Emil Kirkegaard has an interesting blog post showing the polygenic education scores (MTAGeduPGS) of different U.S. ethnic groups. Whites 0.47, Native Americans -0.37 and U.S. blacks -1.37. If we set the white polygenic score to equal IQ 100 and the black one to equal IQ 85, then:

Genetic IQ = 8.15(MTAG eduPGS) + 96.17

Applying this formula to Native Americans gives a genetic IQ of 93.

But since both U.S. blacks and Native Americans are about 25% white, correcting for this suggests “pure” blacks and Native Americans would be 80 and 91 respectively.

From this we might infer that before modern humans left Africa 70,000 years ago, our average genetic IQ was no higher than 80, but by the time we reached the arctic 40,000 years ago, some races were averaging 91.

But it was only after the neolithic transition that triple digit IQ races began to appear. Indeed a recent study by Davide Piffer and Emil Kirkegaard looked at 2,625 European genomes and found that polygenic scores for education (a proxy for genetic IQ) gradually increased over the last 15,000 years.

Assuming Upper Paleolithic Europeans had IQs like Native Americans (both cold climate hunter-gatherers), this suggests genetic IQ in Europe has been increasing by 0.0006 points per year since near the end of the ice age, culminating in the Industrial Revolution a few hundred years ago.

Piffer and Kirkegaard believe the increase was caused by higher IQ farmers from the Middle East replacing the indigenous hunter-gatherers of Europe but Peter Frost argues they were not necessarily replaced by farmers, but may have evolved into them, writing:

There is thus an inevitable confound between hunter-gatherer ancestry and natural selection due to hunting and gathering. If we look at alleles that seem to indicate native hunter-gatherer ancestry, we are excluding alleles from hunter-gatherers who successfully adapted to farming and who thus acquired a genetic profile that converges, to some extent, on that of Anatolian farmers.