News Center



Home News Center News

Three Warnings from the U.S. AI Dilemma

Apr 07-2026   



*This article is based on a piece by Shen Yan, a professor at the National School of Development at Peking University.

Technological iteration and capital investment are growing exponentially and the competitive landscape is becoming increasingly diversified. However, while the economic benefits of adopting technology have yet to be widely realized, its impact on productivity, the job market and energy infrastructure has already begun to manifest. This could potentially trigger profound socio-economic restructuring. Despite unabated enthusiasm for capital investment, the market has already begun to grow wary of the risks associated with realizing returns and the potential for a bubble.

Dramatic Shifts in the Technological Competitive Landscape: Core Infrastructure Becomes Strategic High Ground

The landscape of technological leadership is undergoing a dramatic reshuffle. Energy has replaced capital or talent as the most critical constraint on the expansion of AI computing power. According to the latest data analysis by the International Energy Agency (IEA) and the Electric Power Research Institute (EPRI), the surge in electricity demand driven by artificial intelligence development in the United States has become a significant challenge for the national energy system. According to IEA projections, total electricity consumption by US data centers will reach 133% of 2024 levels by 2030, posing a severe challenge to the country’s ageing power grid. China’s relative advantage in power capacity further intensifies American concerns regarding AI-related competition, standing in stark contrast to the current state of the U.S. grid. The future development of the U.S. AI industry will depend on investments in new energy sources, such as nuclear and geothermal power, and the effectiveness of policy support.

The Impact of AI on the Labour Market Is Becoming Apparent & the Risk of Structural Unemployment Is Rising

AI is now entering a substantive phase of “job displacement”. This impact exhibits two dangerous structural characteristics: Firstly, entry-level white-collar roles are the first to be threatened with unemployment. A significant reduction in positions for recent graduates and young people entering the workforce could mean that the traditional starting point for a generation’s career development is lost, which could lead to long-term misallocation of human capital. Secondly, ongoing layoffs may become the new normal. At a micro level, AI-driven productivity gains are causing companies to view layoffs as an ongoing, incremental process of workforce optimization rather than as occasional events.

From a macro perspective, although AI can create new jobs, these roles often require highly specialized AI skills and cannot offset the negative impact of numerous traditional jobs disappearing, leaving the labour market in a state of prolonged instability. If the income and job stability of ordinary workers decline, there will be insufficient demand for the goods and services produced through AI-enabled means, which in turn will undermine economic growth.

Insights from U.S. AI Development Trends & Their Economic Impact

By 2026, the development of AI in the United States will have reached a critical juncture. Although the technological race and infrastructure development will continue to advance rapidly, the ability of the U.S. socio-economic system to adapt will be severely tested. The main challenges lie in three areas: first, energy and digital infrastructure; second, workforce transition and social safety nets; and third, preventing systemic risks.

Even if the U.S. elevates its focus on these challenges to a national strategic level, significant challenges will remain for its future development. Additionally, the U.S. faces internal institutional friction stemming from intense domestic political and social divisions. The practice of significantly weakening AI ethics, safety governance and environmental standards in pursuit of development has drawn widespread criticism and could put the United States at a disadvantage when it comes to formulating international governance standards.

In summary, while the United States has demonstrated strong strategic resolve and the ability to mobilize resources, deep-seated contradictions such as ageing infrastructure, political and social polarization, and economic structural imbalances constitute fundamental obstacles to its success in this long-term race.

The above analysis of trends in US AI development suggests that technological evolution may outpace the ability of social institutions to adapt. This offers three key lessons for China’s AI development. Firstly, we must guard against capital bubbles and technological myths by strengthening assessments of the commercial viability of AI investment and its integration with the real economy. We must emphasize that AI must empower all sectors of society. Secondly, we must confront the impact on employment and the need for skills transformation. This requires the proactive establishment of vocational training programs focused on human-machine collaboration, as well as social policies that accommodate flexible employment. Thirdly, AI development must be “human-centred” and cannot come at the cost of mass unemployment. Computing power, green energy (especially nuclear energy) and grid upgrades must be integrated into a unified national strategic infrastructure plan to ensure the sustainable momentum of the technological revolution while keeping social costs under control.


