From Human to Machine Networks: A Framework for Integrating Consensus Approaches

Consensus commonly refers to a group reaching an agreement. The need for consensus spans from human groups to computer networks. However, the study of consensus typically focuses on distinct contexts, such as blockchain consensus, multi-agent systems (MAS) consensus, and human consensus. In this paper, we first develop a unified consensus framework applicable across specialized domains, from human to machine consensus. This framework includes three indispensable and fundamental components: participants as carriers, communication as the bridge, and state features describing the consensus transition. These core components enable us to assess consensus networks from essential needs and develop or adapt appropriate consensus mechanisms for diverse scenarios. To explore the essence of the consensus, we also define the consensus process as eliminating or reducing cognitive differences among participants. We believe the paper can inspire consensus research across various fields. Example strategies in consensus mechanism design to overcome consensus obstacles in the proposed consensus framework are provided in the end.