That is nonsense equivalent to arguing height isn’t a measure because it doesn’t predict NBA success, only lack of success.
IQ seems to be one of the best psychological measures we have, and Mr Taleb himself verifies that — if it weren’t, it wouldn’t exclude so strongly at the low end.
If someone is short, they are very, very unlikely to join an NBA team. But if they're tall, you don't know much.
All this is saying is that if someone is missing a necessary ingredient to do something, they can't do it. If they have it, some other ingredient becomes limiting. Or noise drowns out any predictive signal.
It connects nicely with the threshold theory of IQ, basically if you're above some baseline, that's opens up certain possibilities for you. But having loads and loads of IQ doesn't help much.
This is probably a mechanic that works in many areas. Some test will exclude people very well, but not choose "the good ones" well at all.
We all know NBA success requires more than height, however, height is still a measurement — not of total value in a person (even in regards to just basketball), but still of a real and objective difference between people.
Similarly, IQ doesn’t predict success, but is measuring a real and objective difference between people. Mr Taleb seems to confuse “cant predict a multi-factor outcome alone” with “isn’t measuring anything objective”. Again, IQ isn’t a measure of a persons value (or success) — just of a genuine variance in people’s mental abilities.
Mr Taleb unfortunately tries to extend the point that there’s more to success than raw intelligence to some kind of misplaced attack on the notion of measuring intelligence. That argument is as nonsense as insisting I can’t measure height, because not every tall person plays in the NBA.