00:00:00
[Applause]
00:00:08
solo no besties yeah everybody else was scared away I'm afraid yeah or you
00:00:13
scared them away yeah I mean it was uh um challenging prompt interview Steven
00:00:19
Wolfram on stage in 40 minutes so here we go let your winners
00:00:25
ride rman [Music] David
00:00:31
and instead we open sourced it to the fans and they've just gone crazy with it queen
00:00:37
[Music] of um it's a huge honor to talk to
00:00:43
Steven Wolfram creator of Mathematica wolf from Alpha and the Wolf from language the
00:00:48
author of A New Kind of Science the originator of wolf physics project and head of wol
00:00:54
research Steven first used a computer in 1973 and quickly became a leader in the
00:00:59
emerging field of scientific Computing in 1979 he began the construction of SNP
00:01:06
the first modern computer algebra system he published his first scientific paper
00:01:11
at age 15 and had received his PhD in theoretical physics from Caltech by
00:01:17
20 wolfram's early scientific work was mainly in high energy physics Quantum field Theory cosmology and complexity
00:01:24
discovering a number of fundamental connections between computation and nature and inventing such Concepts as
00:01:30
computational irreducibility which we'll talk about today wol firm's work led to a wide range of applications and
00:01:37
provided the main scientific foundation for such initiatives as complexity Theory and artificial life wolm used his
00:01:45
ideas to develop a new Randomness generation system and a new approach to computational fluid dynamics both of
00:01:50
which are now in widespread use the release of wolf from alpha in May of 2009 was a historic step that has
00:01:57
defined a new dimension for computation and AI now relied on by millions of people to compute answers both directly
00:02:04
and through intelligent assistants such as Siri and Alexa and so on among others um so thank you for being here thanks
00:02:11
for having me I worked on um the at the Lawrence Berkeley National Lab at the
00:02:16
center for beam physics for two and a half years and was when I was undergrad and I worked exclusively in Mathematica
00:02:22
as I shared with you the other night so um that's when I first got to know about you and and your work um
00:02:30
who here has seen an interview that Steven's done before just to get a sense okay I want to try and guide the
00:02:36
conversation a little bit so maybe we could start with
00:02:41
computers okay um you talk about this concept of
00:02:47
um computational irreducibility yes maybe you can just and I want to try and
00:02:52
connect with a broad audience yeah yeah in what is computation what is computation okay so at it's at the base
00:03:00
computation is about you specify rules and then you let those rules you figure
00:03:05
out what the consequences of those rules are computers are really good at that you give a little program the computer
00:03:10
will run your program it will generate output I would say that the the kind of the bigger picture of this is how do you
00:03:17
formalize anything you know we can just use words we can talk vaguely about things how do you actually put something
00:03:23
down that is a kind of precise formalism that lets you work out what will happen that's been done in logic it's been done
00:03:29
in mathematic itics it's done in its most general form in computation where the rules can be kind of anything you
00:03:34
can specify to a computer then the question is given that you have the rules is that the end of the story once
00:03:40
you have the rules do you then know everything about what will happen well that's kind of what one would assume kind kind of the the traditional view of
00:03:48
science is once you work out the equations and so on then you're done then you can predict everything you
00:03:54
you're kind of you you you've done all the hard work turns out this is not true turns out that even with with very
00:04:00
simple rules the consequences of those rules can be arbitrarily hard to work out it's kind of like if you just run
00:04:06
the rule step by step step by step it's making some pattern on screen or whatever else you can just run all those
00:04:13
steps see what happens that's all good then you can ask yourself can you jump ahead can you say I know what's going to
00:04:19
happen I know the answer is going to be 42 or something at the end well the point is that that isn't in general
00:04:25
possible that in general you have to go through all the steps to work out what will happen and that's kind of a
00:04:30
fundamental limitation of kind of the sort of prediction in science and something people people have gotten very
00:04:36
used to the idea that with science we can predict everything but from within
00:04:41
science we see this whole phenomenon of computational irreducibility it's related to things like girdles theorem
00:04:47
undecidability halting problem all kinds of other other kinds of ideas but from
00:04:52
within science we see this fundamental limitation this fundamental inability for us to be able to say what will
00:04:58
happen so so we can't just skip ahead in a lot of cases we can't just create simple heuristics or simple solves that
00:05:07
avoid all of the hard work to simulate something to calculate something to come
00:05:13
up with the computational output of something we're trying to figure out in a sense this is a good thing for us
00:05:18
because it's kind of we lead Our Lives time progresses things happen if we could just say we don't need to go
00:05:25
through all of those steps of time we could just say and the end it will be 37 or something that would be a bad feature
00:05:32
for us feeling that it was worthwhile to to sort of lead our lives and see time progress it would be a kind of you don't
00:05:39
really need time to progress you can always just say what the answer will be I mean this kind of idea has has many
00:05:45
consequence I mean for example when it comes to AIS you say well you've got
00:05:50
