Van Hasselt Lecture: "Sharing a World with Artificially - Intelligent Agents"

By Peter Railton, distinguished professor of Philosophy at the University of Michigan in Ann Arbor

With the development of AI, our world will increasingly be shared with highly-capable artificially-intelligent agents, many operating relatively autonomously. This raises a host of safety concerns, and while various restrictions and “guard-rails” are being designed to impose upon such agents, the openness and complexity of the natural and social worlds we will share with AI agents requires that these agents have some capacities of their own for identifying and responding appropriately to ethical features of situations and actions. Current AI systems and agents owe their open-ended capabilities primarily to learning rather than pre-programming. Might learning processes also enable AI agents to acquire open-ended competence with ethically-relevant features, becoming much-needed allies in promoting safety for AI agents and ourselves alike?