@SuperDicq@minidisc.tokyo I have pondered this a bit and I am not sure if an LLM would increase immersion that much while there are risks it could degrade the experience if it makes irrelevant comments or even harm the user if it makes hurtful ones. Given that some use cases have to do with mental health, you gotta be extra careful.
If they're implemented, I would make LLM-generated comments an opt-in feature, but I would say this is a low-priority feature.
Embed Notice
HTML Code
Corresponding Notice
- Embed this notice
(NekoSock) Miya Ironami (iro_miya@mk.absturztau.be)'s status on Saturday, 18-Jan-2025 22:05:35 JST
(NekoSock) Miya Ironami