As you can guess from the title, this is a book about randomness, or if you prefer its popular name, luck. Nassim Nicholas Taleb and his love of Monte Carlo generators will try to convince you that sometimes when you see competence and determinism, you may be looking at randomness.

Crossing the T’s

Yes, randomness favors the ones that are prepared. Preparing oneself is certainly necessary, but surely insufficient for success. Same thing with persistence and perseverance. They are quite necessary as well, but also not sufficient for success.

As more random an environment is, less the aptitudes of a person matter. A dentist’s abilities matter more than a trader’s ones, because the financial market is more random than odontology.

The fact that all millionaires are persistent doesn’t make all persistent people millionaires, as a lot of unsuccessful entrepeneurs were persistent and worked a lot. The same thing applies to the incorrect idea of “taking risks”. You hear a lot about how taking risks is necessary to be successful, but what you don’t hear is that it is also essential to be a huge failure. If instead of looking to a set of millionaires we looked at a set of bankrupt people, “taking risks” would be a common denominator as well.

Alternative stories

10 million earned on russian roulette can’t have the same value as 10 million earned practicing odontology. Both are the same, can buy the same things, but the dependence one has on randomness is higher

This is a really interesting notion: when analysing a result, we should not just consider the observed result, but all the other possible results too. That way, we shouldn’t judge a performance by its results, but by the costs of the alternative results (if history had turned out differently). That way, certainty is something bound to happen in the larger number of different alternative stories, while uncertainty refers to events that occur on the smallest number of alternative stories.

The quality of a decision can’t be judged by its outcome, but it seems like that point of view is only expressed when people fail; when they succeed, obviously it was due to the quality of their decision.

Randomness can be beautiful

The exquisite corpse is a method of collectively gathering words, following some simple rules. One person chooses an adjective, another person chooses an noun, and so on. That weird name is because when people played that game for the first time, the phrase “The exquisite corpse shall drink the new wine” came up.

Sometimes, literature and poetry critics see meanings that the writers themselves missed. If you randomly put together words, some interesting and magical metaphor may arise. Ultimately, if getting fooled by randomness, let it be by the most beautiful (and harmless) one.

Noise and information

An idea that exists for ages has less noise than something new; at least some of the noise was filtered. The difference between noise and information is the same difference between journalism and history.

When it comes to investments, in a short period of time, what is observed is the variability of the portfolio, not the returns. At best, a combination of variability and returns, but definitely not just returns.

Count all the monkeys

Just made a complete examination of my life. During 25 years, in around 9k observations, I was alive. Due to that, I can consider myself immortal, with a high degree of statistical meaning.

If you put one trillion monkeys on a keyboard, it’s likely some of them will type something meaningful. The larger the sample, the most likely it is for you to get a successful item out of it. That’s why it doesn’t matter how often something is successful, if failure has an overly high price.

The car, plane, phone and internet were revolutionary to our world, but that doesn’t mean everything brand new will be; we need to count all the monkeys, not just the successful ones. The monkeys that stay behind are not counted and remain invisible; we only see the winners, and that makes us have a wrong perception of the real chances. That’s survivorship bias.

That also applies to theories. A theory can never be right, even though it seems right and obvious at the time. It could be that we just didn’t observe yet a scenario that falsifies that theory.

Learning with history

It isn’t natural for us to learn with history; as babies, we learn with our mistakes and only stop putting our hands in the fire when we get burn.

He describes an experiment performed by Claparède: a pacient with a debilitating amnesia couldn’t remember who he was (he had to remind her of who he was every fifteen minutes). One day, he hid a pin under his hand and pinched her when shaking her hand. On the next day, she couldn’t recognize him, but retracted her hand when he tried to greet her.

That introduced me to the notions of a declarative memory and a non-declarative memory, which I did not know anything about.

Overall

The main argument in this book is that we shouldn’t take ourselves and our perceived knowledge too seriously and should entertain some intelectual insecurity, which the author judges to be his most important asset in life.

There were some notes I made while reading the book that, even though I couldn’t put them under a specific topic, I am quite sure will serve as advices to myself later on:

  • a mistake isn’t something that is determined after the fact, but rather something determined using the available information you had at the time you made the decision
  • Be careful when you become too leal to ideas; ideas do not deserve your loyalty. They should be falsifiable, and when determined to be false, abandon them.
  • when observing the risks, do you effectively reduce them or that just gives you a feeling of doing your job?

Finally, it left me with a will to revisit stochastic processes and actually understand them (not just superficially know what they are about to pass a college exam); and ergodicity, which I do not remember seeing in college.