The Wisdom of the Network: How Adaptive Networks Promote Collective Intelligence

Abstract

Social networks continuously change as people create new ties and break existing ones. It is widely noted that our social embedding exerts strong influence on what information we receive, and how we form beliefs and make decisions. However, most studies overlook the dynamic nature of social networks, and its role in fostering adaptive collective intelligence. It remains unknown (1) how network structures adapt to the performances of individuals, and (2) whether this adaptation promotes the accuracy of individual and collective decisions. Here, we answer these questions through a series of behavioral experiments and simulations. Our results reveal that groups of people embedded in dynamic social networks can adapt to biased and non-stationary information environments. As a result, individual and collective accuracy is substantially improved over static networks and unconnected groups. Moreover, we show that groups in dynamic networks far outperform their best-performing member, and that even the best member's judgment substantially benefits from group engagement. Thereby, our findings substantiate the role of dynamic social networks as adaptive mechanisms for refining individual and collective judgments.