• mostvexingparse@lemmy.world
    link
    fedilink
    English
    arrow-up
    28
    ·
    1 year ago

    The best comment I read on this topic was “Before AI replaces us, our customers must first learn to precisely formulate what they actually want. We’re safe.”

      • Holyginz@lemmy.world
        link
        fedilink
        arrow-up
        8
        ·
        1 year ago

        Lmao, I’m not a programmer although i know how, and even if I was I wouldn’t be worried for good reason. AI requires explicit instructions for everything. So in order to use it to code you need to be a programmer.

        • Meho_Nohome@sh.itjust.works
          link
          fedilink
          arrow-up
          1
          ·
          1 year ago

          I’m not a programmer and I’ve used it to code. It rarely works the first time around, but I’m sure it will improve quickly to be more accurate.

      • Knusper@feddit.de
        link
        fedilink
        arrow-up
        3
        ·
        1 year ago

        I’m also not worried. Software complexity generally grows proportional to the complexity of the requirements. And most projects I’ve been a part of, no one could have told you all the requirements even after we’ve figured them out.

        The code + test code is usually the only document that describes the requirements. And with high-level languages, there’s not that much boilerplate around the codified requirements either. Besides, we can use LLMs for that boilerplate ourselves.

      • RandomVideos@programming.dev
        link
        fedilink
        arrow-up
        2
        ·
        1 year ago

        If an AI was made that was smarter than programmers, couldnt it make a smarter AI, which could make an even smaryet AI repeating