【By Observer News, Wang Kaiwen】"Will the U.S. Win AI, but Lose the War?" The Financial Times published an article with this title on December 13, pouring cold water on the current AI boom in the United States.
The article's author, Tim Wu, who served as a special assistant to the White House on technology and competition policy during the Biden administration, warns against the U.S. fully betting on AI, pointing out that the idea of a "doomsday battle" with China over this key future technology is both an illusion and a lobbying tactic from Silicon Valley.
In contrast, although China is also vigorously promoting AI development, its attitude is more restrained and pragmatic, and it pays more attention to diversified layout.
The article points out that in the past year, major U.S. tech companies have invested more than $350 billion in AI-related infrastructure, and by 2026, this figure is expected to exceed $400 billion. This number far exceeds that of any other country, including China, which has invested nearly $100 billion in AI so far.
For many Westerners, having sufficiently bold companies and deep capital markets to dominate a "money-spend race" may feel reassuring. If AI is indeed the "One Ring to Rule Them All" as predicted, it seems that the West has already grasped the future in its hands.
However, this is only an optimistic narrative. There is another possibility: the obsession of Silicon Valley with AI means that the U.S. might win the AI race but lose the broader struggle for economic dominance. The reason is that the U.S. has put all its eggs in one basket with AI, while China has spread its bets across several possible futures. The final outcome depends on whether betting on AI is truly the right choice.
The article argues that although AI is very popular in China, the actual investment in AI is not as thorough as it appears to be. China often describes AI as a "national priority" and invests in it, but the government and major enterprises invest far more in other areas, such as electric vehicles, batteries, robotics, solar panels, wind turbines, and other forms of advanced manufacturing, to ensure dominance in these fields. These sectors may not be as glamorous as AI, but their returns are much less speculative.
"It is the U.S. that is truly obsessed with AI," the author wrote, and pointed out that the U.S. investment motives are both commercial and somewhat mystical, especially reflected in the pursuit of general artificial intelligence (AGI) and the "singularity." People strongly believe that technology will continue to progress exponentially, yet this situation is extremely rare in the history of technology. The deeper one looks, the more one realizes that both supporters of AI and doomsayers have become increasingly detached from reality.
At the same time, the U.S. tech industry is highly concentrated, and its near-monopoly structure further amplifies the risks: when huge amounts of money are in the hands of a few companies, the likelihood of collective blind following also increases.
On September 23, 2025, OpenAI, Oracle, and SoftBank jointly announced the construction of five new AI data centers in the U.S., with a projected investment of over $40 billion over the next three years. IC Photo
If the returns from AI are uncertain, then a more cautious strategy should be adopted, such as diversification. However, American venture capital and tech platforms are equally obsessed with AI, and the financial markets have rewarded their current practices.
The U.S. government could have hedged its risks, but during the Trump administration, it cut support for clean energy investments, making the national technology strategy seem like a big bet on a single horse.
The article argues that the current U.S. spending on AI is a gamble on a particular vision of the future, which may prove to be inspired and visionary, but could also trigger a destabilizing stock market crash, leaving the West behind.
The author points out that looking further ahead, the U.S. technology strategy resembles a simple syllogism: AI is the most important technology of the 21st century; it is extremely expensive; therefore, whoever spends the most will become the leading civilization of the future.
But any syllogism has its premises. The key question is whether today's artificial intelligence is truly the most important path to prosperity and a better future.
The article argues that for a long time, the business logic of artificial intelligence has been somewhat unclear, and many of the goals that Silicon Valley has painted for AI are clearly fantastical. They believe that AI can solve most of humanity's problems.
Within Silicon Valley, the end goal of this "race" is to achieve AGI — machines with broad cognitive abilities similar to humans. Believers think that once AGI emerges, it will trigger further breakthroughs, known as the "law of accelerating returns." Sam Altman, CEO of OpenAI, once said that once AGI is invented, "poverty will truly disappear."
Compared to this, the Chinese government's statements about AI are at least more restrained.
The article states that although China also sees AI as a development priority and aims to reach world-leading levels by 2030, unlike the apocalyptic tone common in Silicon Valley, Chinese economic planners seem more concerned with using AI as a tool for industrial processes rather than creating superintelligent systems capable of reaching the "singularity." For example, the State Council's 2025 document on implementing the "Artificial Intelligence +" action plan focuses entirely on improving efficiency through applications, not on the so-called "intelligence explosion."
The article points out that another difference between the development of AI in China and the U.S. lies in the fact that China places more bets on simpler and lower-cost open-source AI models. In the U.S., most leading "frontier" AI models are proprietary and confidential, partly due to business model considerations, and partly due to some half-true, half-false concerns that the wrong players might cause human extinction.
More importantly, China is making large investments in a range of other technologies that can reasonably be called "future technologies," thus hedging its bets. For example, reports indicate that in 2024, China's investment in clean energy capital expenditures was estimated to reach $94 billion. This makes AI investment look modest in comparison, where AI is a complement, a binder, not the main structure.
The article points out that certain aspects of China's broad strategy are already yielding results: in 2024, 70% of global electric vehicles were produced in China; China also holds about 80–85% of the global solar photovoltaic manufacturing, and over 75% of battery production.
"Leave AI aside for a moment. Assuming these technologies represent the future, the leader is obvious," the author wrote.
The article mentions that there is another function to the narrative of a U.S.-China "AI competition": it serves as an excellent lobbying tool for the U.S. tech industry, providing legitimacy for unusually high levels of spending.
Different from China, while the U.S. doubles down on AI, it reduces investments in clean energy technologies and weakens support for basic research in other areas. "The U.S. bet on the future is increasingly like: either AI succeeds, or it fails completely."
In the author's view, one cannot deny that accelerationists might be right, and AGI that benefits humanity may still be achievable, but this outcome remains far from certain, and history and physics remind us to be cautious.
The article warns that betting on AI's continued exponential growth is full of uncertainty, and the current highly concentrated and almost monopolistic structure of the tech industry exacerbates the problem. The U.S. spending decisions have become highly centralized, made by just a few companies, and this way of working is more common in state planning systems, so the risk of groupthink is real. At the same time, despite the ambiguous business models, the stock market continues to generously reward AI investments, and has yet to exert any balancing effect.
Finally, the article analyzes why Silicon Valley is heavily investing in AI. One possibility is cultural factors — Silicon Valley is used to over-investing in new ideas. Another possibility is that, from an economic perspective, spending money on actual projects is better than hoarding profits or repurchasing stocks.
Perhaps there is also a darker explanation: the massive investment in AI by Silicon Valley may simply be a means for tech giants to consolidate their monopoly and strengthen their moats, to prevent startups from challenging their existing businesses. The article points out that if this is the case, the U.S. narrative on AI is less about U.S.-China competition or investment in humanity, and more about the desire of powerful companies to consolidate their positions.
Original: toutiao.com/article/7583630856850375194/
Statement: This article represents the views of the author alone.