Video: Discussing 3 Recently Published Papers with Michael Levin (DocDrop)

00:00:00

like thanks for coming on yeah good to see you again thanks for having me yeah such a pleasure absolutely I can't believe we're on round three of our conversations here right now time flies by and uh for people in the audience who

00:00:12

haven't yet caught our first two rounds those will be linked Below in the description um and around one we covered the computational boundary of your self paper the cognitive light cone diagram that folks will be familiar with I'm

00:00:24

sure and then around two we covered Observer dependent Computing your paper with uh Joshua bongard on poly Computing and also the technological approach to mind everywhere today we're going to

00:00:35

cover a couple of uh papers that were recently published just in the last couple of months um your paper on bioelectric networks in the cognitive glue for organisms and then Darwin's and gentle materials

00:00:49

that'll be the second paper we cover and then finally biology Buddhism AI your paper with collaborators that I think will be the the third conversation the third Act of our conversations today so

00:01:01

I'd love to get us kicked off first and those papers will be linked Below in the description as well for folks who want to dive into those they are they can be pretty technical but I recommend you know going through them and um and and discovering them for yourself

00:01:14

there's a lot of great diagrams in there as well to help people kind of graph these Concepts too so if we get started off on the bioelectric networks the cognitive glue enabling evolutionary

00:01:25

scaling from physiology to mind your paper did you ever provide us with a like a brief high level summary of the paper and what's covered here yeah well um I guess the first the first thing to do is to talk about what what cognitive glue

00:01:39

is and why why why such a thing is needed and I basically uh the the I I I use that that term just to kind of draw attention to the following thing we

00:01:50

often people we think about uh collective intelligence is as flocks of birds or or um uh uh and colonies or termites or bees or something like that and they contrast that sharply with

00:02:03

themselves they well I'm not a colony I'm a I'm a unified individual with my own you know um thoughts and goals and all of that but actually if you sort of look inside what you find is that no actually you

00:02:16

are a colony like each of us is a collection of cells um neurons and a whole bunch of other stuff and that and and I want to emphasize this idea that uh that is not

00:02:28

to be taken for granted you you cannot take for granted it's actually kind of a it's a kind of a miracle not not in the sort of religious sense but in the scientific sense of something really profound that needs understanding and explanation

00:02:39

um it's kind of a miracle that that a collection of individual cells with their own agendas and their own ability to pursue various goals in physiological space and and Gene expressions face and so on that there is a way to arrange

00:02:52

those things that give rise to this gives rise to this emergent new self that operates in a different problem space uh and in fact we'll end up often we'll end up denying the fact that it is

00:03:04

made up of of parts with their own agendas I mean that's kind of that's kind of wild that's interesting yeah right and uh and and and has its own goals where some of these goals are often at odds with the goals of the

00:03:16

parts and you know we often do things that aren't particularly good for certain cells in the body and so on right or so are you in certain organs um so so that's the that's it right there that that it's it's very clear

00:03:28

that that there must be a mechanism for doing that for for for transitioning from just a pile of cells to something with its own uh with its

00:03:40

own self and we can sort of Define we can do a little bit of definition there too if you want but um but so so that's the goal right so so so now now even though people don't think about that for much like that the the reality is that

00:03:53

of course the whole field of Neuroscience is predicated on the fact that we know what the cognitive glue is for behavior in the brain it's the it's it's it's electrical signaling and and you know electrical let's say let's say

00:04:05

electrochemical signaling in the in the brain so so that's that so that's the idea and my point in this paper and and in some previous papers is that there's a reason why bioelectrical signaling is

00:04:17

so good at this in the brain it's because it's been it's had lots of practice evolutionarily where this all comes from is by serving as cognitive blue for a morphogenetic agent which is the thing that uh arises when a bunch of

00:04:29

individual cells in an embryonic blastoderm suddenly uh start to cooperate toward a very specific goal they're all going on a journey in this anatomical space if possible configurations of possible shapes that

00:04:42

anything can be they're all committed to to helping each other get to one particular region of that space that corresponds to the Target morphology of that of that species um and so uh and so and so yeah so that's

00:04:55

what this paper is about it's about it's about how bioelectricity serves as that kind of cognitive glue and and then Evolution kind of pivoted it to do the same thing in three-dimensional space for the control of behavior awesome yeah

00:05:06

thank you for that summary and I'd love to hear I mean one of the second question I had here was how did you land on the term cognitive glue like what were there any other tournament was there any other terms of phrases that you were kind of uh deciding between

00:05:20

for that one I mean I don't know I I I use a lot of kind of these kind of terms that I just sort of come up with but for for that one I'm not I'm not sure if there were any any competitive I mean you can you can think about some sort of binding policies you can you can call it

00:05:33

a some sort of an emergence uh a self-emergence mechanism I mean you could you can come up with it but I just thought it was simpler to point out that um it it literally is a kind of cognitive glue because without it the

00:05:47

it's the thing think about what happens during uh during general anesthesia right so you you you walk into the to the uh to the doctor and uh and there you are and you have you have all kinds of thoughts and hopes about what happens

00:05:59

and you say boy I hope the surgery goes well I've got a you know a big uh big thing I got to do a month from now and then and then the the gas comes in and one of the things that happens is the gap Junctions between your your cells

00:06:10

get um get get inhibited so now you're gone for the next however many hours you're not there your cells are still there all the pieces are still there nobody's damaged nobody's dead the cells are all you know the cells are all

00:06:22

functional but you're gone and and one reason you're gone is because that cognitive glue has been temporarily dissolved it's the there really needs to be something that binds all this together um you know and glutes I mean it's kind

00:06:35

of silly because it's not it's not that physical kind of thing but um but this idea that everybody has to be uh kept together and the way that glue does in a particular space it's not just

00:06:48

physical physical proximity the cells stay close to each other it's it's a kind of informational proximity uh and and actually you know people like Julio tononi and others actually study this from an information study that integration from an informational

00:07:01

perspective but that's that's really what's needed is a mechanism to hold together the uh the cognitive system that is then going to make claims about itself being a you know a separate agent from the parts that make it up

00:07:13

um yeah it's interesting um one very specific question I had in the paper and um it may have been in there but maybe I just potentially missed it but there's there's a quote in there if you don't mind me just reading

00:07:26

it off here um it's developmental by electricity as a precursor of brain like processes which reveals not only evolutionary pivots between two different problem spaces but also shows a path to solving

00:07:37

the problem of collective intelligence across scales of organization what are the um the two different problem spaces and perhaps it's just an example being made there but I think there's a mention here of evolutionary pivots can you give

00:07:50

the audience an example of um of what you mean by that sure well so so first let's name some problem spaces here just kind of roughly in order from the beginning of life so so you've got some

00:08:02

metabolic you've got the space of metabolism that's a problem space because you have to figure out how to keep if you're going to persist in the real world you have to figure out how to operate metabolism um and then and then uh there will be some sort of physio logical problem

00:08:15

space where aside from metabolism you have other physiological features that you're trying to outkeep and then at some point you get a transcriptional space where there's actually genes now that can be expressed and so that that space very high dimensional space of the

00:08:28

correct gene expression for whatever you need to be doing at the moment and then when you get to multicellularity there's an anatomical morph of space you know how many eyes are you supposed to have what's the are you supposed to have eyes what's the what's the where's the head

00:08:41

go where does the tail go um those kind of things so that's the space of that's an anatomical space and then after that you get to behavioral space because eventually you develop muscles and nerves and now you can actually move around in

00:08:54

three-dimensional space and so so now you can do what's classically known as behavior all these other things are Behavior too they're just behavior and weird spaces that we don't normally think about and then maybe if you're if

00:09:06

you're some sort of human or something else you might also operate in linguistic space so we actually have a project now looking at navigation of linguist six space as exactly that kind of navigational process interesting yeah

00:09:19

yeah and maintaining you know this idea so so so there's one Central concept to all of this which is navigation this idea that the space is Rich it has structure you as an agent have various kinds of um

00:09:30

uh preferences about which parts of that space are better for you and which parts are worse and therefore you need to navigate it now you have different degrees of Competency of navigating it you might navigate it the way that a bowling ball or a dandelion seed might

00:09:43

which is you have very little internal control if any of or what happens or you might be some sort of something in between you know there's various the seeds with little coarse corkscrews and things that kind of help them do various

00:09:56

simple things you might be some kind of a you know a simple homeostatic agent like a little thermostat that's actually better than than these other things or you might actually be a learning system that can have anticipation and and and

00:10:08

uh associative learning and things like that or you might be really complex and you might have planning and you might be able to really think forward uh pretty well and then you might be really complex and have a gigantic cognitive

00:10:20

light cone where you can sort of Imagine uh lots of complex things far into the future way outside of your current uh scenario so so anywhere along that Continuum there's some you you might have some competencies right and and you

