Chris_Leong comments on Plan B in AI Safety approach