Latest generation of products not becoming part of people’s “routine internet use”, researchers say.

  • happybadger [he/him]@hexbear.net
    link
    fedilink
    English
    arrow-up
    12
    arrow-down
    1
    ·
    5 months ago

    Student programmers are the only group I’ve seen that seem to consider it useful. It’s useless in plant science without the rigid sourcing that academic/STEM work requires. Everything it says is a surface-level formulaic response which is clearly mixed together from a dozen unreliable blogs/reddit posts/random articles. It’s just having a clueless coworker that googles things for me, but I still have to type in the same prompt I’d google and read an untrustworthy answer of similar length to a good one.

  • jroid8@lemmy.world
    link
    fedilink
    English
    arrow-up
    10
    ·
    5 months ago

    I’m in a country which not only I don’t have access to it but also can’t really afford a subscription. Are GPT-4o or Claude Opus as smart as they say? Can it come up with creative solutions and solve difficult problems which isn’t in it’s dataset?

    • addie@feddit.uk
      link
      fedilink
      English
      arrow-up
      14
      arrow-down
      1
      ·
      5 months ago

      They’re bullshit generators, essentially - it doesn’t matter to them whether they generate something that’s ‘true’ or not, as long as it’s plausible. Depends what you intend to use them for - if you want a throw-away image for a powerpoint slide that will only be looked at once for a few seconds, they’re ideal. They generate shit code and boring, pointless stories, so couldn’t recommend them for that.

      If you’re a D&D GM that’s in need of quite a lot of ‘disposable’ material, they’re alright. Image of a bad guy that you can then work into the story? Great. Names for every single Gnomish villager? Great. Creating intricate and interesting lore that brings your world alive? No, they are not actually intelligent and cannot do that - that’s the part that you provide.

      At the moment, huge amounts of venture capitalist money is making these things much cheaper than their true cost. Can only imagine the price of them is going to go up a lot when that runs out. You might not be able to afford the subscription, but you’ll be in good company soon.

      • kakes@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        3
        ·
        5 months ago

        To address your last point, there have been continual improvements in making LLMs more efficient as well. While I definitely couldn’t run GPT4 on my local machine, I can get decently close to something like 3.5-turbo, which would’ve been unheard of only a year or two ago.

        And from the sounds of it, GPT-4o is another big step in efficiency (though it’s hard to say for certain).

      • Warl0k3@lemmy.world
        link
        fedilink
        English
        arrow-up
        7
        ·
        5 months ago

        They’re a useful tool (unlike me, a useless tool) but man are they being oversold :(

  • Destroyer of Worlds 3000@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    9
    arrow-down
    1
    ·
    5 months ago

    I have more than 80% of my credits on two different imaging Ai products. Prompts and crossing your fingers hoping you get what you want got pretty stale quick. Chat GPT never appealed to me. It seemed less about output than what it could gain from user input. The whole thing seems adversarial to the end user, like “if you don’t like it, your not doing it right” kind of attitude. I think the push to get us to adopt it in everyday use is to lower the responsibility and liability of those pushing it. It’s a bubble like all tech fads, but it will leave some sort of scar on our day to day.

    there is a great episode of Silicon Valley where this shitty billionaire uses a Siri like voice bot to control his son. that way he doesn’t have to be the bad guy when he wants his son to do something. this is exactly the power of Ai our corporate masters crave.

    (Scarlet Johansson voice) Sorry you were denied an extension on you emergency medical loan. How would you like to pay? Don’t forget to take our survey at the end of the transaction! -Medical Credit Authority Assistant powered by Medi-Ai

  • Xavienth
    link
    fedilink
    English
    arrow-up
    5
    ·
    5 months ago

    I find them useful for getting over that initial hurdle of writing documents. I can just write down bullet points, tell ChatGPT to turn it into text, and then revise almost all of what it writes because I don’t like the style and it misinterpreted a bullet point or two. I find that a lot easier than writing from scratch.

  • Daxtron2@startrek.website
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    5 months ago

    Its the classic example of people following the trend and repeating opinions they heard while never actually trying to effectively utilize a thing before deciding to hate it.