Philip Brown, assistant professor of computer science, received a National Science Foundation grant award of $250,472 for his project titled “Socially Networked Autonomy: How Should Machines Interact with Society?” The three-year award will additionally fund a full-time Ph.D. student to assist in research.

The project studies decision design methodologies for independent agents that network with human societal systems. Brown explained that one of the main motivating questions for this project is “how selfishly should self-driving cars behave?” Naive designers might program a vehicle to be altruistic — that is, to behave in a way that attempts to “make life better” for all the cars around it. However, prior work has shown that is not always a good approach, and that in some scenarios it is actually better to make self-driving cars completely selfish and ignore the benefits to others completely. In this project, the mathematics of this and other similar situations are researched to seek a fundamental understanding of when and why altruism can paradoxically backfire and lead to unintended consequences.