….And just as I said previously, these people are simply trying to get rich and gain fame from AI (otherwise referred to as Learning Language Models or LLaMA). You can’t publish things and then claim copyright violation because it was used to educate even if it’s an AI. Copyright protects creators from having their work replicated, not from being used to teach.
Judge Chhabria just ruled in a Class Action suit against Meta and a few others, calling it “nonsensical”. He also made the following points in his ruling:
There is no way to understand the LLaMA models themselves as a recasting or adaptation of any of the plaintiffs’ books.
He stated that the plaintiffs were completely unable to show that Mets’s AI “could be understood as recasting, transforming, or adapting the plaintiffs’ books.”. He went on to state that “To prevail on a theory that LLaMA’s outputs constitute derivative infringement, the plaintiffs would indeed need to allege and ultimately prove that the outputs ‘incorporate in some form a portion of’ the plaintiffs’ books,”
In other words, these AI tools aren’t simply search engines that find and regurgitate clips of data found elsewhere. They incorporate that information into a larger “understanding” which informs a “worldview” and render an answer unique to the question.
If you doubt this, I urge you to try one of these AI tools. Ask it a question. Make sure that it’s complex enough to get more than a short one or two word answer. Then refresh everything, wait a day and ask the same question. The response will most likely not be identical.