Emil O W Kirkegaard

The idea that these studies are spurious simply because they don't identify causal pathways is just crackpot. Sure, causal pathways are good. Why not? And it may be that some of the genes identified actually do not have the right sort of causal role. But all? Most? Why?

despite ridiculously high sample sizes.* Believe what you want, I think I have good reasons to believe nothing interesting will come from this method and already commercializing genetic intelligence predictions is an outright scam.

@KirkegaardEmil Your call. I bet 100 euros that the varaince explained by GRS for IQ or educational attainement will be divided by at least 5 in a couple of years when fine scale population structure is taken into account.

Look, the paper just came out, without peer review, apparently, and hasn't itself been subjected to criticism. God only knows how much of it will survive the process. Seriously, do you have any argument that these sorts of studies are doomed in general, regardless of the trait?

It ain't the first paper that show the same thing. There are published papers and theoretical reasons to expect this problem to be ubiquitous. Naive biological models. Except rare cases of mandelian diseases, partitioning and additive models will fail for most traits.

Are you really going to claim that standard quantitative genetics is in general a failure? I mean, if it is, why does it succeed so well in breeding, say, livestock?

As for quantitative genetics being a failure, it's not a settled down issue. But it's been at least a disappointment. Especially the fact that genomic research technology hasn't been able to support the old models like researchers thought it would.