基于政策建模一致性(PMC)指数模型的中美人工智能政策比较研究  

A Comparative Study of Artificial Intelligence Policies of China and the United States Based on the PMC Index Model

作　　者：朱新超 姜景[1] 陈悦 Zhu Xinchao;Jiang Jing;Chen Yue(School of Management,Anhui University,Hefei 230601,China)

机构地区：[1]安徽大学管理学院,安徽合肥230601

出　　处：《科技管理研究》2024年第8期20-30,共11页Science and Technology Management Research

基　　金：安徽省哲学社会科学规划青年项目“人脸识别技术在地方政府治理中的应用风险及其法律规制研究”(AHSKQ2022D118)。

摘　　要：对人工智能政策文本,特别是从人工智能专项政策的角度开展相关研究,能够为进一步完善中国人工智能政策体系提供决策支持。为此,以中美两国在顶层设计、职能部门两个层面共74份人工智能政策文本作为研究对象(中国51份、美国23份),通过文本挖掘方法识别出两国人工智能政策的关注重点,用政策建模一致性(PMC)指数模型建立人工智能政策评价指标体系,并结合PMC指数值和PMC曲面图,对两国各6份具有代表性的政策文本进行量化评价和对比分析。研究发现在总体上,中美两国的人工智能政策表现较好,能够较为准确地把握当前人工智能的发展前景,具有一定价值导向性和前瞻性;政策制定设计较为合理、政策协同性较高;在政策制定过程中都注重对智能化、科技创新、成果应用等方面的研究,而对于政策领域和效力级别方面则重视程度较低。同时发现两国在政策对象、政策内容和保障措施等方面的差异化显著:美国的政策扎根于技术研发、人才队伍建设、标准体系构建和基础研究投资等方面,在政策内容、保障措施等方面比中国更完善,但政策对象覆盖程度不够广;中国的政策对象兼顾宏观和微观的视角,涉及国家、产业和企业等层面,但在伦理规范、合作发展、核心技术优势等方面存在缺陷。基于此,进一步从保障措施、政策内容与政策领域3个方面提出完善中国人工智能政策的发展对策,强调中国人工智能政策的整体优化路径应为“保障措施→政策内容→政策领域”。Research on artificial intelligence(AI)policy texts,especially from the perspective of AI special policies,can provide decision support for further improving China's AI policy system.To this end,a total of 74 AI policy texts(51 in China and 23 in the United States)at the level of top-level design and functional departments in China and the United States are taken as research objects,and the focus of AI policies in the two countries is identified through the text mining method,and the evaluation index system of AI policies is established with the consistency in policy modeling(PMC)index model.Combined with PMC index value and PMC surface map,this paper makes quantitative evaluation and comparative analysis of 6 representative policy texts of the two countries.It is found that in general,the AI policies of China and the United States perform well,can grasp the current development prospects of AI more accurately,which have a certain value orientation and foresight,and the policy design is reasonable and the policy coordination is high,in the process of policy formulation,the research on intelligence,scientific and technological innovation,and application of achievements is paid more attention to,while the policy field and effectiveness level are paid less attention.At the same time,it is found that the two countries have significant differences in policy objects,policy contents,and safeguard measures.The United States'policies are rooted in technology research and development,talent team construction,standard system construction,and basic research investment,etc.,and the policy contents and safeguard measures are more complete than China's,but the coverage of policy objects is not broad enough.China's policy targets take into account both macro and micro perspectives,involving the state,industry and enterprise levels,but there are shortcomings in ethics,cooperative development,core technology advantages,and other aspects.Based on this,the paper further proposes countermeasures to improve China's artificial intellige

关 键 词：人工智能政策 政策文本分析 文本挖掘 量化评价 政策建模一致性(PMC)指数模型 中美对比 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] D63[自动化与计算机技术—控制科学与工程] G301[政治法律—政治学]