somervta comments on Against easy superintelligence: the unforeseen friction argument