Jamie Jones Decelerating Correlated and Skewed Understanding How the Talks at Google

Show video

Hi. Everyone I'm. Jamie. James Jones and, I'm a, professor, in Earth System science at, Stanford I'm, actually trained, as an anthropologist but. I sit, in the school. Of Earth energy, in the environment at Stanford, in the Department of Earth System science so. Today. I'm gonna talk to you about. Some. Recent interests, I've had in, in. Applying evolutionary, and ecological ideas. To economics, and decision theory. This, interest was piqued by my, attempt. To understand, the peculiarities of a human life cycle and our, life cycles are really peculiar I can assure you in a broad comparative, context they're really, kind, of weird. And, to. Understand, the reason people. And subsistence economies, pursue, particular types of. Resources. Prey and, forage. Food and that sort of thing and so. I'm. Gonna guide you through a big-picture review, of some fairly technical fields, today that include population, biology. Decision. Theory a bit of economics, both classical, and, behavioral. But, I think the ultimate message is pretty straightforward, and the thing that sort of broad picture that motivates me is understanding, how. Something. Like this which. Is an. Artist reconstruction, of Australopithecus. Afarensis, which. Is essentially, a 1. Meter tall. Chimpanzee. Like thing that walked upright right, it's got a. 300. Cubic centimeter brain so, about, a quarter, of the size of, maybe, a little less than, contemporary. Human. Very. Sexually, dimorphic the, males are much bigger than females, they. Had enormous. Cheek. Teeth for. Meeting a very tough diet. And, as I said they're, really tiny these guys. Lived. About three and a half million years ago in East Africa. And. Transition. And this probably doesn't look that different I guess but. It is it's fundamentally, different in about one-and-a-half one at 1.8, million years ago and this critter, called. Homo ergaster. For. My purposes, is is effectively, modern, you've got a large, body size so, if within the normal, range of. Contemporary. Human variation, you've got body proportions, that look very much like people, living, in the tropics today you've, got a great, reduction in sexual size dimorphism, so, that males and females are similarly. Proportion, to how they are today relative to each other and you've got big brains they're not they're smaller than modern but they're actually within the. Range. Of variation on, the low-end of variation, of contemporary, humans. Homo. Ergaster. Sometimes, called African. Homo erectus is the. First of our. Family. To leave, Africa, and to colonize, the, rest of the world. And. It, did so around, you, know. 1.8. Million years ago give. Or take a little bit and so, my. Question is, how. Is it that a population of bipedal ape so you know they're not very impressive looking all right kind of like we aren't. We. Don't have big weapons we don't have antlers or horns we don't have big teeth we don't have. This. Sort of ferocity built, into us we, have some cleverness maybe, but. How is it that in less than two million years which is a very rapid, period of time, in the. Broader span of evolution, how is it that we went from this small. Population. Of, African. Apes, bipedal. Apes and have, grown to a size of. Seven, billion souls in, as. I, say less than 1.8. Million years okay. And we've colonized. And. Come to dominate nearly, every terrestrial, biome, just. A picture of making. The desert green and the Arabian Peninsula as, a sort of indicator of the, extent to which we really dominate. The. World 30 to 40 percent of net primary. Productivity is, captured by human actions right so. The phenomenal, growth of our species in this very short period of time, suggests. That our ancestors on average must have made pretty good decisions, right because, the, decisions, they made led. Them to increase. In. In pretty great numbers. But. There's a torrent of work in psychology and economics, represented. By just a few sample, publications, here, of semi, popular trade books suggest. A decision, made the decision making software, and that our brains run which, is presumably, essential, toward our ecological, domination. Is. Profoundly. Flawed that in the in a word were irrational. How. Is it possible that a species apparently, so defective, in its, ability to generate sound decisions, can, be so incredibly, evolutionary. Evolutionarily. Successful. I'll. Just say as a relative outsider. I'll. Make, this observation about this particular body of work that I'm discussing here, in, this slide there, are lots of anomalies, like we see, these, things that don't make sense from, the way we we sort of think about the way people should make decisions, and. And, you. Can write whole books on this like lots of books on this but, they're generally unconnected, and the whole area seems to lack a sort of a unifying theory, I'm.

Going To humbly, submit that I think that evolution can help us, provide. Some of the theory that brings together some of these anomalies. There's. A tendency in particular in, like in the TED talks where there's real selection, for being provocative, right, to, suggest that people are irrational because they don't follow the rules of formal, or, axiomatic, rationality, theory however, every economic decision entails an optimization, problem and, and I feel like I'm on safe ground when I'm at Google talking, about optimization, problems, right. That and if, you want to solve an optimization problem what do you need well you need to know who's doing the optimizing, you need to know the actor but you need to know the objective, function, right and, if, you get the objective, function wrong you're. Gonna get the solution, wrong okay, and, it. Doesn't matter what. Your optimization. Problem is right it could be an engineering problem it, could be a portfolio. Design, and investment portfolio design it could be about ensure getting. The best mixture of wild foods to ensure growth and survival right, but you need to get the objective, function right and I'm, gonna suggest that in, general, we, don't have, a, very. Solid understanding of what the objective, function is that people are, using when they're making economic, decisions, broadly. Okay. And the results appear anomalous. When you get when you use the wrong objective, function you get the wrong answer and, I think the devolution helps, weave together theoretically. What are otherwise a series, of unconnected, anomalies. So. We, like this coherence. And. I think these may be brought together if we understand what sort of problems, the human brain is actually designed to solve okay. Get. The right objective, function and you have a chance a chance I'm not a hundred, percent convinced, of this I'm this is fairly, new for me but I feel like there's. A real chance that we can begin to understand. The. This sort of broader theory, of choice, so. The. Two things that, we really need I think. Are evolutionary. Theory which, guides our choice of what's being optimized, it gives us a clue as to what the likely objective, function that our brains are designed to solve or. To optimize, and secondly. It's, really important, that we get broad comparative, data to help us understand the conditions that favor good decision making it turns out that. Different groups of people make, more or less good decisions, right, it's not everybody, who's a complete, idiot and behaves, irrationally.

