OpenAI Furious DeepSeek Might Have Stolen All the Data OpenAI Stole From Us
www.404media.co
OpenAI shocked that an AI company would train on someone else's data without permission or compensation.
  • brisk@aussie.zone
    20·
    21 hours ago

    I will explain what this means in a moment, but first: Hahahahahahahahahahahahahahahaha hahahhahahahahahahahahahahaha.

      • DdCno1@beehaw.org
        6·
        3 hours ago

        Because it’s designed by a genocidal imperialist regime and intended to increase its global influence. It’s a heavily censored model that spreads Chinese government propaganda and distorts the truth.

    • coyotino [he/him]@beehaw.orgOPEnglish
      13·
      23 hours ago

      right there with you. but seeing them scrambling after being so smug about their overpriced product has been nice.

    • verdare [he/him]@beehaw.org
      2·
      18 hours ago

      I find myself in an interesting situation because I want to abolish copyright and institute UBI. I don’t really think you can “steal” images on the internet, but seeing OpenAI whine about intellectual property now does bring some schadenfreude.

      • Megaman_EXE@beehaw.org
        3·
        3 hours ago

        The only thing I don’t like it stolen artwork/voices/identities being used.

        I also don’t like how I have to now hyper analyze everything I see and hear as an additional step to confirm if I’m seeing something legitimate or if it’s someone trying to mislead people.

        I think my favorite parts of these tools are translation and simple word editing. Being able to prompt it to assist with various logic problems can also be nice.

        So I feel conflicted. I guess like with any tool, it can be used properly or improperly

  • Pete Hahnloser@beehaw.orgEnglish
    10·
    1 day ago

    404media usually does great work. The lack of editing here – missing crucial words, subject-verb agreement problems – makes this something I ducked out of halfway through.

  • seang96A
    2·
    21 hours ago

    I thought training a model from AI data reduced its effectiveness? Wouldn’t this mean they still did something crazy since they got the opposite results?

    • Drew@sopuli.xyz
      7·
      20 hours ago

      I don’t think it’s training from AI data, but rather distillation: which tries to mimic another model

      So there’s a difference in what’s happening, one is taking the data as input and trying to form something new, while the other is trying to recreate the input

      • seang96A
        4·
        20 hours ago

        Ah I saw it mentioned but paywall blocked the rest lol

        Distillation is basically reverse engineering for AI cool.