Eric Horvitz

Technical Fellow & Director, Microsoft Research Labs

Pursuing research on principles of machine intelligence and on leveraging the complementarities of human and machine reasoning. Deeply curious about the computational foundations of intelligence. Passionate about harnessing computing advances to enhance the quality of peoples' lives through insights, services, and wellness. The Microsoft Research home page is a starting point for learning about people and projects at Microsoft Research labs across the world. Hear directly from folks via our podcast series.

      Publications       For the general reader

Selected readings:   Bounded rationality   Human-AI collaboration   AI & Privacy   Data & public health

Recent events

Research overview

I've long been curious about the computational foundations of intelligence: How do our minds work? What computational principles and architectures underly thinking and intelligent behavior? I've pursued answers via studies of machine perception, learning, reasoning, and decision making. Many questions remain unanswered and much research is to be done. On the way to a deeper understanding, I work to field working systems that can immediately deliver value in the world. Projects include efforts in time-critical decisions, information retrieval, healthcare, urban infrastructure, sustainability, and development--with goals of understanding how computational models perform amidst real-world complexities, and of deploying systems that deliver value to people and society. A key focus of my work has been on opportunities to leverage the complementarities of human and machine intelligence. Related interests include machine learning and decision making for crowdsourcing and human computation, information triage and alerting that takes human attention into consideration, spanning work on notification systems, surprise modeling, multitasking, and psychological studies of interruption and recovery. On the more theoretical front, I've been long interested in offline and real-time optimization of the expected value of computational systems under limited and varying resources. Areas of concentration in this realm include flexible or anytime computation, ideal metareasoning for guiding computation, compilation for reducing real-time deliberation, ongoing, continual computation, and the construction of bounded-optimal reasoning systems--systems that maximize the expected utility of the people they serve, given the expected costs of reasoning, the problems encountered over time, and assertions about a system's constitution. Research in this arena includes tackling hard reasoning problems with learning and decision making methods.


Completed cycle of service as president and then past-president of the Association for the Advancement of Artificial Intelligence (AAAI), and remain active with AAAI strategic planning. Serving on the Computer Science and Telecommunications Board (CSTB), Board of Regents of the National Library of Medicine, on the External Advisory Board of the Center for Causal Discovery, on the board of Partnership on AI, and the scientific advisory boards of the Allen Institute for AI (AI2) and the Simons Institute. Recently served as chair of the Section on Information, Computing, and Communication of the AAAS, President's Council of Advisors on Science and Technology (PCAST) NITRD Working Group, NIH Advisory Committee to the Director NLM Working Group, and National Academies Committee on Information Techology, Automation, and the Workforce.

Fellow of the AAAI, ACM, AAAS, ACMI, the American Academy of Arts and Sciences, and National Academy of Engineering (NAE), and elected to the CHI Academy.

Blast to the relevant past.

Several links       @erichorvitz       One Microsoft Way, Redmond WA 98052