What to do about IQ

How I learned to stop worrying and love the test.

American soldiers taking the Army Alpha Test. Source: US National Archives.

I was administered an IQ test when I was a kid. All I can remember is a nice lady in a green room who asked me how far I thought China was from here. I guessed four thousand miles. I remembered America was two thousand miles across and on a map it seemed like China was about two of those away. Wrong. But she seemed impressed, anyway.

In some ways, IQ as a concept is obvious to the point of tautology. If I sit you down and give you an abstract goal, say, to do well on this test, that would be an indication of your ability to achieve any abstract goal, writ large. I mean, obviously. Turns out, that’s roughly g, or general intelligence. It’s the most stable, replicable, and useful measure that psychology has ever produced.

Despite that clarity, or perhaps because of that, it is rejected, despised, feared, misunderstood, weaponized, and willfully contorted by public perception. We redefine it over and over with nice-sounding bullshit like EQ, trying to escape its cold number. In the shadows, it is inflated beyond what it could possibly indicate, like the size of your soul. To assuage, people claim that there must be another, more mysterious sort of intelligence. You could say, oh, well, I’m emotionally intelligent, or I have common sense, but those are just another means to achieve an abstract goal. If they work, then they load on g, which is, again, merely your ability to achieve a measurable goal through different means.

Then there’s the IQ fanatics, who believe that because so much of what we aim to do loads on this one measure, that must mean the measure itself somehow constitutes consciousness, or what it means to be a person, or intrinsic worth. This is an even dumber mistake than the former, made obvious if you’ve ever spent any time with people with extremely high IQ. They are impressive, yes. Fast, facile. Physically well constituted, even. You’ll also notice that they’re not obviously more human. In fact, the great danger is taking too much pride in IQ and using it for mere personal gain, and not, as it were, giving it away.

The extreme: Ted Kaczynski. Genius IQ, psychotic and resentful. Wants to be elevated above others because of his IQ, becomes the unabomber. Then you have your Albert Einsteins, your Mozarts, who, for whatever reason, develop a commensurate ability to be childlike as a counterbalance to their extraordinary powers, grateful toward the social systems that allowed them to exist, humble enough to work slowly from within a structure set up by people dumber than them. High IQ is only materially useful over multiple generations when its products are given back to the tribe as a gift. Understood this way, IQ isn’t really a phenomenon of one individual, but is a group-level pattern of bets.

Still though, IQ is a scary thing. I can’t quite get over it. It’s so judgmental. You can even get a good sense of someone’s IQ with a quick assessment, with only a small margin of error. Test something like their vocabulary, or how well they can rotate shapes in their head, and there you go. It terrifies, I think, modern individuals who want to believe we are in charge of our own destiny and can “grow” infinitely. We all want to be outliers, the Chosen One. We also hate when it implies certain groups having different average IQs, or whatever other radioactive implication.

IQ is the Logos of the great cult of the modern day. We’ve named our great machines, after all, artificial intelligence. We pride ourselves above all on our score (which most of us are too afraid to measure), and yet fret and pray about the day AGI comes along and renders even the mightiest of our IQ scores puny by orders of magnitude. But already, these things score better on our most well-devised IQ tests than almost anyone. We’re basically just standing around, waiting to see what that means.

Now, I know what it means as little as anybody. But a strange assumption is that you can scale a measure like IQ linearly outside of human relationships without it losing meaning or efficacy, or at least accumulating diminishing returns. High IQ seems to be a contextual interpersonal phenomena, a product of the very tension between society and rebellion. By definition, rare. It cannot become a universal without defeating itself. Even within the animal kingdom, humans aren’t the smartest because we have the most grey matter of all time, but because we developed complex interdependent language systems and then learned to make social sacrifices. I get no sense of what IQ means without what the holder is using his big brains to do and for whom.

An IQ test, converging as it does so perfectly on g, still does not actually tell you all this interplay that gives rise to exceptional human intelligence. Nor does it tell you what that human intelligence is aimed at. It’s like a statistical analysis that can intake a quarter of a second of Mozart and identify it as Mozart. That is impressive, but it doesn’t tell you what gives rise to Mozart or what Mozart is for. In fact, it can give a dangerous illusion of mastery of Mozart and tempt you to do away with anything but piano fingers. The emissary, in McGilchrist’s terms, tries to become the master. The emissary loves abstractions that give him power, the master prefers ideals like virtue and beauty, which we’ve stupidly lost a taste for due to their immeasurability.

