• taiyang@lemmy.world
    link
    fedilink
    English
    arrow-up
    10
    ·
    1 day ago

    I’ve grappled with this before, and even before people were using AI, take home exams were always open-book by default because of course people will look things up. The irony is, since AI, I’ve had a lot worse essay submissions-- they simply get shit wrong a lot more relying on AI (especially in my statistics class).

    And my stats class only needs to do basic stuff to show work; I let them use excel/sheets to check work. They can’t even type in “=correl()” properly and would rather take a photo of the data and send it to ChatGPT. AI gives them wildly wrong steps and it’s so obviously wrong…

    • stoly@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      20 hours ago

      Funny: I’m a doctoral student in a statistical program and I have thought, since I started, that you can’t really use AI to cheat at stats. They can’t even get the math right much of the time. I wouldn’t even want to imagine what an essay would look like.

      ChatGPT is good for ensuring that you understand a concept or, perhaps, help setting up a table or a specific formula. Not so good at actually doing STAT homework, though. (Note: I’ve never tried to cheat, but have seen how bad some of the responses are.)

      • taiyang@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        19 hours ago

        I’ve seen it used very occasionally as a study tool, which is generally fine until it gets it wrong. Just like you can use Khan Academy (or I guess, YouTube but there’s again a chance it’s wrong lol). The craziest is when they use an AI delusional statistical theory that doesn’t exist, lol.

        • stoly@lemmy.world
          link
          fedilink
          English
          arrow-up
          4
          ·
          edit-2
          18 hours ago

          So, as you can see, the Reverse Bayesian Transformation has resulted in a normalized distribution. While n = 3, this logistic regression clearly provides a value of 0.75. Therefore, we conclude that there is a statistically significant relationship between cat owners and people who never pay taxes.