Recently, The Economist published a rather controversial commentary article.
The article argues that while China has made remarkable breakthroughs in robotics, artificial intelligence, new energy, and high-end manufacturing, domestic consumption remains sluggish, corporate profits are under pressure, and market confidence has not yet fully recovered. The article even puts forward a sharp point: to some extent, China's obsession with winning the tech race has distorted resource allocation, directing vast amounts of capital, talent, and policy resources towards cutting-edge technology fields without adequately addressing the insufficient demand problem in economic growth.
This viewpoint has sparked considerable controversy in China. On the surface, it appears to criticize China's development of robotics and AI, but setting aside ideological overtones, the article actually touches on a problem far deeper and more significant than China's economy: when machines become increasingly capable, how do humans share in the value created by machines?
When machines become increasingly capable, how do humans share in the value created by machines?
In fact, this is not just a problem for China, nor is it merely an issue for the AI industry. It is the core political economy question that all countries globally must face in the next two decades.
I. The Fundamental Question Underlying the Controversy: AI is Rewriting the Underlying Logic of Wealth Distribution
The industrial civilization of the past two hundred-plus years was built upon a relatively stable system of wealth creation and distribution: capital provides funds, enterprises organize production, laborers participate in value creation through work, and then share the fruits of economic growth in the form of wages. Whether it's the tax system, pension system, social security system, or the formation of consumer markets, their underlying logic is built on the foundation of sustained growth in labor income.
Although the Industrial Revolution continuously improved production efficiency, machines always remained just tools. Steam engines enhanced physical labor efficiency, assembly lines improved manufacturing efficiency, and computers boosted office efficiency. Regardless of technological progress, humans remained the irreplaceable core component of the production system. Therefore, productivity gains could ultimately translate into employment expansion, wage growth, and consumption growth, thereby forming a virtuous cycle of economic development. The rise of the middle class in Europe and America over the past century is essentially the result of this cyclical mechanism at work.
However, AI is changing all of this.
For the first time in history, the tools created by humans are beginning to possess the ability to replace cognitive labor. Steam engines replaced muscles, robots replaced repetitive labor, while large models and AI agents are now entering the domain of knowledge work. Anthropic has disclosed that over 80% of its internal code is generated by Claude; a significant portion of R&D work at tech giants like Microsoft, Google, and Meta has already begun to rely on AI assistance; and traditional white-collar industries such as law, finance, consulting, education, media, and design are experiencing unprecedented impacts from intelligent automation.
80%+
of Anthropic's internal code
is now generated by Claude
Cognitive Labor
Large Models and AI Agents
First time entering knowledge work domains
If AGI truly arrives in the future, machines will not only be able to undertake physical labor but also a large amount of knowledge work, management work, and even some decision-making labor. At that point, human society will face, for the first time, a situation where the economy continues to grow, corporate profits continue to rise, production efficiency keeps improving, but the number of people directly participating in value creation becomes smaller and smaller—this is the most profound challenge of the AI revolution.
In recent decades, the biggest concern when discussing technological revolutions has been unemployment. But from a longer-term perspective, the issue of distribution may be more critical than unemployment. Because consumption stems from income, and income stems from employment. If more and more labor is replaced by machines, and new sources of income cannot be established, then the entire economic system will face a fundamental contradiction: enterprises become more productive, machines become more efficient, but the purchasing power of consumers grows slowly or even declines. The "insufficient demand" problem that Keynes once worried about may re-emerge in a new form in the AI era.
II. The Real Dilemma: The Global Transmission Disconnect Between Technological Dividends and Mass Income
Looking at The Economist's criticism of China from this perspective, one realizes that its analysis captures the phenomenon but fails to touch the essence. China's problem is not having too many robots, nor is it developing AI too fast. On the contrary, as the old real estate-driven growth model gradually exits the historical stage, China must rely on new quality productive forces like artificial intelligence, high-end manufacturing, robotics, new energy, and semiconductors to find new growth engines. If technological innovation is abandoned, China's economic transformation would face even greater risks.
The real question is how China can connect the transmission chain between "technological innovation - corporate profitability - resident income - consumption growth."
Today, China possesses the world's most complete industrial system, with continuous breakthroughs in areas like new energy vehicles, photovoltaics, lithium batteries, industrial robots, and artificial intelligence. Yet, at the same time, consumer confidence remains weak, some industries are caught in price competition, and corporate investment willingness is insufficient. This indicates that an effective cycle has not yet been formed between technological supply capability and demand creation capacity.
