Black Swan - Chapters II and III
Series of Summaries on Taleb's work
In this series, we are looking at Taleb’s The Black Swan. Chapter I of the book was covered here. In this edition, we look at chapters II and III.
Chapter II - ‘Yevgenia’s Black Swan’ is controversial. Spoilers within, so skip this para if you plan on reading this chapter in the book. Taleb discusses Yevgeniya, a writer whose off-stream book. ‘The Story of Recursion’ finds no publishers, until she posts the manuscript online and is discovered by someone. Subsequently, the book does very well and is translated into 40 languages. However, in Chapter III, he reveals that Yevgenia is fictional. Since Chapter III argues nothing and is not informative, its purpose is unclear.
Resume here if you skipped to avoid spoilers. Chapter III is ‘The Speculator and the Prostitute’. Here, he discusses scalability and two types of randomness.
On Scalability:
Taleb discusses the concept of scalability of professions. Some professions are scalable, where your output is not limited by the number of hours you can perform your job. Some professions are not scalable, where you are paid by the hour and you are limited by the amount of labour you can perform. In this context, I was wondering about whether being a corporate lawyer in India is scalable. Since your pay is limited by the number of hours you can bill, it does not seem scalable. However, once one makes partner, I do not know how the metrics would change and whether you gain some aspects of scalability. For unscalable professions, the kind of work is predictable, varies mildly if at all, and the income or the output of one single day will never be more significant than the rest of your life. Such professions will not be black swan driven. For people who value their laziness, it might be a good idea to become an idea person in a scalable profession, where your revenue depends on you making a few good decisions as against being a labour person in an unscalable profession.
For people who consider laziness an asset (like me) and are eager to free up the maximum amount of time in my day, I should become an 'idea' person and not a 'labour' person. However, Taleb recommends choosing a nonscalable profession, as scalable professions are worthwhile only if you rise to the top and are successful, when you are competent, competitive, and can produce monstrous inequalities in your favour. Scalable professions are random and there are huge disparities between your efforts and rewards. Moreover, scalable professions also seem to be viewed with survivorship bias. Everyone wants to become the next Bezos or Zuckerburg, but we have no idea about the number of failures and the cost of such failure. One needs to consider if they would be comfortable dropping out of a premier institute, say no to a 15LPA job and risk it all on a start-up and then fail five years into the future.
Taleb discusses three important points in human history on scalability - DNA through which humans can pass information across generations (although uni-directionally) without having to be present (or alive), language - where stories can be told without you being physically present in that location (or alive), and technology - like gramophones or printers, which let you scale your work of art.
On Randomness:
Taleb creates two fictional countries to explain randomness - Mediocristan and Extremistan. In Mediocristan, with a large enough sample, no single item will significantly change the total observation. For example, if you arrange 1000 people weighing 70 kg, adding one fat person weighing 150 kg will not drastically alter the average weight. However, in Extremistan, a single item can radically change the aggregate. For example, if 1000 people with average wealth stand in a line, and you add Bezos to the queue, the total wealth will be higher by orders of magnitude. Almost all social matters - such as book sales, academic citations are informational and not physical (like weight).
At this juncture, Taleb states that Extremistan can produce Black Swans and this is the central idea of the book. With Mediocristan style randomness, Black Swan events are not possible (as the chance of their occurrence in his words, is one in a several trillion trillion). However, I found this to be contradictory. The chances of a Black Swan even is perhaps substantially lower in Mediocristan, but isn’t Taleb’s worldview that possibilities may not be ignored merely because they are unlikely?
However, he uses this distinction to explain that in Mediocristan, what you can learn from data increases rapidly with every additional data point you get. Au Contraire, in Extremistan, this does not work as the average across different samples fluctuate massively as each observation has a huge impact. Therefore, what is known and what can be known in Extremistan grows slowly and erratically. Finally, Taleb calls matters of Mediocristan - height, weight, income for a baker, car accidents , etc… as type 1 randomness. Matters of Extremistan - wealth, income, book sales, etc… as type 2 randomness.
I hope you liked this edition. I am looking forward to discussing the next chapter on the problems of inductive knowledge in the next edition.