Otto comments on When “human-level” is the wrong threshold for AI