I think the way to phrase it is something like: We don't know what the heck is happening, actually. We call this "randomness." But it is not randomness in the sense of truly, fundamentally unpredictable, nonsensical things happening. It is randomness only in the sense that almost everybody fail to see important things coming. But usually you could, in principle, see important things coming. And often a handful of people do, but nobody listens to them.

You use bell curves when you radically doubt your ability to see important things coming, and just assume that a sort of vaguely similar distribution of types of things will keep happening.

Taleb argues, correctly, that this is actually a pretty stupid assumption because new, weird, statistically unpredictable important things happen all the time.

I like Taleb's critique but I'm not satisfied with his answer. Radical skepticism in your ability to see important things coming ... probably shouldn't stop you from trying anyway, using whatever tools you decide, using your reason, are the most useful. You can't predict black swans but you have to try. One of the tools here would be leveraging statistical regularities in past performance. Another would be having some kind of thesis about how solar power, or crypto, is gonna be big, for Reasons, which violates past statistical behavior.

It definitely seems suggestive that hedge funds don't seem to outperform index funds, over almost any timescale you care to choose. Everybody is doing the best they can; Taleb points out that this still isn't very good; he's right, but what else are you gonna do?