Nassim Taleb investigates the improbability baked into the world and the expert’s underestimation of it. A quick synopsis of the title is that for many years it was assumed that all Swans were white. This assumption held true over thousands and thousands of observations. It remained true until Australia was discovered and lo and behold a Black Swan was found and overthrew the “scientific theory” that all swans are white. This anecdote is a reminder that we cannot verify anything, only conduct experiments that either confirm or disconfirm theories. Confirmation of a theory should not be considered verification. This problem is the main focus of the book which is: you can for sure know when you are wrong but will never know when you’re right. Taleb defines a Black Swan event as relates to the book as having the following characteristics:

  1. It has a disproportionate effect on reality
  2. It is not predictable using the scientific method
  3. Once the event happens, it is immediately rationalized in hindsight as being something that could have been predicted. Taleb grew up in Lebanon but was forced out at an early age due to an unpredictable civil war that happened there. This probably helped set him off in the direction of this theory. He notes that no one predicted the war, and everyone underestimated how long/damaging it would be. His dad was a high ranking official in Lebanon and his predictions were just as accurate as a cab driver who knew nothing about it. He draws a parallel with the people of WW2 underestimating the effects and severity of the war as evidenced by the markets just prior and during the beginning of the conflict. Taleb attacks people on the left and right in this book but having a background as an options trader he has a special vendetta against those people who right economic forecasts, these forecasters are never held accountable for their poor predictions. These predictions all have a fatal flaw, they presume the ability to quantify using platonic distinctions, without ever realizing how rare a perfect triangle (if one exists) is in nature. His complaint about mathematics is that it is 100% correct and not 99% which makes it dangerous to use in the real world because it gives you a sense of confidence without reflecting the uncertainty inherent in reality. He makes an interesting distinction between two sets of data, one set lives in what he calls Mediocrastan while the other set lives in Extremastan. To illustrate the difference between these two sets, imagine picking at random 1000 people and weighing them and then finding the average. (It might just look like a bell curve ;). Now imagine finding the heaviest human in the world and adding it to the set. How much would it effect the average? Not very much. Next imagine that you took 1000 people at random and averaged their income. Odds are it too would create a bell curve like distribution, but now imagine adding the richest person in the world to the set of 1000 people. This would completely destroy the distribution as the richest person would have over 99% of the wealth leaving the rest of the 1000 people to fight over the remaining percentage, and with that, we have a crack in the liberty bell. People’s weight belongs in Mediocrastan the land of bell curves while people’s income belongs in Extremastan where bell curves lie. He then makes the argument that there are many more scenarios that belong in Extremastan today (economics, history/future events, best sellers, etc.) that are treated as if they belong in Mediocrastan. Another great example he gives was that of a casino that was working on risk management. They spent millions of dollars to catch cheaters that would try to steal money from the casino where what ended up costing them much more was three black swan events that the risk management team had not predicted. A tiger mauled its trainer causing them to cancel a popular show. A lazy employee had inexplicably been hiding tax documents under their desk making them have to pay a huge fine for back taxes and the owner’s daughter was kidnapped making them have to dip into the casino’s earnings to get her back. The major take away here is that people mistakenly define risk in terms of bell curves, but bell curves do not account for the truly improbable which paradoxically happens often. The problem is centered around this conception that randomness can be modeled by a roll of the dice. He would call this mild randomness, because if rolled enough times dice are actually perfectly predictable and therefore not random (he objects to the Heisenberg uncertainty principle as being metaphysically important for the same reason- it averages out). In the real world you will find true randomness that cannot be predicted by any method. He defines randomness as the following:

“In the end Randomness is just unknowledge there is no functional difference between a completely random system and a chaotic system whose outcome we cannot predict”

When I first started this book, the author struck me as conceited and a little annoying, but by the end of it, the tone was somehow a little charming? humans are weird. Either way this book challenges a lot of conceptions in my gourd and gives me some more ammo to flame those scientific positivists. Very interesting read full of anecdotes that will stick around in your head for a while.