ROUGHT TRANSCRIPT: David Krakauer
We talk to the Santa Fe Institute President about the history of complexity science, artificial intelligence and the other technologies making us stupider, and Wall-E.
The following is a rough transcript of our episode with David Krakauer, “The Long Shadow of Complexity.”
Please note that this is a full, rough, unedited transcript. If you’d like us to polish and edit these transcripts, please consider supporting Uncertain Things as a paid subscriber!
Vanessa: Hello, hello, hello Uncertain Things listeners. It is, I, Vanessa. Adaam has lost his voice supposedly from being sick, but I, I gather it was probably from a late night of karaoke, but let's say it, he is sick and karaoke just too hard, and so it's just I today here to introduce our guest this week, David Krakauer, who is the president of the Santa Fe Institute.
So this institute, frankly, does things that my limited brain can't really understand. It is one of the most unique. Academic institutions out there and that it is. Remarkably and conscientiously cross-disciplinary. They do not break things up into disciplines or silos. They rather tackle issues, questions, hypotheses, and gather a group of very smart individuals from cross disciplines to think through them.
So David is incredibly intelligent, uh, evolutionary biologist who has made it his mission to research. Both the question of intelligence and also stupidity. Um, so if you want a nice introduction to David, you might wanna check out Sam Harris's podcast, which is a nice kind of def setting of definitions, type of conversation.
We kind of delve into this, uh, into more of the realm of complexity science and its history. We get into a really interesting conversation about the aesthetics of complexity versus the aesthetics of kind of more minimal minimalist, uh, ways of thinking, including kind of more physics and mathematical based ways of thinking.
And what are the aesthetics of that versus the kind of more maximalist naturalist, uh, aesthetics we go into as is our want, the inadequacies of academia and why their institute has set up to, uh, buck a lot of the trends out there right now. And we also have a really, really, really interesting part about, uh, something that David calls cognitive artifacts.
There are complimentary cognitive artifacts and there are competitive cognitive artifacts. He gets into it, uh, better than I, but suffice it to say that we are living in a world with competitive cognitive artifacts that basically make us dumber. Like your smartphone, your g p s rather than embracing complimentary cognitive artifacts, things like, uh, an abacus or, uh, even a piano.
We kind of get into this, uh, discussion around music and how an instrument like a piano can augment your understanding of musical theory, um, and enhance it even. So, really interesting conversation a little bit outside our field of, um, you know, humanities, uh, vent dripping our toes into the, into the field of science and mathematics, uh, which, you know, frankly scares me a little.
But, uh, these are uncertain things and, um, we, we try to dive into them with, uh, curiosity and questioning. So with that, I hope you enjoy our conversation with David Krakauer,
Adaam: you already caught my, my attention. Ok. Yeah. You wanted talk about the, um, not, not necessarily the, not to start with the current working definition of complexity, but you want to give us a bit of a.
Exploration of the history of it. And by the way, history is something that I, I definitely wanna talk to you about when it comes to, like, I read your essay, um, and, and the collection of essay that Santa Fe Institute Yeah. Published on, on thinking historically. And we recently, um, interviewed Peter Turin.
So the whole area of how to think, um, ok. More analytically on history will be something that we get to, but now we're talking about the more conventional, soft version of history. What is the story of
David Krakauer: complexity? Yeah. Yeah. Okay. Okay. So let's just, and you just interrupt me as you, as you see fit, I mean, and vice versa.
Yes, we'll do, and um, so really this, you know, people often ask me know, what is complexity? And it seems grossly unfair as a question. It's like, what is chemistry? What is English literature? What is history? These are unanswerable questions, right? Um, and. The way that we typically answer them is either we give examples of where they get applied, who represents those areas, you know, who studies them or we give a history.
And I think the history is the most compelling way of trying to tease apart what makes complexity science different. And um, and the way I see it just in brief, and we can jump right into any of these areas, is,
you know, the history of science has been the history of coming to terms with low hanging fruit, right? And every civilization on earth recognizes regularity in the world. They live particularly celestial regularities. And that's sort of the history of physics, right? With an
Vanessa: example being, I know the. Sets and rises every day.
Regularity. That's
David Krakauer: let's, yeah. The orbits of the planets, for example, the, the, the, the, the trajectory of the moon, the trajectory of the sun, the constellations in the night sky. The first move of complexity science, right, was in some sense, or in science generally, was to start to reconcile the minimalist sciences.
Physics, beautifully mathematical and elegant. The kind of mondrian painting of the epistemology epistemological world, um, with a maximalism of natural history, medicine, biology, Linus, all that stuff, right? So very different traditions. Um, one very mathematical, very empirical, both empirical, quite frankly.
Um, but one very descriptive and one very mechanistic and so on. That was one integration. It's still not over. That continues to this day.
Adaam: Can you go into like a, a bit more into how those two fields, um, interact? Cause to me there is almost something instinctively, I mean I shouldn't say irreconcilable, but very difficult to reconcile the sort of mathematical and bias of towards beauty and symmetry that exists within physics, with the study of, of medicine and, um, and biology.
David Krakauer: Yeah, no, that's absolutely right. And so, yes. So one way to say that right, is there are fields that are interested in symmetry or invariant law, things that are, um, symmetric with respect to time and space, you know, and then there are fields which are fascinated with contingent histories. That could be your history, your developmental history, or it could be the history of evolution itself.
And, and the latter seems very messy. You're right. And a whole new set of mathematical techniques had to be developed to deal with what we would call domains of broken symmetry. And we can get there in a bit, but you're completely right. And they, they are very different. And one of the ways, um, that you see that is we build really great computer games to simulate billard tables.
And we, and they're kind of perfect. They're exact, but when we build simulations of society like Sim City, it's a cartoon. We don't, I mean, unless you're weird, you wouldn't use that for policymaking. I mean, probably some people do, but you know, God knows. So, but that you are right. That's one, uh, interesting dichotomy that's been very hard to reconcile.
Another one by the way is the, within the sciences, sort of natural science, social science, both descriptive domains like the biological sciences and economics. For example, anthropology. So there's the sort of biological versus physics, there's the social science versus biological science. And then most recently, how to integrate that into principles that come from engineering.
Right? Um, starting in the forties and fifties, computation, uh, information theory and so forth. That's another integration. So in some sense, complexity science is, if you like, If you have a sort of a paint palette and each of those is a different color and you mish them all together, um, I'm sure that's a word.
Um, mush them all together, I think whatever, um, mix them all together. That's the word. I think I was looking for that. Then how do you prevent it just from being this somewhat unappealing gray? What new structured forms of knowledge emerge when you integrate the physical world? Sometimes called the sciences of simplicity.
Um, the multifarious physical world, sometimes called disorganized complexity and we'll get to that. And then organized complexity, which we tend to all be obsessed with, but human society, the products of human ingenuity and so on. And that has a fascinating history. I just wanted to make a point that those, integrating those is what's at stake.
Right,
Adaam: right. No, so, ok. So I I, I, I've, I also have a
Vanessa: clarifying question, but go
Adaam: ahead. Oh, no, I, I, I have a, a confounding question, so we'll start with the clarifying question.
Vanessa: Ok. So just to make sure I'm following, you're saying that there's different kind paradigms I suppose of, of knowledge seeking and they often.
It's not as if we are going like in some sort of linear progression where one is leading to the other. In fact they're probably overlapping and atten occasionally in tension with each other. But perhaps some of them co become more in vogue or less in vogue over time. And for example, now we're more in an engineering paradigm perhaps of how we approach knowledge.
Um, but I th uh, is what you're saying then that, um, understanding the tensions between those different paradigms is, is what is worth in interrogating and, and, and clarifying what the two different ways of knowledge understanding can
Adaam: continue. I think we, I don't think we even had a prescriptive, uh, argument yet.
Ok. It was just that this is the history of complexity science. Like this is, this is the complexity, right? This is, but I am I
Vanessa: understanding right though that there are kind competing paradigms that are overlapping with each other at least.
David Krakauer: Yeah. I think, you know, historically there are people who've been drawn to these beautiful.
