机器人伦理设计进路及其评价

发布时间：2018-05-10 10:28

  本文选题：机器人 + 伦理设计　； 参考：《哲学动态》2017年09期

【摘要】：在逐渐向我们走来的"机器人时代"里,如何让机器人更好地为人类服务,是机器人伦理研究必须回答的一个重要问题。虽然有的学者反对让机器人成为人工道德行为体,但考虑到机器人的功能日益强大,与人类的关系日益密切,所以让机器人拥有一定的道德判断与行为能力是必然之需,这正是机器人伦理设计所要解决的问题。机器人伦理设计的理论进路有"自上而下"进路、"自下而上"进路、混合进路和关系论进路。关于机器人伦理设计的实践探索,可以分为计算机科学进路、认知科学进路和价值敏感设计等三个方面。道德图灵测试可以作为评价机器人伦理设计的一种方式,但证实方法对于检验机器人伦理设计的效果应该更为合理。
[Abstract]:In the "robot age" that comes to us gradually, how to make the robot serve human better is an important question that must be answered in the research of robot ethics. Although some scholars are opposed to making robots become artificial moral actors, given the increasingly powerful functions of robots and their increasingly close relationship with human beings, it is necessary for robots to have a certain degree of moral judgment and ability to act. This is exactly the problem that robot ethics design needs to solve. The theoretical approach of robot ethical design includes "top-down" approach, "bottom-up" approach, mixed approach and relational approach. The practical exploration of robot ethical design can be divided into three aspects: computer science approach, cognitive science approach and value sensitive design. The moral Turing test can be used as a way to evaluate the ethical design of the robot, but it should be more reasonable to verify the effect of the method on the ethical design of the robot.
【作者单位】： 上海交通大学科学史与科学文化研究院;
【基金】：国家社会科学基金一般项目“机器人伦理问题研究”(15BZX036)的阶段性成果
【分类号】：B82-057
，

本文编号：1868944