How should we understand today’s debate over an AI bubble?

Home > Opinion > Columns

How should we understand today’s debate over an AI bubble?

Published: 05 Dec. 2025, 00:01

Audio report: written by reporters, read by AI

Kim Myung-ja

The author is the chair of the board at KAIST and the former minister of environment.

Whether the current wave of investment in AI represents a speculative bubble or a genuine technological transition has become a central question for policymakers and markets. Governments worldwide are accelerating AI initiatives, driven in part by a fear of missing out. The arrival of ChatGPT and DeepSeek, the volatility of GPU-linked stocks, and comments from the United States Federal Reserve on rates and liquidity continue to stir debate over whether an AI bubble is forming.

A trader works at his desk on the floor of the New York Stock Exchange (NYSE) after the opening bell in New York on Dec. 3. Wall Street's main stock indices dipped on Wednesday after data showed US businesses unexpectedly shed jobs last month. [AFP/YONHAP]

The discussion predates today’s frenzy. In 2016, when Klaus Schwab published "The Fourth Industrial Revolution" (2016) and the Go match between AlphaGo and Lee Se-dol captured global attention, I was serving as president of the Korean Federation of Science and Technology Societies. Observing the match from the next room, I organized a forum series on the fourth industrial revolution and later wrote "Reading World History Through Industrial Revolutions" (2019). A key question was how the world had plunged into the Great Depression despite the unprecedented material growth of the second industrial revolution between 1870 and 1930.

Samsung Electronics Chairman Lee Jae-yong (left), Nvidia CEO Jensen Huang, and Hyundai Motor Group Executive Chair Euisun Chung greet one another after finishing a “chimaek” gathering at a chicken restaurant in Samseong-dong, Seoul, on the evening of Oct. 30. [KIM KYUNG-ROK]

Ben Bernanke, who chaired the Federal Reserve from 2006 to 2014, earned his MIT doctorate by studying why a powerful economy like that of the United States had failed to prevent the Depression. In a 1983 paper, he argued that the collapse of the financial system crippled the real economy and turned a downturn into a disaster. Looking back on the 2008 financial crisis, Bernanke admitted that systemic risk had been underestimated despite years of warning signs. Yet his emergency liquidity measures were credited with preventing a repeat of the Depression, and he received the Nobel Prize in economics in 2022.

A renewed comparison between the 1920s and the 2020s appeared last October with the publication of Andrew Sorkin’s "1929: Inside the Greatest Crash in Wall Street History". Sorkin argues that the Depression emerged from intertwined psychology, institutional weakness and social structure. He asks whether this time is truly different. The United States of the 1920s was captivated by electricity, automobiles, telephones, and radio. Markets soared on the belief that innovation would guarantee rising stock prices. Margin trading that required only 10 percent down, excess liquidity, enthusiasm for financial democratization, weak regulation and official complacency combined to set the stage for collapse. Sorkin warns that the mix of technological innovation, financial democratization and credit expansion offers opportunity but also profound risk. If history is forgotten, he argues, it repeats itself.

Optimists contend that today’s AI boom differs fundamentally from the 1929 crash, the dot-com collapse of 1995 to 2000, and the cryptocurrency bubble of 2017. They argue that the dot-com era lacked the infrastructure needed to convert innovation into productivity. AI and digitalization, by contrast, raise both productivity and quality. Investment in GPUs and data centers is viewed as constructing long-term productivity infrastructure. Stanford professor Erik Brynjolfsson describes AI as a powerful general-purpose technology that may experience a phase of overheating and selective correction before delivering major gains.

Industrial revolutions of the past were engine-based transformations that lifted the productivity of resources, labor, and capital. The AI–data–coin–platform economy of the twenty-first century is an algorithmic revolution characterized by intangible value. The Financial Stability Board estimates that global financial assets total four to five times global GDP, while the nominal value of off-exchange derivatives and other non-physical assets reaches hundreds of trillions of dollars. The IMF warned in 2024 that a digital asset bubble could distort the real economy and weaken monetary policy. Connecting virtual valuations to real productivity has become essential.

The AI economy also moves on expectations and narratives rather than traditional measures of output. Claims that AI will reorganize industries, replace workers or determine geopolitical power shape investment and policy. The notion of “narrative economics,” introduced by Yale professor Robert Shiller in "Narrative Economics" (2019), highlights how economic behavior responds to stories. Managing the side effects of speculative techno-capitalism is emerging as a central policy task.

President Lee Jae Myung heads back to his seat after delivering remarks at the launch ceremony of the National Artificial Intelligence Strategy Committee held at Seoul Square in central Seoul on Sept. 8. The backdrop displays the slogan “Korea, Taking Flight with AI.” [YONHAP]

AI infrastructure is now reshaping labor, science, education, health care and culture while shifting global political and economic influence. Traditional metrics like GDP cannot measure the weight of virtual assets or growth detached from physical industry. I recall reporting to President Kim Dae-jung in 2001 on a plan to introduce a “green GDP” system that would account for ecological preservation and pollution. The initiative ended after two years because the political moment had not yet matured.

Today’s environment requires a techno humanist ecosystem capable of monitoring and integrating the ethical and environmental effects of AI into policymaking. Without GDP systems that reflect soaring electricity demand, grid strain, ecological disruption, climate and financial ethics, inequality and public trust, the sustainability of civilization will come under threat. A new conceptual frame — possibly a form of “civilizational economics” — may be necessary to redefine economic value in the age of AI.

This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.