Mindmasters: The Data-Driven Science of Predicting and Changing Human Behavior

Artificial intelligence monitors internet users’ every move, for good and ill, according to this eye-opening debut exposé. Surveying the disturbing reach of digital surveillance, Matz, a professor at Columbia Business School, notes that smart phones track owners’ movements; that Facebook, Google, and YouTube deduce users’ thoughts and interests, as revealed by searches, posts, and likes; and that Roombas transmit photos of people’s homes, as one woman discovered when someone associated with the vacuum’s manufacturer leaked an image of her in the bathroom. This data enables companies to build elaborate psychological profiles with potentially costly consequences, Matz writes, suggesting that insurance companies might target ads at anxious individuals, and that banks might deny loans to people they determine to be high in “agreeableness,” per the Big Five personality test (studies have shown such individuals default on payment at higher than average rates). Matz argues that while personal data can be put to good use (her own work has used microtargeted ads to persuade people to save more), safeguards should be implemented to protect against abuses. For example, she suggests that data sharing should be opt-in by default and that “data cooperatives” could allow people to voluntarily pool their data for analysis or monetization while maintaining their ownership over the info. Peeling back the curtain on the long lives of user data, this unsettles. Agent: Leila Campoli, Stonesong Literary. (Jan.)