A Ray comments on ML Alignment Theory Program under Evan Hubinger