COllective COordination through ONline Social media

In the coming decades, a defining task for humanity will be to solve global challenges through mass coordination, such as averting the most catastrophic consequences of rapid climate change or curbing the diffusion of pandemics. These challenges can be understood as social dilemmas: the individual benefits are at odds with collective interests, and only concerted action can attain shared gains that outmatch individual costs. The focus of COCOONS is to unveil the prime elements of social interactions that enable spontaneous citizen coordination in the face of social dilemmas. Its ultimate goal is to provide a blueprint to create online participatory platforms that will facilitate their members to find agreement on how to solve problems that can be addressed only with mass action.

The key determinants of cooperation in online communities are largely unknown, and we need more experimental evidence to learn how to build online communities that facilitate cooperation rather than exacerbating polarization. Social scientists have formulated theories that link social dimensions such as knowledge exchange, group identity, and trust to successful coordination. These theories have never been verified at scale because the social dimensions they are based on have been traditionally hard to quantify.

The project will develop AI algorithms for Natural Language Processing that can infer high-level properties of social interactions from conversational text. Guided by existing social theories, we will study how different social dimensions can trigger opinion change and emergence of consensus. Using tools typical of the science of complex systems, we will study large social networks in the light of these dimensions, and design models that can predict coordination dynamics in large communities based on their social network structure and on the social dimensions they voice. Last, we will create a set of guidelines on how to design online communities for promoting grassroots cooperation and agreement, and validate them with controlled experiments.