News for hackers. Stuff that matters.
Re:Meta's new AI image generator was trained on 1.1B Instagr... by ezoe on Thursday December 07, 2023 @08:37PM (#38564319)
Another method of copyright laundering is doing ML learning in a country where it doesn't protected under copyright law.

Personally, I'm on a side of using copyrighted data for machine learning input source doesn't violate copyright. Statistically, learned model for generative Ai doesn't retain even 1 bit of input. It's hard to say NN model data infringe any copyright of the input source. The copyright is applied to the expression, not the process. If the generative AI produces an image that's clearly a copy of a specific Ironman image which existed before the image generation, that's copyright infringement.

Re:Meta's new AI image generator was trained on 1.1B Instagr... by abrookewood on Thursday December 07, 2023 @09:13PM (#38564581)
"learned model for generative Ai doesn't retain even 1 bit of input". If that was true, it shouldn't be possible to trick the models into regurgitating their source material, but cleary that is possible [0].

[0] https://stackdiary.com/chatgpts-training-data-can-be-exposed...

Re:Meta's new AI image generator was trained on 1.1B Instagr... by Kubuxu on Friday December 08, 2023 @08:49AM (#38568996)
Very LLMs are quite a different from diffusion models. The model size vs training set size is skewed the other way.
Re:Meta's new AI image generator was trained on 1.1B Instagr... by AlotOfReading on Thursday December 07, 2023 @09:34PM (#38564717)
Copyright doesn't require a single bit of input to be shared. You can't avoid copyright by using a paintbrush for example, you're simply creating a derived work. You might still be in violation even if you create an entirely new context around the copied elements or substitute for the original in the market, as was the case with Warhol vs. Goldsmith.

Obviously not every generative output is a copyright violation, but it seems equally clear that there are outputs that would be if they were produced by humans.

Re:Meta's new AI image generator was trained on 1.1B Instagr... by WanderPanda on Thursday December 07, 2023 @08:42PM (#38564357)
I agree with you but I think the argument is flawed. If you think about it like this h265 also just steals 10% (or whatever the compression ratio is) of an artifact
Re:Meta's new AI image generator was trained on 1.1B Instagr... by lawlessone on Friday December 08, 2023 @10:10AM (#38569941)
>learned model for generative Ai doesn't retain even 1 bit of input.

It does. The data is just obfuscated.