Nassim Taleb is still wrong about IQ

Blocking your opponents isn't the route to knowledge.

Written by Noah Carl.

Back in 2019, Nassim Taleb posted a diatribe against intelligence research titled ‘IQ is largely a pseudoscientific swindle’. He has since updated the piece a number of times, inserting new paragraphs and screenshots to produce what can only be described as an aesthetic mess. Less important than the aesthetics, however, is the science — which Taleb gets badly wrong.

The central argument in the original version was that IQ “mostly measures extreme unintelligence”. Yet zero supporting evidence was provided. A few cherry-picked datapoints were added to subsequent versions, such as a table showing that the death rate from motor vehicle accidents in Australia flattens out above an IQ of 100. (Taleb wants you to take his word for it that what’s true of motor vehicle accidents in Australia is true in general.) However, it continued to rely heavily on simulations and theoretical points.

As evidence that the piece really was a diatribe, rather than a good-faith critique, Taleb referred to “racists/eugenists, people bent on showing some populations have inferior mental abilities”. He also casually smeared Charles Murray as a “mountebank”. Ironically, just a few months earlier, Taleb had published an article on “bigoteering”, which he defined as a “shoddy manipulation” to force your opponent to spend time explaining why they’re not a bigot. So Taleb says “X is bad” and then a few months later does X himself.

‘IQ is largely a pseudoscientific swindle’ has already been the subject of several rebuttals, including a particularly thorough one by Sean Last. However, I’d like to revisit the article — for two reasons. First, a number of relevant studies have been published since it was originally posted. And second, the damned thing keeps going viral on social media, most recently thanks to a fawning thread by self-proclaimed “statistics person” Kareem Carr.¹