Transcription
Now, a system that I find has pernicious attributes is one like the banking system, in which a bank crash makes the next problem more likely with contagion effects. And, of course, uh, the bankers are being harmed by these bank crashes. These systems should not be tolerated.
And for 5 years, I've been talking, talking into a wall. All right, I got tired of this, that we, the banking system, the finance, anything that governments have been touching has been to transform, uh, a healthy system. Like the restaurant business, there's no bailout. The computer industry, there's no bailout. Anything they touch, they turn it into a crony system of these kind of things, with bureaucrats running companies effectively and having the upside, and of course, a system to blow up.
And I came up with some ideas of how can you counter that.
>> In the most simple terms, just define for all of us here, those that haven't read the book, what do you mean by anti-fragile? Anti-fragility?
>> When you ask people what's the opposite of fragile, typically what they, you know, come up with is solid, robust, all these kind of concepts that are not the exact opposite of fragility. Because if you map fragility mathematically, you get something called concave. And, uh, even in, in, in England or in even in Russia, when you ask people what's the opposite of negative, they don't say zero or neutral. The opposite of negative is positive. No.
If I have a package and the package is fragile, I write, "Please handle with care," and I send it to Siberia with champagne flutes, okay? And you know it's fragile, right? So the opposite of that would not be a package that's robust, resilient, adaptable, adaptable, all that kind of stuff that people say is the opposite of fragile. If it were robust, you write nothing on it. You don't write, "This is robust." I don't really care. You can all, you Russian, uh, custom agents can brutalize it. I don't care. You don't say that. All right? You just write nothing. All right?
The opposite of fragile is, "I beg you to mishandle this package," just like I beg you to, uh, uh, watch out for this package. So the opposite of fragile is something that gains from disorder.
Well, the first thing is, if there's something that gains from disorder, okay, the first conclusion you can derive from it is that historically, something has had to survive in an environment in which disorder is a fuel for that entity, that fuel. Therefore, depriving that element from disorder is harming it.
So, I came up with the idea of the metaphor of the cat versus a washing machine. I don't know if you know what, I'm sure many of you know what a cat is. Some of you may know what a washing machine is, but the difference is monstrous. A washing machine will never self-heal. I, not the washing machines I have, you know, had the privilege of owning. All right.
A cat requires, like your body, your, you know, especially him because he's an athlete, professional athlete. Your body requires some kind of disorder. If you don't have enough disorder hitting your system, then your bones will become brittle. You will start looking like, uh, you know, uh, like Warren Buffett or someone like that, honestly, very quickly. So lack of stress for a, uh, organic something organic is harmful. And effectively, a large, uh, uh, uh, share of what we call aging comes from atrophy, lack of, uh, stress.
So you have two things. Some that respond that like disorder, and some that don't like disorder. And, and one, I'm going to get technical for two minutes. There's something I call the disorder cluster. And actually, no, no, actually, I'm less technical. I call it disorder brothers. All right. Volatility, variability, time, chaos, uh, randomness, uncertainty, all affect these elements in the same way. So, in other words, uncertainty can be very beneficial for a lot of things.
So democracy, it's called democracy. Mess is called democracy. And it's identical to the bank, the banking problem. Banks were good at hiding risk, like the risks in Saudi Arabia. No volatility, but people thought it was no risk. And you had a gentleman here called Golden Brown. No, you remember a little >> Gordon rather than gold. >> Gordon. Okay. So, uh, he, the fellow wanted to eliminate boom and bust. And then on his next assignment, if he were given nature, he would eliminate the seasons. So you'd have 68.7 degrees year-round. So that's the idea.
>> So the point being that what looks sort of on the surface to be unchanging is not remotely antifragile. And we mistake the wrapper. If it's organic, exactly. If it's organic, >> then it will require some variability for it to be in a stable situation. And, and physicists know it. There are a lot of systems that have this property. I know, I'm sure there, I mean, we have these lights, but I'm sure when they beam the lights, where the physicists will read with themselves and give us examples of things called annealing, where, or things that gain from, uh, disorder, or stochastic resonance, things with the physical world that require disorder to function well, require unpredictability to function well.
One of the, the points, the stories you tell in the book that really underlines this is that the, the guy who sells predictive analysis to the US government, yes, about what's going to happen in the world, is a foreign policy expert. And the quote begins, "I think in blackjack, if you get 4% better, you become an awful lot richer. Our job is to be 4% better at predicting what's going to happen in the world." And you sort of make the point that there's no analogy between a casino and foreign affairs.