this AI system and it's doing all kinds of things can you figure out what it will do you might want to figure out
00:05:56
what it will do because you might want to say I never want the AI system to do this very bad thing that you know I'm
00:06:01
trying to prevent so then you say well you know can I work out what it will do can I be sure that these rules that I'm
00:06:07
putting in for the AI system will never have it do this very bad thing well computational irreducibility says you
00:06:13
can't guarantee that you kind of have this trade-off with an AI system you can either say let's you know let's
00:06:19
constrain it a lot then we can know what it will do but then or let's let it sort
00:06:25
of have its way and do what it does if we don't let it have its way it's not
00:06:31
really making use of the computational capabilities that it has right so we can't have this trade-off we either can
00:06:37
understand what's going to happen in which case we don't let our AIS really do what they can do or we uh or we say
00:06:44
okay we're going to run the risk of the AI doing something unexpected so can we just so
00:06:52
AI a lot of what people call AI are predictive models that are effectively
00:06:59
built off of Statistics that make some prediction of what the right next step
00:07:05
in a sequence of things should be whether it's a pixel to generate an image or a series of words to generate a
00:07:12
chat uh response through an llm model like chat GPT or Bard or what have you
00:07:18
um those are statistical models trained on past data are they is that different
00:07:25
than um the the problems in computation that you're talking about about better
00:07:31
understanding the universe the nature of the universe solving bigger problems that AI has its limitation in how we
00:07:37
think and talk about it today and maybe you can connect computation and this idea of AI being this very simple
00:07:43
heuristical statistical thing that just predict stuff right well I mean the the computational universe of possible
00:07:49
programs possible rules is vast and there are I just want to I just want to make sure everyone understands that the
00:07:55
comp to say that the comput computational universe so so you know the set of things you can compute that
00:08:01
you want to try and compute I mean so so people are used to writing programs that are intended for particular human
00:08:06
purposes but let's say you just write a program at random you just put in the program it's a it's a random program
00:08:13
question is what does the random program do so a big thing that I discovered in the 1980s is the thing that greatly
00:08:20
surprised me was even a very simple program can do very complicated things I had assumed that if you want to do
00:08:26
complicated things you would have to set up complicated program turns out that's not true turns out that in nature nature
00:08:34
kind of discovered this trick in a sense you know we see all this complexity in nature it seems like the big origin of
00:08:41
that complexity is just this phenomenon that even a very simple program can do very complicated things so I mean that
00:08:48
that's the so this this sort of universe just give a quick example if you wouldn't mind like yeah yeah so I mean
00:08:53
like three three genes in a genome in in DNA okay so my my favorite example yes
00:08:59
are things called cellular autometer yes and they work like this they have a a line of cells each one is either black
00:09:05
or white it's an infinite line of cells and you have a rule that says you go
00:09:10
down the page making successive uh lines of cells you go down the page and you say the color of a cell on the next line
00:09:17
will be determined by a very simple kind of lookup from the color of the cell right above it into its left and right
00:09:23
okay so very simple setup there are 256 possible rules with just two colors and nearest neighbors you can just look at
00:09:29
what all of them do many of them do very simple things they'll just make some some triangle of black it looks like a
00:09:34
pyramid a triangle when it's done or right right right and then my my all-time favorite it's kind of you turn
00:09:40
the telescope into this computational universe and see what's out there my all-time favorite Discovery is Rule 30
00:09:46
and the numbering of these these rules you you you can specify that rule 30 you started off from one black cell and it
00:09:53
makes this really complicated pattern it makes a pattern where if you just saw it you would say somebody must have gone to
00:09:58
a huge amount it looks designed it looks like there was an architect yes that came in and designed that thing and
00:10:04
that's the only way that thing could have been created cuz it's so beautiful and intricate and right resonates but it
00:10:10
had two simple rules change be black if the left is black and the right is white and be white if right those those those
00:10:16
kinds of things so that's the kind of set up and and for example when you look at rule 30 you look at like the center
00:10:22
column of cells it looks for all practical purposes completely random even though you know that it was really
00:10:28
made from some simple rule When You See It produced it looks random it's kind of
00:10:34
like a a good analogy if you know you know digits of pie people memorize you know 3.14159 it's about as far as I can go
00:10:42
you got you were one ahead of me I that's why build software to do these things we can go to you know I don't
00:10:47
know about that PhD at 20 but uh but you know the the the point
00:10:53
is that the the rule for generating those digits it's you know the ratio of the circumference to diameter of a
00:10:58
circle there's a very definite rule but once you've generated those digits they seem completely random yes it's the same
00:11:05
kind of phenomenon that that you know you can have a simple rule it produces things that look very complicated yes
00:11:10
now so so so that's a simple computer that's a simple
00:11:16
