3 ms·
That is betting LLMs wont get enough non copyrighted training material to be useful. Especially for text thats likely not the case. Pretty sure its a moot point
by cf141q5325 3y ago
That is betting LLMs wont get enough non copyrighted training material to be useful. Especially for text thats likely not the case. Pretty sure its a moot point anyway as you have no way to proof what was in the training set and what wasnt.
I very much expect parallel construction if push comes to shove.
- stonogo 3y agoAn LLM trained on public-domain or liberal-license material would be a great counterexample ... if anyone built it. The popular models in wide use do disclose what they train on (sometimes under NDA) and they all contained copyrighted work.
- cf141q5325 3y agoThere is no need to build it or tell you atm. Once that does change i expect parallel construction. Once you know what behavior you want from the net, recreating that from non-copyrighted material might turn out to be a lot less difficult.