Much of the discussion around the ethics of AI focuses, appropriately, on questions of bias and what AI developers need to know about fairness. But AI, especially of the sort known as machine learning, is also teaching us some lessons about fairness. There are multiple types of fairness. Not all of them require fully transparent AI systems and fair outcomes often require making difficult trade-offs among cherished values. With this in mind, David Weinberger from the Berkman Klein Center explores how can we decide on the values we want AI systems to support.
What fairness can learn from AI