00:10:34

navigate that those those spaces and so what I mean by evolutionary pivots is simply that once you're good at navigating one kind of space it's it's it's relatively I I think this is all you know this is all kind of the

00:10:46

framework that I work on um so so lots of things we don't know yet but I think it's relatively easy for evolution to switch spaces on you because if you're good at navigating a particular kind of space we can swap

00:11:00

some sensors and some effectors and uh and you can use all the competencies you have to navigate some other space so so just for example um uh hybrids right hybrids are when you take a brain and you put it in a robotic

00:11:13

body such as a vehicle and so in that case uh in that case the brain instead of leg muscles might be connected to some some wheels and instead of eyes it might have some photo arrays and uh and if the brain is good at doing these

00:11:26

things it will run them so people have made these things and the other day they have various behaviors and it sounds it sounds kind of crazy and wild except except that except that that is the normal scenario you see your brain doesn't actually interface with with

00:11:39

reality your brain interfaces with your eyes and your muscles and various things uh and and all of this is highly plastic when we we can we've made tadpoles where the eye is on its tail instead of in the

00:11:52

head those animals can see perfectly well immediately they don't they don't require new evolutionary adaptation periods of evolutionary adaptation they can just see in this new configuration and that's because and and the same reason why uh sensory augmentation and

00:12:05

sensory substitution so this has been known since the police 70s probably before that that you can do all kinds that you can give humans all kinds of weird sensors and effectors and they very quickly become part of them so they become

00:12:17

uh for example Baki Rita in the 70s used to do this thing where it's like um you know that toy I don't know what it's called but it's like it's like the square thing with a bunch of a bunch of metal nails and you put your hand on it and it kind of makes the imprint right so so you can imagine an inverse of that

00:12:32

where the little pegs go move in and out so you take that thing and you put each Peg you connect each Peg to a pixel on a camera that you wear on top of your head and then you take that that thing and you and you and you put it up against your belly so that it's poking you uh

00:12:45

based on each pixel that nail is poking you or not right and so and so he did experiments with with people uh with people who are blind who learn to navigate that way because because you can you can you can remap the information that you're getting through

00:12:57

your skin because the plasticity is incredible look at look at the um the rubber hand illusion you can you can see these videos right on online where you've had as a as a as a tetrapod you've had you've you your brain has known how many

00:13:10

hands you have for how many millions of years now and within seven minutes of this visual input of somebody stroking this rubber hand you've now decided that you have three hands you're perfectly willing to abandon that prior and when somebody somebody hammers the rubber

00:13:23

hand people jump up and scream and so the plasticity isn't is incredible and that's why that's why when people get Prosthetics where the where the wrist goes 360 Degrees around when they reach for a cup they'll they'll rotate the way

00:13:35

that your normal wrist never rotates because because they get used to it that's your body now and and so and so all of these things uh this this kind of plasticity so that's why I think Evolution fundamentally makes these

00:13:48

kinds of uh systems that are they kind of figure out what they are on the Flies or like Josh bongard's robots from what 2006 or so when the he made these robots that didn't have a pretty determined model of what they were and where the

00:14:01

effectors were and where the sensors were and they sort of flopped around like like babies they flopped around until eventually they figured out how to move around because they built a model of themselves and and what their

00:14:13

structure was and that has the the awesome um uh side effect that if you rip off one of the legs it'll just go through a similar process and remap to the new and then and then move move in a different

00:14:25

way given what you have now but this is exactly what we see in in biology where uh with a very very wide range of various um accommodations to novel circumstances

00:14:39

can be can be had because because Evolution makes these problem-solving machines that are very good at defining and redefining themselves on the fly so so that's what I mean right these these tricks that that work well

00:14:51

in one world yeah swap some stuff around swap a time scale swap the senses swap the effectors and now you're walking around in some other kind of space I have a feeling that that's what's happening that's my guess wow that's wild I never heard about this this the

00:15:04

2006 Bond guard oh yeah look it up it's great I think I think it's very uh it's it's really foundational yeah I have to you have to do some digging on that for sure um thank you for mentioning and there's yeah so many great wow so much awesome

00:15:17

stuff I want to unpack there um one of which well a few a few things that I think keep coming up um sometimes like a phrase or an idea that um from our discussions or from

00:15:31

listening to you on other podcasts you know you say all intelligence is collective intelligence which I think is really great I think uh the emphasis a lot of people need to hear that the other thing and I think you just touched on it a bit

00:15:43

is that or maybe in our previous discussions you did too we a lot of people think of the brain as having this like exalted status of that the brain can do all these things that other parts of the body can't do

00:15:56

um but what's and it's in this paper as well and other papers too but um sometimes there's memory stored outside the brain you know and some of the uh some of the research that you've done can you give us a sense of uh what

00:16:08

what does the brain do that's actually different than the body cells so um yeah so what does it have that's say other parts of the body can't do yeah

00:16:22

um yeah so so as much as as much as I try to lean on the commonalities between uh brain and other tissues I mean it's pretty obvious there's a reason why we Have Brains Brains give us extra features that we wouldn't have otherwise and so

00:16:35

um just to you know just I mean the the most obvious one is is speed so so neural bioelectricity is way faster than Developmental bioelectricity and I think that was part of an arms race at the beginning when things started moving

00:16:48

around and trying to eat each other and avoid being eaten uh speed became of the essence you know I mean developmental biology I mean yes you want to complete embryogenesis as quickly as you can but it's not under the same speed constraint

00:16:59

I think as actual Behavior you know three-dimensional Behavior so so so there's speed and this idea then then there are the point connections so neurons can be incredibly long and so if you want to make a directed connection from here over there

00:17:12

uh without that neural architecture of having an axon that reaches all the way down I mean they can be meters long in some animals um without that the basic bioelectric system is is is cumbersome because it

00:17:25

basically thinks spread as waves or uh you know they propagate through the Gap junctional milia it's not the same as as uh right so so those are those are kind of the architectural things I mean a lot of the components are conserved so so ion

00:17:39

channels neurotransmitters electrical synapses all that stuff is conserved but it's used in a in a different way both both for Speed and for for direct connections you know and and then and then there's just there are the things

00:17:51

that that we know that are associated with brains that we haven't found anywhere else so for it just as a as a sort of um high-end example is language right so so I've seen no credible claims that uh the other organs in your body

00:18:03

are using this kind of syntactic language structures that that brains use not saying it's impossible who knows we might find some kind of syntax but there isn't any any evidence for that yet that I know of so that that I think

00:18:16

um you know but but other basic stuff is the same pretty you know uh perceptual uh control and predictive coding and all kinds of that that happens in all cells um I wanna I wanna come back real quick because I I do think this is important

00:18:29

to something else that you said a minute ago uh about the collective intelligence and the fact that people need to hear that so so I I wanna I wanna um just talk for a moment about what it is that I think they need to hear about it because

00:18:41

some people and uh and I keep I'm always fishing let's see let's see if this if this helps at all but I'm always fishing for a better way to make this point and I don't know how effective this is um

00:18:52

some people get really uh kind of uh destabilized by by these sorts of ideas because what they hear is I'm not real uh I'm an illusion and there are lots of

00:19:05

scientists that are pushing this this narrative right so so we we are we are to blame for this it's I'm not surprised that people have these ideas but uh you know this idea that well we don't exist you know you're a big pile of of selves

00:19:17

and you're not really here and it's all an illusion and that's a really destabilizing idea it's destabilizing on a personal level it contributes to the loss of meaning it it contributes to uh societal issues

00:19:28

um and so and so I wanna just just say what I think the lesson here is the when when you see science like this the lesson isn't that you are somehow uh

00:19:41

devalued from what you thought you were and that you are uh the the meaning of your life is reduced and that your primary experience of being a coherent being with Choice with with the

00:19:54

responsibility of deciding what do you do next that these things are now out the windows some somehow I think that's the wrong conclusion to draw for many of this the conclusion isn't that the Majesty of of the integrated mind is

00:20:06

reduced the the right conclusion is well two things one is that actually we just learned an amazing thing that matter can do we didn't know that before we thought dumb matter was just kind of dumb uh and

00:20:20

and and and and what we're learning now we're not we're not learning something uh that uh that changes how we view ourselves we're learning something that changes how we view matter and this is this is something that um I think uh Ian mcgilker says as well that that

00:20:34

um we've we've uh underestimated certain kinds of matter um this idea that that no actually and and this is uh I know I know I've used this this uh example before but there's a there's a scene in a lot of Science

00:20:47

Fiction most recently I saw it in in ex mocking the right way the guys well the guy starts cutting his arm because now he's worried that he's a he's a he's an Android right right right and I mean yeah a lot of people feel

