Another example of why you should be using ML monitoring and guardrails in your LLM stack 😬 https://time.com/6564726/ai-chatbot-dpd-curses-criticizes-company/