Kind: captions
Language: en
[Music]
foreign Hoffman welcome back to the show
thanks a lot Thomas great to be here
dude excited to have you back so I'm
obsessed with the Matrix and the idea
that we're living in a false reality now
I know you don't believe that we are
actually in a simulation but do we
recognize the truth of reality
well our best science tells us that
space-time is not fundamental this is
the conclusion of both physics and
evolution by natural selection so the
physicists tell us that space-time is
doomed it's not fundamental and they're
finding new structures Beyond space-time
like the amplitudehedron that actually
make the math easier in space-time for
the things they need to do and then
evolution by natural selection also
agrees with the physicist that
space-time is not fundamental let's
explain that so when you say that
space-time isn't fundamental what do we
mean exactly in like the simplest or
we'll get into the geeky like deep stuff
in a second but for the audience that
hasn't heard you talk before right what
does that mean well we tend to think of
space and time as the basic level of
reality
everything that could possibly be is
inside space and has some some time
the Big Bang was the start of it all and
who knows what the end will be maybe a
big crunch or just petering out in low
entropy and low temperature we don't
know yet but that
we think or we thought is the basis of
all reality so space and time are the
the basic stage on which all of reality
plays out
that's the weird thing yeah does that
mean that whatever is real and we should
probably give people your
um headset metaverse explanation which
speaks dear to my heart but before we do
that does that mean that whatever is
real is non-physical
well so the word real is a little
slippery so
um
in some sense my headache is real right
because it's a real experience but
um it real in the sense that the
physicists are talking about it when
they thought that space and time are
fundamental they were thinking that this
was the fundamental ground of all
possible realities
like in a Newtonian universe and even in
Einstein it's a point of view Einstein
thought that space and time was the
grounding reality for everything
and now we realize that
the four dimensions of space-time
or even 10 dimensions of string theory
or something like that is
not going deep enough there are
structures entirely Beyond space-time
and entirely Beyond quantum theory
so so these new structures are not like
little structures sitting inside at the
small scale system they can get the
structures yeah people are going to be
super lost so okay the idea of the
headset I think is a really core concept
so yeah uh somebody asked you once like
in the future we're going to start using
different metaphors what metaphors do
you think we're going to use and you
said the metaverse about somebody trying
to contribute to the metaverse my ears
perked up on that one right why will
that become such a useful metaphor for
for this moment and how we perceive
things right because the way that
evolution
speaks on this is it says that our
perceptions of of objects in space and
time is really just like a virtual
reality headset it's there to help you
play the game of life
without knowing what's on the other side
of the headset what's on the other side
what's the hardware and software that's
running the game you don't have to know
that to play the game and in fact if you
were trying to play a game of like Grand
Theft Auto and virtual reality and you
know you had to toggle millions of
voltages per second to drive your car
uh you would lose when you were you know
competing with someone who could just
turn a nice little simple steering wheel
and press on a artificial gas pedal so
evolution
gave us senses that allow us to survive
by hiding the truth and just telling us
how to act so as The evolutionary
theorists would say our senses guide
adaptive Behavior why does natural
selection as a theory predict that
because I understand the theory I guess
well enough at a high level but I never
would have guessed that it actually says
that it makes a prediction anyway that
you whatever is real the only thing I
can tell you that Evolution has selected
for is not that so where like would uh
uh is this something that Darwin himself
saw in his theory or would he be
surprised I think Darwin would be
surprised and in fact many um
evolutionary theorists today
are surprised and so how do we know this
isn't just a kooky interpretation of
natural selection by Donald Hoffman
exactly so
the
the way we pursue this is it turns out
that Darwin's theory has been turned
into a mathematically precise Theory
it's called evolutionary Game Theory so
John Maynard Smith
started that in the 1970s and so we now
have instead of you know Darwin's theory
which is you know it's imprecise in the
sense that it's not a mathematical model
evolutionary Game Theory evolutionary
graph Theory are mathematically precise
so we can now prove theorems and we can
ask technical questions so what is the
probability
that natural selection
would shape any sensory system of any
organism
to reveal any true structures of
objective reality that's a clean
technical question and it turns out that
evolutionary game theory is precise
enough to address that question okay so
I know I've gotten hung up on that a lot
and I think
for people of my cognitive ability we
will have to accept that as the miracle
of this conversation otherwise we'll
derail on that because I don't
understand how his theory can be turned
into a math equation and I worry that
for you to explain it to me would take
an entire semester and cause me to tear
my hair out but so if we can accept
unless you're thinking it looks like you
may have a video I can't I can give a
little hint
it's when we say evolutionary Game
Theory it real think about Game Theory
how do you play Monopoly and win how do
you play various games so it turns out
you can look at different strategies
that someone might have you know I'm
going to go for a park place I'm going
to go for Boardwalk I'm going to try to
there's all different strategy and you
can then write down mathematically okay
if you take this strategy what is the
probability that you will do well
