long time to get started and
then things
accelerated really fast uh towards the
end
but there are things that need to be
discovered and they probably will be
over and over again like manipulation
uh of objects uh opposable thumbs and
um and also some way to communicate
uh maybe orally like whether you have
speech it might be some other kind of
sound
and and decision making but also vision
uh
i has evolved many times various vision
systems have evolved
so we would see those kinds of solutions
i believe
emerge over and over again they may look
a little different but they
they get the job done the really
interesting question is would we have
primates would we have
humans or something that resembles
humans
uh and and would that be an apex of
evolution after a while
uh we don't know where we're going from
here but we certainly see a lot of tool
use and
and building our constructing our
environment so
i think that we will get that we get
some
evolution producing some agents that can
do that manipulate the environment and
build
what do you think is special about
humans like if you were running the
simulation
and you observe humans emerge like these
like tool makers they start a fire and
all stuff start running around building
buildings and then running for president
all those kinds of things
uh what would be how would you detect
that
because you're like really busy as the
creator of this evolutionary system
so you don't have much time to observe
like detect if
any cool stuff came up right how would
you detect humans
well you are running the simulation so
you also
put in visualization and measurement
techniques there so if you are looking
for
certain things like communication you'll
have detectors to find out whether
that's happening even if it's a lot
simulation
and i think that that's that's what what
we would do
we know roughly what we want intelligent
agents that communicate cooperate
manipulate
and we would build detections and
visualizations of those processes
yeah it and there's a lot of we have to
run it many times and
we have plenty of time to figure out how
we detect the interesting things
but also i think we do have to run it
many times because we don't quite know
what shape those will take and our
detectors may not be perfect for them
to begin with well it seems really
difficult to build the detector of
intelligent or intelligent conv
communication sort of uh if we take an
alien perspective observing earth are
you sure
that they would be able to detect humans
as the special thing wouldn't they be
already curious about other things
there's way more insects by body mass i
think than humans
by far and colonies
obviously dolphins is the most
intelligent uh creature on earth
we all know this so it could be the
dolphins that they detect
it could be the rockets that we seem to
be launching that could be the
intelligent creature they detect
uh it could be some other uh trees
trees have been here a long time i just
learned that sharks have been here
400 million years and that's longer than
trees have been here
so maybe it's the sharks they go by age
like there's a persistent thing
like if you survive long enough
especially through the mass extinctions
that could be the
the thing your detector is uh detecting
humans have been here for a short time
and we're just creating a lot of
pollution but so is the other creatures
i don't know you do you think you'd be
able to detect humans like how would you
go about detecting
in the computational sense maybe we can
leave humans behind
in the computational sense detect
interesting things
do you basically have to have a strict
objective function by which you measure
the performance of a system or can you
find curiosities and
interesting things yeah well i think
that the first
measurement would be to detect how much
of an effect you can have in your
environment so if you look at
look around we have cities and that is
constructed environments and that's
where a lot of people live most people
live
so that would be a good sign of
intelligence that you
don't just live in an environment but
you construct it to your liking yeah
and that's something pretty unique i
mean certainly birds build nests
but they don't build quite cities
termites build mounds and ice and things
like that
but the complexity of the human
construction cities i think would stand
out
even to an external observer of course
that's what a human would say
yeah and you know you can certainly say
that sharks are really smart because
they've been around so long and they
haven't destroyed the environment
which humans are about to do which is
not a very smart thing
uh but we'll get over it i believe
uh and and we can get over it by doing
some construction that actually is
benign
uh and maybe even enhances uh the um
resilience of nature so you mentioned
that this simulation that we run over
and over might
start so it's a slow start so do you
think uh
how unlikely first of all i don't know
if you think about this kind of stuff
but
how unlikely is step number zero
which is the springing up like the
origin of life on earth
and second how unlikely
is the anything interesting happening
beyond that
sort of like the start that that creates
all the rich complexity that we see on
earth today yeah
there are people who are working on
exactly that problem uh
from primordial soup how do you actually
get self-replicating yeah molecules
and they are very close uh with a little
bit of help you can make that happen
so we of course we know what we want so
they can set up the conditions and try
out conditions
that are conducive to that for evolution
to discover that
that took a long time for us to recreate
it probably won't take that long
and the next steps from there um
i think also with some hand-holding i
think we can make that happen
um but if with evolution what was really
fascinating was eventually the
runaway evolution of the brain that
created
humans and created well also other
higher animals
that that was something that happened
really fast and
that's a big question is that something
replicable is that something that
can happen and if it happens does it go
in the same direction
that is a big question to ask even in
computational
terms i think that it's relatively
possible to
come up here create an experiment where
we look at the primordial soup and the
first couple of steps
of multicellular organisms even but to
get something as complex as the brain
we don't quite know the conditions for
that and how to even get started and
whether we can get this kind of runaway
evolution happening
from a detector perspective
if we're observing this evolution what
do you think is the brain what do you
think is the
let's say what is intelligence so in
terms of the thing that makes humans
special we seem to be able to reason
we seem to be able to communicate but
the core of that is this
something in the broad category we might
call intelligence
so it's uh if you put your computer
scientists add on
uh is their favorite ways you like to
think about
that question of what is intelligence
well my goal is to create
agents that are that are intelligent
not to define what and and that that is
a way of defining it
and that means that it's some kind of an
um
object or or a program um
that has limited sensory and uh
effective capabilities interacting with
the world
and then also a mechanism for making
decisions
so with limited abilities like that can
it
survive um survival is the simplest goal
but it could
you could also give it other goals can
it multiply can it solve problems that
you give it
uh and that is quite a bit less than
human intelligence there are
animals would be intelligent of course
with that definition and you might have
even even some other forms of of life
even so what
so intelligence in that sense is a
survival um
skill uh given resources that you have
and using using your resources so that
you will stay around
do you think death mortality is
fundamental
to an agent so like there's a i don't
know if you're familiar there's a
philosopher named ernest becker who
wrote the
denial of death and his whole idea and
there's folks
psychologists cognitive scientists that
work on terror management theory
and they think that one of the special
things about humans is that we're able
to sort of
foresee our death right we can we can
realize not just as animals do sort of
constantly fear
in an instinctual sense respond to all
the dangers that are out there
but like understand that this ride ends
eventually yeah
and that in itself is the most is a
is the force behind all of the creative
efforts of human nature yeah that's
that's the philosophy
i think that makes sense a lot of sense
i mean animals probably don't think of
death the same way
but humans know that your time is
limited and you want to make it count
and you can make account in many
different ways but i think that has a
lot to do with creativity and the need
for humans
to do something beyond just surviving
and now going from that simple
definition to something that's
the next level i think that that could
be a second decision a second level of
definition that
um intelligence means something and you
do something that
stays behind you that's more than uh
your
existence um something you create
something that
um is useful for others is useful in the
future not just for yourself
and i think that that's a nice
definition of intelligence in a
next level uh and it's also nice because
it doesn't require
that they are humans or biological they
could be artificial
agents that intelligence they could they
could achieve those kind of goals
so particular agent the uh the ripple
effects of
of their existence on the entirety of
the system
is significant so like they leave a
trace where there's like uh
yeah like ripple effects it's the but
see then you go back to the
the butterfly with the flap of a wing
and then you can uh trace a lot of uh
like nuclear wars and all the conflicts
of human history
somehow connected to that one butterfly
that created all the
the chaos so maybe that's not
maybe that's a very poetic way to think
uh that's something we humans in a
human-centric way want to
hope we have this impact
like that is the the the secondary
effect of our intelligence we've
had the long-lasting impact on the world
but maybe the entirety
of physics in the universe
has a very long lasting effect sure but
you can also think of it what if um like
the wonderful life what if you're not
here will somebody else do this is it
is it something that you actually
contributed because you had something
unique to compute
that contribute that's a pretty high bar
though uniqueness
yeah yeah so you know you have to be
mozart or something to actually reach
that level that nobody would have
developed that but other people might
have solved this equation
um if you didn't do it
but but also within limited scope i mean
during your lifetime or next year
you could contribute something that
unique that other people did not see
and um and then that could change
the way things move forward for a while
uh so i don't think we have to be mozart
to be
called intelligence but we have this
local effect that is
changing if you weren't there that would
not have happened and it's a positive
effect of course you want it to be a
positive effect do you think it's
possible to engineer in
to uh computational agents a fear of
mortality
like uh does that make any
any sense so there's a very trivial
thing whereas like
you could just code in a parameter which
is how long the life ends but
more of a fear of mortality
like awareness of the the way that
things end
and somehow encoding a complex
representation
of that fear which is like maybe as it
gets closer
you become more terrified i mean there
seems to be something really profound
about this fear that's not
currently encodable in a trivial way
into our programs
well i think you're you're referring to
the emotion of fear something
because we are cognitively we know that
we have limited lifespan
and most of us cope with it by just hey
that's what the world is like and i make
the most of it but
sometimes you can have a like a a fear
that's not healthy
that paralyzes you you can't do anything
uh and and uh somewhere in between
they're
not caring at all and and getting
paralyzed because of fear
is a normal response which is a little
bit more than just logic
and and it's emotion so now the question
is what good are emotions i mean they
are quite
uh complex and they are multiple
dimensions of emotions
and they probably do serve as survival
function heightened focus for instance
and fear of death might be a really good
emotion when you are in danger that you
recognize it
even even if it's not logically
necessarily easy to derive and you don't
have time for
that logical detection a deduction you
may be able to recognize the situation
is dangerous and this fear
kicks in and you all of a sudden
perceive the facts that are important
for that and i think that's generally is
the role of emotions is it allows you to
focus
what's relevant uh for your situation
