Alien Debate: Sara Walker and Lee Cronin #279

Transcript

00:00:00 I don’t know what it’s like to be an alien.

00:00:01 I would like to know.

00:00:03 Two alien civilizations coexisting on a planet,

00:00:05 what’s that look like exactly?

00:00:07 When you see them and they see you,

00:00:09 you’re assuming they have vision,

00:00:11 they have the ability to construct in 3D and in time.

00:00:14 That’s a lot of assumptions they’re making.

00:00:16 What human level intelligence has done is quite different.

00:00:19 It’s not just that we remember states

00:00:21 that the universe has existed in before,

00:00:23 it’s that we can imagine ones that have never existed

00:00:26 and we can actually make them come into existence.

00:00:29 So you can travel back in time sometimes.

00:00:32 Yes.

00:00:33 You travel forward in time to travel back.

00:00:35 Yes.

00:00:38 The following is a conversation

00:00:40 with Sarah Walker and Lee Cronin.

00:00:42 They have each been on this podcast once before

00:00:45 individually and now for their second time,

00:00:48 they’re here together.

00:00:49 Sarah is an astrobiologist and theoretical physicist.

00:00:53 Lee is a chemist and if I may say so,

00:00:57 the real life manifestation of Rick from Rick and Morty.

00:01:01 They both are interested in how life originates

00:01:04 and develops both life here on earth and alien life,

00:01:08 including intelligent alien civilizations

00:01:11 out there in the cosmos.

00:01:13 They are colleagues and friends

00:01:15 who love to explore, disagree and debate

00:01:18 nuanced points about alien life.

00:01:20 And so we’re calling this an alien debate.

00:01:24 Very few questions to me are as fascinating

00:01:27 as what do aliens look like?

00:01:29 How do we recognize them?

00:01:30 How do we talk to them?

00:01:32 And how do we make sense of life here on earth

00:01:34 in the context of all possible life forms

00:01:37 that are out there?

00:01:38 Treating these questions with the seriousness

00:01:41 and rigor they deserve is what I hope to do

00:01:44 with this conversation and future ones like it.

00:01:48 Our world is shrouded in mystery.

00:01:50 We must first be humble to acknowledge this

00:01:52 and then be bold and diving in

00:01:54 and trying to figure things out anyway.

00:01:57 This is the Lex Friedman podcast.

00:01:59 To support it, please check out our sponsors

00:02:01 in the description.

00:02:02 And now, dear friends, here’s Sarah Walker and Lee Cronin.

00:02:08 First of all, welcome back Sarah, welcome back Lee.

00:02:11 You guys, I’m a huge fan of yours.

00:02:12 You’re incredible people.

00:02:13 I should say thank you to Sarah

00:02:14 for wearing really awesome boots.

00:02:17 We’ll probably overlay a picture later on,

00:02:19 but why the hell didn’t you dress up, Lee?

00:02:21 No, I’m just kidding.

00:02:22 This is me dressed up.

00:02:23 You were saying that you’re pink,

00:02:25 that your thing is pink.

00:02:26 My thing is black and white, the simplicity of it.

00:02:30 Where’s the pink?

00:02:31 When did the pink, when did it hit you

00:02:33 that pink is your color?

00:02:34 I became pink about, I don’t know, actually, maybe 2017.

00:02:41 Did you know me when you first?

00:02:42 I think I met you pre pink.

00:02:44 Yeah, yeah, so about 2017, I think.

00:02:46 I just decided I was boring

00:02:48 and I needed to make a statement

00:02:50 and red was too bright, so I went pink, salmon pink.

00:02:53 Well, I think you were always pink.

00:02:55 You just found yourself in 2017.

00:02:59 There’s an amazing photo of him

00:03:00 where there’s like everybody in their black gown

00:03:02 and he’s just wearing the pink pants.

00:03:03 Oh, that was at the Waggonen University.

00:03:05 It’s totally nuts.

00:03:06 100 year anniversary, they got me to give the plenary

00:03:08 and they didn’t find that outfit for me,

00:03:10 so they were all wearing these silly hats and these gowns

00:03:13 and there was me dressed up in pink

00:03:14 looking like a complete idiot.

00:03:15 We’re definitely gonna have to find that picture

00:03:18 and overlay it, big full screen, slow motion.

00:03:22 All right, let’s talk about aliens.

00:03:23 We’ll find places we disagree and places we agree,

00:03:28 life, intelligence, consciousness, universe, all of that.

00:03:31 Let’s start with a tweet from Neil deGrasse Tyson

00:03:35 stating his skepticism about aliens wanting to visit Earth.

00:03:39 Quote, how egocentric of us to think that space aliens

00:03:44 who have mastered interstellar travel across the galaxy

00:03:47 would give, pardon the French,

00:03:51 would give a shit about humans on Earth.

00:03:54 So let me ask you, would aliens care about visiting Earth,

00:03:57 observing, communicating with humans?

00:04:00 Let’s take a perspective of aliens, maybe Sarah first.

00:04:05 Are we interesting in the whole spectrum

00:04:08 of life in the universe?

00:04:11 I’m completely biased, at least as far as I think right now

00:04:13 we’re the most interesting thing in the universe.

00:04:16 So I would expect based on the intrinsic curiosity

00:04:21 that we have and how much I think that’s deeply related

00:04:25 to the physics of what we are,

00:04:26 that other intelligent aliens would want to seek out

00:04:29 examples of the phenomena they are

00:04:32 to understand themselves better.

00:04:34 And I think that’s kind of a natural thing to want to do.

00:04:37 And I don’t think there’s any kind of judgment

00:04:39 on it being a lesser being or not.

00:04:42 It’s like saying you have nothing to learn

00:04:43 by talking to a baby.

00:04:45 You have lots to learn, probably more than you do

00:04:47 talking to somebody that’s 90.

00:04:48 So yeah, so I think they absolutely would.

00:04:51 So whatever the phenomena is that is human,

00:04:55 there will be an inkling of the same kind of phenomena

00:04:57 within alien species and they will be seeking that same.

00:05:00 I think there’s gotta be some features of us

00:05:02 that are universal.

00:05:03 And I think the ones that are most interesting,

00:05:06 and I hope I live in an interesting universe,

00:05:08 are the ones that are driven by our curiosity

00:05:13 and the fact that our intelligence allows us to do things

00:05:18 that the universe wouldn’t be able to do

00:05:19 without things like us existing.

00:05:22 We’re gonna define a lot of terms.

00:05:23 One of them is interesting.

00:05:25 Yes.

00:05:26 That’s a very interesting term to try to define.

00:05:29 Ali, what do you think?

00:05:30 Are humans interesting for aliens?

00:05:33 Well, let’s take it from our perspective.

00:05:34 We want to go find aliens as a species quite desperately.

00:05:37 So if we put the shoe on the other foot,

00:05:39 of course we’re interesting.

00:05:41 But I’m wondering and assuming

00:05:44 that we’re at the right technological capabilities

00:05:45 to go searching for aliens, then that’s interesting.

00:05:48 So what I mean is,

00:05:50 if there needs to be a massive leap in technology

00:05:52 that we don’t have,

00:05:55 how will aliens prioritize coming to Earth and other places?

00:05:58 But I do think that they would come and find us

00:06:02 because they’d want to find out about our culture,

00:06:04 what things are universal.

00:06:06 I mean, I’m a chemist.

00:06:07 I would say, well, is the chemistry universal, right?

00:06:09 Are the creatures that we’re going to find

00:06:11 making all this commotion,

00:06:13 are they made of the same stuff?

00:06:17 What does their science look like?

00:06:20 Are they off planet yet?

00:06:22 I guess there’s, so I think that Neil deGrasse Tyson

00:06:25 is being slightly pessimistic

00:06:27 and maybe trying to play the tune

00:06:31 that the universe is vast

00:06:33 and it’s not worth them coming here.

00:06:35 I don’t think that,

00:06:37 but I just worry that maybe we don’t have

00:06:39 the ability to talk to them.

00:06:40 We don’t have the universal translator.

00:06:42 We don’t have the right physics,

00:06:43 but sure, they should come.

00:06:45 We are interesting.

00:06:46 I want to know if they exist.

00:06:47 It would make it easier if they just came.

00:06:50 So again, I’m going to use your tweets

00:06:54 like it’s Shakespeare and analyze it.

00:06:56 So Sarah tweeted,

00:06:58 thinking about aliens, thinking about aliens.

00:07:01 So how much do you think aliens

00:07:04 are thinking about other aliens, including humans?

00:07:07 So you said, we humans want to visit.

00:07:13 Like we’re longing to connect with aliens.

00:07:15 Why is that?

00:07:16 Can you introspect that?

00:07:16 Is that an obvious thing that we should be,

00:07:18 like what are we hoping to understand

00:07:20 by meeting aliens exactly?

00:07:23 Asking as an introvert, it’s like,

00:07:25 I ask myself this all the time.

00:07:26 Why go out on a Friday night to meet people?

00:07:30 What are you hoping to find?

00:07:31 I think the curiosity, so when I saw Sarah put that tweet,

00:07:33 I think I answered it actually as well,

00:07:35 which was we are thinking about trying to make contact.

00:07:38 So they almost certainly are,

00:07:41 but maybe there’s a number of classes.

00:07:43 There are those aliens that have not yet made contact

00:07:46 with other aliens like us.

00:07:48 Those aliens have made contact with just one other alien

00:07:51 and maybe it’s an anticlimax and slime, right?

00:07:55 And aliens have made contact

00:07:56 with not just one set of intelligent species, but several.

00:07:59 That must be amazing actually.

00:08:01 Literally there are some place in the universe,

00:08:03 there must be one alien civilization.

00:08:05 Let’s not make contact with not one,

00:08:07 but two other intelligent civilizations.

00:08:10 So they must be thinking about it.

00:08:12 There must be entire degree courses on aliens,

00:08:16 thinking about aliens and universal cultural norms.

00:08:21 Do you think they will survive the meeting?

00:08:24 And by the way, Lee did respond saying,

00:08:26 that’s all the universe wants.

00:08:28 So Sarah said, thinking about aliens, thinking about aliens.

00:08:31 Lee said, that’s all the universe wants.

00:08:34 And then Sarah responded, cheeky universe we live in.

00:08:38 So cheeky is a cheeky version of the word interesting,

00:08:42 all of which we’ll try to define mathematically.

00:08:45 Cheeky might be harder than interesting.

00:08:47 Because there’s humor in that too.

00:08:48 Yes.

00:08:49 I think there’s a mathematical definition of humor,

00:08:51 but we’ll talk about that in a bit.

00:08:52 Oh interesting.

00:08:53 Yeah, sure there is, yeah.

00:08:54 So if you’re a graduate student alien

00:08:57 looking at multiple alien civilizations,

00:09:00 do you think they survive the encounters?

00:09:03 I think there’s a tendency to anthropomorphize

00:09:06 a lot of the discussions about alien life,

00:09:08 which is a really big challenge.

00:09:10 So usually when I’m trying to think about these problems,

00:09:13 I don’t try to think about us as humans,

00:09:16 but us as an example of phenomenon

00:09:18 that exists in the universe that we have yet to explain.

00:09:21 And it doesn’t seem to be the case

00:09:24 that if I think about the features,

00:09:27 I would argue are most universal about that phenomenon,

00:09:30 that there’s any reason to think

00:09:31 that a first encounter with another lineage

00:09:35 or example of life would be antagonistic.

00:09:40 I think, yeah.

00:09:42 And I think there’s this kind of assumption,

00:09:46 I mean, going back to Neil deGrasse Tyson’s quote,

00:09:49 I mean, it kind of bothers me because there’s a,

00:09:51 I mean, I’m a physicist,

00:09:52 so I know we have a lot of egos

00:09:53 about how much we can describe the world,

00:09:56 but that there’s this like,

00:09:57 because we understand fundamental physics so well,

00:09:59 we understand alien life and we can kind of extrapolate,

00:10:02 and I just think that we don’t.

00:10:04 And the quest there is really, you know,

00:10:06 really to understand something totally new

00:10:09 about the universe, and that thing just happens to be us.

00:10:11 I agree, I agree.

00:10:12 There’s something else more profound.

00:10:13 I think Neil was just being, again,

00:10:15 he’s just trying to stir the pot.

00:10:17 I would say from a contingency point of view,

00:10:21 I want to know how many ways

00:10:22 does the universe build structures, build memories, right?

00:10:26 And then I want to know if those memories

00:10:29 can interact with each other.

00:10:30 And if you have two different origins of life

00:10:33 and then origins of intelligence,

00:10:34 and then these things become conscious,

00:10:36 surely you want to go and talk to them

00:10:37 and figure out what commonalities you share.

00:10:41 And it might be that we’re just unable to conceive

00:10:43 of what they’re going to look like.

00:10:44 They’re just going to be completely different,

00:10:46 you know, infrastructure,

00:10:48 but surely we’ll want to go and find out a map.

00:10:50 And surely curiosity is a property

00:10:52 that evolution has made on earth.

00:10:55 And I can’t see any reason

00:10:57 that it won’t happen elsewhere

00:10:58 because curiosity probably exists

00:11:00 because we want to find innovations in the environment.

00:11:03 We want to use that information to help our technology.

00:11:08 And also curiosity is like planning for the future.

00:11:10 Are they going to fight us?

00:11:12 Are we going to be able to trade with them?

00:11:15 So I think that Neil’s just, I don’t know,

00:11:17 maybe, you know, I mean, give a shit.

00:11:19 That’s really, I think that’s really down on earth, right?

00:11:22 How would aliens categorize humans, do you think?

00:11:25 How would we?

00:11:26 So let’s put it the other way around.

00:11:28 Slime category.

00:11:29 Maybe, no, no, no.

00:11:30 Maybe we could, the thing is a bit odd, right?

00:11:33 Look at Instagram, Twitter,

00:11:35 all these people taking selfies.

00:11:36 I mean, does the universe

00:11:38 is the ultimate state of consciousness,

00:11:40 thinking beings that take photographs themselves

00:11:42 and upload them to an internet

00:11:44 with other thinking beings looking at each other’s photos.

00:11:46 So I think that they will be.

00:11:49 What’s wrong with that?

00:11:50 I did not say there was anything wrong with it.

00:11:53 It’s consciousness manifested at scale.

00:11:56 Selfies, Instagram.

00:11:58 It’s like the mirror test at scale.

00:12:00 Yeah, I do think that curiosity

00:12:02 is really the driving force

00:12:04 for why we have our technology, right?

00:12:05 If we weren’t curious, we wouldn’t go out, left the cave.

00:12:08 So I think that,

00:12:11 so I think that Neil’s got it completely wrong, in fact.

00:12:14 Actually, of course they’d want to come here.

00:12:16 It doesn’t mean they are coming here.

00:12:17 We’ve seen evidence for that.

00:12:19 I guess we can argue about that, right?

00:12:21 But I think that we want,

00:12:23 I desperately, and I know that Sarah does too,

00:12:26 but I won’t speak for you, you’re here.

00:12:28 I desperately want to have missions

00:12:30 to look for life in the solar system right now.

00:12:32 I want to map life over the solar system.

00:12:34 And then I want to understand how we can go

00:12:36 and find life as quickly as possible at the nearest stars,

00:12:39 and also at the same time do it in the lab,

00:12:41 just to compensate, you know?

00:12:44 So, sure.

00:12:45 Yeah, just one more point on this.

00:12:47 If you think about sort of what’s driven

00:12:49 the most features of our own evolution as a species,

00:12:53 and try to map that to alien species,

00:12:55 I always think like optimism is what’s gonna get us furthest.

00:12:57 And so I think a lot of people always think

00:13:00 that it’s like war and conflict is gonna be the way

00:13:02 that alien species will expand out into the cosmos.

00:13:06 But if you just look at how we’re doing it

00:13:08 and how we talk about it,

00:13:09 so is our future in space is always, you know,

00:13:12 built from narratives of optimism.

00:13:14 And so it seems to me that if intelligence does get out

00:13:18 in the universe, that it’s gonna be more optimism

00:13:20 and curiosity driving it than war and conflict,

00:13:22 because those things end up crushing you.

00:13:25 So there might be some selective filter.

00:13:28 Of course, this is me being an optimist.

00:13:29 I’m a half full kind of person, but.

00:13:31 Is it obvious that curiosity, not obvious,

00:13:35 but what do you think?

00:13:36 Is curiosity a more powerful force in the universe

00:13:38 than violence and the will to power?

00:13:43 So, because you said you framed curiosity as a way

00:13:47 to also plan on how to avoid violence,

00:13:50 which is an interesting framing of curiosity.

00:13:52 But I could also argue that violence

00:13:55 is a pretty productive way to operate in the world,

00:14:00 which is like, that’s one way to protect yourself.

00:14:02 The best defense is offense.

00:14:05 I’m not qualified to answer this, but I’ll have a go.

00:14:07 I think violence, let’s not talk about violence.

00:14:10 That’s the summary of this podcast.

00:14:12 I would, yeah, maybe, I would, let’s not call it violence,

00:14:16 but I call it erasure.

00:14:17 So if you think about the way evolution works,

00:14:20 or the way, obviously talk about assembly theory,

00:14:23 but I won’t.

00:14:24 So if you say you build,

00:14:26 curiosity allows you to open up avenues, new graphs, right?

00:14:29 So new features you can play.

00:14:33 The ability to erase those things allows you to start again

00:14:36 and do some pruning.

00:14:37 So the universe, I think curiosity gets you furthest.

00:14:39 Curiosity gets you rockets that land.

00:14:41 It gets you robots that can make drugs.

00:14:43 It gets you poetry and art and communication.

00:14:48 And then, I often think wouldn’t it be great in bureaucracy

00:14:52 to have another world war, not literally a world war now,

00:14:54 please no world war, but the equivalent

00:14:57 so we get remove all the admin bureaucracy, right?

00:14:59 All the admin violence, get rid of it and start again.

00:15:02 Do you know what I mean?

00:15:03 Because you get layers and you get redundant systems built.

00:15:06 So actually a reset, let’s not call it violence,

00:15:09 a reset in some aspects of our culture

00:15:13 and our technology allows us to then build

00:15:18 more important things about the,

00:15:19 because how many cookies do I have to click on?

00:15:22 How many extra clicks do I have in the future of my life

00:15:26 that I could remove and a bit of a reset

00:15:29 would allow us to start again.

00:15:33 And maybe that’s how I suppose our encounter aliens will be.

00:15:37 Maybe they will fight with us and say,

00:15:38 oh, we’re not as excited by you as we thought.

00:15:41 We’ll just get rid of you.

00:15:42 So they might want to reset Earth.

00:15:44 Yeah, why not?

00:15:44 To be like, let’s see how the evolution runs again.

00:15:47 This seems like they’ve, there’s nothing new happening here.

00:15:51 They’re observing for a while.

00:15:52 This is just not, let’s keep it more fun.

00:15:55 Let’s start with the fish again.

00:15:59 I like how you equated violence to resetting your cookies.

00:16:04 I suppose that’s the kind of violence.

00:16:06 In this modern world where words are violence,

00:16:08 resetting cookies is the kind of violence.

00:16:10 I don’t know where that came from.

00:16:11 I’m completely, yeah.

00:16:12 That’s poetic, really.

00:16:14 Okay, so let’s talk about life.

00:16:17 What is life?

00:16:19 What is non life?

00:16:20 What is the line between life and non life?

00:16:23 And maybe at any point,

00:16:24 we can pull in ideas of assembly theory.

00:16:27 Like how do we start to try to define life?

00:16:29 And for people listening,

00:16:32 so Sarah identifies as a physicist

00:16:35 and Lee identifies as a chemist.

00:16:37 Of course, they are very interdisciplinary in nature

00:16:41 in general, but so what is life?

00:16:46 Sarah.

00:16:49 I love asking that question

00:16:50 because it’s so absurdly big.

00:16:52 I know, I love it.

00:16:53 It’s my absolute favorite question in the whole universe.

00:16:56 So I think I have three ways of describing it right now.

00:17:00 And I like to say all three of them

00:17:01 because people latch onto different facets of them.

00:17:03 And so the whole idea of what Lee and I are trying to work on

00:17:07 is not to try to define life,

00:17:08 but to try to find a more fundamental theory

00:17:10 that explains what the phenomenon we call life.

00:17:12 And then it should explain certain attributes

00:17:14 and you end up having a really different framing

00:17:15 than the way people usually talk.

00:17:17 So the way I talk about it three different ways.

00:17:20 Life is how information structures matter

00:17:22 across space and time.

00:17:25 Life is, I don’t know, this one’s from you actually,

00:17:29 simple machines constructing more complex machines.

00:17:32 And the other one is the physics of existence,

00:17:35 so to speak, which is life is the mechanism

00:17:38 the universe has to explore the space of what’s possible.

00:17:42 That’s my favorite.

00:17:43 So can I, yeah, yeah, can I add on to that?

00:17:46 Okay, can you say the physics one again?

00:17:49 Physics of existence.

00:17:50 Yeah, the physics of existence.

00:17:52 I don’t know what to call it.

00:17:53 If you think of all the things that could exist,

00:17:55 only certain things do exist.

00:17:57 And I think life is basically the universe’s mechanism

00:18:01 of bringing things into physically existing

00:18:03 in the moment now.

00:18:05 Yeah.

00:18:06 Yeah, and what’s another one?

00:18:08 So we were debating this the other day.

00:18:09 So if you think about a universe that has nothing in it,

00:18:14 that’s kind of hard to conceive of, right?

00:18:16 Because, and this is where physicists really go wrong.

00:18:18 They think of a universe with nothing in it, they can’t.

00:18:20 And you think.

00:18:21 But nonexistence is really hard to think.

00:18:22 Nonexistence, yeah.

00:18:23 And then you think of a universe with everything in it,

00:18:26 that’s really hard.

00:18:28 And you just have this white blob, right?

00:18:30 It’s just everything.

00:18:31 But the fact we have discrete stuff in the universe,

00:18:34 Beyonce planets, so you’ve got stars, space, planet stuff,

00:18:38 right, the boring stuff, but I would define life

00:18:41 or say that life is where there are architectures,

00:18:46 any architectures, and we should stop fixating

00:18:48 on what is building the architectures to start with.

00:18:52 And the fact that the universe has discrete things

00:18:55 and it is completely mind blowing.

00:18:57 If you think about it for one second,

00:19:01 the fact there’s any objects at all,

00:19:03 and there’s, because for me, the object is a proxy

00:19:07 for a machine that built it,

00:19:10 some information being moved around,

00:19:14 actuation, sensing, getting resource,

00:19:18 and building these objects.

00:19:20 So for me, everyone’s been obsessing about the machine,

00:19:23 but I’m like, forget the machine, let’s see the objects.

00:19:27 And I think in a way that assembly theory,

00:19:30 we realized maybe a few months ago

00:19:31 that assembly theory actually does account for the soul

00:19:35 and the objects, not mystically like say,

00:19:37 Sheldrake’s morphic resonance,

00:19:39 or Leibniz’s monadology, seeing souls in things.

00:19:42 But when you see an object, and I’ve said this before,

00:19:45 but this object is evidence of thought,

00:19:49 and then there’s a lineage of those objects.

