Will the AI bubble burst, and will the job market usher in a new era? Nobel laureate economist: Danger and opportunity coexist

AI is rapidly stirring the global economy: on one side, capital is flooding in wildly, supporting economic growth, while Nobel laureate economist Joseph Stiglitz openly states that an AI bubble has already formed, and its burst could shock the macroeconomy, leaving workers facing unemployment crises with no mechanisms to respond. On the other side, he is firmly convinced that once this pain is endured, the AI currently threatening jobs will ultimately become the most effective partner in the workplace. This seemingly contradictory judgment contains the ultimate answer to the economy and the job market in the AI era, and also touches on the survival anxieties of each of us.

1

Contradiction and Unity:

Stiglitz Discusses the Dual Impact of AI

Nobel laureate Joseph Stiglitz believes that people can simultaneously accept two seemingly contradictory viewpoints:

First, the current AI sector is forming a bubble. Once it bursts, it could impact the macroeconomy, and workers may face large-scale unemployment, with society not yet having effective mechanisms to address this issue.

Second, if we successfully navigate this transitional phase, the AI technologies that threaten jobs today will eventually become the most capable assistants and partners in the workplace.

Recently, in an interview with Fortune magazine, Stiglitz pointed out: “Currently, our economic growth largely depends on investments in AI, and this investment is essentially an AI bubble.” He stated that one-third of last year’s global economic growth, and even the economic stagnation, are related to AI. Therefore, in the short term, the AI bubble does play a positive role in macroeconomic stability, but he believes that the AI bubble mainly manifests on two levels.

“This phenomenon will have both short-term impacts and long-term consequences,” Stiglitz emphasized. The current public discourse tends to focus on only one aspect, neglecting the other.

2

The Current AI Investment Boom Is Unstable, and the Risk of Bubble Burst Cannot Be Ignored

In his 2024 book, The Road to Freedom: Economics and the Good Society, Stiglitz deeply analyzes the structural flaws of modern capitalism. He believes that the current wave of AI investment is fundamentally unstable.

He explains that the market’s belief that these investments will yield substantial returns is mainly based on two assumptions: first, that AI technology will succeed; second, that market competition in AI will remain limited. However, the reality is that global competition in AI is now extremely fierce, with major US tech giants and Chinese companies actively involved.

Stiglitz points out that even if AI technology ultimately succeeds, fierce market competition will drive industry profits down to very low levels, making it impossible for companies to realize the previously expected returns. Once the market recognizes this, it could trigger very serious consequences.

He warns that if the AI bubble truly exists, its burst will inevitably cause a huge shock to the macroeconomy in the short term. More alarmingly, this crisis could occur simultaneously with a wave of unemployment across entire industries triggered by AI. In his view, this “worst-case scenario” is entirely possible.

Stiglitz argues that, at present, there are no comprehensive mechanisms—either macro or micro—to effectively respond to AI-induced unemployment. We lack well-developed active labor market policies, large-scale retraining systems, and industry strategies capable of creating high-quality new jobs in sectors where traditional roles disappear. Addressing this requires large-scale retraining programs, but current projects are far from meeting actual needs.

3

AI May Impact White-Collar Workers, but Society’s Response Mechanisms Are Still Lacking

Stiglitz has personally witnessed the consequences when society lacks effective mechanisms to cope with transformative change. Taking the Great Depression as an example, he said: “Back then, agricultural productivity indeed increased significantly; we greatly improved productivity and no longer needed as many farmers, but we had no way to transfer these workers from rural areas. It wasn’t until World War II broke out, and government intervention due to wartime demands, that the problem was finally solved—something that should have been addressed through a well-designed institutional framework.”

This historical analogy serves as a warning today. If AI truly automates many routine cognitive tasks—such as research, copywriting, data analysis, administrative work—supporting millions of office jobs—and the economy lacks mechanisms to reallocate these workers, the consequences will go beyond productivity changes and will profoundly influence the development trajectory of human society.

“A stable macroeconomy will be very hard to sustain,” Stiglitz bluntly states. “I see little chance that such a situation can be maintained smoothly.” He explicitly points out that AI will have a particularly significant impact on routine white-collar jobs, which are mostly held by college-educated office workers. Many of these workers believe they won’t face the same kind of job shocks that the previous generation of manufacturing workers did.

In fact, the sense of security these knowledge workers currently hold may be an illusion.

4

Long-term Perspective on AI’s Role:

Assistant Rather Than Replacement, Human Still Central

However, at this point, Stiglitz’s view takes a crucial turn, making his outlook deeper than mere pessimism or optimism. Looking long-term, beyond the AI bubble and unemployment shocks, the essence of this technology is no longer about replacing human labor but about becoming a tool to enhance work efficiency.

For example, in education—an industry that accounts for about 14% of the total workforce—Stiglitz is clear: AI will not replace teachers. It can assist teachers in optimizing lesson plans and implementing personalized instruction, but it cannot substitute for teachers entirely. Our understanding of learning patterns shows that human interaction remains vital in education.

The situation is similar in healthcare, though the political sensitivity is higher. US healthcare spending accounts for nearly 20% of GDP, yet its efficiency is far below the global average. Despite poor outcomes, costs are much higher than in comparable countries. Optimists hope AI can solve these issues, but Stiglitz disagrees. He points out that the root causes of inefficiency in the US healthcare system are rent-seeking behaviors, lack of competition, and the absence of a robust public health system—fundamentally political problems that AI cannot resolve.

AI can optimize medical records, accelerate drug development, and improve diagnostic accuracy, but it cannot overhaul insurance systems, break hospital monopolies, or make necessary political decisions for systemic reform. The core issue has never been a lack of computational power.

He also uses the example of a plumber to illustrate his understanding of “Intelligent Assistance (IA).” In his view, AI’s future role is to assist work, not replace workers. A plumber, for instance, will not be replaced but will become more professional with AI’s help. AI can diagnose issues based on symptoms—such as detecting a burst pipe inside a wall—and help plumbers locate problems quickly. But this is only an auxiliary role.

Finally, he summarizes his long-term view in one sentence: “But you still need a plumber.”

5

Only by Smoothly Navigating the Pain Can AI Truly Empower the Future

This hopeful long-term outlook is only possible if society can smoothly pass through the initial crises and maintain a sound institutional framework. If a short-term AI bubble burst triggers mass unemployment, and the economy and society lack comprehensive safety nets and retraining programs, and if governments are unable to intervene effectively, then the long-term vision of AI as a powerful assistant to humanity will become distant. This is not due to a failure of the technology itself but because the fair social foundations necessary for its healthy application will have already collapsed before they are truly needed.

Stiglitz’s warning is not an assertion that AI will completely destroy employment prospects but emphasizes that the most dangerous period is the current transition from present to future. We are entering this critical phase unprepared.

This article is translated by Digital Transformation Network. Images sourced from the internet.