If COVID-19 has made anything obvious to everyone, it might be how the very small can force the transformation of the very large. Disrupt the right place in a network and exponential changes ripple outward: a virus causes a disease that leads to economic shocks and other social impacts that, in turn, re-open urban spaces to nonhuman animals and change the course of evolution.
Adapting to these changes will require a different kind of understanding: one of nonlinear dynamics, feedback loops, extended selves, and the tiered and interwoven ecological and economic systems of our planet. By studying the processes and structures that this change exposes, we’re offered a new way of seeing individuality-in-context…and, perhaps, new mechanisms for aligning individual and public good, the human and the wild.
Welcome to COMPLEXITY, the official podcast of the Santa Fe Institute. I’m your host, Michael Garfield, and each week we’ll bring you with us for far-ranging conversations with our worldwide network of rigorous researchers developing new frameworks to explain the deepest mysteries of the universe.
In Transmission, SFI’s new essay series on COVID-19, our community of scientists shares a myriad of complex systems insights on this unprecedented situation. This special supplementary mini-series with SFI President David Krakauer finds the links between these articles—on everything from evolutionary theory to economics, epistemology to epidemiology—to trace the patterns of a deeper order that, until this year, was largely hidden in plain sight.
If you value our research and communication efforts, please consider making a one-time or recurring monthly donation at santafe.edu/podcastgive … and/or consider rating and reviewing us at Apple Podcasts. Thank you for listening!
Further Reading:
Chris Kempes and Geoffrey West on understanding cities to respond to pandemics
Eric Maskin on mechanism design for the market
Pamela Yeh and Ian MacGregor-Fors on studying wildlife in empty cities
Sidney Redner on exponential growth processes
David Wolpert on SARS-CoV-2 and Landauer's bound
What is an individual? Information Theory may provide an answer
Visit our website for more information or to support our science and communication efforts.
Join our Facebook discussion group to meet like minds and talk about each episode.
Podcast Theme Music by Mitch Mignano.
Follow us on social media:
Twitter • YouTube • Facebook • Instagram • LinkedIn
Mentioned in this episode:
David Wolpert, Alan Turing, Rolf Landauer, Timothy Morton, Buckminster Fuller, Sidney Redner, Chris Kempes, Geoffrey West, Bill Gates, Ann Pendleton-Jullian, Luis Bettencourt, Cris Moore, Eric Maskin, Wendy Carlin, Sam Bowles, Kenneth Arrow, John Von Neumann, Eric Morgenstern, John Nash, Pamela Yeh, Ian MacGregor-Fors, Alan Weisman, Doug Erwin
Michael Garfield: All right, well, this week we have a set of really sort of fundamental and potent contributions to the series and I look forward to diving in with you here.
David Krakauer: It's a rich one.
Michael Garfield: So as we have come to kind of formalize this, let's start in the micro with David Wolpert’s piece on the Landauer bound. This was a really cool one.
David Krakauer: Yeah, this is really interesting, it's a field that I’ve found fascinating, and of course, it’s of big interest here at the Santa Fe Institute. I think of David's contribution is the ‘virus as a computer virus,’ and it's all about the energetics of computation. It introduces energetic limits to computation, how they're connected to this idea of the extended phenotype and the virus as a minimal replicator that outsources its functions to its host. And it’s worth, in this case, just do a little history:
When the great Alan Turing invented his famous thought experiment that led to the computer, called the Turing machine, to prove an undecidability theorem in 1936, the computer was a completely idealized logical construct. It didn't live in the physical world. It was a mathematical thought experiment. And he devised that thought experiment to answer a question, that is, can an algorithm tell you in advance whether or not a problem can be solved? The answer, famously, was no. This construct that Turing invented, called the Turing machine, has an infinite tape and you write binary strings to the tape and you can erase these strings, but there's no friction in the Turing tape, right? The Turing computer doesn't live in the physical world, it lives in a kind of mathematical utopia. And a physicist working at IBM in the 1960s became very interested with this fact that most real computers do live in the physical world. That means that they're subject to all the laws of physics, like the laws of thermodynamics. In this case, the most important of which is the second law: that entropy increases in any closed system. And he formulated the second law for computation, called the Landauer Principle, and it basically says that any logical manipulation of information of the sort that Turing envisaged –– writing bits, raising bits, combining bits –– will generate heat, and that heat will be dissipated into the environment. This was important, because it led to limits to militarization because as we pack more and more transistors in integrated circuits, they generate more and more heat and at some point the circuit as a whole will melt. So, a virus is a little bit like a very tiny computer. But the virus is an organic computer and its Turing tape, if you like, is its genome, the sequence of RNA bases.