00:19:51 So I think what is fascinating is that,

00:19:54 you put it much more elegantly,

00:19:55 but the barrier between life and non life

00:19:58 is accruing enough memories to then actuate.

00:20:01 So what that means is there are contingency,

00:20:04 there are things that happen in the universe get trapped,

00:20:06 these memories then have a causal effect on the future.

00:20:10 And then when you get those concentrated in a machine,

00:20:12 and you’re actually able in real time,

00:20:14 able to integrate the past, the present with the future,

00:20:19 and do stuff, that’s when you are most alive.

00:20:22 You being the machine.

00:20:23 Yes.

00:20:24 Wait a minute, why is the object,

00:20:26 so one of the ways to define life,

00:20:29 that Sarah said, is simple machines

00:20:31 creating complex machines.

00:20:34 So there’s a million questions there.

00:20:37 So how the hell does a simple machine

00:20:39 create a complex machine?

00:20:41 By mutation.

00:20:42 So this is what we were talking about at the beginning,

00:20:44 is you have the minimum replicator, so a molecule.

00:20:46 So this is what I was trying to convince Sarah

00:20:47 of the mechanism get there years ago, I think,

00:20:49 but then you’ve been building on it and saying,

00:20:52 you have a molecule that can copy itself,

00:20:55 but then there has to be some variability,

00:20:58 otherwise it’s not gonna get more functional.

00:21:00 So you need to add bits on.

00:21:02 So you have a minimum molecule that can copy itself,

00:21:04 but then it can add bits on,

00:21:05 and that can be copied as well,

00:21:07 and those add ons can give you additional function,

00:21:12 to be able to acquire more stuff to exist.

00:21:15 So existence is weird,

00:21:18 but the fact that there is existence is why there is life,

00:21:21 and that’s why I realized a few days ago

00:21:23 that there must be, that’s why alien life

00:21:25 must be everywhere, because there is existence.

00:21:28 Is there a conservation of cheeky stuff happening?

00:21:33 So how can you keep injecting more complex things?

00:21:38 Doesn’t the machine that creates the object

00:21:41 need to be as or more powerful than the things it creates?

00:21:48 So how can you get complexity from simplicity?

00:21:51 So the way you get complexity from simplicity

00:21:54 is that you, I’m just making this up,

00:21:57 but this is kind of my notion,

00:21:59 and you have a large volume of stuff,

00:22:00 so you’re able to get seeds, if you like,

00:22:03 random cues from the environment.

00:22:05 So you just use those objects

00:22:07 to basically write on your tape, ones and zeros, whatever.

00:22:11 And that is necessarily rich, complex, okay?

00:22:17 But it has a low assembliness,

00:22:19 but even though it has a high assembly number,

00:22:21 we can talk about that.

00:22:22 But then when you start to then integrate

00:22:24 that all into a smaller volume, as over time,

00:22:28 and you become more autonomous,

00:22:31 you then make the transition.

00:22:33 I don’t know what you think about that.

00:22:34 I think the easiest way to think about it is actually,

00:22:38 which I know is a concept you hate, but I also hate,

00:22:40 which is entropy, but people are more familiar with entropy

00:22:43 than what we talk about in assembly theory.

00:22:45 And also the idea that, like, say physics as we know it

00:22:50 involves objects that don’t exist across time,

00:22:52 or as we would say, low memory objects.

00:22:55 So one of the key distinctions that is…

00:22:58 Low memory objects.

00:22:59 Yeah, so physics is all…

00:23:01 Physicists are low memory objects.

00:23:02 Low memory objects.

00:23:05 Physicists are creators of low memory objects

00:23:08 or manipulators of low memory objects.

00:23:10 Absolutely.

00:23:11 It’s a very nice way of putting it.

00:23:13 Okay, sorry, go ahead.

00:23:14 Yeah, no, it’s okay.

00:23:14 Sorry to keep interrupting.

00:23:15 No, no, no, it’s fine.

00:23:16 I like it too, it’s very funny.

00:23:18 But I think it’s a good way of phrasing it

00:23:20 because I think this kind of idea we have

00:23:22 in assembly theory is that, you know,

00:23:25 physics as we know it has basically removed time

00:23:27 as being a physical observable of an object.

00:23:31 And the argument I would make is that

00:23:33 when you look at things like water bottles or us,

00:23:36 we’re actually things that exist

00:23:38 that have a large extent in time.

00:23:40 So we actually have a physical size and time,

00:23:44 and we measure that with something called

00:23:46 the assembly index in molecules,

00:23:49 but presumably everyone should have sort of a,

00:23:53 do you want to explain what assembly?

00:23:54 Yeah, let’s, you know what?

00:23:56 Let’s step back and start at the beginning.

00:23:59 What is assembly theory?

00:24:01 Lee sent me some slides.

00:24:02 There’s a big sexy paper coming out probably.

00:24:05 Maybe, I don’t know.

00:24:07 We’ve almost finished it.

00:24:10 Almost, almost finished it.

00:24:11 That’s also a summary of science.

00:24:13 We’re almost done.

00:24:14 Yes.

00:24:15 Well, no, no, we’re almost done.

00:24:16 It’s the history of science.

00:24:17 We are ready to start an interesting discussion

00:24:20 with our peers.

00:24:21 Right.

00:24:22 You’re the machine that created the object,

00:24:24 and we’ll see what the object takes us.

00:24:26 All right, so what is assembly theory?

00:24:29 Yeah, well, I think the easiest way

00:24:30 for people to understand it is to think about

00:24:33 assembly and molecules,

00:24:35 although the theory is very general.

00:24:36 It doesn’t just apply to molecules.

00:24:38 And this was really Lee’s insight,

00:24:39 so it’s kind of funny that I’m explaining it, but.

00:24:42 I’ll mark you.

00:24:43 Okay, all right, I’m ready, I’m ready.

00:24:45 You have to tell me where I get the check marks minus,

00:24:47 but.

00:24:48 It’s your theory as well.

00:24:49 Yeah, I know.

00:24:49 But imagine a molecule,

00:24:51 and then you can break the molecule apart

00:24:54 into elementary building blocks.

00:24:55 They happen to be bonds.

00:24:57 And then you can think of all the ways,

00:24:58 for molecular assembly theory,

00:24:59 you can think of all the ways

00:25:00 of building up the original molecule.

00:25:02 So there’s all these paths that you can assemble it.

00:25:04 And the sort of rules of assembly is

00:25:06 you can use pieces that have been generated already.

00:25:09 So it has this kind of recursive property to it.

00:25:11 And so that’s where kind of memory

00:25:13 comes into assembly theory.

00:25:15 And then the assembly index is

00:25:17 the shortest path in that space.

00:25:19 So it’s supposed to be the minimal amount of history

00:25:21 that the universe has to undergo

00:25:23 in order to assemble that particular object.

00:25:25 And the reason that this is significant is

00:25:27 we figured out how to measure that

00:25:30 with a mass spec in the lab.

00:25:33 And we had this conjecture

00:25:35 that if that minimal number of steps

00:25:36 was sufficiently large,

00:25:38 it would indicate that you required a machine

00:25:40 or a system that had information

00:25:41 about how to assemble that specific object

00:25:43 because the combinatorial space of possibilities

00:25:45 is getting exponentially large

00:25:47 as the assembly index is increasing.

00:25:49 So it’s just, sorry to interrupt,

00:25:50 so that means there’s a sufficiently high assembly index

00:25:55 that if observed in an object

00:25:59 is an indicator that something lifelike created it

00:26:04 or is the object itself lifelike?

00:26:07 Both.

00:26:08 But you might want to make the distinction

00:26:10 that a water bottle is not life,

00:26:12 but it would still be a signature

00:26:14 that you were in that domain of physics

00:26:16 and that I might be alive.

00:26:19 So there will be potentially a lot of arguments

00:26:22 about where the line, at which assembly index

00:26:26 does interesting stuff start to happen.

00:26:28 The point is we can make all the arguments,

00:26:30 but it should be experimentally observable

00:26:32 and Lee can talk more about that part of it.

00:26:34 But the point I want to make about it is

00:26:36 there was always this intuition that I had

00:26:39 that there should be some complexity threshold

00:26:41 in the universe above which you would start to say

00:26:44 whatever physics governs life actually becomes operative.

00:26:46 And I think about it a little bit

00:26:48 like we have Planck’s constant,

00:26:50 and we have the fine structure constant.

00:26:52 And then this sort of assembly threshold

00:26:54 is basically another sort of potentially constant of nature.

00:26:59 It might depend on the specific features of the system,

00:27:02 but which we debate about sometimes.

00:27:04 But then when you’re past that,

00:27:07 you have to have some other explanation

00:27:10 than the current explanations we have in physics,

00:27:11 because now you’re in high memory.

00:27:15 The things actually require time for them to exist

00:27:18 and time becomes a physical variable.

00:27:20 The path to the creation of the object is the memory.

00:27:23 So you need to consider that.

00:27:25 Yeah, but the point is that’s a feature of the object.

00:27:29 So when I think of all the things in this room,

00:27:34 we see the projection of them as a water bottle,

00:27:37 but assembly theory would say that this is a causal graph

00:27:39 of all the ways the universe can create this thing.

00:27:41 That’s what it is as an object.

00:27:43 And we’re all interacting, a causal graph.

00:27:45 And most of the creativity in the biosphere

00:27:47 is because a lot of the objects that exist now

00:27:50 are huge in their structure across time.

00:27:52 Four billion years of evolution to get to us.

00:27:55 Is it possible to look at me

00:27:57 and infer the history that led to me?

00:28:01 If you, you as an individual might be hard.

00:28:04 You as a representative of a population of objects

00:28:09 that have high assembly

00:28:10 with similar causal history and structure

00:28:12 that you can communicate with, i.e. other humans,

00:28:14 you can infer a lot probably.

00:28:15 Yeah, also with them.

00:28:17 Which we do genomically even.

00:28:18 I mean, it’s not like,

00:28:19 we have a lot of information in us,

00:28:21 we can reconstruct histories from.

00:28:23 Assembly is saying something slightly deeper.

00:28:24 Yeah, one thing to add.

00:28:25 I mean, it’s not just about the object,

00:28:26 but the objects occur

00:28:28 and not just objects with high assembly number,

00:28:30 because you can have random things

00:28:32 that have a high assembly number,

00:28:33 but they must have,

00:28:34 there must be a number of identical copies.

00:28:36 So you know you’re getting away from the random,

00:28:38 because you could take a snapshot.

00:28:40 This is why, it’s not I hate entropy,

00:28:42 I love entropy when used correctly,

00:28:44 but it’s about the problem of entropy,

00:28:46 you have to have a labeler.

00:28:47 And so you can label the beginning and the end,

00:28:50 the start and the finish, you know,

00:28:52 where what you can do in assembly is say,

00:28:53 oh, I have a number of objects in abundance.

00:28:56 They all have these features.

00:28:58 And then you can infer.

00:28:59 And one of the things that we debated a lot,

00:29:01 particularly during lockdown,

00:29:02 because I almost went insane trying to crush the,

00:29:05 produce the assembly equation.

00:29:06 So we came up with the assembly equation.

00:29:08 I had, just imagine this.

00:29:09 So you have a string where,

00:29:13 actually it makes me sick trying to remember it.

00:29:16 It was so, it did my head in for a long time.

00:29:18 Yeah, because I couldn’t,

00:29:21 so if you just have a string of say words,

00:29:23 say, you know, a series of words, series of letters.

00:29:27 So you just have A, A, A, B, B, B, C, C, C, D, D, D.

00:29:29 And you find that object and you just have four A’s,

00:29:32 four B’s, four C’s, four D’s together, boom.

00:29:35 Then, and that you measured that.

00:29:38 So you physically measured that string of letters.

00:29:40 Then what you could do is you can infer sub graphs

00:29:44 of maybe the four A’s, the four B’s,

00:29:46 the four C’s and the four C’s,

00:29:47 but you don’t see them in the real world.

00:29:50 You just infer them.

00:29:51 And I really got stuck with that

00:29:53 because there’s a problem to try and work out

00:29:55 what’s the difference between a long,

00:29:57 you know, a physical object

00:29:59 and this assembly space of the objects

00:30:01 that we realized the best way to put that is infer in time.

00:30:05 So although we can’t infer your entire history,

00:30:08 we know at some point the four A’s were made,

00:30:10 the four B’s was made, the four C’s were made,

00:30:12 the four D’s were made and they all got added together.

00:30:15 And that’s one really interesting thing

00:30:17 that’s come out of the theory,

00:30:19 but the killer when we knew we were going beyond

00:30:24 beyond standard complexity theories,

00:30:26 but incredibly successful is that we realized

00:30:29 we could start to measure these things

00:30:31 for real across domains.

00:30:33 So the assembly index is actually an intrinsic property

00:30:36 of all stuff that you can break into components,

00:30:41 particularly molecules are good

00:30:42 because you can break them up

00:30:44 into smaller molecules, into atoms.

00:30:47 The challenge will be making that more general

00:30:49 across all the domains,

00:30:50 but we’re working on it right now

00:30:51 and I think the theory will do that.

00:30:53 So components, domains,

00:30:54 so you’re talking about basically measuring

00:30:58 the complexity of an object in what,

00:31:00 biology, chemistry, physics,

00:31:03 that’s what you mean by domains.

00:31:05 Complexity of tests.

00:31:06 Sociology. Complexity of computers.

00:31:08 Complexity of memes, you know.

00:31:10 Memes? Yep.

00:31:11 What is that, ideas?

00:31:13 Yeah, I mean, so one of the.

00:31:15 Ideas are objects in assembly theory.

00:31:16 Yeah. They are.

00:31:17 They’re physical things.

00:31:18 They’re just features of the causal graph.

00:31:20 I mean, the fact that I can talk to you right now

00:31:22 is because we’re exchanging structure

00:31:23 of our assembly space.

00:31:27 So conversation is the exchanging structures

00:31:32 in assembly space.

00:31:33 What is assembly space?

00:31:34 When I started working on origins of life,

00:31:36 I was writing about something called top down causation,

00:31:39 which a lot of like philosophers are interested in

00:31:42 and people that worry about the mind body problem.

00:31:44 But the whole idea is, you know,

00:31:45 if we have, you know, the microscopic world of physics

00:31:49 is causally complete,

00:31:50 it seems like there’s no room for higher level causes

00:31:53 like our thoughts to actually have any impact on the world.

00:31:56 And that didn’t, that seems problematic

00:31:59 when you get to studying life and mind

00:32:01 because it does seem that quote unquote,

00:32:03 emergent properties do matter to matter.

00:32:08 And then there’s this other sort of paradoxical situation

00:32:10 where information looks like it’s disembodied.

00:32:12 So we talk about information,

00:32:14 like it can just move from any physical system

00:32:16 to any other physical system.

00:32:17 And it doesn’t require,

00:32:20 like you don’t have to specify anything about the substrate

00:32:22 to talk about information.

00:32:24 And then there’s also the way we talk about mathematics

00:32:27 is also disembodied, right?

00:32:28 Like the platonic world of forms

00:32:31 and I think all of those things are hinging

00:32:34 that we really don’t know how to think about abstractions

00:32:37 as physical things.

00:32:39 And really, I think what assembly theory is pointing to

00:32:43 is what we’re missing there is the dimension of time.

00:32:46 And if you actually look at an object

00:32:48 being extended across time,

00:32:51 what we call information and the things that look abstract

00:32:54 are things that are entangled

00:32:55 in the histories of those objects.

00:32:57 They’re features of the overlapping assembly space.

00:32:59 So they look abstract because they’re not

00:33:02 part of the current structure,

00:33:05 but they’re part of the structure

00:33:06 if you thought about it as like the philosophical concept

00:33:09 of a hyperobject, an object that’s too big in time

00:33:11 for us to actually to resolve.

00:33:13 And so I think information is physical.

00:33:15 It’s just physical in time, not in space.

00:33:18 Too hyperobject, too difficult for us to resolve.

00:33:22 So we’re supposed to think about of life

00:33:26 as this thing that stretches through time

00:33:28 and there’s a causation chain that led to that thing.

00:33:32 And then you’re trying to measure something

00:33:34 with the assembly index about properties of that.

00:33:36 The assembly index is the ordering,

00:33:39 like you could think of it as like a partial ordering

00:33:41 of all the things that can happen.

00:33:43 So in thermodynamics, we coarse grain things

00:33:46 by temperature and pressure.

00:33:48 In assembly theory, we coarse grain

00:33:50 by the number of copies of an object

00:33:52 and the assembly index, which is basically,

00:33:53 if you think of the space of all possible things,

00:33:56 it’s like a depth of how far you’ve gone into that space

00:33:58 and how much time was required to get there.

00:34:00 In the shortest possible version.

00:34:02 The shortest possible version.

00:34:03 Not average, because can’t you just 3D?

00:34:07 You’re gonna kill me with that question.

00:34:09 Not 3D, can’t you always 3D print the thing?

00:34:13 Let’s like stab him in the heart.

00:34:14 No, because I had such a fight.

00:34:16 So Sarah’s team and my team are writing this paper

00:34:18 at the moment and.

00:34:20 It’s so funny.

00:34:21 I think we kind of share the, at the beginning,

00:34:23 you were like, no, that’s not right.

00:34:23 Oh yeah, that’s right.

00:34:24 And we’re doing this for a bit.

00:34:25 And then the problem is when you build a theory

00:34:27 and build the intuition,

00:34:29 there’s some certain features, right,

00:34:31 of the theory that almost felt like

00:34:34 I was being religious about saying,

00:34:35 right, you have to do this.

00:34:37 A good assembly theorist does this, does this, does this.

00:34:40 And Sarah’s postdoc, Daniel, and my postdoc, Abhishek,

00:34:44 and they were both.

00:34:45 They’re both brilliant.

00:34:46 They’re brilliant, but they were like,

00:34:47 no, we don’t buy that.

00:34:49 And I was like, it is, they were like,

00:34:52 well, Lee, actually, I thought you were the first

00:34:55 to say that, you know, you can’t,

00:34:56 if you can’t explain it, it doesn’t,

00:34:58 and you can’t do an experiment that doesn’t exist.

00:35:00 And that saved me.

00:35:00 And I said to Abhishek,

00:35:01 Abhishek’s my postdoc in Glasgow,

00:35:03 Daniel is Sarah’s postdoc in ASU.

00:35:06 I was like, I have the experimental data.

00:35:08 So when I basically take the molecules

00:35:11 and chop them up in the mass spec,

00:35:12 the assembly number is never the average.

00:35:14 It’s always the shortest.

00:35:15 It’s an intrinsic property.

00:35:16 And then the penny drop for Abhishek said, okay.

00:35:19 So I had these things that we had to believe

00:35:22 to start with or to trust,

00:35:23 and then we’d done the math and it comes out.

00:35:24 And they now have the shortest path, actually.

00:35:26 It’s up, it explains why the shortest path.

00:35:29 Here’s why the shortest path is important, not the average.

00:35:32 The shortest path needs you to identify

00:35:34 when the universe has basically got a memory,

00:35:37 not an average.

00:35:38 So what you want to be able to do is to say,

00:35:40 what is the minimum number of features

00:35:43 that I want to be able to see in the universe?

00:35:45 When I find those features,

00:35:46 I know the universe has had a coherent memory

00:35:50 and is basically alive.

00:35:53 And so that gives you the lower bound.

00:35:56 So that’s like, of course there’s going to be other paths.

00:35:59 We can be more ridiculous, right?

00:36:01 We can have other parts, but it’s just the minimum.

00:36:03 So probabilistically at the beginning,

00:36:06 because assembly theory was built

00:36:07 as a measure for biosignatures, I needed to go there.

00:36:11 And then I realized it was intrinsic.

00:36:13 And then Sarah realized it was intrinsic

00:36:15 and these hyperobjects were coming.

00:36:16 And we were kind of fusing that notions together.

00:36:19 And then the team were like, yeah,

00:36:21 but if I have enough energy and I have enough resources,

00:36:25 I might not take the shortest path.

00:36:27 I might go a bit longer.

00:36:28 I might take a really long path

00:36:30 because it allows me then to do something else.

00:36:33 So what you can do is, let’s say

00:36:34 I’ve got two different objects, A and B,

00:36:37 and they both have different shortest paths to get them.

00:36:40 But then if you want to make A and B together,

00:36:43 they will have a compromise.

00:36:44 So in the joint assembly space, that might be an average,

00:36:48 but actually it’s the shortest way

00:36:50 you can make both A and B

00:36:52 with a minimum amount of resource in time.

00:36:54 So suddenly you then layer these things up.

00:36:56 And so the average becomes not important,

00:36:59 but as you literally overlap those sets,

00:37:03 you get a new shortest path.

00:37:05 And so what we realized time and time again

00:37:07 when we’re doing the math,

00:37:08 the shortest path is intrinsic, is fundamental,

00:37:11 and is measurable, which is kind of mind blowing.

00:37:14 So what we’re talking about, some basic ingredients,

00:37:17 maybe we’ll talk about that, what those basic ingredients

00:37:20 could be and how many steps, when you say shortest path,

00:37:23 how many steps it takes to turn those basic ingredients

00:37:28 into the final meal.

00:37:32 So how to make a, what’s the shortest way to make a pizza?

00:37:35 Or a pie.

00:37:36 Or a pie.

00:37:37 An apple pie.

00:37:38 That’s right.

00:37:39 And a pizza and a pie together.

00:37:40 Or a scratch.

00:37:41 So there’s a lot of ways.

00:37:44 There’s the shortest way,

00:37:46 and then you take the full spectrum of ways

00:37:48 and there’s probably an average duration

00:37:53 for a noob to make an apple pie.

00:37:56 Is the average interesting still?

00:37:58 If you measure the average length of the path

00:38:01 to assemble a thing, does that tell you something

00:38:05 about the way nature usually does it?

00:38:08 Versus something fundamental about the object,

00:38:13 which I think is what you’re aiming at

00:38:15 with the assembly index.

00:38:16 Yeah, I mean, look, we all have to quantify things.

00:38:19 The minimum path gives you the lower bounds.

00:38:21 You know you’re detecting something.

00:38:22 You know you’re inferring something.

00:38:24 The average tells you about really how the objects

00:38:26 are existing in the ecosystem or the technology.

00:38:30 And there has to be more paths explored

00:38:34 because then you can happen upon other memories

00:38:39 and then condense them down.

00:38:41 I’m not making too much sense, but if you look and say,

00:38:44 let’s just say, I mean, maybe we’re gonna get

00:38:45 to alien civilizations later, right?

00:38:47 But I would argue very strongly

00:38:49 that alien civilization A and alien civilization B,

00:38:54 they’re different assembly spaces.

00:38:55 So they’re kind of gonna be a bit messed up

00:38:57 if they happen to come one another,

00:38:59 only when they find some joint overlap in their technology,

00:39:02 because if aliens come to us and they don’t share

00:39:05 any of the causal graph we’ve showed,

00:39:06 but hopefully they share the periodic table

00:39:09 and bonds and things,

00:39:11 that we’re gonna have to really think about the language

00:39:14 to talk to us aliens by inferring,

00:39:16 by using assembly theory to infer their language,

00:39:21 their technology, and other bits and bobs.