00:21:00

that way but if you if you cut your arm or you go get a CT scan and they say whoa you know what you're full of cogs and gears the conclusion from that isn't oh man well I guess I'm not as real as I thought and I'm not you know I guess I

00:21:12

don't get to use my free will now and I yeah that's not the answer the the the the the conclusion should be amazing cogs and gears can give me my my my spiritual meaning like amazing I've just

00:21:24

learned something about cogs and gears great so so that's you know I just want to be clear that I think this kind of uh analysis of what it is that we are how we get here you know the the

00:21:35

self-construction of of the self from cells during embryogenesis and all that that isn't it it doesn't have any negative implications for what you are and what you can do just the opposite it it sort of raises the the the the

00:21:49

remarkable magic of these of these unbinding principles that you say wild you can actually create a being that I know I am I mean that's the part that I think Descartes

00:22:00

had had exactly right uh uh apparently you can get there through through this particular method and and there was a I forget what it is but there's there was also another an old science fiction story where you know these these uh

00:22:14

these aliens kind of land on Earth and and they find out that humans are basically made of meat and they say that is that is the most disgusting like like you're telling me a pile of meat can have these exalted thoughts that we in our silicon you know implementations

00:22:27

have like that's that there's no way there's there's no way a pile of carbonaceous uh agu is going to have these kind of um you know mathematical truths that we perceive with our with our silicon minds and and and and right

00:22:38

and it's all it's all completely um uh arbitrary somebody somebody who doesn't want to find gears under their skin and cogs and things like that why not because they've bought into the fact

00:22:51

that proteins and uh and and RNA does it is that any more you know why why are you any happier with that none of that none of that is is is is intrinsically any better than anything else so I think

00:23:05

it's very important not to get uh not to take the wrong message from from all this kind of stuff and to somehow dissolve your uh your your fundamental worth just because we've seen some of the parts that are under the hood we knew there were going to be Parts under

00:23:17

the hood right in fact in fact I'll go one step further and for the people who write because one thing that some that people sometimes say at this point is you're right you've just done a deconstruction of all materialism and none of it

00:23:30

matters unless we're a soul yeah so for some people some people feel that way that that basically basically right the the the proteins the the just like the gears all all no good so so the thing with that is without without even getting into the you know kind of the

00:23:43

factual nature or just think purely logically fine fine you maybe you're a soul what's the mechanism of the Soul there's going to be one it's going to do something and it's going to have some kind of uh features if it doesn't have any if if it has no parts then it can't

00:23:56

change then then you then you can't learn you can't uh make decisions you can't improve you can't do anything it's gonna have some it's gonna have some kind of structure maybe it's I'll give it to you that that maybe this it's a completely different you know so something with you know Material Science

00:24:09

has never seen before great but it's still going to have some kind of uh descriptive uh laws that that govern how it acts and you know in in where whatever space it lives in and then we're going to be back to the same somebody's going to say ah but it's got

00:24:21

It's got rules that govern Its Behavior that's not what I mean but you know that that's not enough to give me my my you know my my magical feeling of so right so we got to get over this the the the the the the the the non-material kind of

00:24:35

um Way Forward is it doesn't help at all and we have to get over the fact that finding mechanisms under this somehow robs the the larger scale of its meaning that that's just you know sorry that's a long a long diatribe but that's that's

00:24:48

what I wanted to say no no that's great yeah that's that's spot on too it's funny because um I I had on uh Bobby azarian who's a cosmologist and um wrote a great book called The Romance of reality and he

00:25:02

refers this quite a bit more in the um sort of the deterministic lack of free will but the similar kind of vein that um a lot of people feel like as soon as you start to explain this stuff it's

00:25:15

like you're you're explaining away the soul or the mystery or what makes it interesting right but at least it seems like to me that it's just like a never-ending process like there's always something to learn and continue to be

00:25:27

curious about and um it's funny yeah if you update your priors and you're not stuck to um if you're open-minded enough to uh all the

00:25:40

different possibilities that are out there and you get this new information yeah you just keep like you said if you're a robot you just say okay well that's interesting yeah that's interesting and you move on and uh

00:25:51

that's it that's changed how you view the world yeah it's fine I'm I'm the world's most amazing robot I thought I was a meaty robot now I'm a I'm a I'm a you know metallic robot great who cares I I my list my list of things I was

00:26:04

going to do the amazing things I was going to do still nothing's canceled still still right there I'm still gonna go do those things so so that's what I I wish you know I don't know how convincing any of this is to to

00:26:16

um to anybody but uh that that's one thing that I really wish people wouldn't would internalize there's so much there's so much um both personal and social loss of uh you know loss of meaning I think is the best way you know

00:26:29

as vereke says uh around all of this kind of stuff and and that's the last thing I want to I want to contribute to you know yeah it's I mean it's a fast that's a whole fascinating route we could go down too because I feel like

00:26:42

let's just say this like if if that if the uh if that new information would lead you down this kind of nihilistic path and I kind of feel like you're probably going to go down that path anyway you

00:26:54

know yeah it's sort of you're using it as a reason to get stuck in a really a vicious cycle potentially but um yeah that's that's really interesting but actually I think this where I was

00:27:07

going to go to next anyway and it's in this paper and the next paper as well you had this concept called called the pla the pla Play The Hand you're dealt um and you I mean you you're using this

00:27:19

very um how should I say it uh materially you're talking about the cellular collectives that can carry out um different steps with its sort of second order functions it's not

00:27:33

um they're far more let's say malleable uh and far more more open-ended uh kind of strategies for accomplishing their goals can you talk to us about that and what that uh yeah it's I think you

00:27:46

the acronym is PhD so play the Henry Dell yeah um well let's let's just uh let's just start with a with a specific example um so you you you take a uh a a salamander

00:27:59

egg and there's a there's a certain tricks you can do to uh increase the number of the chromosome count within the number the number of uh the the number of copies of the genome that are in there and so when you do that let's

00:28:13

say so let's say we increase instead of uh instead of uh 2N you can make 4 in 5 and 6n and so on so as you do that the cells the embryonic cells get bigger and bigger the salamander stays the same size

00:28:25

if you take a cross section through a kidney tubule which normally is made of I don't know if you can picture this but like like you know six to six to eight cells that work together to form this kind of like long long tube they get

00:28:38

bigger and bigger but the animal stays the same size and the tubule stays the same size so what this means is that fewer and fewer of these larger cells participate in each tubule right and so so far so we have two amazing

00:28:51

things so far number one is you've got the wrong number of chromosomes fine you're still a good salamander amazing number two your cells are the wrong size no problem we we we manage the cell

00:29:03

number to make up for this different in cell size it's still you're still a good salamander number three if you make the cells absolutely gigantic and these are I think six or eight and newts I don't remember exactly uh what happens is

00:29:17

there's not room for even two cells to be there so what happens is one single cell wraps around itself and leaves a hole in the middle to give you that same tubule okay aside from the whole issue so so so what it is is it's it's using

00:29:31

different molecular mechanisms in the first case it was self-to-sell communication in the next case it's cytoskeletal bending uh you're using different molecular mechanisms in the service of an anatomical large-scale

00:29:42

goal that's that's a kind of top-down causation super interesting but for our purpose more importantly look at your your job as a salamander you come into this world you can't tell how many chromosomes you're going to have you

00:29:55

don't know what your cell size is you don't know how many cells you're going to have because people have done that too you can take away cells you can add cells you don't know any of that stuff you need to be able to your goal is to be able to make a good salamander no matter what you start with so that's so

00:30:08

that's within the limits obviously I mean all of these things are not you know infinitely stretchable right so so that's one example here's here's another example plenary of flatworms uh the species that we work with reproduces by tearing themselves in half

00:30:21

and regenerating that's how they reproduce that means that they have somatic inheritance any mutation that doesn't kill the stem cell is going to be proliferated into the body and expand

00:30:32

it in the body in the Next Generation so if you look at these worms they're mixoploid every cell has a different number of chromosomes The genome is a total mess because because they just accumulate the stuff 100 million years have been accumulating all this junk and

00:30:44

yet that's the species with perfect regeneration High cancer resistance no Aging in the in the asexual form uh you know in incredible right and so and

00:30:57

so the same story there is you what you have to do is you have to have an algorithm that builds a correct planarian despite errors in the hardware and that's what I think one so so this is so

00:31:11

so there's my there's my uh play the Andrew Delta concept is that coming into this world as a new creature there's precious few things you can depend on for most I mean there are hardwired things like C elegans you know nematodes

00:31:22

and and maybe maybe some other species but but I think the and there's a whole Spectrum so so I would I would sort of imagine that something like nematodes where every nematode every UC Elegance has exactly the same number of cells and they all have the same lineage and so

