PDF(Pubmed)
Abstract：
Direct reciprocity promotes the evolution of cooperation when players are sufficiently equal, such that they have similar influence on each other. In the light of ubiquitous inequality, this raises the question of how reciprocity evolves among unequal players. Existing studies on inequality mainly focus on payoff-driven learning rules, which rely on the knowledge of others\' strategies. However, inferring one\'s strategy is a difficult task even if the whole interaction history is known. Here, we consider aspiration-driven learning rules, where players seek strategies that satisfy their aspirations based on their own information. Under aspiration-driven learning rules, we explore the evolutionary dynamics among players with inequality in endowments and productivity. We model the interactions among unequal players with asymmetric games and characterize the condition where cooperation is feasible. Remarkably, we find that aspiration-driven learning rules lead to a higher level of cooperation than payoff-driven ones over a wide range of inequality. Moreover, our results show that high aspiration levels are conducive to the evolution of cooperation when more productive players are equipped with higher endowments. Our work highlights the advantages of aspiration-driven learning for promoting cooperation among unequal players and suggests that aspiration-based decision-making may be more beneficial for the collective.
摘要：
当参与者充分平等时，直接互惠促进合作的演变，这样他们对彼此有相似的影响。鉴于普遍存在的不平等，这提出了一个问题，即不平等的参与者之间的互惠如何演变。现有的关于不平等的研究主要集中在报酬驱动的学习规则上，这依赖于别人的知识\'策略。然而,即使知道整个互动历史，推断一个人的策略也是一项艰巨的任务。这里,我们考虑志向驱动的学习规则，玩家根据自己的信息寻求满足自己愿望的策略。在愿望驱动的学习规则下，我们探索了禀赋和生产力不平等的参与者之间的进化动态。我们通过不对称游戏对不平等玩家之间的互动进行建模，并描述合作可行的条件。值得注意的是,我们发现，在广泛的不平等问题上，愿望驱动的学习规则比回报驱动的学习规则带来更高的合作水平。此外，我们的研究结果表明，当生产力更高的参与者拥有更高的禀赋时，高期望水平有利于合作的发展。我们的工作突出了志向驱动学习在促进不平等行为者之间合作方面的优势，并表明基于志向的决策可能对集体更有利。