00:39:23 And the shortest path will help you do that quickly.

00:39:26 All right, so all aliens in this causality graphs

00:39:29 have a common ancestor in the…

00:39:32 If the building blocks are the same,

00:39:33 which means they live in the same universe as us.

00:39:35 So this is the assumption.

00:39:36 It depends on how far back in time you go, though.

00:39:38 But the universe has all the same building blocks.

00:39:42 Yeah.

00:39:42 And we have to assume that.

00:39:45 So at least there’s not different classes

00:39:49 of causality graphs, right?

00:39:53 No.

00:39:53 The universe doesn’t just say like,

00:39:55 here you get the red causality graph,

00:39:58 and you get the blue one.

00:39:59 These basic ingredients,

00:40:00 and they’re geographically constrained,

00:40:02 or constrained in space or time, or something like that.

00:40:06 They’re constrained in time

00:40:07 because only by the virtue of the fact

00:40:09 that you need enough time to have passed

00:40:12 for some things to exist.

00:40:14 So the universe has to be big enough in time

00:40:16 for some things.

00:40:17 So just one point on the shortest path

00:40:18 versus the average path,

00:40:19 which I think we’ll get to this,

00:40:20 is you had a nice way of saying it’s like

00:40:22 the minimal compression is the shortest path

00:40:25 for the universe to produce that.

00:40:26 But it’s also like the first time

00:40:28 in the ordering of events

00:40:29 that you might expect to see that object.

00:40:32 But the average path tells you something

00:40:35 about the actual steps that were realized,

00:40:39 and that becomes an emergent property

00:40:40 of that object’s interaction with other objects.

00:40:43 So it’s not an intrinsic feature of that object.

00:40:45 It’s a feature of the interactions with other things.

00:40:47 And so one of the nice features of assembly

00:40:49 is you’ve basically gotten rid of,

00:40:51 you just look at the things that exist,

00:40:52 and you’ve gotten rid of the mechanisms

00:40:54 for constructing them in some sense.

00:40:56 Like the machines are not as important

00:41:00 in the current construction of the theory,

00:41:02 although I would like to bridge it

00:41:04 to some ideas about constructors.

00:41:08 But then you can only communicate with things

00:41:11 as Lee was saying,

00:41:12 if you have some overlap in the past history.

00:41:15 So if you had an alien species

00:41:16 that had absolutely no overlap,

00:41:18 then there would be no means of communication.

00:41:21 But as we progress further and further in time

00:41:25 and more things become possible

00:41:26 because the assembly spaces are larger,

00:41:29 because you can have a larger assembly space

00:41:31 in terms of index and also just the size of the space,

00:41:34 because it’s exponentially growing,

00:41:36 then more things can happen in the future.

00:41:38 And the example I like to give is actually

00:41:40 when we made first contact with gravitational waves,

00:41:43 because that’s an alien phenomenon

00:41:46 that’s been permeating our planet,

00:41:47 not alien in life phenomenon,

00:41:48 but alien like something we had never knew existed.

00:41:51 It’s been like there’s gravitational waves

00:41:55 rippling through this room right now,

00:41:57 but we had to advance to the level of Einstein

00:42:00 writing down his theory of relativity

00:42:03 and then 100 years of technological development

00:42:05 to even quote unquote see that phenomena.

00:42:08 So the, okay, to see that phenomena,

00:42:12 our causal graph have to start intersecting.

00:42:16 Yeah, we needed the idea to emerge first,

00:42:17 the abstraction, right?

00:42:18 And then we had to build the technology

00:42:20 that could actually observe features of that abstraction.

00:42:23 So the nice promising thing is over time,

00:42:26 the graph can grow so we can start overlapping eventually.

00:42:29 Yeah, so the interesting feature of that graph

00:42:31 is there was an event 1.4 billion years away

00:42:34 of a black hole merger that we detected on our detector.

00:42:38 And now suddenly we’re connected

00:42:41 through this communication channel

00:42:42 with this distant event in our universe

00:42:45 that if you think about 1.4 billion years ago,

00:42:48 what was happening on this planet

00:42:49 or even further back in time,

00:42:51 that there’s common physics underlying all of those events,

00:42:55 but even for those two events to communicate.

00:42:56 Now I understand what you were going on about

00:42:58 the other week.

00:42:59 Yeah, I’m sorry, this is a really abstract example,

00:43:01 but it’s sort of.

00:43:03 Your causal graphs are now overlapping.

00:43:05 Well, let’s just say now our causal graphs

00:43:07 are overlapping in the deep past.

00:43:10 Yeah, sorry.

00:43:10 No, I like it, so you made it.

00:43:11 I totally missed it.

00:43:12 Oh, the 1.6 billion.

00:43:13 Yeah.

00:43:14 You made a connection with it.

00:43:15 No, I do like that.

00:43:15 No, no, you can tell me what your epiphany is now.

00:43:17 That’s good.

00:43:18 Because I was.

00:43:19 And I should get the jokes before 30 seconds after, so.

00:43:22 Oh, I get it now.

00:43:25 No, it’s all right.

00:43:25 I was slow.

00:43:26 The joke from two minutes ago.

00:43:27 I’m slow on the uptake here.

00:43:28 I wasn’t able to comprehend

00:43:29 what you were talking about when saying

00:43:31 the channel communicating to the past.

00:43:32 But what you’re saying is we were able to infer

00:43:36 what happened 1.4 billion years ago.

00:43:40 We detected the gravity wave.

00:43:41 I mean, I think it’s amazing that at that time,

00:43:43 we weren’t even, we were just becoming multicellular,

00:43:46 right?

00:43:46 It’s like insane.

00:43:48 And then we progressed from multicellularity

00:43:51 through to technology and built the detector

00:43:55 and then we just extrapolate backwards.

00:43:58 So, although we didn’t do anything back to the graph

00:44:01 back in time, we understood this existence

00:44:03 then overlapped going forward.

00:44:04 And that.

00:44:05 Well, that’s because our graphs are larger.

00:44:06 Yeah, but that means that has a consequence.

00:44:09 One of the things I was trying to say is I think,

00:44:13 I don’t know, Sarah might be, she can correct me,

00:44:15 information first and I’m a object first kind of guy.

00:44:19 So, I mean, there’s things that get constructed.

00:44:21 There has to be this transition in random constructions.

00:44:25 So when the object that’s being constructed by the process

00:44:30 bakes in that memory and those memories then add on

00:44:33 and add on and add on.

00:44:35 So as it becomes more competent and life is about

00:44:39 taking those memories and compressing them,

00:44:41 increasing their autonomy.

00:44:42 And so I think that, like the cell that we have

00:44:46 in biology on earth is our way of doing that,

00:44:48 that really the maximum ability to take memories

00:44:51 and to act on the future.

00:44:53 Oh, I think that’s mathematics.

00:44:55 No, mathematics doesn’t exist.

00:44:58 No, but that’s the point.

00:44:59 The point is that abstractions do exist.

00:45:01 They’re real physical things.

00:45:03 We call them abstractions.

00:45:05 But the point about mathematics that I think is,

00:45:08 so I don’t disagree that I think you’re object first

00:45:11 and I’m information first, but I think I’m only

00:45:14 information first in the sense that I think the thing

00:45:16 that we need to explain is what abstractions are

00:45:20 and what they are as physical things

00:45:22 because of all of human history,

00:45:23 we’ve thought that there were these properties

00:45:25 that are disembodied, exist outside of the universe

00:45:29 and really they do exist in the universe

00:45:32 and we just don’t understand what their physics is.

00:45:34 So I think mathematics is a really good example.

00:45:36 We do theoretical physics with math,

00:45:39 but imagine doing physics of math

00:45:42 and then thinking about math as a physical object.

00:45:44 And math is super interesting.

00:45:46 I think this is why we think it describes reality so well

00:45:48 because it’s the most copyable kind of information.

00:45:50 It retains its properties

00:45:52 when you move it between physical media,

00:45:54 which means that it’s very deep.

00:45:56 And so it seems to describe the universe really well,

00:45:59 but it probably is because it’s information

00:46:00 that’s very deep in our past.

00:46:02 And it’s just, we invented a way of communicating it

00:46:06 very effectively between us.

00:46:08 Isn’t math more fundamental?

00:46:11 Isn’t the assembly of the graph,

00:46:13 isn’t basically, I sound completely boring.

00:46:16 It’s like math, assembly theory invented math, but it did.

00:46:20 It has to be.

00:46:20 Okay.

00:46:23 So what is math exactly?

00:46:28 It’s a nice simplification,

00:46:34 a simple description of what?

00:46:38 So we have a computer scientist,

00:46:39 a physicist, and a chemist here.

00:46:41 Walking to a bar.

00:46:42 I think the chemist is gonna define math

00:46:43 and you guys can correct me.

00:46:45 Go for it.

00:46:46 I would say.

00:46:47 Lay in on us, Lee.

00:46:48 We’re ready.

00:46:50 I think the ability to label objects

00:46:54 and place them into classes

00:46:57 and then do operations on the objects is what math is.

00:47:01 So on that point,

00:47:02 what does it mean to be object first

00:47:04 versus information first?

00:47:07 So what’s the difference between object and information

00:47:09 when you get to that low fundamental level?

00:47:11 Well, I might change my view.

00:47:14 So I’m stuff first, the stuff.

00:47:16 And then when stuff becomes objects,

00:47:18 it has to invent information.

00:47:21 And then the information acts on more stuff

00:47:23 and becomes more objects.

00:47:24 So I think there is a transition to information

00:47:27 that occurs when you go from stuff to objects.

00:47:29 You mean time though, I think.

00:47:30 Yeah.

00:47:31 Information is emergent.

00:47:33 Not emergent.

00:47:34 Information is actionable memories from the universe.

00:47:40 So when memories become actionable, that’s information.

00:47:44 But there’s always memory, but it’s not actionable.

00:47:48 Yeah.

00:47:49 And then it’s not information.

00:47:49 Great.

00:47:50 And actionable is what you can create.

00:47:52 You can use it.

00:47:53 If you can’t use it, then it’s not information.

00:47:55 If you can’t transmit it,

00:47:56 if it doesn’t have any causal consequence.

00:47:59 Falls in the forest.

00:48:00 I don’t understand.

00:48:01 Why is that not information?

00:48:03 It’s not information.

00:48:04 It’s stuff.

00:48:07 It’s stuff happening, but it’s not causal.

00:48:10 Yeah, yeah, we can.

00:48:11 This is cool.

00:48:12 But it’s happening.

00:48:13 Happening requires information.

00:48:15 No, no, no, no, no.

00:48:16 Stuff is always happening.

00:48:18 No, this is where the physicists

00:48:19 and the math petitions get themselves in a loop

00:48:21 because I think the universe, I mean,

00:48:24 I think say Max Tegmark is very playful

00:48:28 and say like the universe is just math.

00:48:30 Well, the universe is just math.

00:48:31 Then we might as well not bother having any conversation

00:48:34 because the conversation already written,

00:48:35 we just might as well go to the future and say,

00:48:36 can you just give us the conversations happened already?

00:48:38 So I think the problem is that math petitions

00:48:41 are so successful at labeling stuff

00:48:43 and so successful understanding the stuff

00:48:45 through those labels,

00:48:46 they forget that actually those labels had to emerge

00:48:49 and that information had to be built on those memories.

00:48:52 So memory in the universe, so constraints, graph,

00:48:56 when they become actionable

00:48:57 and the graph can loop back on itself

00:48:59 or interact with other graphs and they can intersect,

00:49:02 those memories become actionable

00:49:04 and therefore they’re information.

00:49:05 And I think you just changed my mind

00:49:09 on something pretty big, but I don’t have a pen.

00:49:11 So I can’t write, I’m gonna write it down later,

00:49:12 but roughly the idea is like you’ve got these two graphs

00:49:17 of objects of stuff that you have memories

00:49:20 and then when they intersect

00:49:22 and then they can act on each other,

00:49:24 that’s maybe the mechanism by which information is then,

00:49:27 so then you can then abstract.

00:49:29 So when one graph can then build another graph and say,

00:49:32 hey, you don’t have to go through the nonsense

00:49:33 we had to go through.

00:49:34 Here’s literally the way to do it.

00:49:36 Stuff always comes first,

00:49:37 but then when stuff builds the abstraction,

00:49:39 the abstractions can be then teleported

00:49:41 onto other stuff.

00:49:42 And the abstractions is the looping back power.

00:49:45 Okay.

00:49:45 Am I making, I don’t know, I got stuck.

00:49:47 Yeah, so first, God made stuff.

00:49:52 Then after that, when you start to be able

00:49:54 to form abstractions, that’s when the information.

00:49:56 God is the memory the universe can remember.

00:50:00 God is the memory the universe can remember.

00:50:02 Otherwise, there’s no, wait, did you?

00:50:03 Someone’s gonna be deciphering that statement

00:50:05 hundreds of years from now, what the hell does that mean?

00:50:07 What does the humans mean by this?

00:50:09 Look, don’t diss my one liners.

00:50:12 I took me 15 seconds to come up with that.

00:50:16 I don’t know what it means.

00:50:17 What does it mean?

00:50:19 Okay, wait, we need to, how do we get onto this?

00:50:24 We were time, causality, mathematics.

00:50:30 So what is mathematics in this picture of stuff,

00:50:34 objects, memory, and information?

00:50:43 What exactly is mathematics?

00:50:44 It’s the most efficient labeling scheme

00:50:47 that you can apply to lots of different graphs.

00:50:49 Labeling scheme doesn’t make it sound useful.

00:50:52 Can I try?

00:50:53 Yep, sure, please.

00:50:54 Have you rejected my definition of mathematics?

00:50:56 I’m shocked.

00:50:57 Yeah, no, it’s all right.

00:50:59 But it’s correct.

00:51:01 Go on, sorry.

00:51:02 Excellent.

00:51:03 No, I mean, I think we have a problem, right?

00:51:06 Cause we can’t not be us,

00:51:08 like we’re stuck in the shells we are

00:51:10 and we’re trying to observe the world.

00:51:11 And so mathematics looks like it has certain properties.

00:51:13 And I guess the thought experiment I find is useful

00:51:16 is to try to imagine if you were outside of us

00:51:19 looking at us as physical systems using mathematics,

00:51:22 what would be the specific features you associate

00:51:24 to the property of understanding mathematics

00:51:28 and being able to implement it in the universe, right?

00:51:32 And when you do that,

00:51:34 mathematics seems to have some really interesting properties

00:51:36 relative to other kinds of abstraction

00:51:39 we might talk about like language or artistic expression.

00:51:42 One of those properties is the one I mentioned already

00:51:45 that is really easy to copy between physical media.

00:51:48 So if I give you a mathematical statement,

00:51:50 you almost immediately know what I mean.

00:51:51 If I tell you the sky is blue,

00:51:53 you might say, is it gold ball blue?

00:51:54 Is it azure blue?

00:51:55 What color blue do you mean?

00:51:56 And you have a harder time visualizing what I actually mean.

00:52:00 So mathematics carries a lot of meaning with it

00:52:01 when it’s copied between physical systems.

00:52:03 It’s also the reason we use it to communicate with computers.

00:52:06 And then the second one is it retains its property

00:52:10 of actually what it can do in the universe when it’s copied.

00:52:13 So the example I like to give there

00:52:16 is think about like Newton’s law of gravitation.

00:52:19 It’s actually, it’s a compressed regularity

00:52:22 of a bunch of phenomena that we observe in the universe,

00:52:25 but then that information actually is causal in a sense

00:52:28 that it allows us to do things we wouldn’t be able to do

00:52:30 without that particular knowledge

00:52:31 and that particular abstraction.

00:52:33 And in this case, like launch satellites to space

00:52:35 or send people to Mars or whatever it is.

00:52:38 So if you look at us from the outside and you say,

00:52:41 what is it for physical systems

00:52:42 to invent a thing called mathematics

00:52:44 and then to use and then it to become a physical observable,

00:52:51 mathematics is kind of like

00:52:52 the universally copyable information

00:52:55 that allows new possibility spaces

00:52:58 to be open in the future

00:52:59 because it allows this kind of ability

00:53:01 to map one physical system to another

00:53:03 and actually understand that the general principles.

00:53:05 So is it helping the overlap of causal graphs then

00:53:10 by mapping?

00:53:11 Oh, I think that’s the explanation for what it is

00:53:14 in terms of the physical theory of assembly

00:53:16 would be some feature of the structure

00:53:18 of the assembly spaces of causal graphs

00:53:20 and their relationship to each other.

00:53:22 So for example, and I mean, this is things

00:53:25 that we’re gonna have to work out over the next few years.

00:53:27 I mean, we’re in totally uncharted conceptual territory

00:53:29 here, but as is usual, diving off the deep end,

00:53:35 but I would expect that we would be able to come up

00:53:38 with a theory of like, why is it that some physical systems

00:53:40 can communicate with each other?

00:53:43 Like language, language is basically

00:53:45 because we’re objects extended over time

00:53:47 and some of the history of that assembly space

00:53:49 actually overlaps.

00:53:51 And when we communicate,

00:53:52 it’s because we actually have shared structure

00:53:53 in our causal history.

00:53:55 Let me have another quick go at this, right?

00:53:56 So I think we all agree.

00:53:57 So I think we take mathematics for granted

00:54:01 because we’ve gone through this chain, right?

00:54:02 Of, you know, we all share a language now, okay?

00:54:06 And we can, well, we share language,

00:54:07 so we have languages that we can make interoperable.

00:54:11 And so whether you’re speaking, I don’t know,

00:54:14 all the different dialects of Chinese,

00:54:17 all the different dialects of English, French, German,

00:54:20 whatever, you can interconvert them.

00:54:22 The interesting thing about mathematics now

00:54:24 is that everybody on planet Earth, every human being

00:54:26 and computers share that common language.

00:54:29 That language was constructed by a process in time.

00:54:33 So what I’m trying to say is assembly invented math

00:54:35 is those, right from the, you know,

00:54:37 mathematics didn’t occur, it didn’t exist before life.

00:54:41 Abstraction was invented by life, right?

00:54:44 That doesn’t mean that the universe wasn’t capable

00:54:46 of mathematical things.

00:54:47 Wait a minute, can we just ask that old famous question,

00:54:50 is math invented or discovered?

00:54:52 So when you say assembly invented, or whatever.

00:54:58 It means it’s just.

00:54:58 Well, someone might just say assembly

00:54:59 is a mathematical theory, but sorry.

00:55:01 Right.

00:55:02 Are we arguing? Exactly.

00:55:03 Are we arguing now?

00:55:04 That’s what it sounds like.

00:55:05 Are we discovering mathematics?

00:55:06 No, well, yes and no.

00:55:08 I would say.

00:55:09 And you call mathematics a language

00:55:11 that we’re developing. I would say that,

00:55:13 look, I’m pretty sure that there are some very common

00:55:17 seeds of mathematics in the universe, right?

00:55:19 But actually not the mathematics that we are finding now

00:55:23 is not discovered, it’s invented.

00:55:27 And, but even though I think those two terms

00:55:29 are very triggering, and I don’t think

00:55:30 they’re necessarily useful,

00:55:31 because I think that what people do,

00:55:33 the mathematicians that say, oh, mathematics was discovered

00:55:38 because they live in a universe where there is no time

00:55:40 and it just all exists.

00:55:42 But what I’m saying is, and I think in the same way

00:55:45 you can create, let’s say I’m gonna go and create

00:55:47 and make a piece of art, did I make that piece of art

00:55:53 or did I discover it?

00:55:56 Like inventing the aeroplane.

00:55:57 Did I invent the aeroplane?

00:55:58 Let’s stick with the aeroplane.

00:55:59 The aeroplane is a good one.

00:56:00 Let’s say, did I discover the aeroplane?

00:56:03 Well, in a way, the universe discovered the aeroplane

00:56:04 because it’s just chucked a load of atoms together

00:56:06 and a load of random human beings want to do stuff

00:56:08 and then we discover the aeroplane

00:56:10 in the space of all the possibilities.

00:56:12 But here’s the thing, when the space of possibilities

00:56:14 is so vast, infinite almost, and you’re able to actualize

00:56:21 one of those in an object, then you are inventing it.

00:56:24 So in mathematics, because there are infinite number

00:56:27 of theorems, the fact you’re actually pulling,

00:56:29 there’s no difference between inventing

00:56:31 a mathematical structure and inventing the aeroplane.

00:56:34 They’re the same thing, but that doesn’t mean

00:56:36 that now the aeroplane exists in the universe,

00:56:38 it’s something weird about the universe.

00:56:40 So I think that the more, this is the thing

00:56:43 that you probably, the more memory required

00:56:47 for the object, the more invented it is.

00:56:50 So when a mathematical theorem needs more bytes

00:56:54 to store it, the more invented it is,

00:56:56 and the less bytes, the more discovered it is.

00:57:00 But everything then is invented.

00:57:02 It’s just more or less invented.

00:57:03 Absolutely.

00:57:04 Okay.

00:57:05 The universe has to generate everything as it goes.

00:57:07 Yeah.

00:57:09 And it wasn’t there in the beginning.

00:57:11 And the way we’re thinking it,

00:57:12 when you’re thinking about the difference

00:57:14 between invented and discovered,

00:57:15 is because we’re throwing away all the memory.

00:57:17 Yeah.

00:57:18 So if you start to think in terms of causality and time,

00:57:22 then those things become the same.

00:57:23 Everything is invented.

00:57:25 And the idea is to make everything intrinsic

00:57:27 to the universe.

00:57:28 So I think one of the features of assembly theory

00:57:30 is we don’t wanna have external observers.

00:57:32 There’s been this long tradition in physics

00:57:33 of trying to describe the universe from the outside

00:57:35 and not the inside.

00:57:37 And the universe has to generate everything itself

00:57:40 if you do it from the inside.

00:57:41 Assembly theory describes how the universe builds itself.

00:57:46 Did it take you 15 seconds to say that?

00:57:48 Yeah.

00:57:49 And to come up with that also?

00:57:50 No, I’ve thought of that before.

00:57:51 Okay.

00:57:52 That’s a good line.

00:57:52 It’s a, it’s like.

00:57:53 Oh, you’re making fun of me.

00:57:54 No, I’m not making fun, I’m having fun.

00:57:56 There’s a difference.

00:57:57 Oh, that’s good.

00:57:58 All right.

00:57:59 She’s inventing fun.

00:57:59 I’m not all intimidated.

00:58:01 Yes.

00:58:01 And there’s a causal history to that fun.

00:58:06 You mentioned that there’s no way to communicate

00:58:08 with aliens until there’s overlap in the causal graph.

00:58:14 Communication includes being able to see them.

00:58:18 And like, what are we, this is the question is,

00:58:23 is communication any kind of detection?

00:58:26 And if so, what do aliens look like

00:58:29 as you get more and more overlap on the causal graph?

00:58:33 You’re assuming, let’s assume that aliens,

00:58:35 so when you see them and they see you,

00:58:39 you’re assuming they have vision,

00:58:40 they have the ability to construct in 3D and in time.