Symmetries and simplicities and their yield very productively to mathematical description, right? That sort of, and there's been this other domain where we revel in its multifariousness, its sort of baroque quality and natural history and biological science to
Adaam: the point where people like me who come from that world are, are have the b like ha have almost the opposite bias where they say like, you could never, um, lock my field history or the understanding of human behavior or morality into, uh, a, a beautified simplified re reductive model that we just could just never fit.
The chaos of human experience could never fit your, your beautiful models. Um, and then obviously like to jump ahead. Like that's part of, that's how we started the conversation with Peter Turchin, where people from that field will say like, oh, well, well don't be so arrogant to, to think that these things are not
David Krakauer: reconcilable.
Right? No. Okay. So that's very important, right? That, that, that tension and the final one, the things that we build, um, that have actually been putting us all in a shoebox like chat, G B T and so forth, which take all that complexity that you are talking about arm instead of reducing us to a set of very simple rules in a deep neural network.
It's not reducing us to the world of classical mathematical physics, but it is rule-based and, um, a, a certain kind of reduction of complexity. But, but anyway, those you are right. You could call them paradigms. I think what's happening in the modern world is we realize that all of those things are, are sort of laid one on top of the other.
The separation that we achieved in the 17th and 18th and 19th centuries doesn't seem to be working very well in the late 20th, 21st. We are, all of us part of ecology. We are all. Part of some nefarious predictive framework to set us better, you know, shoes. Uh, so we we're living at once in all of these frameworks at the same time.
And I think the question is, is there a science in a sense that allows us to navigate that world without factor it or atomizing it, as you said, into its distinct paradigms, each of which performs well with very restrictive assumptions being made. And, um, so in that sense, that has, is the history. And I think in the early part of the 20th century, mainly in physics there, and an or physicist moving into biology was saying, wait a minute, how can you talk about ecology without talking about energy, you know, without conservation principles, key concept in physics.
And so that, that, that effort to integrate, um, Fundamental conserved variables from the physical world, energy and mass, for example, and then put them in the service of a biological phenomenon like growth and reproduction. Then immediately you are, you are that interface of the paradigms of biology and physics and biology have been cavalier with that.
Well, yeah, you need metabolic energy, but, you know, just eat more. So, you know, how do you connect to the fundamental
Adaam: physical thing? Or, or like, a way to think about this is, I, um, in, in college, um, I, I was a student of, uh, Yuval Noah and he starts his lectures like setting up the different, um, you use the word mysteria, but Maia of, of human sciences says, you know, starts with physics.
It's, it's in his very popular tone. But the, but the idea is like, we start with physics. And then when it turns into the realm of living creatures, it becomes biology. And then ultimately leads to history. Like, and, and the implication is that each of these areas are hierarchically connected, but each discipline tells its own story.
And the
David Krakauer: implication, I consider, I have to interrupt you. I consider that fundamentally wrong. No, exactly.
Adaam: That's what I'm, that's what I'm saying. That this is what you're basically pushing against that the, the intuition that, um, or the concern that these different fields don't have much to say about the other.
Right.
David Krakauer: Lemme give exact, so let me, let me, let me clarify this a bit. So that's the, um, received wisdom. That is you like, that's the standard model of the discipline. Right. And universities and institutions are organized accordingly. Right,
Adaam: right. They're big, they're built into disciplines to teach you how to handle your, even, even sub-disciplines.
And I'll, I'll just even, uh, uh, an anecdote that me and my, a friend of mine we interviewed Chesky, was the, um, now I think the Dean of the humanities in, uh, Hebrew University, or was for a while when he had to, um, decide on budgeting for the linguistics department. The two different factions in the linguistics department, the, uh, the structuralist and the generative Chomsky wouldn't even agree to share any sort of panel because they thought that their maier is so separate that they can't really talk to each other.
Um, so let alone can linguistics talk to, I don't know, um, sociologists or linguistics talk to, um, uh, physicists. Right, exactly.
David Krakauer: But that, that's, so here's, there's a very simple gout like inversion you can perform to eradicate this nonsense. And, and that is that, yes, if you define. Your disciplines by domain, which is what most people do.
If you think about it, I study the non-living world or the non-living. I study very, very big things like stars or very, very little things like quacks and so on. Therefore, I'm a physicist. Um, oddly enough, you go to a physics department, 50% in the studying biology, and then you say, well, what makes your physicist?
And, you know, well, I do back of the envelope calculations, so okay, I do compressive mathematics. And so it becomes so absurd at a certain point that you have the edifice should apart. But one way to do this is say, look, let's take of core principles, energy, information, um, strategic interactions, cooperation, defection, make, and make those primary and then say, okay, I'm gonna study a domain.
It might be a physical one, but I'm gonna use those concepts. So one of the things that complexity science does is it inverts the standard epistemological hierarchy instead of making the domain preeminent. Where every domain, shares, principles, make the principles preeminent and make the disciplinary domains case studies.
And, and outta that, you get this shared language. Cuz if I sit down with an archeologist and say, you know, I'm interested in energy and matter, they'd say, well, so am I. You know, I study Mayan temples and you study atoms, but we're both interested in energy and matter. So that's a really important inversion institutionally actually for us, because all of a sudden, rather than speaking across purposes, we're talking about the same thing.
Adaam: Uh, so how do you avoid, and I, I keep for stalling, for letting you actually tell the story of the evolution of this, um, idea and feel. And I, I, I will relinquish you to go at it, but, um, but how do you avoid in, in, in trying to draw these, uh, foundational principles or as building blocks of ideas? Um, A, the challenge of just being seduced by a really effective metaphor that seems to be, uh, applicable across the domains, but might just be as, um, you know, deceptive as it is, um, congruent, uh, or more, more so.
And then how do you know that even if there, even if it's not just a, a metaphor, like when you think about, um, I dunno, uh, uh, you had a lecture about conflict or you have an idea of, um, Uh, I'm, I'm trying to think of all those like, uh, um, evolutionary psychology, uh, terminologies that you do see crop up across fields.
How do you know that you are not like mating? You have the concept of mating or, uh, um, um, show up in different, uh, fields, but how do you know that you are not just using a, a, a, a borrowed concept really effectively and you're truly talking about similar phenomena?
David Krakauer: No, I mean, the fact is you don't at first, right?
I mean, the reality is you try very hard to instrumentalize the concept. It's, I'll give you an example. Um, the confusion that often comes up between reproduction and replication, you know, we tend to use those terms interchangeably, but when you formalize them, you realize replication is doubling and reproduction is halving.
That's what you do, right? When you, you reduce if half of your genetic material and. Once you've got there, you think, okay, now I have a better understanding sort of formally of what I'm talking about and I can extrapolate to other domains. And so when you talk about culture and people say things like, oh, my idea is reproducing.
Well, is it, is it being halved in each transmission? No, sorry, I didn't mean it was reproducing. I mean it's replicating and so on. So I think the key is to, um, very quickly get to that sort of instrumental definition, which is often somewhat diluted, right? It doesn't capture everything, but at least it gives you this lingua franca that you can use that minimizes some of the metaphorical pitfalls.
Um, they're always there. So
Adaam: as a semantic burden of clarifi of clarification is high priority when you're trying to bridge. And like multiplicity of highly evolved semantic fields, kinda like crunch them together and say like, okay, let's build the, build the lingua franca that is gonna be usable across the board.
Right? Yes.
David Krakauer: I mean that an interesting, I mean, it's not, but this, it isn't a very interesting dimension to the history because, well, institutionally, we are a mathematics and computation institute at our core, and one of the reasons has to do with the reduction of communicative uncertainty. There's a sense in which natural language is much more poly than mathematics.
Right. But, um, just this point that, as you said, if, if, if I use the word reproduction versus I use this mathematical operation of the information, Clearly the former has vastly more ambiguity in it than the latter. Latter still has ambiguity in it. What do you mean by information? You know, on and on. But the, so I think the key to the, to our success in this weird integration of paradigms, right, is finding the language which does the best job, not the perfect job, but the best job of minimizing ambiguity.
