Antifragile
Nov 5, 2022
So far in my life, no book collection has left me with as many insights as Incerto by Nassim Nicolas Taleb. The fourth book in the collection—Antifragile—discusses exposure to uncertainty.
The core message is that in the presence of nonlinearities and asymmetries, the average fails to represent anything meaningful.
Dividing the world in three
Antifragility and fragility mean potential gain or harm from exposure to the extended disorder family: uncertainty, variability, imperfect and incomplete knowledge, chance, chaos, volatility, disorder, entropy, time, the unknown, randomness, turmoil, stressors, errors, dispersion of outcomes, and unknowledge. The robust doesn’t care too much.
When you are fragile, you depend on things following the exact planned course, with as little deviation as possible. Hence, predictive systems cause fragility.
The antifragility of some comes necessary at the expense of the fragility of others. In a system, the sacrifices of some units are often necessary for the well-being of the whole.
While hormesis corresponds to situations by which the individual organism benefits from direct harm to itself, evolution occurs when harm makes the individual organism perish and the benefits are transferred to the survivors, and future generations.
It is often the mistakes of others that benefit the rest of us—and, sadly, not them.
Good systems such as airlines are set up have small errors, independent of each other — or, in effect, negatively correlated to each other, since mistakes lower the odds of future mistakes.
For the economy to be antifragile, every single individual business must necessarily be fragile, exposed to breaking.
No stability without volatility.
With bailouts, governments typically favor a certain class of firms that are large enough to require being saved in order to avoid contagion to other businesses. This is the opposite of healthy risk-taking; it is transferring the fragility from the unfit to the collective.
People have difficulty realizing that the solution is building a system in which nobody’s fall can drag others down — because continuous failures work to preserve the system. Paradoxically, many government interventions and social policies end up hurting the weak and consolidating the established.
Heroism and the respect it commands is a form of compensation by society for those who take risks for others. Entrepreneurship is a risky and heroic activity, necessary for growth or even mere survival of the economy.
Modernity and the denial of antifragility
Modernity here is defined as human’s large-scale domination of the environment, the systemic smoothing of the world’s jaggedness, and the stifling of volatility and stressors. It corresponds to the systemic extraction of humans from their randomness-laden ecology. It’s the spirit of an age marked by rationalization (naive rationalism), the idea that society is understandable, and must be designed, by humans. There is dependence on narratives and a widened difference between the sensational and the relevant.
We hurt systems with the very best intentions by playing conductor. We are fragilizing social and economic systems by denying them stressors and randomness, putting them in the Procrustean bed of cushy and comfortable — but ultimately harmful — modernity.
Avoidance of small mistakes makes the large ones more severe.
It is easy to assess iatrogenics (i.e. causing harm while trying to help) when the surgeon amputates the wrong leg, but when you medicate a child for an imagined or invented psychiatric disease, the long-term harm is largely unaccounted for.
Since no intervention implies no iatrogenics, the source of harm lies in the denial of antifragility, and to the impression that we humans are so necessary to making things function.
Intervening to limit size (companies, airports, or sources of pollution), concentration, and speed are beneficial in reducing Black Swan risks.
The media only report the most anecdotal and sensational of cases giving us a more and more distorted map of real risks. Likewise, by presenting us with explanations and theories, the media induce an illusion of understanding of the world.
Owing to domain dependence, we forget the need to check our map of the world against reality. So we are living in a more and more fragile world, while thinking it is more and more understandable. The best way to mitigate interventionism is to ration the supply of information, as naturally as possible.
The more data you get, the less you know what’s going on, and the more iatrogenics you will cause.
Avoid iatrogenics from forecasting. We understanding childproofing, but not forecaster-hubris-proofing.
A non-predictive view of the world
A man is honorable in proportion to the personal risks he takes for his opinion.
You can’t predict in general, but you can predict that those who rely on predictions are taking more risks, will have some trouble, perhaps even go bust. Why? Someone who predicts will be fragile to prediction errors.
Seneca’s version of Stoicism is antifragility to fate: no downside from Lady Fortuna, plenty of upside.
When you become rich, the pain of losing your fortune exceeds the emotional gain of getting additional wealth, so you start living under continuous emotional threat.
Wealth is the slave of the wise man and the master of the fool.
The modern Stoic sage transforms fear into prudence, pain into information, mistakes into initiation, and desire into undertaking.
If you have “nothing to lose” then it is all gain and you are antifragile.
Fragility implies more to lose than to gain, equals more downside than upside, equals (unfavorable) asymmetry.
Antifragility implies more to gain than to lose, equals more upside than downside, equals (favorable) asymmetry.
The first step toward antifragility consists in decreasing downside, rather than increasing upside; that is, by lowering exposure to negative Black Swans and letting natural antifragility work by itself.
About all solutions to uncertainty are in the form of barbells: the combination of extremes kept separate, with avoidance of the middle.
Optionality, technology, and the intelligence of antifragility
The option is an agent of antifragility. You are only harmed if you repeatedly pay too much for the option. But we don’t pay for the options given to us by nature and technological innovation. The edge from optionality is in the larger payoff when you are right, which makes it unnecessary to be right too often.
Trial and error has one overriding value people fail to understand: it is not really random, rather, thanks to optionality, it requires some rationality. One needs to be intelligent in recognizing the favorable outcome and knowing what to discard.