00:58:43 That’s a lot of assumptions they’re making.

00:58:45 What detection?

00:58:46 All right, let’s step back.

00:58:47 So yes, okay, you’re right.

00:58:49 So when, in the English language,

00:58:50 when we say the word see, we mean visually,

00:58:53 they show up to a party and it’s like,

00:58:54 oh, wow, that’s an alien.

00:58:56 That’s visual, that’s 3D, that’s, okay.

00:58:59 And that’s also assuming scale,

00:59:03 spatial scale of something that’s visible to you.

00:59:05 So it can’t be microscopic or it can’t be so big

00:59:08 that you don’t even realize that’s an entity.

00:59:10 Okay, but other kinds of detection too.

00:59:14 I would make it more abstract and go down.

00:59:15 I was thinking this morning about how to rewrite

00:59:18 the Arecibo message in assembly theory

00:59:21 and also to abandon binary.

00:59:23 Because I don’t think aliens necessarily,

00:59:25 why should they have binary?

00:59:27 Well, they have some basic elements

00:59:29 with which to do information exchange.

00:59:32 Let’s make it more fundamental, more universal.

00:59:37 So we need to think about what is the universal way

00:59:38 of making a memory and then we should

00:59:40 reencode Arecibo in that way.

00:59:43 What’s more basic than zeros and ones?

00:59:45 Well, it’s really difficult to get out

00:59:47 that causal chain because we’re so,

00:59:48 so let’s erase the idea of zero for a moment.

00:59:51 It took human beings a long time

00:59:53 to come up with the idea of zero.

00:59:54 Now you got the idea of zero, you can’t throw it away.

00:59:56 It’s so useful.

00:59:57 To discover the idea of zero.

00:59:59 To discover or invent.

01:00:00 I don’t know, but it took a long time

01:00:02 so it was invented, that’s right.

01:00:04 Yeah, I think zero was invented, exactly.

01:00:07 So it’s not a given that aliens know what zero is.

01:00:10 That’s a massive assumption.

01:00:13 It’s a useful discovery.

01:00:16 You’re saying if you break the causal chain

01:00:18 there might be some other more efficient way of representing.

01:00:20 That’s why I wanna meet him and ask him.

01:00:23 For a shortcut.

01:00:24 But you won’t be able to ask him until.

01:00:27 So I interrupted you and I think you’re making good point.

01:00:29 I was just gonna say, well look.

01:00:31 Thank you.

01:00:32 Sorry.

01:00:34 Rather than saying.

01:00:35 Please internet, tweet at him for the rude interruptions.

01:00:37 Oh, go ahead, I’m sorry.

01:00:38 No, it’s okay.

01:00:40 Maybe it’s change.

01:00:41 How do we, so, oh, I don’t know what it’s like

01:00:44 to be an alien.

01:00:46 I would like to know.

01:00:47 What is the full spectrum of what aliens

01:00:50 might look like to us?

01:00:53 Now that we’ve laid this all on the table of like,

01:00:57 all right, so there has to be some overlap

01:00:59 in this causal chain that led to them.

01:01:03 What are we looking for?

01:01:05 What do you think we should be looking for?

01:01:06 So you mentioned mass spec.

01:01:09 Measuring certain objects that aliens could create

01:01:11 or are aliens themselves.

01:01:15 We show up to a planet or maybe not a planet.

01:01:17 Or maybe, what the hell is the basic object

01:01:20 we’re trying to measure the assembly index of?

01:01:23 Let’s cut ourselves a break.

01:01:23 Let’s assume that they are metabolized.

01:01:27 They’ve got an energy source.

01:01:29 And they’re a size that we can recognize.

01:01:32 Let’s give ourselves a break.

01:01:34 Because there could be aliens that are so big

01:01:36 we won’t recognize we’re seeing them.

01:01:38 There might be aliens that are so small

01:01:39 we don’t yet have the ability to,

01:01:40 we don’t have microscopes that can see far enough away

01:01:44 that just wouldn’t be able to see them.

01:01:45 So what’s a good range?

01:01:46 So let’s just make a range.

01:01:47 Let’s just be very anthropocentric and say,

01:01:50 we’re gonna look for aliens roughly our size

01:01:52 and technology our size.

01:01:53 Because we know it’s possible on Earth, right?

01:01:56 I mean, a reasonable thing to do would be

01:01:57 to find exoplanets that are in the same zone as Earth

01:02:01 in terms of heat and stuff.

01:02:03 And then say, hey, if there’s that same kind of gravity,

01:02:06 same kind of stuff, we could reasonably assume

01:02:09 that alien life there might use a similar kind

01:02:13 of physical infrastructure.

01:02:15 And then we’re good.

01:02:16 So then your question becomes really relevant.

01:02:19 Say, right, let’s use vision, sound, touch.

01:02:23 And then.

01:02:24 So okay, that’s really nice.

01:02:24 So that if there’s a lot of aliens out there,

01:02:27 there’s a good likelihood if you match to the planet

01:02:32 that they’re going to be in the same spatial

01:02:35 and temporal, operating in the same spatial

01:02:38 and temporal domain as humans.

01:02:40 Okay, within that, what do they look like visually?

01:02:47 What do they sound like?

01:02:49 What do they, oh god, this sounds creepy.

01:02:52 Tastes like, what do they, oh, smell like, smell like.

01:02:56 That sounds like our clubhouse.

01:02:57 We was like, can we have sex with aliens?

01:02:59 Which was basically me saying.

01:03:01 Passionate, passionate love.

01:03:01 But it wasn’t actually about sex.

01:03:02 It was about, is our chemistry compatible, right?

01:03:04 Is there some?

01:03:05 Yeah.

01:03:06 Yeah, can we, yeah.

01:03:09 Are they edible too?

01:03:10 They could be very edible.

01:03:12 They could be delicious.

01:03:13 That’s why I want to see some aliens, right?

01:03:14 Because I think, are there, I think evolution,

01:03:19 I mean, evolution exploits symmetry, right?

01:03:21 Because why generate memory?

01:03:23 Why generate storage, the need for storage space

01:03:26 when you can use symmetry?

01:03:27 So, and symmetry is quite, may be quite effective

01:03:30 in allowing you to mechanically design stuff, right?

01:03:33 So maybe alien, you could be reasonable to assume

01:03:36 that aliens could have, they could be bipedal.

01:03:39 They could be symmetric in the same way.

01:03:42 Might have a couple of eyes or a couple of senses.

01:03:44 We can make them, perhaps there’s this whole zoo

01:03:47 of different aliens out there.

01:03:49 And we’ll never get to be able to classify

01:03:50 some of the weird aliens we can’t interact with

01:03:52 because they have made such weird stuff.

01:03:55 But we are just going to look at,

01:03:57 we’re going to find aliens that look most like us.

01:03:59 Why not?

01:04:00 Because those are the first ones we’re likely to see.

01:04:02 Yeah.

01:04:03 Yeah.

01:04:03 But I think it’s really hard to imagine

01:04:06 what the space of aliens is because the space is huge.

01:04:09 Because, you know, like one of the arguments

01:04:11 that you can make about why life emerges in chemistry

01:04:13 is because chemistry is the first scale

01:04:16 in terms of like, you know,

01:04:17 building up objects from elementary objects.

01:04:21 That the number of possible things that could exist

01:04:23 is larger than the universe can possibly make all at once.

01:04:25 Right?

01:04:26 So imagine you have two planets

01:04:29 and they’re cooking some geochemistry.

01:04:31 You know, our planet invented one kind of biochemistry.

01:04:34 And presumably as you start building up

01:04:37 the complexity of the molecules,

01:04:38 the chances of the overlap in those trajectories,

01:04:41 those causal chains being built up is probably very low.

01:04:44 And it gets lower and lower as it gets further advanced

01:04:47 along its evolutionary path.

01:04:49 So I think it’s very difficult to imagine

01:04:51 predicting the technologies that aliens are gonna have.

01:04:54 I mean, it’s so, you’re looking at basically

01:04:57 planets have kind of convergent chemistry,

01:04:59 but there’s some variability.

01:05:00 And then you’re looking basically at the outgrowth

01:05:02 into the possibility space for chemistry.

01:05:04 So do you think we would detect the technology,

01:05:07 the objects created by aliens before we detect the aliens?

01:05:11 Possibly.

01:05:12 So when you’re talking about measuring assembly index,

01:05:15 don’t you think we would detect the garbage first?

01:05:19 Like at the outskirts of alien civilizations,

01:05:21 isn’t this just gonna be trash?

01:05:26 I think I would come back to Arecibo.

01:05:28 The Arecibo message sent from the Arecibo telescope

01:05:30 built by Drake, I think, and Sagan.

01:05:33 How’s Arecibo spelled?

01:05:35 A R E C I B O.

01:05:37 Yes, thank you.

01:05:37 And there we go, they’ve got that up there.

01:05:40 That’s the telescope that sent the message

01:05:41 that you’re talking about.

01:05:42 So that message was sent where?

01:05:45 It was beamed at a star, a specific star,

01:05:48 and it was sent out many years ago.

01:05:52 And what they did, so this is why I was pushing on binary,

01:05:55 it’s a binary message.

01:05:56 I think it’s a semi prime length number of characters.

01:05:59 So I think 73 by 23, I think.

01:06:03 And it basically represents human bit proton,

01:06:07 binary, human beings, DNA, male and female.

01:06:10 And it’s really cool.

01:06:12 But I’m just wondering if it could be done

01:06:18 not making any,

01:06:19 cause it made assumptions that aliens speak binary.

01:06:22 Why make that assumption?

01:06:24 Why not just assume that if the difference between physics,

01:06:27 chemistry and biology is the amount of memory

01:06:29 that’s instead that’s recordable by the substrates,

01:06:33 then surely the universal thing,

01:06:36 I’m gonna make some sacrilegious statement,

01:06:38 which I think is pretty awesome for people to argue with.

01:06:42 So this is, we’re looking at an image

01:06:44 where it’s the entirety of the message encoded in binary.

01:06:49 And then there’s a probably interpretation

01:06:52 of different parts of that image.

01:06:53 There’s a person, there’s green parts.

01:06:57 It looks like for people just listening,

01:06:59 like a game of Tetris.

01:07:01 So it’s encoding in minimal ways,

01:07:03 a bunch of cool information, probably.

01:07:05 Representing all of us.

01:07:06 So the topic’s kind of teaching us how to count

01:07:08 and then it all goes all the way down

01:07:10 teaching you chemistry and then just says,

01:07:12 but it makes so many assumptions.

01:07:14 And I think if we can actually,

01:07:16 so look, I think, I mean, Sarah’s much more eloquent

01:07:19 in expressing this, but I’ll have a go

01:07:20 and you can correct it if you want,

01:07:21 which is like one of the things that Sarah has had

01:07:25 a profound effect on the way I look at the origin of life.

01:07:29 And this is one of the reasons why we’re working together

01:07:31 because we don’t really care about the origin of life.

01:07:34 We wanna make life, make aliens and find aliens.

01:07:37 Make aliens, find aliens.

01:07:38 I think we might have to make aliens in the lab

01:07:40 before we find aliens in the universe, right?

01:07:42 I think that would be a cool way to do it.

01:07:45 So what is it about the universe that creates aliens?

01:07:48 Well, it’s selection through assembly theory,

01:07:51 creating memories, because when you create memories,

01:07:53 you can then command your domain.

01:07:57 You can basically do stuff.

01:07:58 You can command matter.

01:08:00 So we need to find a way by understanding what life is

01:08:03 of how the minimal way to command matter,

01:08:05 how that would emerge in the universe.

01:08:08 And if we want to communicate,

01:08:10 I mean, maybe we don’t want

01:08:11 to necessarily uniformly communicate.

01:08:13 What I would do perhaps if I had,

01:08:15 is I would send out lots of probes away from Earth

01:08:18 that have this magic way of communicating with aliens,

01:08:20 get them quite a far away from Earth, plausibly deniable,

01:08:23 and then send out the message

01:08:25 that would then attract all the aliens,

01:08:27 and then basically work out if they were friend or foe

01:08:29 and how they wanna hang out.

01:08:30 The messages being something has to do with the memories.

01:08:33 Yes, like the assembly version of Arecibo,

01:08:36 so that everyone in the universe

01:08:38 that has been understands what life is.

01:08:40 So aliens need to work out what they are.

01:08:43 Once they’ve worked out what they are,

01:08:44 they then can work out how to encode what they are,

01:08:47 and then they can go out and send messages.

01:08:48 It’s like the universal, the Rosetta Stone

01:08:52 for life in the universe is working out

01:08:55 how the memories are built.

01:08:56 I don’t know, Sarah, you have any, well,

01:09:01 whether you would agree with that.

01:09:02 No, I wanted to raise a different point,

01:09:06 which is about the fact that we can’t see the aliens yet

01:09:09 because we haven’t gotten the technology.

01:09:11 And presumably we think assembly theory

01:09:14 is the right way of doing it,

01:09:15 but I don’t think that we know how to go

01:09:16 from the kind of data you’re describing, Lex,

01:09:18 like visual data or smell

01:09:21 to construct the assembly spaces yet.

01:09:23 And in some ways, I think that the problem

01:09:25 of life detection really is the same problem

01:09:28 at the foundations of AI that we don’t understand

01:09:31 how to get machines to see causal graphs,

01:09:35 to see reality in terms of causation.

01:09:37 And so I think assembly and AI

01:09:40 are gonna intersect in interesting ways, hopefully,

01:09:43 but the sort of key point,

01:09:46 and I’ve been trying to make this argument more recently,

01:09:49 I might write an essay on it,

01:09:50 is people talk about the great filter, right?

01:09:53 And which is, again, this doomsday thing

01:09:55 that people wanna say there’s no aliens out there

01:09:57 because something terrible happened to them.

01:10:00 And it matters whether that’s in our past or our future

01:10:03 as to the longevity of our species, presumably,

01:10:06 which is why people find it interesting.

01:10:07 But I think it’s not a physical filter.

01:10:10 It’s not like things go extinct.

01:10:11 I think it’s literally,

01:10:12 we don’t have the technology to see them.

01:10:15 And you could see that with microscopes.

01:10:17 I mean, we didn’t know there were microbes on this table

01:10:18 or tables for thousands of years or telescopes.

01:10:21 Like there’s so much of the universe we can’t see.

01:10:23 And then basically what we have done as a species

01:10:25 is outsource our physical perceptions to technology,

01:10:29 building microscopes based on our eyes,

01:10:31 and building seismometers based on our sense of feelings,

01:10:34 like feel earthquakes and things.

01:10:36 And AI is basically we’re trying to outsource

01:10:38 what’s actually happening in our thinking apparatus

01:10:40 into machines now and to technological devices.

01:10:43 And maybe that’s the key technology

01:10:45 that’s gonna allow us to see things like us

01:10:47 and see the universe in a totally different way.

01:10:48 But you kind of mentioned the great filter.

01:10:50 Do you think there’s a way through technology

01:10:52 to stop being able to see stuff?

01:10:54 So can you take a step backwards?

01:10:55 I think so, yeah.

01:10:56 Did you imply that with the great, so like?

01:10:58 Well, no, I mean, I think there’s a great perceptual filter

01:11:01 in the sense that a example of life evolving on a planet

01:11:06 over billions of years has to acquire

01:11:08 a certain amount of knowledge and technology

01:11:11 to actually recognize the phenomena that it is.

01:11:14 Well, that’s the sense I have is,

01:11:17 I mean, you talk with physicists, engineers in general,

01:11:20 there’s this kind of idea that we have

01:11:23 most of the tools already to hear the signal.

01:11:27 But to me, it feels like we don’t have any of the tools

01:11:31 to see the signal.

01:11:32 No, we don’t know what we’re doing, yeah, I agree.

01:11:33 That’s the biggest, like to hear.

01:11:35 We don’t have the tools to really hear, to see.

01:11:37 Yeah.

01:11:39 Aliens are everywhere.

01:11:40 We just don’t have the, yeah, well, oh, that’s.

01:11:44 I mean, I got this in part, actually,

01:11:46 because you were like, you know,

01:11:47 last time I was here, he was like, look at the carpet.

01:11:49 You know, if you had an alien detector

01:11:51 where the carpet be aliens.

01:11:52 I mean, I think we really don’t.

01:11:54 I think it would be.

01:11:56 But the aliens would nevertheless have a high assembly index

01:11:58 or produce things of high assembly index.

01:12:01 And those things of high assembly index,

01:12:04 you have to have a detector that can recognize

01:12:07 high assembly index in all its forms.

01:12:09 Yeah. Yes.

01:12:10 That’s it, that’s it.

01:12:10 Take data, construct assembly space.

01:12:13 Yeah. Those patterns, basically.

01:12:15 So one way to think about high assembly index

01:12:17 is interesting patterns of basic ingredients.

01:12:21 I can give you an example,

01:12:23 because I mean, in molecules,

01:12:24 we’ve been talking about in objects,

01:12:25 but we’re also trying to do it in spatial trajectories.

01:12:29 Like, imagine you’re just,

01:12:30 like, I always get bothered by the fact that, like,

01:12:33 when you look at birds flocking,

01:12:35 you can describe that with like a simple Boyd’s model,

01:12:37 or like, you know, people use spin glass

01:12:38 to describe animal behavior.

01:12:40 And those are like really simple physics models.

01:12:42 Yet you’re looking at a system that you know has agency

01:12:46 and there’s intelligence in those birds.

01:12:48 And basically, like, you can’t help but think

01:12:52 there must be some statistical signatures

01:12:53 of the fact that they’re,

01:12:55 that’s a group of agents versus, you know, like,

01:12:58 I don’t know, you know, the physics example,

01:13:00 maybe like, I don’t know, Brownian motion or something.

01:13:03 And so what we’re trying to do

01:13:04 is actually apply assembly to trajectory data

01:13:06 to try to say there’s a minimal amount of causal history

01:13:09 to build up certain trajectories for observed agents

01:13:12 that’s like an agency detector for behavior.

01:13:15 Do you think it’s possible to do some like Boyd’s

01:13:17 or those kinds of things, like artificial,

01:13:21 like cellular automata, play with those ideas

01:13:25 with assembly, with assembly theory?

01:13:28 Have you found any useful, really simple mathematical,

01:13:34 like, simulation tools that allow you

01:13:36 to play with these concepts?

01:13:38 So like one, of course, you’re doing mass spec

01:13:40 in this physical space with chemistry,

01:13:44 but it just seems, well, I mean,

01:13:45 computer science person, maybe,

01:13:47 it seems easier to just.

01:13:48 I agree with you.

01:13:49 It seems even sexier in terms of tweeting visual information

01:13:53 on Twitter or Instagram, more importantly,

01:13:57 to play like, here’s an organism of a low assembly index

01:14:01 and here’s an organism of a high assembly index

01:14:03 and let’s watch them create more and more memories

01:14:07 and more and more complex objects.

01:14:09 And so like, in mathematics,

01:14:11 you get to observe what that looks like

01:14:12 to build up an intuition what assembly index is like.

01:14:15 We are building a toolkit right now.

01:14:17 So I think it’s a really good idea,

01:14:19 but what we’ve got to do is I’m kind of still obsessed

01:14:22 with the infrastructure required.

01:14:24 And one of the reasons why I was pushing on information

01:14:27 and mathematics when human beings,

01:14:29 when human beings, we take a lot of the infrastructure

01:14:31 for granted.

01:14:33 And I think we have to strip that back a bit

01:14:35 for going forward, but you’re absolutely right.

01:14:36 I would agree that I think the fact that we exist

01:14:40 in the universe, this is like,

01:14:42 I can see that lots of people would disagree

01:14:44 with the statement, but I don’t think Sarah will,

01:14:47 but I don’t know.

01:14:48 The fact that objects exist,

01:14:50 I don’t think anyone on earth will disagree

01:14:53 that objects can exist elsewhere, right?

01:14:55 But they will disagree that life can exist elsewhere.

01:14:58 But what perhaps I’m trying to say is that

01:15:01 the acquisition, the universe’s ability to acquire memory

01:15:06 is the very first step for building life.

01:15:10 And that must be, that’s so easy to happen.

01:15:14 So therefore alien life is everywhere

01:15:16 because all alien life is,

01:15:18 is those memories being compressed and minimalized

01:15:22 and the alien equivalent of the cell working.

01:15:24 So I think that we will build new technologies

01:15:27 to find aliens, but we need to understand what we are first

01:15:32 and how we go from physics to chemistry to biology.

01:15:36 The most interesting thing,

01:15:38 as you’re saying to these two organisms,

01:15:40 different assemblies, there’s one you get into biology.

01:15:43 Biology gets more and more weird,

01:15:45 more and more contingent.

01:15:46 Physics is, chemistry is less weird

01:15:48 cause the rules of chemistry are smaller

01:15:50 than the rules of biology.

01:15:51 And then going away to physics where you have a very

01:15:56 nicely tangible number of ways of arranging things.

01:16:00 And I think assembly theory just helps you appreciate that.

01:16:03 And so once we get there,

01:16:04 my dream is that we are just gonna be able to suddenly,

01:16:08 I mean, I’m maybe just being really arrogant here.

01:16:10 I don’t mean to be arrogant.

01:16:11 It’s just, I’ve got this hammer called assembly

01:16:14 and everything’s a nail.

01:16:15 But I think that once we crack it,

01:16:17 we’ll be able to use assembly theory plus telescopes

01:16:20 to find aliens.

01:16:22 Do you have, Sarah, do you have disagreements with Lee

01:16:25 on the number of aliens that are out there?

01:16:28 I do actually, yeah, well.

01:16:30 And what they look like.

01:16:31 So any of the things we’ve been talking about,

01:16:33 is there nuanced, it’s always nice to discover wisdom

01:16:40 through nuanced disagreement.

01:16:43 Yeah, I don’t wholly disagree, but I think,

01:16:47 but I do think I disagree.

01:16:48 It’s kind of, there’s nuance there.

01:16:51 But Lee made it.

01:16:52 You can disagree.

01:16:53 No, it’s fine.

01:16:54 It is nuanced, right?

01:16:55 So you made the point earlier that you think,

01:16:59 once we discover what life is,

01:17:02 we’ll see alien life everywhere.

01:17:04 And I think I agree on some levels in the sense

01:17:06 that I think the physics that governs us is universal.

01:17:09 But I don’t know how far I would go to say,

01:17:11 to say that we’re a likely phenomenon

01:17:12 because we don’t understand all of the features

01:17:15 of the transition at the origin of life,

01:17:17 which we would just say in assembly,

01:17:19 as you go from the no memory physics

01:17:22 to there’s like a critical transition

01:17:26 around the assembly index

01:17:27 where assembliness starts to increase.

01:17:28 And that’s what we call the evolution of the biosphere

01:17:30 and complexification of the biosphere.

01:17:32 So there’s a principle of increasing assembliness

01:17:34 where that goes back to what I was saying

01:17:35 at the very beginning about the physics of the possible,

01:17:38 that the universe basically gets in this mode

01:17:40 of trying to make as much possibilities as possible.

01:17:44 Now, how often that transition happens

01:17:48 that you get the kind of cascading effect

01:17:50 that we get in our biosphere, I think we don’t know.

