V_V comments on AI risk, executive summary