My technical background is in Artificial Intelligence -- Knowledge Representation and Reasoning, in particular. Very early in my career at Bolt Beranek & Newman Inc., I developed a knowledge representation system called KL-One, which was based on my Harvard (1977) Ph.D. thesis work. Ultimately this work became the basis for the field of Description Logics. The work on KL-One was done in collaboration with many outstanding colleagues both at BBN and at USC-ISI, as well as my thesis advisor, Bill Woods. In 1981 I moved to the Fairchild Laboratory for Artificial Intelligence Research, in Palo Alto. There, with Hector Levesque and Richard Fikes, I helped develop the Krypton hybrid representation and reasoning system. We also, with Peter Patel-Schneider, created the Kandor system, a miniature version of Krypton, to experiment with pragmatic KR systems and applications in information retrieval. During my four years in California, I became Program Chair of the National Conference on Artificial Intelligence, and Hector and I published one of our more important papers, on the tradeoff between expressiveness and tractability in knowledge representation systems. This work inspired a rich subsequent literature exploring the complexity of reasoning in description-logic-based knowledge representation systems that continues to this day.
In 1985, I moved east to join AT&T Bell Laboratories and build an AI research group. Over the following years, that group -- the AI Principles Research Department -- grew substantially, and eventually became two departments, the second one focusing on machine learning, information retrieval, and natural language processing. Amongst our more well-known work in knowledge representation at AT&T was the CLASSIC system, a well-founded description logic system with limited representational power in order to support tractable reasoning. With CLASSIC, we implemented the PROSE product configuration system, which was deployed and used in AT&T and later, Lucent Technologies. There were some 15 different PROSE configuration products built, and overall PROSE systems processed more than $5 billion worth of equipment orders. Our AI team at AT&T became one of the stronger groups in the world over the almost two decades of its existence, having counted in its ranks at various times a substantial number of truly outstanding researchers, among whom were David Etherington, Bart Selman, Henry Kautz, Fernando Pereira, Michael Kearns, Rob Schapire, Julia Hirschberg, David McAllester, Rich Sutton, Peter Stone, William Cohen, Michael Littman, Yoram Singer, Yoav Freund, Peter Patel-Schneider, Deborah McGuinness, Charles Isbell, Steve Abney, Michael Collins, Amit Singhal, David Lewis, and many others.
In 1994, I was promoted to become the Director of the lab we were in (Software and Systems Research). In 1996, as AT&T was spinning off Lucent Technologies, I was part of the original senior management of AT&T Labs, and helped form AT&T Labs-Research, at that time led by A. G. "Sandy" Fraser. At that time, my lab became the Information Systems and Services Research laboratory, and besides AI (led by Fernando Pereira and Henry Kautz, and subsequently, Michael Kearns), we had strong teams in HCI (led by Julia Hirschberg and later, Candy Kamm), Secure Systems (led by Dave Maher and then Bill Aiello), IP communications services (led by Larry Jackel), Online platforms and innovative Web services, including secure digital music distribution (led by Gregg Vesonder) and Customer Decision Modeling (led by John Rotondo). We created innovative services ranging from audio-enhanced instant messaging to Universal Message Access (eventually serving almost 3 million customers) to collaborative listening to music. In 2001, we changed the name of our lab to Communications Services Research to reflect our focus on unified communications and related services.
In addition to working in AI, I have been active in the Knowledge Discovery in Databases research community (I have been on the KDD Conference Steering Committee), and have published the information retrieval, database management, and software engineering literatures.
Between 2003 and 2006, I have served as President of what was then the American Association for Artificial Intelligence (AAAI) (now the Association for the Advancement of Artificial Intelligence). Prior to that, for 9 years I was the Secretary-Treasurer of IJCAII -- International Joint Conferences on Artificial Intelligence, Inc. I stepped down in 2002 (yielding to my successor, Samy Uthurusamy, of General Motors) in order to be able to perform my duties as President of AAAI. I have also been a member of the Editorial Board of the journal, Artificial Intelligence, as well as several other boards. I currently co-edit the Synthesis lecture series on Artificial Intelligence and Machine Learning. I am also a member of the Board of Directors of the Computing Research Association and serve as its Treasurer.
Since January of 2015, I have been serving as an advisor for an exciting and innovative startup called Segovia - Segovia builds enterprise-grade technology to support cash transfers to help in the fight against extreme poverty, and also to serve the humanitarian community in times of significant need. Segovia aims to dramatically improve the distribution of essential resources to those who have the least. Segovia has a small but outstanding team of top technical people with exceptional backgrounds, and was founded and is led by the outstanding team of Harvard-trained economics PhD's that created GiveDirectly, whose mission is to allow individuals to send money directly to the extreme poor, mainly in Africa.