Nobel laureate Michael Spence: The AI investment boom is a rational bubble

Spence believes that over - investment in artificial intelligence is actually a rational choice. The logic behind these investment decisions is that the cost of being eliminated from the competition would be much higher than the current investment.

Has the AI (Artificial Intelligence) investment competition among Chinese and American tech giants created a huge bubble?

On November 13th, at the "Taihu World Cultural Forum · Qiantang Dialogue", Huang Yiping, the dean of the National School of Development at Peking University and a member of the Monetary Policy Committee of the People's Bank of China, had a conversation with Nobel laureate in economics Michael Spence about issues such as the AI investment boom.

Spence said that the enthusiasm and investment in the digital field, especially in AI, have sent the US stock market into a frenzy. "The entire market is making large - scale investments in AI models and quantum computing. We are also making large - scale investments in data centers and related power supplies. In the future, there may be large - scale investments in power generation capacity because the demand for electricity is constantly rising."

In Spence's view, there are some bubbles in the market, but it is a rational bubble.

"From the perspective of Google, Microsoft, Meta, or Amazon, the cost of becoming the third - place in the competition is far greater than the losses caused by over - investment or inefficiency. So they are all investing at a very high speed," Spence said. "We are facing an investment competition. This kind of competition exists among large - scale tech companies in China, and also among those in the United States."

The logic behind these investment decisions is that once eliminated from the competition, the cost will be much higher than the current situation. Spence believes that over - investment in artificial intelligence is actually a rational choice.

The competition between China and the United States has also boosted the AI investment boom.

"In terms of the policy framework, both China and the United States, for various reasons such as economy and national security, believe that they cannot fall far behind the other in technology. So, they may not necessarily promote the formation of bubbles in these fields, but they definitely won't hinder it," Spence said. "Whether in China or the United States, the development of artificial intelligence is accelerating. The gap between China and the United States is narrowing, and in my opinion, they are now almost on par."

Under the AI investment boom, more than 30% of the market value of the S&P 500 index is concentrated in the top seven tech companies, reaching an unprecedented level.

The AI investment boom has also greatly boosted the economic growth of the United States.

Huang Yiping mentioned in the conversation that an analysis by Jason Furman of Harvard University shows that in the first half of 2025, the US economic growth was 1.6%. However, if the contribution of (investment in) data centers is excluded, the economic growth rate would only be 0.1%.

This also raises a question: Is the large - scale investment in artificial intelligence and its related fields sustainable? In other words, will artificial intelligence support long - term economic growth?

Spence believes that AI - related investments have no significant impact on productivity growth in the short term, but the potential for future productivity growth is very large. "People often overestimate the short - term impact of major technological breakthroughs and underestimate their long - term impact. One reason for the emergence of bubbles is that people overestimate the speed at which the impact occurs. We saw this during the Internet bubble around 2000."

Spence further said, "The most important thing is that if we want to achieve a surge in productivity, we need to unleash the potential of artificial intelligence in the economic field. This is a problem of technology diffusion. If artificial intelligence brings about great changes in the technology and finance fields, while sectors that absorb a large amount of employment, such as the hotel industry, traditional retail, and education, are only slightly affected, then the surge in productivity will not come. We will only have an economy that was once strong but lacks long - term momentum."

Another related question is whether artificial intelligence is a productivity revolution comparable to the invention of the steam engine and electricity, or just a marginal innovation in technology and methods?

Spence believes that artificial intelligence is very important. At the same time, its progress is very rapid, and the artificial intelligence we talk about today may be completely different from what we will have tomorrow. "If we don't know the development direction of AI agents and how reliable they will be in the future, it is difficult to accurately predict what impact they will bring."

However, Spence said that it can be seen that artificial intelligence has already had a huge impact on the scientific and technological fields. Take the field of biomedical science as an example. The AlphaFold AI launched by DeepMind is being used by more than 2.6 million scientists in over 100 countries.

"Artificial intelligence is revolutionary in promoting the progress of various scientific fields, and its impact on the economy will come more slowly," Spence said. "Because you need to figure out how to use it, conduct experiments, change existing behaviors, learn new skills, change business models and corporate structures, all these things we are familiar with. But I think it is reasonable to speculate that if used properly, AI will have a very huge impact."

Looking to the future, what impact will the rapid development and wide application of artificial intelligence bring to the world?

"My overall judgment is that as the Asian development model centered on labor arbitrage gradually fades away, artificial intelligence will bring significant opportunities to the world. It may be very important to have a powerful potential growth engine as a replacement," Spence said.

At the same time, Spence mentioned that in the future, different countries will have different divisions of labor in the AI industry chain.

Spence believes that in the next period, the development of tools will mainly be concentrated in China and the United States, because other countries cannot compare with China and the United States in terms of computing power, talent, and resource investment. However, the cost of using or adjusting these tools to solve problems in the real world will be much lower. This will bring many opportunities to many countries in the Global South.

In response, Huang Yiping mentioned a distribution problem. In the past process of globalization, some groups in society have benefited relatively less, which has also led to a backlash against globalization policies. Will there be a similar situation of uneven distribution in the AI revolution?

"This may indeed happen," Spence said. "Globalization and previous digitalization were essentially just the automation of routine work, but this did produce unsatisfactory distribution results. I think the situation may change this time. For example, David Autor said that if we do it right, we may not repeat the same mistakes."

At the same time, Spence emphasized that simply relying on the market mechanism and completely believing that it can automatically produce the best results is not the right approach. "This requires our efforts to build, rather than wishful thinking."

(The author is a reporter from Caijing)

This article is from the WeChat official account "Caijing Mayflower" (ID: Caijing - MayFlower). Author: Tang Jun, Editor: Zhang Wei. Republished by 36Kr with permission.