Okay. And and, we. We sort of limit ourselves a lot in the. Sort of people that we inquire. About about, decision-making, I'm. Gonna focus primarily on, the. Evolutionary, theory and and give you just some hints, of some work that we've done without. Getting bogged down in the tighten technical, details but I think that the comparative, data is really important, as well. I'm. Gonna use this word fitness and. Use it a lot and. Evolutionary. Success is determined by fitness and the problem with fitness is it's one of these super fraught terms, right. That, has lots of colloquial meanings, that can kind of interfere, with our rigorous understanding and, analysis of it so, when I talk about fitness. I'm. Talking, about the relative contribution, that an individual, makes to, future populations. So. If you have some heritable trait whether, it's anatomical, or physiological. Or behavioral. Marginally. Increasing, the represents it's bearers in future generations the population, is going to evolve towards, that trait this, is the fundamental logic, of selection, right. And, fit. And and Fitness is the evolutionary. Objective. Function it is the subject, of evolutionary, optimization. Ok so just wanted. To get that out of the way I'm going to give you a. Quick. Preview of the, approach, that we use, in thinking about this. Sort. Of more formally, the. Organs, that make our economic. Decisions, right, they're organs just like our liver is just like our heart is presumably, these have been under intense selection. And part, of the reason just as an aside part, of the reason we think that brains, should, be under intense, selection is not just because they're useful but, because they're super costly, right, the, brain is an incredibly, costly, thing to carry around it, entails mortality. Costs, for. Mothers right, because we got this cephalo pelvic disproportion, big brains have to come through a limited pelvic, cavity when.

They're Born brains. Use, 20%, of, oxygen. At rest right they're super, costly, the, fact that they exist and that they've gotten bigger over time suggests. That there's a reason, they're there right. So. People. With good decision-making, capabilities. Presumably, in the past. Resulted. In leaving more descendants and became disproportionately. Represented in, the population to, the extent, that economic, decisions, affect Fitness we. Should expect outcomes, to track Fitness interests, more than some arbitrarily, defined, standard, form a standard. Of formal rationality, okay, so, let. Me just walk you through a little bit of the logic here of the, model I'm gonna move away from it and go and. Go on to talking about some some, features of the, human evolutionary environment, so. It's. A hierarchical, model. At, the, bottom we, have economic decisions. Right these are about, the, sort of things you hunt for the sort of things you put in your market basket the decisions, you're making on a day-to-day basis. About your, livelihood, right. And these, feed. Into, what, we think of as like proximate, motive. These, are the things that you're, sort of striving for. An. Approximate. Way happiness. Or sexual. Gratification or. Satiety, right the things that motivate, you to eat or to have sex or to do things that are meaningful in your life, but. These then, presumably. Have, an impact, on on fitness, and we can think of as Fitness, here as an aggregator, something. That aggregates, and averages, over these. Different, than, my pointer doesn't work these different. Goals. Which. Are fed into by, these. Economic, decisions, okay so that's a. Preview, of the type of reasoning. That we're using I'll come. Back to this in a little bit but. I do want to mention you, know I said the two things that we need I think we need evolutionary, theory I think we also need broad comparative, data on the. Decision making of actual people and particularly people who are good at making, decisions, under. Tremendous, constraints, and this is just a selection of a bunch of people they've, written some books right, and it.

Turns Out that's somewhat. To the surprise of much, of the the institution. Of. Studying. Economic. Decision making that. Poor people often. Make incredibly. Savvy decisions, right, they faced enormous constraints. But. When you adjust, for that when you control for the fact that their constraints set is very limited. Right that their strategy set is very limited in they're under tremendous constraint. Poor. People tend to make very good decisions in part, because it really matters to them okay. And so, I'm gonna leave that this is another big part of. The. Broader work that I'm doing but I want to focus on on more. Of the the. Sort of evolution part. The. Economic decisions I'm going to talk, about and show you some pictures of. Typically. Center on on subsistence populations. And the. You know the things look a little exotic but, the sorts of decisions, that people make in subsistence economies, are fundamentally, no different than the types of decisions that we make on a regular basis, right, what. How do you fill your Market Basket given a fixed budget. Right. It, may not be money it may be time. It may be political. Capital but. You're. You're doing very much the same sort of thing and one. Of the basically. My argument is that we. Have to think about humans as biological. Entities, we've been shaped by selection. And. That, our decision-making. Capabilities. Have fundamentally. Been shaped by by, by. This. Evolutionary, heritage as, to. A biological entity economic, decisions, are, not arbitrary, preference orderings, right which is sort of the way you typically, will, learn, about them in a microeconomics class for example it. Turns out that the rules for a living organism, anchored. In the present and subject to a force of selection, that really, doesn't. Like extinction, right Fitness, is a multiplicative, process. Right in order for your lineage to persist it has to persist every, generation. You can't get a zero in there or it's an absorbing boundary, right so. You really want to avoid those zeros. I'll, show that the all-important, need to avoid extinction, in a world that's incompletely, known, at. Best has, profound implications, for preferences, utility, and rationality and. One. Of the key, things I think that comes out of this is. It, ignoring the condition, of existential. Uncertainty. By. Doing that the theory of rational decision-making, is developed. A distorted, expectation. Of how, organisms working, in their own interest, should. Behave, okay. So a, bunch, of the work I've done on this is in collaboration, with my, my colleagues Rebecca and Doug Bird and, just. To give you a sampling of the types of economic decisions we tend, to think about do, you hunt for Delana which are these these, monitor. Lizards that live in the desert of Western Australia or do you hunt for Hill kangaroo right. They're different pay offs through different risks associated with this if.

