Journal of Statistics and Information

Artificial intelligence is an important national strategy.Computing power infrastructure serves as the crucial “computing foundation” for the development of artificial intelligence technology,providing massive computing resources to support modules such as large-scale parallel computing of algorithms,algorithm model training,and model operation services,thereby promoting the rapid development of artificial intelligence technology.However,upon reviewing the existing literature,it is found that there are several challenges in the development of artificial intelligence technology,such as pressure on capital investment,lack of human capital,and insufficient technical foundation.Therefore,whether the construction of computing power infrastructure will solve these problems and thereby promote the development of artificial intelligence technology in enterprises is worthy of in-depth study.Based on this,this study takes the listed companies in Shanghai and Shenzhen stock markets from 2010 to 2023 as the research sample,and employs the least squares method(OLS)to conduct an empirical test to examine whether the construction of computing power infrastructure can promote the development of artificial intelligence technology in enterprises and the mechanism of its effect.The result shows that the construction of computing power infrastructure significantly promotes the development of enterprise artificial intelligence technology.The mechanism analysis shows that the construction of computing power infrastructure plays a promoting role by easing the pressure of capital investment,improving the level of human capital and enhancing the capability of digital technology.The heterogeneity analysis shows that the promotion effect is more significant in small and medium-sized enterprises and non-technical board companies.Industry heterogeneity analysis shows that the promotion effect is more significant in manufacturing,technology-intensive and capital-intensive industries.Regional heterogeneity analysis shows that the promotion effect is more significant in the areas with low financial development,low educational resources and high computing power application.Based on the research results,suggestions are proposed from both the government perspective and the enterprise perspective.The policy suggestions from the government perspective include enhancing the regional computing power supply level and the level of computing power application inclusiveness,implementing fiscal policies such as government subsidies and tax incentives,planning regional higher education reforms,increasing the intensity of talent recruitment,supporting the aggregation of digital industries,etc.The development suggestions from the enterprise perspective include actively connecting and integrating computing resources,optimizing production and operation,accumulating relevant technical talents,strengthening digital technology capabilities,etc.