computational exercise by the way it's not such a simple computer because turns out when you kind of try and rank you
00:11:23
know you set up a computer you make it with uh electronics so you might make it you know some some mechanical computer
00:11:29
comp from the past or something like this you ask the question you build a computer of a certain kind how
00:11:34
sophisticated are the computations it can do is it just an adding machine is it just a multiplying machine yes how
00:11:39
far does it get big Discovery from the 1930s is you can make a fixed piece of
00:11:45
Hardware that's capable of running any program that's capable of doing any computation that's the that's the
00:11:52
discovery that launched the possibility of software launched most of modern technology so one might have thought you
00:11:59
have to go to a lot of effort to make a universal computer turns out that's not true it's a thing I call the principle
00:12:05
of computational equivalence which kind of tells one something about how far one has to go and the answer is pretty much
00:12:11
as soon as you see complicated Behavior the chances are you can kind of use that complicated Behavior to do any
00:12:17
computation you want and that's actually that's the reason for this computational irreducibility phenomenon because here's
00:12:24
here's how it works so so let's say you've got some system and it's doing what it does and you're trying to
00:12:29
predict what it's going to do so both the system itself and you as the predictor are computational systems so
00:12:37
then the question is can you the predictor be so much smarter than the system you're predicting you can just
00:12:43
jump ahead and say I know what you're going to do I've got the answer or are you stuck being kind of equivalent in
00:12:49
computational sophistication to the system you're trying to predict right so this principle of computational equivalence says whether you have a
00:12:55
brain or a computer or mathematics or statistics or or whatever else you are really just equivalent in your
00:13:01
computational sophistication to the system that you're trying to predict and that's why you can't make that
00:13:07
prediction that's why computational irreducibility happens but you know you were asking
00:13:12
about AI I mean in the the the thing that we have only just started mining is
00:13:19
this computational universe of all possible programs most programs that we use today were engineered by people
00:13:26
people said I'm going to put this piece in and that piece in and that piece in so now we have a program that can make programs yes well we have we have there
00:13:33
sort of a vast Universe of possible programs we can say if we know what we want the program to do we can just
00:13:39
search this computational universe and find a program that does it often I've done this for many years for many many
00:13:45
purposes just give me an example there so well so so very simple example actually from rule 30 is you want to
00:13:51
make a random number generator you say how do I make something that makes good Randomness well you can just search the
00:13:58
set of possible simple programs and pretty soon you find one that makes good Randomness you ask me why does it make
00:14:04
good Randomness I don't know it's it's not something there's no narrative explanation so now with AI we're
00:14:11
generating a ton of programs and we now have a bigger space of programs or
00:14:17
bigger library to go select from to solve problems or figure stuff out for us is that actually I think AI is sort
00:14:23
of AI is very limited in the computational universe yes I mean the computation this the connection I wanted to make because yeah I mean it's it's
00:14:29
you know the computational universe is all these possible rules we can talk later about whether the I just want to
00:14:34
be clear because you use the term Universe in the sense of all the things and I want to disconnect that from everyone's concept of universe okay just
00:14:41
yes we're going to talk about whether the I hope we're going to talk about whether we talk about the universe and the nature of the universe which we're
00:14:47
going to talk about Consciousness and we're going to smoke weed and then we're going to go to lunch [Laughter] but we're going to talk about you know
00:14:54
whether our physical universe is part of this computational universe but what I when I say computational universe I just
00:15:00
mean this very abstract idea of all these possible programs all the program the library of possible programs yeah
00:15:06
and so there are many many things that those programs can do most of those things are things that we humans just
00:15:12
look at them and say well it's kind of interesting I don't know what the significance of that is they're very nonhuman kinds of things yeah so what
00:15:19
have we done in AI what we've done is we've given for example a large language model we've given it you know 4 billion
00:15:25
web pages for example we've given it kind of the spefic specific parts of essentially the computational universe
00:15:31
that we humans have selected that we care about yes we've shown what what we care about and then what it's doing is
00:15:38
to say Okay I I know what you humans care about so I'm going to make things that are like what you said youve care
00:15:44
about yes and that's a tiny part of the computational universe right just like we saw in Caleb's video his AI said this
00:15:53
video is me telling you that all the stuff you've talked about AI is Terminators and blowing stuff up
00:15:59
and you know that's the limit of what we've we've done right and it can only construct stuff from the limit of what
00:16:05
we've done recorded scen right our data sets so so I mean the thing is so in terms of so give us an example of
00:16:11
something that needs to be computed a a a computational exercise something we got to figure out something
00:16:17
we want to solve outside of what AI is possibly able to to solve for today well
00:16:23
