人工智能与制造业服务化

苗翠芬

（中国社会科学院亚太与全球战略研究院）

　　内容提要：本文通过匹配国际工业机器人联合会（IFR）与2016年世界投入产出表（WIOD）相关数据，利用2000—2014年41个经济体14个制造业行业的数据，考察人工智能对制造业服务化水平的影响及其内在机制。研究结果表明，人工智能显著提升了制造业服务化水平。替换核心指标及考虑内生性问题后，回归结果依然稳健。机制检验结果显示，人工智能主要通过优化人力资本结构、提高数据要素密集度、改变生产和交付模式等途径影响制造业服务化水平。异质性分析结果表明，人工智能显著促进了制造业在岸服务化以及生产性服务化水平。此外，人工智能对不同类型的制造业均具有显著的促进效应，特别是汽车运输、电气机械、计算机、电子以及光学产品等技术密集型制造业。本文的研究可以为进一步推动制造业和服务业深度融合发展、促进制造业转型升级以及实现制造业高质量发展提供经验借鉴和政策参考。

　　

　　关键词：人工智能；制造业服务化；人力资本；数据要素；模块化生产

　　

　　作者简介：苗翠芬，中国社会科学院亚太与全球战略研究院助理研究员，北京，100007。

　　

　　基金项目：中国社会科学院青年科研启动项目“国际数字协定对数字服务贸易发展的影响及对策研究”（2023YQNQD061）

　　

　　引用格式：苗翠芬.人工智能与制造业服务化[J].经济与管理研究,2023,44(7):22-39.DOI:10.13502/j.cnki.issn1000-7636.2023.07.002.

　　

Artificial Intelligence and Manufacturing Servitization

MIAO Cuifen

(Chinese Academy of Social Sciences, Beijing 100007)

　　

　　Abstract: Manufacturing servitization is a new industrial form that integrates manufacturing and services. Based on the importance of manufacturing servitization and the rapid development of new-generation information technology such as artificial intelligence (AI), this paper analyzes the impact of AI on manufacturing servitization and its internal mechanism, using the panel data of 14 manufacturing industries of 41 economies from 2000 to 2014 by matching the International Federation of Robotics (IFR) database with the World Input-Output Database (WIOD).

　　The results show that AI contributes significantly to manufacturing servitization. Specifically, for every 1% increase in the level of AI, manufacturing servitization increases by an average of 0.256%. The result remains valid after robustness tests, including replacing the core variables, and introducing instrumental variables. The mechanism analysis shows that human capital structure optimization, factor intensity improvement, and delivery mode transformation significantly mediate AI, which increases the servitization level of manufacturing. Heterogeneity tests reveal that the driving impact is more potent in the onshore servitization level of manufacturing and producer services. AI significantly promotes the servitization level of three types of manufacturing, especially technology-intensive manufacturing, such as automotive transportation, electrical machinery, computers, electronics, and optical products. In addition, this paper selects 25%, 50%, 75%, and 95% quantiles for empirical regression, and the findings reveal that the positive effect is significant in the four different quantiles. However, as the servitization quantile increases, the regression coefficient gradually decreases.

　　The contributions of this paper are reflected in the following three aspects. First, it discusses the factors that affect manufacturing servitization from the perspective of the application of AI. It complements the quantitative analysis of manufacturing servitization and develops the research on the economic impact of AI. Second, it identifies the channel mechanism through which AI facilitates manufacturing servitization, extending the theoretical analysis of AI on the integration of manufacturing and services. Third, it analyzes the heterogeneous impact of AI on manufacturing servitization from multiple perspectives, such as the source countries of service inputs, the characteristics of services, and the differences in the type of manufacturing, which provides theoretical guidance for the high-quality development of manufacturing and the formulation of relevant policies by the government. This paper provides policy implications for promoting the deep integration of manufacturing and services, achieving structural transformation and upgrading, and promoting the high-quality development of manufacturing.

　　Keywords: artificial intelligence; manufacturing servitization; human capital; data element; modular production