01:17:53 If we did, we would know the likelihood of life

01:17:54 in the universe.

01:17:55 And a lot of people wanna say life is common,

01:17:57 but I don’t think that we can say that yet

01:17:58 till we have the empirical data,

01:17:59 which I think you would agree with.

01:18:01 But then there’s this other kind of thought experiment I have

01:18:04 which I don’t like, but I did have it,

01:18:07 which is if life emerges on one planet

01:18:11 and you get this real high density of things

01:18:13 that can exist on that planet,

01:18:14 is it sort of dominating the density of creation

01:18:17 that the universe can actually generate?

01:18:19 So like if you’re thinking about counting entropy, right?

01:18:21 Like the universe has a certain amount of stuff in it.

01:18:23 And then assembly is kind of like an entropic principle.

01:18:27 It’s not entropy.

01:18:29 But the idea is that now transformations among stuff

01:18:33 or the actual physical histories of things

01:18:36 now become things that you have to count

01:18:37 as far as saying that these things exist

01:18:40 and we’re increasing the number of things that exist.

01:18:42 And if you think about that cosmologically,

01:18:45 maybe Earth is sucking up all the life potential

01:18:47 of the whole universe, I don’t know.

01:18:49 But I haven’t.

01:18:50 How’s that, can you expand that a little bit?

01:18:51 Why can any one geographical region

01:18:54 suck up the creative capacity of the universe?

01:18:57 Just like, I know it’s a ridiculous thought.

01:19:00 I don’t actually agree with it,

01:19:01 but it was just a thought experiment.

01:19:02 I love that you can have thoughts

01:19:04 that you don’t like and don’t agree with,

01:19:07 but you have to think through them anyway.

01:19:09 The human mind is fascinating.

01:19:11 Yeah, I think these sort of counterfactual

01:19:15 thought experiments are really good

01:19:16 when you’re trying to build new theories

01:19:17 because you have to think through all the consequences.

01:19:19 And there are people that want to try to account for,

01:19:23 say, the degrees of freedom on our planet

01:19:25 in cosmological inventories of talking about

01:19:28 the entropy of the universe.

01:19:29 And when we’re thinking about cosmological

01:19:32 arrow of time and things like that.

01:19:33 Now, I think those are pretty superficial proposals

01:19:35 as they stand now, but assembly would give you

01:19:37 a way of counting it.

01:19:38 And then the question is if there’s a certain

01:19:40 maximal capacity of the universe’s speed

01:19:43 of generating stuff, which Lee always has this argument

01:19:45 that assembly is about time.

01:19:47 The universe is generating more states.

01:19:49 Really what it’s generating is more assembly possibilities.

01:19:52 And then dark energy might be one manifestation of that,

01:19:56 that the universe is accelerating its expansion

01:19:58 because that makes more physical space.

01:20:00 And what’s happening on our planet is it’s accelerating

01:20:02 in the expansion of possible things that exist.

01:20:05 And maybe the universe just has a maximal rate

01:20:07 of what it can do to generate things.

01:20:09 And then if there is a maximal rate,

01:20:11 maybe only a certain number of planets

01:20:12 can actually do that.

01:20:13 Or there’s a trade off about the pace of growth

01:20:16 on certain planets versus others.

01:20:18 I have a million questions there,

01:20:19 but do you have thoughts on?

01:20:20 Just a quick, yeah, I’ll just say something very quick.

01:20:22 It’s a thought experiment.

01:20:23 No, it’s good, I think I get it.

01:20:24 I think I get it.

01:20:25 So what I want to say is when I mean aliens are everywhere,

01:20:29 I mean memories are the prerequisite

01:20:34 for aliens via selection

01:20:36 and then concentration of selection

01:20:39 when selection becomes autonomous.

01:20:40 So what I would love to do is to build,

01:20:42 say a magical telescope that was a memory,

01:20:45 a magical one, or a real one,

01:20:48 that would be a memory detector to see selection.

01:20:51 So you could get to exoplanets and say that exoplanet

01:20:54 looks like there’s lots of selection going on there.

01:20:56 Maybe there’s evolution and maybe there’s going to be life.

01:20:58 So what I’m trying to say is narrow down

01:21:00 the regions of space where you say

01:21:01 there’s definitely evidence of memory as high assembly there

01:21:05 or not high assembly, because that would be life,

01:21:07 but where it’s capable of happening.

01:21:11 And then that would also help us frame the search for aliens.

01:21:14 I don’t know how likely it is to make the transition

01:21:16 to cells and all the other things.

01:21:18 I think you’re right.

01:21:20 But I think that we just need to get more data.

01:21:23 Well, I didn’t like the thought experiment

01:21:24 because I don’t like the idea

01:21:25 that if the universe has a maximal limit

01:21:27 on the amount it can generate per unit time

01:21:29 that our existence is actually precluding the existence

01:21:31 of other things.

01:21:32 Well, I’ll just say one thing.

01:21:33 But I think that’s probably true anyway

01:21:34 because of the resource limitations.

01:21:35 So I don’t like your thought experiment

01:21:37 because I think it’s wrong.

01:21:39 Well, no, no, I do like the thought experiment.

01:21:41 So what you’re trying to say is like,

01:21:42 there is a chain of events that goes back

01:21:44 that’s manifestly culminated with life on Earth.

01:21:48 And you’re not saying that life isn’t possible elsewhere.

01:21:50 You say that there has been these number of things,

01:21:52 contingent things that have happened

01:21:53 that have allowed life to merge here.

01:21:56 That doesn’t mean that life can’t emerge elsewhere,

01:21:58 but you’re saying that the intersection of events

01:22:00 may be concentrated here, right?

01:22:04 And I think there’s…

01:22:05 Not exactly.

01:22:06 It’s more like if you look at,

01:22:10 say the causal graphs are fundamental,

01:22:12 maybe space is an emergent property,

01:22:14 which is consistent with some proposals on quantum gravity,

01:22:16 but also how we talk about things in assembly theory.

01:22:18 Then the universe is causal graphs generating

01:22:22 more structure in causal graphs, right?

01:22:24 So this is how the universe is unfolding.

01:22:26 And maybe there’s a cap on the rate of generation.

01:22:30 Like there’s only so much stuff

01:22:31 that gets made per update of the universe.

01:22:34 And then if there’s a lot of stuff being made

01:22:36 in a particular region that happens

01:22:38 to look the same locally, spatially,

01:22:40 that’s an after effect of the fact

01:22:43 that the whole causal graph is updating.

01:22:45 Like it’s…

01:22:47 Yeah, I don’t know that.

01:22:49 I think that that doesn’t work.

01:22:51 I don’t think it works either,

01:22:51 but I don’t have a good argument in my mind about.

01:22:53 But I do like the idea of the capacity that universe,

01:22:55 cause you’ve got the number of states.

01:22:57 Yeah, we can come back to it.

01:22:59 Let me ask real quick.

01:23:00 Like why does different like local pockets

01:23:04 of the universe start remembering stuff?

01:23:07 How does memory emerge exactly?

01:23:10 So at the origin of the universe, it was very forgetful.

01:23:16 That’s when the physicists were happiest.

01:23:17 It was low memory objects,

01:23:20 which is like ultra low memory objects,

01:23:23 which is what the definition of stuff.

01:23:26 Okay, so how does memory emerge?

01:23:29 How does the temporal stickiness of objects emerge?

01:23:38 I’m gonna take a very chemocentric point of view

01:23:42 because I can’t imagine any other way of doing it.

01:23:44 You could think of other ways maybe.

01:23:47 But I would say heterogeneity in matter

01:23:52 is where the memory…

01:23:53 So you must have enough different ways

01:23:55 of rearranging matter for there to be a memory.

01:23:58 So what that means,

01:23:59 if you’ve got particles colliding in a box,

01:24:01 let’s just take some elements in a box.

01:24:06 Those elements can combine in a combinatorial set of ways.

01:24:10 So there’s a combinatorial explosion

01:24:11 of the number of molecules or minerals or solid objects,

01:24:15 bonds being made.

01:24:17 Because there’s such a large number,

01:24:19 the population of different objects that are possible,

01:24:22 this goes back to assembly theory

01:24:23 where assembly theory, there’s four types of universes.

01:24:27 So you’ve got basically, and this is what one was up earlier

01:24:30 where one universe where you’ve just got

01:24:33 everything is possible.

01:24:34 So you can take all the atoms

01:24:35 and combine them and make everything.

01:24:37 Then you’ve got basically what is the assembly combinatorial

01:24:42 where you basically have to accrue information in steps.

01:24:46 Then you’ve got assembly observed,

01:24:49 and then you’ve got the object assembly going back.

01:24:51 So what I’m trying to say is like,

01:24:53 if you can take atoms and make bonds,

01:24:55 let’s say you take a nitrogen atom and add it

01:24:57 to a carbon atom, you find an amino acid,

01:24:59 then you add another carbon atom on

01:25:01 in a particular configuration,

01:25:02 then another one, all different molecules.

01:25:04 They all represent different histories.

01:25:07 So I would say for me right now,

01:25:10 the most simple route into life seems to be

01:25:13 through recording memories and chemistry.

01:25:16 But that doesn’t mean there can’t be other ways

01:25:18 and can’t be other emergent effects.

01:25:20 But I think if you can make bonds

01:25:22 and lots of different bonds,

01:25:24 and those molecules can have a causal effect on the future.

01:25:29 So imagine a box of atoms,

01:25:32 and then you combine those atoms in some way.

01:25:35 So you make molecule A from load of atoms,

01:25:40 and then molecule A can go back to the box

01:25:43 and influence the box.

01:25:45 Then you make A prime or AB or ABC.

01:25:49 And that process keeps going,

01:25:51 and that’s where the memories come from,

01:25:53 is that heterogeneity in the universe from bonding.

01:25:57 I don’t know if that makes any sense.

01:25:58 And it’s beginning to flourish at the chemistry level.

01:26:04 Yeah.

01:26:05 So the physicists have no, like not enough.

01:26:09 Yeah, I mean.

01:26:11 They’re like desperately begging

01:26:13 for more freedom and heterogeneous components to play with.

01:26:20 Yeah, that’s exactly it.

01:26:23 What do you think about that, Sarah?

01:26:24 I mentioned already, I think it’s significant

01:26:26 that whatever physics governs life

01:26:28 emerges actually in chemistry.

01:26:30 It’s not relevant at the subatomic scale

01:26:32 or even at the atomic scale.

01:26:34 It’s in, well, atomic scale because chemistry.

01:26:37 But like when you get into this combinatorial diversity

01:26:40 that you get from combining things on the periodic table,

01:26:44 that’s when selection actually matters

01:26:47 or the fact that some things can exist

01:26:49 and others can’t exist actually starts to matter.

01:26:52 So I think of it like you don’t study gravity

01:26:55 inside the atomic nucleus.

01:26:57 You study it in terms of large scale structure

01:26:59 of the universe or black holes or things like that.

01:27:01 And whatever we’re talking about as physics of information

01:27:04 or physics of assembly becomes relevant

01:27:06 at a certain scale of reality.

01:27:08 And the transition that you’re talking about,

01:27:11 I would think of as just when you get a sufficient density

01:27:15 in terms of the assembly space

01:27:16 of like the relationship of the overlap

01:27:18 and the assembly space,

01:27:20 which is like a feature of common memory,

01:27:23 there is this transition to assembly dominated physics,

01:27:27 whatever that is.

01:27:29 Oh, like when we’re talking about,

01:27:30 and we’re trying to map out exactly

01:27:31 what that transition looks like.

01:27:33 We’re pretty sure of some of its features,

01:27:36 but we haven’t done all of the…

01:27:37 Do you think if you were there in the early universe,

01:27:39 you would have been able to predict

01:27:41 the emergence of chemistry and biology?

01:27:43 And I ask that because at this stage as humans,

01:27:46 do you think we can possibly predict the length of memory

01:27:51 that might be able to be formed later on

01:27:55 in this pocket of the universe?

01:27:56 Like how complex is, what is the ceiling of assembly?

01:28:02 I think as much time as you have in the past

01:28:04 is how much you can predict in the future.

01:28:06 Because it’s actually physical in the system

01:28:09 and you have to have enough time

01:28:10 for features of that structure to exist.

01:28:15 Wait, let me push back on that.

01:28:17 Isn’t there somewhere in the universe

01:28:19 that’s like a shortest path that’s been,

01:28:21 that stretches all the way to the beginning?

01:28:23 Yeah.

01:28:24 That’s building some giant monster?

01:28:26 Maybe, yeah.

01:28:27 Yeah.

01:28:28 So you can’t predict.

01:28:29 The universe has as much memory

01:28:30 as the largest assembly object in the universe.

01:28:32 Yeah. Right.

01:28:33 But so you can’t predict.

01:28:35 You can’t predict any deeper than that, no.

01:28:37 Right.

01:28:38 So like that, I guess what I’m saying is,

01:28:40 like what intuition do you have about complexity

01:28:43 living in the world that you’d have today?

01:28:46 Right, because you just, you can,

01:28:49 I mean I guess how long does it get more fun?

01:28:54 Like isn’t there gonna be at some point,

01:28:55 because there’s a heat death in the universe,

01:28:57 isn’t there going to be a point of the most,

01:29:01 of the highest assembly of object,

01:29:04 with the highest probability being generated?

01:29:07 When is the universe gonna be the most fun,

01:29:08 and can we freeze ourselves and then live then?

01:29:10 Exactly.

01:29:11 And will you know when you’re having the most fun

01:29:14 that this is the best time, you’re in your prime?

01:29:16 Are you going to do what everyone does,

01:29:18 which is deny that you’re in your prime,

01:29:19 and the best years are still ahead of you?

01:29:21 I don’t know.

01:29:23 What option do you have?

01:29:26 I don’t, I mean the problem is there’s lots of,

01:29:29 lots of really interesting features here.

01:29:31 I just wanna mention one thing that might be,

01:29:33 is I do think assembly theory applies all the way back

01:29:36 to subatomic particles.

01:29:37 And I also think that cosmological selection

01:29:40 might’ve been actually, there might’ve been,

01:29:42 I would say it’s a really boring bit,

01:29:43 but it’s really important for a cosmologist

01:29:45 that universes have gone through.

01:29:47 Was it Lee Smolin who proposed this?

01:29:49 Maybe that there is this,

01:29:50 that basically a universe evolves,

01:29:52 you’ve got the wrong constants, we’ll start again.

01:29:54 And the most productive constants

01:29:56 where you can allow particles to form in a certain way,

01:29:59 propagate to the next universe, and we go again.

01:30:01 So actually selection goes all the way back,

01:30:03 and there’s this cycle of universes.

01:30:04 And now this universe has been selected

01:30:06 because life can occur, and it carries on.

01:30:10 But I’ve really butchered that.

01:30:12 There is a much more.

01:30:13 So this is some aspect where through the selection process

01:30:17 there’s parameters that are being fine tuned,

01:30:19 and we happen to be living in one

01:30:20 where there’s some level of fine tuning.

01:30:22 Is there, given that, can you steel man the case

01:30:27 that we humans are alone in the universe?

01:30:31 We are the highest assembly index object in the universe.

01:30:34 Yeah, I can, I guess.

01:30:36 Sad though.

01:30:37 I mean, so from a.

01:30:38 Is it possible?

01:30:39 Yes, it’s possible.

01:30:41 Let’s assume.

01:30:43 Well, we know.

01:30:45 I mean, it’s possible.

01:30:47 So let me, so okay, so there is a particular

01:30:50 set of elements on Earth in a particular ratio,

01:30:54 and the right gravitational constant,

01:30:57 and the right viscosity, you know,

01:30:59 of stuff being able to move around,

01:31:01 the right distance from our sun,

01:31:04 right number of events where we have a moon,

01:31:07 the Earth is rotating.

01:31:10 The late heavy bombardment produced a lot of,

01:31:14 brought in the right stuff.

01:31:16 And Mars was cooking up, you know,

01:31:19 the right molecules first.

01:31:21 So it was habitable before Earth.

01:31:23 It was actually doing the combinatorial search.

01:31:25 And before Mars kind of became uninhabitable,

01:31:29 it seeded Earth with the right molecular replicators.

01:31:34 And there was just the right stuff on Earth,

01:31:36 and that’s how the miracle of life occurred.

01:31:40 Although I find I’m very uncomfortable with that

01:31:43 because actually, because life came so quickly

01:31:48 in the Earth’s past.

01:31:50 But that doesn’t mean that life is easy elsewhere.

01:31:55 It just might mean that,

01:31:57 because chemistry is actually not a long term thing.

01:32:00 Chemistry can happen quickly.

01:32:01 So maybe going on with the steel manning of the argument

01:32:04 to say actually, the fact that life emerged quickly

01:32:06 doesn’t mean that life is easy.

01:32:08 It just means that the chemistry was right on Earth,

01:32:11 and Earth is very special.

01:32:13 And that’s why there’s no life

01:32:15 anywhere else in the universe.

01:32:17 Yeah, so Sarah mentioned this kind of cascading thing.

01:32:21 So what if that’s the reason we’re lucky,

01:32:24 is that we got to have a rare cascading of,

01:32:29 like an accelerating cascading effect

01:32:31 in terms of the complexity of things.

01:32:34 So like, maybe most of the universe

01:32:36 is trying to get sticky with the memory,

01:32:39 and it’s not able to really form it.

01:32:41 And then we got really lucky in that.

01:32:43 And it has nothing,

01:32:44 like there’s a lot of Earth like conditions, let’s say,

01:32:47 but it’s just you really, really have to get lucky on this.

01:32:51 But I’m doing experiments right now.

01:32:54 In fact, experiments that Sarah and I are working on,

01:32:56 because we have some joint funding for this,

01:32:58 where we’re seeing that the universe

01:33:00 can get sticky really quickly.

01:33:01 Now, of course, we’re being very anthropocentric,

01:33:04 we’re using laboratory tools, we’re using theory,

01:33:06 but actually, the phenomena of selection,

01:33:10 the process of developing heterogeneity,

01:33:14 we can do in the lab.

01:33:15 We’re just seeing the very first hints of it.

01:33:17 And wouldn’t it be great if we can start to pin down

01:33:22 a bit more precisely becoming good Bayesianists for this,

01:33:27 for the origin of life and the emergence of life,

01:33:30 to finding out what kind of chemistries

01:33:31 we really need to look for.

01:33:34 And I’m becoming increasingly confident

01:33:36 we’ll be able to do that in the next few years.

01:33:37 Make life in the lab or make some selection in the lab

01:33:41 from inorganic stuff, from sand, from rocks,

01:33:44 from dead stuff, from moon.

01:33:45 Wouldn’t it be great to get stuff from the moon,

01:33:48 put it in our origin of life experiment,

01:33:51 and make moon life?

01:33:53 And restrict ourselves to interesting self replicating

01:33:56 stuff that we find on the moon.

01:33:58 Sarah, what do you think about this approach

01:34:00 of engineering life in order to understand life?

01:34:03 So building life in the machine.

01:34:05 Yeah, so, I mean, Lee and I are trying right now

01:34:09 to build a vision for a large institute

01:34:14 or experimental program, basically, to do this problem.

01:34:17 But I think of it as like, we need to simulate a planet.

01:34:21 So like the Large Hadron Collider was supposed

01:34:23 to be simulating conditions just after the Big Bang.

01:34:26 Lee’s built a lot of technology in his lab

01:34:27 to do these kind of selection engines.

01:34:30 But the question you’re asking is,

01:34:33 how many experiments do you need to run?

01:34:34 What volume of chemical space do you need to explore

01:34:37 before you actually see an event?

01:34:40 And I like to make an analogy

01:34:41 to one of my favorite particle physics experiments,

01:34:43 which is Super Kamiakande that’s looking

01:34:45 for the decay of the proton.

01:34:46 So this is something that we predicted theoretically,

01:34:49 but we’ve never observed in our universe.

01:34:51 And basically what they’re doing is every time

01:34:53 they don’t see a proton decay event,

01:34:55 they have a longer bound on the lifetime of a proton.

01:34:57 So imagine we built an experiment with the idea in mind

01:35:00 of trying to simulate planetary conditions,

01:35:03 physically simulate.

01:35:04 You can’t simulate origin life in a computer.

01:35:06 You have to do it in an experiment.

01:35:08 Simulate enough planetary conditions

01:35:10 to explore the space of what’s possible

01:35:12 and bound the probability for an origin life event.

01:35:15 Even if you’re not observing it,

01:35:16 you can talk about the probability.

01:35:17 But we, hopefully, life is not exponentially rare

01:35:23 and we would then be able to evolve

01:35:27 in an automated system alien life in the lab.

01:35:31 And if we can do that, then we understand the physics

01:35:34 as well as we understand what we can do

01:35:35 in particle accelerators.

01:35:37 So keep expanding physically the simulation,

01:35:40 the physical simulation, until something happens.

01:35:44 Yeah, or just build a big enough volume

01:35:46 of chemical experiments and evolve them.

01:35:48 So if you say volume, you mean like literally volume?

01:35:50 I mean physical volume in terms of space,

01:35:53 but I actually mean volume in terms

01:35:54 of the combinatorial space of chemistry.

01:35:57 So like.

01:35:58 How do you nicely control the combinatorial exploration,

01:36:01 the search space, such that it’s always like

01:36:05 you keep grabbing the low hanging fruit?

01:36:08 Yeah, how do you build a search engine for chemistry?

01:36:10 It’s like for aliens. I think you explained it really well.

01:36:11 We should carry on doing this.

01:36:12 I should pretend the physics, be the physicist,

01:36:14 you be the chemist.

01:36:15 So the way to do it is I will always play a joke.

01:36:18 Cause I like writing grants to ask for money

01:36:23 to do cool stuff.

01:36:24 But years ago I started wanting to build.

01:36:29 So I actually wanted the weather.

01:36:30 So I built this robot in my lab called the computer,

01:36:32 which is this robot you can program to do chemistry.

01:36:36 Now it’s a pro.

01:36:37 I made a programming language for the computer

01:36:39 and made it operate chemical equipment.

01:36:43 Originally I wrote grants to say,

01:36:46 Hey, I want to make an origin of life system.

01:36:48 And no one would give me any money for this.

01:36:51 They said, what this is ridiculous.

01:36:53 Why are you wanting to make, oh, it’s really hard.

01:36:55 It takes forever.

01:36:56 You’re not a very good origin of life chemist anyway.

01:36:58 Why would we give you any money?

01:37:00 And so I turned it around and said, can you,

01:37:02 can instead, can you give me money to make robots,

01:37:05 to make molecules are interesting.

01:37:07 And everyone went, yeah, okay, you can do that.

01:37:09 And that’s, so actually the funny thing is the computer

01:37:13 project, which I have in my lab, which is very briefly,

01:37:18 it’s just basically, it’s like literally an automated

01:37:20 test tube.

01:37:21 And we’ve made a programming language for the test tube,

01:37:22 which is cool, has come as literally came from this.

01:37:28 I went to my lab one day.

01:37:29 So I want to make a search engine to get origin of life

01:37:32 because they don’t have a planet.