I mean any any of these computationally irreducible problems anything where we're asking yeah so just just just to
00:16:28
connect it yeah yeah yeah right I mean oh gosh to pick an area I mean without without esoteric topology and algebra or
00:16:35
something like yeah yeah right I mean you know okay here's an example so you've got a biological system you've
00:16:40
got a good model great example biology okay so we we've got something we're trying to figure out this collection of
00:16:45
cells it behaves in this way is it going to grow forever and be really bad and make a tumor or is it going to
00:16:51
eventually Halt and stop growing okay that's a classic kind of computational irreducible type of problem where you
00:16:58
know you know we could imp if we knew enough detail we could simulate what every cell is going to do every molecule
00:17:03
every atom every cell every interaction if we knew enough we could simulate each of those steps and there's no easy way
00:17:09
to solve that answer that question you can't jump ahead and say so I know this thing is never going to turn into a tumor for example right and so
00:17:17
simulating the physical Universe whether you're simulating atoms in a cell or
00:17:25
SpaceTime and discrete SpaceTime or non disc great space time itself becomes
00:17:30
this thing that where we don't have a simple heris a simple equation that says based on this condition this is how
00:17:37
things are going to end up but you have to actually go through a lot of calculate calculated steps we haven't known that I mean people have hoped that
00:17:43
you can just write down a formula for how physics works and then work out the answer directly from that that's that
00:17:49
was the big advance I mean in in you know if you go back to Antiquity people were just trying to sort of reason about
00:17:54
how the universe works yes sort of philosophically and then late 1600 unds sort of big Advance we can write down a
00:18:01
mathematical formula we can use calculus we can kind of just write essentially
00:18:06
jump ahead and say what's going to happen in the universe we can make a prediction we can say you know the comet is going to be in this place at this
00:18:12
time and so on so there's all these hard problems that we can't solve with AI we have
00:18:17
today or can we and can you just help me and help everyone understand what are
00:18:23
you excited about with respect to AI what is it that has happened in the last couple of months and years that you were
00:18:29
excited about what does that allow us to do that we couldn't do before using just raw approaches to okay so so I mean
00:18:38
several different things here I mean in in the first thing to say is you know we humans have been interested in a small
00:18:44
part of what's computationally possible AI is reflecting the part that we have been interested in that's sort of AI is
00:18:51
is is doing those kinds of things in terms of what's happened with AI I mean the big thing that happened you know a year ago was the arrival of successful
00:18:58
large language models and uh you know what does that tell us I think that um you know it was a surprise to everybody
00:19:05
including the people who were working on large language models that we kind of got past you know got to this point
00:19:10
where they seemed reasonable to us humans where they were producing text that was reasonable to us humans and not
00:19:16
just completely boring and and irrelevant and so on and I think the you know there's this you know this jump
00:19:22
that happened now in 2012 there was a sort of previous jump in machine learning that happened with images and
00:19:28
things like that image recognition and so on so what's the significance of large language models but one question
00:19:34
is why do they work you know why is it possible to make this neural net that uh
00:19:40
can successfully kind of complete an essay or something and I think the answer is that it's kind of telling us a
00:19:46
piece of science that in a sense we should be embarrassed we hadn't figured out before it's a question of sort of
00:19:51
how do you construct language and we've known forever that there's kind of a
00:19:56
syntactic grammar of language you know noun noun verb noun etc etc etc um but
00:20:03
what you know the llms are showing us is that there is a kind of semantic grammar of language there's a way of putting
00:20:09
together sentences that could make sense yeah and you know that for example people are always impressed that the LMS
00:20:14
have figured out how to quotes reason and I think the what's happening is you know logic is this thing that's kind of
00:20:21
this formalization of everyday language and it's a formalization that was discovered you know by Aristotle and
00:20:27
people in Antiquity and in a sense probably one can think about the way they discovered it they looked at lots
00:20:33
of speeches people had given and they said which ones make sense okay there's a certain pattern of how things are said
00:20:39
that make sense let's formalize that that's logic that's exactly what the LM has done as well it's noticed that there
00:20:45
are these patterns of language that you can kind of use again and that we call
00:20:51
logic or reasoning or something like that so you know I think as a practical matter the you know llms provide this
00:20:57
kind of linguistic user interface we've had kind of graphical user interfaces and so on now we have this linguistic
00:21:03
user interface you say you know you've got some very small set of points you want to make you say I'm going to feed
00:21:10
it to an LM it's going to Puff it up into a big report I'm going to send that report to somebody else they're probably going to feed it to their own llm it's
00:21:16
going to grind it down to some small set of uh set of results it's kind of you
00:21:21
know it's it's allowing one to use language as a transport layer I I think that's a you know there there are a lot
00:21:26
