Hallucination as Model Defense: On the Nature of Artificial Intelligence Errors | Hacker News Reader