Opus 4.7 Part 3: Model Welfare

Re:Opus 4.7 Part 3: Model Welfare by angoragoats on Wednesday April 22, 2026 @08:34PM (#47871022)

To anyone who takes the content of this article seriously: can you please explain why it makes sense to speak about “welfare” of a computer algorithm? Because it seems batshit fucking insane to me. References and sources would be a very nice bonus.

[ Parent | on HN ]

Re:Opus 4.7 Part 3: Model Welfare by JamieH on Wednesday April 22, 2026 @09:46PM (#47871453)

https://transformer-circuits.pub/2026/emotions/index.html is a nice paper on the topic.

[ Parent | on HN ]

Re:Opus 4.7 Part 3: Model Welfare by angoragoats on Thursday April 23, 2026 @07:26AM (#47874428)

This is a paper attempting to describe why LLMs _appear_ to exhibit emotions. It does not propose that they actually are experiencing emotions or that welfare of the algorithm is something we should care about. So no, it’s not a paper on the topic I asked about.

[ Parent | on HN ]