China's Strategic Edge in Energy and AI Applications Challenges US Dominance

China's growing strengths in energy and industrial applications could significantly challenge the United States' dominance in artificial intelligence (AI), according to a recent analysis by a leading Chinese scholar. Li Cheng, founding director of the Centre on Contemporary China and the World (CCCW) at the University of Hong Kong, has highlighted how Beijing might eventually undermine Washington’s "AI moat" within the next 10 to 20 years.

The paper, co-authored by Li and two other researchers, outlines how the US currently holds a strong position in advanced chips and AI infrastructure. However, it also points out that long-term sustainability of this advantage is under threat due to structural challenges.

A Five-Layer Framework for AI Competition

Using the "five-layer cake" framework proposed by Nvidia CEO Jensen Huang, the researchers evaluated the AI sectors of both countries across five key pillars: energy, semiconductor chips, AI infrastructure, AI models, and AI applications. They found that while the US leads in high-end chips and infrastructure, China has made significant strides in energy and AI applications.

In terms of AI models, the US still maintains some of the most sophisticated systems globally. However, the gap between the two nations is narrowing, with China making progress in developing cost-efficient, open-source models. These models are helping to drive down adoption costs and accelerate AI integration across China’s economy.

This approach contrasts sharply with the premium, closed-source models favored by major US firms. The researchers suggest that China's strategy could provide a new blueprint for winning the long-term technological competition.

Semiconductor Developments and Research Investments

Even in the semiconductor space, which has been heavily restricted by US-led export controls, the authors predict that China could significantly narrow the "generational" gap within five to eight years. This projection is supported by a more than fivefold increase in chip research and development investments over the past decade.

Despite remaining highly dependent on external suppliers for high-end graphics processing units, high-bandwidth memory, and advanced electronic design automation tools, China is focusing its efforts on achieving breakthroughs in these critical areas through a combination of fiscal subsidies, national funds, and corporate self-financing.

Energy as a Critical Advantage

One of China's most obvious advantages lies in energy, a crucial factor as both countries race to build power-hungry data centers to meet the growing demand for AI. Last year, China's total electricity consumption reached a record high of 10.4 trillion kilowatt-hours, more than double that of the US, according to the National Energy Administration (NEA).

The authors note that this lead is only widening, as China has added more energy capacity every year since 2021 than the entire US energy supply. Data centers in the US are already beginning to face challenges competing with residential users for electricity, driving up prices.

Academic Background and Research Tools

Li Cheng, a renowned scholar, established CCCW in 2023 after leaving the Washington think tank Brookings Institution, where he grew increasingly concerned about the hawkish atmosphere in the US. His latest paper, published only in Chinese in the journal Contemporary China and World, was written with the help of an AI tool called Zhenyan, developed by Beijing-based Fayu Technology for social-science research.

This article originally appeared on the South China Morning Post (www.scmp.com), the leading news media reporting on China and Asia.