Kaj_Sotala comments on Emergent Introspective Awareness in Large Language Models