jayterwahl comments on MIRI announces new “Death With Dignity” strategy

[jayterwahl]

(I predict that would help with AI safety, in that it would swiftly provide useful examples of reward hacking and misaligned incentives)