The Translation Imperative

Keith Sonderling, Commissioner, U.S. Equal Employment Opportunity Commission (EEOC); David Danks, Professor of Data Science & Philosophy, University of California; Cathy O’Neil, CEO, ORCAA; Susannah Shattuck, Head of Product, Credo AI

Ensuring that AI systems are responsibly developed, deployed, and used takes a village. From technical and legal to business and policy, many different perspectives are needed to ensure the consequences of AI systems are effectively accounted for and managed. However, coordinating these diverse perspectives and translating them into one common language for the benefit of AI development can be challenging. Disciplinary silos, varied technical sophistication, and a multiplicity of core values makes strategic alignment difficult and threatens efficient action. The challenge and imperative of including diverse perspectives, point to a specific need for a translation layer - a language that supports the critical reflection and coordination needed by these many stakeholders.