Computerize the Race Problem?: Why We Must Plan for a Just AI Future

Abstract

1960s civil rights and racial justice activists tried to warn us about our technological ways, but we didn't hear them talk. The so-called wizards who stayed up late ignored or dismissed black voices, calling out from street corners to pulpits, union halls to the corridors of Congress. Instead, the men who took the first giant leaps towards conceiving and building our earliest "thinking" and "learning" machines aligned themselves with industry, government and their elite science and engineering institutions. Together, they conspired to make those fighting for racial justice the problem that their new computing machines would be designed to solve. And solve that problem they did, through color-coded, automated, and algorithmically-driven indignities and inumahities that thrive to this day. But what if yesterday's technological elite had listened to those Other voices? What if they had let them into their conversations, their classrooms, their labs, boardrooms and government task forces to help determine what new tools to build, how to build them and - most importantly - how to deploy them? What might our world look like today if the advocates for racial justice had been given the chance to frame the day's most preeminent technological question for the world and ask, "Computerize the Race Problem?" Better yet, what might our AI-driven future look like if we ask ourselves this question today?