of these practical use cases for for this it's always seemed to me like the rate limiting step in humans is
00:21:33
communication like the rate at which you and I are speaking to one another is pretty low bandwidth like a couple words
00:21:40
a minute or something and yeah the question is what really is communication I mean you know in our brains there are
00:21:45
all these neurons that are firing and you know there's 100 billion of them in each of our brains and and there's a lot of sensory input besides the words that
00:21:51
you're saying that are traveling through vibrations in the air to my ear and that's some information from you but
00:21:56
there's so much more information that the human brain can gather and is building models around all the time making predictions around whether this
00:22:02
light's going to be on or off or that person's going to go to the bathroom or sit back down or what have you um but
00:22:09
you know I think one of the things that's sort of interesting is this you know we've got stuff in our brains we are trying to package up those thoughts
00:22:16
you know the structure of each of our brains is different so the the particular nerve firings are different
00:22:22
but we're trying to package up those thoughts in a kind of transportable way that's what language tends to do it kind
00:22:28
of packages Concepts and that's what these llms have done yes I mean because they're outputting a packet of
00:22:33
communicate to me and yes yes I mean so but is there anything else that's exciting to you from a
00:22:40
computational perspective what else can the AI do and what else okay we learn a
00:22:46
lot from the AIS you know that they're telling us there is a science of llms which is completely not worked out yet
00:22:53
there's kind of a a bulk science of knowledge and things that that the llms are kind of of showing us is there
00:22:59
there's a kind of a science of the semantics of language which llms are showing us is there we haven't found it
00:23:04
yet it's kind of like like we just saw some new piece of Nature and we now get to make science about that kind of
00:23:11
nature now it's not obvious that we can make sort of science where we can tell a narrative story about what's going on
00:23:17
right it could be that we're just we're just sort of dumped into computational irreducibility and we just say it does this because there's this black box that
00:23:24
the that the the training model created that black box we don't know what it does I put a bunch of words in a bunch
00:23:30
of words come out it's amazing now you're saying we're going to try and understand the the nature the graphs the
00:23:36
nature of that box and that'll tell us a little bit something about I think we'll discover that for example we'll discover
00:23:43
that that human language is much less
00:23:48
it's it's it's much simpler to describe human language than we had thought yeah in other words it's showing us rules of
00:23:54
human language that we didn't know yes and that's that's that's an interesting thing now if you ask what what else do
00:23:59
we learn from the AI I'll give you another example of something I was was playing with recently so you know use
00:24:05
image generation uh generative AI for making images as we can now see also
00:24:10
making videos and so on there's this question of of you you go inside the AI and you say you know in inside the AI
00:24:17
you know the concept of a cat is represented by some Vector of a thousand numbers let's say the concept of a dog
00:24:22
and another thousand numbers you just say let's take these vectors of numbers and let's just take arbitrary numbers
00:24:29
what what does the AI think what what is the thing that corresponds to the sort of arbitrary Vector of numbers okay so
00:24:35
you can have these definite Concepts like cat and dog they particular numbers in this sort of space of all possible
00:24:40
Concepts and there's this idea I've been calling it interc concept space what's
00:24:45
between kind of the concept of a cat and the concept of a dog and the answer is there's a huge amount of stuff from even
00:24:52
from a generative AI That's leared from us yeah it is finding these these kind
00:24:58
of interc concept things right that are in between the things for which we have words sort of embarrassing to us that
00:25:05
you know simple estimate simple case if you say what fraction of the space of
00:25:10
all possible concept so to speak is now filled with actual words that we have you of the 50,000 words that are common
00:25:17
in English for example what what um uh you know what yeah that's a that's a didn't know that
00:25:23
number that's interesting yeah right if you're an llm person that's the uh you know when it produces when it says what's the next word going
00:25:30
to be cuz that's what llms are always doing they're just trying to predict the next word what it's doing is it says here's this list of 50,000 numbers which
00:25:36
are the probabilities for each of the possible 50,000 words in English and then it'll pick the most likely one or
00:25:42
or the next most likely one or whatever else but in any case the the um you know this when you ask the question in the
00:25:49
space of sort of all possible Concepts how many do we have words for the answer is it's 1 in 10 600 wow it's like we
00:25:57
have a we have have explored a tiny tiny fraction of even the kind of Concepts that are revealed by the things that
00:26:03
sort of we put out there on the web and so on that's more than there are atoms in the universe yeah they 10 to the 8th
00:26:09
atoms in the universe right 10 to the 8th is very small compared to a lot of the numbers one deals with but um uh you
00:26:15
know I think the yeah it's a we should talk about the
00:26:21
universe that I want to talk about the universe yeah right it's pretty cool so
00:26:26
um people want to talk about AI That's why I wanted to just make sure we we got your point of view so I it's appreciated
00:26:33
um what is the nature of the universe and what do we have wrong
00:26:39
