March 30, 2020

What fairness can learn from AI

TL:DR;

  • Fairness may seem like a simple, binary idea — something's fair or it's not — but when dealing with machine learning, it's really a challenging matter of trade-offs.

Most discussions about ethics and artificial intelligence focus on questions of bias, but there’s also much to learn from machine learning about fairness. Sometimes, fair outcomes require us to make difficult trade-offs among cherished values. 

“The requirement that machine learning places on us to tell it what we count as ‘fair’… brings us face-to-face with the complex, complicated nature of fairness.”
In this tech talk from the Harvard Business School Executive Education Advanced Management Program, David Weinberger of the Berkman Klein Center explores fairness in the world of AI, and asks: what counts as relevant, and what trade-offs are necessary to be “fair”?

Related Topics

People in this Article