As the saying goes, the best way to convince yourself of AI risk is to look at the quality of the arguments against it.