• merc@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    102
    arrow-down
    5
    ·
    edit-2
    6 months ago

    PhD level intelligence? Sounds about right.

    Extremely narrow field of expertise ✔️
    Misplaced confidence in its abilities outside its area of expertise ✔️
    A mind filled with millions of things that have been read, and near zero from interactions with real people✔️
    An obsession over how many words can get published over the quality and correctness of those words ✔️
    A lack of social skills ✔️
    A complete lack of familiarity of how things work in the real world ✔️

  • kn0wmad1c@programming.dev
    link
    fedilink
    English
    arrow-up
    85
    arrow-down
    2
    ·
    6 months ago

    Translation: GPT-5 will (most likely illegally) be fed academic papers that are currently behind a paywall

    • twice_twotimes@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      10
      ·
      6 months ago

      I mean, GPT 3.5 consistently quotes my dissertation and conference papers back to me when I ask it anything related to my (extremely niche, but still) research interests. It’s definitely had access to plenty of publications for a while without managing to make any sense of them.

      Alternatively, and probably more likely, my papers are incoherent and it’s not GPT’s fault. If 8.0 gets tenure track maybe it will learn to ignore desperate ramblings of PhD students. Once 9.0 gets tenured though I assume it will only reference itself.

  • HexesofVexes@lemmy.world
    link
    fedilink
    English
    arrow-up
    54
    ·
    6 months ago

    Wow… They want to give AI even more mental illness and crippling imposter syndrome to make it an expert in one niche field?

    Sounds like primary school drop-out level thinking to me.

    • Min7_f43sh_j5@lemmy.dbzer0.com
      link
      fedilink
      English
      arrow-up
      17
      ·
      6 months ago

      I’m planning to defend in October and I can say that getting a Ph.D. is potentially the least intelligent thing I’ve ever done.

      • cabron_offsets@lemmy.world
        link
        fedilink
        English
        arrow-up
        4
        arrow-down
        1
        ·
        6 months ago

        Same, bruh. But I transitioned from biophysics to money & people management and shit’s pretty okay.

    • SturgiesYrFase@lemmy.ml
      link
      fedilink
      English
      arrow-up
      5
      ·
      6 months ago

      I know three people that dropped out of primary and did quite well. They all ended up taking remedial studies later in life. Two were in trades and the other was a postie. All three were smart as fuck. Just because life gets in the way of going to school doesn’t mean a person is dumb, just uneducated.

    • lad@programming.dev
      link
      fedilink
      English
      arrow-up
      11
      arrow-down
      1
      ·
      6 months ago

      You are correct, but there’s a larger problem with intelligence, we don’t have a practical definition, and we keep shifting the goalpost. Then there’s always a question of a philosophical zombie, if someone acts as a human and has a human body you won’t be able to tell apart if they don’t really have intelligence, so we only need to put LLM into humanlike body (it’s not so, but you get the point)

      reminds me of this, although the comic is on a different matter

      https://www.smbc-comics.com/comic/2014-11-25

    • Spacehooks@reddthat.com
      link
      fedilink
      English
      arrow-up
      3
      ·
      6 months ago

      I believe guardian of galaxy 3 did a take of that. Rocket was special because he innovated while the others just mimiced.

  • jas0n@lemmy.world
    link
    fedilink
    English
    arrow-up
    43
    arrow-down
    4
    ·
    6 months ago

    All aboard the hype train! We need to stop using the term “AI” for advanced auto complete. There is not even a shred of intelligence in this. I know many of the people here already know this, but how do we get this message to journalists?! The amount of hype being repeated by respectable journalists is sickening.

    • Ashen44@lemmy.ca
      link
      fedilink
      English
      arrow-up
      20
      ·
      6 months ago

      people have been calling literal pathfinding algorithms in video games AI for decades. This is what AI is now and I think it’s going to be significantly easier to just accept this and clarify when talking about actual intelligence than trying to fight the already established language.

      • jas0n@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        ·
        6 months ago

        While you’re not wrong, I don’t ever recall people en masse believing a game AI was truly intelligent. Everyone was always aware of the truth. There just isn’t a great name for the computer players. I think it’s an important distinction here because people do believe ChatGPT is intelligent.

    • The Bard in Green@lemmy.starlightkel.xyz
      link
      fedilink
      English
      arrow-up
      2
      ·
      6 months ago

      I know many of the people here already know this, but how do we get this message to journalists?!

      Journalists have this information, but articles about it probably generate 10% of the clicks, shares and eyeballs->ad revenue that either the hype or the scaremongering does.

  • Phoenixz@lemmy.ca
    link
    fedilink
    English
    arrow-up
    37
    arrow-down
    1
    ·
    6 months ago

    PhD level of intelligence

    No it won’t. At some point, some AI will, but that point is still far away.

    I’m sure it’ll know how to string words and sentences together real nice, even to the point where it makes sense. It will still not have a clue what it’s talking about, it’ll still not understand basic concepts as “understanding” requires a whole lot more than just an advanced ability of pushing words together.

    • LeFantome@programming.dev
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      2
      ·
      6 months ago

      If she wants it to. At some point, all the chat bots are going to be given bodies. We all know it.

    • Tryptaminev@lemm.ee
      link
      fedilink
      English
      arrow-up
      13
      ·
      6 months ago

      You know the Chineese? They talk about this ChatPT 7. But we Americans. My uncle, very smart man. Smartest in every room except on Thanks Giving. I always had Thanks Giving and my Turkey, everyone loved my Turkey. He said we will soon have Chat 8 and the Chineese they know nothing like it.

  • Matriks404@lemmy.world
    link
    fedilink
    English
    arrow-up
    30
    arrow-down
    2
    ·
    6 months ago

    What a bunch of bullshit. I’ve asked ChatGPT recently to do a morphological analysis of some Native American language’s very simple sentences, and it gave absolute nonsense as an answer.

    And let’s be clear: It was an elementary linguistics task. Something that I did learn to do on my own by just doing a free course online.

    • jubilationtcornpone@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      4
      ·
      6 months ago

      Yesterday, I asked it to help me create a DAX measure for an Excel pivot table. The answers it gave were completely wrong. Each time, I would tell it the error that Excel was displaying and it would respond with “Sorry about that. You can’t use that function there for [x] reasons.”

      So it knows the reason why a combination of DAX functions won’t work but recommends them anyways. That’s real fucking useful.

  • edric@lemm.ee
    link
    fedilink
    English
    arrow-up
    28
    arrow-down
    1
    ·
    edit-2
    6 months ago

    Now it can not only tell you to eat rocks, but also what type of rock would be best for your digestion.

  • intensely_human@lemm.ee
    link
    fedilink
    English
    arrow-up
    25
    arrow-down
    1
    ·
    6 months ago

    A PhD makes a person knowledgeable, not intelligent. And GPT-4 was already extremely knowledgeable.