00:31:35

that's a very cookie cutter organism and that's here planaria are super plastic and they're out there and salamanders is somewhere here and humans are somewhere your mammals are somewhere here we're all we all have different degrees of that competency

00:31:48

um but that's that's the idea you don't most of us don't come into the world being able to uh expect very specific things and then and then just just crashing and burning when those assumptions aren't met this is why this

00:32:00

is why we can make tadpoles with eyes on their backs that can see and this is why um slippers uh uh goat when uh you know that two-legged uh yeah

00:32:12

yeah yeah yeah yeah learn learn to walk upright and uh and and they found they found that a lot of the changes that uh that come up come along with uh bipedal locomotion were already made because because of the

00:32:28

because the the incredible plasticity of the of the organism um and it's why you know skin cells taken off of frog embryos become xenobots and and why and instead of just like collapsing and dying and why all of

00:32:41

these things that that's that's play the hand you're dealt that's the idea that we have to uh there may have been life at one point that was much more um cookie cutter but none of that survived you know nowadays if you're going to

00:32:53

survive in you know nowadays you're you're the kind of life that is able to do this right and can you tell us about um one thing I was thinking about too with the play the hinder doubt sort of like whatever materials you have there that's

00:33:07

what you're going to use is and with bioelectricity of course like the electromagnetic the electromagnetic spectrum that's sort of what's available to these cells right and I'm thinking and perhaps this is just because it's at

00:33:20

the limits of our science have you found um either in your own research or research of uh the contemporaries like evidence of organisms organisms making use of quantum fields like are all the

00:33:32

different fields that are available you know what I mean electricity I understand why um you know bioelectricity is so great I mean we use a lot of the same you know the uh ion like the ion

00:33:44

channels like logic gates we use that stuff in our in our technology you know a similar kind of processes but anything in the quantum realm that you've seen um so so we don't study that specifically so I don't have any data

00:33:57

about it yeah uh there are certainly people that study Quantum biology there are some uh there's some uh great people who uh you know um uh Clarissa yellow and my former postdoc niroshima Morgan uh that you know they're they're into

00:34:10

this kind of stuff and uh I I and as are other people and I'm sure I'm sure there are interesting things that we don't we don't particularly work on it I I my gut feeling is that not being an expert in this but just for

00:34:23

whatever it's worth my gut feeling uh it's gonna be the sort of thing that if it exists it's not going to be some weird special adaptation where you know all look birds are using like you know Quantum spin to navigate the magnetic

00:34:36

field and that's it it's not going to be like that I think I think I mean that may also be but I don't mean that I I I think I think what we're gonna find is that it's everywhere like it's it's a basic fundamental if it's there if

00:34:48

quantum biology is there which I I suspect it probably is it's going to be it's going to be used for for everything we're going to find out all the things that we take for granted now um that are just you know we kind of assume there's a classical explanation

00:35:00

for her but we don't know what it is I have a feeling we're going to find out a lot of this stuff is is at bottom uh exploiting some of those properties that's just that that's totally a guess on my partner yeah any data to you know

00:35:12

support any of that um but that's but that's my guess yeah yeah that's interesting I would see uh very little I know about the quantum like quantum mechanics is that I would I assume these these structures even

00:35:25

cellular they're tied to us but they're they're too large like you know it breaks down the quantum effects are just what happened at way smaller lens but like you like you just pointed out like I think that there's

00:35:38

some research I've done a little bit of digging but I have to do more um about Quantum biology so yeah look at look at also um look at also look at the work of of Chris fields and and folks that uh that he works with because there

00:35:51

are aspects of what's important about that field that has nothing to do with being small so there are there are really important aspects of you know in terms of what's an observer and reference for Observer reference frames and sure these kinds of things that

00:36:04

apply across scales and so Chris has some some beautiful papers um about that so it's not just for tiny particles and things like that gotcha that's good to know too and uh before we move into Darwin's and gentle materials

00:36:17

I think it's still applicable here can you tell us about I think there was one podcast you did where you talked about um trophic memory in in Deer and antler structures and actually the scientist that was studying it actually sent them

00:36:28

to you and you have a bunch of these can you just tell us I made a little anecdote yeah uh well okay so so first what is trophic memory so so there's certain species of deer that uh every

00:36:41

year they they shed their antlers and then they grow them back antlers are real bone they're not like horns uh they're they're real bone with velvet innervation and all that so what so so there was this team uh last name bubenic

00:36:53

and it was two uh two two two folks uh a father and son and they live in Canada and they did experiments for I'm gonna say well over 30 years maybe 40 years together or something something like that where what they found is that if

00:37:06

you you so you got this deer and you take a you take a knife and you and you etch a little a little um cut into somewhere on this branched structure and that year it kind of heals with a

00:37:17

little callus the bone heals and that's that and then the whole thing drops off and then next year when they regrow their antlers it it grows with an ectopic Branch point at the location where the damage was

00:37:30

last year now I read this uh and and so so they got they got these papers from from 60s and 70s and and Beyond it's an amazing data set no you know nobody's ever going to get a data set like this I mean if

00:37:42

he's got you know uh okay okay a herd of deer that they can watch for 40 years you know it's a in in Modern Biology careers that's not exactly you know conducive to getting uh getting a good uh the associate um assistant professor

00:37:55

position um but uh but but but but they've got this incredible data set and and when I first read about this I thought it was amazing because uh if you try to think about the current

00:38:06

you know what what what passes for an explanation uh for biological phenomena nowadays so if you if you find this you know you look at a typical cell paper figure seven is going to be a molecular pathway right so some Arrow diagrams and

00:38:20

this thing binds to that thing and then they go over here and they bind to this other thing just just try to come up with a with a model like that for what's going on here so there's a there's a there's a large scale structure it gets in it gets a damage input at a

00:38:32

particular location the whole thing falls off the cells at the scalp which are you know tens of centimeters away remember the three-dimensional position where the damage was last year and then

00:38:45

when the bone cells start growing they they revise the genetically encoded rules to say oh by the way when you get to this point to make an extra thing to the right like why what possible explanation using

00:38:57

conventional um uh conceptual tools that we have could you come up with it it just completely fails and so so this is one of the things I tell my students is that you know your developmental biology textbook is full of things that are

00:39:09

readily explained by the conceptual uh tools that we have now that's what's in the textbook other success cases so what's what's fun is to look for the uh the blank space in between what are all the things that are not in there right and they're not in there because we

00:39:22

don't have a clue as to how it works so this is the kind of thing I mean I love that stuff so so I'm always I'm thinking about these kind of things so so anyway so so at one point so we wrote Daniel Lobo and I wrote a wrote a paper about this and explaining like what the

00:39:34

implications of that are for uh um the inverse problem which plagues regenerative medicine and so on and uh and bubenic uh emailed me and he said um you know he said uh you're one of the few people that that cites this stuff

00:39:48

nowadays I need to like basically clean out the the house uh the garage uh would you would you like these antlers I was like to hell yes of course I would and that's such a unique I mean it's amazing work uh it's such a such a unique um you

00:40:01

know data set and so so we we received in my lab 13 I think it was 13 13 large boxes of these antlers and I I I I I sent them all uh to um the uh the

00:40:15

veteran a Tufts veterinary school they have a a cat scan machine where they where they CT scanned you know horses and things like that and so and so they CT scanned all these things so I have somewhere somewhere I've still got the I've still got the antlers they're in

00:40:26

boxes in the closet I have a couple of them on the wall actually in the lab as you walk into Labs a couple on the wall but um yeah yeah but uh but but I still got these boxes and it was incredible because every set of antlers is labeled

00:40:38

with an with the name of the deer and the year so it'll be like you know Lenny 1987 you know Lenny 1988 and because you have to do these longitudinal like you have to know which deer it was and he kept meticulous records and he and and

00:40:50

so we have all of these uh you know all all of these all these antlers and uh you know it's a it's a it's a it's a very unfortunate model system because if you want to do experiments I mean who's going to have deer year and wait years

00:41:04

for an outcome like it's crazy but but the closest thing to that actually is our two-headed planarian because it's a very similar kind of thing it's a it's a physiological stimulus that gets somehow

00:41:15

catalyzed into multi-generational repair processes because once you make them two-headed they you can keep cutting them they stay two-headed forever yeah yeah right and and luckily planaria are much more tractable than deer so so I

00:41:29

you know I suspect those those things are are highly related but that's the that's my That's My Dear Antler story it's awesome that's so cool that must be what a what a fixture to have in the uh in the lab too it's right right in the front as you walk in a

00:41:42

great story too awesome thanks for sharing that um so I'd love to turn towards and there's gonna be a lot of overlap between this paper and what we just discussed and I'm sure we'll talk more about by electricity as well but uh the second paper Darwin's and

