Summary of Key Points
Harvard Professor Graham Allison believes that the United States is betting its entire economic, capital, and national security stake on artificial intelligence (AI), particularly general artificial intelligence (AGI), considering it the key to winning the Sino-US technology race. However, this reckless gamble has created a bubble. If the AI productivity miracle does not materialize, it could lead to an economic collapse, defense disarray, and a loss of international confidence. In contrast, China is pursuing a path of low cost, open-source development, and large-scale application, achieving performance close to that of the United States with less investment while rapidly gaining global popularity.
I. How Crazy is America's Bet on AI? They've Staked Half of Their Country
The American economy has almost become the “gambling table” for AI:
- Exaggerated Investment Contribution: AI-related investments have contributed significantly to GDP growth, with AI data center spending accounting for half of all business investment (for every $2 in business investment, $1 is allocated to AI).
- Stock Market Controlled by AI: The market value of AI companies now accounts for over 40% of the total U.S. stock market value; if OpenAI and Anthropic go public, this proportion could exceed 50%. In 2026, 85% of the S&P 500 index’s growth came from AI companies, and ChatGPT added $23.1 trillion in market value to the S&P 500.
- Government-Corporate Alliance: A “national security-AI industry complex” has formed, with the government being a major customer for AI giants (such as the Department of Defense and CIA). There are even alliances pushing for the government to become shareholders of AI companies; AI companies also lobby for policies to suppress domestic and international competitors.
II. Why Is America So Ready to Gamble? Three Forces Are Driving This
This gamble is not accidental; three factors are at play:
1. Companies Seeking Trillions in Wealth: AI companies are targeting the largest “gold mine” in history (谈论的金额 are in the trillions of dollars). CEOs want not only to make money but also to gain influence—e.g., OpenAI’s CEO sits at the same table as G7 leaders. To win, they will say and do anything.
2. Government Seeking to Solve Fiscal Crises: The U.S. annual deficit is close to 6% of GDP, and debt interest exceeds defense spending. AI is seen as a “lifesaver” for the government; if it can bring about super-productivity growth, it could prevent economic disasters.
3. National Security Concerns: The U.S. strategic community believes that “the winner of AI will dominate the world,” with AGI (intelligent beyond all Nobel laureates) providing an irreversible advantage. Former Secretary of State Condoleezza Rice explicitly stated, “This is a race we must win.”
III. What Are the Minefields of This Gamble? Bubble, Security Issues, and Public Opposition
Risks are already emerging:
- Bubble Risk: The valuations and profits of AI companies do not match; they have reached extreme levels since 1929. 45% of fund managers see an “AI bubble burst” as the greatest risk (with prices up 17% from last month). Bridgewater Fund founder Ray Dalio called it an “immature approach,” and Michael Burry warned, “We are entering a high-risk area with unavoidable consequences.”
- AI Security Disasters: Kissinger warned of an AI-related “Chernobyl” or “9/11” scenario—e.g., AI helping villains create biological weapons that could kill hundreds of thousands or paralyze power grids. AI giants also admit a 25% chance of human catastrophe.
- Market Correction Signals: Companies like Meta and SpaceX have built too much computing power and are starting to rent out idle capacity. Bain analysts say tech companies need an additional $2 trillion in AI revenue to break even, but they are currently not making enough profits.
- Public Opposition: New York State has banned the construction of AI data centers for a year; 70% of Americans oppose data center construction in their communities. 77% are concerned about increased electricity costs, and 75% fear job loss due to AI, questioning its reliability.
IV. China Has Taken a Different Path: Low Cost, Open Source, and Using Intelligence to Outperform GPUs
China has not followed America’s AGI bet but adopted a more pragmatic approach:
- Less Money, More Results: Baidu, Alibaba, and Tencent’s AI investments are only one-tenth of those of American giants, yet their model performance is comparable (e.g., Kimi K3 from Moonlight and Anthropic’s top models). DeepSeek used 200 engineers with annual salaries of $150,000 to create a model on par with American ones and made it available for free globally.
- Open Source Like Android: Chinese AI companies share their models (similar to Google’s Android), which is used in 70% of smartphones worldwide. Currently, nearly two-thirds of global AI processing is done using Chinese models, including the five most popular ones. Airbnb uses Chinese models at one-tenth the cost of American alternatives.
- Focusing on Applications Rather than Computing Power: China’s “AI+” initiative aims to integrate AI into society (e.g., improving productivity and governance) rather than building massive data centers. NVIDIA’s CEO said that half of the world’s top AI talent is in China, indicating a focus on “brainpower” rather than just computing power.
V. What Are the Consequences of Losing This Gamble?
If the AI miracle does not happen, America could face:
- Economic Disaster: Worse than the 2008 housing bubble (GDP drop of 4.3%) or the 2000 internet bubble ($5 trillion in wealth evaporated), potentially leading to a situation similar to the Great Depression of 1929 (GDP drop of 30%).
- National Security Chaos: Wealth shrinkage, decreased government revenue, and increased social security needs, with defense investment plans disrupted. The international community will question America’s strength—“losing a race they claimed to be decisive.”
- Global Order Disruption: The 1929 crash led to the Great Depression and World War II; a similar AI-related collapse could cause a global recession. If the AI bubble bursts, China may become more determined in its path, significantly weakening America’s global leadership.
The essence of this gamble is that the U.S. is betting everything on AGI to win the race, while China is quietly taking market share with a combination of low cost, open-source technology, and practical applications. If America loses, it could not only experience an economic collapse but also lose its status as the global technology leader. China’s approach may represent a more stable and sustainable path forward.