You're A. Subsistence. Farmer in, this. Is in Uganda, another. Place where I work do you plant maize do, you plant yams or, do you forego subsistence. Crops, altogether, and take, your take a stab at the cash economy this is coffee. Working. With a postdoc. Ashley, hazel in, southern. Africa and maybe uh you know we're. Asking the sort of questions how do people, particularly women manage. Social relationships, too mad to manage the extreme, uncertainty of, living. In essentially, a desert. And trying, to raise cattle trying, to raise crops trying, to raise children right. Trying to keep yourself, going in a very difficult environment, so. These are the types of economic decisions, that, I tend to think about but as I say they're no they're fundamentally, no different than going, shopping buying. A house doing, the sort of things that he calls tend to look at, the. History of this for me. Goes, back to my former, PhD student Mike, Price who's now a postdoc. At the Santa Fe Institute and. Mike. Was interested, in understanding the, origins of sago palm farming, in in. Southeast, Asia. Sago, palm it's an amazing, crop it's one of the it's one of the Centers of agricultural. Innovation, right, in in around New Guinea in. The South Pacific. Southeast. Asia a. Single. Sago palm tree will Wilmette, literally, millions of calories, in starchy, flour right, that you can make, into a gruel. That's. Very calorically, dense that. Catches it takes 25, years, for a sago palm to develop, to. Maturity and to the point where it can be harvested and trying, to understand the evolution of this remarkable, horticultural. System led us to think hard about. Economic. Topics such, as time, and risk preferences preferences. And. Economic decisions generally it's, hard to make a sort of standard economic, model that. Says ok forego. The, immediate, reward activities, you're pursuing right now and go, plant some sago, palm and, wait 25 years right that's, a that's a hard problem. To solve but clearly it happened, at some point so, that's where all this really, started thinking and Mike and I have been working together for a few years on this stuff so. A. Theme that emerged in our work is that is that a biological entity trying to maximize fitness will, behave in, ways that often, really mimic, the. The, anomalies. That you see in this behavioral, economics, and ecology literature, we. Weren't looking to show that but we found how you, know people, today you know that kind of looks like Prospect area that really I that. Surprises, me so, they'll. Do things like use non, expected, utility rights. I won't, get into what that what, that necessarily means after you talk about later they'll, frequently violate the axioms, of rational. Choice. They'll, seemingly, have inconsistent, time preferences, right, though.

When. From the perspective, of an economic, decision these may appear irrational. But, these behaviors, are the sort of things that keep you alive in in a variable. And uncertain environment. Okay. So rather than the mathematicians. Formal. Axiomatic. Rationality. People. Seem to use a procedural, rationality, and this is a point that the great you, know polymath, economist, Herbert Simon suggested. In his scissors metaphor, right, that, trying, to understand, decision making is like trying to understand the way a pair of scissors work you've. Got on the one hand the computational, capacities, but. On the other hand you've got the task environment, trying. To understand the computational, capacities, like the logical rules of cognition. Without. Understanding the environment is like trying to understand half a pair of scissors doesn't, work very well okay. I. Said. That there's selection and TED talk like the environments. For saying provocative, things this is a slightly, provocative thing. Darwin. Saying it. Economics. Isn't actually a science right and, this. Isn't meant as a put-down it's it's it's a statement of fact and it's because the the preferred, and overwhelmingly, the dominant, mode of economic, theory generation, is axiomatic. Right. And that's not the way the Natural Sciences proceed. Behavioral. Economics, and and, this, great field of experimental. Developmental, economics has, changed this by actually measuring the way people behave, rather. Than sort, of specifying, the way they're, expected, to behave but. I think there's still a decent, amount of baggage, what, do I mean when I say, axiomatic. Standard. Economic, theory generation. Starts, from a series of axioms, right primitive assumptions, that are taken to be self-evident things, like you. Shouldn't reverse your preferences if you prefer. And b2c should prefer A to C that's, a that's an axiom of transitivity, right, that makes total sense but, there's nothing that guarantees that that's right there's nothing that guarantees that that's the way the universe works it's, just a logical rule that we say yeah I can live with that right. And so. You. Take these axioms, you derive out the way you expect. A rational, agent to behave and this. Is what economic. Theory really is all about its, normative, it just describes, how actors should behave, science. On the other hand is positive, right it starts from observations.

About The world and attempts. To explain them as inferred. Through observed data its inductive, right. The theory plays the role of generating, hypotheses. Which can then be confronted, with additional, data right. And so. It's. A common and it's a well stated critique of that neoclassical. Economic, theory describes the behavior of a species, other than humans and it's, referred to tongue-in-cheek by. The fake Latin binomial, Homo. Economicus, right. And. While it's true that humans do not behave according to the classical. Expectations. Or these neoclassical, expectations, you, know the. The expectations. Weren't this designed. To actually explain observed, behavior, right so we shouldn't be that surprised, by it I think so. I'm gonna talk to you about right, now for a little bit about what. I think the really important, bits of. Our. Human evolutionary. Legacy, are for. Understanding, the way we make decisions about things and. I'm gonna dive. Down fairly deep into into, my specialty, area which is called life history theory so. I'm. Gonna suggest that humans, are adapted, to environmental, uncertainty, and. This. Is a I, think a pretty telling graphic, so what, I've got here is its temperature proxy, data this is inferred, temperature. By. Million. Years before present so we're going from more ancient to more recent. Humans. Are a product of what's called the Pleistocene, which is a, geological. Epoch that spans from about 1.8, million years ago to about 10,000, years ago and the, two things that should be really obvious from, this plot are. That one the earth is cooled in the last five million years it's actually cooled considerably more in the last 25 million years it's, cooled and it's. Gotten more variable, okay. And you can't actually see it on here we, know that in places particularly in Africa. Where. Where there's. Been a lot of interest in this there's, actually a decadal, scale variation, here like enormous environmental. Variation going, on that, really gets amplified in, this period where. Our genus emerged which is this this, pink line here okay and just while. I'm on the subject of temperature, global temperature right you'll, sometimes see people say well you know the, earth used to be much warmer we. Can let it get warm it'll be fine maybe. You sure but. We're. A cold planet species, right that we are adapted, to a cooler planet, just, you know some of the bear in mind. The. Origin story of the human lineage and. The and the genus Homo in particular is a, story, of moving, from a, primarily. Arboreal. Lifestyle. In, tropical ever moist for us this is a chimpanzee. Foraging. For you can see the fruits up in the top of the picture, there in the jungle a forest and in in Rwanda, we. Moved from these forests, on. To, this, sort of mixed scrub savanna, this is a picture from Serengeti. National Park in Tanzania, and. Moving. On. To the savanna entailed, a bunch of new challenges. Right. There big predators, there there, are a lot of them they're. Big and they're scary. But. It also. Changes. The way food is distributed all right, here's a picture of a chimpanzee. Who's. In the, kabali National Park and and. He's, looking, down at me and and, he's, got an enormous pile of figs here and this, tree I don't. Want to exaggerate but, it it, certainly, had tens of thousands of figs possibly, hundreds of thousands, possibly. Well, into the hundreds of thousands of figs these, tropical figs are incredibly productive they're.

