如何给机器人设计道德算法  

Design of Ethical Algorithm for Robots

作　　者：赵聪帅 王健 Zhao Congshuai;Wang Jian(Department of Philosophy,Xi'an Jiaotong University,Xi'an 710049,China)

机构地区：[1]西安交通大学哲学系,陕西西安710049

出　　处：《洛阳师范学院学报》2023年第7期15-20,共6页Journal of Luoyang Normal University

摘　　要：机器人伦理学旨在描述一种设计和评价机器人道德算法的方法,该方法可以应用在自动驾驶汽车与军事救援机器人的算法中。依据德里克·勒本(Derek Leben)的道德判断产生自利有机体中合作行为的进化压力的观点,将罗尔斯的契约主义与功利主义、道义论、德性伦理学等道德理论在各种困境游戏中的应用结果进行比较,认为只有契约主义能在各种困境中产生帕累托最优与合作行为,基于契约主义差别原则的Leximin程序适合应用于各种各样的决策空间。Robotic ethics aims to describe the method for designing and evaluating robot ethical algorithm,which can be applied to the algorithm of autonomous vehicles and military rescue robots.According to Derek Leben's view that ethical judgment generates evolutionary pressure of cooperative behavior in self-interested organisms,this paper compares the application results of Rawls' contractualism with that of moral theories such as utilitarianism,deontology and virtue ethics in various dilemma games.The results show that only contractualism can produce Pareto optimality and cooperative behavior in various dilemmas,and the Leximin program based on the principle of difference of contractualism is suitable for various decision-making spaces.

关 键 词：合作 契约主义 道德算法 Leximin 

分 类 号：B82[哲学宗教—伦理学]