large language models (LLMs) are capable of producing high quality
information at unprecedented rates. As these models continue to entrench
themselves in society, the content they produce will become increasingly
pervasive in databases that are, in turn, incorporated into the pre-train