Hacker News Reader

Would that not be the person running the AI? The one giving prompts and verifying that prompts are fulfilled adequately?

If I've learned anything, its people often never verify the correctness of anything automated.