The fundamental questions that motivate me: How can computers (and humans) learn about the world despite our imperfect sensing, limited reasoning, and the known theoretical limitations on induction? What are the necessary and sufficient computational components of intelligent systems? How can AI systems perform robustly in an incompletely-modeled world?
I also seek to develop machine learning and AI techniques that can make the world a better place. These range from applications in ecological science, sustainable development, and biological conservation to problems in law enforcement, national defense, and international peacekeeping.
Most recently, I have joined the discussion of the implications of AI for government and economic policy. I am especially interested in finding ways that AI can empower people in their work and as members of society.