The widespread trust in AI chatbots as all-knowing oracles is a dangerous misunderstanding of what they actually are. Large Language Models are sophisticated pattern-matching tools, not truth machines, a critical distinction defining how we should use them.