US-China AI war boils down to a contest over electricity

The global race for artificial intelligence dominance between the United States and China is shifting focus from model innovation to a fundamental contest over electricity. As AI models become increasingly similar in performance and proliferate rapidly, they are transforming into commodities. This commoditization means the value and pricing of AI services will soon be dictated by their primary cost driver: electricity.

Recent advancements, particularly the emergence of powerful open-source AI models in China from companies like DeepSeek and Zhipu AI, have challenged the notion that US closed-source models hold a significant performance advantage. This rapid development suggests that creating advanced AI is becoming less difficult, diminishing the unique value of owning a frontier model. Consequently, the market is moving towards AI services being priced based on the cost of production, with electricity being the dominant factor.

The competition is now framed as a race between China's renewable energy-focused strategy and America's continued reliance on hydrocarbons. China's commitment to self-reliance and carbon neutrality has propelled investments in solar, wind, and hydropower. The falling costs of solar power, now around two US cents per kWh, make renewables increasingly competitive. Locating data centers near solar farms in western China is becoming a rational economic choice to mitigate transmission costs, potentially giving China an edge in providing cheaper and more reliable electricity for its burgeoning AI industry.