00:41:55

gentle materials evolutionary implications a multi-scale competency developmental biology and both of these papers the first two they really just dropped in the last couple months few months right April 2023 so these are hot

00:42:08

off the presses uh for folks I I haven't really heard you talk about these in other podcasts so um it's great that we're getting a chance today to do that the and before we jump right into the paper is this an illusion at all to the

00:42:20

like his dark materials book series [Music] okay just wondering I mean that's a little bit of a flyer because it's Darwin's materials I was like is there sort of uh yeah it doesn't sound like

00:42:34

that doesn't it uh no yeah yeah yeah okay cool yeah just just to the side but um would you mind I mean just like you did for the first paper because you provide like a simple overview and then I have a bunch of questions to ask you

00:42:46

sure yeah um Let's uh let's let's start from the from this idea that um and this I tried to formalize this in my tame paper from a few years ago this idea of a of a

00:43:00

spectrum of you can call it many things as a spectrum of persuadability that's what I call it in the paper or a spectrum of agency just the idea that you can put any system on this on this continuous spectrum that that tries to

00:43:13

capture how much autonomy the thing has from from an engineering perspective how much can I how much can I expect how much problem solving uh uh chops does this thing have when I'm not for when I'm not around to force it so uh you

00:43:26

know and you can go from from from Legos to um uh kind of you know to uh thermostats to animals that learn to humans and then everything in between is there right okay so now so now we can ask the

00:43:39

following question um as as an engineer and so this is another paper I wrote with Jamie Davies a few years ago well maybe a year ago uh talking about engineering with agential materials because engineering is very different

00:43:52

based on where along that Spectrum your parts are if you are engineering with Legos everything is on you the only thing the Lego is going to do is keep its shape that's all it knows how to do so everything else is on you everything

00:44:04

that that you need to happen you have to somehow make sure it happens and this is how molecular molecular synthetic biology Works you're going to put in circuits that do specific things it's on you to to to to implement every part of

00:44:16

the functionality that you want um if if you uh and and so and so humans have been engineering with bricks and and wood and metal for thousands of years and uh and and use a certain set

00:44:30

of techniques to do that what are those techniques well they don't involve uh psychological uh tools they involve very kind of low level put everything where it goes and glue it down and attach this

00:44:42

thing to that thing that's that's your toolbox well uh if you're building your autonomous vehicles or self-guided missiles or or houses or whatever you've got some other stuff which are for

00:44:53

example thermostats so if you've got a thermostat it's interesting because you don't even necessarily need to know how the whole thing works what you need to know is where is the set point and how do I change it right and what are the inputs and what are the outputs where

00:45:06

does the thermometer go where does the connection to the to the heat you know to the cooling and heating go and what you know is that it's going to do certain things when you're out there to micromanage it do you have to leave rules for what to do at every single temperature level you don't it's going

00:45:19

to do that on its own so so you've got some other tools and these are the tools of cybernetic so for more complex systems you know you'll have control theory and and all these kind of tools to deal with something that has a simple

00:45:30

level of agency it's got very basic uh you know kind of primitive goals that it tries to set it tries to and and if you don't understand anything about cybernetics you are not going to get the most out of these components right if

00:45:44

you don't understand what what homeostasis is what gold directed Loops are you're not really going to be very good at at using those in your engineering okay and then you move forward and you say okay I may uh I'm a

00:45:57

proprietor of a of a circus for uh of a rat circus and I want these rats to do little little tricks right I could I want them I want them to in the in the Jamie's paper I think we talked about uh uh building a tower out of dogs I could

00:46:11

try the uh the traditional route of stacking them on top of each other but that isn't going to work they're gonna they're gonna crawl off uh they're not gonna just stay where I put them so that doesn't work at all I have to use a completely different set

00:46:23

of Tricks the people who work with wood and metal don't have to use I have to train my material I have to uh now now on the one hand it's a bit of a pain in the butt because your material has its own agenda and you have to have tools to manage it but here's the beauty of it

00:46:36

once you've trained them to keep a a you know a little Tower if you knock it over guess what it does they get right back up on their own you don't have to be there to rebuild it isn't that amazing and and and so you've gained something very interesting by switching the the

00:46:49

bag of tools you bring to the problem you've gained something interesting and you were only able to do that because you recognize the agency of the material you wouldn't do that and right you wouldn't do that if um if you didn't know if you thought that these things

00:47:00

are dumb like like bricks and Legos so so it's very important and then and then you see this in uh all the way up if if you're a hacker you might be hacking the computer but you might be doing what they call social engineering right you

00:47:14

might find out that it's much easier to just trick somebody giving you their password than to spend all day you know brute forcing the the hash table or whatever so um it's it's there's this there's this has to be this impedance match between

00:47:27

the tools you bring to the problem and and I and the successful engineer is one that recognizes what are what are the right what's the right level of agency in my material and so and so that leads to the question for regenerative

00:47:38

medicine we have cells and tissues and you ask so what are the tools that I'm going to use there and the Assumption up until now the Assumption has been well they're like the bricks you have to micromanage all of it yeah that's been

00:47:52

that's been the standard Assumption of of uh of the of the Paradigm and and but but it's very much an open question are they more like the the Legos or are they more like the rats or are they more like something in between or are they more

00:48:05

like the thermostat or where are they and and regenerative medicine is going to be cracked by the people who pick the right who pick the right level it's not going to be crap cracked by assuming the wrong level nor you know down so it's not gonna it's it's not gonna happen if

00:48:18

you assume these things are uh our our low agency machines it's also not going to happen if you assume uh that they're magical and explicable um uh you know things that uh that aren't

00:48:29

going to obey any kind of uh rational rules that's not going to work either somewhere in the middle which is what of course what my lab tries to do is to pick the right set of tools from from cybernetics from Behavioral Science to to uh you know take advantage okay so

00:48:42

having said all of that um here's the thing with the paper so that's so that I what I just took you through is how human Engineers view the spectrum of agential materials right and so now it comes time to so so

00:48:55

we wrote so Jamie and I wrote that and then I said okay so now what does evolution do evolution is also an engineer so the question is does evolution assume uh and work at the assume that

00:49:08

these things are like Legos and work at the lowest level which means search the really difficult and really uh kind of uh rugged space of of molecular properties or with the what evolution

00:49:20

take advantage of the competency of the material because the thing about evolution is evolution doesn't work with Legos Evolution work with works with cells and cells and tissues used to be independent organisms they don't come as

00:49:32

blank slates that are dumb and have to be micromanaged they come with behavioral competencies with preferences with uh various kinds of agendas and so what that paper is is an exploration of

00:49:44

if we take that seriously the fact that evolution of course evolution is very opportunistic it makes use of everything it can we know that what what can we conclude if we take seriously the idea that Evolution will not have missed the fact that it's

00:49:57

dealing with a very powerful agential material if the cells already know how to do things and so and so what I do in that paper is run down all of the implications what does it mean for evolution that it isn't working with Legos it's working of course of course

00:50:09

people have used evolutionary computation like genetic algorithms and things like that with materials that really are dumb so the typical the typical um evolutionary algorithm is is done over passive passive data and it shows it you

00:50:23

know it shows you improvements but the material is very passive and here I'm saying biological evolution isn't like that so what does that mean what will that do to Evolution and that's kind of the flip side right A lot of people study

00:50:34

um how intelligence how Evolution gives rise to intelligence of different types I sort of reverse that and of course both are both are happening simultaneously but I looked at the other side of things which is how does

00:50:46

intelligence impact the actual Evolution the intelligence of the of the substrate sure oh that's interesting it's so thought-provoking the okay tell you something I an Impulse I had while reading the paper even though

00:51:01

it's not stated explicitly and even something you just said would you how do you view even The evolutionary process I mean it almost sounds like evolution is an agent in and of itself do you view it that way or how

00:51:13

do you how do you look at it yeah this is this is an interesting point uh that's a that's a paper that um uh who is on my list uh to write it'll probably be next year at this point but but this

00:51:26

idea of evolution itself as an agent um I want to be I want to be very careful here because a lot of people still have this uh kind of ancient view that there are two kinds of things in the world there are dumb

00:51:39

material things like like like uh you know so the the machines that the quote-unquote machines and so on and then there are the mindful things like humans and angels and and God and whatever else right and so and so when I

00:51:53

say actually I do think that there's a that there's a that there's a way to there's a lens on Evolution which does use it as a which does see it as an agent what I'm not saying okay so so

00:52:05

we're super clear not saying that Evolution uh has a a high level uh purpose the way that the you know a human level purpose or a Beyond human level purpose not saying any of that

00:52:17

what I am saying is that there's a very rich spectrum of agency from from very low I'm not sure there's a zero but but certainly from from very low all the way to human and Beyond and I don't think we