against someone who's taking this other
stream it's all about most Offspring and
the so the strategies are ways to
survive long enough to reproduce and so
you can look at different strategies for
playing the game of life so for example
some organisms will have millions or
thousands of Offspring and
but they don't care about The Offspring
most of them will die
but if one percent of them make it
you're good humans tend to have just a
couple a handful of Offspring and we put
a lot of effort into them so those are
different strategies and so as you look
so some strategies for example in
perception humans really have focused in
our Evolution on vision and hearing and
less on smell and taste and so forth
other organisms focus on things that we
don't even have like echolocation and
bats so different organisms will take
different strategies The Game of Life is
how do I live long enough to reproduce
and how do I raise my Offspring to
maturity no do I do I just make lots of
them and let them fend for themselves
and most of them die but a fraction will
make it or do I make just a few of them
and really help them for 20 or 30 years
until they can go on their own or more
of these days or more of those days so
buy from evolutionary game theories
perspective what is the most successful
creature on planet Earth
uh well probably bacteria
um interesting right well yeah there
there's a lot more bacteria than not a
good answer than us and and maybe
viruses if they're more so from that
point of view
um
right the the winner is the one who
um
you know survives long enough to
reproduce and reproduces for a long
period of time and you know
cyanobacteria been around for billions
of years so you know they're they're
certainly candidates I'm not saying that
they're the final answer but that kind
of thing would be humans are you know
relative newcomers and I actually really
like the theory that humans are
bacteria's way of moving around which is
pretty interesting when you think that
we're outnumbered by the bacteria in our
guts on our skin and all of that stuff
it's pretty interesting I should have
guessed that answer but I didn't but
that makes a lot of sense right right so
so this gives you the idea of when
you're playing a game
there's lots of strategies especially in
a complicated game there's lots of
strategies and it's not that there's
going to be one best strategy it's
rather that if so you know if Tom is
using this strategy what should what
strategy should I use to counter Tom's
strategy and and so forth same thing in
business right it depends on who your
competition is what strategies you're
going to take and what is the governing
system and so forth like with the laws
and so forth that will all determine
your strategy so
you can use Game Theory and turn it into
a tool for studying
Evolution as a game where your bacteria
are trying to play the game of Life One
way humans are playing the game of Life
another way every different organism
every different plant is playing the
game of life with a different kind of
strategy that's really interesting it's
funny I I this is the third time I've
interviewed you and I've never pushed on
this because it there was something
about I couldn't wrap my brain around it
so I'm glad you took the time yeah uh
what's fascinating to me is every
species has its own umvelt yes which is
a really fascinating concept so I looked
this up once and every time I say this
that I think I must be wrong because it
just seems way too far off but humans
are able to perceive point zero zero
three five percent of the
electromagnetic spectrum and I was like
how is that that's so like every
everything that we see and think of as
the the known world is
.0035 percent right that is like
vanishingly small exactly right so our
our window on the on the world is
Trivial compared to what could in
principle be available and so
the question that you can then ask in a
technical fashion is
what is the probability that a strategy
of seeing
truth true structures about objective
reality would that strategy help you to
survive long enough to raise kids
[Music]
and
so we can ask that as a technical
question Evolution has the tools to do
that and the key concept is something
called a fitness payoff so it's Fitness
payoff is like if you're playing a game
there's a certain way that you get
points in the game if you're playing a
video game right you have to shoot
things down or avoid getting hit and to
get points and if you get enough points
you get to the next level of the game
well Fitness payoffs um if you get
enough Fitness payoffs what that
corresponds to is you're surviving long
enough to reproduce
and you don't go to the next level of
the game but your Offspring and your DNA
in your Offspring go to the next level
of the game
so here's the here's the big idea
we can ask these Fitness payoff
functions that govern our Evolution they
do depend on whatever the world is and
the world structure so they do depend on
the world they depend on the organism
you know what's fit for me is not fit
for a benthic fish being 5 000 meters
under the water would kill me it's just
what the benthic fish wants so so the
fitness payoffs depend on the true
structure of the world depends on the
organism you know Hoffman versus a fish
and the um the action feeding fighting
fleeing and mating and and so forth and
you can then ask what is the probability
this is now this is the key technical
question what is the probability that a
randomly chosen Fitness payoff function
that's governing my Evolution has
information
about the true structure of the world
right because it's that fit Evolution
tells us those Fitness payoffs are what
determine how your senses are going to
evolve what's the base assumption there
that the that reality is so complex in
fact I want to press I want to take a
second to really elucidate the example
you have about Grand Theft Auto which I
think is so brilliant
what's actually happening in Grand Theft
Auto is uh electrical currents are
toggling on and off Gates on the
computer and that somehow makes things
happen on your screen that you can
interact with and score points and all
that right
but at like if you look at a chip it is
so complicated that uh trying to like
