人工智能治理的三个基本问题:技术逻辑、风险挑战与公共政策选择  

Three Fundamental Issues of Artificial Intelligence Governance: Principle, Challenge and Public Policy

作　　者：贾开 蒋余浩[2] Jia Kai Jiang Yuhao(Lecturer at School of Politics and Administrative Management, University of Electronic Science and Technology of China, Chengdu 611731 Post-doctor at School of Public Policy ＆ Management, Tsinghua University, Beijing 100084)

机构地区：[1]电子科技大学政治与公共管理学院 [2]清华大学公共管理学院

出　　处：《中国行政管理》2017年第10期40-45,共6页Chinese Public Administration

基　　金：首都师范大学文化研究院重大研究项目"‘互联网+城市治理’:智慧城市2.0的理论"(编号:ICS-2016-A-03)

摘　　要：人工智能的兴起不仅是技术或产业领域的重大发展,更是经济、社会、政治各领域的基础性和综合性变革,由此产生的治理挑战要求公共政策框架的重构与创新,而这又体现为三个基本问题。首先在技术层面,人工智能的核心是算法与数据,其内在逻辑是基于机器学习形成自主决策。其次,传统科层治理结构、方法以及范围不能应对新的变化,从而构成了人工智能时代的治理挑战。各国现有政策所遵循的"无需批准式监管"或"审慎监管"逻辑固然有其合理性,但却局限于"创新"与"安全"的二维选择而失于片面。最后,本文提出综合性公共政策框架,力图为人工智能时代的崛起建立制度基石。The rise of Artificial Intelligence （ AI ） is not limited in the field of technology or industry, but rather a fundamental and comprehensive revolution of economy, society and politics, causing three critical issues. From the technical perspective, the core of AI is the code and data, utilizing the algorithm of machine learning to make decisions. The traditional governance structure, method and scope fail to respond to the governance challenges proposed by AI. It is reasonable to follow the logic regulation. However, it fails to recognize the comprehensiveness of permissionless regulation or precautionary and fundamentality of AI. This article proposes a new path focusing on three mechanisms, including the governance of code and data, the reform of social governance and the global governance.

关 键 词：人工智能 治理 公共政策 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]