"Imagine with Meta AI" turns prompts into images, trained using public Facebook data.

For those who think copyright approaches to Generative AI are going to be a solution. They won’t be

#Facebook and Adobe already have the rights to all the images legally they need, which means that they will make GenAI models and there’s nothing you can do about it.

Copyright attacks will only kill the Open Source versions will mean that everyone will be pushed to these priorietary models instead, benefiting only the rich corpos.

  • heavyboots@lemmy.mlEnglish
    3·
    1 year ago

    Image-based AI doesn’t bother me nearly as much although I strongly believe it should be required to have digitally embedded markers to tag it as AI-generated.

    Also, it is somewhat hilarious to me that all those people who willingly gave their content to Meta are now reaping what they sowed.

    Text-based AI is the real danger though as it pollutes the Internet at large with AI-generated misinformation and is much harder to tag as AI-generated.