I don't know why people make such a huge deal over what is essentially a simple statement that Taleb makes:
Educators posit that linear and other models of statistics are capable of dealing with highly non-linear systems, and then use the same tools to manage risk. He claims this is wrong and worse, is being taught to naive students at major universities, who then go and run corporations and are regulators, run hedge funds, etc. I agree.
Where I disagree with Taleb is that he doesn't propose a solution, only a criticism. He simply cites Mandelbrot and leaves it at that. Further, for years he took the other side of trades that were put on by the same people that derive value and manage risk by models like CAPM. He put on these trades not because the opposing trades were ill framed, but because he simply applied the negation operator to their philosophy and therefore he must be trading truth. So, if Neiderhoffer is putting on this trade, Taleb should be on the other side of it for no reason other than Neiderhoffer uses linear statistics.
That is as naive as the people/methods he criticizes.