OpenAI says models are programmed to make stuff up instead of admitting ignorance
AI models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result ...
theregister.co.uk