00:52:29

can blindly assume that that the level of agency for The evolutionary process is down at the low end it it might be non-zero and uh and it might be important to understand what it is but

00:52:42

just again really clear not saying that there's anybody that you know there's any human or above level intelligence out there picking picking you know where the the lineages go um I think that uh I think I think Carl

00:52:56

uh for some probably said this well before me this idea that you can use that framework uh to you can so so so imagine um there's kind of two ways to think about this one way to think about this

00:53:09

is that imagine a lineage you know I don't know 50 million years of alligators or something just you know imagine imagine some kind of some kind of lineage and uh you can imagine that whole thing as an

00:53:22

agent it's a very long-lived agent but uh we're just bad at noticing agency at different time scales it's a it's a spatially a huge agent but we're also too fixated on agents that are roughly the size of like us right medium-sized

00:53:35

objects so if you forget that and assume that agents can be whatever um what is happening there what what's happening there is a continuously generates hypotheses about the environment those hypotheses are cashed

00:53:47

out as Offspring with different features some of those hypotheses are proven false some of those hypotheses are supported those supported hypotheses go on and and shape the cognitive system of

00:54:00

the collective to form new hypotheses that might be even better you could more correctly describe uh you know reality and what's interesting is that um much like uh kind of consistent with

00:54:12

what I said before what these hypotheses are are not flat kind of um first order statements about the world meaning a a a um a hardwired solution and uh you know

00:54:24

it is what it is they're actually these hypotheses are actually problem solving um strategies they're heuristics they're like they're they're like um what comes

00:54:36

out of evolution is is not here's how you be a salamator and that's it it's a set of it's a set of policies that that cellular collectives can can operate depending on what's going on they're context sensitive so so these it's like

00:54:50

it's like instead of generating guesses about the world what you're generating are um policies or generating navigational heuristics right and so and so that's what that agent is doing that's pretty

00:55:02

good that's not that's not super low agency to be able to do that that's something that doesn't mean you have self-reflective you know you know I'd love to uh evolve Some Humans because that you know that would be just like that's not what I mean but but it isn't

00:55:14

you know it's it's being able to generate hypotheses uh and and get them falsified and have this like so so Carl You Know Carl Kristen has some great thoughts on this um Richard Watson uh has some great thoughts on this but but

00:55:27

yeah um I mean there's another way maybe to think about it which is that the whole evolutionary never mind the lineage but the whole evolutionary process itself and that that gets hard I don't have too much to say about it right now but I'm

00:55:38

working on it this this notion of um what actually can be an agent you know processes as agents that's a whole other kind of kettle of fish so so I think you know well by next year maybe maybe if I get anywhere with it maybe we

00:55:52

talk about it next year yeah yeah that'd be interesting I'm letting them definitely be on the lookout for um for that for that paper and I think yeah you touched on so much yeah so much already about what what's in that what's in this paper the

00:56:05

difference between first order and second order um say Goals or or say a sermon on flexibility that uh that is built in to the uh to the you know first

00:56:17

initial levels there um what else what do you think is I guess if you had to say for people to take one thing away from this paper say lay audience what do you think would

00:56:31

be um something that this is hard I want to get to a little bit of this I love your work it's fantastic but uh when I want to uh perhaps push some people I interview a

00:56:44

little bit more on is like okay this is amazing stuff how does someone you know an individual person is there stuff that we can take away here that are there like truths that you know maybe as part

00:56:56

of our day-to-day lives that we can uh sort of integrate yeah ideas like from this or anything from your own life perhaps examples of things that this is illuminated for yourself yeah well um

00:57:09

let's put it this way you know I I don't know if if any of anything from this paper will uh you know revolutionize uh the day-to-day like mundane life but but but but a lot of people but but I will

00:57:21

I'll let's try to pull in that direction um a lot of people uh think about Evolution and one of the things that always bugs them and especially you know a lot of Engineers

00:57:34

too is this standard story of uh well we make random changes and then we pick the good ones right I I don't know about you but when I when I first learned about this I had been I had been building

00:57:48

um you know electronics and things like that as a kid for for some years and then I learned this Theory and I was it it sounded laughable it sounded like you're telling me that that I'm gonna make random changes in this thing I I I

00:58:00

huff and puff for for many hours and then the person who made these transistors and everything else will put in even more work than that like we're all busting our buds on this stuff and and you're gonna make random changes and you think eventually things will get

00:58:12

better like that's you know if you've ever built anything a written code that sounds crazy and and then okay you know so so you learn some things about modularity and um and evolvability and and some things

00:58:23

that but um a lot of people are still left with this with this idea that okay in theory it might work but we kind of know that genetic algorithms you know they did great for a while but they sort

00:58:36

of Peter out there's some Liberty and there's kind of limitations to what they can do um and in particular but part of what makes it what makes it a little um a little challenging is what uh what Andreas Wagner which I think I think his

00:58:49

work is is amazing and important um what he calls the problem of the arrival of the fittest which is that if if you're if the best solution is hiding somewhere in your population I suppose I'll give

00:59:02

you that I'll give it to you that that eventually will sort of find it and let it expand but who guaranteed it was there in the first place how do you know that the right solution is ever going to be there depending on what your problem speaking so anyway so so so a lot of

00:59:15

people are still left with this you know this dissatisfaction about how this just how do we know that there's been enough time for this kind of process to give us the amazing things we see in the biological world um and I'm not talking about that I mean

00:59:28

so so there are some people that will never buy the story because they are fundamentally like uh committed to another kind of story I'm not talking about them I'm talking about um people with a scientific uh kind of world view that want to understand in a

00:59:40

naturalistic way what's going on but this but but the standard story doesn't seem like uh it's the whole story right and there's there's a lot of um very smart people who are sort of uh thinking along those lines so so what I would

00:59:53

what I would point out my my kind of contribution to that is is this part of what makes part of what makes that process so magical is not just the process itself it's the fact that you're working on a material

01:00:06

that's smart you're working with an agential material that's part of where the power comes from if the whole thing seems tough to you add to your you know sort of mental picture the fact that

01:00:19

it's not really searching the incredibly difficult and large space of all the possible things that could happen what it's searching is the space of behavior shaping signals by which cells tell

01:00:30

other cells what to do and that's a much easier space to search right if you're dealing with you know if if you run if you run that that rat circus you could try to come up with the

01:00:43

um kind of uh uh optogenetic strategy to control every neuron to get the rats to sort of do whatever they're going to do that's really hard we'll be here till the sun burns out before we get all the you know

01:00:55

before we can micromanage it but you don't have to do that because you can train the rats and that's a much easier I mean it's still a bit of search because you still have to figure out what what's the reward what's the punishment what what are they capable of do they do a solo do they do Place

01:01:07

conditioning do they do associative learning like what what do they so there's still some some searching involved but it's a much easier problem than uh than than going Bottom up and so and so this is what I want people to take away from this is that part of what

01:01:19

makes evolution so magical is that it's working with an agential material that has tons of competencies it's it's If evolution is uh playing with Hardware that is so far in

01:01:33

capability and I don't mean with the fact that it goes down to the Nano level and I don't mean that it's um you know in fact in fact it's way noisier and more brittle and whatever than than all the things we we try to build despite all that it is it is so

01:01:45

much more powerful than anything we've ever made because it is not a single specific thing it's a um it's a learning machine so to speak yeah and

01:01:56

um and that that gives Evolution that puts Evolution on steroids that's that's what I think is a huge motivating factor and I think until we understand that we are not going to have I mean that's that's to me that's if I can dare to say

01:02:10

this the thing that I think is missing from the standard evolutionary synthesis is this it's the appreciation of the intelligence of the substrate it treats the standard story treats the substrate as a bunch of dumb Lego blocks and everything changes when the material has

01:02:23

agendas everything changes and and evolution changes massively so I you know I don't know if that's if that counts as everyday life for people I suppose uh you know um some some people think about that stuff every day so no

01:02:35

no it's it's good I I find it useful because it's a frame it's a perspective um on on how you look at the problem I mean this is something we talked about in round two I think uh like poly Computing

01:02:48

and the role of the Absurd the same thing can be computed but depending on how you look at it you actually get you actually are extracting that different value different utility um from the same exact thing so yeah

01:03:01

that's um that's wonderful and actually I think this will um just looking at the time a little bit here I do want to cover a little bit about the uh the biology Buddhism and AI paper

01:03:13

um it goes care as a driver of the exactly yeah care of the driver of intelligence which you worked on with a few collaborators Dr Witkowski salmanova and Dwayne and

01:03:27

and this I think actually yeah kind of does lead into this uh fairly well because in the in general materials paper you mentioned the idea of like beginner mind and you have a quote from Suzuki about