I think it's absolutely true that that's a very interesting move. And I think it's, um, there have been histories written about the, the, the obsession with metrics and quantity, and there was a beautiful book called The Reign of Quantity. And the main point being, even though quantity is approximate, it allows for consensus formation more quickly.
So,
Adaam: So did you have an example from your work or work that you've oversaw, uh, at Santa Fe where you've encountered an, an idea or, or a area where you felt, felt like you were making progress, but in fact it was semantic, fallibility or there was some, some something that fell on, on poorly defined ideas?
David Krakauer: Yeah, I mean, uh, this is work actually many years ago that I did with Jessica Flat. We were working on power, social power and, you know, not the physics sense of power, but the, but the power in society of power from individual and all that nonsense. And the question there was what do we mean by it? And um, in some of our earliest papers, I realized and just realized that there were, people were interpreting that very differently from our intention.
So we quickly formalized it using information theory and in the process, as one always does, losing. Certain nuances, which culture has stored in language use in language communities, but gaining, um, clarity and the ability therefore to coordinate on a shared meaning, it's always this way. I mean, that's sort of the, the, the, the bind of formalization.
You, you, you gain clarity and you lose complexity.
Adaam: Do, do you remember a specific idea that fell through or fell apart the mo in the context of power that you had to give up on?
David Krakauer: Well, for us, I, I don't know. I mean, oh, you mean what, what subtleties do you mean? Yeah. What subtle subtlety? Yes, exactly. Well, okay, so this formalization of power is based on the idea that power is conferred.
Power cannot be claimed or arrived at. Alone, it's conferred, it's conferred attribute. Right. And so like being voted for that would be the simplest case. I'm powerful by virtue of a democratic electoral system. But clearly there are other notions of power that nothing to do with that. You could say what despotic power that's not in the definition.
Adaam: Interesting. Right. Why? And and does that mean that when you are thinking about things like, um, like a coup, are you imagine, is, is that something that has to be understood within terms of con or is that something that just, that doesn't exist within the realm of your power
David Krakauer: model? It doesn't exist within our definition.
And I think I would either say to you there are two responses. One would be you're not talking about power, you're talking about coercion or something like that. Or give me your definition. And so we can call it power one and power two. That's also okay. It's just that I want to make sure that. The thing I'm talking about is well understood and can be quantified, right?
And we can measure it. And who can assign different power values to different individuals based on that definition? I mean, all of the history of physics and biology once, for example, to a history, again, once evolutionary theory was mathematized in the 1920s and thirties, huge amount of i, of, of, of ideas, a huge quantity of ideas that Darwin had first and Huley and others had proposed were lost because there was only one thing that the modern synthesis fixated on.
And that was the fixation of alleles. You know, so that's something that's often not carefully discussed. I think that, um, it's why history should be in textbooks, by the way, because it's in the history that you fess up to what you assumed. If you present the theory in its final form, you don't quite know what was swept under the carpet.
Adaam: Although a great deal of the work of, um, history in general when you're trying to understand society changing over time, but especially when you're talking about the evolution of ideas and the intellectual history fa falling on these definitional mismatches over time. What did, what did the word liberty mean in, in, in the Roman Republic and, and it's incongruity with what we're thinking about when we are touting freedom in, at a political rally today.
That's, that's something that's sometimes is deceptively transparent because they, the, a lot of the language and the the, um, The assertions around that word would be very similar. It wouldn't be just that they use, they use the same word liberals and liberty, but they also make very similar proclamations.
Partly, partly because our modern political world has like Drew inspiration and had some emetic relationship with the Roman Republic, and yet the underlying value or, or a phenomenological quality of whatever liberty might have meant is completely lost.
David Krakauer: Yeah. I mean that's why, I mean, of course that's why CSIRO made that dis distinction between the signifier and the signified.
It, it, they're not the same thing, you know? And I think the, the way I view language essentially is a series of, of, of, of memory addresses, um, that point to, uh, memory that stores the actual semantic content. Right. In other words, so when I say alligator, According to your experience, you'll have more or less bits assigned to that concept.
Right. And I think that, and of course, you know, um, as Alberto Echo writes about the platypus, you know, what was the early, what was that thing early on? And now we have, we now understand it's a monitoring and it's a different, different group of animals. Oddly enough, in the late nineties I started working on this problem.
I started working on, um, the evolution of language and. Sign signifier relations based on wittgenstein's philosophical investigations. And we started developing game theoretical models for arriving at shared meanings. And that's a whole other discussion perhaps, but it's a very hard problem and it, it grows exponentially more difficult as the number of objects and signs increases.
So you're constantly settling into approximate local minimum where you think you know what someone else means. But the reality is, if you were to interrogate it at a higher resolution, it would be revealed that that was
Adaam: not the case. And by the way, we were talking about, we're talking about this in terms of, um, time, like over time, but also now with communication being so widespread and you, you can have a conversation, like say a political conversation with somebody on Twitter who across the country, across the world, You don't even understand you, you can't even be sure that you are talking about the same idea when you're talking about values or for that matter, any kind of use of language.
And I, especially when language is stretched out on such scale as opposed to being communicated just like a a by a local town or a local, uh, community. But when you have English being used as the language of the, of the globalized world, you know that there's just complete incongruity in terms of how we are we, are we understanding what we're saying, even if the, the language has been reduced enough so that we can kind of understand, approximate our understanding of each other.
Yeah.
David Krakauer: No, it's, it's a good example from complexity. Science is entropy, right? And that's a concept that has its roots in thermodynamics and statistical mechanics in relation to the production of heat. That is sources of energy that are not useful for performing work. Okay. And then slightly modulated via Shannon in information theory to be uncertainty.
That's a statistical inferential concept, not a physical concept, formalized eventually by Edwin James in the theory of maximum entropy. But that's a beautiful example actually, of a concept that starts in physics of molecules and particles bumping into each other and before too long becomes a theory of knowledge, a theory of inference.
So this field is just suffused with, um, these migrations of concepts, sometimes productive and sometimes dis simulating the, the real meaning. Mm-hmm.
Vanessa: So you said earlier that the way that your. Organizing yourselves as an institute is around these principles in order to hopefully have the kinds of conversations that will break out of the paradigms that people have been raised in and locked into their domains.
How do you know if it's successful? How are you, how do you, and not to, you know, fall under the reign of quantity here, but how are you measuring, uh, uh, a successful collaboration around a principle?
David Krakauer: Right. Okay. So this gets into a very gnarly wood. It's like Fang Horn Forest of this discussion. Um, the, I make a very strong distinction between judgment and impact.
And I think what's been happening in the modern world is we've been using impact as a metric for good judgment and some mistake. Ok? Um, so let's just make that more concrete, um, book sales. That's impact, right? You are an author, um, and your publisher wants your advance to be made back, you know, and, but no one would be so stupid as to assert that book Sales is proportional to quality.
Because without naming names, the best selling books we would rarely describe as the best book popular books in a thought, logical sense, fun books. Okay? And the same goes for everything. It's for music, you know? I mean, again, I don't want to, you know, but just look at the top five musicians, and I don't even think they would say they're as.
Good as Miles Davis or you know, Johann Sebastian Barr. Um, so impact is one thing and judgment is another. And the way I, judgment is a much harder thing to define and it can suffer from cronyism and selectivity and bias. And so the move towards the metric is driven in part by an understandable desire for equity.
On the other hand, it suffers from popularism and lock-in, right? So the way I've thought about it as an institute is sfi as a judgment inte institute. We are unapologetically driven by our interests and we don't care either. Those interests are very successful. The, the, the, the, the application of those interests, for example, a theory or a model, the world makes that determination.
Right. In other words, let others make that determination, not us. If I have a theory that's wrong, no one's gonna listen to me, right? And in the long run, we'll sort of we'll shrivel up and die. I consider that separation of powers crucial in any creative enterprise. If you are a truly creative artist, musician, the last thing you would do unless you were sort of a coward, is poll society and say, what would you like me to sound like?
And, you know, any artist listening to this would be, you know, feeling deeply uncomfortable by that idea because that's not how they operate. They, they assert their own crazy egos in the hope that the world likes them. No different in science, no different in any creative enterprise. So my view is think very carefully about judgment.
