• rumschlumpel@feddit.org
    link
    fedilink
    English
    arrow-up
    6
    ·
    2 days ago

    Any job that is actually productive rather than doing stuff that no one wants or needs is an “actual” job, and I just don’t see how you can replace so many people with LLMs without severely affecting the quality of the work, even if it’s grunt work done by juniors.

    • agamemnonymous@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      3
      ·
      2 days ago

      There are tasks that are necessary but tedious. These are tasks that either consume the time of experienced professionals, or are offloaded to inexperienced professionals when payroll allows.

      Tedious tasks are perfect candidates for automation, especially when the result is much easier to verify than to find. This frees the experienced professional to do interesting work.

      • TheAgeOfSuperboredom@lemmy.ca
        link
        fedilink
        English
        arrow-up
        6
        ·
        2 days ago

        Tedious tasks are also great for assigning to juniors to help them learn.

        I’ll often use tedious tasks as a testing ground for new technologies as well. Tedium leads to innovation.

      • rumschlumpel@feddit.org
        link
        fedilink
        English
        arrow-up
        2
        ·
        2 days ago

        ‘automation’ and ‘AI’ are currently very different things. There are certainly applications where machine learning is very effective, sometimes even better than senior professionals, but LLMs are mostly bullshit machines.

        • agamemnonymous@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 days ago

          They generate a lot less bullshit when deliberately trained on a specific dataset, and they’re only getting better with time.