Insufficient demand has never been because ordinary people are unwilling to consume, but because income expectations are unstable, balance sheets have been impacted, and the social security system still needs improvement. In other words, the problem China faces today is essentially not a technological one, but rather the problem of how productivity dividends diffuse throughout society.
In fact, this problem is beginning to manifest globally.
Over the past few years, the US tech giants have contributed the vast majority of the gains in the S&P 500 index. Companies like OpenAI, Anthropic, NVIDIA, and SpaceX have created astonishing wealth myths. The combined market capitalization of the world's top ten tech companies now exceeds $20 trillion. Yet, at the same time, labor's share of income in the US GDP has been declining long-term, with wealth increasingly concentrated among the minority who control data, algorithms, computing power, and capital.
$20 Trillion
World's Top Ten Tech Companies
Total Market Capitalization
Labor Income Share
data-check-id="505721">Long-term decline in the USWealth concentrating among the few
AI is creating an ever-larger economic pie, but the number of people sharing it is getting smaller. This is also why more and more economists are beginning to discuss the issue of "sharing the AI dividend."
AI is creating an ever-larger economic pie, but the number of people sharing it is getting smaller.
AI makes the economic pie bigger, but fewer people are sharing it
III. The Path Forward: Building a Shared Distribution System is the Core Competitiveness of the Intelligent Era
Three different development paths may emerge in the future world:
The first is the traditional capitalist path, where machines are owned by capital, and profits created by AI primarily belong to shareholders, with secondary distribution occurring through tax and welfare systems. The US is largely proceeding along this path.
The second is the state capitalist path, where the state participates in AI infrastructure construction through sovereign funds, public capital, and strategic investments, allowing the public to indirectly share in AI benefits. Discussions proposed by Trump about the government holding shares in top AI companies essentially carry the shadow of this thinking.
The third path is more innovative, involving the establishment of digital sovereign funds, universal shareholding mechanisms, or even new basic income systems for the AI era, allowing members of society to directly share in the incremental value created by the intelligent economy.
In a sense, oil defined the twentieth century, while data, computing power, and algorithms may define the twenty-first century. If oil revenues can become the source of sovereign wealth funds for entire populations, then whether the excess profits created by AI should also be partly returned to society will become one of the most important public policy issues of the future.
For China, this could precisely be a new strategic opportunity.
China's greatest advantage lies not only in its vast market, complete industrial chain, and rapidly developing AI technology, but also in its institutional capacity to coordinate industrial policy, social security, and long-term development strategies. What China truly needs to consider for the future is not how to slow down the development of robots, but how to transform the wealth created by robots into growth in resident income, enhancement of consumption capacity, and improvement of social security.
This means that future policy priorities must not only continue to support the development of artificial intelligence, robotics, and advanced manufacturing, but also simultaneously advance income distribution reform, social security system improvement, vocational retraining system construction, and the exploration of AI dividend-sharing mechanisms. Technological investment must gradually shift from solely pursuing capacity expansion and industrial competition towards creating broader social value. Only when ordinary people can share in the benefits brought by technological progress will the technological revolution not become a feast solely for a few enterprises and capital.
Ultimately, the real question posed by The Economist is not "Should China develop robots?" but rather "To whom does the wealth created by robots ultimately flow?" This question applies equally to the US, Europe, Japan, and all future countries entering the intelligent society.
Conclusion
The competition over the next two decades may not only be about who possesses the strongest large models, the most robots, and the largest computing power centers. More importantly, it will be about who can establish a new distribution system adapted to the era of the intelligent economy. Because the ultimate question the AI era must answer is not what machines can do, but rather: when machines start working, what do humans rely on for income? When AI creates value, how do humans possess value? When productivity grows at an unprecedented speed, how can the fruits of growth be translated into broader social well-being.
The ultimate question the AI era must answer is not what machines can do, but rather: when machines start working, what do humans rely on for income? When AI creates value, how do humans possess value? When productivity grows at an unprecedented speed, how can the fruits of growth be translated into broader social well-being.
This is the biggest political economy question of the AI era, and the ultimate proposition that will determine the future world order and the form of civilization.
For China, this question is particularly important, because China must not only win the technological competition but also rebuild its economic circulation; it must not only become an AI power but also a country where the majority shares in the AI dividend. Whoever can solve this problem first will not only seize the initiative in the next round of technological revolution but also grasp the discourse power in the next round of civilizational evolution.
💡 Interactive Topic
Do you support letting ordinary people directly share in the dividends created by AI through means like digital sovereign funds, universal shareholding, or new forms of basic income? Feel free to leave your views in the comments section.
This article is from the WeChat public account: My Second Half in the AI Era, Author: Dr. Xi Chunying
