The central subject of this paper is the relation between the macro-level of the Society and the micro-level of its individuals. In complex real-life systems, this relationship is not trivial and even the result at the macro-level can be counter-intuitive. One of the first models that highlights this is the Sakoda and Schelling’s model of spatial segregation where a person uses a selfish rule to decide whether or not to leave his current place; on the basis of the ratio of “compatible” agents in the neighborhood of each individual, simulations have shown that tolerant agents group together beyond what their own tolerance imposes. This paper compares the macro-behavior arising from either selfish or altruist micro-motives; for this purpose, we propose agent-based computational models based, respectively, on selfishness, altruism and a mixture of both. Computational simulations show that (i) in a totally altruistic population, a vast majority becomes satisfied as if the agents were themselves selfish and end up spatially segregated; (ii) altruist micro-motives allow to reduce both the micro-macro gap and the segregation phenomenon; (iii) in a mixed population, the macro-effect depends both on the respective weight of each micro-motives and the allocation strategy.

中文翻译：

---

从自私利他到宏观行为

本文的中心主题是两者之间的关系宏观层面协会和微观层面其个人。在复杂的现实生活系统中，这种关系并非微不足道，甚至宏观层面的结果也可能违反直觉。最早强调这一点的模型之一是 Sakoda 和 Schelling 的空间隔离模型，其中一个人使用自私的规则来决定是否离开他现在的位置。根据每个个体附近“兼容”智能体的比例，模拟表明，宽容的智能体组合在一起超出了它们自身的容忍度。本文比较了两者产生的宏观行为自私要么利他主义者微观动机；为此，我们提出了基于代理的计算模型，分别基于自私、利他和两者的混合。计算模拟表明（i）在一个完全利他的人群中，绝大多数人感到满意仿佛代理人本身是自私的，最终在空间上被隔离；(ii) 利他主义的微观动机可以减少微观宏观差距和隔离现象；(iii) 在混合人群中，宏观效应取决于每个微观动机各自的权重和分配策略。