>> Exactly. I mean, I call that the ludic fallacy. It's very fallacious to think that the environment was nonlinear dynamics. You can use the same can predictive methods that you use in, uh, in, in, uh, in blackjack. And what you call cosmetic, uh, I know you're going to hit me with something because [snorts] I, I heard in the green room, I specialize in something called nonlinear, uh, exposures. In other words, if the stock market goes down 10%, you're hurt a lot more than it went down twice 5%. Or more than twice if you went down 5%. That kind of thing.
So I thought about it, and then one day I realized that if I, if humans jump 10 meters, they're hurt more than twice than if they jump five meters. You see? Then if you recurse, then you realize that they hurt a lot more than if they jumped a thousand times, 1 centimeter, uh, 10,000 times, 1 millimeter, and so on. And it has to be that way because anything that survives has to have nonlinear exposure. Because if, if harm, cumulative harm, were linear, then you would die just walking into the office. So fragility comes from the fact that every incremental, uh, degree of stressor harms you more than the previous one.
So once I realized that the fragile can be mapped to the nonlinear, and there's some theorem that it took me 15 years to to write, all right, that the nonlinear is what doesn't like stressors, volatility, whatever it is, I, I said, "Okay, you can measure fragility, and you can measure antifragility." Because once you can measure fragility, you can measure its opposite, antifragility. And it's very simple. There's something called convexity. A convex is simply if you make more, if something goes up a dollar, then you lose, and something goes down a dollar. If you are in that situation, if you're convicted to a source of harm or a source of randomness or source of anything, then you benefit a lot more from variability than otherwise. And, and that obsessed me for a while, and then I realized it was very applicable.
So I'm starting this, you know, it's not the solution to the problem, it's the beginning of the problem. I want to ask you about one of the metaphors you use, because as you were saying to the about the firemen who understood your work very quickly, you try lots of different metaphors, and once it clicks, it clicks. One I really enjoyed was about the barbell strategy. So you're trying to introduce a a visual idea to say what you mean by an antifragile approach or strategy to low.
>> Uh, okay. The solution is very simple. There's something that flows automatically from the following. In life, you can't predict. And the reason they keep predicting is because they're not harmed by their errors. We are harmed by their errors. All right. If you can't predict, there's an optimal strategy. The strategy is to not have to predict your losses in a bank portfolio or a personal portfolio. To take, instead of having medium risk, to have one large portion that's zero risk, if you can find something equivalent [snorts], and one portion that's very speculative.
Then I thought about it and discovered that it was literally practiced by every risk management, uh, entity in nature. So look at monogamous birds. I don't know if you heard about monogamous birds. They're not as monogamous as you think. One of the birds, all right, one of the members of the couple, uh, instead of having a medium-term, uh, spouse in bird terms, what I don't know what they call that. All right, mate. All right, whatever. It's not official documents. Instead of having someone of equivalent genetic quality, uh, shoots for a linear combination of 95% accountant, 5% rockstar, right? So just like in a portfolio, 95% low risk but boring stuff, 5% extreme, you know, happiness, whatever it is. These strategies seem to be, you know, they're like 23 in nature. 23 barbells in nature where you combine extremes together, and these represent, help you become, move away from fragility and become antifragile.
>> Two different, uh, models, if you like. You talk about the French model where people have synicures, all writers are in favor of synicures, um, and then they write, if you like, in their spare time. And then in England, you, you provide a wonderful list of 18th and 19th century innovators, um, uh, original thinkers, and you say what do, what do they all have in common? And they were all parish priests.
>> Exactly. The point being not that we should take up holy orders in order to think better, but that having an 18th century house and logs and tea and stability enables you to have, >> and free time. No, no, you have stability and free time. It's a lot better than being a writer being forced to write like Trollope and Dickens for a living. I'm an academic now after being a trader, and I tell you, academia is the closest thing I know to prostitution. I, I've never, you know, engaged in prostitution. So I tell you, see, you know, you can't do anything. Everything is about perception and, uh, how someone's going to like it or so on. You can't produce any serious work.
>> But I want to ask you about your, uh, if you like, your personal history and your ideas and how they interact.
