2 pointsby SuboptimalEng2 hours ago4 comments
  • serf3 minutes ago
    "dislike ai? prove so by using an alternative ai."

    I don't think this train is going to stop, but to pretend that the anti-AI proponents are suggesting moving away from frontier labs and into open-source is short-selling their desires; these people want literal bans, arms exclusions, and penalties for use.

  • newsomix9xl2 hours ago
    AT&T is already locking in open source to save money.

    Not sure how using open source AI means "hating AI and proving it".

    Are you implying open source is like an economic boycott?

    • SuboptimalEng2 hours ago
      Exactly. Be prepared to switch companies/models at moments notice.

      I mean, if someone truly disliked AI, the only natural course would be to ensure that they can switch to open source and not willingly give money + data to the frontier labs.

      • newsomix9xl2 hours ago
        Well... I'm not sure that's how boycotts work.

        What if open source is like radio or YouTube acting as a free ad for a product that still has commercial value in other ways?

        And if the OSS product doesn't "kill off" its competition it may simply appeal to a lower cost or price sensitive segment.

        Once students learn the AI on free models they graduate to paid versions.

        A boycott, "for the AI haters" would need to be against all AI, free or not, maybe?

        • bigyabaian hour ago
          Boycotts can absolutely mean looking for an equivalent alternative somewhere else. If you think the mainstream is unsustainable, there's nothing wrong with supporting an alternative business model.

          OpenAI and Anthropic have billions of dollars in outstanding debt, and neither Fable nor Sol is pulling them out of the red. As the gap between Chinese and American benchmarks closes, it seems like the safest route is to let America's frontier labs default on their debt. The absolute worst-case scenario would be nationalizing America's labs and socializing the losses from their failure to compete.

  • toomuchtodo2 hours ago
    “If you don’t believe in God, prove he doesn’t exist.”

    The burden of proof lies with the proponents. AI is useful, it’s just not going to materially change the world outside of a few domains it is very well suited for or be worth what folks are hoping it is. On prem self hosted solutions are certainly a great way to protect the value the capability can provide for consumers of it, versus having to live through constant churn in the marketplace for companies chasing quick economic returns before the capital bubble pops.

    • SuboptimalEng2 hours ago
      I see you didn't read the description under the title :)
      • toomuchtodo2 hours ago
        I think you’re asking the wrong questions, which you’ve somewhat tried to merge together, and so I answered the more optimal question.

        Does AI have value? Yes. Is it smart to maintain optionality for using it? Yes. Is it rational for folks to hate the AI vendors grifting the capital market frenzy and look forward to them potentially failing? Yes. Is it worth trillions of dollars? No.

  • khangtong9882 hours ago
    [flagged]