The hazard with modern industrial LLMs is that they do not have a method to signal "inadequate information:" their solutions are usually self-confident. This also relates to using ChatGPT for coding: It might create code that is non-useful or insecure. My guideline will be to rely on the chatbot only https://julianm431ltb9.evawiki.com/user