Right after making my crux claim, I reference some of Tobias Baumann’s (2022a,2022b) work which gives some examples of how significant amounts of suffering may be instrumentally useful/​required in cases such as scientific experiments where sentience plays a key role (where the suffering is not due to it being a strong motivator for an efficient algorithm, but for other reasons).
I think it would be helpful if you provided some of those examples in the post.
I think it would be helpful if you provided some of those examples in the post.