zap electrodes in the right order
literally impossible right and so
everything that we we as the average
non-computer programmer think of as a
computer is really just the GUI it's the
interface and so you're there at a
really ABS really abstracted level it is
so abstract is to be nonsensical
compared to what's actually happening at
the electrical communication level with
the Machinery itself sending signals to
your TV exactly and if real life has
that same level of complexity then I get
why it would need to be so abstracted
that as to be just nonsensical compared
to what reality really is
something I think breaks some people's
intuition it certainly breaks in my
intuition when I think though that there
has to be some sort of mapping so the
example that you've said many times I
think is really on point is uh if people
are going to make fun of you what they
will say is oh you don't think any of
this is real go ahead and step in front
of that train and see if it kills you
right and of course it's going to so the
representation of the train
is pointing at something
that will change your state from alive
to dead that's right now whether all of
that is is so again abstracted from
what's actually happening at a larger
level I don't even know what to liken it
to
um but nonetheless stepping in front of
a train will flip you from alive to dead
whatever that means in the underlying
reality so do you think at all about
like do you care what it's mapping to or
are you just like eh it doesn't matter
it's too complicated we're not there yet
well I do care and that's why I'm
interested in this particular theorem
right because my interest is I'm seeing
a world of space and time and objects
with colors and shapes and motions how
is is that the true world is that the
the true structure of objective reality
or is this as divorced from reality is
what we're seeing as divorced from the
fundamental reality as my Grand Theft
Auto VR headset is from the voltages
inside the super computer that's running
it that's the that's this simple
question right so when I talk about
things outside of space-time it's just
like suppose someone had played Grand
Theft Auto
since they were one day old and their
parents had left them in a headset their
whole life and when they're 25 the
parents say guess what you've been in
the headset your whole life and and that
that person probably can't even what
could possibly be outside of my headset
I've lived my whole life inside this
headset and you pull it off and you
realize oh wow there's a whole world
that's entirely outside of what you're
in that's the question we're asking has
has Evolution shaped us with just a
little headset of VR headset that that
guides adaptive Behavior but shows us
none of objective reality that's that's
the technical question and the answer is
is very very clear the probability is
one
that we don't see the truth at all
meaning 100
okay so if the probability is a hundred
percent that you are seeing a very false
version right the the thing that that
seems to predict to me is that the
underlying reality is so complicated
that at least in this form you know a
house refer to that in this form it
would with our umvelt our ability to
process data whatever it would not make
sense to try to
[Music]
um
to deal with the reality that it's far
more efficient to create an abstraction
layer but if underlying reality is dead
simple that doesn't seem like it would
hold true so do we just presume that
there is Extreme complexity
well it turns out that the extreme
complexity isn't necessary for this
theorem to be true interesting why would
you need such an elaborate abstraction
if it isn't complicated well so it turns
out when you actually just look at the
math so suppose the world has some
number of states a billion States or 100
States whatever it might so there's some
number of states in the world
and you have some number of states of
perception I can see green red there's
lots of things I can see when you just
do a simple count look at all the
possible functions from the states of
the world to the states of my percentage
just count them so it doesn't the world
does not be complicated it could have
just you know 100 points or a thousand
points when you count those all the
functions and that are the fitness
functions
and ask how many of those functions
actually contain information about the
structures in the in the world it turns
out that very quickly the proportion
goes to zero it's just just so even if
the structure isn't that complicated
maybe there's only one structure in the
world that's all it has like a total
order something you know one is less
than two is less than three is like what
is the probability that that total order
so the world could be very simple it
only has one simple structure total
order and and the world only has you
know maybe a million States so it's not
a very complicated World a million
States
what is the probability that um the
fitness payoff functions that govern my
my Evolution would Preserve
the total order of information would
actually be able to tell me about the
toll order and the math is quite simple
and the answer is zero if it has to
predict something
like so when when I make the base
assumption that it's it's because it is
too complex so to get people I want to
start putting definitions to some of
these words so when you say State let's
say lights on lights off so we all live
where Earth has two states the Sun is up
the sun is down that's one uh
temperature would be another state could
be hot could be cold barometric pressure
could be high could be low could be wet
could be dry like we can just so there's
a lot of different things and so to your
point about the fish they're dealing
with massive pressures right if they
were to come up where there's no
pressure they would disintegrate or not
be able to move or whatever just like we
Crush down to the you know like a tiny
can so they would explode and we would
crush right exactly right right so okay
that when you say States that's one
example exactly I don't understand how
if everything were static it were one
state
that we would need an abstraction layer
to navigate it more effectively than
somebody that sees objective reality so
