Riccardo Volpato comments on Risks from Learned Optimization: Conclusion and Related Work