Summary of Key Points
The surge in generative AI has disrupted the steady growth of corporate values. Over the past decade, the global value of companies has shifted by $27 trillion, with one-third of this shift occurring in the two years following the rise of AI (2022-2024). Investors are increasingly willing to pay a premium for the future potential of companies that utilize AI. Chinese companies are accelerating their digital transformation, but only 10% have successfully transformed their AI investments into competitive advantages, and even fewer have turned these advantages into tangible performance improvements. Meanwhile, the cost of using AI has been significantly underestimated, leading to budget overruns and concerns among investors about valuations. However, in the long term, capital remains patient (at least for the next three years).
Detailed Analysis
1. Generative AI Doubles the Speed of Corporate Value Shift
The migration of corporate value can be seen as money flowing towards companies with greater future potential. Over the past decade, the total global corporate value shifted by $27 trillion, which is like redistributing 270,000 smaller amounts of wealth. In the two years when AI gained momentum (2022-2024), this shift accounted for one-third of the total, doubling the previous rate.
Why? Investors now place more emphasis on a company's ability to create value through AI. For example, the ratio of a company's total value (including equity and debt) to its current earnings before interest and taxes has increased by 25% over ten years. This means that even if a company doesn't generate much profit now, investors are willing to pay more for its stock if they believe it has a strong AI strategy and the potential for future growth.
2. Chinese Companies Are Moving Fast, but Their Internal Capabilities Aren’t Keeping Up
China’s digital transformation index increased by 10 points year-on-year in 2026, the largest increase in recent years, indicating a surge in AI adoption. There has been a 27% improvement in digital twins for intelligent manufacturing and a 22% increase in intelligent research and development capabilities, as well as a 14% improvement in data governance maturity. 88% of Chinese companies have moved beyond the pilot phase of AI applications, and 44% have even started to redesign job requirements.
However, only 10% of these companies have successfully transformed their AI investments into competitive advantages, such as autonomous decision-making, enhanced product value, or global collaboration (meeting at least two of these criteria). Most companies are still competing in their existing businesses without truly opening up new growth opportunities; only 14% have achieved significant value improvements through AI.
3. AI Is Not a Cost-Saving Tool, but Rather an Expense-Generating One
Initially, people thought AI would reduce costs, but it has turned out to be a major expense. Cao Qifeng points out that the cost of using AI is greatly underestimated. For example, the cost per call during testing may be low, but in real-world scenarios (high frequency, multiple interactions, across different tools), it can become a continuous burden. The cost of the same task can vary by up to 17 times depending on how AI is used.
A typical example is the “token tax”: each word, phrase, or segment processed by AI incurs an additional cost. Uber ran out of its annual AI budget in just four months and had to set limits on token usage; Meta and Amazon have even eliminated internal evaluations based on token consumption. Chinese companies are facing similar issues, with rising costs for computing power and API calls, which are affecting their profits and return rates.
4. The Capital Market Is Worried: Mismatch Between AI Investments and Outcomes, Could a Bubble Be Forming?
The high cost of building AI infrastructure (such as purchasing computing power and developing models) contrasts with the limited ability to commercialize these technologies, leading to concerns among investors. They wonder when they will see returns on their investments. If there are no results or if regulatory policies change, capital may withdraw.
As a result, there is growing concern about “bubbles” in the market, and global tech giants are even starting to slow down investment in digital infrastructure. This reflects capital’s concerns about whether the expected benefits will be realized quickly enough.
5. Long-Term Optimism, but Short-Term Prudence Required
Cao Qifeng believes that AI will definitely bring value to companies, but in the short term, businesses need to carefully consider the economic benefits of their investments. For example, they must quantify the marginal cost of using AI and determine the potential returns.
Regarding investor patience, he suggests that capital will not lose confidence for at least the next three years. However, if companies continue to invest without generating results or fail to control costs, bubbles may burst. Companies need to balance long-term strategic investments with short-term profitability and avoid simply “burning money” without creating sustainable value.
In Conclusion
AI is the future, but we first need to address the issue of balancing expenses and profits. Without resolving this, no matter how great the potential of AI is, it won’t be enough to overcome investor concerns.