Mad Math

Mathematical models or algorithms envelop our modern lives, from Facebook showing you things it thinks you’ll like, to dating apps showing you people it thinks you’re a good match for. But are these powerful algorithms always acting in our best interests? That’s what Harvard PhD and data scientist Cathy O’Neil investigates in her book, Weapons of Math Destruction. Weapons of math destruction, or what she calls WMDs, are algorithms that claim to quantify important human traits like performance, risk, and credit worthiness, but can instead backfire and produce harmful outcomes instead. O’Neil explains that WMDs often have three things in common: Opacity, scale and damage. Nobody but their creators know how they work, they’re employed in massive scale, and thus have the ability to do a lot of damage if something goes wrong. As a Wall Street quantitative analyst on Wall Street from 2007 to ’09, O’Neil saw first hand how WMDs contributed to the Great Recession. And while people often trust that algorithms are unbiased and purely logical, O’Neil explains in her book how “models are opinions embedded in mathematics,” that is, they can carry on the biases of their creators. Weapons of Math Destruction has been named one of the best books of 2016 by The Guardian, Boston Globe, New York Times, Kirkus Reviews, Boston Globe and more. Even if you’ve never been very good at math, the book is an accessible read about the math that increasingly directs our world.