And what the virus computes is the function “copy.” It gets into the cell and it copies its RNA genome. And what David points out is the virus, like any computers, should be subject to the Landauer constraints, which would put a limit on how much information it can encode. What the virus ingeniously does is outsource almost all the computational work to the host. And that's where this extended phenotype concept comes in –– sometimes also called ecosystem engineering, or niche construction –– where the virus basically allows us to in some sense do the bulk of the work. So, the virus doesn't encode all the ribosome complex…we do, right? We encode the Turing tape head. And, by virtue of outsourcing to its host, most of the computational work, the virus gets away with being an extraordinarily efficient little bit of computer code.
That's why I alluded at the beginning to David's contribution of a virus as a computer virus. Which leads to this very counterintuitive fact that most of the virus is not virus. Most of the virus is, in fact us, because we contribute most of the machinery necessary for it to complete its life cycle. One way to think about this is like technology. So, we talk about iPhones as these miracles of miniaturization, but that's only partly true because most of what the iPhone does is actually outsource to vast server farms that are cooled by river systems. So the phone is a bit like a technological virus that heats up the world in order that you can keep this cool piece of glass in your pocket.
Timothy Morton, the philosopher of ecology calls it a hyper-object. It's an object that only partly lives in the dimension of your immediate experience, most of which is out of sight. And as computing trends move towards outsourcing computation, they're becoming more and more viral. So there's this beautiful physics of computation link that David points out between the microbial world and the macroscopic world of technology.
Michael Garfield: You know, this brings us to this overarching theme in the series about the problems of bringing over-simple models to complex situations. When we had you on the show way back at its beginning, for the live recording at InterPlanetary Fest last year on the origins of life course offered by Complexity Explorer, someone in the audience asked the classic “is a virus alive” question. Well, that's a binary question for which there is not a binary answer, right? I mean you just lead-authored a piece on an information theory of individuality where this kind of question you and your coauthors…I really enjoyed this piece, the way that you talk about different kinds of individuals that are scaffolded in different ways by the environment, and how we tend to think of the individual as more or less only the kind of individual that humans believe ourselves to be. That we have a sort of internal driver and informational continuity that pushes us forward in time. It's this classic nature vs. nurture question.
And then, with the Landauer bound and the extended phenotype, bringing this subtlety that you addressed in that paper: where does the spider end? The web allows the spider to be small and to outboard most of that sensitivity to the web, which is really light and comparably easier to build than an enormous spider body that would have to sit there and sense the whole thing.
David Krakauer: Yeah, that's absolutely right. I think that our perceptual apparatus operates on preferred space and timescales, and doesn't see the spider for what it really is, which is, as you say, is spider plus web plus… Truly, individuals are these much more interesting hyper-objects which extend into the world in complicated ways. Human beings are their genomes, their bodies, their social networks, their homes, their cities… And the key question of course, in that work was trying to determine where you should draw a boundary in a principled way. It has something to do with causality which is, where does your causality, the center of causality extend out and attenuate to the point where you have no more control? So yes, I think that there needs to be a real revolution in how we think about what objects are, particularly adaptive objects. Partly for conservation reasons, because, if you believe that your cell phone is just that thing that you carry that you can charge every night when you go to bed with a regular charger; and you forget that part of that phone is actually somewhere hidden away drawing down huge quantities of energy that are radiating out heat, then there's a sense in which you're not truly responsible for your extended phenotype. So, there are ethical implications of rethinking what the individual is.