Exercise your judgment internally. Allow others to exercise selectivity to evaluate impact. Now, what happens, by the way, is I, I talk to a lot of foundations and they'll say, David, that's all very good, but I have to give out money. And you are telling me I'm just gonna exercise my judgment. But then when I come back to the board meeting, they say, yes, but what happened is your judgment sound.
And um, and so what I start doing is in a slow diffusive way, integrating and dogen my predicted impact into my judgment. After all I've learned from my experiences, right? And before long, you are just a mirror of the market. You're not a creative mind. Um, so I guess my job in part is to say, you know, Pursue your passion, pursue your interests.
Clearly they're a part of history. You can't completely isolate yourself from what was effective, but for a period of time, for a period of years, ignore what your peers say is good and bad. We will not internally evaluate you according to peer metrics, like citation metrics. We don't care. It's not a part of our game a hundred years, hence they might be informative or whatever, 20 years hence, or some period of time.
But, so it does require an active strategy to keep the kinda impact metric obsession out your hair
Adaam: that's, uh, uh, there is so much latent or, or not so latent. Um, Uh, rage in what you were saying against the
David Krakauer: Yes. Not mentioned. No, it's totally manifest, it's manifest
Adaam: rage. I mean, well, you know, you know, let, lemme put a pin in it.
Let, I, I wanna let you, uh, go back to your history and then we're gonna get to our current state of, of academic thinking. And this is, this is as structured and logical as our conversations get. So let me go back to you and, and the story of this idea of complexity.
David Krakauer: Yeah, no, I'm totally happy for us to be oblique and digress, but, okay.
So let's just go back a little bit. In the forties that was an effort to sort of try and understand the relationship between formal logic of the kind that girdle had worked on and Turing had worked on, and how the brain and mind works. So that interface interests me a great deal. Um, the, the interface between logic and life.
In particular, it's been extremely instrumental in the development of complexity science. One other like that, just to throw into the mix, is cybernetics. That was the interface between dynamic systems and control theory and homeostasis and selection. Um, so over time, I guess what I'm trying to do is enumerate a whole series of collisions or interfaces, and they keep going every decade and multiplying, um, and complexity in some sense.
Is this dialectical enterprise to try and resolve apparent tensions between these, the thesis and the antithesis in each particular instance. When you're
Adaam: saying the thesis antithesis on a particular question, you're, you're talking across disciplines, right? A question that is an an emergent question that, that is not restricted to one field.
David Krakauer: Yeah, I mean, for example, this idea that in economic theory in the forties developed in the Austrian school by people like Hayek who are now associated, I think largely with neoliberal thought, but who were some of the people who said, you know, how do we resolve the mechanism through which the invisible hand of Adam Smith works and Hayak said is resolved through the pricing system, through markets clearing, and we need to understand that as information.
So this was the first interface between, and a formal effort to define what information was. Oh, that's interesting. And, and a sort of an 18th and 19th century political economic theory. Another one. Right. Um, of course that wasn't, even to this day, that isn't fully resolved. Right.
Adaam: On a personal level that this question of information is one of the things that has.
Perplexed me and fascinated me for, I'd say a decade. The, the, it's one of those concepts that is just in common use and the moment you try to think of it in, in a more effective, definitional way, I dunno, at least my small brain explodes from that,
David Krakauer: but we probably should discuss information. I just sort of preoccupied with this idea that over the course of time, either ideas have been independently discovered and did and discovered to have the same form or two fields have come into collision and progress has been made in their reconciliation and at the largest scale for me in complexity, it is the reconciliation of the universal laws of the universe.
Dominated by symmetry and the contingent rules of life dominated by broken symmetry. And that is
Adaam: the game when you give in. If you ever do to more simplistic fairytale, um, or I should say simplistic, simplified tale version of your, uh, journey or your mission as you see it, do you imagine the, the place, the locus where this reconciliation happens?
Do you, can you imagine it, can you see like a certain type of compromise between the overly idealized beauty minded symmetry of the physicist and the, the, the, the chaos devotion of the
David Krakauer: humanist? I do believe that complexity in that respect is the third. Aesthetic option. Hmm. And um, in fact in the founding sfi, this was discussed at length.
Actually, there are those who enjoy the pristine, perfect regular crystalline, the sort of apollonian right worldview. And then there are those who enjoy reveling in mess and uncertainty of a certain kind. And they're the Ians where are the odys? And this is this idea for people who have an ultimate ordering principle, but need to pass through chaos.
And I think it's an aesthetic preference. I do believe that there are some of us who like them both, you know, and in a way you could argue that's all of culture because. If you ask people, you know, but we've talked about artists. If you think about his kandinski, his Pollock, and his Ian, well, you know, it's a smaller crew who liked the other two, right?
Um, the realism with some degree of abstraction seems to be preferred. It's true also in music, I mean, You know, um, some people find listening to
Adaam: Wait, you, so your attribute Ian, to the, to the third aesthetic.
David Krakauer: To the third aesthetic. Third to the, to the,
Adaam: it's a bit of a stolen base, I'd say.
David Krakauer: Well, ok. Ok. What's wrong with that?
Adaam: No, base stealing is legitimate. But, but if anything, I would imagine Titian as a little more leaning towards the apollonian.
David Krakauer: Fair enough. I mean Okay. And we could decide, we could have a wonderful dinner cocktail party where we face people in this space. Right. And usually I put the South American surrealist matter in the middle cause he an interesting case.
But I feel that, um, even though it feels like a novel alternative, To the stereotypes of the two extremes, the biologist, physicist, or, you know, mathematician, social scientist. Um, it's actually what most people like best, most music. For example, if I just place you one note forever, right? Or if I played you random notes and I what I somehow in the middle between the two.
I want repetition, but I also want symmetry. Uh, so whether it's visual or auditory or even in mathematical theories, we're, we're sort of drawn to things which can accommodate themselves to a degree of chance like quantum mechanics, but have an underlying determinism, like the wave equation. So I would actually argue in a weird sort of way, complexity of what we have always been, but it's the hardest one to theorize and the two extremes we know how to mathematize, right?
So we sort of, we've carved out the world according to expediency rather than reality.
Vanessa: You brought up in your, bring up in your writing, often you reference art, you reference books, your print literature. In, in this conversation you've brought up music, uh, you've brought up art. Right? And this is something that I think is atrophying, I think in our education there, there seems to be a lot of emphasis in academia on the hard sciences, and we can, you can see the humanities kind of fewer, smaller budgets, less enrollees, less focused from the, the academic world.
I mean, uh, I'm assuming from your perspective as someone who tries to grapple with problem sets, drawing from a bunch of different mindset sets and perspectives, what does it mean for the future of us solving these thorny problems when we are essentially cutting off or devaluing a whole way? Whole ways of thinking about our world.
David Krakauer: Yeah, I, I mean, I think it's tragic. You know, the way I think about it at least is that, and Richard Feynman made these points in his lectures at Cornell that if you're dealing in the world of identical, numerous particles that obey a small set of universal laws, they lend themselves to a form of description in mathematics.
But as you add contingencies and histories, parameters, the language has to grow and dealing with broken symmetries, right? Things that have to be described in their own terms, if you like.
If I wanted to characterize your life as opposed to a human being's life, or a mammals or a multicellular organism or a life form, those sort of increasingly low resolution concepts as you get high resolution to a single life narrative is the right description.
It is the correct theorization of life. Dickstein made that point in his late writings on psychology. He said, if I really want to explain this concept to you with any fidelity, I'd have to do it through a narrative, through a story. Cause it's the embodiment of the concept, right? Certain concepts need to be embodied.
Others are so generic that they don't want to be embodied at the speed of light. So for me, it's actually a spectrum of forms of understanding. They are all actually in the same spectrum. They lie on the same electromagnetic spectrum of ideas like, but some of them are very high frequency, high resolution, the novel, the history, and some are low resolution Maxwell's equations.
They're
Adaam: high low resolution in your metaphor, in our ability to perceive and parse them out.
