Comment on OpenAI confirms that AI writing detectors don’t work

GlendatheGayWitch@lib.lgbt ⁨1⁩ ⁨year⁩ ago

• vale@sh.itjust.works ⁨1⁩ ⁨year⁩ ago

  Then you have that time that a professor tried to fail his whole class because he asked chatGPT if it wrote the essays.

  wgntv.com/…/professor-attempts-to-fail-students-a…

  source

  • wedeworps@sh.itjust.works ⁨1⁩ ⁨year⁩ ago

    Could you please provide a brief overview? This article is not available in my country/region.

    source

    • T156@lemmy.world ⁨1⁩ ⁨year⁩ ago

      It cites this article, which might work you.

      source

      • wedeworps@sh.itjust.works ⁨1⁩ ⁨year⁩ ago

        Thank you very much

        source
• 4AV@lemmy.world ⁨1⁩ ⁨year⁩ ago

  It doesn’t have “memory” of what it has generated previously, other than the current conversation. The answer you get from it won’t be much better than random guessing.

  source

  • randint@lemmy.frozeninferno.xyz ⁨1⁩ ⁨year⁩ ago

    Maybe it should keep a log of what was generated? Would that even work though?

    source

    • sep@lemmy.world ⁨1⁩ ⁨year⁩ ago

      Ignoring the huge privacy/liabillity issue… there are other llm’s then chatgpt.

      source
    • zikk_transport2@lemmy.dbzer0.com ⁨1⁩ ⁨year⁩ ago

      Llama 2 also exists. You can’t control that.

      source
    • BetaDoggo_@lemmy.world ⁨1⁩ ⁨year⁩ ago

      The model is only trained to handle 4k tokens, roughly 2000 words depending on complexity. Even if it had a log of everything asked it wouldn’t be able to use any of it.

      source
• mwguy@infosec.pub ⁨1⁩ ⁨year⁩ ago

  No. The model doesn’t have a record of everything it wrote.

  source
• echodot@feddit.uk ⁨1⁩ ⁨year⁩ ago

  That doesn’t really work because it just says whatever half the time. It’s very good at making stuff up. It doesn’t really get that it needs to tell the truth because all it’s doing is optimising for a good narrative.

  That’s why it says slavery is good, because the only people asking that question clearly have an answer in mind, and it’s optimising for that answer.

  Also it doesn’t have access to other people’s sessions (because that would be hella dodgy) so it can’t tell you definitively if it did or did not say something in another session, even if it were inclined to tell the truth.

  source
• Kazumara@feddit.de ⁨1⁩ ⁨year⁩ ago

  Obviously not. Its a language generator with a bit of chat modeling and reinforcement learning, not an Artificial General Intelligence.

  It doesn’t know anything, it doesn’t retain memory long term, it doesn’t have any self identity. There is no way it could ever truthfully respond “I know that I wrote that”.

  source