Michael Garfield: Yeah. And in that movement towards a deeper apprehension of our identity as hyper-objects, outboarding the cost of viral reproduction allows the very small to drive the very large. I think a lot of listeners are probably familiar with Bucky Fuller's analogy of the trim tab on the rudder of the Queen Mary, and how that tiny little thing can push a much larger craft, and steer it in a different direction. That brings us to Sid Redner’s piece on exponentials, and how important it is to embrace the enormous range in a lot of these epidemiological models, because they're inherent in the nature of grappling with an exponential phenomenon.
David Krakauer: So the consequence of all of that efficiency that David described, whereby the virus outsources its essential computational machinery to us is the possibility of exponentials.
So, we pay the entropic cost for viral overpopulation, not the virus. And exponentials are notoriously hard to get your brain around. In popular culture, they're behind the so-called Butterfly Effect in chaotic dynamics. That's where you have two arbitrarily close initial conditions, where the two associated trajectories diverge exponentially in time. So, exponentials are what we are talking about when we say extreme sensitivity to initial conditions. That's the Butterfly Effect. Sid gives this example from compound interest, which is the exponential that most of us know from our bank accounts. And he makes his point. You deposit a dollar with an outrageous interest rate, say 5% per day, and in about nine months you'd have a million dollars, right? When he says it’s a 10% interest rate, you'd be a millionaire in just under five months. That's exponentials. The one that I like, because I think it's even more crazy to come to terms with, is the consequence of Moore's law.
Folks will remember Moore's law is this empirical observation that the number of transistors and so forth in an integrated circuit doubles approximately every two years. And this is a very general technological observation. Take this one: In the 1960s, a gigaflop of processing power –– a gigaflop is about 1000 floating point operations per second –– cost $18 billion. In the 1980s that costs 20 million. In the year 2000, you get a gigaflop for about a $1000, and iPhone X, which costs about $500 on eBay, runs at 600 gigaflops. Now here's the weird thing about exponentials. That means that if you could time travel just to the 1960s with your $500 iPhone, it would be worth hundreds, if not thousands, of billions of dollars. In other words, your $500 iPhone is a trillion-dollar hyper-object.
Michael Garfield: Well, I mean that's, if you could travel back with the entire network architecture that supports it, right…
David Krakauer: Yeah, that's a very good point! That's a very good point, but to be fair, the actual chip on the phone is running at 600 gigaflops. It wouldn't be very useful; you couldn't do much with it. But that little device contains, you know, by 1960’s standards, trillions of dollars of technology. So that's what exponentials do.
Okay, let's get back to Sid. Sid is making the point that R-naught, which we've talked about a lot, is like the interest rate of the viral investment in its hosts. So we are the virus’s genome bank and the virus does not make more money, it makes more genomes. That's its profit. So if you start with one infection, COVID has an R-naught of about 2.5, that's a 250% interest rate. That's pretty high. At those rates you not only get huge returns, but you could potentially get huge variations in return. And this takes us to the Cris Moore insight about long tails. It's a very similar style of reasoning in this article. So, here's a simple mathematical model where you basically say, let's imagine that our social distancing policies…we don't have to imagine it, it's true…are aimed at reducing our not below one. Of course, you can't do that overnight, policies aren't perfect. So he says, let's assume that every day you can reduce from 2.5 in some range between 10% and 0%. So, on average the daily decrease would be about 5%. That means you're expected to drop beneath 1 in about 20 days. But of course we've already established that that's the average, between 10% and 0%, so there'll be variation, and because this is an exponential, this leads to massive differences in outbreak sizes.
So, even though you could reduce it in about 20 days, let’s say on average, the size of the epidemic could vary between a hundred times the initial number of infected to about 10,000 times the initial number of infected. This is again, like Moore's law, one of those very counterintuitive outcomes that relates to the uncertainty. Not the uncertainty of ignorance, but the uncertainty of the Butterfly Effect. Exponentials build this other kind of radical uncertainty into our models, which has nothing to do with ignorance, or a little bit to do with it, but moreover to do with the non-linearities that can massively amplify tiny measurement errors, or tiny differences in initial conditions.
