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.
It’s a complicated answer I’m unqualified to answer but essentially there exists some sort of baseline either for people or for how gpt responds usually and then they can figure out which way the text “leans”