• 0 Posts
  • 15 Comments
Joined 13 days ago
cake
Cake day: August 8th, 2026

help-circle


  • antianarchist@sopuli.xyztoTechnology@lemmy.worldDon't paste the AI.
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    8
    ·
    17 hours ago

    This is such a bizarre take on AI.

    If I ask you a question, I want a useful answer. I don’t care whether you got there by remembering it, Googling it, reading documentation, asking a colleague, or asking an LLM.

    Obviously, blindly pasting 2,000 words of irrelevant AI slop is annoying. But that’s a problem with bad answers, not with AI.

    The particularly stupid part is suggesting that you should take a perfectly good AI-generated answer and rewrite it in your own words. What exactly has been achieved then? The information is identical, except somebody wasted five minutes manually paraphrasing it so the recipient can feel sufficiently exposed to human keystrokes.

    Use AI. Give it the relevant context. Verify important claims. Remove irrelevant garbage. Send the useful answer.

    Judge communication by its accuracy, relevance and usefulness—not by how inefficiently somebody produced it.

    /s










    • Only the key-holder can check. Your teacher, editor, or favourite “AI detector” website cannot run this test; a genuine check needs the provider’s secret key, or a checking service the provider runs. Google runs an early-access detector portal for SynthID; Anthropic says detection tooling is forthcoming.

    I am not so sure about that. The amounts of words is finite and with enough text, you will see that certain words are used more often, especially in certain combinations. I believe people will brute force this and then create a way to destroy the watermark again.




  • Technically interesting, at the same time a bit far fetched with calculations that do not make much sense. For example they say that the resolution I use is rare and that the Do Not Tracl feature is off, makes it even rarer.

    However, this is just a regular iphone with regular Safari. Based on those two factors, you cannot determine a user, as they are too broad, even in combination with others.

    So yeah, fun numbers, fun to see how much information the browser exposes to the server, just to increase comfort.