Sure, just turn left.
Hallucination is when an AI makes up confident-sounding but false information.
Imagine your GPS confidently telling you to turn left into a lake because it 'thinks' there's a bridge. That's what AI hallucination is like: the AI creates plausible-sounding but completely made-up facts, often because it's trying too hard to give an answer.
Hallucination matters because it can spread misinformation or cause real harm when people trust AI outputs blindly. You encounter it in chatbots, search summaries, and AI-generated content that sounds right but isn't.
Many think hallucination is a bug or glitch, but it's actually a feature of how large language models work — they predict the next likely word, not truth. This means even the most advanced AI can hallucinate confidently without 'knowing' it's wrong.
In AI, hallucination refers to the generation of outputs that are factually incorrect, nonsensical, or ungrounded in the input data, yet presented with high confidence. It arises from statistical language models that prioritize coherence and plausibility over factual accuracy, especially when the model lacks relevant training data or is prompted beyond its knowledge boundary.