Shared by JohnH
Nice intro to NNT article found at Reddit, by sixbillionthsheep
Nassim Taleb is the best-selling author of The Black Swan. He is a former Wall Street trader and a current University Professor of risk engineering whose grandfather and great-grandfather were former deputy prime ministers of Lebanon.
Taleb describes himself as a skeptical empiricist. Philosophers of science would probably describe him as an extreme scientific anti-realist. He attacks modern intellectuals for what he calls their platonicity – a tendency to create underlying pictures of reality that go well beyond the data. His gripe is primarily with the social sciences, historical studies, the media and every day thinking. He is particularly savage towards finance, attacking Nobel Prize winners such as Robert C Merton (including to his face) for devising and promoting economic models which assume that market price movements behave in a log-normal fashion. It is the widespread belief in these models which he believes are the cause of financial market crashes.
The problem is, he says, we inhabit two worlds of probability which are often confused. He labels these two worlds – mediocristan and extremistan. Mediocristan governs much of the human physical world – it can be described neatly by the Gaussian (bell) curve where no single observation can affect the properties of a large sample much – height, weight, IQ, mortality rates. Extremistan describes the more random nature of human constructions and some natural processes – wealth, book sales, inflation, deaths from terrorist incidents, population of cities, size of planets. Western intellectuals have tried to fit normal curves to Extremistan processes.
His book is frequently punctuated by descriptions of common fallacies, some of which I list below.
Ludic fallacy : Trying to fit probability distributions to data without enough information and reasoning based on these false assumptions. Sometimes we have to reason without even knowing anything about the underlying distribution.
The problem of silent evidence: Deriving causality from an analysis of the winners e.g. assessing the traits (such as risk-tasking appetite) in common of millionaires to explain their success without examining the many more people with those same traits that failed to be millionaires.
Narrative fallacy: Fitting a story or pattern to a series of facts such as what happens in historical explanation and ignoring the many counterfactuals.
Confirmation bias: Seeking confirmatory evidence to validate preconceptions and hypotheses.
For a video lecture detailing his thoughts in The Black Swan, see here (thanks to rjefferson). His (rather unconventional) website can be found here. Long podcast interviews with Taleb about the financial crisis found here. One of his great heroes is Benoit Mandelbrot.
Taleb’s influence has been substantial. He is credited with predicting the present financial crisis (and profitting from it substantially as a result). He recently testified in Congress. He draws crowds that have Bill Gates and Michael Dell waiting in line. At times he is a pompous and haughty read – unquestionably controversial – but as of 2009, he is probably the most influential and widely read philosopher of science alive.