>> Uh, before talking about that, I mean, I will get to that, but before talking about that, uh, I want to mention one thing concerning antifragility. So antifragility is someone who gains from disorder. And the biggest problem we have today is when some people have the upside, and they transfer the downside to others. And society, modernity has moved into a mode where you have bonus earners. What I call bonus earners. They have the upside and, and no downside. Like bankers made in 2010 more than they ever with our money. You know, we bailed them out, and they paid themselves bonuses. So the upside, absolutely no downside. Regulators, all these bureaucrats are not harmed by their, uh, by their, uh, decisions where, and we're harmed by it. So the incentives are are warped. But in a small village, if you're the head of the village, and you have a bad decision, people are going to see you.
>> So you're in favor of making things on a smaller scale. This is one of your points about banking. It's also one of your points about other spheres of >> Yes, exactly. I mean, what happened is that you have two systems. A system that is, uh, calibrated. I call it calibrated. It means it matches nature. Is a system in which no failure is in vain.
Now, a system that I find has [clears throat] pernicious um attributes is one like the banking system, in which a bank crash makes the next problem more likely with contagion effects. And, of course, uh, the bankers are being harmed by these bank crashes. These systems should not be tolerated. And for five years, I've been talking, let's talk into a wall. All right, I got tired of this, that we, the banking system, the finance, anything that governments have been touching has been to transform a healthy system. Like the restaurant business, there's no bailout. The computer industry, there's no bailout. Anything they touch, they turn it into a crony system of of these kind of things with with bureaucrats running companies effectively and having the upside, and of course, a system to blow up.
And I came up with some ideas of how can you counter that? But it's very simple. If a company is to ever be bailed out by the taxpayer, if it fails, it should be declared bailable out. Hence, no employee should ever earn more than a civil servant of the same category. Because effectively, civil servants, I mean, all these bankers in New York, a city bank, they're paying themselves millions. They're civil servants. If I, as a taxpayer, have to finance them, they're civil servants. Okay? That would force the system to become small and work more like organically like nature. So if you have lots and lots of small banks, they effectively become, it's the ecology of the the economy is more like a restaurant business where if one of them goes bust, no >> Exactly. Except that I don't like banks overall because banks tend to create debt. So it's not like they're not like restaurants. If you let, you know, if you lower the debt in society, and debt is not a very good thing because it fragilizes. It makes mistakes large, lending makes mistakes large, and gains small.
>> Now, you are very anti-planning, aren't you? And you even say, "I hate having a plan for a book."
>> Exactly. No anti-planning because you get locked. It's like a highway. Think of a highway with no exits. All right. There's nothing wrong with having a plan or having a prediction, except you always end up going, getting on a highway without an exit. And they are, uh, there are reasons, and we have enough evidence that if you have a prediction, if I give you a prediction, you're going to take more risk.
>> You also talk about tinkering and experimenting rather than planning. So you, you, you, you quote Terence Keely's work about the industrial revolution, where it didn't actually was not inspired by someone innovating in a vacuum, if you like, and think, "We're going to invent stuff." It was a lot of it came from workmen on the floor actually tinkering and trying to find practical solutions. Technology leads to science, and, and, and just like people think that education leads to wealth, and the exact opposite. Countries get wealthy first. We had the illusion that wealthy countries, because they're educated, that education leads to wealth. But typically, it's education that comes after wealth. That science comes after technology. And often, when you have beautiful cathedrals, some schmuck will come up with the idea that people use Euclidean geometry when most of cathedrals in Europe were built at a time when only five persons knew how to divide, you know, longhand. You see, when I say trial and error, trial and error is a misnomer. It is not trial and error because an error can, uh, uh, crash a plane. Ah, it is trial with small error, large upside, very different. It's a convex trial and error.
So what we have is a mechanism. We have to encourage people to take a lot of risk when their losses are small. And effectively, we know now that the, the more you raise education in society, the less trial and error. So the, the problem we have of everything can be mapped antifragile. You can be stupid and antifragile and do much better than if you're very, very intelligent and fragile. And the more uncertainty there is in an environment, the more you can do, uh, you can capture that that volatility through trial and error. It's like a free option nobody's paying for.
