Commentaire positif le plus utile
14 sur 14 personnes ont trouvé le commentaire suivant utile
An interesting, but highly polemic argument (2nd edition)
le 26 février 2013
Taleb's book is certainly one of the most interesting books I've read in a while. In it he argues a very simple point- many things in life happen on exponential scales, and because huge events on those probability distributions are also very rare, we're ill-equipped to understand or deal with them. That's really the central point of this book and he does a very good job presenting a very forceful argument for its adoption. Taleb focuses frequently on economics and business as areas that lack such knowledge. Rightly so. For years, I could not believe that an academic discipline could even exist when it was founded on the argument that human behavior was rational. It took a psychologist to teach them it wasn't (versus them reading a psychology article or two). Clearly, we have ample evidence from all financial walks of life that markets are all but unpredictable. Financial "experts" are generally just lucky. Throw 100,000 people into business and sheer dumb luck will ensure that some will achieve regular success. Taleb says that's all well and good, but not if we start to believe that it's skill and prediction that makes the difference. Because we get fooled into thinking we can predict the future, and then 2007/8 comes along and smacks the financial world into near-oblivion.
Taleb discusses when and where Black Swans (unknown, unlikely, but powerful events) are likely to occur and what we should do about them. If they are expected to be positive (e.g., scientific discoveries) then we should do things to maximize our exposure to them (i.e., have diverse and open research groups). If they are negative, we should do things to guard ourselves against them (e.g., purchase insurance, temper average predictions, etc.). There could be more depth here, and the additional chapters of this 2nd edition to add some of that. But if you're looking for a book with specific predictions about how to live your life or conduct your finances, this isn't it.
My biggest problem with this book is the relentless polemics and strong doses of arrogance. It's ironic in many cases. Taleb repeatedly bashes financial fields (especially economics) and then turns around and says one of his biggest sources of consolation against attacking academics is that he's made far more money than them. So which is it- knowledge is key or money is key? He also bashes the social sciences at every turn. I agree that many social "sciences" are in fact social arts (most economics being a great example). They are poor at predicting and fail to use the scientific methods. But many social sciences (e.g., psychology) as well as biology (he frequently mentions evolutionary psychology positively, but somehow misses this connection in his polemics) do indeed use rigorous scientific methods and their use of the bell curve is appropriate for their measures (e.g., personality). Black Swans are most common in large, complex structures such as weather, history, or economics. Also, any good scientist should know that we never know everything, that any theory can be toppled, and to expect the unexpected from research. Taleb also attacks tenured professors, but the system of tenure allows for the open research Taleb claims as necessary for detecting or even just thinking about Black Swans. His name-dropping, of both foes and friends, is annoying and detracts from his otherwise intellectual arguments.
Overall then, I think this is a fantastic book about how we think. It's very similar to Malcolm Gladwell's or Daniel Kahneman's works on why our minds are easily fooled, only the topic is larger and more elementary (the basic nature of probability and knowledge). It's not a simple book to read, but it is engaging as the ideas within are very powerful. Unfortunately, it is marred by a deeply aggressive, reactive, and polemic style of writing. Fortunately, in the additional 2nd edition chapters, he realizes that he can make much more headway by telling researchers and practitioners when their arguments/theories/models work (i.e., when the distribution is bell curved) versus when they need to realize they aren't in Kansas anymore (i.e., exponential-type distributions with Black Swans). That approach fits much better with the human psyche and is thus a more effective way of delivering one's message. And it's how I'll wrap up this review. Buy this book if you want to know what we don't know and why that really matters. Because surprisingly, what we don't know can sometimes be much more important than what we do know!