Gunnar_Zarncke comments on How to get into AI safety research