and maybe if fear of death plays the
same kind of role
uh but if it consumes you and it's
something that you think
in normal life when you don't have to
then it's not healthy and then it's not
productive
yeah but it's fascinating to think how
to uh
incorporate emotion into a computational
agent
it almost seems like a silly statement
to make
but it perhaps seems silly because we
have such a poor understanding of the
mechanism of
emotion of fear of uh
i think at the core of it is another
word that we know
nothing about but say a lot which is
consciousness
do you ever in your work or like maybe
on a coffee break think about what the
heck is this thing
consciousness and is it at all useful in
our thinking about ai systems
yes it is an important question
you can build representations and
functions
i think into these agents that act like
emotions
and consciousness perhaps so i mentioned
emotions being something that allow you
to focus and pay attention
filter out what's important yeah you can
have that kind of a filter mechanism
and you can it puts you in a different
state your computation is in a different
state certain things don't really get
through and others
are heightened now you label that box
emotion i don't know if that means it's
an emotion but it acts
very much like we understand what
emotions are
and we actually did some work like that
um modeling
hyenas who were trying to steal a kill
from lions
which happens in africa i mean hyenas
are
quite intelligent but not really
intelligent
and they they have this behavior that's
more complex than anything else they do
they can band together if there's
about 30 of them or so uh they can uh
coordinate their effort so that they
push the lions away from a kill
even though the lions are so strong that
they could kill a lion
kill a hyena by by striking with a paw
but when they work together and
precisely time this attack the lions
will leave and they get the kill
and probably there are some
states like emotions that the hyenas go
through the first they
they call for reinforcements they really
want that kill but there's not enough of
them so they vocalize and
there's more peop more people more
hyenas that come around
and then they have two emotions they're
very afraid of the lion
so they want to stay away but they also
have a strong affiliation
between each other and then this is the
balance of the two emotions and
and also yes they also want the kill so
it's both rebelled and attractive and
then but then this
affiliation eventually is so strong that
when they move they move together they
act as a unit
and they they can perform that function
so there's an
interesting behavior that seems to
depend on these emotions
strongly and makes it possible um
important reactions
and i think a cr a critical
aspect of that the way you're describing
is emotion there is
a mechanism of social communication
of a social interaction maybe that
maybe humans won't even be that
intelligent or most
things we think of as intelligent
wouldn't be that intelligent without
the social component of interaction
maybe most
much of our intelligence is essentially
in our growth of social interaction
and maybe for the creation of
intelligent agents we have to be
creating
yes fundamentally social systems yes i i
strongly believe that's true
and uh yes the uh communication is
multifaceted i mean
they they vocalize and call for friends
but they also
rub against each other and they push and
they do all kinds of
gestures and so on so they known act
alone and i don't think people act alone
uh very much either at least normal most
of the time
and social systems are so strong for
humans
that i think we build everything on top
of these kind of structures and
one interesting theory around that
bigger this theory for instance for
language but language origins is
that where did language come from and
and it's a plausible theory that first
came social systems that
you have different roles in a society
and then those roles are exchangeable
that you know i scratch
your back you scratch my back you can
exchange roles
and once you have the brain structures
that allow you to understand actions in
terms of roles that can be changed
that's the basis for language for
grammar and now you can start
using symbols to refer to uh objects in
the world
and you have this flexible structure so
there's a social structure
that's fundament fundamental for
language to develop
now again then you have language you can
you can refer to things that are not
here
right now and that allows you to then
build all the
all the good stuff about uh planning for
instance and
building things and so on so yeah i
think that very strongly
uh humans are social and that gives us
ability to structure the world
but also as a society we can do so much
more because we don't
one person does not have to do
everything you can have different roles
and together achieve a lot more and
that's also something we see in
computational simulations today i mean
we have multi-agent systems that
can perform tasks this fascinating uh
demonstration marco dorico i think it
was
um these robots little robots that had
to navigate through an environment and
there were
there were things that are dangerous
like maybe a
a big chasm or some kind of groove a
hole
and they could not get across it but if
they grab each other
with their gripper they formed a robot
that was much longer on
the team and this way they could get
across that yeah
so this is a great example of how
together we can achieve things we
couldn't otherwise like the hyenas
you know alone they couldn't but as a
team they could uh and i think humans do
that all the time we're really good at
that
yeah and the way you describe the the
system of hyenas
it almost sounds algorithmic like the
the problem with humans is they're so
complex
it's hard to think of them as algorithms
but with hyenas
there's a it's simple enough to where it
feels like
um at least hopeful that it's possible
to create
computational systems that mimic that
yeah that's exactly why why we looked at
that
as opposed to humans um like i said they
are intelligent but they are not quite
as intelligent
intelligent as say baboons which would
learn a lot and would be much more
flexible that hyenas
are relatively rigid in what they can do
and therefore you could look at this
behavior like this is a breakthrough in
evolution about to happen
yes that they've discovered something
about social structures communication
about cooperation and and it might then
spill over to other things too
yeah in thousands of years in the future
yeah
i think the problem with baboons and
humans is probably too much is going on
inside the head
we won't be able to measure it if we're
observing the system with hyenas is
probably
easier to observe the actual decision
making and the various
motivations that are involved yeah they
are visible
and we can even um quantify possibly
their
emotional state because they leave
droppings behind
and there are chemicals there that can
be associated with uh
with neurotransmitters and we can
separate what emotions they might have
experienced in the last 24 hours yeah
what to use the most beautiful speaking
of hyenas
uh what do you use the most beautiful uh
nature inspired
algorithm in your work that you've come
across something
maybe early on in your work or maybe
today
i i think that evolution computation is
the most amazing method so what
fascinates me
most is that with computers is that you
can
you can get more out than you put in i
mean you can write a piece of code
and your machine does what you told it i
mean
this happened to me in my freshman year
i it did something very simple and i was
just amazed i was blown away that it
would
it would get the number and it would
compute the result and i didn't have to
do it myself
very simple but if you push that a
little further
you can have machines that learn and
they might learn patterns
and already say deep learning neural
networks they can learn to recognize
objects sounds um patterns that humans
have trouble
with and sometimes they do it better
than humans and that's so fascinating
and now if you take that one more step
you get something like evolution
algorithms
that discover things they create things
they come up with solutions that you did
not think of
and that just blows me away it's so
great that we can build
systems algorithms that can be in some
sense
smarter than we are that they can
discover solutions that we might miss
a lot of times it is because we have as
humans we have certain biases we expect
the solutions to be a certain way
and you don't put those biases into the
algorithm so they are more free to
explore
and evolution is just absolutely
fantastic explorer
and that's what what really is
fascinating yeah i think uh
i get made fun of a bit because i
currently don't have any kids
but you mentioned programs i mean
um do you have kids yeah so maybe you
could speak to this but there's a magic
to the creation
creative process like i uh with spot
the boston dynamic spot but really any
robot that i've ever worked on
it just feels like the similar kind of
joy i imagine i would have as a father
not the same perhaps level but like the
same kind of wonderment like
exactly this which is like you know what
you had to do
initially uh to get this thing going
let's speak on the computer science side
like what the program looks like
but something about it uh doing
more than what the program was written
on paper
is like that somehow connects to the
magic
of this entire universe like that's
that's like i i feel like i found god
every time i like it's like uh because
you're you've really
created something that's living yeah
even if it's
it has a life of its own has the
intelligence of its own it's beyond what
you actually thought
yeah and that is i think it's exactly
spot on that's exactly what it's about
uh you created something and has a
ability to
uh live its life and and do good things
and um
you just gave it a starting point so in
that sense i think it's that may be part
of the joy actually
uh you see but you mentioned creativity
in this context uh
especially in the context of
evolutionary computation
so you know we don't often think of
algorithms as creative
so how do you think about creativity
yeah
algorithms absolutely can be creative um
they can
come up with solutions that you don't
think about i mean creativity can be
defined
a couple of requirements has to be new
it has to be useful and it has to be
surprising
and those certainly are true with say
evolution computation discovering
solutions so maybe
an example for instance we did this
collaboration with mit media lab kelp
harvest
lab where they had a
hydroponic food computer they called it
environment that was completely computer
controlled nutrients water
light temperature everything is
controlled now
um what do you do if you can control
everything
farmers know a lot about how to do how
to make plants grow in their own batch
of land
but if you can control everything it's
too much and it turns out that we don't
actually know very much about it
so we built a system evolution
optimization system
together with a surrogate model of how
plants grow
and let this system explore recipes
on its own and initially now we were
focusing on light
uh how strong what wavelengths how long
the light was on
um and we put some boundaries which we
thought were reasonable
for instance that there was um at least
six hours of darkness like night because
that's what we have in the world
and very quickly um the system evolution
pushed all the recipes to that limit uh
we were trying to grow basil
um and we had initially have some 200
300 recipes
exploration as well as known recipes but
but now we are going beyond that and
everything was like pushed at that limit
so
we look at it and say well you know we
can easily just change it let's have it
your way
and it turns out uh the system
discovered that bazel does not need to
sleep
uh 24 hours lights on and it will thrive
it will be bigger it will be tastier and
this was a big surprise
not just to us but also the biologists
in the team
that anticipated that this is some
constraints that
that are in the world for a reason it
turns out that evolution did not have
the same bias
and therefore it discovered something
that was creative it was surprising it
was useful and it was new
that's fascinating to think about like
the things we think that are fundamental
to
living systems on earth today whether
they're actually fundamental or they
somehow shape
uh fit the constraints of the system and
all we'll have to do is just remove the
constraints
do you ever think about um
i don't know how much you know about
bringing computer interfaces in your
link
the the idea there is you know our
brains are very limited
and if we just allow we plug in
we provide a mechanism for a computer to
speak with the brain
so you're thereby expanding the
computational power of the brain
the possibilities there sort of from a
very high level philosophical
perspective is limitless but i wonder