David Krakauer: Both, you know, in other words, if, if I said to you
as a mammal, why are you in a mood? And you could say, well, my endocrine system is out of whack. I've been eating too many hamburgers. You know, there's a sort of level of description that seems appropriate when answering that question with mammal. But if you said, Vanessa, why are you in a mood? And you said, oh, cause you are endocrin you, you'd be rightly offended.
Right? It's not the right level of granularity. It's because the jerk who lives next door, you know, peed through my mailbox again. You know? And so that's the sense in which the, the level of granularity that you are making, the observation determines the kind of theory and description you need. And I'm saying that if society only sees the lowest resolution regularities, which can be answered with reductive, neurophysiological, or physical theory, that would be a problem, right?
Because you are not making distinctions between individual lives and experiences. So for me, the novel and music are important, not cause they're in entertainment outside of my work, but cause they're the right way to think about high resolution phenomena in the human domain.
Adaam: Do you see any, um, interface in that sense or with that world in a way that maybe even it's.
It shapes your ability to, to to think in other fields or not, not to turn art into your utilitarian uh, product. But is there, do
David Krakauer: you have to be careful, Adam? You have to be very careful with that because what you might be saying is, the reason why I read Toto is to do better physics.
Adaam: That's what I was trying to avoid and
David Krakauer: that's not what you wanna say.
The read I wanted to, to read is to, I understand humanity better at the resolution of the lived experience.
Adaam: No, but, but I am interested that, that's why I was trying to be careful in, in saying, cuz I don't want to talk about art as a, as for utility, but I am interested in the way that things interface. So I, I do wonder if you find that connection from the, forget about the art itself, but the way that thinking about the human experience reflects back into the world of, of l lofty symmetries.
David Krakauer: Well, lemme give you an example from the history of mathematics. So there was a movement, primarily European, primarily French, associated with people like Andre Vey, Groden Deek, uh, in mathematics that came under the fictional authorship of a, a abbreviated name called Bo Barky. And this was a group of people they wrote under this assumed name, and it was actually their, their initials that were concatenated.
And they wrote very formal mathematics. They were, they were very despondent about the fact that the calculus as developed by Newton and Leitz was a little bit too, um, unprincipled. They wanted to really ize it and derive it from first principles, Manian Set Theory. And they wrote these books, Bob books, which are unreadable.
They're unreadable because they're instantly operating at that most clinical absolute level, you know, consider. A home amorphic contraption of the S nine variety, that sort of language. And you think now pore great mathematician writing didn't write that way at all. In fact, if you pick up a hu paper, it's beautiful narrative and he leads you into the ideas as a human being.
He recognizes not that you need to be entertained, but how the mind actually works. So we have a history which is actually illuminating on exactly this topic. No one really writes bull barking like texts anymore. They're, they're, they're anti-human. They're, they're fascinating as, as artifacts in a misguided ambition to turn humans into autometer.
That's my view. I, I'm sure people listening to this will, but death threats in response to this in binary
Adaam: makes me think, and maybe that's a segue for, for us to return to the, the points of rage. Um, Actually, I, I find that a culprit or, uh, and a better in the decline of the humanities is the humanities.
Yeah. And I think that a lot of the production, I, I'm using this Marxist terminology, a lot of the work that come out of, um, my field of history and sociology and, uh, political science tend to be, first of all, unreadable, but also completely, to use your phrase, anti-human. Um, and I, I think guided by not much more than the institutional corruption of play playing terminological games that don't serve neither soul nor social results, the social results that they purport to be, um, advancing, which is in itself a problem.
But
I want to hear your thoughts about the, um, like a bird's eye view of the, of American Academy.
David Krakauer: Oh, no. You're dossing the wrong person. And so I sort of escaped from it. You know, I, I'm an, I'm an exile from the academy,
Adaam: which is why it's interesting.
David Krakauer: Yeah. So that's why I'm unreliable. An unreliable witness, I will say a few things about your, I'll answer your question, but I want address your issue there of the, um, turn towards theory in the humanities and a part of me when I read or did in the past Deida or Laan or Fuko and others, um, I read them as stylists.
Yeah. You know, I read them as very creative stylists with incredible ideas, not unlike James Joyce, you know? For sure. And what happened, perhaps, and I'm not the, the history
of Madness is one of my favorite works of history in the 20th
century. Right? Know, it's a, it's a wonderful book and I. What happens is that kind of humanistic inquiry that has the same relative descriptive properties of novels and poems, um, started to pose as orthodox theory at the end of the spectrum, typically occupied by physics.
It became an absolute right and I don't think that's necessarily the fault of the creatives of those frameworks. Maybe it is in part with lacan and I don't know. I think when they become orthodoxies they become dangerous and especially when they have i some kind of, um, confusion of identity.
Adaam: And like you said, that when the pretense is we are doing this because this is our formal Exactly.
Way of inquiry.
David Krakauer: Exactly. As opposed that the formal way of in inquiry has been around for a long time, which is close reading. But, you know, I'm, but that's, that's what I tend to like, so I don't want diminish that work. Cause to your point, you know's histories of, of power and, and sexuality and punishment are deep, really deep works.
Um,
Adaam: but, but, but before you get to answer my question more broadly about the academia, I think that the reason it became formal had some sense to it. The idea was we can't accept languages transparent because this is the main tool that we're using. It was kind of like our conversation. You need to be very cautious about the word you use and the meanings they carry.
So sometimes even using grammar to destabilize the reader and make him, uh, uh, approach language in a more inquisitive way is important. You don't want a complacent reader, like you said, that's the, the, it's, it's a more active attempt at close reading. But I think that where we are now is not only far away from that, but we are no longer even gesturing towards that as the reason and the truth is, uh, institutional corruption and the understanding that this is a style that that is used.
I as a, as whether knowingly or not what rather wittingly or not as the shibas of institutional power.
David Krakauer: Ah, God, you're using all these terms and we don't, do we agree on them meaning power. So I think that I will say status, no, I'll, uh, so I don't disagree. And
just to make it slightly less, um, confrontational, um, not you with me, but you with the academy and me with the academy actually.
Um, it has something to do with concepts we don't like to use very often, like delight and joy. And I think that the reason why I love science and literature is because I derive huge amounts of joy from the pleasure. These kind of loose, complex concepts, right? Which refer to a, a lot of memory somewhere in some register.
Uh, and
what I've often found aesthetically off-putting apart from, as you say, the kind of power play of conformity, um, is this lack of joy and playfulness. And
when I'm around good artists and scientists, I mean, that's what I feel. They're, they're kind of, they're riffing, they're, they're improvising. It's like being in a jazz quartet.
They're sort of playfulness within discipline. So I'm completely accepting of any of these kinds of ideas if they're presented in that sort of joyous way where, here's what, here's what makes me happy. Tell me what makes you happy. And I think that this is a part life that's not sufficiently appreciated or certainly academic life.
Adaam: Absolutely. I mean, If, if anti-human is, the thing we're trying to avoid here is this, or at least the thing we're trying to be wary of, a way to interpret pro humanity or active humanity is a a, a simple heuristic for it at least is recognizing the exchange of delight or joy in the process. Process.
David Krakauer: Exactly. No, exactly. And I, that's why we play instruments. It's why we play sports, and the reason we stop is when we have a coach who is exercising despotic power. Right. It's sort of, we don't like it very much, you know? And unless people are, that misery is offset by a proportional salary, you know, which somehow sometimes pays for pain.
I think that would be my test. And that's, that's true by the way, as humanities, people give the humanities a hard time, but these Orthodox are all over the sciences as well, and. You go into departments and it's, it's like entering into a morgue of dead ideas and people defending history that doesn't need to be defended.
So I think there's a more general critique to be made of the academy. So that takes us to that point. So I think it's lost a sense of its own purpose, and I think that's because it's been totally infiltrated by the market, quite frankly. I mean, that's my view. I think that you're attracting students to make money through tuition.
You are building your, you know, bigger and bigger sports facilities. They've got nothing against exercise, you know, but you know what I mean. And, um, faculty are being relied on to bring huge sums of money in through grants and indirects that are imposed on those grants. So the whole system has a impact.
