I don't really see support for that in the article, but most of the details are in the individual studies. The study does say
Kanazawa notes that intelligence correlates positively with evolutionarily novel activities, but the correlation with ancient activities is zero or even negative. That is also evident in Table 25.1 , which mostly lists novel school- or job-related forms of success that have the expected positive correlation with intelligence, but one of the most ancient forms of success, number of children, has a negative correlation (−0.11).
I would guess that you are reading "social success" as "social skills" while in the article's terminology they mean by "social success" things like career success, education level etc.
As for mathematics, it's true that this is a very special case, and people select into mathematics based on both ability and personality. All I can say for certain is that I can identify a distinct group of people who are both intelligent and have autism spectrum traits, and they seem to occur more often than if these traits were uncorrelated.
Kanazawa notes that intelligence correlates positively with evolutionarily novel activities, but the correlation with ancient activities is zero or even negative. That is also evident in Table 25.1 , which mostly lists novel school- or job-related forms of success that have the expected positive correlation with intelligence, but one of the most ancient forms of success, number of children, has a negative correlation (−0.11).
I would guess that you are reading "social success" as "social skills" while in the article's terminology they mean by "social success" things like career success, education level etc.
As for mathematics, it's true that this is a very special case, and people select into mathematics based on both ability and personality. All I can say for certain is that I can identify a distinct group of people who are both intelligent and have autism spectrum traits, and they seem to occur more often than if these traits were uncorrelated.