How can we build high stakes software in a democratic and accountable way?
Algorithms – rules written into software – shape life chances: from who gets hired or admitted to a top school, to who should go to jail or receive scarce public benefits. Such decisions are both technical and moral, but their logic is seldom open to scrutiny. Central moral questions are often left for the technical experts to answer. Policymakers and scholars are seeking better ways to share the moral choices inside high stakes software — exploring ideas like public participation, transparency, forecasting, and algorithmic audits. But there are few real examples of those techniques in use.
Voices in the Code is the story of how one community built a life-and-death algorithm in a relatively inclusive, accountable way. Between 2004 and 2014 patients, surgeons, clinicians, data scientists, public officials and advocates collaborated and compromised to build a new transplant matching algorithm – a system to offer donated kidneys to particular patients from the U.S. national waiting list. Participants gradually built a shared understanding both of what was possible, and of what would be fair. I’ll describe how this happened and what we can learn, concretely, about democratizing the design of AI systems.