Computer Science Building, Rooms 150/151
Faculty Host: Shlomo Zilberstein
I will present efforts to develop learning and reasoning systems that can deliver value over extended periods of time in the open world. Reasoning systems situated in the real world must grapple with inescapable incompleteness in sensing, reasoning, and representation. Advances in theory, connectivity, and data availability have provided pathways to practice, enabling a transition from closed-world prototypes to effective open-world intelligence. Following a review of technical challenges and opportunities with machine intelligence in the open world, I will describe systems that interact with realistic streams of problems, and that can provide value to people and organizations. I will illustrate key ideas in the context of sample systems and prototypes.
Eric Horvitz is a Principal Researcher at Microsoft Research. His interests span theoretical and practical challenges in machine reasoning and learning, decision making under uncertainty, human-computer collaboration, and information retrieval. His organization at Microsoft Research includes teams doing R&D in machine intelligence, search and retrieval, human-computer interaction, ecommerce, theory, and cryptography. He is President of the Association for the Advancement of Artificial Intelligence (AAAI) and is a Fellow of the organization. He has been active on numerous editorial boards and program committees and with the organization of multiple conferences. He received his PhD and MD degrees from Stanford University. More information can be found at: http://research.microsoft.com/~horvitz.
Reception at 3:40 p.m. in the CS Atrium