A viral AI-doom video and one hedged research paper explain why overclaiming wins, and why calibrated uncertainty is honesty wearing an anxious face.
