previewed. About a two minute read, opp to participate in ongoing research.
Comment: ok, it will be interesting to see how this develops. The fundamental problem is LLM’s (like ChatGPT) are trained on billions of human authored documents, so its not totally clear how they can model generated detected content.