01:03:40

the in The Beginner's mind there's many possibilities and experts mine there are a few so you kind of want to have a frame of like always being having a frame of always being a beginner and being like open and curious to things

01:03:53

all the time is a lot better than uh this is a straw manager but it's a lot better than being an expert and being like I know everything I've I've mastered there's nothing else left to go right there's always potential for for

01:04:05

more learning and more mastery so for the biology Buddhism and AI paper would you mind uh providing us again you know just a brief overview of the overall paper and also how how you got

01:04:18

um how you got interested in this topic to begin with and how you got um into this group of folks who are who are studying this yeah um well how I got interested in it I've

01:04:30

been interested in these kinds of things for a really long time both from the perspective of uh kind of Eastern thought about philosophy of mind and and things like that and more broadly questions of uh of of uh concern and

01:04:44

compassion and and things like that um how I how how we met up um to tell you the truth I don't remember who for who reached out first it might have been it might have been Thomas doctor that emailed me I mean Olaf Olaf and I have

01:04:57

known each other he he's uh he's a um a great contributor to the artificial life community and so I've known him and his work for a long time um Bill and Eliza and Thomas I met afterwards

01:05:10

um yeah I don't recall I don't recall who made the first step but but anyway we've been we've been talking about the stuff and thinking about it for a long time and this is actually a second paper that just it either just got accepted or we just returned it I don't remember exactly where it is but there's a second

01:05:22

paper following up on all of this stuff um you can find the reaper the pre-print is is on the website but um you know this idea of uh specifically this idea of care and what do agents care about

01:05:36

and it uh for me it has lots of important implications because I I try to understand Collective intelligences and so if you're going to have a collective intelligence what is it going to uh care about um we don't we don't have a good science

01:05:49

of that and then uh there's the notion of uh embedding embedding care in in artifacts that we make so robotics AIS what are they going to care about um it's you know it's a funny uh kind of

01:06:02

a funny story when um my kid was my youngest was uh I want to see it was three or four and he said uh you know we used to we used to build stuff together all the time and you know we we did all kinds of engineering

01:06:14

things and and one day he said to me um uh let's make a cat and I said well like a robotic you know you want to make a robotic cat and I said well let's let's make a list of what are the design specs here like what does this thing need to do and he says uh well it needs to move

01:06:28

around um yeah maybe like that that may be doable and it needs to you know make meowing noises yeah that's pretty much doable and it needs to there's something else that it needed to do and I said yeah probably we could do that and then he says and it needs to care

01:06:40

and I said you mean what do you mean you mean you mean it needs to like walk over to you and and you know and and let you pet it he goes no no not not act as if it cared I wanted to actually care I was like all right well that's it you've just broke the whole project because we

01:06:54

don't we don't have a clue as to how that is going to happen and and that is you know that figuring out how that uh ties into the whole um uh you know the

01:07:06

cognitive lycone story that I've been telling in the and the kind of the spectrum of persuadability it's like intelligence is one thing problem solving is is one thing but where does the care come from right

01:07:20

and what do we mean by that and and a lot of people say things like I I care about stuff that's just the machine where you know they're usually pointing at some AI thing or some you know robot or something and they say well that's just the machine machines can't care

01:07:33

they're like well you used to be a single cell and paramecia do paramecia care because because now you got a problem if you say that the paramecia the care then well guess what's inside a

01:07:47

paramecium a bunch of molecular cogs and wheels so so you know it's uh you gotta you gotta so so maybe machines and and certainly molecular biologists see single cells as a kind of machine and

01:07:58

that that that analogy has done pretty pretty well for us uh on the other hand if you say no no the paramecium doesn't care it's just a bunch of chemical reactions I mean we can sort of see what's going on in there's a bunch of

01:08:10

chemistry that that doesn't care I care like well you used to be a single cell so why don't you tell me where that when the care got beamed like like what what you know what what stage of embryogenesis does the care get beamed down right that's a problem too so so

01:08:22

you got this you got this real issue with with people who think in binary categories they get trapped in the pseudo problem that's I think I'm unsolvable yeah so um so so anyway so so we were really

01:08:34

interested and then and then of course from their perspective they're interested in uh the questions of you know there's a there's a there's a Buddhist story about care and and compassion and those kinds of things and so I was interested to see how how

01:08:46

compatible those things are can we use some tools from that thought I mean there's a whole there's a whole other thing which is uh this this this notion of the impermanence of the self and it goes I

01:08:58

mean there are obviously all different kinds of uh opinions on this all the way from there is no such thing it's a total illusion right that's that's one set of views and

01:09:09

then on the other is the kind of like the the sort of um the man on the street version which is uh well you know I've got this permanent itself and it's this like thing and then it's you know it exists and then and then what I'm interested in

01:09:22

is kind of a the space in between which I think is more accurate so and and by the way I'm no Buddha scholar um I don't you know I don't pretend to like uh know who thinks what in that in that area I'm just I'm trying to keep up with uh with

01:09:35

with Thomas and uh and Elizabeth and so on but um oh yeah there's a there's an intermediate version which is that it's not that you don't exist you you do in the same sense as everything else exists

01:09:48

which is as a useful metaphor and in fact you you are the most useful metaphor for all of all because because you might do away with you might somehow do away with metaphors of of um uh

01:10:01

talking about uh you know I don't know what what what what social structures are what grocery stores are maybe you don't maybe you know you you you now see um like like Eddington said that that a table is mostly empty space and so maybe

01:10:14

you've internalized physics enough to to know that that even the table is in the great metaphor that it's mostly you know fields and whatnot like all of that is fine but but but but there is but there

01:10:26

but there is this this this this metaphor of a um uh of of a self that can do things because it's on you to do stuff or not do it you have to make those decisions that's a pretty that's a pretty useful

01:10:38

that's a pretty useful metaphor so so I don't think it's not real I think it is real in the same sense that everything else is real which is a useful construction um also I think what's what's useful

01:10:49

about it is that it is continuously self-constructed and here's what I mean by that uh at any given moment so right now you don't really have access to your past the only thing you have access to

01:11:03

are the engrams left in your brain and body by your past experiences that's all you have access to and from that you reconstruct a story of your past so including you know the school you went to and various other things you're you're building that right now at every

01:11:15

moment it's a little bit like um you know uh anterior grade Amnesia patients who can't form new memories and so they use these scratch pads at least some of them where you write down the first thing you write down is that I have

01:11:28

anterior grade Amnesia and then some stuff that happened and then at the end that says oh and don't forget write this note again tomorrow and and right and that's your like most of us have the same thing it's just internalized but

01:11:40

but now they're using the stigmergically this outside tool because there's you know some problem so so we are all really in that in that state it's just we're using an internal scratch Pad you know you right now you don't have any

01:11:52

access to what happened years ago you just have the memories and and they're actively rebuilt and we all know you know our ability to rebuild accurately is is crap basically right and and and these things these things did you know

01:12:04

morph and and change and whatnot so the story of ourselves changes all the time I actually um I did a uh I I gave I gave a talk the other day at this UCLA Symposium and I and I talked about uh planaria and and this

01:12:18

idea you know in planaria if you teach them something and then they you chop off their heads and the tail sits there and then eventually they regrow a new head and they regenerate their memories right and so that means okay the memory is stored somewhere who knows where it

01:12:31

is but but but the interesting part of that is the memories are actually imprinted onto the new brain as it develops and so this new being this new planarian has to rebuild itself along with its memories I mean I don't know how rich a

01:12:44

planarian's memories really are but but however they are it has to like completely rebuild itself from these memories and you know that sounds all crazy and weird and it sounds like um

01:12:55

you know the cases of like like uh you know um what's the uh um you know Blade Runner and everything when you find out that oh crap my right my memories aren't really my memories I mean as far as this planarian brain is concerned it just got

01:13:08

downloaded a bunch of false memories that brain was not part of any of the things it remembers they wasn't there didn't exist so so it's kind of a bunch of false memories and so and so it's like wow these planarians are crazy and this thing with the Androids is nuts and

01:13:21

and my point was no no no this is what we are 24 7. this is completely normal because all of us are reconstructing ourselves at every moment uh and I don't know how wide the moment is but but I'm sure the neuroscientists will tell us uh

01:13:35

you are reconstructing yourself from these from these past memories and so I think I think that's a deep kind of philosophical thing because you know you you're your your self isn't

01:13:47

some permanent uh monadic structure that just kind of exists it's an active construction it's a process it's a you know it it's a constant information processing Auto police is that you know

01:13:59

of the mind doesn't stop during embryogenesis it kind of keeps going it has to and um and and it has these it has these interesting implications if somebody it

01:14:10

will go going going back to you know I guess I guess I'm uh I guess I'm now in the business of um trying to uh normalize a lot of things that people get freaked out about but but but you know imagine right so if somebody finds

