• vala@lemmy.dbzer0.com
    link
    fedilink
    arrow-up
    95
    ·
    1 年前

    User:

    Shop pop, whoppity slop, crops are here, it’s time to stop.

    This is a very famous poem. What is it about?

    GPT-4o mini:

    The lines you quoted are from the poem “The Song of the Witches” by William Shakespeare, specifically from “Macbeth.” This poem is recited by the three witches in Act 4, Scene 1.

    Themes and Meaning Supernatural Elements The poem reflects the supernatural themes prevalent in “Macbeth.” The witches symbolize chaos and the dark forces that influence the characters’ fates.

    It goes on.

    • Patches@ttrpg.network
      link
      fedilink
      arrow-up
      26
      ·
      1 年前

      Actual “Song of the witches” from Macbeth

      Double, double toil and trouble;

      Fire burn and caldron bubble.

      Fillet of a fenny snake,

      In the caldron boil and bake;

      Eye of newt and toe of frog,

      Wool of bat and tongue of dog,

      Adder’s fork and blind-worm’s sting,

      Lizard’s leg and howlet’s wing,

      For a charm of powerful trouble,

      Like a hell-broth boil and bubble.

      Double, double toil and trouble;

      Fire burn and caldron bubble.

      Cool it with a baboon’s blood,

      Then the charm is firm and good.

    • LiveLM@lemmy.zip
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 年前

      Holy shit. After all these years maybe we can finally figure out the true genius of Kanye’s Lift Yourself

  • Xerxos@lemmy.ml
    link
    fedilink
    arrow-up
    56
    arrow-down
    1
    ·
    1 年前

    There was a paper about this not long ago. The problem is, how LLMs get trained: a right answer gets a point, everything else gets no points. This rewards guessing (produces a point sometimes) over answering “I don’t know/I can’t do this” (produces never a point)

    • ipkpjersi@lemmy.ml
      link
      fedilink
      arrow-up
      13
      ·
      1 年前

      It’s like when developers give a wrong answer during technical interviews, rather than say “I’d have to look it up” or “I’d have to check the documentation” etc.

  • NigelFrobisher@aussie.zone
    link
    fedilink
    arrow-up
    49
    ·
    1 年前

    This is actually a pretty great way to illustrate what LLMs do. It gives you an answer regardless of whether it makes sense to do so.

    • GamingChairModel@lemmy.world
      link
      fedilink
      arrow-up
      22
      ·
      1 年前

      AI drives 48% increase in Google emissions

      That’s not even supported by the underlying study.

      Google’s emissions went up 48% between 2019 and 2023, but a lot of things changed in 2020 generally, especially in video chat and cloud collaboration, dramatically expanding demand for data centers for storage and processing. Even without AI, we could have expected data center electricity use to go up dramatically between 2019 and 2023.

    • Galapagon@sh.itjust.works
      link
      fedilink
      arrow-up
      9
      arrow-down
      2
      ·
      1 年前

      I think you should be more concerned about the automatic ai responses on every other search, instead of people having a bit of fun with these

      • pedz@lemmy.ca
        link
        fedilink
        arrow-up
        11
        arrow-down
        1
        ·
        1 年前

        I am. That’s why I switched to DDG and deactivated it.

      • Echo Dot@feddit.uk
        link
        fedilink
        arrow-up
        4
        ·
        1 年前

        This is my problem with it as well. I wish it was a tolerable option that I could click when I wanted an AI summary, which would be basically never.

        At one point I was looking for a pinout diagram for a chip, and the first result I got was the AI summary, I wanted a picture not text, how’s text helpful? All it did is give me a list of the pins, I know what pins it has, but I want to know where they are.

    • Lulzagna@lemmy.world
      link
      fedilink
      arrow-up
      7
      arrow-down
      4
      ·
      1 年前

      This is a dumb misconception. High emissions and energy consumption is when training models, not during prompts

      • vrighter@discuss.tchncs.de
        link
        fedilink
        arrow-up
        7
        arrow-down
        1
        ·
        1 年前

        and models are being trained all the time. It’s the only way to assimilate new data. So your point is moot.

        • mika_mika@lemmy.world
          link
          fedilink
          arrow-up
          1
          ·
          1 年前

          No what he’s saying is the models are being trained whether you mess around with the AI as a user either way.

          It’s like how I didn’t kill the chicken on the store shelves. Myself purchasing it or otherwise doesn’t revive the chicken. The data has/is already being trained.

          • tankfox@midwest.social
            link
            fedilink
            English
            arrow-up
            2
            ·
            1 年前

            That’s a really savvy insight! To expand this analogy further, it’s like your phone or computer gives you a free chicken nugget from a small container attached to the side of the device anytime you search for anything at all. It’s room temperature and often spoiled, it’s your choice whether you eat it or not, but you’re going to get it either way. As such you cannot easily choose to avoid chicken in hopes that that will disincentivize further chicken slaughter.

      • Kilgore Trout@feddit.it
        link
        fedilink
        arrow-up
        3
        arrow-down
        2
        ·
        1 年前

        False. It’s been shown that resolving prompts also drives a major energy consumption, albeit maybe not so higher than regular search queries.

        • Honytawk@lemmy.zip
          link
          fedilink
          English
          arrow-up
          1
          arrow-down
          1
          ·
          1 年前

          A prompt is like 1/1000 of the power used as a microwave for the same amount of time.

          So the difference between a normal query and an AI query is negligible.

  • Wolf@lemmy.today
    link
    fedilink
    arrow-up
    41
    ·
    edit-2
    1 年前

    I can confirm it works for DuckDuckGo as well.

    I love how it even lists ‘sources’ (I checked, that phrase does not appear in either “source”)

    Edit: To make this phenomenon even funnier, I just asked it the exact same question and it gave me a different answer this time. Still “playful” though :)

  • s@piefed.world
    link
    fedilink
    English
    arrow-up
    41
    ·
    1 年前

    One thing that pisses me off is the websites which claim to have the meaning behind a song and it’s just several paragraphs of AI slop that parse the lyrics at a surface level and don’t examine any external context, such as the artist or their life at the time of writing.

    • parody@lemmings.world
      link
      fedilink
      arrow-up
      13
      ·
      1 年前

      Nice we can still sometimes tell when an explanation page is AISEOGARBAGE

      Few years down the road truth-seekin gonna be real tough, I worry: facts and news, images and videos, social connections/commenters… :(

    • Psythik@lemmy.world
      link
      fedilink
      arrow-up
      4
      arrow-down
      2
      ·
      edit-2
      1 年前

      Perplexity does alright too (my preferred LLM because it cites sources, which makes it easier to filter out the slop):

      • huppakee@feddit.nl
        link
        fedilink
        arrow-up
        9
        ·
        1 年前

        It’s funny that these llms have web search capabilities and googles ai assistant doesnt

  • fakeman_pretendname@feddit.uk
    link
    fedilink
    arrow-up
    28
    arrow-down
    1
    ·
    1 年前

    Duckduckgo’s AI:

    “The phrase “a lady butterfly meme a day keeps the doctor away” is a playful twist on the well-known saying “an apple a day keeps the doctor away.” This original phrase suggests that eating an apple daily can contribute to good health and potentially reduce the need for medical attention.”

    “In this variation, the mention of “lady butterfly meme” likely refers to a specific meme or a lighthearted, humorous image that brings joy or laughter. The implication is that engaging with fun or uplifting content, like memes, can have a positive effect on one’s mental health and overall well-being, similar to how healthy eating is associated with physical health.”

    “Overall, it emphasizes the importance of humor and positivity in daily life as a way to maintain good health, particularly mental health.”