Comment on AI’s Unpaid Debt: How LLM Scrapers Destroy the Social Contract of Open Source

<- View Parent
fodor@lemmy.zip ⁨1⁩ ⁨week⁩ ago

Several explanations exist. First ShareAlike or GPL require re-sharing of downstream content. But if things are used for training, and later the model produces new code, that isn’t implicated by copyright legislation… And we could discuss whether that’s ethical or moral, of course, and there are various opinions…

I feel like this is similar to the weakness of BSD licenses, Public Domain releases, and CC BY licensing. Someone can come along and take all of your work, polish it a little better, and sell their new service. Then you don’t get recognition or support, they get some contracts from their friends’ companies for a few years, and you feel sad.

The other commonly-remarked angle is the death of the Web. Because so many websites are script-generated copy/paste, and they are tweaked to fit SEO, and Google doesn’t give a fuck, it’s hard for real website authors to get seen, and without any visibility, their work ends up being personal or pointless. This isn’t limited to open source, but it’s closely connected.

source
Sort:hotnewtop