how limitless it is are the constraints
we have like features that are
fundamental to our intelligence
or is this just like this weird
constraint in terms of our brain size
and skull
and uh lifespan and the
senses it's just the weird little like
quirk of evolution and if we just open
that up like add much more senses
add much more computational power the uh
intelligence will be will expand
exponentially
do you have a do you have a sense about
constraints the relationship of
evolution computation to the constraints
of the environment
um well at first i'd like to comment on
on that like
changing the inputs uh to human brain uh
yes and
flexibility of of the brain i think
there's a lot
of that uh there are experiments that
are done in animals like megan kasir
um the mit is switching the um auditory
and visual
information and going going to the wrong
part of the cortex and the animal
was still able to hear and perceive the
visual environment
and there are kids that are born with
severe disorders and sometimes they have
to remove
half of the brain like one half and they
still grow up they have the functions
migrate to the other parts
there's a lot of flexibility like that
so i think it's quite possible to
hook up the brain with different kinds
of sensors for instance
and something that we don't even quite
understand or have today
on different kind of wavelengths or or
whatever they are um
and then the brain can learn to make
sense of it and that i think
is um this good hope that these
prosthetic devices for instance work
not because we make them so good and so
easy to use but the brain adapts to them
and can
learn to take advantage of them um and
so in that sense if there's a trouble a
problem i think that brain
can be used to correct it now going
beyond what we have today can you get
smarter
that's really much harder to do uh
giving
the brain more more input probably might
overwhelm it it would
have to learn to filter it and focus um
and in order to use the information
effectively
and augmenting intelligence with some
kind of external devices like that
might be difficult uh i think but
replacing what's lost
i think is quite possible right so our
intuition
allows us to sort of imagine that we can
replace what's been lost
but expansion beyond what we have i mean
we are already one of the most
if not the most intelligent things on
this earth right so it's hard to imagine
um if the brain can hold up with an
order of magnitude greater set of
information
thrown at it if it can do if you can
reason through that
part of me this is the russian thing i
think is uh i tend to think that the
limitations is where the
the superpower is that
you know immortality and uh
huge increase in bandwidth of uh
information by connecting computers with
the brain is not going to produce
greater intelligence
it might produce lesser intelligence so
i don't know there's something about
the scarcity being essential
to uh um fitness or performance
but that could be just because we're so
uh
limited no exactly you make do with what
you have but you can
uh you don't have to pipe it directly to
the brain i mean we already have
devices like phones where we can look up
information at any point
yeah and that can make us more
productive you don't have to argue about
i don't know what happened in that
baseball game or whatever it is because
you can look it up right away and i
think in that sense
we can learn to utilize tools and that's
what we
we have been doing for a long long time
um
so and we are already the brain is
already drinking from the water
fire hose like vision there's way more
information
in the vision that we actually process
so brain is already good at
identifying what matters yeah and
that we can switch that from vision to
some other wavelength or some other kind
of modality but i think that the same
processing principles probably still
apply uh but
but also indeed this uh ability to
uh have information more accessible and
more relevant i think
can enhance what we do i mean kids today
at school
they learn about dna i mean things that
we discovered just a couple of years ago
and it's already common knowledge and we
are building on it and we don't see
a problem where um where
there's too much information that we can
absorb and learn maybe people become a
little bit more
narrow in what they know they are in one
field
but this information that we have
accumulated it is passed on and people
are picking up on it
and they are building on it so it's not
like we have reached the point of
saturation
um we have still this process that
allows us to be selective
and decide what's interesting um i think
still works
even even with the more information we
have today yeah it's fascinating to
think about
like wikipedia becoming a sensor like uh
so the fire hose of information from
wikipedia so it's like you
integrate it directly into the brain to
where you're thinking like you're
observing the world with all of
wikipedia directly piping into your
brain
so like when i see a light i immediately
have like the history of
who invented electricity like integrated
very quickly into so just the way you
think about the world might be very
interesting
if you can integrate that kind of
information what are your thoughts if i
could ask
uh on uh early steps on that on the
neurolink side i don't know if you got a
chance to see but
uh there's a monkey playing pong
yeah through the brain computer
interface and uh
the dream there is sort of you're
already replacing the thumbs essentially
that you would use to play video game
the dream is to be able to
increase further the the interface by
which you interact with the computer
are you impressed by this are you
worried about this what are your
thoughts as a human
i think it's wonderful i think it's
great that we could we could do
something like that i mean you can
there are devices that read your eeg for
instance
and and you and humans can learn um
to control things using using just their
thoughts
in that sense and i i don't think it's
that different i mean those signals
would go to limbs they would go to
thumbs
uh now the same signals go through a
sensor to some computing system
it still probably has to be built on
human terms
uh not to overwhelm them but but utilize
what's there and sense the right kind of
um
patterns that are easy to generate but
oh that
i think is really quite possible and and
wonderful and could be very much more
efficient
is there so you mentioned surprising
being a characteristic of
uh creativity is there something you
already mentioned a few examples but
is there something that jumps out at you
as was particularly surprising
from the various evolutionary
computation systems you've worked on
the solutions that were
come up along the way not necessarily
the final solutions but
maybe things have even discarded is
there something that just jumps to mind
it it happens all the time i mean
evolution is so
creative uh so good at discovering
uh solutions you don't anticipate a lot
of times they are
taking advantage of something that you
didn't think was there like a bug in the
software for instance
a lot of there's a great paper uh the
community put it together
about uh surprising anecdotes about
evolution computation
a lot of them are indeed in some
software environment there was an
a loophole or a bug and the system uh
utilizes that by the way for people who
want to read it's kind of fun to read
it's called the surprising creativity of
digital evolution a collection of
anecdotes from the evolutionary
computation and
artificial life research communities and
there's just a bunch of stories from all
the seminal figures in this community
uh you have a story in there uh that
released to you at least
on the tic-tac-toe memory bomb so can
you can you uh
i guess uh describe that situation if
you think that's yeah
that was that's a quite a bit smaller
scale than our
um basil doesn't need to sleep surprised
but
but it was actually done by students in
my class um
in a neural net evolution computation
class uh there was an assignment
uh it was perhaps a final project where
people built game playing uh ai
it was an ai class uh and this one and
and it was for tic-tac-toe or
five in a row in a large board uh and uh
this one team
evolved a neural network to make these
moves
uh and um they set it up the evolution
they didn't really know what would come
out
but it turned out that they did really
well evolution actually won the
tournament
and most of the time when it won it went
because the other teams
crashed and then when we look at it like
what was going on
was that evolution discovered that if it
makes a move that's really really far
away
like millions of squares away
the other teams the other programs just
expanded memory
in order to take that into account until
they ran out of memory and crashed
and then you win a tournament by
crushing all your opponents
i think that's quite a profound example
which it probably applies to most
games from even a game theoretic
perspective
that sometimes to win you don't have to
be
better within the rules of the game you
have to come up with ways to
break your opponent's uh
brain as a human like not through
violence but through
some hack where the brain just is not um
you're basically uh how would you put it
you're
the you're going outside the constraints
of where the brain is able to
to function expectations of your
opponent i mean yeah this was even
kasparov pointed that out that when the
blue was playing against kasparov that
it was not playing
the same way as kasparov expected uh and
this has to do with
you know being not having the same
biases uh
and that's that's really one of the
strengths of of the
ai approach yeah can you at a high level
say
what are the basic mechanisms of
evolutionary computation
algorithms that use something that
could be called an evolutionary approach
like how does it work
uh what are the connections to the it's
what are the echoes of the connection to
is biological
a lot of these algorithms really do take
motivation from biology but they are
carry catches you try to essentialize it
and take the elements that
you believe matter so in evolution
computation
it is the creation of variation and then
the selection
upon that so the creation of variation
you have to have some mechanism that
allow you to create new individuals that
are very different from
what you already have that's the
creativity part and then you have to
have some way of measuring how well they
are doing
uh and using the uh that measure to
select
uh who goes to the next generation and
you continue so first you
also you have to have some kind of
digital representation
of an individual that can be then
modified so i guess humans
i mean biological systems have dna and
all those kinds of things and so you
have to have similar kind of encodings
in a computer program yes and that is a
big question how do you encode
these individuals so there's a genotype
which is that encoding and then
a decoding mechanism just gives you the
phenotype which is the actual individual
that then
performs the task and in an environment
can be evaluated how good it is so even
that mapping is a big question and how
do you do it
but typically the representations are
either they are strings of numbers or
they are some kind of trees those are
something we know very well in computer
science and we try to do that but they
and you know dna in some sense is also a
sequence
um and it's a string um
so it's not that far from it but dna
also has many other aspects
that we don't take into account
necessarily like this folding and
and interactions that are other than
just the sequence itself
and lots of that is not yet captured and
we don't know whether they are
really crucial um evolution biological
evolution has produced
wonderful things but if you look at them
it's not necessarily the case that every
piece is irreplaceable
and essential there's a lot of baggage
because you have to construct it and it
has to go through various stages and we
still have
appendiciti appendix and we have
tailbones and things like that that are
not really that useful
if you try to explain them now it would
make no sense
very hard but if you think of us as
productive evolution you can see where
they came from they were useful
at one point perhaps and and no longer
are but they're still there
so um that process is complex
uh and your representation should
support it
uh and that is quite difficult
if if we are limited with strings or
trees
and then we are pretty much limited what
can be constructed
and one thing that we are still missing
in evolution computation in particular
is
what we saw in biology major transitions
so that you go from for instance single
cell to multi-cell organisms
and eventually societies there are
transitions of level of
selection and level of what a unit is
and that's something we haven't captured
in evolution computation yet
does that require a dramatic expansion
of the representation
is that what that that is most likely it
does but it's
quite we don't even understand it in
