Gunnar_Zarncke comments on Selfishness, preference falsification, and AI alignment