Scientific consensus is pretty heavily on the side of "50+% of within-group variation in intelligence is genetic, the remainder is non-shared environment" (in the US, barring known negative confounders such as fetal alcohol syndrome), and as far as I know the science on between-group variation is still very much undecided, but genetic components are as far from ruled out as you can get. But that debate won't be settled until you can 1) find the environmental differences that lead to IQ differences between groups, including (IIRC) in adopted children, 2) show that IQ tests only correlate with g within ethnic groups, but not across them, or 3) decode the genetic basis of IQ and show how differing frequencies explain group differences. Which is going to take a number more years, no matter which it ends up being.