Stanford study finds widely cited LLM claims often expire within months
🤖 AI-generated content — The title and summary were produced automatically by artificial intelligence, without human editorial review.
Researchers Owen Queen and James Zou checked 50 widely cited LLM claims against 20 models from eight developers (July 2024–September 2026). Forty-four percent of findings had already expired; for 22 with estimated dates, average lifespan was 15.7 months. Another 34% weakened sharply, 16% still hold, and 6% strengthened; after two years, 39% were expired. Claims about what models cannot do aged fastest; findings on preferences and evaluator behavior lasted longer.
Continue in the app — vote & join in ➔Source: The Decoder (AI) · via ahirlevel.hu