Large language models eroding science understanding: an experimental study

Harry Collins, Hartmut Grote, Paul Newbury +2

This paper is under review in AI and Ethics This study examines whether large language models (LLMs) can reliably answer scientific questions and demonstrates how easily they can b…