site banner

Nassim Taleb is likely wrong about IQ and talent

greyenlightenment.com
11
Jump in the discussion.

No email address required.

Taleb's claim is that IQ is meaningful for lower than average but not higher than average values

Taleb is right in probably a strict sense — there is diminishing marginal returns to intelligence if the standard of success is wealth.

For any given task, I would expect ability = f(IQ) to be a logistic function, because everything is a logistic function.

Income is a result of supply and demand.

IQ is by definition a normal distribution, which defines what the supply is.

But, what's the demand? If there just aren't many tasks (relative to supply of smart people) where the linear region is at or above say IQ=150, then increasing IQ above 150 shouldn't increase income.

if you want to want to be a billionaire before 35, it seems like a preq. having an iq above 150. Look at the founders of Google, Facebook, etc. It was like this in the late 90s too except put the cutoff at $100 million and instead of apps it was websites, IT. If you want to completely master a new and complicated technology at an early age, i don't see how there is an IQ ceiling.

I've not seen a study that doesn't top out too low to confirm or refute that.

Probably because very high levels of intelligence correlate with interests that aren’t super compensated in a pecuniary sense (eg pure math). Being 130 IQ with a great work ethic and not super interested in technical skills might play better than 145 IQ.

which would be true in a society that does not have division of labor or social stratification, which excludes every modern society. Even in antiquity smarter people were valued or had higher status, such as politicians, priests, generals, etc.. A comatose person has an IQ of zero. Yes, when taken to its extreme Taleb's argument is correct. To say that high IQ only measures test-taking ability ignores that such skills correlate in almost all aspects of life in which technical ability is an asset. People who score below 20 on the wonderlic, corresponding to an IQ of < 100, will almost always do worse at g-loaded activities than those who score above 30 (IQ above 125). It sucks that it is this way it is. It would unlock a lot of economic growth and productivity if IQ 100 people can be trained, at no added cost ,to do what 130 IQ people can do, but this does not seem to be the case.