Every once in a while, a book comes out with ideas that are like fireworks. Nassim Nicholas Taleb's book "Antifragile" is a explosion of life changing ideas. Taleb gains his prominence as one of the very few who predicted the banking collapse before it happened. He took a great deal of adverse social pressure until it actually happened.
Antifragility is the property where something benefits from risk, uncertainty, volatility, randomness, disorder and stressors. It is not resilience or robustness because that is just staying strong where antifragile actually gets better. Examples of antifragile from the book and from me are:
- Iron is tempered by fire into steel.
- Maximum weight lifting improves human muscle mass the fastest.
- Good computer system developers stress test their systems in every way.
- Venture capitalists make multiple small investments needing only one big win.
- Human Learning is best done dynamically not in rote ways.
- Mythologically a hydra gains more heads than it loses.
- Germs get more drug resistant if not totally killed off.
- Football games score the most per minute in the last two minutes of a game.
- Every plan crash is examined and used to make all flying safer.
Nature and evolution recover and grow so well that people think there is nothing that nature cannot handle.
Even more significant is the idea that we can make things, people, ideas and human institutions more antifragile by reducing fragile characteristics. Anti-fragile resists the harm from errors, and then recovers and learns from errors. To see how revolutionary this idea is, just imagine a mother telling a child to embrace risk and just handle all the possibilities better.
And this is just the beginning, the best is after the fold.
In this discussion, antifragile is good, like a hulk superman who gets better at every fight. Fragile is bad, like a superhero who has every wound from previous battles and who fights worse each time.
Predictable Single Path vs Antifragile Ready for Anything
Even thinking that things are predictable makes us more fragile for we invest in that vision of the future and optimize for that single future. "Life is lived forward, but remembered backward." We as humans are prone to single path thinking. Our human minds make up narratives to fit circumstances retroactively. Even a false narrative will make us feel better. Predictions are based on normal variations and the recent past which ignores a Hurricane Sandy. The author describes Hurricane Sandy events as Black Swan events. These are events that are so rare that no one anticipates them.
A very-smart corporate turkey is being fed for a thousand days by a butcher. A staff of analysts say that butchers love turkeys and the future looks brighter and hopeful with each passing day. The positive outlook has a greater statistical significance each day. The turkey assumes from the past that since there was no harm so far that there will be no harm in the future. Then butchering day comes!
So then how can we win as turkeys in this world, when we cannot know who is the butcher and when butchering day happens? Here is where Taleb's book shines with practical suggestions.
1) Do less with grand theories more with tinkering/testing. Knowing the negative is very helpful. Every failure is information. A simple rule of thumb is better than a complex theory. Knowing and being able to predict a pattern is more important than knowing the actual story behind the pattern.
2) Simple rules: Smaller is better. Slower is better. Older is better. Anything that gone through volatility is more trustable that anything that has not. Explanation: Companies work better if they are able to make small errors that are easily fixable. The longer that something has existed the more likely that item, design, or institution will around for much much longer, no matter what you think of it. Think of wine, coffee, chairs and religions. Small can "squeeze" through, large never can.
3) Trust people who have skin in the game. Be especially wary of people who will lose nothing if wrong. Be wary of a situation where benefits are small and visible and the side effects potentially severe and invisible, like testing a new drug. An agent moves risk to others, a hero takes risks on to oneself. CEOs have very little downside risk, hence no skin in the game.
4) Less is more and usually more effective. Be more likely to do omission instead of acting. Make actions small and reversible. Make small reversible errors to find the negative spaces.
5) Mastery is doing, not the study of doing. Learn from the best doers not the best theorists.
6) Create stressors if you don't have them. Police do this by creating practice disaster events.
7) Create redundancies like Nature does. For me, Just-In-Time inventory is one of the most fragile creating choices. At the very least, a company ought to have the extra critical parts of anything that would shut down the business. Local is more anti-fragile. Think insurance and back up systems.
8) Designing fragile parts can make a whole mechanism or whole institution antifragile. Think fuses.
9) Bottom up is better than top down. Bottom up is the fight for success of little small parts, easily fixed and changed. Top Down is good as long as it is always right. However the errors are huge and institution threatening. Taleb actually makes the case that freedom is more important than peace.
10) React to signal not noise. Only look at very large changes in conditions not small ones. Otherwise you can get problems from intervening, causing unintended consequences. Information has a nasty property of hiding failures because of the great opportunities to cherry pick results. Detail gives us comfort when it should not. We react to change more than direction.
11) Have a surplus of options. Get out of debt. It would be better to live in a smaller house with no debt than a fancy house with a huge debt. Have multiple career options and a backup plan. Need and want less, then you are not susceptible. Avoid stupid risks like smoking. Use emotions to make one successful by transforming fear into prudence, pain into information, mistakes into initiation and desire into undertaking.
12) Avoid banking on the mean or the middle, instead build a bimodal strategy that builds for maximally safe and maximally speculative. For example, in every economy the super rich have money to waste. Better yet instead of doing a single path analysis, do multiple path assessments. Taleb gives some great examples where one can win big with a 20% chance of winning as long as the winning is huge and one invests in multiple 20% big win chances. Asymmetrical relationships are key here. A 10% increase in New York Traffic does not cause a 10% slowdown, it causes total blockage. Track payoffs not probabilities.
13) Variability matters as much as the average. Minnesota weather is a good example of that. If you plan for the average, you could never live in Minnesota.
14) Trust in real numbers not marketing. We would consider a person who self-marketed the way that corporations do as a bore and probably deceptive. Then why do we accept this from corporations? Corollary is do not trust in the words of a person who is not free, like Lawrence of Arabia. Greed can be more than money, it can be greed for credentials.
Technical if you can go there
1) Antifragile - If you gain more than you lose by variability. This is where you want to be. If you graph a variable to its effects, the graph will be convex, like a smile.
Samples are exercise and finding shipwrecked treasure.
2) Fragile - If you lose more than you gain by variability. This is where you don't want to be. If you graph a variable to its effects, the graph will be concave, like a frown.
Samples are number of cars in traffic and storm loss.
The book is very readable for the non-technical with great stories. The book is layered so if you want to go technical you can. Taleb also goes into the technical modeling errors of portfolio theory showing that people think they are spreading risk when it may be a false illusion.