now I'm going to use an example to
further illustrate what I mean I'm going
to use an example you gave me the first
time you cannot imagine how many times
I've quoted you on this okay
you said uh Tom you have to understand
that objective reality isn't like oh
here's a table and it's got this nice
swirly grain pattern it's the number of
photons reflecting off of that desk and
the the amount of reflectivity and all
that
now irony of ironies as I have started
working in the metaverse you realize how
complicated the visual world is the the
0.0035 of the visual spectrum that we
actually see is insanely complicated to
replicate right right Donald right right
it's the hardest thing I've done in my
life it's crazy and I don't even have to
fully understand it I just have to guide
the team understands it anyway when you
said that I was like whoa what reality
is is very different than how I
experience it so cool complex right so
now I get why the math works out right
but if it isn't complex
so you don't seem to be struggling with
this what is it that you understand that
I don't or what is your base assumption
right that's different than mine
that makes it
makes sense to you that
to achieve maximum Fitness payoff
you would a hundred percent not retain
elements of reality right so so first I
don't deny that I I suspect that reality
is very complicated so so my point isn't
necessary that's not necessary for this
that's right it's just simply accounting
things so if you if you look at all the
functions from one set to another set
like so I have
functions for say I have numbers one
through ten and that's my base set and
I'm going to map them into numbers 1
through 10. so I can map one to three
and two to five and so forth so you know
if you just do okay if you think about
that problem you I could probably figure
okay how many different functions are
there right so you can write the write
down all now you can say okay how many
of those functions have the property
that um you know
they preserve that one is less than two
is less than three and less than four
how many of them scramble that order how
many preserve that order how many
scramble how many contain information
about the one less less than two less
than three less than four so this is
called combinatorics it's a branch of
mathematics 92 percent of people that
set a New Year's goal fail to achieve it
which is why I've created a 90-day
challenge designed specifically to
ensure that you hit your goals you
really can radically transform yourself
just click the link below to join me and
the entire impact Theory university
community to kick off 2023 right with
the impact 90 challenge right guys now
back to the episode
oh I'm unfortunately all too aware of it
because of nfts yes which require you to
understand this because you're making
you have to your point and maybe this is
what you're saying and so maybe I
actually now I'm understanding it let me
walk you through what we had to discover
in nfts okay so you create all these
traits right all these categories I
should say and then within each category
you have maybe 10 possible eyebrows that
it could be eyeball types hairstyles uh
facial hair so on and so forth
that outputs let's say 2 billion right
potential permutations exactly right but
you want to maintain a distribution in
the 10 000 that you're actually going to
show so we were all trying to do the
math and we're working it out and I'm
like
there's no way it's as simple there's
some problem and then we showed it to
physicists and they fell out laughing
and they're like yeah it's not that
simple and so they're like for you to
maintain the right the um the percentage
likelihood to get gold eyes let's say
right out of your two billion
combinations they're like you have to
force it down into this thing which they
called combinatorial or whatever and so
I was like okay
and so that's that really is the point
here that even though I agree with you
that the universe is probably the real
Universe whatever it is is very
complicated I I believe that
combinatorics blow up so quickly got it
by the time you just get to a few
hundred elements you know that as you
found the thing the explosion of
possibilities is so great that when they
ask how many of those possible Fitness
functions would actually be so special
that they contain information about the
structure of where they came from
out of all of the possible Fitness
functions that so it's not an overly
complicated world it's just the number
of potential mapping points and
combinations exactly right very
interesting because evolutionary theory
puts no restriction on the fitness
payoff functions
there could be as many as you can
imagine and there's no restrictions
there's no restriction that says they
have to show you the truth that's not
part of the theory right so until so and
and by the way no one knows how to put
that into the theory right so I mean to
say that it requires that only the
fitness functions that preserve the
truth would be a major revision to
evolutionary theory it would be
unrecognizable
so so when you look then and say okay
every Fitness payoff function is is
equal likely as any other Fitness payoff
function they're all in equal footing
and then you count the ones that
actually have information about the
truth
they go to zero probability in fast
order now there is one I should bring
out there's a group at Yale that has
recently published a paper that's trying
to push back on this
and what they say is if you have say a
bunch of like thousands of Fitness
payoff functions they're all radically
different
then they say that she'll be forced to
um to go to the truth and and the the
argument that they make is that if our
high level cognitions our beliefs our
goals and so forth are not going to
interfere with our perceptions
they claim that then our perceptions
have to map have a single mapping from
the state of the world into the state of
our senses has to be a single mapping
you can't have more so because one thing
I could do with a lot of Fitness
functions is to say well this this
function is different from that one so I
will do this kind of mapping from the
world into my senses with this Fitness
path function then I'll do another
