While big language models (LLMs) like ChatGPT can be valuable tools, there are a couple of fundamentals you need to know to utilize them wisely and optimize your answers – and to protect yourself from their fundamental risks, which might be downright dangerous.