I've been asked by several people to opine on Nassim Taleb's recent Twitter war against IQ as a concept. You can read a preliminary write up of his critiques on a Medium article he wrote. Not too refined, but if you listen you get the point. I'll see if I can sum some of the ideas up:
- IQ is a very effective measure of outcome... on the dumb side of the distribution, but not the "intelligent" side. People of measured IQs of 70 are well predicted to not get far in life. The difference between 70s and 100s is massive, but the difference between 100s and 130s is not on average.
- Because of the above, many very strong correlations between IQ and other things are spurious because IQ isn't really so predictive/causative at all levels.
- Thus, IQ tests are pretty much effective in that they can detect utter morons fairly well, but so can any other arbitrarily defined test; the IQ test isn't especially effective or some grand discovery.
- IQ tests have been designed to award a particular type of abstract thinking that is not generally helpful and can lead people to be oblivious, naïve, boring and dangerous. The Talebian IYI.
On that last point, you might want to listen to the podcast I did on Gerd Gigerenzer's work, which is related: we, especially psychologists, have this fetish for cargo-cult rationality, often when it leads us to be dumber/bigger chumps IRL. Taleb is also keenly aware of the sociological aspect of academia, the replication crisis, the wishful thinking and confirmation bias ubiquitous in the soft (i.e. non-) sciences. Many people will correctly call IQ or g "the most successful" finding of psychological sciences in the 20th century. I say "correctly" because everything else psychological "sciences" have found is either totally non-reproduceable or the result of smoke-and-mirrors, so, sure IQ is number 1 in a field of continual failure.
Anyway, on the topic, a lot of Boomers will deny any potential reality of IQ/g or any other kinds of cognitive tests out of a deep-seated faith in total human cognitive uniformity (otherwise that would make them literally Hitler). This isn't really Taleb's motivation: his, if I my psychoanalyze for a moment, amounts to a disappointment with the pseudo-rationalist "funcionary" thought-patterns that an IQ test favors being celebrated as a cognitive ideal, when in reality these mental habits aren't really so correlated with success outside of bureaucracy. Taleb certainly loves to dab on n*rds and I support his efforts. That and a conviction to sound understanding of statistics drive him.
Is "IQ" utterly meaningless? No I don't think so, nor do I think Taleb thinks so, but it's only a clumsy and introductory mosaic used to approach human cognitive differences and we shouldn't necessarily treat it as a mentalistic holon without careful caveats. As he puts it in the above Medium article:
"If you renamed IQ, from "Intelligent Quotient" to FQ "Functionary Quotient" or SQ "Salaryperson Quotient", then some of the stuff will be true. It measures best the ability to be a good slave. "IQ" is good for @davidgraeber's "BS jobs"."
"Autism Quotient" might be another candidate. IQ shows something, but I think a lot of Taleb's critique comes down to assumption that "high IQ" tendencies are not necessarily either good or indicative of intelligence in a meaningful sense. I think there is a strong correlation between IQ what we intuitively think of as "intelligence", but you must take that with Taleb's qualifiers mentioned above and the fact that for some cultural reasons (and pity), we think of autistic shut-ins and losers in high-school as being "intelligent" (thus the concept of IQ has affected what we think of as being smart). In the past, autistic shut-ins were understood to be utter morons and the ideal intelligent person might be an outgoing, socially-aware, but "irrationally" cautious, high-agency person with little attention to detail. If "IQ" measured that kind of person, which we could easily do, we would likely find some of the same correlations since like our "IQ" it rules out the extremely dull, but such a test might be likely to have better life-performance correlations as well.
To sum up, I mostly concur with what Taleb says, but I also don't particularly mind using the term "IQ" in the way it's conventionally understood so long as it's not understood as some eternal metric of goodness. I did speak somewhat about the selection of cognitive ability, measured by IQ in one of my other podcasts, but my editorial stance was that that selection was not necessarily "good". Like Taleb says in the quote above, a person with a high IQ tends to be a good functionary, a good cog in the system. Perhaps to slavishly obey social orders one does necessitate some level of real intelligence, but that servile and autistic mentality not what true intelligence causes per se, nor should we celebrate than kind of mindset.
If anything Taleb says sounds like gibberish, I recommend you either to read his books (Skin in the Game, Antifragile, etc. etc.) or wait for his next inevitable book which I'm sure will include the IQ discussion. Taleb has a tendency to put things in superficially coarse phrasing online, but his books are lucid beyond comprehension.