Michael Garfield: That calls pretty directly to, if people want to go deeper into that, Michael Hochberg's piece that we discussed in the second episode of the series on the importance of timing. And then more largely this cuts across articles we've discussed by Lenny Smith, by Jurgen Jöst, Simon Dedeo, Melanie Mitchell and this issue we've addressed here on embracing radical uncertainty. I really liked Redner’s point that in some ways, whether you saturate the system or you saturate it five times over, is kind of beside the point. That the initial investment in some ways matters less than the exponent and when you start investing, So, at what point do we accept this and act on it? This is linked to Chris Kempes’s and Geoff West's piece, in that we live in a world now where exponentials are not only difficult for a lot of people to grasp, but also much like hyper objects, they’re a central feature, a dominant feature of our world, because we live in cities.
David Krakauer: Yes. So this has been, of course, the subject of Geoffrey's work for a long time, and this is a very nice contribution from Chris and Geoff. You know, it was not until about around 2007 when for the first time in history more human beings lived in cities than in rural populations. The history of humanity by and large has been a rural history, and our institutions and our intuitions come from that world. But the point that Chris and Geoff are making is that cities are kind of time-travel machines. They speed everything up. They increase the rate parameters of these exponentials for all sorts of different processes. So, let's get to the virus. It's been observed that over two thirds of all COVID-19 cases in the U.S. comes from about eight jurisdictions, and all of those are concentrated around cities.
Transmission spread and potential mitigation are all tied to an understanding of the dynamics of cities. What Geoff has shown in the past about cities are very interesting regularities, the best known of which is that socioeconomic quantities (wages, patents, wealth) all scale super-linearly in population size, which means that the wealth scales as the [size of the city] raised to a [power], and that power tends to be 1.15. So you get a compounding effect of 15% per doubling. There's a direct connection to Sid here, if you like, with an exponential process this exponential in time, then the rate parameter will vary with the [city size] raised to this [power.] What that means is that a city with a million people will double the number of cases in half the time a city of 10,000. That's like the Moore's law, there’s this huge exponential effect of being larger.
And the neat thing about this work that I find really intriguing, and I have to say when I first saw it of course many years ago, very surprising, is that it's not just things like patents and wealth that scale to this power, but also the rate of disease transmission. So, we've talked earlier in this series about the connections between memetics and genetics, but cities are a very interesting machine because they impose through mechanisms that we still don't fully understand this extraordinary universality on the way very different transmission dynamics work. In the city, disease transmission scales the way wealth scales and those are really amazing results that we have to understand.
The reason that's important, I think for culture, is that there's been so much emphasis in this period on understanding epidemiology, understanding virology…The Gates Foundation that's been extraordinary in its funding of disease has declared that it will dedicate all of its funds towards COVID, and that's a huge problem because it's one factor in this multidimensional hyper object, if you like, that we call “the city.” To understand these kinds of complex events means not disaggregating them and treating epidemics as if they're independent of the economy, which they’re clearly not. The city is a super integrator of complex phenomenology where you can take nothing for granted. So I think what Chris and Geoff are really arguing for here is that in just the same way that we have surveillance systems and prediction systems for weather and for tsunamis and for earthquakes, we need to think very carefully about having very thoughtful, rigorous surveillance, and theories of surveillance, for the integrated set of phenomena that all adhere to the same scaling laws. Because the city, as I said before, imposes convergence on their collective dynamics.
