As for "intelligence" selecting for IQ points doesn't make sense. There's little reason to beleive g is anything but an artifact from factor analysis that's completely unsupported by any neurological data or known causal genetic mechanisms, and in fact a g-like measure can be proven to exist for any set of variables that positively correlate with each other regardless of existing causal links among them. The non-additive interactions with the environment are also just too high. Selecting for IQ would just result in random artificial genome drift (beyond the first obvious gains from developmental defects etc. but we sort of already do this).
I think there's a real disconnect between how HN techies think about genetics and biology and the actual state of the art - it's much less grandiose than they think. Science is generally boring.