Epiphenomena refers to a set of illusions whereby you observe both A and B and think that A causes B or B causes A. For instance, lecturing birds on flying and believing that the lecture is the cause of their flying skills.
The idea of educating people to improve the economy is novel. In ancient times, learning was for learning’s sake, to make someone a good person, worth talking to.
Green lumber fallacy refers to a situation where one mistakes a source of necessary knowledge for another — less visible from the outside, less tractable, less narratable.
In theory there is no difference between theory and practice; in practice there is — Yogi Berra
Expert problems (in which the expert knows a lot but less than he thinks he does) often bring fragilities, and acceptance of ignorance the reverse.
Practitioners don’t write, they do. Birds fly and those who lecture them are the ones who write their story. So it is easy to see that history is written by losers with time on their hands and a protected academic position.
Governments should spend more on non-teleological tinkering, not research. As in venture capital, you bet on the jockey, not the horse.
We accept the domain-specificity of games, the fact that they don’t really train you for life, that there are severe losses in translation. But we find it hard to apply this lesson to technical skills acquired in schools, that is, to accept that what is picked up in the classroom stays largely in the classroom.
Only the autodidacts are free. And not just in school matters — those who decommoditize and detouristify their lives.
We do not study for life, but only for the lecture rooms — Seneca
People who build their strength using modern gym machines can lift extremely large weights and develop impressive-looking muscles, but fail to lift a stone. Their strength is extremely domain-specific and their domain doesn’t exist outside of ludic constructs.
The same with people who were selected for trying to get high grades in a small number of subjects rather than follow their curiosity.
Trial and error is freedom. Avoidance of boredom is the only worthy mode of action. Life otherwise is not worth living.
Example of a barbell: play it safe at school and read on your own, have zero expectation from school.
Much of what other people know isn’t worth knowing
What is not intelligible to me is not necessarily unintelligent. — Nietzsche
A fragilista mistakes what he does not understand for nonsense.
The nonlinear and the nonlinear
The difference between a thousand pebbles and a large stone of equivalent weight is a potent illustration of how fragility stems from nonlinear effects. Every additional weight of the stone harms more than the previous one.
For the fragile, shocks bring higher harm as their intensity increases (up to a certain level). The cumulative effect of small shocks is smaller than the single effect of an equivalent single large shock.
For the antifragile, shocks bring more benefits (equivalently, less harm) as their intensity increases (up to a point).
Black Swan events are increasing as a result of complexity, interdependence between parts, globalization, and the push toward “efficiency”. The world is getting less and less predictable, and we rely more and more on technologies that have errors and interactions that are harder to estimate.
The nonlinear is vastly more affected by extreme events.
To measure fragility, we need to detect acceleration of harm — figure out if our miscalculations or misforecasts are on balance more harmful than they are beneficial.
The function of something becomes different from the something under nonlinearities. The more nonlinear and volatile the something, the more the function divorces itself from the something.
We may never know x, but we can play with the exposure to x, barbell things to defang them; we can control a function of x, f(x), even if x remains vastly beyond our understanding. The more uncertainty, the more role for optionality to kick in.
Via negativa
Charlatans are recognizable in that they will only give you positive advice, exploiting our gullibility and sucker-proneness for recipes that hit you in a flash as just obvious, then evaporate later as you forget them. In life, antifragility is achieved by not being a sucker.
We know a lot more what is wrong than what is right. Negative knowledge is more robust to error than positive knowledge. Knowledge grows by subtraction much more than by addition. Since one small observation can disprove a statement, while millions can hardly confirm it, disconfirmation is more rigorous than confirmation.
A modern paradox is that we have more data than ever, yet have less predictability than ever.
More data — such as paying attention to the eye colors of the people around when crossing a street — can make you miss the big truck. When you cross the street, you remove data, anything but the essential threat.
Time and fragility
Antifragility implies that the old is superior to the new, and much more than you think.
Lindy effect: for the perishable, every additional day in its life translates into a shorter life expectancy. For the nonperishable, every additional day may imply a longer life expectancy. Hence, the longer a technology lives, the longer it can be expected to live.
Paradox: longer-term predictions involving the currently fragile are more reliable than short-term ones, given that one can be quite certain that what is Black Swan-prone will be eventually swallowed by history since time augments the probability of such an event. In contrast, typical predictions degrade with time; in the presence of nonlinearities, the longer the forecast the worse its accuracy.
People in risk management only consider risky things that have hurt them in the past, not realizing that, in the past, these occurrences that hurt them severely were completely without precedent, escaping standards.
Principles of iatrogenics:
We do not need evidence of harm to claim that a drug or an unnatural via positiva procedure is dangerous.
It is not linear. We should not take risks with near-healthy people (rather, Mother Nature should be the doctor); but we should take a lot more risks with those deemed in danger.
What Mother Nature does is rigorous until proven otherwise; what humans and science do is flawed until proven otherwise.
When the inhabitants of Mother Nature want to do something counter to nature, they are the ones that need to produce the evidence, if they can.
The good is mostly in the absence of bad. — Ennius
Many people benefit from the removal of products that did not exist in their ancestral habitat.
Distributed randomness is a necessity, not an option. Everything big is short volatility. So is everything fast. Modern times don’t like volatility.