• FoxAlive@lemmy.zip
    link
    fedilink
    English
    arrow-up
    21
    ·
    edit-2
    2 days ago

    In no possible way can this be construed as a good or positive thing.

    Going to call it now. Its going to be a csam generator like grok. Mark zuckerburg won’t face any consequences for his actions, people will make a fuss out of it. He will do some blank appology promising to fix it, but ultimately because you can’t stop a prediction model without more LLMS and more ai moderating the main ais, you will never be able to assure safey.

    Like elon musk it will get whisked away, partially hiding it or rebranding it so only paid users will be able to abuse the llm for whatever weird stuff they do with it, and people will move on and no one will get in trouble.

    Also you are in the training data, you have and always been in their training data. Even if they are using something similar to a qwen model for this Instagram bs, meta has 100% used your photos, conversations etc to train their other shitty LLMS that no one uses.

    • only paid users will be able to abuse the llm

      J’accuse! You cannot abuse something which does not think, which is why it’s totally fine for me to be like “what’s the most fucked up thing you can think of” to it, a million times in a row