3 ms·

I believe we are currently measuring recent functional selection on humans and it's non-zero. That is, empirically you are incorrect.