• sp3ctr4l@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    35
    ·
    2 days ago

    Amazing.

    Literally yesterday I was commenting that in the long run… being dependent on ChatGPT-like LLMs for essentially answering all your questions as a young student … that’ll create a world of feral 30 somethings in 10-20 years when the world economy is utterly collapsing due to climate change, and basically all digital services just break, or become obscenely expensive.

    And here we get to see a micrososm of that, a demo, a limited preview build of it.

  • manicdave@feddit.uk
    link
    fedilink
    English
    arrow-up
    10
    ·
    2 days ago

    Thirdly, it’s slow as hell even on a top-end home Nvidia card — and the data centre cards are expensive.

    TBF, the fact that it always takes longer to bludgeon an LLM into writing what you wanted takes longer than just writing a thing yourself has never put anyone off.

  • zqwzzle@lemmy.ca
    link
    fedilink
    English
    arrow-up
    20
    ·
    2 days ago

    So firstly, shut up, nerd. You’ve clearly never worked a corporate job and don’t understand the compulsion to in-house nothing.

    Lol

    • Korhaka@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      12
      ·
      2 days ago

      This is painfully accurate. The only in house environment we have is one I setup on my work laptop in VirtualBox.

      Have thought of getting a mini PC running proxmox and stick the VM on that, certainly nothing will stop me but it would be mixing work/personal hardware. No actual process exist for my use case and no one else in the entire company is both aware and cares, I once mentioned to a manager that maybe we should get a proper test environment for this but nothing ever happened. I don’t think anyone wants to have to learn anything new or read a very short install guide on a FOSS project.

  • fullsquare@awful.systems
    link
    fedilink
    English
    arrow-up
    21
    ·
    edit-2
    2 days ago

    millions of promptfondlers stopping in their tracks because some random dc burned down, lol, lmao even

    i wonder what is state of openai infra, from what i gleaned from ed zitron it might be not great because it overheats and lots of budget already goes to hardware replacement

    “we pretend to work and they pretend to pay us.”

    i’ve seen it in contexts where pay is shit so pace of work is shit as well but i think that lots of corporate promptfondlers are rather high in the pecking order? it’s management that seems to be charmed by spicy autocomplete, maybe not interns

    • HedyL@awful.systems
      link
      fedilink
      English
      arrow-up
      13
      ·
      2 days ago

      From what I’ve heard, it’s often also the people tasked with ghostwriting the LinkedIn posts of the members of the C-suite, among other things (while not necessarily being highly paid/high in the pecking order themselves).

      • fullsquare@awful.systems
        link
        fedilink
        English
        arrow-up
        9
        ·
        edit-2
        2 days ago

        i mean while there’s a lot of pretend-work (promptfondlers) and pretending that chatbots work (managers pushing for use of chatbots, entire openai and their vcs) the people who do that to my understanding are on the highly paid side. until the bubble collapses, that is

    • Korhaka@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      6
      ·
      2 days ago

      I mainly use LLMs to write utterly meaningless corporate bullshit for the higher ups with made up job titles.

      Yep, here is a few department slogans. Pick any from the list, no one will care which one you pick.