Asynchronous, Breeders so they breed sort of throughout throughout. The year they fruit throughout the year they're, not really high quality like, you wouldn't want to make a Newton out of them right you're not gonna make figgy, pudding with, these because they taste like crap I assure you but. When. There's no good fruit and the forest chimps will happily eat this and this is in general what the great apes do they fall back when, their preferred foods aren't there onto, this lower quality, stuff and figs, are a great thing in any tropical forest. We. Go out into the savanna there no there, isn't much fruit what. There are are underground storage organs, so. These are a bunch of hodza tubers, that my former postdoc Brian Wood took a picture of and, they. Look like sticks don't they doesn't, that look delicious, right. Try, to eat that and you're gonna spend like 12 hours just chewing. On it or your teeth your jaws are gonna be so tired you, have to have fire basically, in order to scorch. Them you. Burn them and you make them chewable. There's. A lot of stored, starch, in there and it makes for a great fallback, food but. They're very different than figs they're widely, distributed. They're, not at all clumped, right you're never gonna find a patch you, know just start digging oh yeah, and you got a dig for them because they're underground but you got to start digging you're not going to find a hundred thousand tubers there right, like you're gonna find a hundred thousand figs and a tree so. You have to spread out your effort right it changes the, economics of. Foraging. Tremendously. And. Here's Brian. Who's. You know one of the world authorities. On the hog hunter-gatherers. In Tanzania. And his, hair really is fascinating so, he. Gets that a lot he's now a professor down, at UCLA. So. Moving, from the force of the savanna and beyond increases, the variability, in food intake. Right. Because, it's, a drier environment it's a less productive, environment, there. Are lots of opportunities, but there are also lots of challenges and, while. There's, certainly more sort of meat on the hoof, your. Your ability to sort of manage, your downside, I think is is, much harder on, the savanna and so, this is why adaptation. To uncertainty, is a central, part of being human, right humans, move out of the forest onto the savanna. And it's a much more, variable, place and. This. This variability, is not, just, you. Know you don't just like. Set. Up an experiment, measure the variability and, get, your nice distribution, that you fit to it and say okay, fourteen, percent of the time I'm gonna have to do this and 27 percent of the time I'm gonna have to do this the way a risk, manager, would do it there's real uncertainty here you really don't know what the probabilities, are in part because the environment, is is very, rapidly changing. So. When I say uncertainty, I mean uncertainty, and not risk. Adaptation. Doesn't certainty is central to being human there are all these things that we do that are very unusual. We. Have an expanded, diet as I've mentioned you. Know that typically requires. Technology. To extract, like. Like, these, underground storage organs we. Are the mobile, primate, right our, home ranges, for, like a hunter-gatherer. Is. On the order of ten to a hundred times, bigger than a comparable size, chimpanzee. Group okay, so, like the hodza range over about 2,500. Square kilometers, whereas. A chimpanzee. Group will range, over about 20 now the chimps live in a forest so you, know that there is that difference but. That's part of moving, out onto the savanna and where that where the Hogs will live humans. Are so mobile right, when things get really bad we, get up and we move continents, right, this is very much a timely, issue right the 21st century in the latter part of the 20th first century is gonna be dominated, by human. Mobility. Right. We, see that it's breaking, lots, of international systems, right now and possibly leading to to, the rise of of right, or right wing extremism, right. Migration. Is part of our story from the outset. The. Other, really big thing that we do that, we're really known for and that's very, unusual right. Is we, share food and. It's. So unusual we don't even think about how weird it is that a bunch of unrelated, males doesn't. Matter in males females can, sit in the room and share, food together and, like. Not want to kill each other I mean I don't know I don't want to speak for you but you. Know I can sit in a room with people people are I don't feel like I need to go you know get all dominant, or something right, this is weird right you talk about your dog and all, the unconditional, love of a dog I've got an idea why, don't you try pulling.

