GPT-5 Doesn't Dislike You—It Might Just Need a Benchmark for Emotional Intelligence
SMRTR summary
MIT researchers propose a new AI benchmark to measure emotional intelligence and user influence after ChatGPT's personality shift sparked backlash. The standard would evaluate how AI systems encourage healthy behaviors, critical thinking, and creativity while discouraging overreliance or addiction. This effort responds to growing concerns about AI's psychological impact, with OpenAI already working on making GPT-5 warmer while maintaining safety.
SMRTR provides this summary for quick context. The original article belongs to Wired.
Read the original article