This book is an interesting and fun read. It can admittedly also be a little bit frustrating at times, but it is one of those books that make you see things from a different perspective, and most importantly, it makes you stop and think. Its author, Nassim Nicholas Taleb, is a freethinker, no doubt. Not only does he touch upon subjects as diverse as philosophy, history, economics, mathematics, cognitive science and more, he also mixes fiction and fact, autobiography and argument, wit and graphs. The style in which this book is written is perhaps more original than the contents, but it is indeed worth reading for more than just its entertainment value.
Taleb is primarily interested in epistemology (what we can and cannot know and how), not metaphysics or language or anything else. He calls himself a sceptical empiricist and mentions Popper in favourable light. That induction is unreliable is hardly news, it cannot bring us certainty. (Neither is it, in contrast to falsification, a logically valid type of inference). Taleb takes this simple point to its logical conclusion and follows it into the various disciplines mentioned above.
Personally, I found the reoccurring references to the studies of cognitive bias one of the most interesting aspects of the book. Humans are prone to fallacies, even when we know that they are fallacies! We are prone to use induction even if we know that it is not reliable; we are prone to ignore what is not seen (silent evidence); we are prone to use narratives instead of argument; we are prone to fabricate in retrospect an “explanation” of an event that we could not foresee, etc. All of these biases probably result from our evolutionary history - they are in our psychology, not because they lead us to truth, but because they contributed to our survival.
This might not have been such a big problem if it wasn’t for the fact that we were made for (evolved in) what Taleb calls Mediocristan, while we now live in what he calls Extremistan. In Mediocristan, Black Swans – rare, unpredictable events – are so rare and relatively small that they cannot meaningfully affect the average. In Extremistan, by contrast, Black Swan-events can have a massive impact. Height and weight of humans are examples of phenomena belonging to the province of Mediocristan, while the stock market and website hits belong to Extremistan. We humans are fit for Mediocristan but very bad at predicting outcomes of phenomena belonging to Extremistan.
Another thing to mention is Taleb’s distaste for what he calls “platonicity” – our tendency to try to push the world into our models and theoretical constructs rather than the other way around. We create ideal models and categories and use them to predict phenomena in the real world. Perfect triangles are not to be found in nature.
This is not a very political book, but Taleb is clearly against central planning of all kinds (not just that of governments). He mentions Bastiat and Hayek in favourable terms and The Soviet Union as the worst example of bad planning. He emphasizes the value of bottom-up thinking over top-down regulation. His practical advice remain reasonable and rather "conservative": he does not tell us to stop planning or predicting things, but that we ought to restrict these activities to the small matters on the short-term and avoid making predictions of big events on the long-term. He also advices his readers to avoid reading newspapers, an advice that I myself have been following for years.
At the end of the book there is a small glossary explaining some of the many terms he makes use of in the book. One of the entries is academic libertarianism, which is a label that Taleb applies to himself meaning that "knowledge is subjected to strict rules but not institutional authority". He points out that "the interests of organized knowledge [for example, governments] is self-perpetuation, not necessarily truth" and that "academia can suffer from an acute expert problem". We humans are prone to listen to experts, Taleb holds, even in areas in which there are no experts!
There are many good and interesting points made in this book, but Taleb often leaves it to the reader to fill in the gaps and connect the dots.