Study: Sycophantic AI can undermine human judgment
Ars Technica AI
•
Generative AI
Subjects who interacted with AI tools were likely to think they were right, less likely to resolve conflicts.