01:37:33 And I thought about doing in a microfluidic format.

01:37:36 So microfluidic is very nano, very small channels

01:37:39 in device where you can basically have all the pipes

01:37:41 lit dump produced by lithography.

01:37:43 And you can have a chamber, maybe say between say 10

01:37:46 and a hundred microns in volume.

01:37:48 And we slot them all together like Lego,

01:37:50 and we can make an origin of life system.

01:37:52 And I could never get it to work.

01:37:55 And I realized I had to make, do chemistry at the kind of

01:37:59 test tube level and what you want to be able to do.

01:38:03 Yeah, it goes back to that tweet in 1981.

01:38:06 1981, the computer, we’re looking at a tweet from Lee.

01:38:10 In 1981, the computer was a distant dream in,

01:38:13 oh wow, this is the scientist looking back.

01:38:15 It is the young boy who dreamed.

01:38:19 In 2018, it was realized, spelled in a British way,

01:38:23 realized, which is the wrong way.

01:38:25 Yeah, I’m starting with Z, but not.

01:38:27 So now there’s a system that does the physical

01:38:30 manifestation or whatever the programming language,

01:38:34 the spec tells you to do.

01:38:37 Yeah, well in 1981, I got my first computer, ZX81.

01:38:40 What was the computer?

01:38:41 ZX81.

01:38:42 ZX81.

01:38:44 Sinclair ZX81.

01:38:46 It was, and I got a chemistry set.

01:38:49 And I liked the chemistry set and I liked the computer

01:38:53 and I just wanted to put them together.

01:38:54 I thought, wouldn’t it be cool if I could just use

01:38:56 the computer to control the chemistry set.

01:38:58 And obviously that was insane.

01:39:00 And I was like, you know, eight years old, right?

01:39:04 Nine years old, going on nine years old.

01:39:06 And then I invented the computer

01:39:12 just because I wanted to build this origin of life grid.

01:39:15 Which is like literally a billion test tubes

01:39:17 connected together in real time and real space,

01:39:20 basically throwing a chemical dice.

01:39:23 Throw dice, throw dice, throw dice.

01:39:24 You’re gonna get lucky.

01:39:26 And that’s what we, I think Sarah and I

01:39:28 have been thinking very deeply about.

01:39:30 Because, you know, there’s more money being spent

01:39:32 on the origin of the gravity

01:39:37 or looking at the Higgs boson than the origin of life, right?

01:39:39 And the origin of life is the, I think the biggest question

01:39:43 or not the biggest question, it is a big question.

01:39:46 Let’s put it that way.

01:39:47 It is the biggest question.

01:39:48 You’re okay saying that.

01:39:49 Okay, all right.

01:39:50 Isn’t it possible once you figure out the origin of life

01:39:52 that that’s not going to solve,

01:39:55 that’s not actually gonna solve

01:39:58 the question of what is life?

01:39:59 Because you’re kind of putting a lot of.

01:40:02 Yeah, I think they’re the same problem.

01:40:04 But you’re putting, is it possible

01:40:06 that you’re putting too many,

01:40:09 too much bets into this origin part?

01:40:12 Maybe the origin thing isn’t,

01:40:14 isn’t there always a turtle underneath the turtle?

01:40:16 Isn’t there a stack of turtles?

01:40:17 Because then if you create it in the lab,

01:40:19 maybe you need some other stuff.

01:40:21 Well, let’s not think about the origin.

01:40:23 Like in the lab, there’s still memory.

01:40:27 Yeah, yes.

01:40:28 Right?

01:40:29 So the experiment is already the product of evolution.

01:40:31 Right, in some maybe really deep way,

01:40:33 not an obvious way, in some very deep way.

01:40:36 So maybe the haters are always going to be like,

01:40:39 well, you have to reconstruct the fold.

01:40:42 You have to build a new script.

01:40:42 Fortunately for us, the haters are not aware

01:40:44 of that argument.

01:40:46 Well, no, I know, I just.

01:40:48 We’re the one making that argument usually, but yeah.

01:40:51 I just think that if we create life in the lab,

01:40:54 it’s not obvious that you’ll get

01:40:56 to the deep, deep understanding of necessarily,

01:41:00 what is the line between life and non life?

01:41:03 No, I think, so there’s so much here.

01:41:05 I’m just like playing devil.

01:41:06 So much here, but let me play devil’s advocate

01:41:08 back in a previous conversation, right?

01:41:10 And say, yeah, I will.

01:41:14 Why not?

01:41:15 Why not?

01:41:16 We’ve got time.

01:41:16 Yeah, let’s go.

01:41:17 Cellular automata.

01:41:18 Cellular automata, these very, very simple things

01:41:21 where you color squares black or white

01:41:24 and implement rules and play them in time.

01:41:26 And you can get these very, very complex patterns coming out.

01:41:30 You know, there’s nice rules.

01:41:31 There are Turing complete rules and I would argue

01:41:36 that cellular automata don’t really exist on their own.

01:41:41 They have to exist in a computing device.

01:41:45 If that, well, that’s computing devices,

01:41:46 a piece of paper and abstraction,

01:41:47 a mathematician drawing a grid or a framework.

01:41:52 Now, so I would argue CAs are beautiful things,

01:41:56 simple, going complex, but the complexity is all borrowed

01:41:59 from the lithography, the numbers.

01:42:03 Right, now let’s take that same argument

01:42:05 with the chemistry experiment origin of life.

01:42:11 What you need to be able to do is go,

01:42:12 and I’m inspired to do this,

01:42:13 to go out and look for CAs occur in nature.

01:42:17 You know, let’s kind of, let’s find some CAs

01:42:21 that just emerge in our universe and.

01:42:24 For people just, sorry to interrupt,

01:42:26 for people just listening and in general,

01:42:29 I think what we’re looking at is a cellular automata

01:42:33 where again, as Lee described,

01:42:35 there is just binary black or white squares

01:42:38 and they only have local information

01:42:40 and they’re born and they die.

01:42:42 And you would think nothing interesting would emerge,

01:42:45 but actually what we’re looking at is something

01:42:47 that I believe is called glider guns

01:42:49 or a glider gun, which is moving objects

01:42:54 in this multi cell space that look like they’re organisms

01:42:58 that have much more information,

01:43:01 that have much more complexity

01:43:03 than the individual building components.

01:43:05 In fact, look like they have a long term memory

01:43:09 while the individual components don’t seem like

01:43:12 they have any memory at all, which is fascinating.

01:43:15 The argument here is that has to exist

01:43:17 on all this layer of infrastructure, right?

01:43:20 And though it looks simple.

01:43:21 And then what I would make, the argument I would make

01:43:23 if I were you, say, well, I think CAs are really simple

01:43:26 and everywhere, is say, show me how they emerge

01:43:28 in a substrate.

01:43:29 Now this goes to the origin of life, machine.

01:43:32 I don’t think we want to do the origin of life,

01:43:34 just any origin is good.

01:43:36 So we do, so we literally have our sand shaker,

01:43:39 shake the sand like massive grid of chemistry experiments,

01:43:42 shaking sand, shaking whatever.

01:43:45 And then because we know what we’ve put in,

01:43:47 so we know how we’ve cheated and the same way with CA,

01:43:49 we know how we’ve cheated,

01:43:50 we know the number of operations needed,

01:43:52 we know how big a grid we want to get this.

01:43:54 If we could then say, okay,

01:43:56 how can we generate this recipe in the lab

01:44:01 and make a life form?

01:44:03 What contingency did we need to put in?

01:44:06 And we’re upfront about how we cheated, okay?

01:44:09 Say, oh, you had to shake it, it was a periodic,

01:44:12 planet rotates, it’s tried, comes in and out.

01:44:15 So, and then we can start to basically say,

01:44:17 okay, how difficult is it for these features to be found?

01:44:23 And then we can look for extra planets and other features.

01:44:25 So I think Sarah is absolutely right.

01:44:27 We want to explain to people we’re cheating.

01:44:29 In fact, we have to cheat.

01:44:30 No one has given, I’m good at writing grants,

01:44:32 well, I used to be, I’m not very good right now,

01:44:34 I keep getting rejected,

01:44:35 but writing a grant for a planet in 100 million years,

01:44:38 no grant fund there is going to give me that,

01:44:39 but maybe money to make a kind of a grid,

01:44:42 a computer grid, origin of life computer grid.

01:44:45 In physical space.

01:44:46 In physical space, and just do it.

01:44:49 So Sarah said something which is,

01:44:51 you can’t simulate the origin of life in a computer,

01:44:56 so like in simulation, why not?

01:44:59 What are your, you said it very confidently,

01:45:02 so is it possible and why would it be very difficult?

01:45:07 Like what’s your intuition there?

01:45:08 I think it’s very difficult right now

01:45:12 because we don’t know the physics,

01:45:13 but if you go based on principles of assembly theory

01:45:15 and you think every molecule is actually

01:45:17 a very large causal graph, not just the molecule,

01:45:20 then you have to simulate all the features

01:45:21 of those causal graphs,

01:45:22 and I think it becomes computationally intractable.

01:45:24 You might as well just build the experiment.

01:45:26 Because you have, in the physical space,

01:45:28 you have all the objects with all the memories.

01:45:31 Yes.

01:45:32 In the computer, you would have to copy them

01:45:35 or reconstruct them.

01:45:36 Yes.

01:45:37 Yeah, that’s a beautifully put,

01:45:38 and I would say that lots of people,

01:45:41 you just don’t have enough resources.

01:45:44 It’s easier to actually do the physical experiment

01:45:47 because we are literally,

01:45:49 I would view the physical experiment

01:45:51 almost like a computational experiment.

01:45:53 We’re just outsourcing, it’s just basically,

01:45:55 we’re just outsourcing all the matrix algebra.

01:45:57 And on your point about the experiment

01:46:00 being also an example of life,

01:46:03 it’s almost like you want to design,

01:46:05 it’s like all of us are lineages

01:46:08 of propagating information across time,

01:46:10 and so everything we do becomes part of life

01:46:12 because it’s part of that causal chain.

01:46:14 So it’s like you want to try to pinch off

01:46:15 as much as you can of the information

01:46:17 from your causal chain that goes into the experiment,

01:46:20 but you can’t pinch off all of it

01:46:21 to move it to a different timeline.

01:46:23 It’s always going to be part of your timeline.

01:46:25 But at least if you can control

01:46:26 how much information you put in,

01:46:27 you can try to see how much does that particular trajectory

01:46:30 you’ve set up start generating its own assembly.

01:46:33 So you know where it starts,

01:46:36 and then you want to try to see it take off on its own

01:46:38 when you try to pinch it off as much as possible.

01:46:42 Got it.

01:46:43 Quick pause, bathroom break.

01:46:45 Yes.

01:46:46 All right, cool.

01:46:47 And now we’re back.

01:46:49 All right.

01:46:50 We talked about the early days of the universe

01:46:53 when there was just stuff and no memory,

01:46:55 not even causality.

01:46:57 I think Lee at least implied

01:46:58 that causality is immersion somehow.

01:47:00 We could discuss this.

01:47:01 What happened before this all originated?

01:47:06 What’s outside the universe?

01:47:10 Divided by zero.

01:47:12 Okay, so it’s not relevant, not understandable.

01:47:16 Is it useful to even ask the question?

01:47:18 No.

01:47:19 Just because it’s so hard?

01:47:22 No, it’s not hard.

01:47:23 It’s just not a question.

01:47:24 If I can’t do an experiment or even think of an experiment,

01:47:27 the question doesn’t exist.

01:47:29 Well, no, you can’t think of a lot of experiments,

01:47:32 no offense.

01:47:33 What I mean is I can’t.

01:47:35 Your causality graph is like,

01:47:36 this is what we’re talking about.

01:47:38 It’s like there is limits to your ability

01:47:43 to construct experiments.

01:47:45 I agree, but I was trying to be facetious

01:47:48 and I’m trying to make a point

01:47:49 because I think that if there is a causal bottleneck

01:47:56 through which information can’t propagate in principle,

01:48:00 then it’s very hard to think of an experiment,

01:48:03 even in principle, even one that’s beyond

01:48:05 my mediocre intellect, right?

01:48:08 Which is fine.

01:48:09 I’m happy to accept that.

01:48:10 But this is one of the things I actually do think

01:48:12 there was something before the Big Bang

01:48:14 because I would say that I think the Big Bang

01:48:17 just couldn’t occur and create time.

01:48:19 Time created the Big Bang.

01:48:21 So there was time before the Big Bang.

01:48:23 Yeah.

01:48:24 There was no space, but there was time.

01:48:25 Yeah.

01:48:26 Yeah.

01:48:27 But I mean, I’m just making that stuff up

01:48:29 just to make all the physicists happy,

01:48:31 but I think it’s…

01:48:32 Do you think that would make them happy

01:48:33 because they would be quite upset, actually.

01:48:35 And why would they be upset?

01:48:37 Because they would say that time can’t exist

01:48:40 before the Big Bang.

01:48:41 Yeah, I mean, this goes back to an argument

01:48:43 that you might not want to have the argument here.

01:48:46 I was talking to Sarah earlier today

01:48:48 about an argument we had about time a long time ago.

01:48:51 Yeah.

01:48:52 A long time in time.

01:48:53 And what I would, it’s like, I think there is this thing

01:48:55 called time or state creation.

01:48:57 The universe is creating states and it’s outside of space,

01:49:00 but they create space.

01:49:02 So what I mean is you can imagine there are states

01:49:04 being created all the time.

01:49:05 And there is this thing called time.

01:49:08 Time is a clock, which you can use to measure

01:49:11 when things happen, but that doesn’t mean,

01:49:13 because you can’t measure something,

01:49:15 that states aren’t being created.

01:49:17 And so you might locally refer to the Big Bang

01:49:22 and the Big Bang occurred at some point

01:49:25 when those states were there.

01:49:27 Probably there had to be enough states

01:49:28 for the Big Bang to occur.

01:49:30 And then, but I think that there is something wrong

01:49:34 with our conception of how the universe was created

01:49:37 and the Big Bang because we don’t really get time.

01:49:40 Because again, I don’t want to become boring

01:49:44 and sound like a broken record, but time is a real thing.

01:49:48 And until I can really explain that more elegantly,

01:49:52 I’m just gonna get into more trouble.

01:49:54 Well, we’re gonna talk about time

01:49:55 because time is a useful measuring device for experiments,

01:49:59 but also time is an ideal, okay.

01:50:02 But let me first ask Sarah, what do you think?

01:50:05 Is it a useful question to ask what happened

01:50:07 before the Big Bang?

01:50:08 Is it a useful question to ask what’s outside the universe?

01:50:14 So I would think about it as the Big Bang

01:50:17 is an event that we reconstructed

01:50:19 as probably happening in the past of our universe

01:50:21 based on current observational data.

01:50:24 And so the way I like to think about it

01:50:25 is we exist locally in something called the universe.

01:50:33 So, and going back to like the physics of existence,

01:50:35 we exist locally in the space of all things that could exist

01:50:39 and we can infer certain properties of the structure

01:50:42 of where we exist locally.

01:50:43 And one of the properties that we’ve inferred in the past

01:50:46 is that there is a thing we call the Big Bang.

01:50:50 There’s some signatures of our local environment

01:50:52 that indicate that there was a very low information event

01:50:58 that started our universe.

01:50:59 I think that’s actually just an artifact

01:51:02 of the structure of the assembly space

01:51:05 that when you start losing all the memory in the objects,

01:51:10 it looks like what we call a Big Bang.

01:51:13 So I think it makes sense to talk about

01:51:14 where you are locally.

01:51:16 I think it makes sense to talk about

01:51:18 counterfactual possibilities,

01:51:20 what could exist outside the universe

01:51:22 in the sense that they become part of our reasoning

01:51:24 and therefore part of our causal chain

01:51:26 of things that we can do.

01:51:28 So like the multiverse in my mind exists,

01:51:31 but it doesn’t exist as a multiverse

01:51:33 of possible universes.

01:51:34 It exists as an idea in our minds

01:51:36 that allows us to reason about how physics works

01:51:38 and then to do physics differently

01:51:39 because we reason about it that way.

01:51:42 So I always like to recenter it on things exist,

01:51:47 but they don’t always exist like we think they exist.

01:51:50 So when we’re thinking about things outside the universe,

01:51:53 they absolutely exist because we’re thinking about them,

01:51:55 but they don’t look like the projections in our minds.

01:51:59 They’re something else.

01:52:00 And something you said just gave an idea

01:52:02 to go back to your question.

01:52:04 If there was, I mean, if something caused the Big Bang,

01:52:09 if there was some memory or some artifact of that,

01:52:12 then of course, to answer your question,

01:52:13 it’s worth going back to that

01:52:15 because that would imply there is something

01:52:17 beyond that barrier, that filter.

01:52:19 And that’s what you were saying, I guess, right?

01:52:21 I’m agnostic to what exists outside the universe.

01:52:23 I just don’t think that.

01:52:24 I think the most interesting things for us to be doing

01:52:26 are finding explanations that allow us to do more,

01:52:31 like that optimism.

01:52:32 So I tend to draw the boundary on questions I ask

01:52:35 as being scientific ones because I find

01:52:38 that that’s where the most creative potential is

01:52:40 to impact the future trajectory

01:52:42 of what we’re doing on this planet.

01:52:43 It’s an interesting thing about the Big Bang

01:52:45 is basically from our current perspective

01:52:48 of what we’re able to detect,

01:52:50 it’s the time when things were forgotten.

01:52:52 Yes.

01:52:54 It’s the time to reset from our limited perspective.

01:52:58 And so the question is, is it useful to ever study

01:53:01 the thing that was forgotten?

01:53:05 Or should we focus just on the memories

01:53:07 that are still there?

01:53:08 Well, the point I was trying to make about the experiment

01:53:10 is I was trying to say both things.

01:53:11 And I think perhaps yes, from the portfolio point of view,

01:53:14 if you could then imagine what was forgotten

01:53:17 and then work forwards,

01:53:19 you will have different consequences.

01:53:20 So then it becomes testable.

01:53:22 So as long as we can find tests,

01:53:24 then it’s definitely worth thinking about.

01:53:26 What I don’t like is when physicists say

01:53:28 what happened before the Big Bang

01:53:29 and before, before, before,

01:53:31 without giving me any credible conjecture

01:53:34 about how would we know the difference?

01:53:38 But the way you framed it is quite nice.

01:53:39 I like that.

01:53:40 It’s like, what have we forgotten?

01:53:43 Is there room for God in assembly theory?

01:53:47 Who’s God?

01:53:49 I like arguments for a necessary being better than God.

01:53:52 Well, I think I said it earlier.

01:53:53 What’s a necessary being?

01:53:54 What’s a necessary?

01:53:55 Like something that has to exist.

01:53:58 Oh, so you like, I mean, you like the shortest path.

01:54:01 Like does God need?

01:54:02 No, no, no.

01:54:03 I mean, well, you can go back to like Thomas Aquinas

01:54:05 and arguments for the existence of God.

01:54:08 But I think most of the interesting theological arguments

01:54:11 are always about whether something has to exist

01:54:14 or there was a first thing that had to exist.

01:54:17 But I think there’s a lot of logical loopholes

01:54:19 in those kind of arguments.

01:54:19 Well, so God here, meaning the machine

01:54:24 that creates, that generates the stuff.

01:54:29 But God, so what I was trying to say earlier is that.

01:54:31 Isn’t that just the universe though?

01:54:32 Yeah, yeah.

01:54:33 Well, yeah, well, but there’s a difference between,

01:54:36 I sort of imagine like a black box, like a machine.

01:54:40 Yeah.

01:54:41 I mean, I would be more comfortable calling that God

01:54:44 because it’s a machine.

01:54:45 You go into a room and there’s a thing with a button.

01:54:47 Yeah, I don’t like the great programmer in the sky version.

01:54:50 Yeah, but if it’s more kind of like,

01:54:54 I don’t like to think of, if you look at a cellular automata,

01:54:59 if it’s the cells and the rules,

01:55:03 that doesn’t feel like God

01:55:04 that generates a bunch of stuff.

01:55:06 But if there’s a machine like that does,

01:55:10 that runs the cellular automata and set the rules,

01:55:13 then that feels like God.

01:55:15 That sort of, in terms of terminology.

01:55:19 So I wonder if there’s like a machine

01:55:21 that’s required to generate this universe.

01:55:23 That’s very sort of important for running this in the lab.

01:55:26 So as I said earlier, I think I said this earlier,

01:55:29 that I can’t remember the phrase, but something like,

01:55:32 I mean, does God exist in our universe?

01:55:34 Yes.

01:55:35 Where does God exist?

01:55:36 God at least exists in abstraction in our minds,

01:55:40 particularly of people who have religious faith

01:55:44 they believe in.

01:55:45 But let’s then take, but you’re talking a little bit more

01:55:47 about generics, say, well,

01:55:49 is there a mechanism beyond the universe you’re calling God?

01:55:51 I would say God did not exist at the beginning,

01:55:55 but he or she does now.

01:55:59 Because I’m saying the mechanism.

01:56:00 Well, you don’t know that he didn’t exist in the beginning.

01:56:03 So like this could be us in our minds trying to,

01:56:07 like just listening to gravitational waves,

01:56:10 detecting gravitational waves.

01:56:12 It’s the same thing.

01:56:13 It’s us trying to go back further and further

01:56:16 into our memories to try to understand the machines

01:56:19 that make up, that make up us.

01:56:23 And so it’s possible that we’re trying to grasp

01:56:27 at possible kind of what kind of machines could create.

01:56:32 There’s always a tweet.

01:56:33 There’s always a tweet.

01:56:36 If the universe is a computer, then God must have built it

01:56:39 because computers need creators.

01:56:41 There you go.

01:56:43 And then Joshe Bach replied,

01:56:46 since there’s something rather than nothing,

01:56:48 perhaps existence is the default.

01:56:52 If existence is the default, then many computers exist.

01:56:56 Creator gods are necessary computers,

01:56:59 unnecessarily computers too.

01:57:00 I’m very confused by that, but that’s an interesting idea

01:57:03 that existence is the default versus nonexistence.

01:57:05 I agree with that, but the rest is not.

01:57:07 And then Lee responds,

01:57:08 perhaps this reasoning is incomplete.

01:57:11 That’s how scientists talk trash each other

01:57:14 on Twitter apparently.

01:57:16 Which part don’t you agree with?

01:57:18 When he said if existence is default,

01:57:20 then many computers exist,

01:57:22 this comes back to the inventor and discovery argument.

01:57:26 I would say the universe at the beginning wasn’t capable

01:57:29 of computation because there wasn’t enough technology,

01:57:33 enough states.

01:57:34 So what you’re saying is if God is a mechanism,

01:57:38 so I might actually agree,

01:57:40 but then the thing is lots of people see God

01:57:43 as more than a mechanism.

01:57:44 For me, God could be the causal graph in assembly theory

01:57:47 that creates all the stuff that the memories we know.

01:57:50 And the fact that we can even relate to each other

01:57:52 is because we have the same, we share that heritage.

01:57:54 And why we love each other

01:57:56 or we like to see God in each other

01:57:58 is it’s just we know we have a shared existence.

