Monday, December 15, 2008

Thoughts on "The Black Swan"

Having read "The Black Swan" by Nassim Nicholas Taleb (his website is here) I have come to a few conclusions:

  1. I spend too much time reading newspapers and on trivial and timewasting exercising (writing this blog post does not qualify as it counts as practice in writing and helps me organise my thoughts).
  2. Much of what I believe about the world isn't based on any kind of objective reality, but rather a collection of superstitions, cognitive biases[link], and predjudices.
  3. In certain non-empirical fields there is little value in deferring to experts if your intention is to accomplish something (like building a bridge, or making money).
  4. The value in the "free market" has less to do with competition and more to do with the resultant levels of "stochastic tinkering" that have the potential to lead to hugely influential but unpredicted developments.

The Book Itself

Taleb is arrogant, irreverent and amusing - all qualities I value in anyone. He writes engagingly, occasionally dipping into anecdote (for illustrative purposes, not necessarily to support his argument).

The Black Swan

A Black Swan is an unpredicted event that has a huge impact. Human beings tend to ignore Black Swan events when making decisions about the future, despite the fact that Black Swans tend to have an overwhelming effect in history, science, business, finance, and individual lives.

Black Swan events are outliers, they have an extreme impact, and third human nature has the tendency to attempt to "explain" the Black Swan after the fact, making it seem predictable and obvious.

Extremistan and Mediocristan

Taleb defines two areas of human experience: in Mediocristan things tend to behave in a fairly orderly and predictable manner. The distributions of height in a large population, for example, remain relatively close to a normal Gaussian bell curve even if you were to add the world's tallest man to the population - because he is not 3 kilometres tall he does not effect the overall distribution.

In Extremistan, on the other hand, outliers have a disproportionately large effect. The distributions of personal net worth in a large population, for example, will be completely thrown out of whack if you add Bill Gates to the population.

Power Laws and Guassian Bell Curves

In Extremistan, power laws (Pareto's Principle, Zipf's Law) and fractal relationships are the norm. In Mediocristan Guassian bell curves are the norm.

Taleb believes the bell curve is misused in it's application in investing (I wasn't aware the bell curve was used extensively in finance).

Scalable and Non-Scalable Professions

Inequality tends to follow a power law. The richest 1% own 50% of all assets, whereas the poorest 10% own substantially less than 10% of all assets.

With regard to income some professions lend themselves to black swans and power laws, and others don't. Medical doctors and priests will tend to earn roughly a certain amount, which will fall somewhere on the bell curve of income for that particular profession.

Writers, venture capitalists, entrepreneurs, and scientists, on the other hand, all occupy extremistan, or black swan territory. Some writers (like J.K. Rowling or Terry Pratchet) earn collossal amounts of money per hour worked whereas the vast majority will struggle to earn anything.

Empiricism and the Problem of Induction

How do we know that what happens in the past will continue to happen in the future? Taleb likes empirical philosophers like Francis Bacon and Karl Popper, and Sextus the Empirical. In trying to find knowledge about the world it is better to prove conclusively that something doesn't work than that it does - in fact you can't prove that something always works.

Knowledge, therefore, emerges from negatives. Disproving something adds to knowledge.

Platonism and Theories

Taleb rails against creating a theory and then selecting evidence to fit the theory. He dislikes the application of game theory to economics, and portfolio theory to investment. He describes the mistake of focusing on elegant, tractable mathematics at the expense of empirical knowledge as platonicity, and any knowledge divined from such abstract mathematical theorising is "nerd knowledge."

Cognitive Bias and the Narrative Fallacy

Correlation does not equal causation. Post hoc ergo propter hoc.

Terry Pratchett and Taleb

Pratchett comments extensively on the human desire to create stories (the Discworld is a world that runs on narrative, rather than physical laws) and on the fact that million-to-one chances occure nine times out of ten.

Also Pratchett's uber-politician Lord Vetinari is the embodiment of the knowledge that in an unpredictable world what people really, really want is for tomorrow to be pretty much the same as today.

Arthur Koestler and Taleb

Skimming through The Sleepwalkers by Arthur Koestler (which I have read) I can see many parallels between Taleb's dislike of platonism and Koestler's similar criticism of excessive theorising in physics.

Taleb refers to Koestler's book in The Black Swan in the context of the inadvertant nature of much scientific discovery.

Koestler was also intensely critical of the dogmatic nature of theoretical physics (as he perceived it). Koestler's problem was that, unlike classical physics, quantum electrodynamics involved fields and quanta that could be one or the other or something else, depending on your perspective. I personally think the problem is more to do with the fact that QED is counter-intuitive. We humans have evolved in a quasi-classical environment and just aren't set up for thinking in terms of the quantum environment of the very small.

Hayek and Stochastic Tinkering

Taleb believes the value in free markets lies in their ability to generate new idea in a process of stochastic tinkering. This apparently is what Hayek thought.

Taleb is also a fan of Francis Bacon and Karl Popper. Does not like Platonism - mistaking the map for the territory and overextending the use of models.

Whither singularity?

Both Kurzweil and Taleb are obsessed with the fact that humans tend to have an "intuitive-linear perspective."

Taleb doesn't comment on the singularity directly. He does refer to the importance of power laws and such.


There is a lot of food for thought in this book, and I intend to read a lot more about the various topics that Taleb raises.

As ever at this stage in my education any new knowledge raises more questions and than it answers.

However as Taleb's central point is that the world is in many ways fundamentally unknowable and unpredictable perhaps this is a good thing.

The objective of learning is not simply to know a load of facts, but rather to become comfortable with the limits of your understanding. To transform unknown unknowns into known unknowns and to cope as best you can with those things that remain unknown unknowns.

No comments: