Taleb: IQ is largely a pseudoscientific swindle (Wrong again!)

Nassim Taleb is at it again with a Medium post that expounds on his anti-IQ twitter rant, which I critiqued earlier.

Taleb’s argument, essentially, is that IQ tests only work on the ‘low end’ and not ‘high end’, and that IQ is not a predictor of above-average ability.

Taleb’s post is mostly incoherent. As a commenter puts it, “Did the people that clapped on this read the article? Because I cannot make heads or tails out of anything he’s saying, except that he seems to be shilling stuff he’s written elsewhere.”

It begins by calling Charles Murray a charlatan and implying he’s an enabler of racism (which is a word that is meaningless and devalued due to its definition encompassing everything but actual racism. Now it means the mere suggestion that some races are better at certain things than other races.):

racists/eugenists, people bent on showing some populations have inferior mental abilities based on IQ test=intelligence who have been upset with me for suddenly robbing them of a “scientific” tool (as evidenced by the bitter reactions to the initial post on twitter/smear campaigns by such mountebanks as Charles Murray). (Note: there were close to 3.1 million views of the tweetstorms).

This proves what I have suspected for years, which is that Taleb is a leftist, or, worse, a social-justice warrior, or at the very least is sympathetic to them. Even Sam Harris, who is a liberal, defended Murray against slander by Ezra Klein. And his tweets conflating nationalism with imperialism further confirms that Taleb is not not right-wing, as many mistakenly still believe he is. He’s not right-wing or even of the Ayn-Rand variety of libertarianism, but rather subscribes a sort of socially-egalitarian equalitarian variant anarchism.

psychometrics peddlers looking for suckers (military, large corporations) buying the “this is the best measure in psychology” argument when it is not even technically a measure — it explains at best between 13% and 50% of the performance in some tasks, minus the data massaging and statistical cherrypicking by psychologists; it doesn’t satisfy the monotonicity and transitivity required to have a measure. No measure that fails 60–95% of the time should be part of “science”.

The U.S. military uses IQ for screening recruits. The National Guard, for example, has a much higher IQ cut-off than the Army. This is backed by a huge sample size of recruits and over a century of research going as far back as World War 1, so it’s not just some ‘data massaging and statistical cherrypicking’. In fact, as further evidence of the value of IQ for predicting recruit performance, during the Vietnam War the Army tried lowering the IQ requirement to enlist (called Project 100,000), and the results were disastrous.

If you want to detect how someone fares at a task, say loan sharking, tennis playing, or random matrix theory, make him/her do that task; we don’t need theoretical exams for a real world function by probability-challenged psychologists. Traders get it right away: hypothetical P/L from “simulated” paper strategies doesn’t count. Performance=actual. What goes in people’s head or reaction to a screen image doesn’t exist (except via negativa).

That is the point of having IQ tests in the first place, to avoid having to create a test for every conceivable task. IQ measures an ability to learn. Recruits with higher IQs can lean drills faster than less intelligent recruits and are better at adapting to unpredictable conditions as one would encounter on a battlefield.

“IQ” is most predictive of performance in military training, with correlation~.5, (which is circular since hiring isn’t random and training is another test).

Rather than correlation with job performance, another way of looking at IQ is in terms of minimum required intelligence. If you’re a headhunter for Google, you know a competent coder is almost certainly going to have a high IQ at a minimum, but not all high-IQ people will make good coders. But a low-IQ person will never make for a good coder, no matter what, because he simply will never be able to understand the abstractions. That is why prestigious firms recruit at selective universities, because they know they are choosing from a high-IQ pool of potential candidates.

But the “intelligence” in IQ is determined by academic psychologists like the “paper trading” we mentioned above, via statistical constructs s.a. correlation that I show here (see Fig. 1) that they patently don’t understand. It does correlate to negative performance (as it was initially designed to detect learning special needs) but then any measure would work there. A measure that works in left tail not right tail (IQ decorrelates as it goes higher) is problematic. We have gotten similar results since the famous Terman longitudinal study, even with massaged data for later studies. (To get the point, consider that if someone has mental needs, there will be 100% correlation between performance and IQ tests. But the performance doesn’t correlate as well at higher levels, though the psychologists will think it does.)

Wrong again. Scores on the sub-tests such as verbal and digit recall also follow a normal distribution for all range of scores , so it’s not like IQ tests only adhere to a normal distribution at the low-end or only differentiate ability at the middle and low-end. Furthermore, studies show that children with IQs above 130 need 1/4 the repetition to learn material as average-IQ children, which explains why they are far beyond their peers:

The above passage taken from the book Losing Our Minds: Gifted Children Left Behind, By Deborah L. Ruf, also refutes the argument that IQ is not predictive of ‘real world’ abilities, because the ability to learn is invaluable, especially in an increasingly technical and competitive economy. This can explain why IQ is also highly correlated with job performance. If you’re a waitress, it helps having a good working memory for remembering customers’ orders. At the ‘high end’, success at programming requires making sense of abstractions and being able to memorize certain functions.

A high IQ is predicative of success for professions that can be considered cognitively demanding such as academia and programming. How many successful fiction writers or STEM people are there with IQs below 100? Few, if any, compared to IQs above 120. This is corroborated by studies that show how 13-year-olds who score in the top .25% on the SAT are much more likely to be successful in terms of creative output and educational attainment than those only in the top 1%, which is further evidence that intelligence can be ascertained, differentiated, and quantified at the high-end. Mathematicians, physicists, fiction writers, artists, etc., who have IQs above 120 or so, enrich the world with their creations and discoveries. Is Terrance Tao an underachieving bureaucrat? I think not. Although not all high-IQ people are creative and successful, the odds of an average-IQ person advancing the cannon of human knowledge are slim to none. This shows how IQ, in terms of being a predictor of exceptional ability, is actually more predictive on the high-end than even the low and medium-end.

The most common criticism of IQ some is variant of the claim that IQ does not measure ‘true’ intelligence, but rather is a construct of what the test creator deems to be intelligence. Or that ‘g’ does not exist, or is affected by biases by the test creator, or is not predicative of IQ. However, the critics can never provide a better theory or explanation that explains differences in individual outcomes, except sweeping them under the rug of ‘cultural factors’ even though IQ is better predictor of upward mobility than even background socioeconomic status. Or why so many sub-tests, despise the best efforts to bypass g, correlate (this is called the ‘positive manifold’ as discussed in further detail here).

You can look at it inductively: give every working person an IQ test. Assume you don’t know what IQ is or what it is supposed to test. What you’ll find , with few exceptions, it that people with higher IQs gravitate to certain set of jobs, and those with lower IQs do another set of jobs. People who work in tech, medical, legal, academia, and finance tend of to have a greater amount of this quantity called ‘IQ’ and those that work in lower-paying jobs such as retail have less of it. If IQ is not supposed to predict anything, especially on the high-end, why does this sorting effect exist? It is so consistent that it cannot be a coincidence.

The same people hold that IQ is heritable, that it determines success, that Asians have higher IQs than Caucasians, degrade Africans, then don’t realize that China for about a Century had one order of magnitude lower GDP than the West.

China was able to pull itself out stagnation because of having a high-IQ population and by abandoning market-communism. To get an idea of what happens when a low-IQ country tries a mixed economic system, look at Brazil. It’s better than communism, but still pales in comparison to smarter countries. Low-IQ countries and regions will never catch up to smarter regions and countries on a per-capita basis. There is actually a term for this phenomenon of some countries always being stuck, which is called the ‘middle income trap’, as I discuss in further detail.

I would like to see Jordan Peterson or someone who is an expert refute Taleb in further detail. Dr. Peterson is too polite to confront Taleb beyond the occasional tweet. I am not. We need more people to expose Taleb for the Intellectual-Yet-Idiot he really is (to borrow his own expression).