That's what he is saying, yes. It's a dangerous thing to think because we do not understand if there is actually some hidden purpose to the genes that modulate

That's what he is saying, yes. It's a dangerous thing to think because we do not understand if there is actually some hidden purpose to the genes that modulate autistic behavior, same with depression. It's a strange thing to bring up after assortative mating, really. If it comes up in assortative mating it implies there is a reason for the genes to be there. One example: In various experiments related to resource exhaustion, antisocial and/or "autistic" mice were the ones to survive the collapse of mouse society as they were more likely to not be involved over resource squabbles. Of course, there are some people who are very autistic who do not have very happy lives, and some people who are just kind of autistic.