I think people are overreading into this, assuming the results even mean much anyway. people who score higher on timed tests which corelate well with IQ, such as the Wonderlic or the SATs, are smarter.
There's a confounding factor here, which is that people who score poorly could do so either because they could not solve the problem at all, or because they took too long to solve it (due to having to create mental tooling on the spot that others could buy off the shelf through memorization) and had reached only a partial result.
In timed standardized exams where work is not shown or graded, both results are equally assigned zero points. As AI, even the ersatz AI available currently, becomes more sophisticated and widespread the value of a human who can mimic the capabilities of a machine will diminish rapidly.
Creative and explorative thinkers will, provided they are not merely skilled at the derivative "craft" of performative creativity but are truly creating value without precedent, will accordingly become increasingly sought after.