>> One of the big targets in your book is medicine, or, and you're very critical of modern medicine. Do you want to tell us about >> Okay, let me come up. I mean, I don't know how technical the audience is, but it's okay to be boring once in a while. There's some property in mathematics called Jensen's inequality, discovered in 1905 by someone who is not actually a professional mathematician. Jensen's inequality tells you the following. If you gain more, okay, in an accelerated way, if you have acceleration of gains, like for example, if one glass of water is, uh, sorry, if two glasses of water are a lot better, more than twice better than one glass of water, then you have acceleration, then you, you, you, you like randomness.
So I started, I read about, uh, the diet, and someone said, "Okay, very naive, what do the Cretans eat? They live long. Let's eat like the Cretans." Okay? So needless to say that every single cab driver in New York became a restaurateur that worked, okay? And they thought it was healthier because they live longer. Then I thought about it, I said, "Well, these guys are, I mean, I'm not going to use the word Cretans, but, uh, they're not very intelligent." Because there's a second-order effect, that the Cretans matching what they eat doesn't mean you have to match the frequency of what they eat. And effectively, the Cretans have traditionally 200 days of some kind of fast, and then you take, uh, so I started looking at, uh, convexity in medicine, completely missed. And convexity is so trivial. I was an option trader for 20 years, but don't tell people because by now I'm trying to make people forget it. But when you're an option trader, you understand convexity. You understand that what's convex likes volatility. What is non-convex dislikes volatility. And, um, and, and see naturally that, hey, you know what, maybe we do better if you had a lot of food one day and no food the other day. And I looked at it, and effectively, a lot of people documented, but they didn't put the mathematics together. And it's exactly the same as an option. [snorts] You want an option. I'd rather have 10% one day, zero the other, than 5% two days in a row. It's the same thing with food, up to a point.
I haven't been bothered by doctors because they don't mess with statisticians. But it's a statistical property of things that if something likes randomness, you got to feed them with randomness. Same thing with the human body. This comfort, five meals a day, two meals a day, whatever, all that is destroying humans. And, and I can't believe that the Romans had one meal a day. People still tell you have breakfast. You know, like food deprivation is ne necessities to be calibrated so you can start drawing on have a stressor, necessary stressor. People go to the gym to work out, lift muscles on some stupid Nautilus machine. They don't understand that hunger is a stressor, just like any other stressor, or sleep deprivation, any, any stressor, provided you have time to recover from it.
>> It's very unusual for a book to actually put you off your breakfast, but yours one does because it's, it tells you pretty much it's forbidden to have breakfast. There's no evidence. There's no evidence. And you see the lion eats breakfast and then hunts for pleasure. No. I mean, you, you hunt and then you eat breakfast.
>> I want to talk about the lion. I want to talk about the lion.
>> The lion's vision improves after they miss a hunt, but from starvation.
>> The lion also sleeps for about 20 hours a day and then hunts, which is a very nice way to live, I think. And you, you advocate something similar. One of the, one of the metaphors we, we use a lot.
>> That's a barbell. That's a barbell.
>> That's a barbell strategy.
>> Barbell sleep 20 hours and frenetic for 4 hours. Yeah. Our ancestors lived like like that. They played played played and then hunted a little bit, which was fun. They had absolutely no duty feeling of to be effortful. Things came naturally to them. You run, you know. So we in modern life had to minimize this torture called the effort. You have to learn to adapt life so you don't have to make an effort.
Um, I want you just to talk about the, the, the title of your book, Aphorisms, because it's a lovely story, the Bed of Procrustes, which appears again in, because I read the, the Aphorisms first, and then, and then Antifragile, in some ways you can see a lot of the strands are already there in The Bed of Procrustes, and then you can see nobody like having all the footnotes, it's, it's quite fun. So tell us about that mythical Greek story.
>> Procrustes was, uh, the, uh, an inmate, a lot of legends, but he was a nasty fellow, right fellow. He would invite people, force him to, to sleep in his house after a great meal. And though he had a bed, those who were too short, he would stretch them to fit the bed. Procrustes, a stretcher. And those who were too tall, he would cut their limbs, okay? So, uh, so I came up with this idea. I said, "Okay, there we go. The, the modern life changes us. Medicate children so they can fit the curriculum, rather than changing the, you know, and you have to be in sync with nature, not by changing nature, but changing you." All right, changing your own habits. And Antifragile is, of course, a more nerdy approach to the problem where, you know, uh, you have this, um, this, uh, you know, where we have to match the randomness of life, uh, embrace the disorder and stuff like >> [applause]