Yeah, AI chatbots are notorious at not understanding their own limitations. I wonder how that could be fixed?
Easier to think about if LLM means Large Limitations Mirror.