• mlg@lemmy.world
      link
      fedilink
      English
      arrow-up
      15
      ·
      23 hours ago

      Train it exclusively on Linus’s entire mailing list history so that it actively insults poor merge requests and code lol.

    • TehPers@beehaw.org
      link
      fedilink
      English
      arrow-up
      1
      ·
      16 hours ago

      This is how many spam filters have worked for a very long time. Using an LLM to help with classification could be a real option.

      It wouldn’t help with determining which security reports are slop and which are real, though.

    • Denixen@feddit.nu
      link
      fedilink
      arrow-up
      10
      arrow-down
      3
      ·
      1 day ago

      Ironically, this is probably the way it is going to go. As we keep going deeper into the information age, managing all the information thrown at us is becoming unmanageable. With the introduction of AI it will get worse, even as AI becomes the only way of handling the influx of information. Using machine learning has been a way to sort information for a long time (spam sorting is a type of AI after all) and present generative AI are an evolution of that.

      • fizzle@quokk.au
        link
        fedilink
        English
        arrow-up
        6
        arrow-down
        1
        ·
        22 hours ago

        I dont think thats necessarily true.

        FWIW im team “AI can be a useful tool”, but on the day when other people’s chat bots start emailing my chat bot who summarises it for me im just going to go throw my laptop in the river.

        I think Linus has kind of addressed it: if you found it with a tool then we already found it.

        Maybe im being naive. IDK. Im so sick of tech just generally.