• DragonBallZinn [he/him, they/them]@hexbear.net
    link
    fedilink
    English
    arrow-up
    26
    arrow-down
    1
    ·
    edit-2
    22 hours ago

    If you frowned at that, remember that Linus Torvalds also supported AI-code in Linux, saying: “AI is a tool, just like other tools we use. And it’s clearly a useful one. It may not have been that ‘clearly’ even just a year ago, but it’s no longer in question today.”

    a-guy

    • hello_hello [hy/hym]@hexbear.net
      link
      fedilink
      English
      arrow-up
      3
      ·
      4 hours ago

      They always cite this and not Greg KH (co-maintainer with Torvalds) saying that a third of all LLM submissions for security bugs are dead ends and trying to restrict the overwhelming amount of machine generated patches which have only doubled or tripled their normal workload. Or that these models are being used to review a immensely large C codebase where blinking the wrong way means someone can pwn an entire system in 100 lines of code.

      It’s because they’re fucking posers and should take the penguin operating system that they’ve purposefully underfunded out their mouth. AI boosters need to be purged.

    • SnoopSqueak@lemmy.today
      link
      fedilink
      English
      arrow-up
      17
      ·
      22 hours ago

      Microsoft is magnitudes worse at using “AI”, but I still question whether it’s a worthwhile tool. I wonder how long it will be before someone gets a little lazy and vibe codes severe vulnerabilities into the Linux kernel. Yes, humans can also make mistakes, but there is more accountability. There is reason behind their decisions, at least a little more than a pseudorandom number generator.