Tony StubblebineMar 26, 2023

Is there any benefit to the author of letting a for profit LLM like OpenAI spider and consume their writing? I can’t think of any.

If that’s true, then should we be rethinking Creative Commons licenses? I’m wondering if the current one we use at Medium is obsolete.

At least with something like Google there was an exchange of value: contribute to their search results and get traffic back.

Show threadmcc
@coachtony The problem is bigger than that: LLMs don't follow the licenses, so it doesn't really matter what license you use or what special clauses you add to your license. I expect LLM companies do not even make an effort to track what license the material they are spidering
Mar 26, 2023 at 5:41pmWeb
Show threadTony StubblebineMar 26, 2023

@mcc I wonder if any other content companies, i.e. WordPress, Substack, StackOverflow, care?

In matters of law there are almost always two alternative paths: relationship or power.

I don't think the platforms are powerless here.

Show threadmccMar 26, 2023
@coachtony Several lawsuits with various filers and targets are ongoing https://www.theverge.com/2023/1/17/23558516/ai-art-copyright-stable-diffusion-getty-images-lawsuit
Getty Images is suing the creators of AI art tool Stable Diffusion for scraping its content

Getty Images is suing Stability AI, creators of generative AI art model Stable Diffusion. The stock photo company claims Stability AI ‘unlawfully’ scraped millions of images from its site.

The Verge
Show threadBart Veldhuizen 🚀Mar 26, 2023
@coachtony @mcc Objaverse just ingested 800k CC licensed 3D models from Sketchfab in their training set. We now have noAI tags for users and no-scraping language in our TOS, but this happened before those went into effect.
Show threadmccMar 26, 2023
@BartV @coachtony Which CC license?
Show threadBart Veldhuizen 🚀Mar 26, 2023
@mcc @coachtony different types, but all CC-BY. We also let creators add clauses like NC, SA etc. I’m reading the the attribution requirement might be a blocker for AI use, but I haven’t found any lawsuits that prove this.
Show threadmccMar 26, 2023

@BartV @coachtony It seems to me that if a court held a derivative work of your scraped data was unbound by license requirements like attribution unless you add a magic additional "noAI" tag (which there probably wasn't even a standard for at the time the data was uploaded), this would be utterly absurd.

I am also not aware of any lawsuits proving any of these scraped large models are *legal*.