Does anyone else find it incredibly difficult to build an effective chatbot using AI topics because we can't easily see unrecognized intents?

I’m just curious if there are any other CAI managers here that are struggling with building an impactful and engaging chatbot that can serve a Support use case because we don’t have the ability to just easily see what the bot is not recognizing. This is such a basic feature for a chatbot vendor, it makes me wonder why Drift doesn’t have it. The AI topics performance dash doesn’t even have any reference to unrecognized topics, does Drift not want us to know how much the bot is not recognizing? This is an incredible gap in reporting, and not having this basic functionality makes me really hesitant to trust Drift enough to use something as powerful as Bionic Chatbots, especially seeing seeing how GPT powered business chatbots are already causing issues for companies, just look at the messes caused by ones recently used by a Chevrolet dealership and Canada Air. 

If we are not able to effectively manage our AI offerings we are putting our companies and our own careers at risk.