• 0 Posts
  • 89 Comments
Joined 5 months ago
cake
Cake day: July 22nd, 2024

help-circle















  • While this example is somewhat easy to corect for it shows a fundamental problem. LLMs generate output based on the data they trained on and by that regenerate all the biases that are in the data. If we start using LLMs for more and more tasks we are essentially freezing the status quo with all the existing biases making progress even harder.

    It’s not gonna be “but we have always done it like that” anymore it’s going to become “but the AI said this is what we should do”.





  • killingspark@feddit.orgtoMemes@lemmy.mlForest of trees
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    4
    ·
    29 days ago

    I’m saying that modern practitioners of Shinto don’t consider the emperor divine.

    So they were able to continue to live their culture without being individually forced to do anything? Great, then thats not genocide.

    What an interesting perspective. So what you’re saying is, if the Chinese government were to recognize Islam as one of its major, protected religions, but restrict certain radical teachings and versions of it, then it wouldn’t be genocide.

    The better analogy would be to allow the chinese government to force one person to say “I am not divine”. Let’s say they were able to revive the prophet and make him say these words. Yes that would shatter the faith of a lot of people and how they deal with that is pretty much no-ones business. And it wouldn’t be genocide if they all turned away from their faith after that.