Filter bubbles and affective polarization in user-personalized large language model outputs

Open in new window