Summary of Key Points
This article challenges the common perception that "major infrastructure projects" merely involve building roads, bridges, and developing the real estate industry. Comparing the "old 4 trillion yuan" stimulus package in 2009, which focused on railway and infrastructure projects, the current investments in information and communication technologies, computing power networks, and new types of power grids amount to several more "4 trillion yuan" packages. The total scale of these investments over the next five years is expected to reach 25 trillion yuan, all aimed at supporting the underlying infrastructure for the AI era. This new wave of infrastructure development not only fundamentally changes the logic of how regional economic benefits are distributed but also transforms it into a global competition among major economies such as China, the United States, Japan, and South Korea to bet on the next industrial revolution. Ultimately, it will reshuffle the development opportunities for all cities and even ordinary individuals.
---
Detailed Analysis
1. The "new 4 trillion yuan" is not about real estate; it's about laying the foundation for AI in China
Many people mistakenly think that "major infrastructure" means constructing roads and buildings. However, the 2009 stimulus package was designed to address the shortcomings of the industrial era: at that time, China had less than 10,000 kilometers of high-speed railways, and many counties lacked even access to highways. The investment in roads and bridges helped China's transportation network become one of the best in the world, thereby boosting the entire steel, cement, and real estate industries. The current "4 trillion yuan" packages are being invested in areas that are invisible to the general public:
- 3.8 trillion yuan is going into communication networks to ensure seamless data transmission for AI;
- 4 trillion yuan is being invested in computing power networks to connect AI supercomputers nationwide like a public power grid;
- 5 trillion yuan is being invested in new power grids to supply energy to high-energy-consuming computing centers, preventing power outages when large AI models are running.
The total investment of 25 trillion yuan over the next five years aims to provide the basic resources needed for AI operations as easily accessible public services—just like water and electricity. In the future, starting a small company that uses AI will not require buying servers or installing fiber optic connections; you can simply plug in a "computing power plug."
2. The benefits of infrastructure are no longer evenly distributed; only certain types of cities will benefit significantly
In the era of traditional infrastructure, the benefits were more widespread. For example, the construction of highways and high-speed railways in remote areas boosted local logistics, tourism, and housing prices. Ten years ago, many western provinces saw their GDP growth rates soar thanks to infrastructure investment, with Guizhou even achieving highway access in every county. However, in the new infrastructure era, the benefits are concentrated in two types of cities:
- The largest and most advanced cities in the east, which already host 90% of China's AI companies and internet giants, have a high demand for computing power. By locating computing centers near their customers, they minimize data transmission delays and maximize profits from the entire AI industry chain (Shenzhen, Shanghai, Beijing).
- Small cities in the west with cheap renewable energy (such as wind and solar power) can significantly reduce costs. For example, the total cost of building a computing center in Inner Mongolia, Guizhou, or Gansu is half that of the east. These cities can convert their renewable energy into digital products for sale to the east, transforming from being suppliers of raw materials to providers of advanced services.
Cities that lack both the AI industry demand of the east and the renewable energy advantages of the west will not benefit much. For instance, the eastern region currently accounts for only 0.9% of the total computing power capacity in China.
3. This is not just China acting alone; the world is betting on AI infrastructure
Many believe that this new infrastructure initiative is a Chinese economic stimulus measure. In reality, major economies around the world are investing heavily in AI infrastructure, as it represents a significant gamble on their future economic success. For example, the U.S. saw its GDP growth exceed expectations in the third quarter of this year, largely due to AI investments. Tech giants like Amazon, Microsoft, Meta, and Google are spending 5 trillion yuan on AI infrastructure. Experts estimate that AI investments will contribute 25% of the U.S. GDP growth by 2026. South Korea has seen even more impressive growth, with its GDP growing by 47% in the second quarter, thanks to investments in AI data centers and semiconductors. This competition is on par with the industrial and internet revolutions of the past; countries that cannot develop the necessary infrastructure (computing power, chips, optical modules) will be left behind and may become suppliers for developed nations.
4. Infrastructure is not just about short-term GDP growth; it will transform people's lives
Some criticize infrastructure as a waste of money on redundant construction. However, the 4G network construction in China ten years ago was met with skepticism, but it led to the emergence of new industries like TikTok, food delivery, and mobile payments, creating trillions of new market opportunities. Similarly, the new AI infrastructure will significantly reduce the cost of using AI. For example, intelligent driving will no longer require expensive chips in vehicles; instead, it will rely on public computing networks. Thousands of drones in cities will be coordinated by a widespread communication network, and household robots will use cloud-based computing. Small businesses can use AI to create marketing strategies at low costs. New industries and entrepreneurial opportunities will emerge in the coming years, just as they did during the mobile internet era.
5. The race for urban development is in its final stages; those who choose the wrong path will fall behind
In the era of traditional infrastructure, cities could catch up by investing in transportation and airports. However, the new AI infrastructure era offers limited time for such catch-up. Cities that have invested in AI-related infrastructure are leading in GDP growth. In the first eight months of this year, China's integrated circuit exports increased by 103.9%, driving a 17.6% rise in foreign trade. Cities that focus on traditional real estate and transportation will see their local youth and businesses migrate to more profitable areas, widening the economic gap. The next five years will see a reshuffle of urban economic rankings, and those that lag behind will struggle to regain their position in the next decade.