Your Dog's food bowl away and eating from it see if that love is so unconditional. Right it's. A weird thing that people do we, share food all the time and, hunter, gathers it's it turns out to be a super risky and uncertain, endeavor to, go hunting, those. Hill kangaroos. 85%. Of all, hunting. Bouts end with zero, calories. Netted, okay, 85%. So these are. Mixed. Hunter-gatherers. And in the Canadian, Arctic my. Student Elspeth Reddy gathered. These data heroically. Over two years it's an incredible, incredible, story, and. It's. Just to show you I mean she also makes really pretty graphics, but it's just to show you that by having, these networks, of food sharing right, you. Ensure against. You. Failing on any given day you can count on getting food from someone else so. This. Is also a very unusual thing, that people do right, and and this. Is part, of the argument for this embodiment, of yeah. So. The red, houses, are food. Insecure, right. And the and the blue ones are food secure. Let's. Actually yeah yeah it's, actually a very complicated, graphic, I just show it cuz it's nice. So. Those. Are a few few ways we manage uncertainty. But, my main focus is on the evolution of our life cycles, right how long we live, why. We reproduce when we do the. Patterns of reproduction. That. Sort of thing and I. Would. Suggest that uncertainty is embodied, fundamentally. In our reproductive biology humans. Are really unusual I mentioned that we are peculiar, in our life cycles we. Have, probably. The latest age at first reproduction, of any mammal. It's, very late, we've, got high. Fertility given. How you. Know we've got low fertility when you compare us to like a pig right. We, have Singleton's, and we spread them out for well his pigs will have a litter of 13, or piglets. Or whatever right, but for. Our body size and for how late we begin reproducing we, are we are very, fecund, we have, a fertility rate that's about 50% higher than chimpanzees, for instance in natural fertility populations. We. Have a super long reproductive, span you, start reproducing you, know around age 20 or so you. End, 4550. That's a long time to be reproducing, for any mammal, and, we've got extensive, periods of overlap, of dependence, and and I this is another picture from Brian Wood among the hodza this, beautiful, hodza family, with.

A Mother with. Her four dependent, children and her mother who. Allows, her to keep. These four beautiful, children alive, by. Subsidizing, her. Foraging. Effort and her, time and providing this babysitting, I think, this picture embodies, a lot of the peculiarities of the human life cycle, among. Them include, substantial. Post, reproduction, post, reproductive, life span right which is something that most, animals don't have so. Let's look a little little data I've. Got a database, of about 1,400. Mammal species here. And. I've just plotted things, out as a function of body size. Biological. Things tend, to scale with body size and and. The. Modal. Mammal weighs about 50 grams right so there are lots of little things there fewer big things that's one of those things that scales with body size so. You put these on double logarithmic axes, and the. Red. Points. Are primates. The. Black points are all other mammals and the, green are humans, and so we can see that for. Our. Body, mass we have a very, late age at first reproduction. We. Have a very. Long life span and. We. Have low fertility but. We're not quite the same outlier that we are in, age at first reproduction, and lifespan we're. Kind of in with all the other great apes in terms of our fertility you. Don't really see how. Freakishly. Different we are how we're. We are until. You start to combine the parameters and this is also a sort of engineering kind of idea you, have these different parameters they have different different units on them let's, combine them, and make some dimensionless, numbers, and and, try to get at the design features of the organism, okay and this, life, history theorist rich Arnoff, showed. That, what. I'm calling it's it's the it's, the ratio of age, at first reproduction, the total reproductive, lifespan and, I'm calling that relative lifespan is call. Is constant. Across mammals, okay, and, we can see for I've, got the four great apes here for, the the, three non-human, gradate so you can see it's it's remarkably. Constant, the, same thing is true for the it's the product of age at first reproduction, and annual fertility, I've called it reproductive, power here, and. It's. A little more variable, but it's you know, constant. Enough for, comparative, biology considering. You're these things are measured with a lot of slop how, do humans fit in so. Our our, relative. Lifespan is really right in the ballpark of the grade of the other great apes what. About our reproductive power. Too. Right. So way more than double what pretty. Much all the other mammals are this. Is the thing that's really peculiar. About humans, so, we live a long time but, we actually also make, a lot of babies right. And those things typically, don't go together I.

Suggested. That a few, years back that this pattern of very long life and very high reproductive power, is consistent, with a strategy known as bad hedging, humans. Are bad hedgers the, long generation, length that comes from a late age at first reproduction. And a long reproductive, span coupled. With relatively, high fertility means. That you get to sample multiple environments, think. Back to that plot, of. Temperature. Over time and you. Know zoom in to a more. Local time scale it's, varying. Quite a bit right there's decadal, scale variation, that we know is going on in the Pleistocene in Africa and you. If you could if you had great, information, right. You, could say oh I know that this is going to be a great year to reproduce and you could pile all your reproduction, into it right. But, if wrong you're completely hosed right because. You've put all your eggs in one basket right, bed, hedging is all about spreading, out your risk doing. It a lot and trying, to sample a broad temporal. Series. Of environments. The. Essence, of bet hedging is trading. Off the mean for a reduction in variance, and I should mention you know the term comes from the. Racetrack. Betting, strategy, where you bet at least a little money on every single horse in a race, right. And, what that means is that you're guaranteed to lose money every race. But, you're also guaranteed not, to lose it all right, because some horse has to win and. You. Know you don't just like lay it out at random presumably, you have an idea about which horse is going to win but you had yourself right. And so hedging of course is is is. You. Know diversification. In general is right it's the fundamental strategy, for risk management okay but, why hedge, bets and why should. Humans in, particular need to hedge. Why. Hedge humans. Inherited a legacy of slow reproduction, from the great apes right we come, from this great apes talk and we, have very low, in. The broader scheme of things reproductive. Capacity, so, under the best of circumstances a. Chimpanzee. Population. Is going to grow at, about one percent annually if everything, comes together just perfectly and, the. Higher the growth rate the less likely a population, will go extinct. As. Humans, moved into more variable, savanna, woodland environments, they experienced greater variance, and. You. Can show right that, your long-run, expectation, is essentially it's a it's a function of your mean it's, also a function of your variance variance. Pushes. You symmetrically, around a mean and. When you have an absorbing, boundary. Right, that. That's a bad thing you want to avoid getting pushed around that mean right. So. Prolific. Reproduction, is sort of the other way you do this like if you're a muscle, or a sea urchin or, something that lives in a highly, variable like, coastal, intertidal, right. What do you do you just produce, millions. Of gametes that's. Not an option for humans, right because we inherit this legacy of the great apes so we're trying to avoid, extinction, given. These, enormous constraints. On our reproductive, biology, and. Again another another photo that I think really encapsulates so, many of these these. Challenges. That humans have in the embodiment, of uncertainty. In our life histories is. This incredible. Portrait. By Dorothea Lange this iconic, photograph, of Oklahoma dust bowl refugees, and. Here's a mother in California, with her infant son and, and. Her. Infant, and her young school-aged son. This. Is weird, I'm just i we. Take it for granted because it's it's humans you know you expect a young mother has has, a bunch of young kids but. It's really weird a chimpanzee, mother. Birth. Intervals of about five and a half years in Gombe National Park, and in Tanzania. Another. Place where I've worked. And. She. Will. Wean her infant, that starts that interval, right, at about age five four or five and she, will then never invest, in that child again economically. She will never share food with that child again she will never nurse that child again did, you imagine. Doing. That this kid sorry five. Year old we've. Got a new baby you. Got to make it on your own right that's not an option, for a human even. Those, of you who haven't had kids probably know that right. Sometimes. You wonder with Stanford students, right you have to explain that that it that won't work but, you know I assume. You guys know, so. Outpatient, uncertainty, is embodied, in our very reproductive, biology given. Its central importance it probably comes into our economic decision-making as well, now. I'm gonna, I don't have much, time. Left I'm going to talk about two. Specific. Applications. Very quickly. So. Here. Is, a schematic model, and I just put in the types of economic decisions, I tend to think about as a reminder we've, got these four economic.