Dimension to it as opposed to a judgment dimension to it. Impact, he largely measured in terms of money. Right? Unfortunately not exclusively. Um, and I just don't think the creative life and monetary impact maximization are that well aligned. I really don't. And one of the sequelae of this style is excessive specialization because it's a means of minimizing audits, right?
So it's a kind of uncertainty reduction synonymous with risk minimization, and it leads to the state we're in today. You know, it's a complex system, right? It's, it's there. There are these market forces, there's human psychology, there are. Strategies for maximizing payoff, which typically require specialization and risk aversion.
So all of these things are leading to this repulsion away from the the academy.
Adaam: And do you think that this problem exists broadly in the, in the academic world in the sense of its ability to apply judgment and sift through knowledge accumulated the, say, peer review magazines? Is that, is that part of the problem?
Problem
David Krakauer: peer review is one of those impact metric obsessed constraints on creative work. No doubt. And everyone would tell you that no one would deny it for a second. This is no less true for working in private companies. I mean, if you work for Google, you feel exactly the same way. I think the difference is in, in a sense, in terms of what you're trying to produce.
And, um, I mean, Google is in the business of producing marketable products that make a profit. I mean, that, that's what they do. We are not, you know, we are in the business of generating ideas that annoy you because they're at odds with orthodoxy. That is the mission of scholarship to defy conventional wisdom.
We're irritants in the system. Now, it doesn't mean we're nasty people. It doesn't mean we're pish, but an ill tempered. But the job of generators, of ideas is, is not to replicate the ideas that came before them. So, so by definition you're gonna be annoying somebody who's defending that idea. Do
you
Adaam: have any encounter, personal encounter with instances where you've.
Where an interaction that you have either witnessed or have known about constituted a form of, of suppression due to these bad incentives?
David Krakauer: Have I ever witnessed it? Yes, infinitely all the time.
Adaam: Can, can you gimme an example of, uh, of somebody trying to explore a, a new avenue, but you know, then it's like, you know, this is not gonna, this is not gonna work.
David Krakauer: We can't, well, I mean, honestly, I mean, it's sort of my life and my job as a director of an institute. I mean, serving on grant committees. I mean, they are, this kind of exchange is, is very frequent. Um, and you have to guard against it in yourself and in others. I mean, it's, it's in every selection process basically.
This tendency is, Present. When Darwin and Alfred Russell Wallace presented the theory of natural selection for the first time in London, they neither was present, by the way.
Adaam: Wallace. Wallace was the guy who independently Yes. Came with the theory and in fact prompted Darwin to publish. Yes. Cause fuck, he's gonna beat me to it.
Yes.
David Krakauer: And, and, and exactly. And, and Wallace at the time was, was in the Malay Archipelago and Darwin was at home cause he didn't go to London anymore cause he was grumpy about going into the city. And the, the paper was read, I think it was, was kinda malinger, I think. Yeah. I think the paper was read by Charles Lyle, the great geologist, someone like that.
And in the minutes, so this was the first presentation, right. Of this idea. And in the minutes you have to look for any reference to this paper being delivered because, you know, professor. Sive showed us how delightful the hind legs of the red and jar, that sort of thing. And then, so you're looking for this revolutionary idea, and when you get to the revolutionary idea, it says, oh and such and such.
Read the works of, you know, Wallace and Darwin and something like, and everything. Knew that they presented was wrong, um, and everything right. They presented. We already knew. Huh. And so, you know,
Adaam: institutional thinking, right?
Vanessa: Wrap head around exactly how your, your institute is attempting to be the irritants the world needs. Um, and I'm just wondering if you can give me an example. It's clear. Clear. It's very clear in my mind to say, like, I, I see the failures of academia. I see how institutional thinking is just replicating.
There's very little. Innovative thinking or, or out of the box thinking. I also can have a basic intuitive sense that bringing cross disciplinary people into a room would probably have some sort of magic or spark, but it's, it's harder for me to understand what is it exactly? Could you give, maybe perhaps you could give me an example of something your institute has produced that because of the, the interactions, the, the sparks that were created, we were able to put something forth in the world that could spur some sort of change in our thinking Impact
Adaam: notwithstanding,
David Krakauer: no.
Right. Let me give you two examples from the early history, which cause it, the dust is settled. So one would be the field of artificial life. If you look at origin of life research up until seventies and eighties, it's completely dominated by biochemists, and the idea was to recreate in the laboratory the chemical reducing atmosphere of the.
Under those conditions, could you produce enough carbon, hydrogen, oxygen, nitrogen, phosphates, or simple amino acids? That would given us in an insight into how the earliest forms of life came about. Okay? But very early people were saying, all you're trying to do is recreate life on earth under conditions that you're conjecturing.
Maybe life on earth is not life. Maybe life is a more general concept. And there was an instance which evolved on earth, maybe we could do this on computers, maybe we could explore artificial forms of life in simulation. And this work had started with people like John Von Neuman when he was at Princeton, and then Esfi, people like Chris Langton.
This generated a huge amount of extraordinary dismissive. Response in the established origin of life community. I mean, extraordinary. Um, this is rubbish. You don't know anything about life, you know, because
Vanessa: it was inorganic and there
David Krakauer: was an assumption. Cause it was inorganic, it was based in code. It wasn't being done in laboratory.
You didn't have to be a biochemist to do it. It was an alternative approach to life. And it generated, you know, uh, a backlash. Now, probably they had some very good points that it was idealistic in a certain way. But, but you know, who cares? I mean, that's the point. You try to do it right, it's now, and that field sort of died for a period of time, and now it's coming back to life actually a little bit the way neural networks died, and it came back to life with deep neural networks that have now culminated in things like, you know, GBT three.
Um, I'll give you another example. Agent-based modeling economic theory, neoclassical economic theory. Um, we have our own podcast on this. If listeners want to listen, listen to that one, because I'm not gonna repeat all this. But, you know, neoclassical theory makes a set of assumptions, which basically come from mathematics and, um, and the need to arrive at some analytical tractability.
So, in other words, the expediency of solving equations is more important than the encoding of reality, economic reality. And so a group of researchers asked, I said, well, maybe we shouldn't make those assumptions. Maybe not everything is at equilibrium. Not everyone has perfect knowledge, et cetera, et cetera.
Let's write a different kind of computational economics and that giver rise to agent-based models, which now that's a success story. So this is the opposite. It was so self-evidently required. That, right? That you give up some analytical tractability, you move towards an understanding of algorithms, but much more parameter rich and therefore less causally intuitive.
So there are trade-offs all the way along, but that was another one. And I, myself found a lot of those early models, deeply frustrating. They look like video games, you know, really is this really serious science and, and it's proven to be serious science. But those are just two stories of SFI experiences, one of which was killed by the opposition and one of which one.
But let me just make one more point, which I think is very underappreciated, including by my own community, um, which is if you look at some of the foundational papers in our field by people like Herb Simon, like John Holland, genetic algorithms and other area, uh, like people like Ed Lorenz. The theory of deterministic chaos.
People like Benmar Mandel, bro, in the development of fractals, all of these people's careers were aligned with the emergence of the digital computer. Um, one way to talk about complexity science is it is that field whose origin story is inseparable. Mm. The origin story of the digital computer we're indebted to it.
The founding of SFI in 1984 is coincident with the apple Macintosh, right? And which is symbolic of the move of computing power outta national labs and computing centers into people's homes. And much could be said about this is fascinating, but you know, when Ed Lorenz was working at m i t on meteorology, He heard about John Von Neuman developing numerical meteorology on his own computer, the i a s computer, the maniac.
And that was the thing that gave him this, um, epiphany that you could do short term forecasting by simulation, not by statistics. And, you know, Mandel bro who invented co-invented with Taylor the fractal, he was working at ibm, he, you know, so he had access to things like the IBM system 360 and its prequels, um, before anyone had those kinds of resources.
And it was through playing visual games with simple iterated dynamical systems that he first developed an intuition for self-similar geometry. So that's a really fascinating history and I think it's one of the things that distinguishes us a bit from physics and English and history who don't have that.
