[sulam]

The idea of screening for a software engineering gene is patently absurd. In an ideal scenario, we screen for the ability to do the work the job requires. In the normal scenario, we screen for some mix of that and the ability to apply cookie-cutter techniques to (hopefully) novel problems in the space of ~45 minutes. Some companies also ask you to do a sample practical problem either on site or on your own time. None of this indicates anything about someone's "innate genetic talent", if there is any such thing.

[Blazespinnaker]

To imagine that genetics do not play a role in intelligence (or any talent) is horribly naive, otherwise we could teach our dogs to program.

[sulam]

And it’s illegal to try and screen on anything except ability to do the job. Be very careful if you find yourself thinking in terms of software engineering interviews being a general intelligence test — you are setting yourself up for a discrimination suit.

https://en.wikipedia.org/wiki/Griggs_v._Duke_Power_Co.

[Myrmornis]

As you clearly demonstrate, intelligence has a genetic basis. However, it does not follow that there is variation among humans at the relevant genetic loci. The variation may have become fixed in an ancestral population of modern humans.

[Myrmornis]

What I gave there is an absolutely standard assessment of the situation from the point of view of population genetics. Of course intelligence "is genetic" in some sense: that's why dogs and cucumbers are less intelligent than humans for most definitions of "intelligent". The question is about the phenotypic significance of current intraspecific genetic variation.