Living With Risk


Finally, I’ve finished the Incerto. The Black Swan and Antifragile are the best two titles in the volume (including Skin in The Game). I’d go as far as to say those are the only two that are really worth reading (unless you somehow happen to have loads of free time). Antifragile is a little bit more of a chonker than its brethren, but as usual, NNT keeps his exploration concise, but also broad. He covers very many facets of the same general theme: things that benefit (or detriment, in the case of fragility) from randomness. Antifragile things are in context and usually have a maximum amount of volatility they can take before they break, but the recipe of benefiting from harm is widely applicable. To stay in shape, you want to have days to rest and days of extremely hard exercise. In nutrition, variability in the quantity and types of food that you eat (i.e. fasting or eating a lot of salad for a while, then gorging on meat) can have health benefits [1]. When you are only mildly sick or injured it’s best not to go to the doctor, to let your body fix itself (often better than the doctor would) [2]. In the economic and political landscape, optionality [3] and opportunism often trump long-term planning. “Convex tinkering” (trial and error) tends to be the source of most technological and social improvement, rather than specifically directed research. Recalling our discussion on the types of randomness [4] from The Black Swan, antifragility is the way to live life in extremistan. In extremistan we cannot successfully predict since small errors can butterfly out. Instead, we benefit more from focusing on exposing ourselves to randomness everywhere where it may benefit us, and shielding ourselves from it where it may detriment us. This is to some extent common sense, but easy to forget, and also, NNT is writing partly in an effort to stop a current trend of modernity: a desire to reduce, to model, and to predict (often with disastrous consequences)—a desire that usually falls into the trap of thinking we are in mediocristan when we really are in extremistan. Related to these reductionist tendencies, he explores various antipatterns that are often seen today. Roughly these include variations on lack of skin in the game [5], neomania [6], the need to “do something” when the best thing is to do nothing (which often leads to iatrogenics [7]).

I’ll leave it to you to explore what he has to say and find what is relevant to you, but I did find some things that I want to try (and/or keep in mind) for the future from Antifragile.

The main one here, however, is to look for options with high upside and low downside and do NOT eschew optionality. Similarly, to look for variability, not only since it will do me good, but it will make life more exciting. Along these lines, I also have a short list of quotes that I’ve found to neatly encapsulate broader points relevant to me (and hopefully you too). They may overlap a little with the previous points. Here they are:

[1] The fasting has been verified by scientific studies, while I’m not sure if the same holds for the other forms of variability. However, due to our generally antifragile nature, and mother nature’s ability to weed out detrimental behaviors, it makes sense to give it a shot. You don’t need “scientific proof” to try it though. (return)

[2] Of course, kids should also be exposed to pathogens early on, and be overly protected. This would weaken them later in life. (return)

[3] This roughly means having the option of doing something with little to no cost, keeping your options open without a cost, etcetera. In finance, option contracts allow you to do something along these lines for stocks, but the idea is more broadly applicable. Someone stuck to a plan is fragile, in that if something goes wrong they are kind of screwed, but someone who keeps their options open can take advantage of whatever random event occurs that may benefit them. (return)

[4] There are roughly two types of randomness: mediocristan and extremistan. In mediocristan the bell curve works and in extremistan it does not. Extremistan has distributions where single outlier events tend to be so large as to outdo the sum of the rest of the events, while in mediocristan the number of events needed to see such an outlier are so big, that no single event can change more than the sum of the rest of the events. In extremistan outliers (extremes) make or break the course of history, while in mediocristan it’s the average (mediocre) than defines it. Extremistan includes anything with popularity (virality), modern finance, social networks, startups (especially software), modern history, wealth distributions, and other situations with snowball, winner-take-all, or non-linear effects (feedback loops too). Mediocristan includes the distribution of human height, physical phenomena (thermodynamics, for example), the earnings of a dentist, casino probability, and things that are linear and straightforward. (return)

[5] Bailouts for example screw taxpayers by helping fragilistas in positions of power siphon away the shekels from them. They are not held accountable since they have nothing to lose from this. Same goes for interventionist (and militaristic) makers of foreign-policy. NNT believes that if we left the middle east alone it would probably be ok by now, and that many issues stem from going in and messing things up because we don’t know how they work and are overconfident. Sadly, it is these people (as well as cheap “tawkers” in the media) that receive the most recognition and status in our society, leading to a massive agency problem. (return)

[6] We desire new things all the time, but often the time-tested things are the ones that are most suitable and will probably outlive their successors. In information-based objects (i.e. DNA, books, religions, ideas, technologies) it is often the case that what has existed for the longest time, will exist for the longest time in the future. The wheel, weightlifting, cars, capitalism, and the literary classics are not going to go away any time soon. Meanwhile, flying cars, expensive new gym machines meant to be more “efficient” than free weights, and utopian economic ideals are less likely to work. Part of the reason is that the older ones have been tested by time. (return)

[7] Hurting without meaning to (usually by experts, often medical experts, at least historically). The idea here is, say, doctors, who insist on doing surgery when it might not really be necessary, exposing the patients to further risk. This can be related to neomania, in which experts start deploying new technologies that may be harmful since they haven’t been tested for long enough (i.e. blasting acne with radioactivity in the fifties, leading to cancer way down the line). (return)