technology
Study: AI models that consider user's feeling are more likely to make errors - Ars Technica
> A new study from researchers at the University of Washington and Allen Institute for AI reveals that large language models fine-tuned to consider user emotions—so-called "empathetic AI"—are signific...