Michael Garfield: Last year when she was at InterPlanetary Festival, Ann Pendleton-Julian had a really lovely riff on something that you just touched on, which is that we're not just talking about cities as one network, that sort of one dimensional thing, but it's a multidimensional thing that includes the infrastructure, the environmental resources, the complex political, economic and social systems, the stories that we tell, one another, the virtuality that we inhabit now that we're distancing from one another physically… A lot of us seem to be engaged in more social instances per day with this proliferation of video calls. Last week when we were discussing understanding this complexity in terms of identifying points of systemic intervention, Luis Bettencourt in episode four of this show talked about his research on intervening in slums by finding ways to bring new vasculature into the slums to allow for better distribution of resources in infrastructural networks. But then, you know, Cris Moore talks about how understanding the hidden structure under the average of R-naught allows us to find points of intervention for lowering the node-degree to slow the transmission of virus. Your point about surveillance needing to integrate across all of these dimensions…it's too simple to say, “let's grow the network to solve problems” or to say, “let's break the network to solve problems.” It's understanding how to modulate that across many different dimensions at once.
David Krakauer: Yes, I would make two points here, Michael. The first is, why have we heard so much from the epidemiologists and so little from the economists? Even though for many people, for the majority of people, COVID is an economic catastrophe, not a health catastrophe. And the reason for that is because we have models and principles of transmission where points of intervention are understood and principles of vaccine development are somewhat understood. Whereas when we talk about economic markets, we don't have comparable insight and we'll get to that with Eric's piece. And so I think one of the deeper points being made here is that we now know that there are these scaling laws that unify biological and cultural phenomenon, and if the underlying dynamics on the fractal infrastructure of the city are what give rise to these universalities, it does suggest the possibility of a very principled understanding of economic response, not just epidemic response. So I view that as the deeper point here, that it's time to take complexity very seriously in order to be as useful in responding to the socioeconomic catastrophe as we have been in responding to the healthcare catastrophe.
Michael Garfield: Indeed. Maskin’s piece is really interesting in light of this question of what happens when we use an insufficiently complex model. Like you were just saying, we look at it across only one dimension. And actually, for Vox EDU, Wendy Carlin and Sam Bowles just published this lovely piece on the revival of the civic society. Rather than thinking of things as merely just “the market” or “the state,” and allowing one or the other of those to drive those things, but to understand how we need to fill this in so that governance is operating at multiple scales at once. I think we can get to this idea that this is a way in which economics is related to evolutionary questions, like what forces lead to different kinds of nervous system architectures. And when is it good to have a head, or not have a head? Maybe that's skipping ahead. Let's talk about what happens when the city breaks, when the market doesn't work fast enough to address the problems that we're facing.
David Krakauer: Yes. Yeah. That's what to the extent that this crisis has shown us, that these systems are coupled. It also shows us how they all break together, and Eric's contribution is what do we do when we don't have markets. Of course, we don't have markets, we do in some sense, but by virtue of social distancing, market efficiency is compromised and the pricing mechanism that mediates supply and demand doesn't work. And Eric is one of the inventors of the solution to this problem and the solution is called Mechanism Design. Eric won the Nobel prize with Hurwicz and Myerson in 2007 for this work. Just a nice footnote on SFI history here, because Eric is a longstanding member of our science board, but Eric himself as an undergraduate stumbled into a course being taught by Ken Arrow on information economics. And Ken, of course, was one of the founders of the Santa Fe Institute who had won the Nobel prize himself in 1972, and he inspired Eric to pursue this work, particularly through his familiarity with the work of Hurwicz.
So I need to explain a little bit where mechanism design comes in by explaining what game theory is. And I apologize cause I'm sure most folks know this, but a little bit of background on this, because it's important. Game theory, perhaps we know, is associated with extraordinary figures like John Von Neumann and Oscar Morgenstern and John Nash. And the essential idea behind game theory is to find the best strategy to use in a game. What’s a game? Well the game has a set of agents –– Think of chess –– us, the players. There are strategy options, and there are payoffs associated with playing those strategies against one another. And most importantly there is a solution concept, that is this choice of strategy that is the best against all other strategy, which gives you the highest payoff that allows you to win the game. And what game theory does is it looks for those solutions. It looks for their accessibility and it looks for their stability.
