Comment by callc

A good time to teach a hard lesson about the trustworthiness of LLM output