Three Warnings from the U.S. AI Dilemma

Apr 07-2026   



*This article is based on a piece by Shen Yan, a professor at the National School of Development at Peking University.

Technological iteration and capital investment are growing exponentially and the competitive landscape is becoming increasingly diversified. However, while the economic benefits of adopting technology have yet to be widely realized, its impact on productivity, the job market and energy infrastructure has already begun to manifest. This could potentially trigger profound socio-economic restructuring. Despite unabated enthusiasm for capital investment, the market has already begun to grow wary of the risks associated with realizing returns and the potential for a bubble.

Dramatic Shifts in the Technological Competitive Landscape: Core Infrastructure Becomes Strategic High Ground

The landscape of technological leadership is undergoing a dramatic reshuffle. Energy has replaced capital or talent as the most critical constraint on the expansion of AI computing power. According to the latest data analysis by the International Energy Agency (IEA) and the Electric Power Research Institute (EPRI), the surge in electricity demand driven by artificial intelligence development in the United States has become a significant challenge for the national energy system. According to IEA projections, total electricity consumption by US data centers will reach 133% of 2024 levels by 2030, posing a severe challenge to the country’s ageing power grid. China’s relative advantage in power capacity further intensifies American concerns regarding AI-related competition, standing in stark contrast to the current state of the U.S. grid. The future development of the U.S. AI industry will depend on investments in new energy sources, such as nuclear and geothermal power, and the effectiveness of policy support.

The Impact of AI on the Labour Market Is Becoming Apparent & the Risk of Structural Unemployment Is Rising

AI is now entering a substantive phase of “job displacement”. This impact exhibits two dangerous structural characteristics: Firstly, entry-level white-collar roles are the first to be threatened with unemployment. A significant reduction in positions for recent graduates and young people entering the workforce could mean that the traditional starting point for a generation’s career development is lost, which could lead to long-term misallocation of human capital. Secondly, ongoing layoffs may become the new normal. At a micro level, AI-driven productivity gains are causing companies to view layoffs as an ongoing, incremental process of workforce optimization rather than as occasional events.

From a macro perspective, although AI can create new jobs, these roles often require highly specialized AI skills and cannot offset the negative impact of numerous traditional jobs disappearing, leaving the labour market in a state of prolonged instability. If the income and job stability of ordinary workers decline, there will be insufficient demand for the goods and services produced through AI-enabled means, which in turn will undermine economic growth.

Insights from U.S. AI Development Trends & Their Economic Impact

By 2026, the development of AI in the United States will have reached a critical juncture. Although the technological race and infrastructure development will continue to advance rapidly, the ability of the U.S. socio-economic system to adapt will be severely tested. The main challenges lie in three areas: first, energy and digital infrastructure; second, workforce transition and social safety nets; and third, preventing systemic risks.

Even if the U.S. elevates its focus on these challenges to a national strategic level, significant challenges will remain for its future development. Additionally, the U.S. faces internal institutional friction stemming from intense domestic political and social divisions. The practice of significantly weakening AI ethics, safety governance and environmental standards in pursuit of development has drawn widespread criticism and could put the United States at a disadvantage when it comes to formulating international governance standards.

In summary, while the United States has demonstrated strong strategic resolve and the ability to mobilize resources, deep-seated contradictions such as ageing infrastructure, political and social polarization, and economic structural imbalances constitute fundamental obstacles to its success in this long-term race.

The above analysis of trends in US AI development suggests that technological evolution may outpace the ability of social institutions to adapt. This offers three key lessons for China’s AI development. Firstly, we must guard against capital bubbles and technological myths by strengthening assessments of the commercial viability of AI investment and its integration with the real economy. We must emphasize that AI must empower all sectors of society. Secondly, we must confront the impact on employment and the need for skills transformation. This requires the proactive establishment of vocational training programs focused on human-machine collaboration, as well as social policies that accommodate flexible employment. Thirdly, AI development must be “human-centred” and cannot come at the cost of mass unemployment. Computing power, green energy (especially nuclear energy) and grid upgrades must be integrated into a unified national strategic infrastructure plan to ensure the sustainable momentum of the technological revolution while keeping social costs under control.