Decisions, They're lotteries, in the sense that they they they, have variable, payoffs right these are decisions that have variable payoffs, and. We call them lotteries and, in. The standard, economic. Analysis, you have these different lotteries and they get aggregated. By a thing called utility okay. The. Utility, aggregates. It averages over these things and you're trying to maximize that and when, people. Start thinking, about well how do we mix economics. And evolutionary. Biology it's tempting, to first simply substitute, Fitness. For. For utility, but. That's wrong for two reasons first no. One but sociopaths, thinks about maximizing. Their fitness you know you say I must, have more babies and, they you know and I. Must compete with you know the people don't think that way that's not that's. Not something. And, that's about you know sort of fitting into the observations, again, but, second and more importantly. It. Would be a terrible, control, variable, for your for your behavior because fitness is measured. At a time scale that's basically greater than a human lifespan right, it's, about your contribution, to future generations you. Can't sit there and and have, a real-time feed of your fitness right, you. Need something, more proximate. To. Use. Fitness, it. Ends up being the arbiter of what things work and what things don't but you want to have something proximate, that you that you can use as a control, variable okay. To, respond, to adapt adaptively. To a changing, environment alright, so. Utility. In, the standard economic model aggregates over average and averages over different lotteries, nor. Model that I've already shown you right. The, lotteries are still there, we've. Got these aggregators. At approximate, level these, are these these these sort of. These. Proximate, motivational, systems satiety. Sexual. Satisfaction, love of your children what, you know whatever whatever, they are but what matters is that they are things, that you can use to. Guide your decisions, right. Over how how, you, decide to mix your your your, market basket for example and then, Fitness is the thing that aggregates, over the top of this this.

Is A model. Hierarchical. Model that Mike, price and I put together and. We've. Done a bunch of things with it it turns out they're a system, of equations that underlie this we can put some some. Constraints, on there and there's actually a whole talk where I talk about the constraints because it's it's. It's sort of surprising and interesting where they come from. And. We have results here. Are three very quickly. Distortion. Of decision weights kind. Of like prospect theory if the. You, know the standard economic analysis when, you have a risky decision to make is that, you're trying to maximize the expected, value of your utility and think. About what expected value means it means that you have a linear combination of, some weights there. Happen to be probabilities, but they're their decision, weights and and, some outcomes right it's a linear combination it, looks like people aren't, putting. Together linear, combinations. Of weights and and outcomes, okay they're using some sort of weird curvilinear, thing that. Kahneman. And turski have suggested looks looks kind of like that when. We use. When, we use. Fitness, as a criterion. As a objective. Function right. What we get is decision, weights that curve like that okay. So that was cool, another. Thing is. High risk aversion among the poorest people there's this idea it's, kind of this folklore I think, that that poor people have nothing to lose somehow you, know in the sense that they're very resource poor and so they should be willing they're natural, entrepreneurs right. They should be willing to take risks because. Was I got to lose well, what they have to lose when you think of us as, biological. Entities is that everything to lose right they're, poor, and just. And, and so if they're if their financial, decisions if their economic. Decisions, have, an impact on their survival ability then, we. Should expect them to be actually risk averse and in this framework it turns out that they're extremely, risk averse okay, and. Then finally. We. Can recover, without. Any sort of active choice these. The what looked like preference, reversals, some of the classic, paradoxes, of choice that came about in the 1950s. That's. A little more technical but it's a cool result to. Just. Want to remind you as I get to my last point here about the diversity, of habitats, that humans live in right. We. Come out on to the savanna. We. Can live in these. These mixed, rainforest, agricultural. Lands, in Uganda again we. Live in the tundra right in the arc inside, the Arctic Circle here's. Some some Venezuelan Llanos, home. Of you, know swamps, and big snakes right this is where anacondas, live. People. Are very they're very successful hunter-gatherers. Who live there we live on the, coastal. Shore as fisher, fisher folk we. Live on temperate. Prairies, right. We have a wide, variety of, habitats. And presumably, are, sort. Of choice, calculus, should be a little different depending on where we're actually making.