biology very well where it's coming from
so it would be really good to look at
major transitions in biology try to
characterize them a little bit more in
detail what the processes are
how how does a so like a unit a cell
is no longer evaluated alone it's
evaluated as part of a community
organism right even though it could
reproduce now it can't
alone and it has has to have its
environment so there's a
there's a push to another level at least
the selection
and how do you make that jump to this
yes how do you make the jump that's part
of the algorithm
yeah yeah so we haven't really seen that
in computation
um yet and there are certainly attempts
to have
open-ended evolution things that could
add more complexity and start
selecting at a higher level but it is
still not
um quite the same as going from single
to multi to
society for instance in in biology so so
there essentially would be as opposed to
having one agent
those agent all of a sudden
spontaneously decide to
then be together and then your entire
system
would then be treating them as one agent
something like that
some kind of weird merger building but
also so you mentioned i think you
mentioned selection so basically there's
an
agent and they don't get to live on
if they don't do well so there's some
kind of measure of what doing well is
and isn't
and uh does the mutation come into play
at all
in the process and what the world does
it serve yeah so
in again back to what the computational
mechanisms of evolution computation are
so
um the way to create variation uh you
can take multiple individuals too
usually but
but you could do more and you exchanged
the part of the representation you do
some kind of recombination
it could be crossover for instance um
in biology you do have dna strings that
that are
cut and put together again we could do
something like that
um and it seems to be that in biology
crossover is really the workhorse in in
biological evolution in computation
we tend to rely more on mutation
and that is making random changes into
parts of the chromosome you try to be
intelligent and target certain areas of
it
and make the mutations also
follow some principle like you collect
statistics of performance and
correlations
and try to make mutations you believe
are going to be helpful
that's where evolution computation has
moved in the last
20 years i mean evolution competition
has been around for 50 years but a lot
of the recent
um success comes from mutation comes
from comes from
using statistics it's like the rest of
machine learning based on statistics we
use similar tools
to guide evolutionary computation and in
that sense it has diverged
a bit from biological evolution and
that's one of the things i think
we could look at again
having a weaker selection more crossover
large populations
more time and maybe a different kind of
creativity would come out of it
we are very impatient in evolution
competition today we want answers right
now right quickly
and every if no somebody doesn't perform
kill it yeah
uh and biological evolution doesn't work
quite that way
uh and and it's more patient yes much
more patient
so i guess we need to add some kind of
mating some kind of like dating
mechanisms like marriage may be in there
so to uh in into our algorithms to
improve the the
the the combination mechanism as opposed
to all mutation doing all of the work
yeah and many ways of being successful
you know
usually in every competition we have one
goal you know play
this game really well uh and compared to
others but
in biology there are many ways of being
successful you can build niches you can
be stronger faster larger or smarter
or you know eat this or eat that or you
know so
so there are many ways to solve the same
problem of survival
and that then breeds creativity um
and um it allows more exploration and
eventually you get
solutions that are perhaps more creative
rather than trying to go from
initial population directly or more or
less directly to your
maximum fitness which you measure that's
just one metric
so in a broad sense
before we talk about newer evolution
do you see evolutionary computation as
more effective than deep learning in
certain contexts
machine learning broadly speaking maybe
even
supervised machine learning i don't know
if you want to draw any kind of lines
and distinctions and
borders where they rub up against each
other kind of thing or one is more
effective than the other in the current
state of things
yes of course they are very different
and they address different kinds of
problems and
the deep learning has been really
successful in domains where we have a
lot of data
and that means not just data about
situations but also what the right
answers were
so labeled examples or they might be
predictions may be weather prediction
where the data itself
becomes labels what happened what the
weather was today and what will
be tomorrow so they are very effective
deep learning methods on that kind of
tasks
but there are other kinds of tasks where
we don't really know what the right
answer is
uh game playing for instance but many
robotics tasks and
actions in the world decision making um
and actual practical applications like
treatments and healthcare
or investment in stock market many tasks
are like that
we will we don't know and we'll never
know what the optimal answers were
and there you need different kinds of
approaches reinforcement learning is one
of those
uh reinforcement learning comes from
biology as well
agents learn during their lifetime they
eat berries and sometimes they get sick
and then they don't
and get stronger and then that's how you
learn
and evolution is also a mechanism like
that
at a different time scale because you
have a population not an individual
during its lifetime
but an entire population as a whole can
discover um
what works and there you can afford
individuals that don't work out
they will you know everybody dies and
you have a next generation and it will
be better than the previous one
so that's that's the big difference
between these methods they apply
to different kinds of problems um and
um in particular there's often a
comparison that's kind of interesting
and important between reinforcement
learning and evolution and computation
and initially um reinforcement learning
was about
individual learning during the lifetime
and evolution is more
engineering you don't care about the
lifetime you don't care about all the
individuals that are tested you only
care about the final result
the last one the best candidate that
evolution produced
in that sense they also apply to
different kinds of problems
and now that boundary is starting to
blur
a bit you can use evolution as an online
method and reinforcement learning to
create engineering solutions but that's
still
roughly the distinction and
from the point of view what algorithm
you want to use
if you have something where there is a
cost for every trial reinforcement
learning might be your choice
now if you have a domain where you can
use a surrogate perhaps
so you don't have much of a cost for
trial
and you want to have surprises you want
to explore more broadly
then this population-based method is
perhaps a better choice because you you
can try things out that you wouldn't
afford when you're doing reinforcement
there's very few things as entertaining
as watching either evolution competition
or reinforcement learning teaching a
simulated robot to walk
i maybe there's a
higher level question that could be
asked here but
do you find this whole space in of
applications in the robotics
interesting for evolution computation
yeah yeah very much
um and indeed that's the fascinating
videos of that
and that's actually one of the examples
where you can contrast the difference
so between reinforcement learning
evolution yes so
if you have a reinforcement learning
agent it tries to be
conservative because it wants to walk as
long as possible and be stable
but if you have evolutionary computation
it can afford
these agents that go haywire they
fall flat on their face and they could
take a step and then they jump and then
again fall flat yeah and eventually what
comes out of that
is something like a falling that's
controlled yeah
and you take another step another step
and you no longer fall
instead you run you go fast so that's a
way of discovering
something that's hard to discover step
by step incrementally
because you can afford these
evolutionists
dead ends although they are not entirely
dead ends in the sense that they can
serve as stepping stones
when you take two of those put them
together you get something that works
even better
and that is a great example of of this
kind of discovery yeah learning to walk
is a
is fascinating i talked quite a bit to
russ tedron because mit
there's a there's a community of folks
who who just roboticists who love
the elegance and beauty of uh movement
right and uh walking bipedal
robotics is um
beautiful but also exceptionally
dangerous in the sense that like
you're constantly falling essentially if
you want to do elegant movement
and uh the discovery of that is uh
i mean it it's such a good example of um
that the discovery of a good solution
sometimes requires a leap of faith and
patience and all those kinds of things
i wonder what other spaces where you
have to discover those kinds of things
in
yeah yeah yeah and another interesting
direction is um
learning um for for uh virtual creatures
learning to walk
uh we did a study in in simulation
obviously that
um you create those creatures not just
their controller but also their body so
you have
cylinders you have muscles you have
joints
and sensors and you're creating
creatures that look quite different some
of them have multiple legs some of them
have no legs at all
and then the goal was to get them to
move the walk to run
uh and what was interesting is is that
when you evolve
the controller together with the body
you get movements that look
natural because they're optimized for
that physical setup
and and these creatures you start
believing them that they're alive
because they walk in a way that you
would expect somebody
with that kind of a setup to walk yeah
there's a
there's something subjective also about
that right i've been thinking a lot
about that especially in
the human robot interaction context
you know i mentioned spot the boston
dynamics robot
there is something about human robot
communication
let's say let's put in another context
something about
human and uh dog context
like like a living dog where there's uh
there's a there's a dance of
communication first of all the eyes you
both look at the same thing and you
dogs communicate with their eyes as well
like if if the
if you and a dog want to uh
like deal with a particular object you
will look at the person
the dog will look at you and then look
at the object and look back at you all
those kinds of things
but there's also just a elegance of
movement
i mean there's the of course the tail
and all those kinds of mechanisms of
communication it all seems natural
and often joyful and for robots to
communicate that
is it's really difficult how to figure
that out because it's it's almost
seems impossible to hard code in you can
hard code it for a demo purpose with so
you know something like that but it's
essentially choreographed
like if you watch some of the boston
dynamics videos where they're dancing
all of that is choreographed by human
beings
but to learn how to with your movement
demonstrate a naturalness and elegance
that's fascinating of course in the
physical space that's very difficult to
do
to learn the kind of at scale that
you're referring to but
the hope is that you could do that in
stimulation and then transfer into the
physical space
if you're able to model the robots
efficiently naturally
yeah and and sometimes i think that that
requires a theory of mind
on the yes on the on the side of the
robot that
that they as they understand what you're
doing because they
themselves are doing something similar
and uh that's a big question too
uh we talked about how intelligence in
general
and and the social aspect of of
intelligence and
i think that's what is required that we
humans understand other humans because
we assume that they are similar
to us um we have one simulation we did a
while ago ken stanley
um did that um two robots that were
uh competing um simulation like you said
they were foraging for food to gain
energy and then when they were really
strong they would
bounce into the other robot and win if
they were stronger
and we watched evolution discover more
and more complex behaviors they
first went to the nearest food and then
they started to
plot a trajectory so they get more get
more but then they started to take
pay attention what the other robot was
doing and in the end there was a
behavior
where one of the robots the most
sophisticated one
you know sensed where the food pieces
were
and identified that the other robot was
close to uh
two of a very far distance uh and there
was one more food
near by so it faked
that's now i'm using anthropomorphized
terms but it made a move towards those
