OpenAI transcribed over a million hours of YouTube videos to train GPT-4

Comments

• lvxferre@mander.xyz ⁨7⁩ ⁨months⁩ ago

  The “fun” part is that they’ll get away with it. Google likely did the same with other platforms; so if it sues OpenAI, it’s creating a big precedent against itself.

  source
• Defaced@lemmy.world ⁨7⁩ ⁨months⁩ ago

  I fucking hate how this company is just taking data and metrics without any permissions and repercussions. OpenAI and Sam Altman can fuck right off. Same with Microsoft and copilot and every other company rushing for the AI/ML arms race, its disgusting and irresponsible.

  We joke about skynet and terminators and whatnot, but the reality is OpenAI is legitimately moving towards that end with no safety precautions, no thought put into the economic and humanitarian impacts they’re going to cause. Capitalism in general (and yes I’m going to be that guy and say it) simply cannot survive the AI/ML age of humanity without evolving.

  source

  • Dkarma@lemmy.world ⁨7⁩ ⁨months⁩ ago

    How clueless are you. Everything “taken” was available for free. Provided for free for any web crawler to consume and now you’re acting like consuming it is a crime?

    I get that you’re really jealous because you didn’t think of LLMs but you don’t get to claim something is a crime in one specific instance just because you don’t like what they’re doing after their program consumes content.

    Google has done the same thing for years and no one said a peep. What does everyone think search results even are???

    source

    • Defaced@lemmy.world ⁨7⁩ ⁨months⁩ ago

      You completely miss my point, are you saying data such as copyrighted published works and medical records are free? Because I did not in any way consent to sharing medical records to OpenAI businessinsider.com/openai-chatgpt-generative-ai-…

      Now I realize this is an alleged offense, but it’s still fucked up.

      source
    • EurekaStockade@lemmy.world ⁨7⁩ ⁨months⁩ ago

      Fuck Google too

      source
  • afraid_of_zombies@lemmy.world ⁨7⁩ ⁨months⁩ ago

    Going to start keeping score. Mark you down in the AI is going to be amazingly powerful camp.

    source
• csm10495@sh.itjust.works ⁨7⁩ ⁨months⁩ ago

  I’ll be that guy here and start a small fire.

  If it’s not OK for OpenAI to violate YT’s terms of service, then it’s not OK to use an alternative frontend or ad blocker that also violates ToS.

  You should be for enforcing the ToS or against it. Allowing one version of breaking it without the other is fickle.

  Some people think OpenAI should stop, but they should still have their ReVanced, which boggles my mind a bit.

  source

  • patatahooligan@lemmy.world ⁨7⁩ ⁨months⁩ ago

    No, the intent and the consequences of an action are generally taken into consideration in discussions of ethins and in legislation. Additionally, this is not just a matter of ToS. What OpenAI does is create and distribute illegitimate derivative works. They are relying on the argument that what they do is transformative use, which is not really congruent with what “transformative use” has meant historically. We will see in time what the courts have to say about this. But in any case, it will not be judged the same way as a person using a tool just to skip ads. And Revanced is different to both the above because it is a non-commercial service.

    source
  • sugar_in_your_tea@sh.itjust.works ⁨7⁩ ⁨months⁩ ago

    ad blocker

    I can see an argument for an alternative FE, but I’ll never agree that an ad blocker is “wrong.” It’s my choice what to install on my machine, and it’s YouTube’s choice how they choose to deliver their content.

    YouTube controls their API (as in, what a third party FE uses), I control what I do with what I receive from the API. They’re different things.

    I do agree that ReVanced doesn’t have any “rights” here, but it’s on YouTube to block their requests, it shouldn’t be something they sue over.

    All of that being said, I’m actively looking to eliminate YouTube from my life. Nebula, Odysee, and others have enough content that I think I can replace most of my usage of YT. The main thing left is music, but I don’t listen to music all that often anyway.

    source

    • SuperSynthia@lemmy.world ⁨7⁩ ⁨months⁩ ago

      I actually use funkwhale now for music. It’s really decent so far. A little buggy though, but it should iron out

      source
• gunpachi@lemmings.world ⁨7⁩ ⁨months⁩ ago

  Doesn’t that violate Youtube’s ToS ?

  source

  • realharo@lemm.ee ⁨7⁩ ⁨months⁩ ago

    It does - it’s also mentioned in the article.

    source

    • dotMonkey@lemmy.world ⁨7⁩ ⁨months⁩ ago

      We don’t do that around here

      source
• afraid_of_zombies@lemmy.world ⁨7⁩ ⁨months⁩ ago

  Oh no. Big corporation didn’t follow the ToS of another big corporation and “stole” content that the first didn’t create and paid pennies for. Clearly the greatest injustice of all time.

  source