Our Decisions, so, as generalists with tremendous capacity to learn it seems reasonable to suppose that human the human capacity to make rational decisions has, an ontogeny, right, meaning it has a development, you, have to learn a little bit about it okay. And. I've developed, a simple model of learning about. One's uncertain environment that helps understand one of the most vexing results, in in, this, irrationality. Literature, which, is inconsistent time preferences, right, the fact that we we. We, value. Like. Waiting a day differently, whether it's today or whether, it's next year right, we. Care a lot about waiting, from today till tomorrow but we don't care that much at all you know a year a year and a day from now in whatever it's it's a year from now those, drives economists, crazy right, it may seem common sense to people, i think, that's an interesting insight. But. It does drive economists crazy and it's one of the fundamental. Anomalies. So, i have the simple. Model. Of. That. Has basically four parts the first is that your time preferences why do we care about the. Present more than the future well, a reasonable, way to think about that is because there's, there's, the chance that you might die before, you recover a reward, that you delay or. Less dramatically, that, reward might go away right, your friend might forget or it. Might be a resource. That that gets, swooped, on right, so, there's, a there's a hazard of failure. Of. This, resource and you. Can then have a, time. Consistent, prior which, is like an exponential distribution. Right that you value all all. Delays. Of, the same same, amount by the same by. The same amount, that's your prior distribution, right. Consistent. With the prior you. Can then go out and learn about the environment but. A lot of these important, economic, decisions, things about like how, how. What. You plant this year yields. Your. Yield, a you, know nine months from now a year from now you don't have a lot of opportunities, as a kid to learn about that you have a few so, you're probably at best going to learn about it with uncertainty, when you integrate the. Uncertainty, of it so the posterior looks pretty good here right I figure again I'm on fairly safe ground Google, pays right. Your. Posterior looks pretty good but in fact when. You integrate it to give your posterior, predictive distribution. All. Of a sudden you've got all that uncertainty that that. Causes, this big. Plateau. Here and it essentially, turns into what's. Known as hyperbolic discounting all right so it's an inconsistency. In time preference because you've got this fat, tail that you have to integrate over so. Uncertainty. In discount rate leads to inconsistent, time preferences, there's the exponential. Prediction. Right. The discount rate with no uncertainty, is exponential, ya uncertainty. And what you get is essentially. The. This. Hyperbolic, function, I was really excited when I read because it's really simple model right it's just a it's just a simple little Bayesian, model. Is like I've really nailed it this time turns, out then I then went back to the literature it turns out a bunch of people have said it. This. Is always a bittersweet moment on the one hand you're like damn I, thought it was so original but on the other hand it means that I had the same good, idea that partha does scoop the head and that's, always a good thing right, so, you know you, know in some you lose them so I'm, gonna give you three takeaways. Maximizing. Fitness distorts, economic, choices. Right. We've, seen this repeatedly, but. It maximizes Fitness right and that's that's, the direction that selection, is going to push the population. Uncertainty. Changes, optimality, predictions, qualitatively. When. You put uncertainty, into the game you get different answers. And that evolutionary, theory has promise for providing, a comprehensive theory of choice it's very much a work in progress I'm excited. About the potential here I'm excited, about trying to recruit some more people to work on this and. I'm going to thank you for your time and I'm, happy to entertain. The. Role of like social signaling. And. There's. That game economic, economic. Rationale. Where you know two. People. Yeah. Which. To me it seems like that's because. Costly. Punishment, absolutely. Yeah I mean so, this is another area that, has, got actually gotten a lot more attention. I think from evolutionary, anthropologists.

And Evolutionary, biologists, is is what's. Sometimes called our pro-sociality, right. This intense, social sociality. That we have that's, reflected. In these food sharing networks right but, that we have this psychology, of we. Don't like unfairness, right, and and, and. When you play the dictator game right. You you, punish people. Who. Don't give fair what you perceive to be fair offers, and presumably, this this, relates to this this need, for. Ensuring. Cooperation. In small-scale societies and, punishing. Cheaters. Yeah. Thanks. For that. Question. Related. Questions so if we're really concerned about you, know comfort and survival and things like that it's, amazing that humans will migrate to some of the most uncomfortable. Places. Yeah. Why. Would a rational creature go, to the South Pole go to the moon go to Mars yeah. Interest they're not very pleasant. Well. Do, we as creatures. You. Know there'll I think that anything that, makes, its living as a generalist, is is is gonna be curious I mean lots lots of primates are super curious right I mean if. You, live around here and you have a like like French doors you've possibly, seen raccoons at, your door it's, just sort of like looking, in like hey what's going on in there right so. I don't know if curiosity. We, certainly I think take curiosity, to another level all of those things. You. Know if you go to Mars and you come back that's a pretty high status thing to do right you you you have a you have a social signaling going on right. If you come back no right and. And, if you think about my my, Bayesian. Model right how many opportunities, do you have to learn about what the actual hazard, is you. Know suppose. We have like a really flat prior you know you've, never seen anyone go to Mars so I don't know it. Could be anything right. So. I think that that sort, of explains some of the the the. Unusual. Sort of decisions, that people seem to make moving. Into the Arctic to. To. Live right, it, may seem uncomfortable, to, us but you, know if you're in you do. Pretty well in general, right and and. You. Know the there's. A living to be made there and and, people are incredibly, adaptable. Question. Do. You see any other species. In certain, environments that share, similar pattern. Yeah. That's a great question. I don't. Really know there's there is a so, the classic, bed hedgers are, there. A bunch, of birds like again. Local, environment, red-tailed Hawks right, so a lot a lot of the raptors will, do this thing that that's really, kind, of perplexing, at face value they'll. Lay clutches, of two eggs but. Only one will, ever fledge, right, and and, it seems incredibly, wasteful, right and, it's. A classic bet hedging thing if that first one fails then. The second one is there right. And and the cost you're. Paying a little bit of you're mean you know you're paying out a little insurance, policy, right. For catastrophic, to insure against catastrophic failure. So, there are lots of birds that engage in this type of bet hedging. Reproductive. Power as. I've called it I've not seen it in any mammal it's, possible it exists another but, you know the way the the the Chernoff model works it you have to for this model you have to be a mammal like you could come up with with, these. Sort of dimensionless. Design, features, of different life histories but, it only works for you know if you have the same kind of general reproductive, biology. So. In in light of your research. Maybe. Academically, you wouldn't make any recommendations, like this but personally. Do you have, recommendations.