other pieces
in order for the other robot to actually
go and get them
because it knew that the other the last
remaining piece of food was close
and the other robot would have to travel
a long way lose its energy
and then lose the whole competition
so there was like emergence of something
like a theory of mind
knowing what the other robot would do
guided towards bad behavior in order to
win
so we can get things like that happen uh
in in simulation as well but that's a
complete
natural emergence of a theory of mind
but i feel like if you
add a little bit of a place for a theory
of mind to emerge
like easier then you can go really far
i mean some of these things with
evolution
you know you add a little bit of design
in there
it'll really help and i think i tend to
think that
a very simple theory of mind
will go a really long way for
cooperation between agents
and certainly for human robot
interaction like it doesn't have to be
super complicated um
i've gotten a chance to in the
autonomous vehicle space to watch
vehicles interact with pedestrians or
pedestrians interacting with vehicles in
general
i mean you would think that there's a
very complicated
theory of mind thing going on but i have
a sense it's not well understood yet but
i have a sense it's pretty dumb
like it's pretty simple there's a social
contract
there where between humans a human
driver and a human crossing the road
where um the the human crossing the road
trusts that the human in the car is not
going to murder them
and there's something about again back
to that mortality thing
there's some dance
of ethics and morality that's built in
that you're mapping your own morality
onto the
the person in the car and even if
they're driving
at a speed where you think if they don't
stop they're going to kill you
you trust that if you step in front of
them they're going to hit the brakes
and there's that weird dance that we do
that
i think is a pretty simple model but of
course it's very difficult
to introspect what it is and autonomous
robots in the human robot interaction
context have to
have to build that current robots are
much less than what you're describing
they're
currently just afraid of everything
they're they're more they're not the
kind that
fall and discover how to run they're
more like
please don't touch anything don't hurt
anything
stay as far away from humans as possible
treat
humans as ballistic objects that you
can't
uh that you do uh with a large spatial
envelope
make sure you do not collide with that's
how like you mentioned elon
musk thinks about autonomous vehicles i
tend to think
autonomous vehicles need to have a
beautiful dance between human and
machine
where it's not just a collision
avoidance problem but a weird
dance yeah i think that you these
systems need to be
able to predict what will happen what
the other agent is going to do
and then have a structure of
what the goals are and whether those
predictions actually meet the goals and
and you
can go probably pretty far with that
relatively simple setup already
but to call it a theory of mind i don't
think you need to i mean it
it doesn't matter whether a pedestrian
has a mind it's an object
and we can predict what we'll do and
then we can predict what the states will
be in the future and whether they are
desirable states
stay away from those that are
undesirable and go towards those that
are desirable so
it's a relatively simple functional
approach to that
where do we really need the theory of
mind
maybe maybe when you start interacting
and
you're trying to get the other agent to
do something and jointly so that you can
jointly
collaboratively achieve something then
then you then it becomes more complex
well i mean even with the pedestrians
you have to have a sense of where their
attention
actual attention in terms of their gaze
is but also like
a tent i mean there's this vision
science people talk about this all time
just because
i'm looking at it doesn't mean i'm
paying attention to it so
figuring out what is the person looking
at what is the sensory information
they've taken in
and the theory of mind piece comes in is
what are they
actually attending to cognitively
and also what are they thinking about
like what is the computation they're
performing
and you have you have probably maybe a
few options
you know for the pedestrian crossing it
doesn't have to be
it's like a variable with a few discrete
states but you have to have a good
estimation which of the states that
brain is in
for the pedestrian case and the same is
for attending with a robot
if you're collaborating to pick up an
object you have to figure out
is the human like uh like there's a few
discrete states that the human could be
and you have to
you have to predict that by observing
the human and that seems like a machine
learning problem to figure out
uh what's how the human is uh what's the
human up to
it's not as simple as sort of planning
just because they move their
arm means the arm will continue moving
in this direction
you have to you have to really have a
model of what they're thinking about and
what's the
motivation behind the moment and here we
are talking about
uh relatively simple physical actions
yeah but you can take that
the higher levels also like to predict
what the people are going to do you need
to know
what uh what their goals are uh what are
they trying to are they exercising are
they
starting to get somewhere but even even
higher level i mean you are
predicting what people will do in their
career what their life themes are do
they want to be famous
rich or do good and that takes a lot
more information
but it allows you to then predict their
their actions what choices they might
make
so how does uh evolution computation
apply
to the world of neural networks because
i've seen quite a bit of work
from you and others on the in the world
of neural evolution so maybe first
can you say what is this field yeah a
new evolution
is a combination of of uh neural
networks and evolution computation
in many different forms but the early
versions were simply using evolution
the way um as a way to construct the
neural network
instead of say stochastic gradient
descent or back propagation
because evolution can evolve these
parameters weight values in a neural
network just like any other string of
numbers you can you can do that
and that's useful because some cases you
don't have
those targets that you need to um back
propagate from
and it might be an agent that's running
a maze or a robot
playing a game or something you don't
again you don't know what the right
answer says you don't have backup
but this way you can still evolve in
your own hand and
neural networks are really good at this
task because they um
they recognize patterns and they and
generalize
interpolate between known situations so
you want to have a neural network in
such a task
even if you don't have the supervised
targets so that's a reason and that's a
solution
and also more recently now when we have
all this deep learning literature
it turns out that we can use evolution
to optimize
many aspects of those designs the deep
learning
architectures have become so complex
that there's little hope for
as little humans to understand their
complexity and what actually makes a
good design
uh and now we can use evolution to give
that design for you and it might be
mean um optimizing hyper parameters
like the depth of layers and so on uh or
the topology of the network
um how many layers how they're connected
but also other aspects like what
activation functions you use where
in the network during the learning
process or what loss function you use
you could generalize that generate that
even data augmentation all the different
aspects of the design of deep learning
experiments could be optimized that way
so that's an inter interaction between
two mechanisms
but there's also when we get more into
cognitive science and the topics that
we've been talking about
you could have learning mechanisms at
two level time scales
so you do have an evolution that gives
you
baby neural networks that then learn
during their lifetime
and you have this interaction of two
time scales and i think that can
potentially be really powerful
now in biology we are not born with all
our
faculties we have to learn we have a
developmental period in humans it's
really long
and most animals have something and and
probably the reason is that
evolution and dna is not detailed enough
or plentiful enough to describe them we
can't describe how to set the brain
up but we can
evolution can decide on a starting point
and then have a learning algorithm that
will construct
the final product and this interaction
of
you know intelligent um well
evolution that has produced a good
starting point for the
specific purpose of learning from it
with the interaction of
uh with the environment that can be a
really powerful mechanism for
constructing brains and construction
behaviors
i like how you walk back from
intelligence so optimize
starting point maybe uh yeah
uh okay there's a lot of fascinating
things to ask here and this is
basically this dance between neural
networks and evolutionary computations
could go into the category of automated
machine learning so where you're
optimizing whether it's hyper parameters
of the topology or
hyper parameters taken broadly but the
topology thing is really interesting i
mean that's not really done that
effectively or throughout the history of
machine learning has not been
done usually there's a fixed
architecture maybe there's a few
components you're playing with
but to grow a neural network essentially
the way you grow in their organisms
really
fascinating space how how hard it is it
do you think
to grow in your network and maybe what
kind of neural networks
are more amenable to this kind of idea
than others
i've seen quite a bit of work on
recurrent neural networks is there some
architectures
that are friendlier than others and is
is this just a fun
small scale set of experiments or do you
have hope
that we can be able to grow powerful
neural networks
i i think we can uh and most of the work
up to now is taking architectures that
already exist that humans have designed
and tried to optimize them further and
and you can totally do that a few years
ago we did an experiment we took a
winner of the
uh image captioning competition um
and um the architecture and just broke
it into pieces and
took the pieces and and that was our
search effects see if you can do better
and we indeed could fifteen percent
better performance by just
searching around the network design that
humans had come up with oriovenials
and others uh so but that's starting
from a point of
point that humans have produced but we
could
do something more general it doesn't
have to be that kind of network
the the hard part is just a couple of
challenges one of them is to define the
search space
what are your elements uh and how you
put them together
and the space is just really really big
uh so you have to somehow constrain it
and have some hunch of what will work
uh because otherwise everything is
possible and another challenge is that
in order to evaluate how good your
design is
you have to train it i mean you have to
actually try it out
and that's currently very expensive
right i mean deep learning networks may
they
take days to train well imagine having a
population of 100 and have to
run it for 100 generations it's not yet
quite feasible computationally
um it will be but but also there's a
large carbon footprint and all that i
mean we're using a lot of computation
for doing it
so intelligent methods and intelligence
i mean
we have to do some science in order to
figure out what the right
representations are
and right operators are and how do we
evaluate them
without having to fully train them and
that is where the current research is
and we're making progress on all those
fronts
um so so yes there are certain
architectures that are more amenable to
that
uh approach but also i think we can
create our own
architecture and whole representations
that are even better
do you think it's possible to do like uh
like a tiny baby network that grows into
something that can do state of the art
and like even the simple data set like
mnist
and just like it uh just grows into a
you know gigantic monster that's the
world's greatest handwriting recognition
system
yeah there are approaches like that
esteban rail and cochlear for instance
have worked on
evolving a smaller network and then
systematically expanding it to a larger
one
uh your elements are already there and
scaling it up will just give you more
power
so again evolution gives you that
starting point yes and then there's a
mechanism
that gives you the final result and a
very powerful approach
um but you know you could you could also
um simulate the actual growth process
and like i said before evolving a
starting point and then evolving it
uh or training the network there's not
that much work that's been done
on that yet uh we need some kind of a
simulated
simulation environment so there are
