I think this framing also aligns with the discussion of whether we should focus on making AI corrigible or virtuous.
I think this framing also aligns with the discussion of whether we should focus on making AI corrigible or virtuous.