Do as I Say, Not as I Do: Can We Avoid Imparting Human Biases to Computers?

As the AI revolution is poised to impact nearly every industry, as well as government, education, family, and social life, concerns are mounting about how training computers to think like us can amplify some of the worst parts of ourselves. Is our training data teaching machines to adopt human biases on race, gender, age, and more? We'll look at several concrete examples of these issues, as well as examine how we got here in terms of workforce and educational systems. This starts with basic questions of who decides who gets to see themselves as "a computer person," and how early childhood and educational experiences shape our perceptions of our relationship to technology.

Speaker: Cynthia Lee, Stanford