Predicting Poverty by Satellite

Guest: Michael Xie, Computer Science PhD Student, Stanford University Within 12 years, the United Nations hopes to have eliminated world poverty. One hurdle toward that goal is measurement: how to identify poor areas that need support and account for economic improvement happening in other areas? Right now, the basic approach is pretty archaic – survey takers go door-to-door asking people about their assets and income.  But the Stanford’s Artificial Intelligence Lab is working on training computers to identify poverty by looking at satellite images such as Google Earth.