[gburt]

This is what I get out of his writing. He is a great pop-culture writer, expresses some important (and mostly, relatively old, repackaged) ideas about convexity and natural mathematical implications thereof.

He is one of my favorite authors, but I am unconvinced by nearly everything he has to say. He has this cult-leader style of "everyone else is wrong, but here, you can be part of the enlightened group" that I find somewhat grating, but more importantly, he simply does not provide "extraordinary evidence for extraordinary claims."

Almost nothing he asserts is empirically testable due to his reliance on fat tails and Knightian uncertainty, and when it is, it either has not been empirically tested or the entire "traditional" industry is built around the empirical results. His response to this is generally of the form "systemic failures can dominate normal effects." [1]

In the cases where he leans on statistical modelling, especially modelling with fat tails, his writing breaks down. I find his approach at that point to feel sort of "baffle them with bullshit" in the sense that he is indifferent to whether it is even possible for the reader to understand what he is saying, once he has convinced himself.

[1] e.g., https://www.facebook.com/nntaleb/posts/10155509264828375