OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling’s Harry Potter series::A new research paper laid out ways in which AI developers should try and avoid showing LLMs have been trained on copyrighted material.

    • stevedidWHAT@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 year ago

      https://medium.com/@rxtang/the-science-of-detecting-llm-generated-texts-e816a14c18d

      But yes it is tho ;)

      “Existing detection methods can be roughly grouped into two categories: black-box detection and white-box detection, black-box detection methods are limited to API-level access to LLMs. They rely on collecting text samples from human and machine sources, respectively, to train a classification model that can be used to discriminate between LLM- and human-generated texts. An alternative is white-box detection, in this scenario, the detector has full access to the LLMs and can control the model’s generation behavior for traceability purposes. In practice, black-box detectors are commonly constructed by external entities, whereas white-box detection is generally carried out by LLM developers.”

    • stevedidWHAT@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      1
      ·
      edit-2
      1 year ago

      Are you going to bother explaining or are you just here to be contrarian and pedantic.

      If so, please do, I love learning new things! Otherwise, I’m not really interested, my general over arching point stands

      From my understanding training a model on a dataset of real world password permutations would make guessing those passwords significantly faster (by factors) as opposed to doing a full brute force.

      • redw04@lemmy.ca
        link
        fedilink
        English
        arrow-up
        1
        arrow-down
        1
        ·
        1 year ago

        Nah, you can go research it yourself and stop spreading misinformation online. Just because you don’t understand how something works doesn’t give you the right to be angry at it.

        • stevedidWHAT@lemmy.world
          link
          fedilink
          English
          arrow-up
          1
          ·
          1 year ago

          I don’t need to, I use models to do this stuff regularly lmao.

          You have a misunderstanding which is apparent by your inability to actually explain. This type of response is also commonly used by trolls to just “no you” repeatedly.

          Modeling predictive text based on a data set is literally LLM basics lmao.