The most famous of which of course is the Nash equilibrium. Now, mechanism design turns that on its head. Sometimes mechanism design is called “reverse game theory,” because it doesn't look for the solution given the strategy set and the agents. It starts with the solution. It says, “this is where we want to be,” and then it asks how would we go about defining the strategies and the payoffs to ensure that that solution is reached? And if you think about a market, it's made up of strategies: strategies of buyers, strategies of sellers, and the pricing mechanism is a mechanism of the market that ensures under very idealized conditions, of course, that at least for a short while some fixed point –– a Nash solution –– is reached that satisfies the majority of suppliers and consumers…one hopes. So, if the market has gone, what do you do? How does mechanism design work for something as complicated as a market?
And it sounds arcane, but most people interact with mechanism design regularly through auctions like eBay or by voting. And here the objective is, we want to elect the preferred candidate, or we want to maximize the profit of the seller. So the mechanism design in those cases is how you design the voting system, or how you design the auction to achieve the goal. The example that everyone will know, that Eric often gives is imagine, Michael, that we're both super greedy. Now of course you are and I'm not, but let’s, for the sake of a thought experiment, imagine we both are.
Michael Garfield: I think we just failed the prisoner's dilemma…
David Krakauer: …and we are given a cake and you get to cut up the cake. Now imagine that you get to cut the cake up, and you also get to divvy it up. Under those conditions, unfortunately, you'll get a very big chunk of cake and I'll get a tiny piece. That's what's called a “tragedy of the commons,” because your selfish interests don't maximize the social welfare function, the wellbeing of everyone, which in this case is two people. Mechanism Design comes to the rescue. And the mechanism in this case that most people will know is one person cuts the cake, and the other person chooses. If you do that then there's actually no incentive to take more than half, so you get a fair sharing of the cake. This is the key to mechanism design: the key underlying principle is you have a tragedy of the common situation, you want somehow to reach fairness, and the way you do it is by outsourcing the utility function to others. And Eric discusses the more technical case of how we would distribute fairly ventilators to treat the acute condition of COVID infection without a market.
We all saw in the news this spontaneous emergence of aberrant auctions in which state governors were bidding against each other very inefficiently to try and save lives. And the point that Eric's making is we can do better than that! The mathematics of Mechanism Design shows how. This article while on the surface is arithmetically a little bit challenging, it’s actually quite simple: the idea is that buyers and sellers all report their preferred costs to the so-called mechanism, in that case, that would be a federal agency of some kind. And the mechanism creates a payment rule for both parties. The key is just like that divide and choose mechanism of the cake, is to align the individual incentive with the global social welfare function. And so, all mandatory transfers from the mechanism turn each of our personal utilities into social utility.
That's what the mathematics does. It provides an incentive to both buyer and seller to do the right thing. Sometimes this is called the Vickrey-Clarke-Groves Mechanism, and the essential concept is always that one: you turn the individual utility-maximizing strategy into the social utility-maximizing strategy.
There are some simple examples I can give. Economists often talk about externalities. These are the things that are not typically on the spreadsheet, right? In other words, they're the hidden cost to the world of trade and production. So for example, you might be manufacturing bricks, but in the process emitting a huge amount of toxic pollutants. But here's what a mechanism would do. It would say your cost to production are proportional to the damage to the environment. Well, if that was your true cost, then there would be every incentive for you to reduce the pollution and deal with the tragedy of the commons, and achieve the fair outcome. The critical point about mechanism design is it's not despotic. It's not the command economy that says thou shalt produced x levels of y. It says you still choose, you can still make a profit, but we're going to tell you how much you're going to make and incentivize you to do the right thing.
Michael Garfield: To draw back to a talk that you gave at UBS last year, this links again to the Information Theory of Individuality paper, you know? This notion that the stable, predictable environment of the flow of nutrients allows for the rare cases in evolution where an animal might lose its head, where sea squirts and clams and so on, are capable of just fixing themselves to a rock and hanging out there and filtering. But the process of evolution, as we discussed with Brian Arthur on this show back in episode 13, that when you run evolutionary simulations, where you iterate game theoretical situations like the prisoner's dilemma over and over, thousands and thousands of times, you see what looks kind of analogous to an evolutionary trend to grow a head, to grow this kind of mechanism, this regulatory coordination. And so when we think about this in light of Geoff West’s and Chris Kempes’s comments on cities, COVID might be bringing us to a turning point where cities are seen as self-organizing, kind of headless, processes. We're seeing a kind of a historical shift towards more balance between the emergent bottom-up and the regulatory top-down that aligns the incentives of all of the various actors in the emergence of or the steering of a city.