01:58:04 So if the God is the mechanism

01:58:06 that created this whole thing,

01:58:08 I think a lot of people see God in a religious sense

01:58:12 as that mechanism also being able to communicate

01:58:15 with the objects it creates.

01:58:18 And if it’s just the mechanism,

01:58:19 we won’t be able to communicate with the objects it creates.

01:58:23 It can only create.

01:58:24 You can’t interact with the…

01:58:29 Well, there’s versions of God that create the universe

01:58:30 and then left.

01:58:33 Yeah, like spark.

01:58:35 For some religions.

01:58:36 The first spark, yeah.

01:58:37 But I think I liked your analogy

01:58:40 of the machine and the rules, right?

01:58:42 But I think part of the problem is,

01:58:47 I mean, we have this conception

01:58:48 that we can disentangle the rules

01:58:50 from the physical substrate, right?

01:58:51 And that’s the whole thing about software and hardware

01:58:53 being separate or the way Newton wrote his laws

01:58:56 that there was some,

01:58:57 like they exist outside the universe.

01:58:58 They’re not actually a feature of the universe.

01:59:00 They don’t have to emerge out of the universe itself.

01:59:03 So I think if you merge your two views,

01:59:06 then it gets back to the God is the universe.

01:59:08 And then I think the deeper question

01:59:10 is why does it seem like there’s meaning and purpose?

01:59:12 And if I think about the features of the universe

01:59:16 that give it the most meaning and purpose,

01:59:17 those are what we would call

01:59:19 the living components of the universe.

01:59:21 So if you wanted to say God is a physically real thing,

01:59:24 which you were saying

01:59:25 is like an emergent property of our minds,

01:59:26 but I would just say the way the universe

01:59:30 creates meaning and purpose,

01:59:31 there is really a physics there.

01:59:32 It’s not like a illusory thing.

01:59:35 And that is just what the physics of life is.

01:59:39 Is it possible that we’ve forgotten

01:59:41 much of the mechanisms that created the universe?

01:59:45 So like, so basically, you know,

01:59:48 whatever, if God is that mechanism,

01:59:50 we just leave parts of that behind.

01:59:52 Well, but the universe is constantly generating itself.

01:59:54 So if God is that mechanism,

01:59:56 it would be that that would still be active today.

01:59:58 I don’t, like, I’m agnostic,

02:00:00 but if I recall the things I believe in God

02:00:05 in the way that some people talk about God,

02:00:07 I would say that God is, you know,

02:00:09 like in the universe now, it’s not an absent thing.

02:00:16 So I think there’s a mislabeling here

02:00:18 because you’re, I mean, I mean, I’m a professional idiot,

02:00:25 actually, but, but, um.

02:00:27 You should put that on your CV.

02:00:29 Yeah.

02:00:30 Professionally, not recreationally or amateur,

02:00:33 but professionally, you’re paid for it.

02:00:36 I would say if you were talking about God,

02:00:38 I mean, again, I’m way out, way out of my depth here,

02:00:40 and I almost feel uncomfortable.

02:00:41 Yeah, but I feel quite uncomfortable articulating,

02:00:43 but I’ll try.

02:00:45 For me, a lot of people that think of God as a consciousness

02:00:47 and a reasoning entity that actually has causal power,

02:00:52 and you’re, and so you’re, it’s like,

02:00:56 then you’re saying like gravity could be God

02:00:57 or time could be God.

02:00:58 I mean, I think for me, for my conception of time

02:01:02 is probably as fundamental as God

02:01:04 because it gave rise to human intelligence and consciousness

02:01:07 in which we can have this abstract notion of God.

02:01:12 So I think that you’re maybe talking about God

02:01:15 in a very mechanistic, kind of unsophisticated sense,

02:01:19 whereas other people say that God is more sophisticated

02:01:21 and got all this, you know, feelings and love

02:01:23 and, you know, and this abstracting ability.

02:01:27 So is that what, or do you mean that?

02:01:29 Do you mean God as in this conscious entity

02:01:32 that decided to flick the universe into existence?

02:01:36 Well, one of the features that God would have

02:01:40 is the ability to flick the universe into existence.

02:01:46 I, you know, like Windows 95,

02:01:49 I don’t know if God is Windows 95 or Windows XP

02:01:52 or Windows 10, I don’t know the full feature set.

02:01:55 So at the very least, you have to flick the universe

02:01:58 into existence, and then other features might include

02:02:03 ability to interact with that universe in interesting ways,

02:02:07 and then how do you interact with the universe

02:02:11 in interesting ways?

02:02:11 You have to be able to speak the language

02:02:13 of its different components.

02:02:15 So in order to interact with humans,

02:02:17 you have to know how to act humanlike.

02:02:24 So I don’t know, but it seems like

02:02:31 whatever mechanism created the universe

02:02:33 might want to also generate local pockets of mechanisms

02:02:40 that can interact with that.

02:02:42 Like inject.

02:02:44 Like God was lonely?

02:02:46 Yeah, I mean, it could be just a teenager

02:02:49 and another just playing a video game.

02:02:51 Yeah, maybe.

02:02:52 Well, I was gonna say, I mean, I don’t,

02:02:54 so this is referring to our origin of life engine.

02:02:57 It’s like, I don’t believe in God,

02:02:59 but that doesn’t mean I don’t wanna be one.

02:03:01 Right.

02:03:02 I don’t wanna make a universe and make a life form,

02:03:04 but that may be rude to people who have

02:03:07 dear religious beliefs.

02:03:08 What I mean by that is if we are able to create

02:03:13 an entirely new life form, different chemistry,

02:03:16 different culture, what does it make us?

02:03:20 By that definition, it makes us gods, right?

02:03:23 Well, there is.

02:03:24 I mean, like when you have children,

02:03:24 you’re like one of the magical things of that

02:03:28 is you’re kind of mini gods.

02:03:29 I mean, first of all, from a child’s perspective,

02:03:32 parents are gods for quite a while.

02:03:34 And then, I mean, in the positive sense,

02:03:38 there’s a magic to that.

02:03:39 That’s why I love robotics,

02:03:41 is you instill life into something,

02:03:43 and that makes you feel godlike in a sort of positive way.

02:03:50 Being a creator is a positive feeling.

02:03:51 Creator, yeah, exactly, on a small scale.

02:03:54 And then God would be a creator

02:03:57 at the largest possible scale, I suppose.

02:03:59 Okay, you mentioned offline the Assembletron.

02:04:05 Assembletron.

02:04:06 Assembletron.

02:04:07 Yep.

02:04:08 What’s an Assembletron?

02:04:09 This is an early idea of something you’re thinking about.

02:04:12 So Sarah’s team, well, I think Sarah’s team

02:04:15 are interested in using AI to understand life.

02:04:19 My team is, and I’m wondering if we could apply

02:04:23 the principles of assembly theory,

02:04:25 that is the causal structure that you get

02:04:28 with assembly theory, and hybridize it,

02:04:30 and make a new type of neuron, if you like.

02:04:34 I mean, there are causal neural networks out there,

02:04:37 but they are not quite the architecture

02:04:40 of what I would like.

02:04:41 I would like to associate memory bits with,

02:04:45 basically, I’d like to make a,

02:04:47 rather than having an ASIC for neural networks,

02:04:50 I wanna make an ASIC for assembly networks, right?

02:04:54 And.

02:04:56 So can you say that again?

02:04:57 Assembly networks.

02:05:00 So what is a thing with an input and an output,

02:05:05 and it’s like a neural network type of thing,

02:05:07 what does it do exactly?

02:05:08 What’s the input, what’s the output?

02:05:09 So in this case, so if you’re talking about

02:05:11 a general neural network, I mean,

02:05:13 in general neural network, you can train it

02:05:15 on any sort of data, right, depending on the framework,

02:05:20 whether it’s like text, or image data, or whatnot.

02:05:25 And that’s fine, but there’s no causal structure

02:05:30 associated with that data.

02:05:31 Now just imagine, rather than, you know,

02:05:33 let’s say we’re gonna classify a difference

02:05:35 between cat and dog, right?

02:05:36 Classic cat and dog neural network.

02:05:38 What about if the system understood the assembly space,

02:05:42 it created the cat and the dog,

02:05:44 and rather than guessing what was happening,

02:05:47 and training on those images,

02:05:49 and not understanding those features,

02:05:53 you almost like, you could imagine doing a,

02:05:56 going back a step and doing, and training,

02:05:59 going back a step and doing the training,

02:06:01 going back a step, back a step, back a step,

02:06:02 and I wonder if that is actually the origin of intelligence,

02:06:06 or how we’ll crack intelligence.

02:06:08 Because we need to, because we’ll create

02:06:11 the entire graph of events, and be able to kind of

02:06:17 look at cause and effect across those graphs.

02:06:19 I’m explaining it really badly,

02:06:20 but it’s a gene of an idea, and I’m guessing

02:06:23 very smart, very rich people in AI are already doing this.

02:06:28 Trying to not generate cats and dogs,

02:06:31 but trying to generate things of high assembly index.

02:06:35 Yeah, and I think, and also using causal graphs

02:06:39 in neural networks, and machine learning, and deep learning,

02:06:42 maybe building a new architecture.

02:06:44 I’m just wondering, is there something

02:06:45 we can get out of assembly theory

02:06:47 that allows us to rebuild current machine learning

02:06:49 architectures to give causation more cheaply?

02:06:54 I mean, I don’t know if that’s what you,

02:06:56 we’ve been inventing this for a little while,

02:06:58 but we’re trying to finish the theory paper first

02:07:00 before we do anything else.

02:07:01 Yeah, you also wanna have, say, goal directed behavior

02:07:05 in neural networks, then assembly theory

02:07:07 is a good framework for doing that.

02:07:08 Daniel’s been thinking about that a lot.

02:07:10 And I think it’s a really interesting idea

02:07:12 that you can map concepts from how neural networks learn

02:07:16 to thinking about goal directed behavior

02:07:18 as a learning process.

02:07:20 That you’re learning a specific goal.

02:07:21 The universe is learning a goal

02:07:22 when it generates a particular structure,

02:07:24 and that you can map that physical structure

02:07:25 in a neural network.

02:07:29 What’s the goal?

02:07:31 Well, in a neural network, you’re designing the goal.

02:07:34 In biology, I mean, people are not supposed to use

02:07:39 teleological language in biology, which is ridiculous,

02:07:41 but, because goals are real things.

02:07:45 They’re just post selected.

02:07:47 So you can talk about goals after the fact.

02:07:51 Once a goal emerges in the universe,

02:07:52 that physical entity has a goal.

02:07:54 But Lee and I came up with a test for,

02:07:57 like a Turing test for goal directed behavior

02:08:00 based on the idea of assembly.

02:08:01 We have to formalize it still,

02:08:03 but I would like to write a paper on it.

02:08:04 But the basic idea is if you had two systems

02:08:09 that were completely equivalent,

02:08:11 you know, like in the instantaneous,

02:08:14 like physical experimental setup.

02:08:16 So Lee has to figure out how to do this.

02:08:18 But there was something that would be different

02:08:19 in their future.

02:08:21 And there was a symmetry breaking you observe

02:08:23 in the present based on that possibility

02:08:25 of that future outcome.

02:08:26 Then you could say that that system

02:08:28 had some representation of some kind of goal in mind

02:08:31 about what it wanted to do in the future.

02:08:33 And I, so goals are interesting

02:08:35 because they don’t exist as instantaneous things.

02:08:38 They exist across time,

02:08:40 which is one of the reasons that assembly theories

02:08:42 may be more naturally able to account

02:08:45 for the existence of goals.

02:08:48 So goals are, they only exist in time,

02:08:52 or they manifest themselves in time

02:08:55 through, you said, symmetry breaking.

02:08:58 So it’s almost like, imagine,

02:09:01 like if representations in your mind are real, right?

02:09:05 And you can imagine future possibilities,

02:09:08 but imagine everything else is physically equivalent.

02:09:11 And the only thing that you actually change

02:09:13 your decision based on is what you model

02:09:15 as being the future outcome.

02:09:17 Then somehow that representation in your mind

02:09:20 of the future outcome becomes causal

02:09:21 to what you’re doing now.

02:09:22 So there’s kind of like retro causal effect,

02:09:24 but it’s not actually retro causal.

02:09:25 It’s just that your assembly space

02:09:28 is actually includes those possibilities

02:09:31 as part of the structure.

02:09:32 It’s just, you’re not observing all the features

02:09:34 of the assembly space in the current moment.

02:09:36 Or the possibilities exist, but they don’t become a goal

02:09:40 until they’re realized.

02:09:42 So one of the features of assembly space

02:09:44 that’s super interesting,

02:09:45 and it’s easier to envision with like Legos, for example,

02:09:48 is if you’re thinking about an assembly space,

02:09:50 you can’t observe the entire assembly space

02:09:52 in any instant in time.

02:09:54 So if you imagine a stack of Legos,

02:09:55 and you wanna look at the assembly space of a stack of Legos,

02:09:58 you have to break the Legos apart,

02:10:00 and then you look at all the possible ways

02:10:03 of building up the original object.

02:10:05 So now you have in your mind

02:10:06 the goal of building that object,

02:10:08 and you have all the possible ways of doing it.

02:10:10 And those are actual physical features of that object,

02:10:13 but that object doesn’t always exist.

02:10:14 What exists is the possibility of generating it.

02:10:17 And the possibilities are always infinite.

02:10:20 Well, for that particular object,

02:10:22 like it has a well defined assembly space.

02:10:25 And I guess what I’m saying is that object

02:10:27 is the assembly space,

02:10:28 but you actually have to unpack that object across time

02:10:31 to view that feature of it.

02:10:33 It’s only an observable across time.

02:10:35 The term goal is such a important

02:10:39 and difficult to explain concept, right?

02:10:41 Because what you want is a way is like,

02:10:44 I think only conscious beings can have conscious goals.

02:10:49 Everything else is doing selection.

02:10:51 And but selection does invent goals.

02:10:53 And in a way that the way that biology reinterprets the past

02:10:58 in the present is kind of helped you to understand

02:11:03 there was a goal in the past now, right?

02:11:05 It’s kind of like goals only exist back in time.

02:11:09 So first of all, only conscious beings

02:11:14 can have conscious goals.

02:11:17 I’m not even gonna touch that one.

02:11:19 Why?

02:11:21 Go for it, come on.

02:11:23 The line between conscious goals

02:11:25 and non conscious goals, exactly.

02:11:28 And also maybe just on top of that,

02:11:31 you said a Turing test for goal directed behavior.

02:11:33 What does a Turing test potentially look like?

02:11:39 So if you’ve got two objects, we were thinking about this.

02:11:40 So we actually got some funding

02:11:42 to work to go on two teams.

02:11:43 So I’m trying to do, and part of this

02:11:46 is I’m trying to do a bit of theory

02:11:47 and Sarah is teaching me a bit of theory

02:11:49 and Sarah is trying to design experiments

02:11:51 and I’m teaching experiments.

02:11:52 Cause I think it’s really good for us to have that.

02:11:53 So say, when would a, so that’s good.

02:11:58 I like this, I’m sure we’re using Dan Dennett essay.

02:12:02 Yeah, and I can explain why we wouldn’t want to call it

02:12:04 a Turing test after, but Dan Dennett.

02:12:05 So Dan Dennett wrote this really nice essay

02:12:07 about herding cats and freewill inflation.

02:12:12 The title is so brilliant.

02:12:14 That’s the actual title?

02:12:14 That’s the title, yeah.

02:12:15 Herding cats and freewill inflation.

02:12:18 Yeah, something like that.

02:12:19 I mean, it’s not, maybe not.

02:12:20 And so.

02:12:21 No, I think that’s right.

02:12:22 So if you’ve got a, let’s imagine you’ve got two objects

02:12:23 on a hillside, okay?

02:12:26 And it just happens to be a snowy hill

02:12:28 and let’s just say you see an object

02:12:29 get rolling down the hill.

02:12:31 Or you, and the rock rolls down the hill

02:12:34 but the start goes to the end.

02:12:36 How do you know that object’s had a goal?

02:12:38 Now you unveil the object

02:12:39 and you’ll see it’s actually a skier.

02:12:41 And the skier starts at the top and goes down the bottom.

02:12:44 Great.

02:12:45 Then you look at the rock.

02:12:46 Rock rolls down the hill and goes to the bottom.

02:12:48 How can you tell the difference between the two?

02:12:51 So, and what Dan says is like, well, this is clear.

02:12:54 The skier’s in control.

02:12:56 And because they’re adjusting the trajectory,

02:12:58 so there’s some updating going on.

02:13:00 Then the only way you could really do that

02:13:01 is if you put the skier back to the top of the hill again,

02:13:04 they would tend to start roughly in the same space

02:13:06 and probably take all that complex set of trajectories

02:13:09 and end up pretty much at the same finish point, right?

02:13:12 With plus or minus a few meters.

02:13:13 Whereas if it was just a random rock

02:13:15 going down to a random trajectory, that wouldn’t happen.

02:13:18 And so what Sarah and I were kind of doing

02:13:20 when we were writing this grant,

02:13:21 we were like, we need to somehow instantiate the skier

02:13:25 and the rock in an experiment.

02:13:27 And then say, okay, when does the object,

02:13:31 so for an object to have a goal, it has to have an update.

02:13:34 It has to have some sensing and some kind of,

02:13:36 you know, inbuilt actuation to respond to the environment.

02:13:41 And then we just have to iterate on that.

02:13:44 And maybe Sarah, you can then fill in the Turing test part.

02:13:47 Well, yeah, I guess the motivation for me

02:13:49 was slightly different.

02:13:50 So I get really frustrated about conversations

02:13:52 about consciousness as most people do.

02:13:54 You know, a lot of people are,

02:13:56 which is not necessarily related to free will directly

02:13:59 or to this goal directed behavior.

02:14:01 But I think there’s a whole set of bundled

02:14:03 and related topics here.

02:14:04 But I think for me, I was, you know,

02:14:07 everybody’s always interested in explaining

02:14:09 intrinsic experience and quantifying intrinsic experience.

02:14:12 And there’s all sorts of problems with that

02:14:14 because you can never actually be another physical system.

02:14:17 So you can’t know what it’s like

02:14:18 to be another physical system.

02:14:20 So I always thought there must be some way

02:14:21 of getting at this problem about

02:14:23 if an agent or an entity is conscious

02:14:26 or at least has internal representations

02:14:28 and those are real physical things,

02:14:30 that it must have causal consequences.

02:14:34 So the way I would ask the question of consciousness

02:14:38 is not, you know, what is it like intrinsically?

02:14:40 But if things have intrinsic experience,

02:14:43 is there any observable difference from the outside

02:14:45 about the kind of causation

02:14:47 that that physical system would enact in?

02:14:50 And for me, the most interesting thing that humans do

02:14:52 is have imagination.

02:14:53 So like we can imagine rockets centuries

02:14:56 before we build them.

02:14:57 They’ve become real physical things

02:14:59 because we imagine them.

02:15:00 And people might disentangle that from conscious experience,

02:15:02 but I think a lot of the sort of imagination we do

02:15:04 is actually a conscious process.

02:15:06 So then this becomes a question of

02:15:09 if I were observing systems and I said,

02:15:12 one had an internal representation,

02:15:15 which is slightly different

02:15:16 than a conscious experience, obviously.

02:15:17 So I’m entangling some concepts,

02:15:19 but it’s a loose set of thought experiments.

02:15:22 Then how, and I set them up

02:15:24 in a physically equivalent situation.

02:15:28 Would it be the case that

02:15:30 there would be experimental observables associated with it?

02:15:34 And that became the idea of trying to actually

02:15:38 measure for internal representation and conscious.

02:15:40 So Turing basically didn’t wanna do that.

02:15:42 You just wanted the machine that could emulate

02:15:44 and trick you into having the behavior,

02:15:46 but never dealt with the internal experience

02:15:48 because he didn’t know how to do that.

02:15:51 And I guess I was wondering,

02:15:53 is there a way to set up the experiment

02:15:55 where you could actually test for that?

02:15:57 For imagination that led to the thing.

02:16:00 That there was something internal going on,

02:16:02 some kind of inner world, as people say,

02:16:06 or you could say, it actually is an agent,

02:16:09 it’s making decisions, it has an internal representation.

02:16:13 And whether you say that’s experience or not

02:16:15 is a different thing, but at least the feature

02:16:18 that there’s some abstraction it’s doing

02:16:20 that’s not obvious from looking at the physical substrates.

02:16:23 Do you think it’s possible to do that kind of thing?

02:16:25 One of the compelling things about the Turing test

02:16:27 is that defining intelligence,

02:16:30 defining any complicated concept

02:16:32 as a thing like observing it from the surface

02:16:37 and not caring about what’s going on deep inside

02:16:40 because how do you know?

02:16:42 That’s the point.

02:16:43 So the idea is exactly that.

02:16:45 So what we’re trying to do,

02:16:46 the Turing test for goal directedness

02:16:47 is literally take some objects

02:16:50 that clearly don’t have any internal representation,

02:16:52 grains of sand blowing on the beach or something,

02:16:55 and I don’t know, a crab wandering around on the beach

02:16:59 and then generating an experiment

02:17:00 where literally the experiment generates an entity

02:17:04 that literally has no internal representation to sand,

02:17:07 these are oil droplets actually,

02:17:08 what we’ve got in mind, a robot that makes oil droplets.

02:17:11 But then what we wanna try and do

02:17:13 is train the oil droplets to be like crabs,

02:17:16 give them an internal representation,

02:17:18 give them the ability to integrate information

02:17:21 from the environment so they remember the past,

02:17:27 are in the present, and can imagine a future.

02:17:31 And in a very limited way, their kind of game engine,

02:17:33 their limited simulation of the world

02:17:36 allows them to then make a decision.

02:17:38 They’re objects across time.

02:17:40 So then you would run a bunch of crabs

02:17:43 like over and over and over and over?

02:17:44 How many crabs, Lee?

02:17:45 How many, is there, what’s,

02:17:47 because you have to have a large number of crabs,

02:17:49 what does your theory say, is there a mathematical?

02:17:52 We’re working on it, I mean, this is literally.

02:17:54 Limit, crab limit.

02:17:55 There’s literally a.

02:17:57 Excellent.

02:17:57 There’s literally a.

02:17:58 What’s the herding cats have to do?

02:17:59 Oh, that’s random, wait, what’s cats,

02:18:02 in the title by Daniel Dennett,

02:18:04 Herding Cats and the Free Will Inflation.

02:18:06 So, I love this.

02:18:07 What does herding cats mean,

02:18:08 what does free will inflation mean?

02:18:10 So this, I love this essay,

02:18:13 because it explained to me

02:18:14 how I can live in a deterministic universe,

02:18:18 but have, not free will, but have freedom.

02:18:24 And also it helped me explain

02:18:26 that time needed to be a real thing in this universe.

02:18:30 So what basically Dan was saying here is like,

02:18:33 how do you, how do these cats appear

02:18:35 to just do what they want, right?

02:18:38 And if you live in a deterministic universe,

02:18:41 why do the cats do these things?

02:18:43 You know, aren’t they just, isn’t it all obvious?

