 AI Foundation Models that Re-Shape Macroeconomy: Mechanisms and MeasurementC I C C Research CICC Global Institute ()
Chapter Chapter 1 in The AI Economy, 2026, pp 1-43 from Springer Abstract: Abstract Artificial Intelligence (AI) is a system that simulates intelligence, currently referring to computer systems in particular. The idea of humans simulating intelligence has existed for a long time, but it was not until advances in electronics that the field substantially developed. The development of AI has not been a straight path: Throughout several rounds marked by circuitous progress, a persistent challenge of an absence in generalizability has emerged. Today, the success of deep learning algorithms based on transformer architecture marks AI’s entry into the era of general-purpose models. AI technology has made a milestone breakthrough in simulating intelligence across different scenarios and extrapolating patterns in the real world. Supported by high-performance computing and high-quality data, AI performance also demonstrates the scaling law, leading to continuously rising levels of intelligence. AI is characterized as a general-purpose technology (GPT) with broad applicability, innovation-spawning potential, and capacity for continuous improvement, demonstrating the potential for broader integration into the economy and enabling significant cost reductions. In this chapter, we analyze the macroeconomic impact of AI using the analytical framework of “meta-task”, which refers to the division of human work into general task elements according to different functions. (The total working time equals the sum of time spent on accomplishing all types of meta-tasks.) As technology advances, the scope of meta-tasks performed by AI progressively expands from simple to complex types with declining costs. The business sector decides when to integrate AI meta-tasks based on cost-benefit considerations. In this report, we plot the expected cost curves of AI meta-tasks on the R&D side and the application side, and we compile the wage and meta-task compositions of industries representing 92.5% of China’s GDP. This provides the basis for our estimation of the market value, industry shock, and economic impact of AI products. Compared with the baseline scenario, we estimate that AI has the potential to increase China’s GDP by Rmb12.4trn by 2035, equivalent to an increase of 9.8% and an average annual growth rate that is about 0.8 percentage points higher. In the next decade, we expect AI to introduce greater productivity gains in industries such as mining, healthcare, resource processing, information, leasing, and business services, and smaller productivity gains in industries such as wholesale and retail, accommodation and catering, and light manufacturing. Due to the impact of AI, China’s employment over the next five years might shift towards light manufacturing, accommodation and catering; however, in the next ten years and beyond, employment may shift towards finance, information services, real estate, leasing and business services, and healthcare. Date: 2026 There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it. Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-981-92-3270-3_1 Ordering information: This item can be ordered from DOI: 10.1007/978-981-92-3270-3_1 Access Statistics for this chapter More chapters in Springer Books from Springer |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.