• XLE@piefed.social
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    2
    ·
    3 hours ago

    Just assuming this is all true (i.e. that AI can do good and bad code outputs), why would Linux development be able to succeed at something that Microsoft (which has an insider track with AI, far more money, and far more maturity) failed at?

    • kreskin@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      2 hours ago

      Could be a lot of reasons. A big one i see working at a large company myself is that AI needs to draw from a lot of data to do its work. A huge amount of contextual data too. A company like MSFT inevitably needs to provide AI with a walled-off curated set of data, and prevent any of it from leaking. Its AIs will not have the same amount of data an AI can draw from outside MSFT.

      • XLE@piefed.social
        link
        fedilink
        English
        arrow-up
        1
        ·
        43 minutes ago

        Leaking? Microsoft basically owns OpenAI. They pull the data in and don’t need it to go out. The whole industry is fighting to close off competition, meaning they know they’re on top.

        So do you have any reason to assume the open-source community’s use of these (closed-source) other models is somehow bucking all real-world evidence to the contrary, or are we just hoping and praying?