, University of Maryland
Pages: p. 15
It was with great sadness that I learned of Push Singh's death on 28 February this year. This remembrance is in lieu of the essay we expected to publish for his "AI Ten to Watch" award. It would be hard to write even if I knew him only through the paperwork submitted with his nomination. In Push's case, however, it's much harder, for I had recently begun talking with him about research ideas, and he had some exciting thoughts that I was eager to pursue during an upcoming sabbatical.
Push started at the Massachusetts Institute of Technology as an undergraduate in 1988, staying on as a graduate student and completing his thesis in MIT's Department of Electrical Engineering and Computer Science in 2005. His advisor was Marvin Minsky, one of AI's founders. Push was slated to begin a position as a faculty member in the MIT Media Laboratory in 2007 after, as he described it to me, a much-needed year off "to think."
Push's research was based partly on Marvin's society-of-minds approach, exploring what common sense was and how it could develop. His thesis, "EM-ONE: An Architecture for Reflective Commonsense Thinking," included the creation of Open Mind Common Sense ( http://csc.media.mit.edu), a Web-based approach to acquire commonsense knowledge from the general public. He also developed a layered cognitive architecture aimed at exploring how reasoning about physical, social, and mental domains could work.
In remembering him, Marvin Minsky said, "Push had just been appointed to become a new professor at MIT, to pursue what we knew would be a brilliantly productive career. For several years we have been designing an ambitious project to develop [commonsense] theories, which have slowly been gaining the interest of many researchers inside and outside the Media Lab. To us his loss is indescribable because of how we could communicate so much and so quickly in so very few words, as though we were parts of a single mind."
Henry Lieberman, who worked with Push on the Open Mind project, said, "Push and his students worked on collecting common sense and developing architectures and tools surrounding the knowledge. My students and I concentrated on applying this knowledge to improve all kinds of interactive applications: browsers, editors, games, phones, etc. Push helped me teach my course, where we taught students how to understand and work with commonsense knowledge. It has now grown to the point that there are numerous projects involving common sense in many, if not most, of the other groups at the Media Lab, and also elsewhere. This collaboration has been one of the most productive and fruitful of my career, and I thus owe him a tremendous debt that I shall never have the opportunity to repay." (For these and many other tributes to Push, see http://pedia.media.mit.edu/ wiki/Push_Singh.)
In my own discussions with Push, we were exploring how to put his commonsense work, especially the Open Mind project, together with the Semantic Web work I do. From the first moment we talked, I could tell he was a brilliant young man with inspiring ideas. I looked forward to working with him, fully expecting to learn more than I could teach. We planned a visit for him to discuss his research and how we could proceed. While I didn't know him as well as Marvin, Henry, and the many students, faculty, and friends he had at MIT and beyond, I can't believe I'll never get a chance to pursue these ideas with this talented young scientist. He will be missed.