Indeed. It's probably what I would do if I were him. Or direct OpenAI entirely into AI safety research and stop doing capabilities research. I watched his interview with Lex Fridman,[1] and I didn't think he seemed very sincere. On the other hand I think there are a lot of people who are very sincere, like Max Tegmark.[2]
[1] https://www.youtube.com/watch?v=L_Guz73e6fw&pp=ygUWbGV4IGZya...
[2] https://www.youtube.com/watch?v=VcVfceTsD0A&pp=ygUXbGV4IGZya...