论人工智能的自反性伦理治理  

On the Reflexively Ethical Governance for AI

作　　者：杜严勇[1] 

机构地区：[1]上海交通大学科学史与科学文化研究院,上海200240

出　　处：《新疆师范大学学报（哲学社会科学版）》2018年第2期111-119,共9页Journal of Xinjiang Normal University(Philosophy and Social Sciences)

基　　金：国家社科基金一般项目"机器人伦理问题研究"(15BZX036)的阶段性成果

摘　　要：在人工智能、机器人科技飞速发展的时代背景中,鉴于其可能产生的深刻社会影响,近年来世界各国纷纷发布治理规划与报告,突显了人工智能伦理及其治理问题的重要性与紧迫性。文本在澄清自反性伦理治理概念的基础上,分析了新制度主义经济学、节点治理、协同治理与实验主义治理等四种代表性治理理论,认为各种治理理论都有不同程度的自反性。要实现自反性治理,需要克服有限理性、打开认知闭合,培养行动者治理能力,重视治理过程中的开放性与多样性。The rapid development of artificial intelligence and robotics will result in significant social influence and ethical issues. Many governments release projects and reports of governance for the robotics and artificial intelligence, which highlight its importance and urgency. This article clarifies the concept of reflexively ethical governance, then elaborates different theories of governance and their reflexivity. For the success of governance, we need overcome bounded rationality and open cognitive closure while pay attention to capacitation of actor and openness or diversity in the process of governance. This article also develops a framework of reflexively ethical governance for artificial intelligence.

关 键 词：人工智能 机器人 自反性 伦理治理 

分 类 号：N31[自然科学总论]