Streaming now: https://linktr.ee/jorvex609

A video essay examining the AI alignment debate through a structural lens. Combining Max Tegmark’s 12 possible AI futures with Yudkowsky and Soares’ alignment arguments, this video argues the real alignment problem isn’t technical — it’s about who controls the most consequential technology in human history.

Who is building superhuman AI? A handful of unelected executives in Northern California. For whom? Shareholders, not humanity. Under what pressures? A competitive race where the most reckless actor sets the pace for everyone.

Drawing on historical parallels from leaded gasoline to asbestos, the video argues the same pattern repeats: a small group captures benefits while everyone pays costs. The solution isn’t more alignment research — it’s democratic governance of AI development.

  • Dort_Owl [they/them, any]@hexbear.net
    link
    fedilink
    English
    arrow-up
    5
    ·
    5 months ago

    Look, I’m not smart at all, but I’m pretty sure we’re seeing the beginning of this bubble bursting. No one is making any money on this stuff and the costs keep on going up.

    Copilot has recently started charging people for previously free use, from what I understand (again, my understanding LLMs and how their sold is limited)

    But yeah that signals to me that they’re starting to get desperate to make money. Except no one wants to pay for this stuff.