• untorquer@quokk.au
    link
    fedilink
    English
    arrow-up
    117
    ·
    3 days ago

    Work paid for ChatGPT and asked us to all use it. Not a hard requirement but they were insisting we try to use it to see if it could speed up work.

    I’m in a Nordic country. I haven’t had an email longer than a single paragraph. Documents are already as concise with good organization. The documentation is optimized so AI “summaries” only translates it into longer form tech bro speak. Just looking it up directly is faster.

    I needed to do batch operations on database files without providing admin access. The same database we all work on. It suggested i could do so with a script. This was a flat out lie, with multiple responses hitting the same block. I felt sick about using it so i dropped it for fear of them attempting to try it with admin access.

    Haven’t used it since those few prompts. Yet i get monthly emails talking about how we’re running out of credits every month. I’m guessing one of my coworkers has an AI romance…

    • vithigar@lemmy.ca
      link
      fedilink
      English
      arrow-up
      71
      ·
      3 days ago

      The company I work for paid for Claude and has made it mandatory that we use it. It’s actually insane.

      We also have “Code Rabbit” leaving LLM reviews on our PRs. It’s hilarious how often it completely misses the actual change and just waxes eloquent on the formatting changes that our linter enforces when we commit a change to a file that hasn’t been touched since prior to our current linter rules.

      Even crazier, I’ve had a VP complain to me directly how much he hates reading LLM generated text that someone just spat out at him.

      Yet here we are, still required to use it.

      • untorquer@quokk.au
        link
        fedilink
        English
        arrow-up
        5
        ·
        2 days ago

        Damn that sucks.

        In glad I’m still mostly insulated from it by sector and good old enterprise class lack of investment in modernization.

      • Carrot@lemmy.today
        link
        fedilink
        English
        arrow-up
        3
        ·
        2 days ago

        My favorite part of AI PR reviews is that I’ll fix one of its comments, which leads to a new comment. I fix that one and the new review suggests I actually try just doing the implementation that I started with

    • Visstix@lemmy.world
      link
      fedilink
      English
      arrow-up
      30
      ·
      3 days ago

      “see if it could speed up work” is at least a reasonable request. I do wonder what the rest uses it for.

      • untorquer@quokk.au
        link
        fedilink
        English
        arrow-up
        7
        ·
        2 days ago

        Yeah the company is treating it as an “exploration” but the experiment turned into a regular sub without much scrutiny like a streaming service you forgot you had opted into the trial for.

    • fizzle@quokk.au
      link
      fedilink
      English
      arrow-up
      11
      ·
      3 days ago

      This is my experience, but it seems polarising.

      In the recent thread here about the doctolorow article an army of AI dweebs were proselytizing our great AI overlords.

      I want to be open minded, and will remain so, but as yet I haven’t seen any significant clear benefits to the use of AI.

      My observations are telling me it’s all emperors new clothes, but listening to the commentary of anyone not on Lemmy it would seem I am mistaken. I’m not arrogant enough to believe I’m right and everyone else is wrong, but this remains a mystery to me.