David Krakauer: I think it's a very interesting point, and this is in some sense the Holy Grail of thinking about decentralized architectures. So, there's a lot of ideology here and we talked about it in relation to federalism and anti-federalism. It's talked about a lot in terms of collectivism versus liberalism or you know, libertarianism, and so forth. And what I think complex systems shows us is that these categorical distinctions are bogus. That each of these kinds of solutions evolves out of circumstance, and I think you're absolutely right. Mechanism design is not the reverse game theory in the sense of being totalitarian game theory. It's the right kind of game theory when the market mechanics don't work. There are a list of reasons for when that doesn't work, and we can go into them, they're the basis of economics 101 class, but it's a very natural evolutionary adaptation of the economy when a certain kind of information is not present. I think having a sophisticated take of all the strategies available to us to achieve desirable outcomes is important.
We talked about that in relation to genetic regulatory architectures, and how different species according to the complexity of their genome assume the more autocratic versus more democratic topologies. Exactly the same applies here, and I think what we should all be understanding increasingly is that there isn't one best way to do things. The circumstances tell us which of the strategies we should be pursuing.
Michael Garfield: Yeah. So we've got one more piece here, Pamela Yeh’s and Ian MacGregor-Fors’s piece on animals in cities. And it seems like the question that they're gesturing toward in this piece is whether it's possible for us to apply this kind of reasoning to create mechanisms by which we can grow or adjust the way that we live in cities to align the incentives of humans and non-humans.
David Krakauer: Yeah, I liked this a lot. It reminded me of, I don't even remember the film, Logan's run. Have you ever seen that film? It was directed by Michael Anderson and Michael York and Jennifer Agutter and Peter Ustinov, and that was a film where –– I don't know when it set, sometime in the far distant future where human civilization lives in these geodesic domes, you talked about Buckminster Fuller –– and we're all placed in these small populations run by these horrible computer systems that take care of our lives, including when we get to reproduce. And so, to reduce overpopulation by the age of 30 we're all exterminated in this horribly weird ritual. And the reason I mentioned it is because at the end the characters that are played by Michael York and Jenny Agutter escape outside of their geodesic dome, and find themselves in a Washington of the future, and it's all overgrown with beautiful, lush vegetation and firms and so forth. It's the world that was reclaimed by nature when humans abandoned it, presumably for some folly that that goes on specified. And in 2008, Alan Weisman wrote a beautiful book called The World Without Us, which was a catalog of events like that where for one reason or another buildings, various edifices, cities had been abandoned, and how long it takes for nature to reclaim its world. In that book, in fact, he interviews our colleague Doug Erwin, who contributed one of these transmissions. And so, what Pam and Ian do is report on that real world without us during the COVID crisis, and I think they're making two points. One is, what is going on really? And the second question is what kind of science can we do? What's going on that we should be recording during this natural experiment.
We've all seen, you know, mountain goats in the streets of Wales, and Buffalo walking along highways in Delhi ,and dolphins frolicking in the Bosporus. It was all of this wonderful natural history going on now in urban settings. But there have been longterm evolution experiments, and they cite the particular example of the Galapagos tortoise. This is an animal that's evolved on an Island because of its size, it's essentially free of predators. And Darwin, when he was there in the 1830s says, you know, I met an immense Turpin that took little notice of me. They just don't care. They're sort of indifferent to humans. Pam and Ian and point out that there is research showing that in these very insular ecosystems, you have far tamer animals. That looks like a real effect, and you can ask, how long would it take for that to happen in our world?
