obviously what I said is not an actual court case and is meant to be a facetious condensation of what would happen. so after the several days when Disney shows

obviously what I said is not an actual court case and is meant to be a facetious condensation of what would happen. so after the several days when Disney shows how this LLM was obviously trained on the dataset of available Disney content, and then the defendant responded that they trained on a corpus of pseudo-Disney but then could not produce this corpus it would be reasonable to conclude they were lying. The parent suggested "No one will be able to prove the source of the data" and that is the kind of thing that programmers for some reason often think is some fantastic gotcha so one can really get away with anything, but it is these kinds of things that the law is generally pretty good in handling.