02:18:46 And how does free will inflate the universe?

02:18:48 And for me, I mean, probably I love the essay

02:18:51 because my interpretation of the essay

02:18:53 in assembly theory makes complete sense.

02:18:57 Because you need an expanding universe

02:19:00 in assembly theory to create novelty

02:19:04 that you search for,

02:19:07 that then when you find something interesting

02:19:09 and you keep doing it because it’s cool

02:19:10 and it gives you an advantage,

02:19:13 then it appears in the past to be a goal.

02:19:15 So what does, in assembly theory,

02:19:17 the expansion of the universe look like?

02:19:20 What are we talking about?

02:19:22 Why does the expansion of the universe

02:19:26 give you more possibilities of novelty and cool stuff?

02:19:30 So for me, I don’t think about the universe

02:19:32 in terms of big bang and space.

02:19:34 I think about it in terms of the big memory expansion.

02:19:37 That you have one, you only have the ability

02:19:39 to store one bit of information,

02:19:41 so then you can’t do very much.

02:19:43 So what the universe has been doing since forever,

02:19:46 it’s been creating more,

02:19:49 it’s been increasing the size of its RAM, okay?

02:19:53 So it’s like one megabyte, two megabyte,

02:19:55 three megabyte, four megabytes, all the way up.

02:19:57 And so the more RAM you have,

02:20:00 the more you can remember about the past,

02:20:04 which allows you to do cooler things in the future.

02:20:07 So if you can remember how to launch a rocket,

02:20:09 then you might be able to imagine how to land a rocket,

02:20:12 and then relaunch, reland, and carry on.

02:20:15 And so you’re able to expand the space

02:20:19 and remember the past.

02:20:21 And so that’s why I think it’s very important.

02:20:23 But not a perfect memory.

02:20:25 It’s an interesting question,

02:20:27 whether there’s some forgetting that happens

02:20:28 that might increase.

02:20:30 Is the expansion of the forgetting, at some point,

02:20:34 accelerate faster than the remembering?

02:20:37 I think that that’s a very important thing

02:20:38 that probably intelligence does,

02:20:40 and we’re gonna learn in machine learning about,

02:20:42 because you want machine learning right now,

02:20:44 or artificial intelligence right now,

02:20:45 doesn’t have memory right,

02:20:47 but you want the ability to,

02:20:49 or not for, if you want to get to human like consciousness,

02:20:52 you need to have the ability, I suppose, to remember stuff

02:20:55 and then to selectively forget stuff

02:20:57 so you can re remember it and compress it.

02:20:59 Arguably, the way that we come up with new physical laws.

02:21:02 I think that there is a great deal to be gained

02:21:12 from having the ability to remember things,

02:21:15 but then when you forget them,

02:21:18 you can then have a,

02:21:18 you can basically do the simulation again

02:21:20 and work out if you get to that compressed representation.

02:21:23 So that’s in cycles.

02:21:25 So cycles of remembering and forgetting

02:21:29 are probably important,

02:21:31 but there shouldn’t be excuse to have a universe

02:21:33 with no memory in it.

02:21:34 The universe is gonna remember that it forgot,

02:21:37 but just not tell you.

02:21:40 I’m looking at this paper

02:21:42 and it’s talking about a puppet controlling a puppet

02:21:44 controlling a puppet controlling a puppet controlling

02:21:46 a puppet controlling a puppet,

02:21:48 conceptually easy to understand,

02:21:49 but physically impossible,

02:21:50 as physically impossible as predicting a fair coin toss.

02:21:53 I don’t know what he’s talking about,

02:21:54 but there’s pictures of puppets controlling puppets.

02:21:58 Let me ask you,

02:21:59 there’s a few things I wanna ask,

02:22:01 but we brought up time quite a bit.

02:22:04 You guys tweet about time quite a bit.

02:22:08 What is time in all of this?

02:22:10 We kind of mentioned it a bunch.

02:22:13 Is it not important at all in terms of,

02:22:15 is it just a word?

02:22:16 Should we be talking about causality mostly?

02:22:18 Like Sarah, what do you think?

02:22:20 Is, we’ve talked about like memories.

02:22:24 Is that the fundamental thing

02:22:25 that we should be thinking about?

02:22:26 And time is just a useful measurement device or something

02:22:29 like that.

02:22:30 Well, there’s different concepts of time, right?

02:22:32 So I think in assembly theory,

02:22:33 when we’re talking about time,

02:22:34 we’re talking about the ordering of things.

02:22:36 So that’s the causal graph part.

02:22:38 And so then the fundamental structure of the universe

02:22:40 is that there is a certain ordering

02:22:42 and certain things can’t happen till other things happen.

02:22:44 But usually when we colloquially talk about time,

02:22:48 we’re talking about the flow of time.

02:22:51 And I guess Lee and I were actually debating

02:22:52 about this this morning.

02:22:53 So in talking on it, walking on the river here,

02:22:56 which is a very lovely spot for talking about time,

02:23:00 but that when the universe is updating,

02:23:02 it’s transitioning between things that exist now

02:23:05 and things that exist now.

02:23:09 That’s really the flow of time.

02:23:11 So you have to separate out those concepts at bare minimum.

02:23:15 And then there’s also an arrow of time

02:23:17 that people talk about in physics,

02:23:19 which is that time doesn’t appear to have a directionality

02:23:21 in fundamental physics, but it does to us, right?

02:23:25 Like we can’t go backwards in time.

02:23:26 And usually that would be explained in physics

02:23:29 in terms of, well, there’s a cosmological arrow of time,

02:23:32 but there’s also the thermodynamic arrow of time

02:23:33 of increasing entropy.

02:23:36 But what we would say in assembly theory

02:23:37 is that there is a clear directionality.

02:23:39 The universe only runs in one direction,

02:23:40 which is why some things, it’s easy to make,

02:23:42 if the universe runs in one direction,

02:23:45 it’s easy to make processes look reversible.

02:23:47 For example, if they have no memory,

02:23:49 they’re easy to run forward and backwards,

02:23:51 which is why the laws of physics that we have now

02:23:52 look the way they do, because they involve objects

02:23:55 that have no memory.

02:23:56 But when you get to things like us,

02:23:58 it becomes very clear that the universe

02:23:59 has a directionality associated to it.

02:24:01 So it’s not reversible at all.

02:24:03 It’s the no man ever steps in the same river.

02:24:06 I just have to bring that out

02:24:07 because you walked on the river.

02:24:08 No man ever steps in the same river twice

02:24:11 before he’s not the same river and he’s not the same man.

02:24:14 So it’s not reversible, any of the same.

02:24:16 No, but reversibility is an emergent property, right?

02:24:19 So we think of the reversibility laws as being fundamental

02:24:22 and the irreversibility as being emergent.

02:24:23 But I think what we would say from how we think about it,

02:24:26 and it certainly seems to be the case

02:24:28 for our perception of time,

02:24:29 but also what’s happening in biological evolution,

02:24:33 you can make things reversible,

02:24:35 but it requires work to do it.

02:24:37 And it requires certain machines

02:24:39 to run it forward and backward.

02:24:40 And Chiara Marletto is working some interesting ideas

02:24:42 on constructor theory related to that,

02:24:44 which is a totally different set of ideas.

02:24:46 You can travel back in time sometimes.

02:24:49 Yes.

02:24:52 You can’t travel actually back in time,

02:24:54 but you could reconstruct things

02:24:56 that have existed in the past.

02:24:58 You’re always moving forward in time,

02:24:59 but you can cycle through.

02:25:01 Like, I mean, I can…

02:25:02 Can I clarify what you just said?

02:25:02 Yeah, yeah, go for it.

02:25:03 Quickly, you travel forward in time to travel back.

02:25:06 Yes.

02:25:07 Thank you, that really clarified it.

02:25:09 What Sarah’s saying is you don’t go back in time,

02:25:12 you recreate what happened in the past in the future

02:25:14 and inspect it again.

02:25:15 So in that local pocket of time,

02:25:17 it’s as if you travel back in time.

02:25:19 So I don’t…

02:25:20 How is that not traveling back in time?

02:25:22 Because you’re not going back

02:25:23 to your same self back in time.

02:25:24 You’re creating that in the future.

02:25:27 But everything else is the same as it was in the past.

02:25:29 No, no, no, no, it’s not in registry.

02:25:32 I mean, it goes back to the big question I’m saying.

02:25:35 I mean, this is something I was trying to look up today

02:25:38 when we first had this discussion

02:25:40 and I was talking to Sarah on Skype and said,

02:25:42 by the way, time is the fundamental thing in the universe.

02:25:45 She almost hung up on me.

02:25:47 Right, but you can even…

02:25:48 I mean, if you wanna make an analogy to computation,

02:25:51 and I think Charles Bennett actually has a paper on this

02:25:52 like about reversible computation

02:25:55 and reversible Turing machines.

02:25:57 In order to make it reversible,

02:25:58 you have to store memory to run the process backwards.

02:26:01 So time is always running forward in that

02:26:04 because you have to write the memory.

02:26:05 You can’t erase the memory.

02:26:06 You can erase the memory,

02:26:08 but the point when you go back to zero, right?

02:26:11 But the whole point is that in order to have a process

02:26:15 that even runs in both directions,

02:26:17 you have to start talking about memory

02:26:19 to store the information to run it backwards.

02:26:21 I got it.

02:26:22 So you can’t really then…

02:26:24 You can’t have it exactly how it was in the past.

02:26:27 Yeah, exactly.

02:26:27 You have extra stuff, extra baggage always.

02:26:30 Okay.

02:26:31 A really important thing that I want to say on this,

02:26:33 I think if I try and get it right,

02:26:34 I have to say that if you can think

02:26:36 that the universe is expanding

02:26:38 in terms of the number of boxes

02:26:41 that it has to store states, right?

02:26:44 And this is where the directionality of the universe

02:26:47 comes from, everything comes from.

02:26:48 You could erase what’s in those boxes,

02:26:51 but the fact you’ve now got so many boxes at time now

02:26:54 in this present, there’s more of those boxes

02:26:56 than there were in the past.

02:26:57 See, but the boxes aren’t physical boxes.

02:27:00 They’re not space or time.

02:27:03 Why is the number of boxes always expanding?

02:27:05 It’s very hard to imagine this

02:27:07 because we live in space.

02:27:09 So what I’m saying, which is I think probably correct,

02:27:14 is that we just, let’s just imagine for a second,

02:27:17 there is a nonlocal situation,

02:27:20 but there are these things called states

02:27:23 and that the universe irrespective

02:27:26 of whether you measure anything,

02:27:28 there is a universal, let’s call it a clock

02:27:31 or a state creator.

02:27:33 Maybe we can call it, that’s why maybe you can call it God,

02:27:36 but let’s call it a state creator

02:27:37 where the universe is expanding

02:27:39 in the number of states it has.

02:27:41 Why are you saying it’s expanding though?

02:27:43 Is that obvious that it’s expanding?

02:27:44 It’s obvious because that’s where the,

02:27:46 because we, we.

02:27:49 That’s a source of novelty.

02:27:50 It’s a source of novelty

02:27:51 and it also explains why the universe is not predictable.

02:27:56 How do you know it’s not predictable?

02:27:57 I just like interrupting you.

02:27:59 Sorry, it’s fun, because you’re struggling.

02:28:01 I’m struggling because I’m trying to be

02:28:03 as concrete as possible and not sound like I’m insane.

02:28:06 Yeah.

02:28:07 And I’m not insane.

02:28:08 It’s obvious because you,

02:28:13 I’m a chemist.

02:28:14 So as a chemist, I grew into the world

02:28:17 understanding irreversibility.

02:28:19 Irreversibility is all I knew.

02:28:22 And when people start telling me

02:28:24 the universe is actually reversible, it’s a magic trick.

02:28:27 We can use time to do it.

02:28:28 So what I mean is that the second law

02:28:32 is really the magical.

02:28:38 But why does it need to be magical?

02:28:40 The universe is just asymmetric.

02:28:41 All I’m saying is the universe is asymmetric

02:28:43 in the state production and we can erase those states,

02:28:48 but we just have more computational power.

02:28:50 So what I’m saying is that the universe’s

02:28:53 deterministic horizon,

02:28:55 this is one of the reasons we can’t live in a simulation,

02:28:57 by the way, you can’t live in a simulation.

02:29:01 The irreversibility.

02:29:02 Yeah, yeah, so basically every time

02:29:04 you try and simulate the universe,

02:29:06 in this, you know, I live in a simulation,

02:29:07 the universe is expanded in states.

02:29:09 You’re like, oh damn it,

02:29:10 I need to make my computer bigger again.

02:29:11 And every time you try and contain the universe

02:29:13 in the computation,

02:29:14 because it’s got bigger in number of states.

02:29:16 And so I’m saying the fact the universe has novelty in it

02:29:21 is going to turn out experimentally to be proof

02:29:25 that time, as I’ve labeled it, is fundamental

02:29:30 and exists as a physical thing that creates space.

02:29:34 Okay, so if you can prove that novelty

02:29:37 is always being created,

02:29:39 you’re saying that it’s possible to also then prove

02:29:41 that it’s always expanding in the state space.

02:29:45 Those are things that have to be proven.

02:29:48 That’s what we’re working on experiments for, yeah.

02:29:50 And you’re trying to, like by looking at the sliver

02:29:52 of reality, show that there’s always novelty being generated.

02:29:56 Yeah, because if we go live in a universe

02:29:59 that conventional physicists would live in,

02:30:01 it’s a big lookup table of stuff and everything exists.

02:30:05 I want to prove that that book doesn’t exist,

02:30:10 it’s continuously being added pages on.

02:30:12 So all I’m saying, if the universe is a book,

02:30:14 we started, the universe at the beginning only had no pages

02:30:17 and had one page, another page, another page.

02:30:19 Whereas a physicist would now say all the pages exist

02:30:22 and we could in principle access them.

02:30:24 I’m saying that is fundamentally incorrect.

02:30:27 Do you know what’s written in this book?

02:30:29 The free will question.

02:30:32 Is there room for free will in this view

02:30:38 of the universe is generating novelty

02:30:40 and getting greater and greater assembly structures built?

02:30:45 Sarah.

02:30:46 Yes.

02:30:48 Okay.

02:30:49 Done.

02:30:49 Next question.

02:30:51 Why, what’s the source of free will in this?

02:30:53 Well, I think it depends on what you mean by free will, but.

02:30:58 Yeah, well, please, there’s a lot.

02:31:00 I think what I’m interested in

02:31:02 as far as the phenomena of free will is,

02:31:06 do we have individual autonomy and agency?

02:31:10 And when I do things, is it really me

02:31:13 or is it my atoms that did it?

02:31:16 And that’s the part that’s interesting to me.

02:31:19 I guess there’s also the determinism versus randomness part.

02:31:23 But the way I think about it is like each of us

02:31:25 are like a thread or like an assembly space

02:31:30 through this giant possibility space.

02:31:33 And it’s like we’re moving on our own trajectory

02:31:37 through that space and that is defined by our history.

02:31:39 So we’re sort of causally contingent on our past,

02:31:42 but also because of the sort of intersection

02:31:46 of novelty generation, it’s not completely predetermined

02:31:50 by the past.

02:31:52 And so then you have the causal control

02:31:55 of the determinism part that you are your causal history

02:31:59 and there’s some determinism from that past,

02:32:01 but there’s also room for creativity.

02:32:04 And I think it’s actually necessary

02:32:06 that something like free will exists

02:32:08 if the universe is gonna be as creative as possible.

02:32:11 Because if I were all intelligent being,

02:32:16 inventing a universe, and I wanted it to have

02:32:19 maximum number of interesting things happen,

02:32:22 again, we should come up with the metric of interesting,

02:32:25 but generating, yes, I know,

02:32:28 generating maximal possibilities,

02:32:31 then I would want the agents to have free will

02:32:33 because it means that they’re more individual,

02:32:38 like each entity actually is a different causal force

02:32:41 in the universe.

02:32:43 And it’s intrinsic and local property of that system.

02:32:46 There’s a greater number of distributed agents?

02:32:50 Like are you always creating more and more individuality?

02:32:54 Kind of.

02:32:56 I would say you’re creating more causal power, but.

02:32:59 So causal power, the word consciousness,

02:33:01 is the causal power somehow correlated with consciousness?

02:33:05 I mean, that’s why I have this conception

02:33:07 of consciousness being related to imagination,

02:33:09 because the more that we can imagine can happen

02:33:11 and the more counterfactual possibilities you have in mind,

02:33:14 the more you can actually implement.

02:33:16 And somehow free will is also at the intersection

02:33:18 of the counterfactual becoming the actual.

02:33:21 So can you elaborate on that a little bit,

02:33:23 that consciousness is imagination?

02:33:25 I don’t know exactly how to articulate it,

02:33:27 and I’m sure people will aim at certain things I’m saying,

02:33:31 but I think the language is really imprecise,

02:33:33 so I’m not the best way to.

02:33:35 It’s really interesting, like what is imagination

02:33:37 and what role does it play in the human experience,

02:33:42 in the experience of any agent?

02:33:44 Yes, I love imagination.

02:33:46 I think it’s like the most amazing thing we do.

02:33:49 But I guess one way I would think about it is,

02:33:51 we talked about the transition to life

02:33:52 being the universe acquiring memory,

02:33:55 and life does something really interesting.

02:33:56 You just think about biology generally.

02:33:58 It remembers states of the past to adapt

02:34:00 to things that happen in the future.

02:34:01 So the longer life has evolved on this planet,

02:34:04 the deeper that past is, the more memory we have,

02:34:06 the more kinds of organisms and things.

02:34:08 But what human level intelligence has done

02:34:11 is quite different.

02:34:12 It’s not just that we remember states

02:34:14 that the universe has existed in before,

02:34:16 it’s that we can imagine ones that have never existed,

02:34:19 and we can actually make them come into existence.

02:34:22 And I think that’s the most unique feature

02:34:25 about the transition to whatever we are

02:34:27 from what life on this planet has been doing

02:34:30 for the last four billion years.

02:34:32 And I think it’s deeply related to the phenomenon

02:34:33 we call consciousness.

02:34:34 Yeah, I was gonna, I mean, just agree with that.

02:34:36 I think that consciousness is the ability

02:34:38 to generate those counterfactuals.

02:34:40 Now, whether you can say, you know,

02:34:41 are there degrees of consciousness?

02:34:43 I mean, I’m sorry, panpsychist,

02:34:47 but electrons don’t have counterfactuals,

02:34:49 although they do have some kind of,

02:34:51 they are able to search a space and pathways.

02:34:53 But I think that there is a very concrete,

02:34:58 or concrete, there’s a very specific property

02:35:01 that humans have, and I don’t know if it’s unique to humans.

02:35:05 I mean, maybe dogs can do it, and birds can do it, right?

02:35:09 And where they are basically solving a problem,

02:35:11 because consciousness was invented,

02:35:13 or this abstraction was invented by evolution

02:35:16 for a specific reason.

02:35:19 And so look, one of the reasons why I came to the conclusion

02:35:23 that time was fundamental was actually because

02:35:26 Sarah and I had a completely different.

02:35:30 The most heated debate on Skype chat ever.

02:35:32 No, no, no, we had to, I was like.

02:35:34 Stop it.

02:35:35 No, no, it goes back to the free will thing.

02:35:37 So I think that, although I’ve changed my view a bit,

02:35:40 because there’s some really interesting physicists

02:35:42 out there who talks about how the measurement problem

02:35:45 in Newtonian space, but I don’t want to go there just now,

02:35:49 because I think I’ll mess it up.

02:35:50 But briefly, I could not see how the universe,

02:35:54 how we can have free will.

02:35:56 And I mean, this is really boring,

02:35:57 because this is like, this is a well trodden path,

02:35:59 but I mean, not so boring.

02:36:01 I suppose it’s kind of, we just want to be precise.

02:36:04 If the universe is deterministic,

02:36:05 how can we have free will, right?

02:36:07 So Sarah’s a physicist.

02:36:11 I think she, not believe, can show that most of the laws

02:36:15 we have are deterministic to some degree.

02:36:17 Quantum mechanics onto Newtonian stuff.

02:36:20 And yet there’s Sarah telling me she believes in free will.

02:36:22 And I’m like, your belief system’s broken here, right?

02:36:26 Because you’re demanding free will

02:36:29 in a deterministic universe.

02:36:32 And then I realized that I agreed with her

02:36:36 that I do think that free will is a thing

02:36:39 because we are able to search for novelty.

02:36:41 And then that’s where I came to the conclusion

02:36:43 that time, the universe is expanding in terms of novelty.

02:36:46 And it goes back to that Dan Dennett essay

02:36:48 that we’re talking about, the free will inflation.

02:36:51 Free will, so you have, so the past,

02:36:55 it did not exist in the past.

02:36:56 The past exists in the present.

02:36:58 What I mean is like, you are the, there was no past.

02:37:01 There is only present.

02:37:03 So that means you are the sum total.

02:37:05 Everything that’s occurs in the past

02:37:06 is manifestly here in the present.

02:37:09 And then you have this little echo state

02:37:12 in your consciousness because you’re able to,

02:37:14 you’re able to imagine something without actualization.

02:37:18 But the fact you imagine it, that occurs in electrons

02:37:20 and potassium ion flows in your neural network

02:37:23 in your brain.

02:37:24 Maybe consciousness is just the present.

02:37:27 So somehow you imagine that, and then by imagining,

02:37:31 oh, that’s good, yeah.

02:37:33 I’m gonna make a robot to do this thing and program it.

02:37:36 And then you physically then go and do it.

02:37:39 So that then changes the future, sorry.

02:37:42 What’s imagination?

02:37:44 Does it require the past?

02:37:46 Does it require the future?

02:37:47 Does it require memory?

02:37:48 Does it?

02:37:49 It’s imagination.

02:37:51 Does it only exist in the moment?

02:37:53 So imagination is, well, yeah, probably

02:37:57 it’s an instantaneous readout of what’s going on.

02:37:59 You can maybe, your subconscious brain

02:38:01 has been generating all the bits for it,

02:38:03 but no, imagination occurs when you,

02:38:07 in your game engine, you remember the past

02:38:10 and you integrate sensory to the present

02:38:11 and you try and work out what you want to do in the future.

02:38:14 And then you go and make that happen.

02:38:16 So the imagination is this, it’s like, imagine,

02:38:19 asking what imagination is about, asking what surfing is.

02:38:22 You can see, you can surfboard, surfer, wave coming in.

02:38:26 When you’re on that wave and you’re surfing,

02:38:28 that’s where the imagination is.

02:38:31 I think imagination is just accessing things

02:38:34 that aren’t the present moment in the present moment.

02:38:36 So like I’m sitting here and I’m looking at the table

02:38:38 and I can imagine the river and things or whatever it was.

02:38:43 And so it seems to be that it’s like,

02:38:45 it’s our ability to access things that aren’t present.

02:38:50 So conjure up worlds, some of them might be akin

02:38:53 to something that happened to you recently.

02:38:57 Right, but they don’t have to be things

02:38:58 that actually happened in your past.

02:39:00 And I think this gets back to assembly theory.