I found this the worst book from those three, time after time he describes an idea that sounds somewhat interesting and then every single concrete example he gives seems completely wrong, the more so the more real-world experience I have in a topic he describes. For example he goes on and on about how "random tinkering" beats science when it comes to engineering - I would love to see an engineer that would agree with him on this. What he says about evolution seems completely misguided.
I like Taleb's books. His ideas are far from novel -- especially to people who work with complex systems -- but his books are engaging. When I read them, I rarely find that I'm impressed by anything he said, yet I'm _always_ inspired to explore some tangent, which I think is very valuable. However, you have to take him with a grain of salt. The arrogance is a bit much.
I read the first book. The arrogance was a bit much, especially in a book where he's talking about the importance of randomness. But, like you, I did go off an explore about a lot of what he said. So it's a fun exercise.
There are a few great books which are too repetitive. "Black Swan" is one; he could have got all of that into a book a quarter the size and not lost much important. Another one is "Reckoning with Risk" by Gerd Giggerenzer - he takes many chapters to say people don't understand percentages, and you should really use "X people out of 10,000" when explaining risk to people.
I'm only about 1/6th of the way through (listening in the car). I did have some disagreement on the specific examples, but they seemed to just be badly chosen examples (since he only really understands finance and some liberal arts, not engineering). So he's a bad writer, but not in a way which directly detracts from his philosophical argument. (and yes, the evolution parts sounded whacko too, but I think he was quoting someone else's beliefs, and had an argument that even if the explanation was wrong, it would still be a valid conclusion due to another argument)
Why would you include examples if they aren't necessary to understand the argument, or arguments that aren't required to make your point? That sounds confusing and tedious.
Because he likes to hear himself talk (well, read his own words) and show off superficial tangential knowledge (or fairly deep tangential knowledge about some humanities things, especially classics and lebanese culture).
It's kind of grating, but works for an audiobook in the car, where I might lose a few seconds due to temporary attention increase on the road. He'll reliably spend 3-4 minutes saying the same thing with minor variations when 20 seconds would do.
Nice to hear your opinion, hearing some praise of the book I was wondering what similar-minded people think of it, I wish he just stayed closer to fields he is more familiar with, as you said.
I did really like Fooled by Randomness, so I'm a bit biased toward him. Once I finish listening to the whole Antifragile I'll probably write a review and maybe send it to him -- if it's just bad examples, coming up with some replacement examples which are actually valid seems like an improvement, and not too hard.
It's essentially the culmination of Fooled By Randomness and The Black Swan.