OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling’s Harry Potter series::A new research paper laid out ways in which AI developers should try and avoid showing LLMs have been trained on copyrighted material.
Stupid. No it isn’t. Establishing legal precedent or, in countries that don’t work on precedent, a preponderance of legal cases, prohibiting this practice is what is needed.
It will be even worse if you must pay for all data to train an AI because it will make the systems even more exclusive. Copyright as a law is incompatible with AI and the change must be to require models trained on controlled works to be provided free.