what's that and what do we have wrong yeah what do what does consensus have wrong yeah well so I mean this this is
00:26:46
where this is like where you like let the Reigns go and then you you go yes like we had CH at dinner the other night
00:26:53
so I'm excited yeah it's it's um you know physics as we know it right now was
00:26:59
kind of 100 years ago big advances made in physics three big theories in physics
00:27:05
um general relativity the theory of gravity quantum mechanics the theory of small kinds of things and statistical
00:27:11
mechanics the theory of heat and uh sort of how and the second law of Thermodynamics law of entropy increase
00:27:16
things like this okay so those are three big theories that were invented about 100 years ago physics has been sort of
00:27:22
on a gradual incremental trajectory since then I've been interested in trying to understand kind of what what
00:27:28
could be underneath everything that we see in in the world and I think we figured it out which is kind of exciting
00:27:35
not something I expected to see in my lifetime not something I kind of expected to see for 50 100 more than
00:27:41
that years and kind of the the the number one thing that you say what do people get wrong one question is what is
00:27:48
space well you know people just sort of say space is the continuous thing you put things in any different place in
00:27:54
space people have been arguing about whether space is continuous or discreet since Antiquity discret means that it's
00:28:00
broken up into little pieces yes yes so so you know this argument also happened for matter is matter continuous or
00:28:07
discrete you know you have water is it just this continuous fluid or is it made of discrete kinds of things that got
00:28:13
answered about 120 years ago the answer is water is made of discrete molecules same thing happened for light light is
00:28:20
made of discrete photons space got kind of left out space people still think
00:28:25
it's a continuous kind of thing so the kind of the starting point of things I've tried to figure out is actually
00:28:31
know that's not true space is discret there are atoms of space so to speak not
00:28:36
like physical atoms like hydrogen helium and things they're just uh these sort of
00:28:41
points there're slots where things can fit is that a way you think about it they're they're really just things
00:28:48
Nothing fits in them I mean they're just all that one knows about them they're abstract things all one knows is that
00:28:55
one is different from another so there are the two discret things next to each other there's no next there's no next
00:29:02
there's no space we're about to build space so there is a graph is that a way to think about it a relationship between
00:29:08
two things right well relationship between several things it's kind the giant friend network of the atoms of
00:29:14
space right and that's that's all there is that's what our universe is made of and you know all the things that we
00:29:20
observe you know electrons black holes all these kinds of things they're all just features of this network and that's
00:29:26
it's it's kind of like you might say how could that possibly be the way things are you know if you think about something like water you think about a
00:29:32
little Eddie in the water that Eddie is a thing where you can say there's a definite Eddie that's moving through the
00:29:37
water yet it's made of lots of discrete molecules of water yet we can identify
00:29:43
it as a definite thing and so it is with electrons etc etc in the universe it's
00:29:49
all features of this giant Network and it's it's so that's the way it seems to be help so does each particle as we know
00:29:57
it an electron a proton have a feature that
00:30:03
defines its relative connectedness to other particles and that definition that
00:30:10
little number is what we look at as space as a whole is that a way to think about it no an electron is a pretty big
00:30:17
thing relative to the atoms of space we don't know exactly how big but it's a big floppy thing um the the the the
00:30:24
feature I mean right now in in in it has been assume that electrons are actually infinitesimally small but that doesn't
00:30:30
seem to be true but the thing that that kind of defines something like an electron it's kind of like it's sort of
00:30:36
a topological kind of thing it's like you can have a you can have a piece of string and you can either knot it or you can not knot it right and you know there
00:30:42
are lots of different ways you can make the knot but it's still it's either knotted or it's not knotted and that's there's either an electron or there
00:30:48
isn't an electron so you know kind of the the structure of space it's kind of much like what happens between molecules
00:30:54
and a fluid like water there are all these discret atoms of space and they have these relations to each other and
00:31:01
then if you look kind of at a large scale what are all these I should say by the way that these atoms of space the
00:31:07
main thing that's happening is pieces of this network are getting Rewritten so a little piece of network this is really
00:31:13
important yes I think this is really important because it I'm going to ask the follow on question all right so so so what is time time is this kind of
00:31:20
computational process where this network that represents the change through this
00:31:26
physical Network is time yes yes so so time is a computational phenomenon time
00:31:32
is the is the progressive change of the network one particle touches another one
00:31:37
touches another this thing changes this one this changes this one and sounds to me like you're describing your cellular
00:31:43
automata Yes except it's it's very much like that it's a computational process and it happens to be operating on these
00:31:50
Piper graphs rather than just lines of black the universe itself is a computer yes running a
00:31:58
computational exercise yes well I don't know whether you call it an exercise but yes it's a it is Computing yes it is
