OpenAI’s Santosh Vempala Explains Why Language Models Hallucinate

Published October 24, 2025