Societal intelligence - a new perspective for highly intelligent systems

Abstract

A novel concept of intelligence called “societal intelligence” and its related architecture for solving complex problems are introduced. The idea is based on what we consider on the “intelligence of human society”. For illustrative purposes, a case study is realized, which involves the solution of a difficult problem in a societal multi-agent system where the agents operate in an unknown environment. In the simulated robotic mobile multi-agent system, the agents adapt their movement control in the environment based on some global knowledge constructed by the system. Besides the proposed architecture, a novelty presented in the paper is the demonstration that even in a simplified knowledge-based multi-agent system, if the principles of societal intelligence are followed, powerful global intelligence emerges at the system’s level.