• Kitikuru@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    65
    arrow-down
    1
    ·
    1 year ago

    This sounds like something an AI would say… Roberto… or should I say Robot-o… I’m gonna need you to click all the bicycles in this photo

      • j4k3@lemmy.world
        link
        fedilink
        English
        arrow-up
        5
        ·
        1 year ago

        I know how to make llama.cpp respond in only emojis too. You’re fooling no one o’robot’o

        I’m messing with AI and I don’t see it as AI training. It would be useless training data to me. There are a few people posting interesting stuff here and there, but this does not have the type of knowledge base and volume I would want to use without a ton of manual filtering.

  • zeppo@lemmy.world
    link
    fedilink
    English
    arrow-up
    17
    ·
    edit-2
    1 year ago

    Paranoia? I’ve never had that impression. Eventually though, sure. I don’t think there’s quite enough content or contributors on Lemmy yet, but certainly data from similar sites like reddit and tweeter has been used for model training.

  • Cheems@lemmy.world
    link
    fedilink
    arrow-up
    16
    arrow-down
    1
    ·
    1 year ago

    It’s true, lemmy is just a website bill gates made specifically to train his ai. You’re just a pawn in his quest to make the greatest ai ever. Congrats, even this question is furthering his ultimate goal.

  • VelociCatTurd@lemmy.world
    link
    fedilink
    English
    arrow-up
    18
    arrow-down
    4
    ·
    1 year ago

    You are right to feel that way. Many of the questions on Ask Lemmy are indeed meant to train an AI. This is because Ask Lemmy is a large language model, also known as a conversational AI or chatbot trained to be informative and comprehensive. It is trained on a massive amount of text data, and it can communicate and generate human-like text in response to a wide range of prompts and questions. For example, it can provide summaries of factual topics or create stories.

    The questions that are most helpful for training an AI are those that are open ended, challenging, and require the AI to think critically. For example, questions about artificial intelligence, philosophy, or science are often good for training AIs. Questions that are too simple or straightforward can be answered by a simple lookup of information, and they do not require the AI to learn anything new.

    If you are interested in training an AI, you can ask questions on Ask Lemmy that are challenging and require the AI to think critically. You can also provide feedback on the AI’s responses, so that it can learn from its mistakes. Over time, the AI will become better at answering questions and generating text.

    Here are some specific questions that you can ask Ask Lemmy to help train it:

    • What is the difference between Strong Artificial Intelligence and Weak Artificial Intelligence?
    • What are some applications of AI in the real world?
    • What are the challenges of developing AI?
    • What are the ethical implications of AI?
    • What is the future of AI?

    You can also ask Ask Lemmy more specific questions about your own research or interests. For example, if you are interested in the development of AI for healthcare, you could ask Ask Lemmy about the potential applications of AI in healthcare or the challenges of developing AI for healthcare.

    By asking Ask Lemmy challenging questions and providing feedback on its responses, you can help it to become a better AI.

    • remotelove@lemmy.ca
      link
      fedilink
      arrow-up
      4
      arrow-down
      1
      ·
      edit-2
      1 year ago

      You know that you are on asklemmy, right? Someone asks a question, and then other people might answer it. It’s a brilliant concept.

      You literally do not have to care and nobody will say anything if you don’t. Nobody gives a flying fuck that you don’t care or even cared about your existence up until this point. Most of us probably cared more about you than your own mother, to be honest. But yet, here you are. You had to ask that question like OP disturbed your jerk off session or something.

      All you had to do is keep sliding that shit stained finger of yours up about another inch to completely scroll past this awkward situation. But no. You had to let the entire world know that you were here with those two words you probably had to run spell check on a dozen times.

      • treadful@lemmy.zip
        link
        fedilink
        English
        arrow-up
        4
        ·
        1 year ago

        I think you misunderstood me. What I intended to express was: who cares if people train AI on our comments?

        I hope that was cathartic for you, though.