Always Validate LLM Output

Never blindly trust the output of an LLM, especially for factual information or code. Always have a validation step, whether it's manual review, fact-checking against reliable sources, or running generated code in a sandboxed environment. LLMs can and do make mistakes.