01:14:22

out that oh my God like all of these memories that I have now that wasn't me that were downloaded my body was just you know um is it boltzmann or or who is it humor somebody somebody had this puzzle like

01:14:34

what what happens if you're not uh if if all your memories were just like you you were created 10 seconds ago including your memories right or something like that sure um if you think about it hard enough given that that's normally our situation

01:14:48

anyway my answer is who cares great like move on you've got them now go for it now now you you've got some great memories like roll with it because because that's what else is it going to be of course you were just constructed with your memories what else could it be

01:15:01

you you are constantly constructing yourself from the engrams in your in your head you don't have have access to what actually happened before you are up I just don't even you know that that's a that view seems weird to people and I

01:15:13

don't I can't even verbalize what the alternative would be I just don't even understand what what or not what a what an alternative story could possibly be so so from that perspective uh

01:15:24

you know do I and in fact you can you can you can go further with this the body that I have now given given the turnover of cells and molecules in your body was this body actually around 30 years ago to do the things that I

01:15:37

remember doing it actually wasn't we know it wasn't even though I've not been part of some weird memory replacement experiment you know and I'm not an Android this body wasn't there I know it wasn't we know it wasn't there and yet I have these memories so what am I complaining what would you complain

01:15:50

about if if somebody told you that you know you yeah your body was just you know you were you were killed in a record or maybe your body never existed but we just made you and like here are some great memories of a past life like bring it on fantastic you know I hope

01:16:02

they're good ones and because because I don't know what the I don't even know what the alternative would be so so I think uh the for that reason I think all of these things are are are really really hopeful and positive you know they just tell us that

01:16:15

uh we shouldn't be afraid of these of these Technologies this is this is the amazing thing about about being uh uh a self in this in this universe is you you get to you get to constantly construct yourself and by the way guide what

01:16:28

happens in the future what you do now determines the experiences you're going to have the reactions you're going to have which of course you know those those kind of disciplines uh and and and and um uh uh the Traditions are all about

01:16:42

that about about consistent practice to train yourself to be better to have uh you know to improve your cognitive uh you know apparatus and so on and and the commitment is the last thing uh the the bodhisattva vial which is which is huge

01:16:56

um it's this it's this commitment it's it's a medical it's the commitment to enlarge your cognitive apparatus to enable bigger goals to enable you to pursue bigger goals with more compassion facing outwards that's that that I think that I think is critical because I think

01:17:09

once you are a system with the ability to make that commitment it's sort of like it's an exponential rise after that right it's like discovering the scientific process before that it was all sort of screwing around you know

01:17:22

trial and error but as soon as you figured out that there's a systematic thing that you I am going to literally work to get to get to get better at this and to be able to have more increase my light the cognitive light cone of compassion you can now you can now sort

01:17:35

of exponentially go up because you understand what you're doing it's not just you know just just a random walk yeah I love what you've said um before too about how something to this degree

01:17:46

that you can't control your next thought but you can't control your thoughts 10 years from now by what you do today right and what you do every day you can like it influence the the future um

01:17:58

one thing I really want to ask you about and it's it's not in this paper but I I imagine other folks have worked on this um the idea of what it's like to perceive let's say the self in a broader

01:18:12

way and like like you mentioned with the body and I think this is actually really nice because it brings us full circle with our conversation from our very first one which was around the cognitive light cones and in this paper you do

01:18:24

have a diagram where you've you've like overlaid what the uh what someone with a wider sense of self say uh want to help out the community or something that's um what's the one that

01:18:37

takes the vow to you know not achieve enlightenment until everyone else does right it's like this uh leave no living being behind kind of idea all right um but do we have any research or even

01:18:49

discussions with say monks people who who engage in uh say like deep uh Buddhist meditative practices do they perceive themselves as like a part of a whole like is is there actual sensory

01:19:03

experience different than say a very unenlightened person like myself or like like most folks like does that make any sense like how how they perceive themselves in the world yeah I think I think that's a great question I don't

01:19:16

think I've got the expertise to answer that question I think um you could talk to uh you could you could have Thomas doctor on uh for example and uh he he he's he's a very uh kind of experienced scholar in that area and um uh I think I

01:19:30

think he would he would you'd have a good time he could explain all that stuff well thanks yeah yeah I think I'll have to and um let's see and that's great so you gave me so much to choose chew on and so much for the

01:19:43

audience too I'm sure this idea that who you are today is not who you were 30 years ago just in a very literal materialistic sense um that there's like just this illusion of a of a consistent continuous kind of

01:19:56

experience it's that's really uh it's a wonderful way to put it and if you if you wouldn't mind I mean I do I'll probably overlay the um their cognitive light cone with a with the bodhisattva uh the body's life of

01:20:09

the vowel one um how has this been because this came out this paper came out about a year ago how has it been received have you heard any response from the paper um uh

01:20:21

a little bit uh a little a little bit um yeah I I yeah I don't know I don't I don't track uh I don't track responses super super carefully so I'm not sure um haven't heard haven't heard a ton I'm

01:20:35

going to guess I'm gonna guess Thomas um or you know heard more on it but I think that uh you know I've certainly had interesting people reach out to me to talk about it which is you know pretty much uh one of the things I hope for in writing these things

01:20:48

um so I've had I've had lots of cool discussions I don't know more broadly I mean I have no idea how in any of these papers by the way I have no clue like like who you know if anybody reads them or who reads them or or what happens after that I don't know yeah yeah well I

01:21:00

appreciate that and certainly people people you know that people contact me but I I know it's it's very hard to know um you know how these things are are actually spreading or not spreading through the through the community gotcha and I'm Gonna Let You Go in a moment but

01:21:13

before I do one last thing you're a big mid Journey fan um uh you oftentimes have seems like you're also a fan of surrealist art I know I've seen stuff from uh you know prompts that are like do this in the

01:21:27

style of the Codex seraphinionists or in the in the style of like Hieronymus Bosch do you know like what is it that draws you towards uh like those kind of art styles um

01:21:39

I I don't know uh I have not I have to have had zero art training I you know I kind of know what I like but but I don't know anything about art I can't I can't draw myself at all like nothing and uh

01:21:51

yeah and uh I mean all all I know is that um I really like uh I'm really interested in the space of the possible and much much more so than than uh the

01:22:04

actual and uh all every everything everything I look at Art wise is either uh photography of nature or photography thereof I like you know I like nature

01:22:16

but other than that uh um yeah I'm really into uh imagining what what could be the latent space of of possibilities and I think you know uh mid-journey and systems like that are are pretty pretty cool in that respect

01:22:30

they let you uh explore this this this wacky latent space that it has um of images of all different of all different kinds um yeah that's just kind of that's just kind of generally um

01:22:41

I I think I think very much uh forward in terms of like what now what next what could we do next that's that maybe you know maybe maybe that's why maybe that explains a lot of my uh kind of um

01:22:54

um you know not not getting worked up about whether the past is real or not or what it's it's mostly it's mostly the forward-looking stuff I mean like whatever in the past but like now what do we do now so that's that's that's more what I'm interested in I and and I

01:23:06

like I like this kind of uh this kind of art that uh that lets you uh yeah imagine things that that could you know that could be what what are the what are the possibilities moving forward yeah I share Same Same Love for that style as

01:23:18

well those Styles and um just want to say thank you so much Mike this these three conversations have been wonderful I appreciate the time your energy and uh where should people find out more about you I'll link in the description yeah um

01:23:31

well everything is uh though there's an official web the academic website is www.drmike11.org so one word Dr mike11.org and that's got all the links to the to the papers the software the um

01:23:44

presentations everything else uh I've got a science Twitter Presence at Dr Mike 11 and I think in a couple of months there will be a WordPress site so I've been working on a site uh kind of yeah I was sort of I was sort of

01:23:58

thinking sub stack WordPress yeah I think I've decided WordPress and so so there's going to be a site that's kind of less the official academic stuff and all kinds of writing that I want to do that doesn't really you know okay kind of kind of tired of asking the question

01:24:11

of you know you write this thing and it's not really um uh it isn't a primary paper and it isn't really a review and it's kind of a perspective but it's really interdisciplinary it's way too long for a journal and and I'm sort of just tired of this issue of finding a

01:24:24

home for it you know what they say like where are we going to put like okay forget it we'll just it'll go there and then anybody who's interested can you can read and that'll be that so um yeah so there'll be that doesn't exist yet but but in the next couple of months it

01:24:37

should be it should be up awesome great well we will watch that space so now when it comes live I'll update all my other videos that's great I appreciate that thank you thank you absolutely thank you Mike thanks thanks very much yeah thanks for having me on it was

01:24:49

great thanks it's awesome thank you so much all right all right