Re:Uncensored (Language) Models
by commandlinefan on Monday May 15, 2023 @03:09PM
(#35953133)
> Given how easy "hurt feelings" escalate into real-world violence
Well, not that I'm at all convinced, but that's more an argument for extreme soviet-style censorship rather than a minimalist "make sure AI has a political slant".
Well, not that I'm at all convinced, but that's more an argument for extreme soviet-style censorship rather than a minimalist "make sure AI has a political slant".