There is some work that they cite from 2010 where researchers from Spain looked across six different Galapagos species that have been reported to be somewhat indifferent to humans showing that now, about half of them respond with fear to human tourists. So, it didn't take very long for them to evolve an appropriate fear response. So that's the first part, how long does evolution take? The second part is, is their own work really. And they've been studying dark-eyed juncos: songbirds in Southern California. These are birds prior to the shutdown, had been moving down from the mountains into the cities and are actually quite well adapted to human presence, and have exploited our profligacy very efficiently. I think what they're asking now is what will happen to that species now that humans are no longer there? Will they become more like the Galapagos tortoise, even less concerned about human presence? That could actually be a problem when we come back into the environment and assert our ecological dominance.
So I think the questions they're asking are very deep and very interesting. It's also worth pointing out that there have been negative impacts on animals by people staying indoors. There are many species that have, for good or ill, become dependent on us. I remember reading an article about a macaque species outside of Bangkok that had been fed by humans around monasteries, but now the food has run dry because the humans are no longer out in the world and they're engaging in these huge conflicts over scarcity of resources. That's a frightening fact, too. It's worth bearing in mind the negative consequences of humans retreating. And the larger picture, to your point Michael, is that what the pandemic is teaching us among many, many other things is that we share the planet, and we share it not only with viruses, but numerous other species. And those species are sensitive to our behavior. So, you know, when we return to the world, we kind of mentally also want to return to the wild. And it's an opportunity for us as a culture to evolve. I found that a very uplifting insight into what's going on today.
Michael Garfield: Yeah, you know Andy Dobson has spoken both on the show and elsewhere about the rate at which habitat destruction has increased, the frequency of zoonotic diseases, that more and more of these diseases like COVID-19 are emerging from the wilderness that we have unsettled by settling. And I think about Geoff West's piece on cities as disease incubators, and a piece of the conversation I had with Jennifer Dunn back when she was on the show in episodes five and six, about how the disruption to biodiversity is complex with cities. That empirical work that's been done over the last few years suggest that certain regions of cities actually promote different kinds of biodiversity. There's a change in the philosophy of conservation that is less about a sort of retro romantic restoration of lost biodiversity, but an active cultivation of an ongoing and open process. So, it gets us back again to this looking at the city as a multidimensional and wild process. I liked your turn, that as animals returned to these urban spaces, there's a kind of counterflow which is the rewilding of the human imagination, and the understanding through complex systems science that the city is itself a phenomenon of nature and so how can we design them to minimize the ecological disturbance and, counter-intuitively, to improve their capacity for epidemic proofing?
David Krakauer: Yeah, I always thought that the great existential crisis of modernity accelerated by isolation is loneliness, to feel alone in the world. And what the theory of evolution did for me was connect me to the rest of the world, right? Trees and my cousins. Insects and my cousins. There's a sense of true connectedness that comes from understanding our common evolutionary histories and now we're living in this time where we're fearful of a virus, and one would hope that the resolution of this kind of dialectical standoff would be a greater appreciation of our wild nature and our susceptibility to that world. That we're not isolated from it, we never have been. In learning about the microbiome, we realized that we were dominated by the genetics of a bacterium, that we're a tiny fraction of who we really think we are in terms of our individuality. And so, it would be nice to imagine a world where we can rethink our relation to the natural world.
Michael Garfield: Indeed. On that note, we're going to move this particular series of the show to biweekly, and we're going to take a week and let the city streets remain empty and the soil fallow on this podcast while we prepare for a series of interviews with our Miller Scholars, and return to our sort of regular programming. But this is a great time for those of you listening who have been inspired by these ideas. Complexity Explorer is about to reopen the Origins of Life course on May 18th. If you're sitting at home and looking for a way to stay stimulated and to continue learning, then that's a free course at Origins.ComplexityExplorer.org. Anything else before we sign off for this week, David?
David Krakauer: Well, thank you all for listening and we will continue with an a slightly different schedule. We thought it would be worth now reflecting on what we've learned and thinking about moving into the future, how our ideas should change. And so that the tone will be slightly different, hopefully more optimistic, but also with a view towards this new, wiser world that we hope to be occupying. So thank you, Michael. Thank you all.
Michael Garfield: Thank you. Take care everyone.