Jan comments on Reflections on Deception & Generality in Scalable Oversight (Another OpenAI Alignment Review)