Seth Herd comments on Some biases and selection effects in AI risk discourse