Evolution of confrontation and cooperation in simple organisms as a function of environmental resources and cost of a conflict

Abstract

AbstractThe root cause of human conflict needs to be understood but it is currently unknown whether the decision to engage in conflict is an inherited or acquired trait. This article reports two experimental simulations which demonstrate that the level of confrontation in a population of simple organisms can be explained by the evolution of a simulated gene pool. Game theory and evolutionary algorithms were combined in a novel way to examine how six variables influenced the decision to confront in the competition for resources. The main variable was how the genetically determined rate of confrontation evolved as a function of environmental resources and cost of a conflict. The additional modulatory effects of four other variables were also considered in the first round of simulations. Two variables were responsive to the difference between resources and cost. Two other variables were responsive to the organism's health status. Taking a systematic approach, we examined how a population of 1000 organisms were evolving in environments with different levels of reward and punishment. During each cycle, each organism was paired with another organism and thus needed to decide whether to confront or cooperate. We used a genetic algorithm to simulate the evolution of the gene pool over 500 cycles. The first series of simulations demonstrated that the baseline rate of confrontation was very responsive to environmental conditions. Our results also indicate that the decision to confront or cooperate depended not only upon the immediate competitive conditions, in which the organisms evolved, but were also responsive to their own health status. The second series of simulations used zero‐sum games to explore how risk levels varied as a function of the potential cost of engaging in a confrontation. In the second round of simulations, a simple form of memory was implemented. The results indicated that memory had a limited, but significant effect, while the cost of a conflict was highly predictive of the level of risk taken by the organisms. Our two series of simulations show that AI could contribute to answering psychological and societal questions. Our unique combination of techniques has brought to light several new insights into the mechanisms that drive the population towards cooperation and confrontation. The degree of generalizability of our results and future avenues for deepening our understanding of these evolutionary dynamics are discussed.