mapping with this Fitness payoff
function and and they say you know if
you're going to have what we call
cognitive impenetrability so what you
believe cognitively cannot affect what
you see okay that's that's the argument
then you must have only one mapping
well it so that's their assumption so
hold on let me make sure I understand
that so they're saying that
basically so that your delusions don't
create the exterior world or at least
your perception of it you have to have
this mapping so that you're actually
detecting and seeing what is real
they're saying that if what you believe
doesn't affect your senses in a
fundamental way yep
then they claim
that that entails that you can only have
one mapping from the world the fitness
the the mapping of your senses from the
whatever the world is into what what
you're seeing the colors and the shapes
and so forth there can only be one map
um that that holds regardless of what
the fitness payoff functions that was
their claim
so and another reason to bring this up
is because this is a recently published
paper the claim is false it's it's
trivial to show counter examples they're
fundamental claim is false please do as
a way just to make sure that I actually
understand what they're saying because
this sounds like what they're trying to
protect against is
um hallucinations basically becoming
subjectively real
right so so I actually think that it's
true probably to a large extent that
what we believe does not really affect
fundamentally what we see so technical
term we use the geek term is cognitive
and penetrability of perception that's
what the philosophers of science will
talk about in cognitive scientists that
are and you can think about scientists
might like this because they'll say look
we want to
use our senses in our experiments I want
to my theory makes a prediction I have
to go look and see if the prediction is
true well if my theory that I'm holding
would change what I see then science
isn't going to really be objective right
I mean if I believe this Theory and it
changes how I see the data then I might
just see the data that confirms the
theory and I can't escape so that's why
there's the philosophy of science has
been very interested in this question
are are high level theoretical beliefs
and just our beliefs as Everyday People
do they get in there and somehow
fundamentally affect how we see the
world and there is a sort of a way you
could say that you know I the way I
believe things does change my world but
not they don't change like the color I
see or the three-dimensional structure
of the cube here that I'm seeing I mean
they might change it in some way but but
not fundamentally like that so that's
the that's the question
and so it's trivial I mean so when the
group at Yale makes this point that you
know if you have lots of different
Fitness payoff functions and you don't
have your high level beliefs interfering
with
the process of perception then you can
only have one one map from the world
into your senses and of course they
don't prove that they just state it
without proof and so it's trivially
false we have made counter examples it's
very very easy to make counter examples
I can design a system in which I have
say two Fitness payoff functions
and I I use onefitnesspal function to
make one map from the world into my
perceptions use the other Fitness
function to make another map and if I
have a system that has no high level
beliefs then the high level beliefs
aren't interfering with it there's the
counter example right there no cognitive
penetration of perception multiple Maps
but then I can add beliefs and say I
know I can have beliefs there as long as
they don't interfere with this mapping
here I could have two two maps why not
so it's they're they're the guys the the
group at Yale they're brilliant
experimentalists and you know one of
them is a really good friend of one of
my collaborators I mean they're they
were post-docs at MIT together and so
forth so they're brilliant
experimentalists but the fundamental
assumption that they're making is just
trivially false and so so then what how
do we see this in our perceptions the
way we see it in our perceptions is
we have probably hundreds of thousands
if not millions of Fitness payoff
functions that are governing our our
Behavior so what do we do with all that
complexity what we do is we group The
Fitness payoff functions into groups
that are similar
and we take the and we make simple
little data structures out of them and
those data structures are what we call
Objects
so this object
is good for drinking can you what what
is a data structure when you say that
it's an object meaning my mind
groups it so that I can differentiate
the cup from the coaster from the desk
what I'm saying is we're making all this
stuff up as a simple way to represent
the fitnesses fitness payoffs and how to
get them so so for example in when
you're playing Grand Theft Auto
you're playing a game
um if you looked inside the super
computer
there there is no red Porsche there is
no steering wheel there is no gas pedal
in some sense those are what I call
Simple data structures they're coding
for you know the gas pedal and pushing
on the gas pedal is coding for who knows
countless millions of voltage changes
happening in in exactly the right
sequence in the computer I have this
trivial data structure gas pedal push on
it that triggers this whole other thing
that I don't want to know about it's
really too complicated so that's what I
mean by these simplifying data
structures my steering wheel is this
simple data structure that I can use to
interact with who knows how many
billions or trillions of voltages and
make them do exactly the right sequence
in the right order
could I say representation instead of
data structure sure absolutely data
structure is the computer science term
so computer scientists would would be
very happy with that but but
representation is is perfectly good and
so the idea then is
what evolution has done from an
evolutionary point of view is it takes
all these Fitness payoff functions that
govern us the governor our survival and
that we need to respect in order to play
the game of life and we organize them so
an apple is an object it's a
