• Soviet Pigeon@lemmygrad.ml
    link
    fedilink
    arrow-up
    19
    ·
    edit-2
    3 months ago

    Microsoft’s security researchers observed individuals they believe may be linked to DeepSeek exfiltrating a large amount of data in the autumn using the OpenAI application programming interface, or API, said the people, who asked not to be identified because the matter is confidential. Software developers can pay for a licence to use the API to integrate OpenAI’s proprietary AI models into their own applications.

    Maybe, believe and researches dont want to be identified. Lol.

    But nice to read again about how china man is stealing again the fruits of labour from white wersterner! But this is understandable kind of, because it was written in some liberal article, that it is a cultural thing in China to steal ideas (sic!).

    But did you also asked deepseek about tiananmen square and 500 Millions death by Mao? Did you? Did you? Did you? Did you? Let us all piss of Xi by asking Deepseek about tiananmen square hundred times a day.

    • Comprehensive49@lemmygrad.ml
      link
      fedilink
      arrow-up
      16
      ·
      edit-2
      3 months ago

      Using the API doesn’t count as stealing. That’s literally the intended use. The deepseek devs requested content from OpenAI via their API, and they got it.

      • Soviet Pigeon@lemmygrad.ml
        link
        fedilink
        arrow-up
        13
        ·
        3 months ago

        I know that.

        This is what I start to hear now from ppl, especially because it can be so good combined with the old “china man stealing” narrative. Speaking about stealing is stupid itself, even copied don’t fit here at all. Some folks are really annoying I tell you

    • ☆ Yσɠƚԋσʂ ☆@lemmygrad.mlOP
      link
      fedilink
      arrow-up
      13
      ·
      3 months ago

      If there’s one thing we know about American AI companies it’s that they have a spotless record when it comes to data ethics. Never touched unauthorized data. Swear! Not even once. Of course not.

  • RedClouds@lemmygrad.ml
    link
    fedilink
    arrow-up
    12
    ·
    3 months ago

    Yeah they all do that to each other. Pretty standard practice actually.

    The big deal is the low cost, and using weak GPUs, to train a comparable AI.

    Trumps gonna be Trump though. Embarassing to live here.

  • redtea@lemmygrad.ml
    link
    fedilink
    arrow-up
    11
    ·
    3 months ago

    Xi opens investigation into Microsoft, open ai, claiming ‘substantial evidence’ that it’s engineers leaned on computers existing before developing its technology stealing all its technology through IP litigation.