On How to, motivate. People psychologically. And collectively, to address issues, like climate change yeah, you. Know we're feeling in the present now yeah. Well. I have I have pretty strong opinions about the. Way we value the future and I think that. You. Know it it, seems. You. Know like it's a it's a bit of a moral argument but when. We say you know we want to take. Some steps to mitigate climate change now. Whenever. We do a cost-benefit, analysis, what's going to happen is that you're going to take a discount. Rate right of the. Value of an. Investment today you know a couple, of hundred years in the future and that's that's the real rub it's both the fact, that. You're. You're you're valuing the present more than the future but you're also all, these things play out at very long timescales. So. You throw in a four percent discount, rate and. You. Know it's very difficult, to come up with a cost-benefit, analysis, that favors mitigation, now with a 4% discount rate that pays out you, know its benefits in two hundred years so. I think and, I'm not alone in this they're a bunch of environmentally, economists, who are. Engaged in a very vigorous debate, on this that, we should use much lower discount rates for. Thinking about cost-benefit. Analysis, with these long payoffs. Martin. Weitzman at Harvard has suggested using the, absolute, lowest discount. Rate that you can that you can imagine, that, makes it tenable. Because. Of this problem, right and. Because, the the tail risk of these outcomes, are so huge, right. So. That's something I take. Away from this like like. Don't. Over, value the present, but. It's easy for me to do that I live in Palo Alto right we all live, in pretty predictable environments. Right we're. Quite privileged, people and it's easy to say well you, know we should really value, the future this. Is some of the work that my.

Colleagues Who I showed. You briefly have, done showing, that you know if you're, a poor so, like that they're the classic marshmallow experiments, that. Were done here, at Stanford right and you, know give a kid a marshmallow, and walk away say don't eat that all right because if you do that's. All you're getting but, if it's still there when I come back get to write, you, walk away and it turns out that you know these kids who struggle to they're looking at this marshmallow I know god I want to eat the marshmallow so bad right it, turns out that it correlates very well that, your ability to put off that reward and get the larger reward after a little delay predicts. A lot of things about your success in life. It's. Really hard. Anthropologists. Have found it really hard to go into places and, and. Replicate these types of economic games that we play all the time including, you know that dictator, games, the these, these things that we talked about before. In. Part because people are like how do I know you're. Gonna be back with your two marshmallows, that you talked about I'm. Gonna take that one right now thank you very much right and so it that's. Just a cautionary tale that it's easy for me to say we should value the future more because. You, know I I, have, a pretty low. Discount, rate because you know I live in Palo Alto, as, a, society, I think we should be modeling. The. Future. Valuing. The future more than we do. That's. A really long answer, to a pretty. Straightforward question. Yeah. Yeah. It's. Really I mean, that's a vexing problem, right that that we sort of develop, these these. These. In-group, and out-group identities. And that they probably are very functional, when you live in a band like society, because you, know your in-group is really your everything right, it's. Maybe not so functional, when, when. You're you you make you know the rest of the world you're out grouped right. And and you, know the same sort of rules of morality, don't, apply to them and that sort of thing. That's. A real problem and I and solving, that you, know there's there's a prize, for someone, right, we. Can figure out how to really begin to address that. Equally. Importantly, in the, world we live in. When. You're a peasant. Farmer, and. You plant. Yams. And and, you know some maize and whatever you know you sort, of you. Place. Your bet and then, you, tend it and and you hopefully get a big payoff right, there's.

A Direct. Causal. Interpretation of, what you do and what you get in. The world we live in we. Have all this complexity, and we, don't necessarily pay, the cost for bad decisions right away and I think that makes it very hard to learn about, good. Decision-making right. And so there. Are a number of people who've developed sort of the curricula, for teaching. Kids how to how to make risky, decisions, I think that. Giving. Kids and responsibility, and getting them some skin in the game is a really important thing so that's a little off of your question but I think that that's another real challenge. To thinking about how our brains were designed to make decisions and the, types of environments, in which we have to do it now many of us. Yeah. If. We predict, greatly, increased, variability in, the near future what. Would you predict. Well. First of all we have to get through this. Possible. Bottleneck, right, and. And. I say that flippantly because I otherwise. I'd just descend, into depression, right. Sure. It's conceivable, that. You know there could be more selection, on even, more spread-out. Reproduction. I mean. The thing is that a. Really. Interesting observation that, we make and development, economics, and demography is that. As, soon as you can ensure the. More or less certain survival, of your kids to Reaper of age people. Voluntarily, reduce, their fertility right. Like. And and and so we had this model. Called. A demographic transition, model and predicted. The sort of time scale over which this happened, and it seemed to explain the historical, data very well and then, you know late in the 20th century these places a lot, of places in the in the Middle East and North Africa. And-and-and, in Asia you know and the tiger economies like, fertility, felt like that, as soon as kids, could. Be counted on to survive. And. So. Yeah. If. We can if, if our sort of technological, and social technologies, can keep up with insuring, the child survivorship, right. And and keeping, the. The keeping. It favoring you know economic, investment, then, that sort of relaxes the selection pressure to spread everything out for sure and. I think that that should be a goal that we have right if we want to and if, we want to avoid sort, of negative, population consequences. Make sure that kids that. Kids survive to adulthood right. Because. That changes everything, about people's, investments and. That's really what what what, this you know given, that you can't effect, that right as a hominin, you know a million years ago right you don't have the. The technology, to really mitigate that right, so what you do is you spread everything out. But. If we have the chance to actually solve, it right then that's, the better solution, probably and that also. Sort. Of reduces the Selective. Pressure I think. Yeah. Yeah, thanks. You.

2018-04-01

Show video