representation of a bunch of Fitness
payoffs for example the Apple if I'm
interested in mating
Apple's no good if I'm interested in
eating great if I'm interested in a
weapon
so so I mean I could throw it to
someone's head but it's not going to do
much damage you know if I'm you know so
there's if but if I have a sword
a sword well for for mating no good for
eating not really I mean I could use it
to cut a coconut in half but but I can't
eat this I can't eat the sword for
fighting great but not if you're
fighting against a you know a gun and
things like that so every object and we
can recognize I would say on the order
of 30 or 40 000 different objects basic
kinds of objects so what that indicates
is that Evolution has taken all these
hundreds of thousands maybe millions of
Fitness payoff functions and it's not
making one map from the world into our
senses it's making a bunch of different
maps and those different maps are what
we call Objects and our high level
cognition all it does is I I'm hungry
okay or I won't be looking for tables I
won't be looking for the moon I'll be
looking for apples and bananas and
things like that those data structures
those representations that have high
Fitness payoffs for for the action of
eating and so visual attention paying
attention to different objects is our
way of switching from this
representation of Fitness payoffs to
this representation of Fitness payoffs
as I need to be able to to do to survive
long enough to reproduce and so that's
so this sort of technical but it's the
reason I bring it out is because this is
brand new it's gotten a lot of attention
from Yale and so it's an important thing
from the scientific side to to Really
lay to rest that that you know there's
not one mapping that's required from the
world into our senses by Evolution even
if we assume that our our beliefs don't
you know interfere with our cognition
our cognitions don't interfere with our
perceptions
that doesn't entail that we have to have
one mapping it's just a false assumption
when she let go of that false assumption
then you're opened up to realize that
objects every object is just a data
structure coding for a whole group of
Fitness payoffs and that's how Evolution
deals with this okay so the reason that
I find this so endlessly fascinating is
I
um in trying the whole reason I stepped
in front of the camera in the first
place was I made a very profound change
in my life and I thought hey anybody can
do this but it really is about reframing
the world so recoding re coming up with
new references or seeing the cup in a
different light whatever so it's
interesting so the idea of our beliefs
don't influence our cognition or
influence the mapping to the
um the real world it's probably only at
the margins it's pretty minor as you
said but I think that there is a lot of
um difference in outcome in The Game of
Life as we think about it in a modern
context depending on how you code things
but I've I've struggled with this so at
one point I was going to write a book
and I was working with a Ghostwriter and
I was saying like it doesn't matter
what's true all that matters is that
it's effective and that the way that you
view the world is moving you towards
your goals and this was like at the
height of trump and the Ghost Writer was
like yo I'm not writing that and she was
like you need to tell me that you don't
believe in like a post-truth world and I
was like that's interesting because no I
don't mean just lie and make things up
but what is guiding my decision making
isn't a quest for what's true it's a
quest for what works
and so as I think about Fitness payoffs
I get that I'm going to put a pin in the
following for a second when I hear you
talk it feels like you think the level
of abstraction is is like being in a
game headset versus what the game
machine is doing itself that that is
so different right and so we'll get to
that in a minute because that's what
we're talking about but even like at the
layer of okay I've got my headset on I'm
locked in like even there how you can
influence things by how you perceive
them is interesting and we're living in
a moment where saying post-truth
triggers a lot of things I want to strip
all that away but get people to focus on
because really really truly in life what
you're talking about with Fitness
payoffs is how people should look at
their own belief system of like okay I
believe the way that I tell people to
judge what is true is what is the thing
that allows you to better predict the
outcome of your actions and so if I
believe in gravity that allows me to
better predict how to handle this cup
right because if I hold it over here and
let go and expect it to stay there I'm
going to be very disappointed when it
crashes to the floor and so believing in
gravity even if it's fake is very useful
stepping on the gas pedal even if there
really is no Porsche even if there is no
gas pedal if I'm in the game like just
assuming that that's how it works even
though it isn't true it's a total
abstraction it's going to help you get
towards your goal if your goal is to win
that game so all of that is very
interesting I do think that we can
even take something like synesthesia
where would you say that that's they're
intentionally
um using cognition no but their
perception is like I don't know if you
know who Dave Grohl is but uh drummer
for Nirvana lead singer Foo Fighters and
he is a synesthet of his own
um claim my own admission yeah and he
said that I forget if he sees or shapes
I think there might be shapes and for
him and he said that's why it's so easy
for him to remember songs because they
have these literal shapes yes and so he
just has to remember the sequence of the
shapes and he can play the song and
that I mean that really has an impact
he's able to remember things that I
wouldn't be able to remember for
instance because his perception is being
influenced by the way that his brain
processes data so for whatever reason
two areas of his brain trigger when he
hears something whereas in mine only one
triggers
and
so that to me when I again going back to
why I find this so interesting
that to me says hey I don't know how
much of what you're perceiving is real
but I know that there are consequences
