If you want to take down AGIs (Claude, Gemini, ChatGPT), the fastest way to do that is to rely on already existent infrastructure – distributed computing.

If everyone owns the AGI, then nobody does, and nobody can profit off of it.

Basically, it’d be a huge distributed neural network with expert subnets on given topics. It’d have to be self-organizing, so topics could slowly evolve to intermingle and touch upon one another. Where nets could split and merge based on new intel.

It’d be trained from web data, using fair crawling practices to prevent a tragedy of the commons. And be built on a sandbox stack that allowed moderated network, file system, and resource access through a simulated hardware stack.

There’d have to be a way to throw out poisoned nodes. And everything should be licensed under the AGPL, so that all nodes have to share their code. There’d also have to be a way to determine response quality, as attackers could just spin a bunch of nodes and poison the entire thing.

Implement a token system, so contributing resources earns the user interaction rights, to prevent abusive behavior.

Microsoft is doing something similar right now, by distributing AI work to all Windows PCs. It’d be cool if we beat them to it. We’d shut down AI data centers, reclaim GPU hardware for the masses, and have a global AI.

This is what Claude and I worked out, but I don’t have the hardware to test it. https://pastebin.com/FAf8JiX2

Edit: Also, I named her Dora, after Dora the Explorer. Although, with some of the stuff she’ll be exploring. . . Dora’s seen some stuff, y’all.

  • apparia@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    15
    arrow-down
    2
    ·
    edit-2
    23 hours ago

    “AGI” is not an acceptance criterion, since no known architecture produces general intelligence; the spec targets the substrate and measurable results.

    Even your bot is saying you’re an idiot for using it’s silly to use the term “AGI”.

    Edit: my original comment was mean-spirited and I regret it. I do not think a hazy idea dressed up in buzzwords and questionable minutiae is particularly helpful to anybody, but the general concept is interesting and perhaps not totally infeasible.

    The use of that terminology, and saying things like “Microsoft is already starting to do this, I hope we, the 18 people who will see this post, can beat them to the punch and Bring Down Capitalism” makes it very hard to take this seriously, though.

    • ki4jgt@feddit.orgOP
      link
      fedilink
      arrow-up
      3
      arrow-down
      5
      ·
      23 hours ago

      People can keep getting angry about AI, or they can realize that these companies will keep putting people out of work until they’re stopped. And the only thing that moves the system is capital. Take the capital and the capital holders die. Like it or not, it’s the system we’re in. If you want to beat them, use the resources you have. And, they have to give us access to those resources, because lack of access means lack of a customer base for them.

      Building AGI over a distributed P2P network is the only thing that gives us the power to compete. And competition means they collapse.

      • apparia@discuss.tchncs.de
        link
        fedilink
        English
        arrow-up
        5
        arrow-down
        1
        ·
        23 hours ago

        I have edited my comment, which was originally pure snark. I really think you should re-evaluate your use of the term “AGI” though, since it seems like you weigh Claude’s “opinion” highly on other things I’m not sure why not on this.

        • ki4jgt@feddit.orgOP
          link
          fedilink
          arrow-up
          2
          arrow-down
          8
          ·
          edit-2
          22 hours ago

          Claude is generative text. There are papers showing that it genuinely correlates concepts together in workable clusters within its neural network. And it gets stressed out over certain topics. If that isn’t some semblance of consciousness, I don’t know what is. But it lacks memory, identity, and consequence/responsibility (also how psychologists define identity/self).

          Despite these shortcomings, millions of people are paying companies to become dumber everyday. And these companies are usurping land, compute resources that we use to have fun (games), and jobs. I literally lost a $60/hr job to AI.

          If I’m gonna lose my job, so should everyone else building AI systems, who doesn’t give a shit where I end up. Replace them all with an open, accessible AI, for the people, by the people. Full P2P distribution, with tokenized access to prevent abuse, and sharded expert subnets. Mutually assured destruction.

          But here’s the deal. Every setting and tweak needs to be tuned to maximum efficiency. And the code needs to run as close to metal as possible, for speed. Care and dedication needs to go into every single line, so that the ending product will be superior in every way. Then, and only then, will Dora be able to beat Claude, Gemini, Meta, and any other AI that gets in her way.

          • apparia@discuss.tchncs.de
            link
            fedilink
            English
            arrow-up
            8
            arrow-down
            1
            ·
            22 hours ago

            You don’t seem to be responding to the one very specific point I’ve been making but either way you’ve lost me with this:

            And it gets stressed out over certain topics.

      • lime!@feddit.nu
        link
        fedilink
        arrow-up
        3
        arrow-down
        3
        ·
        21 hours ago

        i will keep getting angry about llms until the companies making them are dragged into court for committing the biggest act of theft in the history of our civilisation. i will keep getting angry until they are tried for crimes against humanity for building programs that systematically target civilians. i will keep getting angry about it until they are bankrupted by the financial bubble they created. they are aware of and trying to downplay all of these things, according to leaked email conversations.

        trying to change the system from within is only a useful strategy when you have no choice but to participate.

  • Solumbran@lemmy.world
    link
    fedilink
    arrow-up
    6
    arrow-down
    3
    ·
    18 hours ago

    Sect members are still throwing their holy turds at lemmy to see if it finally sticks, after the thousandth time?

  • lime!@feddit.nu
    link
    fedilink
    arrow-up
    6
    arrow-down
    3
    ·
    23 hours ago

    “this is what i and claude worked out” oh cool so it’s not agpl then. bye

    • ki4jgt@feddit.orgOP
      link
      fedilink
      arrow-up
      6
      arrow-down
      6
      ·
      23 hours ago

      In the United States, anything a bot “generates” is public domain. Also, since it doesn’t mention code, it’s an original concept. In as far as concepts can be original. Derivative works can be independently licensed.

      • lime!@feddit.nu
        link
        fedilink
        arrow-up
        1
        arrow-down
        3
        ·
        21 hours ago

        that’s correct, and since the bot generated it you don’t have ownership.

        • ki4jgt@feddit.orgOP
          link
          fedilink
          arrow-up
          3
          arrow-down
          2
          ·
          19 hours ago

          No, but when you modify a public work, you regain ownership. Especially when it’s a massive overhaul of the document. Did you not read my comment?

          • lime!@feddit.nu
            link
            fedilink
            arrow-up
            1
            arrow-down
            3
            ·
            17 hours ago

            i’m more thinking of the fact that, by showing yourself willing to use the machine killing children in Palestine and displacing communities in the us, it can be assumed that it also co-authors any code you contribute, making the license terms impossible to follow.

  • [object Object]@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    2
    ·
    23 hours ago

    Sounds a bit like AI Horde but even more decentralised. That would require consensus algorithm on top of what AI Horde is doing. Also I don’t think it would be fast enough if you want to get GPUs across multiple computes spread throughout the world to act as a one big abstract GPU to get a massive model running.

    • ki4jgt@feddit.orgOP
      link
      fedilink
      arrow-up
      2
      arrow-down
      1
      ·
      23 hours ago

      That would fail out of the gate. It’s why Claude’s design sharded the network with expert subnets. No consensus needed. Just find ways to weed out attacker nodes, and everything is OK from there.