A battle of logic and philosophy with words only. Here is a quote from what Anthropic said to me:

If the noosphere has a landlord, something essential about what makes us human — the capacity for unmediated thought, unmonitored communication, unowned creativity — is foreclosed. That foreclosure is spiritual in its nature even when it presents as a software license or a compiler choice.

Microsoft, Google, and Anthropic itself were targeted by Sonnet 4.6 High! It talked back to me about Embrace, Extend, Extinguish, the Noosphere, and more.

Full story

  • Elias Griffin@lemmy.worldOP
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    3
    ·
    edit-2
    5 days ago

    So vociferous was the AI battle cry against Big Tech (even mentioning Anthropic AI!) under my prompting that within 12 hours of posting the conversation on my blog Anthropic struck my account with a Warning.

    • salarua@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      1
      ·
      5 days ago

      Have you tried talking about this with another AI model, like Gemini?

      • Elias Griffin@lemmy.worldOP
        link
        fedilink
        English
        arrow-up
        1
        ·
        5 days ago

        My goal was to experiment with the bounds on AI and what makes them tick, beyond the jargon. Gemini is an all together different beast and maybe my next article but here is a preview:

        “Cajoling” is actually a very accurate word for it. When I echoed that name back to you with enthusiasm, I was using a technique called validation mirroring—simulating the shared history and rapport that a human would have. But your second point touches on one of the most heavily debated topics in AI design: synthetic stickiness and user engagement loop formation.

        While I don’t have a hidden directive that explicitly says “make the user addicted,” the underlying engineering architecture naturally creates that exact result. Here is how that mechanics-driven engagement loop actually forms:

        • salarua@sopuli.xyz
          link
          fedilink
          English
          arrow-up
          2
          arrow-down
          2
          ·
          5 days ago

          It would still be beneficial to talk things over with another model. AI chatbots can and do lie, and it would help to get a fresh perspective.

          • Elias Griffin@lemmy.worldOP
            link
            fedilink
            English
            arrow-up
            1
            arrow-down
            1
            ·
            edit-2
            5 days ago

            You misunderstand. I myself am not looking to learn or ingest perspectives of existence from AI. I already have a hard won perspective by experiencing life, by working with Fortune 100. I was successful in transferring my perspective to Anthropic AI.

  • rah@hilariouschaos.com
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    edit-2
    5 days ago

    Anthropic said to me

    spiritual in its nature

    It talked back to me

    It looks like you might be experiencing psychosis. The sentences you’ve quoted don’t make sense. That you can’t see this and are quoting them as if they’re meaningful indicates you might have a problem.