to how you categorize so your idea of
data structures is going to matter a lot
and so if you can categorize something
as shapes and sound it's going to be
easier to remember if you categorize
like for instance the thing I'm always
trying to get people to understand is
if you have what I call the only belief
that matters that you can if you put
time and energy into getting better at
something you actually will get better
right if you believe that then you'll
pursue Improvement if you don't believe
that then you won't because it wouldn't
make any sense right so you miss out on
Fitness payoffs based on your cognitive
assessment of how the world works
right so all of that's fascinating okay
absolutely and important understand
where my brain breaks with your thesis
is how different
what you perceive is and what the world
is like and I know and this is where it
gets hard because I think you would say
we don't know what's under space-time
right
but what's your best guess like as we
strip away
this layer and this might be the time to
talk about Consciousness but I don't
want to lead the witness what what do
you if it isn't space time
stab in the dark for me what the hell is
it well I'll tell you what the
physicists are doing on this because the
physicists are the ones who are saying
space time is not fundamental
so it's there it's a pointer it's a
representation it's a data structure
it's a data structure to something
deeper that's right but it's it happens
to be the human brain which is already a
data structure you're already making
that up exactly right but that data
structure
represents things through space-time
exactly right that's our headset space
time is just our headset and it only
goes down to the is that the plank
length I always hear you quote a a size
plank length is 10 to the minus 33 so
that is where you're quoting right
because that's the smallest thing that
we can measure yeah that's the smallest
thing that's the smallest scale at which
space time has any operational meaning
if you try to go smaller space time
ceases to make any operational sense at
all because gravity insists that below
that things have condensed to too fine
of a point it becomes a black hole
exactly right you create a black hole
Okay so
so and if you think about it and we know
that isn't true like why can't they just
be true smaller than that is a black
hole yay you know well we know it's we
know that at the plank scale you you
um
space time stops and you get you you get
black holes so what's the problem well
black hole is a singularity it means we
don't know what's happening wait so you
get Infinities popping up
um but black holes are real right
they're they're real as a data structure
they're they're they're real stopping
points in our understanding but they're
in the universe well they're um
I know this gets complicated because the
universe is a representation oh yeah so
and so where I want to start Penrose and
others have been studying the properties
of black holes right Penrose won the
Nobel Prize very recently for his his
wonderful work on black holes and so
there's a lot of work that's being done
to understand the properties of black
holes for example the amount of
information you can store in a black
hole doesn't depend on its volume only
the surface area
yeah I don't understand that yeah right
right this is it's very very strange but
that turns out to be true in everyday
space the amount of information that you
can store in this volume here is not
dependent on the volume it depends on
the surface the surface area
that's the universe we live on it's it's
so that's LED people to this holographic
kind of idea oh every word out of your
mouth I'm like we actually are in a
simulation we haven't even talked about
the non local things are not locally
real
right we'll get to that because that's
the new Nobel Prize this year which is
insane and literally just says you're in
a simulation and it's the same as
rendering and when you look at something
it renders when you look away it it
doesn't and we can prove it
mathematically yeah that's right way too
fascinating we'll get to that but first
I want to understand like black holes
the word real gets very slippery in this
conversation right but black holes are
observable yes
so so the idea is that the notion of
space-time at like instead of 10 to the
minus 33 centimeters say 10 to the minus
40 centimeters what would that mean it
does it has no meaning it has there's
nothing you can do with it so so black
holes are fine they're they're objects
there that are at the end point of what
space-time can do
but if we say but I thought space time
was fundamental that means I should be
able to talk about what's happening at
10 to the minus 50 centimeters intended
and you just cannot there's no
operational meaning and in that sense so
you're saying whatever is fundamental
will be able to tell you exactly what's
happening inside of a black hole
well or it will tell you that this whole
framework in which black holes appear is
the wrong framework and thusly black
holes are just a data structure for
something else that is describable once
you get outside once you get out of
space time and and you know it's hard
for us to think outside of space-time
like yeah can we can we beat this point
to death for a second because this one
was a a breakthrough for me when I
realized I always thought of the plank
plonk length as like so infinitesimally
small that like we should all be in awe
and you're like
like that space-time breaks down that
early is just ridiculous and I was like
okay that's a different frame of
reference yeah it's it's a very shallow
data structure if it was 10 to the minus
33 trillion centimeters that it broke
down I'd be I'd be impressed 10 to the
minus 33. we got cheated this is a
really shallow data start it's only four
dimensions I can't even imagine
something in five dimensions
I can't even imagine a new color that
I've never seen before so so we've been
given this really we think that we're in
many cases we think we're the epitome of
intelligence and the the smartest thing
in the universe my feeling is
we've been shortchanged really shallow
data structure only three dimensions of
space one dimension of time we got a
