Humans also make mistakes in ways that other humans can understand or expect. Sometimes LLMs make mistakes in a way that makes you say “no human would have ever done that”.

You can not trust human output without verification either. That's why you have tests, qa, staging envs, A/B tests..