interactions uh at will
uh the supervised environment doesn't
really it's not as easily
uh usable here sorry the interaction
between neural networks yeah the neural
networks that you are creating
interacting the world uh and learning
from these uh
sequences of interactions perhaps
communication with others
[Laughter]
that's awesome we would like to get
there but just the task of simulating
something is
at that level is very hard it's very
difficult i love the idea
i mean one of the powerful things about
evolution on earth is the predators and
prey emerged
and like there's just like there's
bigger fish and smaller fish and
it's fascinating to think that you could
have neural networks competing against
each other one yellow network being able
to destroy another one
there's like wars of neural networks
competing to solve
the mnist problem i don't know yeah yeah
oh totally yeah yeah yeah
and and we actually simulated also that
uh prayer the prey
and it was interesting what happened
there but budget but minnie roger poland
did this
and um kay holcomb was a zoologist so we
had
again um
we had simulated hyenas simulated zebras
nice uh and initially you know
the hyenas just tried to hunt them and
when they actually
stumbled upon the zebra they ate it and
we're happy
um and and then the zebras learned to
escape uh and the hyenas learned to team
up
and actually two of them approached in
different directions and now the zebras
their next step
they generated a behavior where they
split
in different directions just like
actually gazelles do
in in when they are being hunted they
confuse the predator by going in
different directions
that emerged and then more hyenas joined
and and kind of circled them uh and
and then when they circled them they
could actually hurt the zebras together
and and eat
multiple uh zebras so there was a like
an
arms race of predators and prey
and they gradually develop more complex
behaviors some of which we actually do
see in nature
uh and and this kind of co-evolution uh
that's competitive evolution it's a
fascinating topic because there's a
a promise or possibility that you will
discover something
uh new that you don't already know you
didn't build it in
it came from this arms race it's hard to
keep the arms race going it's hard to
have
reits enough simulation that that
supports all of these complex behaviors
but at least for several steps we've
already seen it in the spread of the
prey scenario yeah
first of all it's fascinating to think
about this context in terms of uh
evolving architectures so i've studied
tesla autopilot for a long time
it's one particular implementation
of an ai system that's operating in the
real world i find it fascinating because
of the scale at which it's used out in
the real world
and uh i'm not sure if you're familiar
with that system much but
you know andre kapathi leads that team
on the machine learning side
and there's a multitask
network multi-headed network where
there's a core
but it's trained on particular tasks and
there's a bunch of different heads that
are trained on that
is there some lessons from
evolutionary computation or neural
evolution that could be applied to this
kind of
multi-headed beast that's operating in
the real world yes
it's a very good problem for new
revolution
and the reason is that when you have
multiple tasks
they support each other so let's say
you're
learning to classify x-ray images uh
different pathologies so you have one
task is to classify
this disease and another one this
disease another on this one and when
you're learning
from one disease that forces certain
kinds of internal representations and
embeddings
and they can serve as a helpful starting
point for the other tasks
so you are combining the wisdom of
multiple tasks into these
representations
and it turns out that you can do better
in each of these tasks
when you're learning simultaneously
other tasks than you would by one task
alone which is a fascinating idea in
itself yeah
yes and and people do that all the time
i mean you use knowledge of domains that
you know
in new domains uh and and certainly
neural networks can do that
when your evolution comes in is that um
what's the best way to combine
these tasks now there's architectural
design that allow you to decide where
and how the the embeddings the internal
representations are combined
and how much you combine uh them uh and
uh
there's quite a bit of research on that
and and my team eliot madison has worked
on that
um in particular like what is a good
internal representation
that supports multiple tasks uh and
we're
getting to understand how that's
constructed and what's in it
uh so that it is in a space that
supports multiple
different heads like you said um and
and that i think is fundamentally how
biological intelligence works as well
uh you don't build a representation just
for one task you try to build something
that's general
not only so that you can do better in
one task or multiple tasks but also
future tasks and future challenges
so you learn to learn the structure of
of the world
um and and that helps you uh in all
kinds of future future challenges and so
you're trying to design a representation
that will support
an arbitrary set of tasks in a
particular sort of class of problem
yeah and and also it turns out and
that's again a surprise that elliot
found
was that those tasks don't have to be
very related
you know you can learn to do better
vision by learning language
or better language by learning about dna
structure
no somehow the world
yeah it rhymes the world rhymes even
it's very uh
very desperate fields um i mean
on that small topic let me ask you
because you've also on the competition
your science side
you worked on both language and vision
what's what's the connection between the
two uh
what's more maybe there's a bunch of
ways to ask this but what's more
difficult
to build from an engineering perspective
an evolutionary perspective
the human language system or the human
vision system or
the equivalent of in the ai space
language and vision
or is it the the best is the multi-task
idea that you're
speaking to that they they need to be
deeply integrated
yeah absolutely learning both at the
same time
i i think is a fascinating direction in
that in the future so you have data sets
where there's visual component as well
as
verbal descriptions for instance and and
that way you can learn
a deeper representation a more useful
representation for both
uh but it's still an interesting
question of um
which one is easier eventually i mean
recognizing objects or
even understanding sentences that's
relatively possible
but where it becomes where the
challenges are is to understand the
world
like the visual world the 3d uh what are
the objects doing and predicting what
will happen
uh the relationships that's what makes
vision difficult and language obviously
it's it's what's the mean what what is
being said what the meaning is
and the meaning doesn't stop at who did
what to whom
um there are goals and plans and themes
and
you eventually have to understand the
entire uh human society and history
in order to understand the sentence very
much fully that
there are plenty of examples of those
kind of short sentences when you bring
in all the world knowledge
uh to understand it uh and that's the
big challenge
now we are far from that but even just
bringing in the visual world
uh together with the sentence will give
you
already a lot deeper understanding of
what's happening
and i think that that's where we're
going very soon i mean
we've we've had imagenet for a long time
and now we have
all these uh text collections but having
both together
uh and then learning a semantic
understanding of what is happening
i think that that will be the next step
in the next few years yeah you're
starting to see that with
all the work with transformers was the
the community
the ai community started to dip their
toe into this idea idea of
having uh language models that are now
doing stuff with images with vision and
then
connecting the two i mean right now it's
like these little explorations we're
literally dipping the toe in
but like maybe at some point we'll just
like dive into the pool
and it'll just be all seen as the same
thing i i do still wonder what's more
fundamental
well their vision is um whether we
don't think about vision correctly maybe
the fact because we're humans and we see
things as beautiful and so on
that and because we have cameras that
taking pixels is a 2d
image that we don't sufficiently think
about
vision as language you know maybe
maybe chomsky is right all along that
vision is fundamental to
uh sorry that language is fundamental to
everything
to even cognition to even consciousness
like the base layer is all
language not necessarily like english
but some
weird abstract representation uh the
linguistic representation
yeah well earlier we talked about the
social structures and that may be what's
underlying
the language and that's the more
fundamental part and then language has
been added on top of that language
emerges from the social
interaction probably yeah that's a very
good guess um
via visual animals though a lot of the
brain is dedicated to vision and
and also when we think about various
abstract
concepts uh we usually reduce that to
vision
uh and and images and that's
you know go to a whiteboard you draw
pictures of very abstract concepts
so we tend to tend to resort to that
quite a bit and that's a fundamental
representation it's probably
possible that it predated um you know
language even i mean
animals a lot of they don't talk but
they certainly do have vision
uh and and language is interesting
development in um from for mastication
from eating
you develop an organ that actually can
produce sound to manipulate them
maybe that was an accident maybe that
was something that was available and
and then allowed us to to do that
communication or maybe it was
gestures sign language could have been
the original proto-language
we don't quite know but they're the
language is more fundamental than the
medium
in which it's uh communicated and i
think that it comes from those
representations
now in in current
world they are so strongly integrated
it's really hard to say which one is
fundamental
you look at the brain structures and
even visual cortex
which supposed to be very much just
vision well if you are
thinking of semantic concepts you're
thinking of language visual cortex
lights up
it's still useful even for language
computations
so there are common structures
underlying them so utilize what you need
yeah and and when you are understanding
a scene you're understanding
relationships
well it's not so far from understanding
relationships between words and concepts
so i think that that's how they are
integrated yeah and there's dreams and
wants to close our eyes there's still a
world in there somehow operating and
somehow
possibly the visual visual system
somehow integrated into all of it
i tend to enjoy thinking about aliens
and thinking about uh
the sad thing to me about
extraterrestrial intelligent life
that if it was if it visit us here on
earth
or if we came on mars and or maybe
another other solar system another
galaxy one day
that uh us humans would not be able to
detect it
or communicate with it or appreciate
like it'd be right in front of our nose
and we're too self-obsessed
to see it not self-obsessed but
our our our our tools
our frameworks of thinking would not
detect it
as a good movie arrival and so on where
stephen wolfram and his son i think were
part of developing this alien language
of how aliens would communicate with
humans do you ever think about that kind
of stuff
where if humans and aliens would be able
to communicate with each other
like if we uh met each other at some
okay we could do
seti which is communicating from across
a very big distance
but also just us you know
if you did a podcast with an alien do
you think we'd be able to find a common
language
uh and a common methodology of
communication
i think from a computational perspective
the way to ask that is
is you have very fundamentally different
creatures agents that are created would
they be able to find a common language
yes that's i do think about that i mean
i think a lot of people who are in
computing they
uh and ai in particular they got into it
because they were fascinated with
science fiction and and all of these
options i mean
star trek generated all kinds of devices
that we have now they they envisioned
it's true first and and it's a great
motivator um
to think about things like that um and i
so one and again being a computational
scientist and and
trying to build intelligent agents
what i would like to do is have a
simulation
where the agents actually evolve
communication not just communication
we've done that
people have done that many times they
communicate they signal
and so on but actually develop a
language and language means grammar it
means all these
social structures and on top of that
grammatical structures
and we do it in under various conditions
