72
Why your local LLM feels dumber than it is (forum.level1techs.com)
submitted 2 months ago by to c/technology@lemmy.world
103 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] -1 points 2 months ago (12 children)

Sounds like you haven't tried modern AI yet. I could train an AI on your comment history and copy you, and nobody would be able to tell the difference

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 12 child comments
    •
  • [–] 3 points 2 months ago* (7 children)

    I've tried. They're adamant against AI here. It's not even worth the discussion. There's an army of belligerent people here who have made up their minds that LLM are "stochastic parrots" and they're convinced they understand exactly how they work. No one can tell them otherwise. Doesn't even matter your credentials. They're the experts here. If you argue, they'll only belittle you.

    Save yourself the frustration and heartache; just silently pity them.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 7 child comments
    •
  • [–] -1 points 2 months ago (6 children)

    I work in tech

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 1 point 2 months ago* (5 children)

    Uh huh... That makes you specifically qualified on the topic? Far more than myself, certainly, with my MSc in Machine Learning.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] -1 points 2 months ago (4 children)

    As far as understanding the technology goes? It makes me more qualified than 99% of the AI glazers I encounter, but you may be an exception.

    To say that an LLM is “less dumb” is a misnomer tho, because it implies that the LLM is “thinking” rather than “calculating”.

    My issue here is the anthropomorphic language being used to intentionally mislead people who don’t have even a tertiary understanding of the underlying mechanics.

    Surely you know that under the hood it’s all just algorithms. 1s and 0s run through a fluctuating probability equation.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 1 point 2 months ago (3 children)

    Surely you know that under the hood we're all just algorithms? Electrical impulses determining what we're going to do before our own consciousness are aware of the impending decision.

    You make the mistake of ascribing more importance to the "1s and 0s" of neuron activity we ourselves encompass, and imagining that a different arrangement of that exact same activity could not possibly yield a similar outcome. It's like saying plants can't be alive because they don't have brains.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 0 points 2 months ago (2 children)

    It’s nothing like saying “plants can’t be alive”, that’s preposterous.

    You make the mistake of thinking that brains and computers are exactly the same when we only really understand the workings of one of those things (computers).

    All of the world’s top neuroscientists would tell you that we don’t really actually know how the human brain works.

    People who are far more qualified than you to make such a comparison would not do so.

    Your argument is built upon a logical fallacy known as “false equivalence”. If you’re not familiar, it means that you are presenting two things as equivalent based on flawed logic.

    You’ve probably heard someone use an idiom about comparing apples and oranges right? Doing so would be a false equivalence, that’s where the idiom comes from.

    Another example of false equivalence would be your comparison of my logically sound argument to a completely unrelated and nonsensical assertion about plants.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 2 months ago (1 child)

    Sorry, 'h0tbeef', I forgot that only you understand consciousness and only you know how it is defined. So only you know what qualifies, because of your superior understanding that no one else can possibly hope to have.

    Of course, only humans are capable of being conscious, you're absolutely right as always. You're so clever, you should go reward yourself with a cookie.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 0 points 2 months ago

    Oh, have we reached the stage of the debate where you project your insecurities onto me in a desperate bid to avoid conceding your palpably untenable position? How dignified.

    If you wish to gain more knowledge and understanding, I would suggest you stop outsourcing your cognition to a chatbot.

  • permalink
  • fedilink
  • source
  • parent
  • [–] -3 points 2 months ago (3 children)

    You’re underestimating most people

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 1 point 2 months ago* (2 children)

    You overestimate most people. Here's a study from 2025

    https://arxiv.org/html/2211.13087

    Among AI models, Blenderbot stood out: in AI-AI conversations, it was judged human 67% of the times - more often than actual human-human conversations. In human-AI conversations, human participants were labeled as humans 68% of the time, and AIs were classified as AI 56% of the time.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 2 months ago* (1 child)

    This is an interesting study. I wonder why there’s such a disparity between the number of AI ‘participants’ (10) and human participants (1,916). I wonder how adding additional AI models would impact the data.

    Anyway, this is just benchmarking the AI on 6 language specific tasks. They were programmed and set up specifically to perform whatever specific one task they had to do at the time, that is what computers do. The fact that computers are 1% better at recognizing other computers than humans are is irrelevant.

    AI cannot pass the Turing Test, this article is about “6 Turing like tasks” and it can’t even pass those yet, lmao. You know they’re already seeing diminishing returns in further training right?

    Even in this article it says that even if AI could pass a real Turing Test it wouldn’t be a sign of intelligence or cognition.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 1 point 2 months ago

    Something I should point out just in case is that 50% success rate is the same as random chance. So when the study says:

    AIs were classified as AI 56% of the time.

    That really means that humans could not tell the difference between AI and human, since their success rate was almost the same as random chance.

    6 language specific tasks. They were programmed and set up specifically to perform whatever specific one task they had to do at the time, that is what computers do.

    Sure, and that task was "imitate humans", and these computers seem remarkably good at it. That's what I'm saying. I don't care that these AI work differently than humans. They can imitate humans so well that humans can't tell the difference.

  • permalink
  • fedilink
  • source
  • parent