Kinda intimate relationship with digital computing and actually aligns us with the modern machine learning movement who do
Adaam: can, can you talk about machine learning? Um, so for starters, one of the problems when you're depending on machine learning is that the, the, the, the internal systems are, are just completely opaque to you as a, as a user.
You get, uh, um, outputs from the inputs that you plug in and they're instructive and predictive. But whatever logic governs these results, you are still completely blind to. Whereas when you have a theory, the theory as a simplifying effect, which the benefit of which is, um, credibility, but the downside is that it tends to bend reality, uh, on the margins if not further down.
Right. That's basically the dichotomy.
David Krakauer: Yes. So I mean, a good example, a history of this, um, which I like to give is the history of. Astronomy cosmology.
So ever since Aristotle, the world was committed to the idea that the most perfect shape is the circle, and therefore, in a God created universe, all the orbits of the planets need be circular. This is a, this is not an empirical statement. This was a kind of, um, theological geometry and observations were made by people like Tycho Brahe 2000 years after this proposition had been asserted, and it still seemed as if it was acceptable that orbits were circular, right? More observations were made until eventually Kepler said, they're not circular, they're elliptical and conjectured, his famous laws. But these were what we would call phenomenological laws. He didn't know where they came from, that came from Newton, and the ellipse is the solution to. The fact that gravitation operates by the inverse square law. Okay, so long story, ideology, observation, anomaly theory.
Now, if we had given someone like Tycho Brahe, a supercomputer or like server farms of GPUs, he, we would've stayed with circular orbits because we know from. The so-called furrier series that you can approximate any orbit if you have enough circles. That's what we mean by things like epicycles,
Adaam: right? It just adds on more epicycles.
David Krakauer: You add more deference in epicycles and you, you can always fit it. Cumbersome, but you can fit it, right.
Adaam: So in theory, uh, if you have enough processing power, there is no need for you to leave the Ptelomic system and you can just, just expand epicycles until you get more in line with observation.
David Krakauer: Correct.
Adaam: Right.
David Krakauer: So, so under a alternative history, kind of a steampunk history of astronomy and cosmology, the Tychonic system would still be with us and we would be doing perfectly well right? Now. Fortunately, we were too stupid at the time or didn't have the computational resources to do that, and our own cognitive limitations forced us. To go down a different path.
And the question for us now, I think is,
Adaam: which is, which is a great point for people who, who, who were not familiar with that history, is that one, the original commitment to Epicycle was based on beauty and divine symmetry. Wasn't because there was any commitment to scientific truth as we understand it today, but rather that the, that the symmetry and beauty itself was beginning to collapse and, and, uh, the elliptical, uh, solution was more elegant.
David Krakauer: Right. And, and I think that, yes. So it's always this complicated mixture of biases and preferences, symmetries and empirical reality. It's always some complicated. It always is. It's true in economics, it's true in every field, right? Um, but the reason I give that example is because you look at current, you mentioned the opacity of. Large models is
the open question for us intellectually is, are we missing Newton? Right? Because at that point in history, right, our limitations forced us to find a simpler solution, which was more ingenious, right? It was simpler, but cleverer in a certain way. And does computing power kinda shut us off from other more sophisticated ways of being in the world?
I mean, for example, right? Let's imagine that you had a super vacuum cleaner of nano bots in your home. So you could make a total mess every day and you'd never see it. There's something aesthetically repulsive about that idea, isn't there? Because you think you'd never really learn how to live in a kinda harmony with your environment Cause you wouldn't have to.
So there's this weird fact, I think it's more general than science that. Effort. Right? And limitations and constraints force us down virtuous paths or imaginative paths that having infinite power wouldn't. I mean, it's like being a corrupt billionaire. You right? You can just exploit everyone else forever or a monarch.
So I think there is this deeper principle in here, which is both social and scientific and humanistic about the creative value of constraint. That is everywhere, right? I mean, that's one of those principles that I would like to think about more in terms of complexity science, actually.
Adaam: So make use of these big models, but beware of what you're missing out.
David Krakauer: Yeah, beware. Yes. Because
what was the film? I think it might have been Wall-E, was it Wall-E? When they're all up in that spaceship. In their little chairs drinking huge amounts of fizzy drinks. Right. And that technology is enabling a certain lifestyle, which we all recognize is rather undesirable. And,
Adaam: and you're, you're imagining this almost as a metaphor to the, to the intellectual impact.
David Krakauer: Yes. Of Yes. Yeah. Yes. It's partly why I'm very interested in slide rules. I'm very interested in the abacus, I'm interested in ancient calculating machines. Cause they bring you a little closer to the essence of the problem and the nature of one's own limitations. Right. So let me make a more explicit example.
We all learn, learn how to use maps when we're quite young, right? And the, and most of us become obsessed with maps at some point in our lives. I dunno when, but you know, maps on our walls and remembering where every capital is and so forth. And I think. There's something incredibly generative about that constraint.
And if we move to a world of entirely of GPS and Google Maps, you would definitely lose some insight into the geometry of the world. You'd gain something else practically, no doubt. And I use them all the time. I'm not a Luddite, but it's further away from the spatial reckoning of your brain. And I kind of want to feel my limit.
I think it's like people who do sports seriously, like surfing you. You wanna feel how bad you are and say, oh my God, that wave is terrifying. You know? And I like this sensation of being close to my cognitive and physical limits. And I think a lot of these technologies are denying us that sensuous experience somehow, I dunno what to call it, but um, which is another one of those anti-human vices that we've invented.
And that's, you, you call these artifacts at some point, right? Cognitive artifacts.
Yeah, cognitive artifacts, yeah.
Adaam: Uh, which I love the idea because to me it, it resonates as a, as a pianist. I, my friends, when I play music with my friends who are guitar players, would have a very easy time transposing any, any piece of music, any, or not any piece of music, but definitely the simpler rock songs that we'd play together.
And what about me
Vanessa: as a vocalist, though? Way
Adaam: easy. For me, way easier. But, but, but it's interesting for, for guitar players that they transposing chords, uh, along scale is pretty easy. You just move along the frets. Whereas for a pianist, It requires a little more, um, um, calculation and thinking, well, not only that, but
Vanessa: the guitarist has a capo, which makes it even easier,
Adaam: but there just cheats, um, cheats.
But, but the benefit of the piano is that because it's, it was ordered to, to, to give you an express, a visual expression of a music theory. Basically, once you absorb the visual model of the piano, I can do sophisticated musical theory calculations in my head. Yeah. In ways that my guitar player friends cannot.
Yeah. Cause the, the, they don't understand the logic of necessarily going up to f Fred. But when you think of transposing chords and the relationship, when you think about a third or, or a perfect fifth, you actually think 1, 2, 3, 5. Here's the fifth. Um, and, and, and those relationships, um, in music, Become incur, like get absorbed on a much deeper level?
I think so to me that's an art, one of those artifacts in a way that losing it would be, would be just a tragedy. If I had, if I had a machine that like, I dunno how you can outsource music in the same sense, but, but, but, but, but the tragedy of it would be clear to me. I wonder if you were thinking if there are any other ways in which you think that the, um, reliance on these, um, I dunno if you have a word for these external, like the, the substitute to the artifacts?
Yeah.
David Krakauer: Yeah. I call that general domain embodiment. And um, I call these pianos complimentary cognitive artifacts as opposed to something like gps competitive cognitive artifact. Interesting. And I. A piano is your collaborator. Mm-hmm. To the point. And in fact, we've been, I have a post now, Mel Cullins, and we've been formalizing this.
I started working on these things years ago actually. But, um, I got very interested in memory palaces and blind rubrics, cue solving, um, how people do things like that. And, um, we now have a more general theory of a memory palace called the memory lattice. And a piano is one, a violin is one. Um, as you say, if you ask me to recollect a melody, I can either do it because I have perfect pitch and can sing it back, or I can visualize it on the instrument I play.
And actually, if I sit down, I also play the piano. If I sit down to play an instrument, most of my recollection is visual, not harmonic. When I'm playing, I correct. According to tone, but the recollection is very much motoric and visual. And we've been very interested in how these different, I mean, this is really off course now, but how these different memory systems are integrated so as to allow you to memorize whole symphonies, for example.