and actually try to identify what
conditions are necessary for it to
come out and then we can start asking
that kind of questions
are those languages that emerge in that
those different simulated environments
are they understandable to us can we
somehow make a translation
we can make it a concrete question so
machine translation of evolved languages
and so like
languages that evolve come up with can
we translate
like i have a google translate for the
evolved languages
yes and if we do that enough we have
perhaps an idea what an alien
language might be like the space or
where those languages can be
because we can set up their environment
differently there doesn't need to be
gravity
you know you can you can have all kinds
of societies can be different they may
have no predators they may have all
everybody is a predator
all kinds of situations and and then see
what the
space possibly is where those languages
are and what the difficulties are
they'll be really good actually to do
that before the aliens come here
yes it's good practice yeah uh on the
similar
connection you know you can think of ai
systems as
aliens is there uh ways to evolve a
communication scheme
for there's a field you can call like
explainable ai
for ai systems to be able to communicate
so you have a but you evolve a bunch of
agents but
for some of them to be able to talk to
you yeah also
so to evolve a way for agents to be able
to communicate
about their world to us humans do you
think that there's
possible mechanisms for doing that we
can certainly try
and if we um if it's an evidence
competition system for instance you
reward
those solutions that are actually
functional that that communication makes
sense it allows us to
together again achieve common goals i
think it's possible
but even from that um paper that you
mentioned the the anecdotes it's quite
likely also that the
uh the agents learn to you know lie
and fake and do all kinds of things like
that yes i mean we see that in in
even very low level like bacterial
evolution there are there are cheaters
um and who's to say that what they say
is actually what they think
it um but but that's what i'm saying
that there would have to be some common
goal
so that we can evaluate whether that
communication is at least useful
um you know they may be saying things
just to make us feel good
or or get us to do what we want whatever
not turn them off or something
but but uh so we would have to
understand their internal representation
is much better to really make sure that
that translation is
political um but it can be useful and i
think that it's possible to do that
there are examples where visualizations
um are automatically created so that we
can look into the what
the system uh and the language is not
that far from it i mean it is a way of
communicating and logging
what you're doing in some inter
interpretable way
um i think a fascinating topic yeah to
do that
yeah you're making me realize that it's
a good scientific question
whether lying is an effective mechanism
for integrating yourself
and succeeding in a social network in a
social
in a world that is social i tend to
believe
that honesty and love are evolutionary
advantages
in us in a in an environment
where there's a network of intelligent
agents but it's also very possible that
dishonesty
and manipulation and
uh even you know violence all those
kinds of things might be more beneficial
that's the old open question about uh
good versus evil but i tend to
there's some i mean i don't know if it's
a hopeful maybe i'm delusional
but it feels like karma is a thing
which is like if
long term the agents that are just kind
to others sometimes for no reason
will do better in a society that's not
highly constrained on resources it's
like people start getting weird and evil
towards each other and bad
when the resources are very low relative
to the needs of the
the populace especially at the basic
level like
survival shelter uh food all those kinds
of things but um
i i tend to believe that uh once you
have those things established then
well not to believe i i guess i hope
that ai systems would be honest
but it's fun it's scary to think about
the touring test you know ai systems
that will
eventually pass the touring test will be
ones that are exceptionally good at
lying
that's a terrifying concept yeah i mean
i i don't know first of all so from uh
from somebody who studied language and
obviously
are not just the world expert in ai but
somebody who dreams about
the future of the field do you hope do
you think
there will be human level or superhuman
level intelligences in the future
that we eventually build
well definitely hope that we can we can
get there
one i think um important perspective is
that we are building ai to help us
uh that it is a it is tool like cars or
or or language or
communication uh ai will help us be more
productive
uh and that is always a condition
it's not something that we build and let
run and it
becomes an entity of its own that
doesn't care about us
now of course really far in the future
maybe that might be possible but not in
the foreseeable future when we are
building it
uh and therefore we are always in a
position of
limiting what it can or cannot do uh
and the um
your point about lying is very
interesting
um even even in these highness societies
for instance
uh when a number of these hyenas band
together and they
they still they take a risk and steal
the kill
they're always hyenas that hang back and
don't participate in that
uh risky behavior but they walk in later
and
and join the party after the after the
kill
and there are even some that may be
ineffective and
cause others to have harm so and
like i said even bacteria cheat and we
see it in biology
there's always some element an
opportunity if you have a
i think that is this because if you have
a society in order for society to be
effective you have to have this
cooperation and you have to have
trust uh and and if you have enough of
agents
who are able to trust each other you can
achieve a lot more
but if you have trust you also have
opportunity for cheaters and liars
and i don't think that's ever going to
go away
there will be hopefully a minority so
that they don't get in the way and we
studied in these high-end simulations
like what the proportion needs to be
before it is no longer functional
and you can point out that you can
tolerate a few cheaters and a few
liars and the society can still function
and that's probably going to happen um
when we build these systems that
autonomously learn
um the really successful ones are
honest because that's the best way of
getting things done
um but there probably are also
intelligent
agents that find that they can achieve
their goals by
by bending the rules of cheating so
there could be a huge benefit to
uh as opposed to having fixed ai systems
say we build an
agi system and deploying millions of
them
it'd be that are exactly the same
uh there might be a huge benefit to um
introducing sort of from like an
evolution computation perspective a lot
of variation
yeah sort of uh like diversity in all
its forms is beneficial
even if some people are or some
robots are
so like it's it's beneficial to have
that because
i uh because you can't always at
pre-order i
know what's good what's bad but
uh there's that that's a fascinating
absolutely diversity
is the bread and butter i mean if you're
running away you see diversity is the
one fundamental thing you have to have
and absolutely it also it's not always
good diversity
right it may be something that can be
destructive we had in these heinous
simulations we have hyenas that just
are suicidal they just run and get
killed but they
form the basis of those who actually are
really fast
but stop before they get killed and
eventually turn into this mob uh
so there might be something useful there
if it's recombined with something else
right
so i think that as long as we can
tolerate some of that it may turn into
something better
you may change the rules because it's so
much more efficient to do something that
was actually against the rules before
yes
uh and we've seen society change uh over
time quite a bit along those lines that
there were
rules in society that we don't believe
are fair anymore
even though they were you know
considered
proper behavior before yes um so things
are changing and
i think that in that sense i think it's
um it's a good idea to be able to
tolerate some of that
some of that cheating because eventually
we might turn into something
better so yeah i think this is a message
to the trolls and the of the
internet that you two
have a beautiful purpose in this uh
human ecosystem so we
i appreciate you guys watering
quantities yeah moderate quantities
uh so there's a whole field of
artificial life
i don't know if you're connected to this
field if you uh pay attention
is do you think about this kind of thing
uh
is there a impressive demonstration to
you of artificial life do you think of
the agents that you work with in the
evolutionary competition
at perspective as life
and where do you think this is headed
like is there interesting systems that
we'll be creating more and more
that uh make us redefine maybe rethink
about the nature of life different
levels of
definition and goals there then i mean
at some level artificial life
can be considered multi-agent systems
that build a society that again achieves
a goal
and it might be robots that go into a
building and clean it up or or
after an earthquake or something you can
think of that as an artificial life
problem
in some sense um or you can really think
of it
artificial life as a simulation of life
and a tool to understand what life is
and how life evolved
in on earth and like i said in
artificial life conference
there are branches of that conference
sessions of people who really worry
about
molecular designs and and the start of
life like the
like i said primordial soup where
eventually you get something
self-replicating
and they're really trying to build that
um so it's a whole range of
of uh of topics um
and i think that artificial life is a
great tool
uh to understand life and there are
questions like sustainability
um species we're losing species
uh how bad is it is it natural
uh is there a tipping point um
and where are we going i mean like the
hyena evolution we may have
understood that there's a pivotal point
in their evolution they discovered
cooperation and coordination
you know artificialized simulations can
identify that and
maybe encourage things like that um so
and and also societies can be seen as a
form of life itself i mean we're not
talking about biological evolution we
have all evolution of societies maybe
some of the same
phenomena emerging in that uh domain and
unders and having artificial life
simulations and understanding could help
us
build better societies yeah and thinking
from a
meme perspective of of uh from
richard dawkins that
maybe the organisms ideas of the
organisms not the humans
in these societies that from
it's almost like reframing what is
exactly evolving
maybe the interesting the humans aren't
the interesting thing is the contents of
our minds is the interesting thing and
that's what's multiplying
and that's actually multiplying and
evolving in a much faster time scale
and that maybe has more power on the
trajectory of life on earth than
this biological evolution yes the
evolution of these ideas yes
and it's fascinating like i said before
that
we can keep up somehow biologically yeah
we
have we belong to a point where we can
keep up with this
meme evolution literature you know
internet
um we understand dna and we understand
fundamental particles
we didn't start that way i mean thousand
years ago and we haven't evolved
biologically very much but
somehow our minds are able to uh extend
um
and and therefore ai can be seen also as
one such
step that we created and it's our tool
uh and it's part of that meme evolution
that that we create
even if our biological evolution does
not progress as fast
and us humans might only be able to
understand so much we're keeping up
so far or we think we're keeping up so
far but we might need ai systems to
understand
maybe like the physics of the universe
is operating
like a string theory maybe it's
operating in much higher
dimensions maybe we're totally because
of our cognitive
limitations are not able to truly
internalize
the way this world works and so our
limit we're running up against the
limitation of
our own minds and we have to create
these next level organisms like ai
systems
that would be able to understand much
deeper like really understand what it
means to live in a
uh multi-dimensional world that's
outside of the four dimensions the three
of space and
one of them yeah translation and and
generally we can deal with the world
even if you don't understand all the
details we can use computers
yes even though we don't most of us
don't know all the
structures underneath or drive a car i
mean there are many components
especially new cars that you don't quite
fully know but you have the interface
you have an abstraction of it
