Not long ago, I shared my quest to get AI vendors to openly acknowledge that LLMs (sometimes) disregard RAG/context in favor of their own pre-training. I found one vendor in particular that has an ...