And, um, but, but that's a whole well, domain of discussion.
Adaam: So I think for conclusion, I'll just, I, I do want to hear your, your, uh, your thoughts if you have any, if we are becoming in certain areas more, um, mentally lazy or mentally flabby to take the Wall-E imager, how are you, what are you imagining the consequences of our increased reliance on competitive, uh, cognitive artifacts?
David Krakauer: I think it's pretty ob. I mean, there's so many studies now suggesting. How dangerous they are. And it's a very difficult thing because the defense of the situation is, this has always been true. It's sort of what culture is, pencil and paper with this in a way. Now I believe that they're making a mistake when they make that comparison, because that for me is complimentary, not competitive.
It amplifies your reasoning, it doesn't destroy it. So it's a bit sloppy on their part. But nevertheless, uh, it is true that we're all dependent on these things and they give us time to do other things, you know? Um, I don't think we've reached that singularity yet where it's obvious. Right. In other words, when I talk to young students, there's no, I mean, they seem quite brilliant to me, and I'm not convinced that they're not thinking every bit as carefully as I did, or probably thinking more carefully.
Um, But I do think now with these machine learning technologies, there's a real prospect that people will give up on being discerning. And if, if you stop making decisions, you just follow recommendations based on your history, that will be a significant phase transition, if you like, in the style in which humans have reasoned in the past.
Um, and I do think it will make us more controllable by correlation, right? In other words, if you become accustomed to being told what to do in one domain, you're probably gonna be willing to be told what to do in another.
Adaam: I, I, your point about curation, I find, I find fascinating cuz I was in a bookstore, um, and looking for, I think the Great Gatsby.
Couldn't find it. It was a secondhand bookstore. But on the way out, I, I just, I literally stumbled on a, a shelf of books that, that looked kinda interesting and, and I ended up getting, uh, uh, buying an essay that I've never heard of about, um, cliche in, in the cliches in the, uh, third Rike and, and a theory about cliches in the Third Reich, which is totally down my alley.
But I can't imagine an algorithm really figuring out and proposing it to me. Certainly not when I'm searching for the Great Gatsby and, and the, and the joy of that moment, the serendipity of it and, and the sort of intellectual excitement that it gave me was unlike anything that you encountered when you go on Amazon and, and see all the recommendations on the side.
It, it was thrill of discovery and something new immediately made me wanna rush home and read the whole thing. Um, And in contrast, I reme recently I got a work computer from my day job and uh, my account there was completely new. My Google account was completely new, and when I entered YouTube, the algorithm didn't know me at all at that point.
So it threw at me a selection of videos completely. Not random, I suppose, but, but from disparate topics. Feel like there was a sport, there was a cartoon, there was a, a music video. I was like, look at all this diversity. I didn't realize that YouTube is so full of variety. Mm-hmm. I was like, it never occurred to me that YouTube has so much to offer it.
YouTube felt so small in my experience of curation and that suddenly terrified me. It really, like, I,
David Krakauer: but the good news here is that we're ultimate, like I said earlier, we're drawn to complexity aesthetically. Mm-hmm. And that excitement, the, your conveying now right over being treated as a YouTube virgin.
Right. Um, is your complexity centers being rewarded? Like, wow, this diversity center is being rewarded. And I think that's what sees us in the end. I mean, we're talking about a few decades of convergence on homogeneity, on a nervous system. That's been an evolution for millions that's been rewarded for complexity.
And everyone you talk to, however channeled they are, as you say, into this curated monoculture of culture in YouTube will, will derive pleasure from the experience you described. Because between the random and the perfectly regular is this sweet spot that we enjoy. I mean, you know what you wouldn't have enjoyed, I imagine, if it had been the absolute opposite, right?
It was. All over the place. There's probably within YouTube itself, some set of constraints and ensures that's not true. Regardless of you, you're a human. So I think that complexity comes to the rescue in our preference for it. And I mean,
Adaam: I mean, granted, my timescale is much shorter than human evolution.
Mm-hmm. But I, in, in the past 10 years, I've been writing about the, the, that this is why cliches interest me, but the cliches that have permeated public thinking and public discourse, and I, I made an argument, I think seven, eight years ago, that if there's gonna be, uh, a way to breach intellectual silos and, and.
The, the increasingly reductive political part, uh, uh, polarization that we are experiencing discursively, it would be because people will be bored of the cliches that, that infiltrate their sides, and they will, they, they'll just recognize their banality and, and seek out alternatives. And I, I feel like in many cases, I feel like, again, timescale of to seven years is not, um, is not a match for human evolution.
But I feel like the, while the human brain reward is rewarded or the, the, the person who's rewarded for complexity, it also is addictive to complacency. And I'm, I, I am not sure I'm as optimistic as I would've been seven years ago that the former would triumph.
David Krakauer: Well, I think, you know, there's lots of data.
I think we enjoy comp complexity, but we also conserve energy. Huh. I mean, that's what you called complacency. And, um, and it turns out that the modern world has found a way of squaring the circle of maximizing the delivery of complexity while minimizing your energy. And that's through essentially a kind of empathic third person experience.
So watch someone else play basketball is almost as good as playing it. And there's something about the way, actually our nervous system works that makes that an active process, not a purely passive process like reading, right? It's like here we are translating a narrative into some version of our own life, to the extent that sometimes when I'm asked to recollect or an event I, I mistake, you know, I was like, we've all had the experience.
Like I dreamt it, or I did. Was that me or did I dream it or did I read it? The fact that that happens is, is revealing, isn't it? And I think that the problem with these technologies of advanced, say gaming technologies, they really do play to that what was historically and. A trade off, and they've actually eliminated it, right?
Because you can conserve energy while experiencing
Adaam: complexity. Uh, and do you really see, uh, a possible dystopia of increased control through these habits?
David Krakauer: It's already true. I mean, there's no doubt about it. I think that, um, I do believe that there's an argument to be made that every day you wake up and you should say, today I'm going to defy something.
And, and it, you know, it shouldn't be an act of violence, but it should be an act of defiance, defiance of habit. And I think it would be a very interesting exercise, um, because I think, look, but, but you
Adaam: get the illusion. But that's part of the complacency making system that we. Indulgent give us a sense of defiance where it's unearned, but it's reshifted people's idea of what defiance means and getting angry at somebody on Twitter for saying something bad could, could, gets translated as I've, I've, I've, I I, I just ate my vegetables for, uh, social activism and defiance for today.
Whereas self defiance, and this is something that we just talked to with, uh, bill, uh, de um, defiance against Oneself, um, or defiance against One's tribe is a little less encouraged, but, but that distinction is eroded,
David Krakauer: you know? Yeah. That's all true. I mean, I, there's a very famous physicist, John Wheeler and, um, he wrote the standard textbook on Gravity.
It's called Gravity. And he, um, he made this point when he was asked how. He thinks, he said, I'm just counterfactual to a fault. Right? When I go to a meeting, I hear a talk. I say, what would the implications be if that had a negative sign in front of it or what if it wasn't three? It was nine, what have you.
And that's the kind of defiance I'm interested in. Um, not acts of outrage as you say, which are kinda, uh, but yeah, it's a basically, and presumably just lead to despair in the long run, but these kinds of acts of intellectual defiance where you ask if the world could be otherwise in a deep structural sense and, and then pursue in a methodical way that thought experiment to its conclusions.
That's, that's the kinda defiance I'm interested
Adaam: in. I wonder if some of the Twitter warriors would claim that that's what they're doing.
David Krakauer: I would say more power to you.
Adaam: David, thank you so much. As expected, we only got to about a quarter of our intended questions, but we really appreciate you playing ball with us.
Good. Yeah. Thank you
Vanessa: for your time.
David Krakauer: Well, yeah, well, good luck and let me know when it all gets edited down into three seconds.
Adaam: Thank you for listening to Uncertain Things. We are uncertain.substack.com or wherever you get your podcasts. If you wanna support us, please give us a five star review on Apple Podcasts because that makes a world of difference, and also share us
David Krakauer: with your friends and enemies.
Adaam: Until next time, stay sane.