Study: Meta AI Model Can Reproduce Almost Half of Harry Potter Book
Study: Meta AI Model Can Reproduce Almost Half of Harry Potter Book
The research could have big implications for generative AI copyright lawsuits.
Why is plagiarism an accomplishment? I can reproduce the entire Harry Potter book with a scanner and some OCR software.
That's the point of the article, I think. They're saying AI models are copyright violators.
It isn't. Its a sign of poor-quality work.
It's a sign that the AI companies are complicit in industrial level copyright violation. There was a recent US court decision that a book publisher brought again Anthropic. In it the judge rules that the use of the book contents was "spectacularly transformative" (and therefore "fair use") because the resulting machine did not copy the work.
People are trying to prove the judge wrong. It does copy the work.
Its also interesting that this is the most conservative, pro “its not just memorizing” estimation possible : they multiplied the probabilities of consequent tokens. Basically it means if it starts shitting out a quote it will not be able to stop quoting until their anti copy the whole book finetuning kicks in after 50 words or so.
It can probably output far more under a realistic test (always picking the top token, temperature =0)
Which half?
Just the pronouns and articles, some of the verbs and adjectives