No-training Creative Commons manton.org
@manton CC was designed so the content creator could indicate their intention of diluting the default (copyright). "CC-NT" doesn't make sense because the default covers it.
@samgrover My reading of CC-NT would be something unique to default copyright: you can use the text for any purpose you want, including commercial use, but you can't train LLMs with it. Not sure if that's exactly what Tantek intended, though.
@samgrover Actually, the Creative Commmons organization has explicitly stated they don't think any of the CC licenses prohibit AI training. They see it as "fair use".. So the default doesn't cover it. See creativecommons.org/2023/08/1...
@manton Maybe the intention is to create an NT clause that could be applied to an existing license, e.g. one could do CC-BY-NT to allow companies to use with attribution but not for training, and CC-BY-NC-NT would allow a researcher to use for any purpose except training. That makes more sense.
@dvdlite Right, but the default I'm referring to is one where a creator doesn't use a CC license, or any other. In that case copyright offers all the protections including "no training", IMHO. Of course, the companies using it for their profit disagree with that, and we'll see how the lawsuits go.
@samgrover Oh I 100% agree with you, but read their paper. Until the lawsuits sort it out, Creative Commons believes AI training is fair use whether you are talking a CC License OR regular copyright. Fun times! 🤣
In the age of AI, one of those benefits is now letting me contribute in a small way to something bigger, in the same way that someone finds an answer in one of my blog posts when they search on Google.
This has been my line for a while. If we are not profiting from our online writing​, why not contribute to a product that can be accessible? We did it for Google, and they made tons of money indexing and presenting everyone's content, so how is AI different?​