Can machines spot a lie better than we can? I talk with psychology researcher Bennett Kleinberg, whose work brings together psychology, machine learning, natural-language processing, and deception detection. We explore why humans are bad at identifying lies, which verbal clues may reveal deception—and why nervousness is often misleading. Bennett explains how liars use detail and how level of detail can be a clue to deception, the fact that deception usually involves a good amount of truth in the mix and what that means, and what happens when human judgment is combined with AI. We also discuss the risks of the mysterious black-box nature of some machine learning systems, and how smarter, research-informed interviewing techniques could transform criminal investigations. For more details and a transcript, see this episode’s entry at www.behavior-podcast.com. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related