I was pondering how detection of AI generated text might work. Since GPT and the like are large language models (LLM) with varying context awareness they are basically Markov chains on steroids. Ofcourse none of the detect-AI-generated-text slingers really want to disclose their secret sauce as it w...