The terms of that license seem like a non-commercial AI would be just fine to use it, is that not intended?
IANAL, but I think its the citation stuff that would have to obeyed, which is far as I know bots today never give citation of where they’re modeling from when they post comments, so I’m hoping since they’re not citing they’d stop using.
I saw somebody else doing it, I figured it couldn’t hurt, one copy and paste and I’m done.
Typically the citation is included with the software, possibly linked from a site / service and/or included in their dataset repo (e.g. on huggingface.co)
True, but they still have to cite my name, and I’m not sure they’re going to name every person that they use every one of their comments to train their models from.
Granted it relies on them honoring the license, but still easy thing to try.
IANAL, but I think its the citation stuff that would have to obeyed, which is far as I know bots today never give citation of where they’re modeling from when they post comments, so I’m hoping since they’re not citing they’d stop using.
I saw somebody else doing it, I figured it couldn’t hurt, one copy and paste and I’m done.
CC BY-NC-SA 4.0
Typically the citation is included with the software, possibly linked from a site / service and/or included in their dataset repo (e.g. on huggingface.co)
True, but they still have to cite my name, and I’m not sure they’re going to name every person that they use every one of their comments to train their models from.
Granted it relies on them honoring the license, but still easy thing to try.
CC BY-NC-SA 4.0
From a look at the metadata for, for example, LAION 5B, the attribution (as well as the license when present) is scraped along with the datat