Assessing the nature of large language models: A caution against anthropocentrism

Open in new window