that allows you to operate it and
utilize it and i think that that's
that's perfectly adequate and we can
build on it and ai can be
play a similar role i have to ask
uh about beautiful artificial life
systems
or evolutionary computation systems uh
cellular automata to me
like i remember it was as a game changer
for me early on in life when i saw
conway's game of life who recently
passed away unfortunately
it's beautiful how much
complexity can emerge from such simple
rules i i just don't
somehow that simplicity is
such a powerful illustration and also
humbling because it feels like
i personally from my perspective
understand almost nothing about
uh this world because because like my
intuition fails completely how
complexity can emerge from such
simplicity like my intuition fails i
think is
the biggest problem i have
do you find systems like that beautiful
is there
do you do you think about cellular
automata because cellular tama don't
really have um
and many other artificial life systems
don't necessarily have an objective
maybe
maybe that's a wrong way to say it it's
almost like
it's just evolving and creating
and there's not even a good definition
of what it means to create something
complex and interesting and surprising
all those words that you said
um is there some some of those systems
you find uh beautiful
yeah yeah and uh similarly evolution
does not have a goal
uh it is responding to uh
current situation uh and so survival
then
if it creates more complexity and
therefore we have something that we
perceive as progress but that's not what
evolution is inherently
said to do uh and yeah that's that's
really fascinating
how how a simple set of rules or simple
uh mappings can um from from
how from such simple mapping complexity
can emerge
so it's a question of emergence and
self-organization uh and
um the game of life is one of the
simplest ones
and very visual and therefore it drives
home the point that it's possible that
non-linear interactions uh and
and this kind of complexity can emerge
emerge from them
and biology and evolution is along the
same lines we
have simple representations dna if you
really think of it
it's not that complex um it's a long
sequence of them there's lots of them
but it's a very simple representation
and similar evolutionary computation
whatever string or tree representation
we have
any operations you know the amount of
code
that's required to manipulate those is
really really little and of course came
alive even less
so how complexity emerges from such
simple principles that's that's
absolutely fascinating um the challenge
is to be able to control it
and guide it and direct it so that it
becomes useful
and like game of life is fascinating to
look at and and evolution all the forms
that come
out is fascinating but can we actually
make it useful
for us and efficient because if you
actually think about
each of the cells in the game of life as
a living organism
there's a lot of death that has to
happen to create anything interesting
yeah
and so i guess the questions for us
humans that are mortal and then life
ends quickly we want to kind of hurry up
and make sure
we make sure we take evolution uh uh
the trajectory that is a little bit more
efficient than uh
the alternatives and that that's one
something we talked about earlier that
evolution computation is very
uh impatient yeah we had we have a goal
we want it
right away whereas this biology has a
lot of time and
and deep deep time and weak pressure and
large populations
uh one great example of of this is the
novelty search
uh so evolutionary computation where you
don't
actually specify a fitness goal
something that is your actual thing that
you want
but you just reward solutions that are
different
from what you've seen before yeah
nothing else yeah and
you know what you actually discover
things that are interesting and useful
that way
um guess danny and joe lemon did this
one study where they actually
tried to evolve walking behavior on
robots and that's actually we talked
about earlier where
your robot actually failed in all kinds
of ways and eventually discovered
something that was
a very efficient walk uh and and it was
because they if
they rewarded things that were different
that you were able to discover something
uh and i think that this is crucial um
because in order to be really different
from what you already have
you have to utilize what is there in a
domain to create something really
different
so you have encoded the uh fundamentals
of your world
and then you make changes to those
fundamentals you get further away
so that's probably what's happening in
these systems of emergence
uh that the fundamentals are there
and when you follow those fundamentals
you get into points and some of those
are actually interesting and useful now
even in that robotic walker simulation
there was a large
set of garbage but among them there were
some of these
you know gems and then those are the
ones that somehow you have to outside
recognize and make useful but these kind
of productive systems
if you code them the right kind of
principles i think that they
that encode the structure of the of the
domain then you will get to these
solutions and the discoveries
it feels like that might also be a good
way to live life so let me ask
do you have advice for young people
today
about how to live life or how to succeed
in their career
or forget career just succeed in life
form an evolutionary computation
perspective yes
yes definitely explore
diversity exploration yeah and i mean
individuals take classes in music
history philosophy
yeah you know math engineering uh
see connections between them travel
you know learn a language i mean all
this diversity is fascinating and we
have it at our fingerprint
fingertips today it's possible you have
to make a bit of an effort because it's
not easy
but the rewards are wonderful um
yeah there's something interesting about
an objective function of new experiences
so try to figure out i mean uh
what what is the maximally new
experience that could have
today and that so like that novelty
optimizing for novelty for some period
of time might be a very interesting way
to sort of uh
expand the the sets of experiences you
had
and uh then ground from that perspective
um like what you what would be the most
fulfilling trajectory through life of
course
the flip side of that this is where i
come from again maybe russian i don't
know
but the the choice has a
choice is a has a detrimental effect i
think
from at least from my mind where
scarcity is has a empowering effect
so if i sort of
if i have very little of something and
only one of that something
i will appreciate it deeply until i came
to
texas recently and i've been picking out
on
delicious incredible meat i've been
fasting a lot so i need to do that again
but
when you fast for a few days that the
first taste of
of a food is is incredible so
the downside of exploration is that
uh somehow maybe maybe you can correct
me but
somehow you don't get to experience
deeply
any one of the particular moments but
that could be a psychology thing
that could be just a very human peculiar
flaw yeah i didn't mean that you
superficially
explore i mean you can explore deeply
yeah so you don't have to
explore 100 things but maybe a few
topics where you can take a deep enough
time
a dive that you gain an understanding
um you yourself have to decide at some
point that this is deep enough
and i i unders i i've obtained what i
can
from this topic uh and now it's time to
move on
and that might take years um people
sometimes switch careers and they may
stay on some
career for a decade and switch to
another one you can do it
you're not pretty determined to stay
where you are but
you know in order to achieve um
something you know 10 000 hours makes
you need 10
000 hours to become an expert on
something uh so you don't have to become
an expert but to even
develop an understanding and gain the
experience that you can use later you
probably have to spend
like i said it's not easy you got to
spend some effort on it
now also at some point then when you
have this diversity and you have these
experiences exploration
you may want to um you may find
something that you can't stay away from
uh like for as it was computers it was
ai it was
you know that you i just have to do it
you know and i uh you know and then
we'll
it will take decades maybe and you are
pursuing it
because you figured out that this is
really exciting and you can bring in
your experiences
and there's nothing wrong with that
either but you asked what's the advice
for young people
that's the expiration part and then
beyond that if
after that expiration you actually can
focus and and build a career
and you know even there you can switch
multiple times but
but i think the diversity exploration is
fundamental to having a
successful career as is concentration
and spending an effort where it matters
and and
but you're in better position to make
that choice when you have done your
homework
so exploration precedes commitment but
both are beautiful
uh so again from an evolutionary
computation perspective
we look at all the agents that had to
die
in order to come up with different
solutions in simulation
what do you think from that individual
agent's perspective is the meaning of it
all
so far as humans you're just one agent
who's going to be dead
unfortunately one day too soon
what do you think is the why
of why that agent came to be
and uh eventually will be no more
is there meaning to it all yeah in
evolution there is meaning
everything is a potential direction
everything is a potential stepping stone
um
not all of them are going to work out
some of them are foundations for
further um improvement
and even those that are perhaps going to
die out
uh where potential energies potential
solutions
in biology we see a lot of species die
off naturally and you know like the
dinosaurs i mean they have a really good
solution for a while
but then it didn't turn out to be not
such a
good solution in the long term uh when
there's an
environmental change you have to have
diversity some other solutions become
better
it doesn't mean that that there was an
attempt it didn't quite work out or last
uh but they're still dinosaurs and
mountains at least they're relatives
uh and they may one day again be useful
who knows so from an individual's
perspective you've got to think of a
bigger picture
that it is a huge engine
that is innovative and these elements
are all
part of it potentially innovations on
their own and also as
as raw material perhaps or um stepping
stones for other things that could come
after
but it still feels from an individual
perspective that i i matter a lot
but even if i'm just a little cog in the
giant machine
well is that just a silly human notion
in uh individualistic society and though
she'll let go of that
do you find beauty in being part of the
giant machine
yeah i think it's meaningful um i think
it adds
purpose to your life that you are part
of something bigger
[Laughter]
that said are you uh do you ponder your
individual agent's mortality do you
do you think about death do you fear
death
well certainly more now than when i was
a
youngster and did skydiving and
paragliding and
you know all these things you've become
wiser
um there is a reason
for this uh life arc that
younger folks are more fearless in many
ways it's part of the exploration
you know they are the they are the
individuals who think hmm
i wonder what's over those mountains or
what if i go really far in that ocean
what would i find
i mean older folks i don't necessarily
think that way
but younger do and it's kind of
counterintuitive so yeah this is uh
and biologically it's like you know you
have limited amount of time what can you
do with it that matters
so you try to you have done your
exploration you
committed to a certain direction and you
become an expert perhaps in it
what can i do that matters uh with with
the
limited resources that i have that's
what how you i think
a lot of people myself included start
thinking later on in their career
and uh like you said leave a bit of a
trace and a bit of an impact even though
after the agent is gone yeah that's the
goal
well this was a fascinating conversation
i don't think there's a better way to
end it
uh thank you so much so first of all i'm
very inspired of how vibrant the
community at ut austin in austin is it's
really exciting for me
uh to see it and this whole field
seems like profound philosophically but
also the path forward for the artificial
intelligence community so
thank you so much for explaining so many
cool things to me today
and for wasting all of your valuable
time with me oh it was a pleasure
thanks i appreciate it thanks for
listening to this conversation with
and thank you to the jordan harbinger
show grammarly
belcampo and indeed check them out in
the description to support this podcast
and now let me leave you with some words
from carl sagan
extinction is the rule survival
is the exception thank you for listening
i hope to see you
next time
you