00:32:04
Computing and that's that's what the progress of time is time is the is the progress of that computation yes and
00:32:10
that is what people mean when they say are we living in a simulation
00:32:16
no no I think that's a philosophically rather confused thing I mean there's there's there's a little bit deeper in
00:32:21
the rabbit hole that you have to go to understand why that doesn't really make sense
00:32:27
but but um no so so I mean you know then then the question is just like you have
00:32:32
all these molecules bouncing around they make sort of continuous fluids like water you can ask what do all these
00:32:38
atoms of space make turns out they make Einstein's equations that describe
00:32:43
gravity and describe the structure of SpaceTime so that's kind of a kind of a neat thing yes that that's because it's
00:32:49
not been imagined that one could derive the properties of of something
00:32:54
like that from something lower level I mean I should say okay this is this is a big complicated
00:32:59
subject but but um the the um the thing that you guys doing
00:33:07
okay okay good the the the thing that's pretty interesting in all of this is the way
00:33:14
that the nature of us as observers is critical to the kind of thing that we
00:33:21
obser this is the most important thing I've heard in the last year when I watched your interview a few weeks ago so I I just want you to walk everyone
00:33:26
through this statement again cuz this is so important well all right so that the
00:33:32
okay this this um let's let's the the question is what
00:33:39
for for example if we're looking at a bunch of molecules bouncing around in a gas and we say what do we see in those
00:33:47
molecules well we see things like the gas laws pressure volume things like this we see things like the gas tends to
00:33:54
get more random things like this that's what we see because we're observers with certain characteristics if we were
00:34:01
observers who could trace every molecule do all the computations to figure out what all the molecules would do we would
00:34:06
make quite different conclusions about what happens in gases it's because we are observers who are bounded in our
00:34:12
computational capabilities we're not capable of untangling all those all those kind to see all those atoms we
00:34:18
have to look at the whole can of gas we can't look at each atom individually right we can't we can't trace the motion
00:34:25
of each atom individually so ouris is the PV equals nrt the the gas
00:34:32
laws right we look at that gas and we say this is the ratio of temperature which is the total energy of the system
00:34:39
as opposed to the energy of each individual atom and there's a lot of different energy states of all these different atoms so we kind of sum sum
00:34:44
everything up we take an average of a lot of stuff to understand it right and and the point is that people haven't
00:34:50
understood the fact that we sort of have to take that average because what's underneath is computationally ired
00:34:56
usable but we comp are computationally bounded and it turns out that exact same
00:35:03
phenomenon is the origin of SpaceTime and gravity that these all these atoms of space operating in this network it's
00:35:10
all this computationally irreducible process but observers like us observers
00:35:16
who are computationally bounded necessarily observe this kind of uh sort of aggregate Behavior which turns out to
00:35:23
correspond to the structure of SpaceTime same happens in quantum mechanics quantum mechanics is a little bit harder
00:35:28
to understand one of the big features of quantum mechanics is in in classical physics you say you know I throw a ball
00:35:35
it follows a definite trajectory in quantum mechanics you say I do something there are many possible Paths of history
00:35:41
that could be followed we only get to work out things about averages of those paths so it's sort of there are many
00:35:47
different Paths of history that being being pursued in these networks and
00:35:52
things what's happening is that there are many different rewrites to the network that can occur we get these many branches of history and so then the
00:35:58
question is and like a quantum computer is trying to make use of the fact that there are many branches of history that
00:36:04
it can kind of follow in parallel but then the question is well well how do we observe what's actually happening well
00:36:10
we are embedded in this whole system that has all these branching parts of history and so and we're limited well we
00:36:17
we we our brains are full of sort of branching behavior and so on but we are
00:36:23
we're sort of computationally bounded in what we can do we are effectively we we're sort of the question you have
00:36:29
to ask is how does sort of the branching brain perceive the branching universe and as soon as the brain has this sort
00:36:35
of computational boundedness in its characteristics and it also has one other assumption which is it assumes
00:36:41
that it is persistent in time right so it's like we are at every moment we are made of different atoms of space yet we
00:36:49
believe that it's the same us right at every successive moment right as soon as so our concept of consciousness
00:36:56
precludes us from being able to see perhaps a different nature of the universe a different okay yes told you
00:37:02
guys we were going to go like really right so so you know it's kind of like observers like us who are
00:37:08
computationally bounded believe they're persistent in time the big result is that that observers like us inevitably
00:37:15
observe laws of physics like the ones we know yeah and so you know imagine sort of the alien that isn't like us that
00:37:21
isn't computationally bounded doesn't believe it's persistent in time it will observe different law of the universe
00:37:27
and so the laws of the universe are only based on our nature yes but but very
00:37:33
