The Black Swan

Taleb identifies 3 key features of Black Swans:
• Rarity: Black Swans are outliers that can’t be reasonably expected to happen based on past events.
• They have a profound impact on our society/world.
• Retrospective predictability: Although the events are unpredictable, humans tend to explain them on hindsight as if they could be perfectly understood and predicted.

The author calls an even a "Black Swan" if it's unpredictable not because it's random, but because our outlook on what's possible was too narrow. Imagine you’d known about the 9/11 attacks in advance. You wouldn’t have been surprised. In some cases, a Black Swan is only a tragedy for a single person. However, often it affects entire societies. Just think of Copernicus’s discovery that the sun is the center of the universe, not the earth, or when Armstrong set foot on the moon.

Based on the only things we can be certain of – what has happened in our lives in the past – we weave a narrative that makes sense and expect that the future simply must unfold this way. For example, imagine you’re a turkey and for years you live on a farm, get to roam free every day and are fed great food by a farmer. Why would you expect anything to change? But if tomorrow’s Thanksgiving, you’re just 24 hours away from getting killed, stuffed and roasted.

Another fallacy Taleb describes is called the ludic fallacy. This one explains why we do such bad jobs at getting the right insurance policy, for example.

When faced with the task to assess risk in the real world, we usually try to imagine the risk like a game, where there’s a set of rules and probabilities that we can determine up front, in order to then make the right decision.

• In Mediocristan environments, there’s a limit to the amount of randomness or deviation from the average. Inequalities exist, but they’re mild or controlled. Usually, there’re some physical constraints (e.g. height, weight, running speed) which limit the amount of variability. For example, if you add the tallest or heaviest man in history to a sample size of 1,000 people, the outlier won’t make a real difference to the average.

In Extremistan environments, there can be wild randomness and extreme deviations. Typically, there’re no physical constraints and no known upper/lower limits (e.g. knowledge, financial markets, e-book sales, social media “likes”). Thus, the outliers can make a big difference—if you add the net worth of Jeff Bezos or Bill Gates to a group of 1,000 people, it will drastically shift the average.