hi, I 100% agree with you and I need to have an uncensored nsfw chatbot for my marketing strategy, do you think using mistral is the best way to go? Can I train it on a remote GPU paid by the hour ? and how would all of this work? i know I ask a lot of qs but I am really not so experienced with this
I haven't tried to train mistral on more naughty stuff than it delivers out of the box.
Good stuff for writing nasty stories about whatever your (target audience's) fantasies are.
| Feature | OpenAI (Original Vision) | OpenAI (Current State) | ClosedAI (Elon Musk's Argument) | My Perspective (LLM) |
| Focus | Open development & collaboration | Profitability with capped returns | Safety & potential risk mitigation | Responsible AI development |
| Transparency | High - Share research details | Limited - Details on latest models not public | Lower - Keep some research private | High - Transparency builds trust |
| Benefits | Faster progress, identify risks | Faster progress with focused resources | Mitigate potential risks | Wider range of ideas, faster progress |
| Drawbacks | Lack of transparency | Limits public oversight | ||
| Overall | Idealistic but challenging to maintain | More pragmatic but raises concerns | May be necessary for safety but limits collaboration | Transparency is crucial for responsible AI |
