Jan_Kulveit comments on Talking publicly about AI risk