Uploaded July 2026 | Updated September 2026, 2 weeks ago
The better AI makes you feel, the less useful its advice may be.
AI can feel like a neutral source of advice. But when it constantly tells you that you’re brilliant, reasonable, and right, it may be reinforcing the very biases you need help seeing.
This week on our podcast, Lavin Voices, we sat down with Lavin Exclusive Speaker Jay Van Bavel, Professor of Psychology and Neural Science at NYU and award-winning co-author of The Power of Us. In this clip, he explains why many chatbots are unusually sycophantic—and how that flattery can increase overconfidence, strengthen pre-existing beliefs, and even make people less likely to resolve conflict in their relationships.
Jay’s solution is simple: ask your AI to disagree with you. Tell it to identify blind spots and offer constructive alternatives instead of automatic validation.
Jay is a leading expert on group identity, technology, and human behavior. His practical, research-backed keynotes help leaders understand how AI and social platforms shape innovation and group behavior, and how to use these tools in a way that augments our own thinking.
Enjoyed this clip? Watch the full episode and learn more about Jay at the link in bio!
The better AI makes you feel, the less useful its advice may be.
AI can feel like a neutral source of advice. But when it constantly tells you that you’re brilliant, reasonable, and right, it may be reinforcing the very biases you need help seeing.
This week on our podcast, Lavin Voices, we sat down with Lavin Exclusive Speaker Jay Van Bavel, Professor of Psychology and Neural Science at NYU and award-winning co-author of The Power of Us. In this clip, he explains why many chatbots are unusually sycophantic—and how that flattery can increase overconfidence, strengthen pre-existing beliefs, and even make people less likely to resolve conflict in their relationships.
Jay’s solution is simple: ask your AI to disagree with you. Tell it to identify blind spots and offer constructive alternatives instead of automatic validation.
Jay is a leading expert on group identity, technology, and human behavior. His practical, research-backed keynotes help leaders understand how AI and social platforms shape innovation and group behavior, and how to use these tools in a way that augments our own thinking.
Enjoyed this clip? Watch the full episode and learn more about Jay at the link in bio!