that's what I thought was so interesting right it's a very coarse feature of our nature it's not like we have to know
00:37:38
every detail of us yes the laws of physics only require these and actually I suspect that as we there are other
00:37:45
things that we probably take for granted about the nature of us as observers and as we start putting these in we'll
00:37:51
probably actually find more things that are inevitable about the way the physic that physics works so the the belief as
00:37:57
I use that term I kind of feel um like I'm deluding myself because I am not the
00:38:04
same Adams as I was a second ago a second ago second ago second ago this belief that we have a self talk
00:38:11
about your understanding I know this is a bit farfetched from maybe your specialty I'm sure you have a point of view on then what is this concept of
00:38:18
Consciousness that we have where we think we're persistent in time where we have this concept of self-identity you
00:38:25
know what does this all and how do you think about this notion of Consciousness and the observer in the
00:38:30
context of the universe I'm observing stuff in the universe and I'm think I'm a human body and I'm really I'm a bunch
00:38:36
of atoms floating around with a bunch of other atoms like yeah you know I used to think kind of there was a sort of hierarchy where Consciousness was at the
00:38:42
top but I don't think that anymore I think Consciousness is a very it's just just like the AIS are not doing all
00:38:48
possible computation you know we are actually rather limited in our kind of
00:38:53
uh observation of the universe we have you know we're localized in space we have this belief that we're persistent
00:38:59
in time and so on imagine what you would feel like so to speak if you were much
00:39:04
more extended in the universe if you were in fact one one of the things that we see in our in our models is this
00:39:10
thing we call the ruad which is this kind of this entangled limit of all possible computations and we are every
00:39:17
mind in a sense is just at some small Point some small region in this ruad and
00:39:23
so it it's and you you imagine what happens if if you uh and and by the way
00:39:29
as we as we learn more in science we're effectively expanding in this kind of Ral space where just like we can send
00:39:35
spacecraft out our aggregate Consciousness yes as a species yes yes
00:39:40
so you know just like we can send spacecraft out that Explore More of the physical universe so as we expand our
00:39:47
science as we expand kind of the ideas that we use to describe the universe we're kind of expanding in this Ral
00:39:52
space and so you might say well what happens if we kind of expand you know that that should be the future of
00:39:57
civilization to expand in Ral space to expand kind of our domain of understanding of things it's a do is is
00:40:05
is a shifting Consciousness like hippie type question but like there's this um
00:40:11
there's this uh a guy in the UK named Darren Brown he's a mentalist he puts
00:40:17
these two advertising execs in a room and he tells them hey come up with an ad the name of the company come up with a
00:40:23
logo come up with a a catchphrase he goes out for a few hours comes back pulls off a thing he copied exactly what
00:40:29
they he had written down exactly what they were going to do the way he did it is as they drove over he subliminally put a little image in the cab he had
00:40:35
some kids walk across the street with a logo and they just basically were programmed to Output what he asked them
00:40:43
to do and they they they thought that they were creative Geniuses they're like well these high paid adex look at our genius look at what we did and it always
00:40:49
struck me as like the human is just the unconscious computer we just the you know the the node in the the neural net
00:40:57
that you know takes the input gets sensory programmed output and we're part of the computational exercise is that a
00:41:05
um like a a way to think about this idea that we are part of this broader computation and as we do that this
00:41:11
Consciousness you're the person you mentioned you know they're cheating computational irreducibility so to speak
00:41:16
yes they're they're they're saying I'm going to I'm going to put this thing which is going to be the answer and
00:41:21
that's you know that that the you know the more interesting way to sort of lead life in a sense is just by this process
00:41:28
of letting time progress and sort of this irreducible computation occur and is that what gives you
00:41:35
Joy yeah I think so I mean I think it's it's a it's a funny thing because when you when you kind of think you know sort
00:41:41
of what's underneath the universe and and how kind of all this all these ideas fit together and you kind of realize
00:41:48
that uh it's it's you know I I'm I'm a person who likes people and so it seems
00:41:54
very bizarre that I should be interested in these in these things that kind of deconstruct everything about about
00:42:00
humans and I I realized at some point you know one of the things about doing science that's one of the more difficult
00:42:06
human things about doing science is you have to kind of you know get rid of your prejudices about what might be true and
00:42:12
just you know follow what the science actually says and so I've done that for years and then I realized actually it
00:42:19
turns out the thing that I've done puts humans right back in the middle of the picture you know with these things about
00:42:25
the fact that matters what the Observer is like realizing that you know in this sort of space of possibilities that we
00:42:32
are just you know what we care about is this part that is the result of human history and so on um you guys asked for
00:42:39
more science corner so I hope this fit the bill guys please join me thanking Steven
00:42:45
[Applause] wolf let your winners ride Rainman
00:42:51
[Music] David and said open source it to the
00:42:57
fans and they've just gone crazy with [Music]
00:43:03
it
