The Emergence of Cooperation in Public Goods Games on Randomly Growing Dynamic Networks

Abstract. According to evolutionary game theory, cooperation in public goods games is eliminated by free-riders, yet in nature, cooperation is ubiquitous. Artificial models resolve this contradiction via the mechanism of network reciprocity. However, existing research only addresses pre-existing networks and does not specifically consider their origins. Further, much work has focused on scale-free networks and so pre-supposes attachment mechanisms which may not exist in nature. We present a coevolutionary model of public goods games in networks, growing by random attachment, from small founding populations of simple agents. The model demonstrates the emergence of cooperation in moderately heterogeneous networks, regardless of original founders’ behaviour, and absent higher cognitive abilities such as recognition or memory. It may thus illustrate a more general mechanism for the evolution of cooperation, from early origins, in minimally cognitive organisms. It is the first example ...

• Type: Journal
• Year: 2016
• Where: EVOW
• Authors: Steve Miller, Joshua D. Knowles
• Keywords: Artificial Intelligence, EVOW 2016