Collins Conference Room
Working Group
 

Meeting Summary: A longstanding concern of the social sciences has been to understand why various social organizations have the particular (network) structures that they do. A broad conclusion—voiced by people from Coase to Arrow to Radner in economics, and from Wright to Johnson in anthropology—is that one of the determining factors for how a social group is organized is the need for the organization as a whole to transform information from its environment into the actions that the group members take, while obeying constraints on how fast and accurately the (human) members of the group can communicate with one another.

Despite wide-spread agreement with this broad conclusion, much of the actual associated research has been semi-formal or limited in scope. In this working group we propose to revisit this topic, in a more principled manner. Our ultimate goal is to marry real-world and simulation-based data with recent breakthroughs in (telecommunication) network coding theory, deep learning, dynamic Bayes nets, game theory on networks, and other modern tools for analyzing communication networks.

Purpose:
Research Collaboration
SFI Host:
David Wolpert and Tim Kohler