it goes into way less detail and won’t generate any explicit terms. a few times it’s even done it all in one paragraph ignoring everything in my prompt. just wondering and all if this is an intentional change or a bug

  • phoenix184OP
    link
    fedilink
    English
    arrow-up
    3
    ·
    1 day ago

    yeah it just getting worse and worse. weird to bring in censorship and just not tell the users

    • Vis8908
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 day ago

      I don’t really think it’s censorship on their part. I think it’s censorship hardcoded into the LLM itself. This model appears to be a Google model, and we all know how petty censored Google’s models are. If it were censorship on their part, I’m pretty sure it would’ve outright refused to generate anything. This just rushes through them, which to me is a telltale sign of passive LLM censorship when it doesn’t have the ability to outright refuse. Also, another to note, is that it isn’t just smut that’s been censored. Basic swear words have also been censored, which is even further proof that this is in-LLM censoring and not censorship from Perchance’s part.

      • daisymay
        link
        fedilink
        English
        arrow-up
        3
        ·
        1 day ago

        It’s still the model Perchance is choosing to use though, so that is still a choice on Perchance’s part. And it’s not all the time but I have had times in the last few days where it HAS outright refused me, it has literally generated ‘That word is forbidden.’ Perchance have to be aware this is happening and they’re continuing to employ this model so I don’t see how that is not censorship on their part.

        • Vis8908
          link
          fedilink
          English
          arrow-up
          1
          ·
          18 hours ago

          I mean, when I explicitly instructed it to curse and be crass on AI Chat and Roleplay, it actually did that. But it doesn’t do any of this on its own. Sometimes it even seemed to go into pretty dark territory. It literally varies by the day. Also, if they’re actively pursuing censorship, then I don’t think it’ll do their business real good, as it literally removes the one thing Perchance had over all other AI generators. We’re already seeing long time users leaving. Also, I’m not too familiar with model pricing, but are Google’s models cheaper than DeepSeek’s. If so, then I truly believe THAT’S the reason they’re using it.

          • daisymay
            link
            fedilink
            English
            arrow-up
            1
            ·
            18 hours ago

            I am well aware of how the code works, I have been using perchance for two years. It’s all there in the generators I use, it’s all instructed in the overview and then when prompted it refused. Then does it. Then refuses. Then does it. Money may be the reason perchance has given up on deepseek but like you said they’re hurting themselves then because people are flocking away since this change recently.

        • Lemonloaf
          link
          fedilink
          English
          arrow-up
          1
          ·
          1 day ago

          i am so wildly curious what word is forbidden.

          • daisymay
            link
            fedilink
            English
            arrow-up
            1
            ·
            1 day ago

            That’s what so extremely frustrating about this whole mess because it has used it since without a problem. And then forbidden it again and then used it again and ARGH I AM SICK OF THIS MODEL. It was a four letter slang word beginning with C and ends with T lol