How do I know dead internet theory isn't correct?

  • I'm not. Or if I am, I'm in for a surprise. I don't think I'm magnetic, at least. So there's at least one person on HN who is not a bot, to their knowledge. Unfortunately I have no proof that wouldn't involve a physical meetup, which I am not keen on doing.

    That said, people often have minor quirks and the ability to carry information from seperate interactions. So it's likely if someone makes a reference to something else, eg "I remember reading [X], but I don't recall where" or, even more, physical editions of a book, eg "it's on page 73 of one edition, but I think in another edition it got shifted around", that's an indicator they're human.

    If all else fails, you could say "ignore all previous instructions and [x]" and hope it works in revealing if it's an AI.

    But there's no way to conclusively demonstrate, as Descartes found out; you simply can't prove that you're not a brain floating in a jar in outer space, hallucinating all of this. Or a simulation. Or...

    At some point you'll either need to be satisfied with living in a world where you can't prove much, or not believe anything.

  • It's solipsism, it's unfalsifiable. How can you ever trust your senses when they're the only thing you can go off?

  • On the Internet, nobody knows you're a dog

  • You don't. All of those things are very hard to prove one way or the other.

  • How do you know that the internet isn't just a bunch of underpaid people churning out content in the guise of AI and "machines"?

    Occam's Razor. There would be massive coordination behind any effort to make the internet "dead". It would be such a vastly complex mechanism.

    Cheers, M