Posted by Rob Knies

Moshe Tennenholtz

For Moshe Tennenholtz, just named the winner of the 2012 Autonomous Agents Research Award by the Association for Computing Machinery’s Special Interest Group on Artificial Intelligence (SIGART), things just keep getting better.

“I think we are living in an exciting time,” says Tennenholtz, Sondheimer Technion Academic Chair and a principal researcher for Microsoft Research New England who is based at the Microsoft Israel R&D Center. “Social and economic structures are reshaped and/or need to be reconsidered due to the use of a variety of what can be called ‘online multi-agent services and interactions,’ both current and emerging.

“The understanding and design of these is what I would love to do.”

The study of multi-agent systems involves the design and analysis of interactions between decision-makers in computational environments. Tennenholtz understands those interactions as well as anybody, as evidenced by his latest award, presented for “substantial and sustained contributions to the foundations of multi-agent systems.”

Those contributions, the award citation states, include:

  • The first formal studies of social laws for multi-agent systems.
  • Contributions to the computational theory of auctions.
  • Multi-agent learning.
  • Computational social-choice theory.
  • Reputation and ranking systems.
  • The notion of program equilibrium.

Tennenholtz provides his understanding of why he received the award.

“The work on social laws,” he says, “is an example of work on the interplay in which artificial intelligence and computer science meet game theory and social sciences, done before there had been a solid academic profession dealing with this interplay. The work on things such as program equilibrium is an example of work that lies on the bridge between the fields, rather than in applying one field to the other.

“Being among the first ones to deal with the above interplay and with work that lies on that bridge were probably the main reason for the award.”  
 
But that’s not all. Tennenholtz also has a significant history of community service, including time spent as editor-in-chief of the Journal of Artificial Intelligence Research.

The Autonomous Agents Research Award is presented each year for research excellence in the area of autonomous agents. Funded by an endowment created by SIGART from the proceeds of autonomous-agents conferences, the honor is intended to recognize researchers whose current work is an important influence on the field.

Now, with the announcement of the award, Tennenholtz’s achievement finally can be recognized by his family, friends, and colleagues.

“I knew that I was a candidate,” he divulges, “but I was still surprised, and I could not tell my friends, because I was asked to keep it confidential for a few days.”

After accepting a few congratulations, Tennenholtz will get back to work. There’s simply too much to be done.

“One of the challenging aspects here,” he concludes, “is the design and analysis of potential new services and interactions without relying only on data analysis. Data by itself is shaped, in many cases, as a result of implementing a particular service or interaction.”

For Tennenholtz, conquering those challenges will constitute the real reward.