cheap headset and so when that's a fun
way to say it
when data breaks down like that right
what so I always forget the guy's name
so I wrote it down but Nema or Connie
Hamed right
so I've heard you talk about them a lot
so I started doing some research on him
and if I'm understanding what he's
saying correctly is basically when you
have a data structure that falls apart
that early right which was again a total
reframe for me because I thought of that
as like oh my God uh but apparently when
you understand this better you realize
that's that's a pretty early tap out so
when a data structure falls apart that
early that that tells you that it's
proximal right which I'm interpreting as
a it's the finger pointing at the Moon
it is not the moon itself exactly and so
now you know you're looking at a pointer
and so that seems to be the thing that
his whole case rests on for uh space
time being doomed that if your data
structure falls apart that early you
know there's no way this is the
fundamental thing
that's one of the big pointers the other
big pointer a couple other big pointers
he gives is that when you let go of
space-time and you start Computing
particle interactions like two gluons
hit each other and four gluons go
spreading out the kind of thing that
happens at the Large Hadron Collider all
the time if you compute it inside of
space-time
that one I mentioned two gluons in four
gluons out hundreds of pages of algebra
for one interaction
why is it so complicated because it's
the wrong data structure it's an ugly
nasty data structure and the thing that
you're doing the algebra on is in what
way they scatter inside Space time you
have to do to make all the math work out
you have to have these Feynman diagrams
with virtual Parts people are trying
that they're trying to say okay a theory
of everything right which you are saying
does not exist and will never exist but
we'll get to that later right uh so if
there were a Theory of Everything though
we should be able to know everything so
finely that I can tell you oh if they
Collide at this energy with this
directionality it will scatter exactly
like this with these probabilities you
get your probabilities of their of their
scattering okay and so they're just like
oh my God it's a dizzying amount of math
that's right
until you let go of space time and then
that one that I mentioned two blue ones
in four gluons out it's one term
you can compute it by hand it's like
when they hit they'll be a diamond
yeah well because you need to start
talking in shapes right well yeah so so
it's a shape Beyond space-time whose
volumes so yeah it's a shape outside of
space-time
outside of our headset and the volumes
of this shape actually tell you the
probabilities of the various kinds of
particle interactions okay so and so it
turns billions of terms into a handful
of terms and it shows you new symmetry
so that's what the physicists really
love it's simpler math which is great
and then all of a sudden you see new
symmetries that you can't see in space
time okay I'm going to try to draw an
analogy which is already going to break
things but let me see how close I get
you're in Grand Theft Auto right you
step on the gas and you go forward and
we're just like oh my God the math to
predict in what way the car is going to
move when you step on the gas pedal is
ridiculous
but if we were to be actually looking at
the electrical pattern that's stepping
on the gas which would be pressing
buttons on your controller
in a certain context if we understood
that there's a pattern outside of the
headset so in the the PlayStation or the
Xbox there's an electrical pattern
inside of that that looks so if you know
chess and I don't but I'm familiar with
the the idea of chunking so apparently
what Chess Masters do is they're not
looking at the individual pieces on the
board they just know the patterns so
they're like oh that image of where the
pieces are in this order that's this
setup so they've chunked the whole board
into like oh I know where we're at in
the game and I know what the right next
move is so basically what you're saying
is you step on the gas and it gives you
an image of a shape of electrical
patterns outside of the headset
if that's what you're saying I at least
understand
I I don't I could not give you the math
or any of that but I get like this
representation this data structure
which you think of as being real
stepping on the gas and the red portion
goes is actually this
chunk of electrical impulses if we think
of it as a shape or a pattern or a
rhythm or however we're going to think
of it is that what we're saying that
could be a helpful metaphor and I've got
another metaphor that may also try to
help people on the because that's an
important point that you're raising so
suppose here's another way to think
about this suppose that I'm looking at a
video
and are seeing all these pixels
and the pixels are moving in really
complicated ways you know there's the
red pixels and green pixels and light
pixels and dark and I'm just and and
I've and I know that there's something
interesting going on and so I write down
all these equations for the Motions of
these pixels and but but someone says
you know what there is just this I've
got this little Rubik's Cube and I'm all
I'm doing is rotating a Rubik's Cube
and but but you're only seeing the pixel
projection of if you just could see this
3D object you would realize how simple
it is but when you only see the pixels
and see all the then it's oh man I got a
I've got to model all the pixels moving
on my screen how do I do that well if
you can just let go of the screen behind
it there's this unified geometric object
the Rubik's Cube and if you just see oh
it just rotates rigidly that's and that
rigid rotation is the only motion I need
it's a rotation here I have to look at
all the pixels and this pixel I'm paying
attention to the dots rather than the
shape space time it's paying attention
to the dots right so in space time we're
we're stuck on the video screen and
we're trying to model all the pixels
moving around the video screen and what
the physicists have said if you let go
of the video screen
take it off you see that these geometric
objects like that Rubik's Cube