• 0_o7@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    36
    arrow-down
    1
    ·
    1 day ago

    I knew this was a clickbait when it’s gizmodo, but oof

    The mods in the Wired story explain how they detect AI content, and unfortunately their methods boil down to “It’s vibes.”

    • Psythik@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      ·
      edit-2
      1 day ago

      Hey man, if the vibes aren’t right, it’s probably AI. Uncanny Valley and shit, bruh. Seems legit.

  • blockheadjt@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    7
    ·
    1 day ago

    If AI is modeled after intellectuals, there will inevitably be a swath of non-intellectuals who conclude the post title… because the idea of intellectuals predating AI is unthinkable to them.

    • ragepaw@lemmy.ca
      link
      fedilink
      English
      arrow-up
      1
      ·
      5 minutes ago

      I used an AI to analyze a piece of writing I did years ago, long before AI was a thing. It determined that there was some huge margin of my work was likely written by AI, and when I asked why, it stated by use of sentence structure, words spelt using British spellings, oxford commas, and emdashes indicated I was AI — which I am not.

  • hedge_lord@lemmy.world
    link
    fedilink
    English
    arrow-up
    11
    arrow-down
    1
    ·
    edit-2
    1 day ago

    What logically follows is that I need to be as weird and eccentric as possible in order to counteract the memetic contagion of a lovecraftian averaging machine. I bet I could make a cult out of this!

    • vithigar@lemmy.ca
      link
      fedilink
      English
      arrow-up
      7
      ·
      1 day ago

      I’m one of those deranged few who actually used em dashes in my normal typing habits. Not super often the way LLMs are prone to, maybe once a month tops. Alt+0151 or Compose, dash, dash, dash.

      Now a find myself reluctant to use what I felt was a useful bit of punctuation out of concern people might think what I’m typing was LLM generated. It sucks.

      • cecilkorik@piefed.ca
        link
        fedilink
        English
        arrow-up
        4
        ·
        edit-2
        1 day ago

        I always just used – instead, because I’m too lazy to remember weird codes and I don’t know what a compose key is, but the intent is the same.

        Edit: oh and I forgot lemmy does weird stuff with em-dash too, what I originally wrote there was hyphen-hyphen

    • pHr34kY@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      ·
      2 days ago

      I always liked the dramatic…

      …pause.

      I’ve never used em-dashes to represent them. Am I doing it wrong?

    • corsicanguppy@lemmy.ca
      link
      fedilink
      English
      arrow-up
      5
      arrow-down
      1
      ·
      2 days ago

      People have been speaking with punctuation since language was invented. If you only now realize that pause is an em-dash, that’s on your schooling. :-p

    • cecilkorik@piefed.ca
      link
      fedilink
      English
      arrow-up
      6
      ·
      1 day ago

      Autistic people were the original LLMs. I don’t care how much data you shove into it, there is still no LLM on Earth that knows more about trains than an autistic guy who knows everything about trains.

      The difference is the autistic person isn’t completely sycophantic and might be completely disgusted that you don’t know the first production date of the C40-8W.

    • anake@piefed.zip
      link
      fedilink
      English
      arrow-up
      69
      arrow-down
      3
      ·
      2 days ago

      That’s a very thoughtful thought! Here’s what I thought you could think about when these thoughtful thoughts come up:

      · Thoughts help thinking brain go poo poo every
      · Sometimes brain say “hmm” but actually “yikes”
      · When idea go bonk against skull wall, that’s innovation
      · Overthinking is just cardio for neurons
      · Deep philosophical questions like “Why am I?” and “How?” share 87% of the same ingredients
      · If you can’t stop thinking, try unplugging brain and plugging it back in (this is also called “a nap”)

    • MDCCCLV@lemmy.ca
      link
      fedilink
      English
      arrow-up
      4
      ·
      2 days ago

      Using a larger vocabulary doesn’t make you sound like an llm, it’s more about the tone.

  • RememberTheApollo_@lemmy.world
    link
    fedilink
    English
    arrow-up
    29
    ·
    2 days ago

    So perversely chatbots are increasing people’s vocabulary because people read so few books anymore and most of their word usage comes from what they read online?

  • Allero@lemmy.today
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    1 day ago

    The worst part - through human communication, texts on the internet, etc. - it seems to propagate towards people who don’t even use AI.

    I myself have strong aversion to it, but found myself using much more bullet points, cliché constructions, and yes - even em-dashes - when I don’t actively pay attention to how I write.

    I’ve always been sensitive to language around me, naturally adapting to the environment around - and now it backfires.

      • oh_@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        ·
        1 day ago

        My huge ADHD sentences love the dash action, but with AI using it now people think I use it to write.

        • SoleInvictus@lemmy.blahaj.zone
          link
          fedilink
          English
          arrow-up
          7
          ·
          edit-2
          1 day ago

          I see the point being made; however, how the fuck else might I jam several clauses into one compound megasentence - overwrought and peppered with purple prose - if not with multiple types of punctuation? Should I summarize, streamline? Clearly not: all thoughts should be expressed in full, replete with all irrelevant details. Perhaps remove unneeded, intra-sentence explanations and asides (like this one, which I’m quite fond of) of details obvious to the reader?! Never!

          No AI could write with matching convolution — especially when, unlike me, it cannot do so on the toilet.

          • Allero@lemmy.today
            link
            fedilink
            English
            arrow-up
            3
            ·
            21 hours ago

            Your eloquence strikes me as unnecessary but beautiful - truly, it is in human nature to overcomplicate, but in such a special way that makes it unique and almost spiritual instead of verbose and obtuse. I compliment your decisiveness and ability - a distinctively human ability - to see through this thin line and turn your toilet habit into something so profound. Yet, I cannot help but note that taking out the phone while unloading your bowel - or bladder, for that matter - is yet another symptom - a very alarming symptom - of the modern age interfering with our ability to distance ourselves from a neverending steam of distractions, one that comes hand in hand with modern AI-powered enshittification. So, if you’d like to truly reconnect with your roots, to find human in yourself - perhaps it might be bright to create an intimate space where the only connection you experience is a connection with yourself.

            Perhaps, I couldn’t express myself as vividly as you do; for that I apologize; yet, I’m not even operating in the domain of my mother tongue, so I hope I can be forgiven. Besides, one other distinctively human thing is to try regadless.

  • Gsus4@mander.xyz
    link
    fedilink
    English
    arrow-up
    6
    ·
    edit-2
    2 days ago

    Hahaha, now even the source of new data is starting to be poisoned by LLMs…good luck trying to outproduce LLM slop to train LLMs…and ending up with goop real fast.