By naming g, we feel like we have a grasp of it, as it converges around this one handy number. We easily succumb to the delusion that the abstraction means everything that what it is abstracted from means. Of course, it doesn’t. IQ is just a snapshot of many converging phenomena, some of which are “computational power” but much of which is vast combinations of known and unknown histories of ancestral telos and unconscious pantheons of psychologized gods that deem a multiple-choice test worth enormous cognitive strain. The problem is not that IQ is wrong, it’s only when we assume that the snapshot is the thing.

This is the motivation of the honorable, but usually technically incorrect, impulse to discredit IQ as a measure. Nassim Taleb, for example, tries to prove statistically that it reliably measures unintelligence but not high intelligence. Anywhere above average it fades into noise. In cases like lifetime earnings, for example, the top 1% of earners score slightly worse on IQ than the ones just below them. But money is not the only thing about IQ. There are scientific and creative benefits all the way up the scale, as the top .1% of scorers win more nobel prizes and such than do the top 1%. As long as you’re still within the relational interplay of other people, there seems to be only a soft upper limit to how meaningfully smart you can get.

Sociologists and psychologists have famously failed to do anything to reliably raise the intelligence of individuals, too. Study and practice only make you better at what you study and practice, but don’t raise g. There are some ecological indications that it can be raised across populations. The Flynn effect over the 20th century might be attributable to better nutrition, or to more years of schooling. Both seem to be plausible from the data. Genetics, certainly, is a strong factor. But there are also things going on at a much larger scale that confound the measurers of IQ, namely that if you try to optimize for IQ through the measurement of IQ, you will only do damage to your ability to produce exceptional IQs. We’ve had, for example, far fewer geniuses over the last 100 years or so, just as measures of IQ have become standardized. You’d assume the opposite would happen after the discovery of a measure as stable and useful as g. Really, you’d hope it would trigger a renaissance of genius. It didn’t.

After all, people with exceptionally high IQs have always emerged, somehow, from normal people. In fact, people with exceptionally high IQs might even be dangerous to the well-being of general populations. They are only tolerated and provided for as an anomaly because they occasionally produce something exceptional. The family lines of Shakespeare and Joyce are not only gone, but perhaps destroyed by their own generational genius. They can at least make up for it by being a father to us all, spiritually. A great sacrifice of exception that we cannot make the general rule without killing ourselves.

Fixation on the IQ of a single individual makes us fail to notice and appreciate these larger patterns that are not causal in only one direction: the human animal quickly developed a high IQ much faster than selective pressure on genetics would seem to indicate is possible — we shape the environment that shapes us, the patterns of culture put pressure on our genes. From Levin’s recent work, even in individual bodies, patterns above the genome select what the genome builds. This is not something that suddenly ended, so your “genetics” is not the only limit on your potential for transformation, nor your children’s, just like it obviously didn’t for our ancestors. The felt limits are likely more socio-spiritual than wetware.

Too much obsession with high IQ also devalues normal experience and ordinary people. It plays this game of trying to optimize for outliers, which is disruptive for the very stablity that gives rise to outliers, and so is also self-defeating. Now measurable in the 21st century: the reverse Flynn effect. A culture obsessed with everyone being a eccentric genius contributes to lower familial stability, for example, paradoxically lessening the chance of fostering an exceptional outlier here and there.

I wanted to be one of these outliers, but even in the microcosm of my being, I find that whatever genius tendencies I may have are much more likely to emerge when going about living an ordinary life. If not in myself, then maybe in my children, or in their children.

IQ, in this way, is too short-term a time horizon to build a society around. It’s an interesting measure, and it’s probably most useful when measuring diminished intelligence, so we can better accommodate those people. But no truly great outlier would need a test to confirm his outlier status. The numbers provide no theory of true transformation, which is the minimum required of greatness: geniuses will always be those few who confound and reshape measurement trends that the rest of us later assume to be inevitable. Even though IQ may be quite predictive of those extremes in hindsight, contrary to Taleb’s thesis, knowing it is much less useful toward achievements of true genius, affirming his intuitions.

On top of that, artificial intelligence might be about to fully commoditize intelligence. As the apotheosis of the IQ-as-religion worldview, it will either be the end of what makes us human, or much more likely, clearly reveal what exactly we are besides intelligence.

Strangely, now is the best time ever in human history to quit optimizing for IQ, in ourselves and in our children. If commoditized with no subsequent apolocapyse, it was perhaps never quite as valuable as we thought.

Focus on what remains: What increases over time when consciously aimed at: Virtue.