[Gangnam Perspective] The Pitfalls of the AI Bubble Debate

There is a bubble issue that seems ready to burst at the slightest touch: the debate over artificial intelligence (AI) bubbles. The prevailing narrative that AI is taking over industries has been quickly followed by confusion as concerns about an AI bubble emerge. The confusion stems from the fact that bubble theories and conspiracy theories are intertwined. Rational arguments about 'performance verification relative to investment' are mixed with suspicions that 'someone is deliberately manipulating the market.' This makes it all the more important to distinguish between genuine concerns and unfounded speculation.
First, the bubble theory serves as a 'warning light,' not a doomsday scenario. It does not suggest that the technology itself has failed. Rather, it signals that expectations have outpaced reality. The core of the AI bubble debate also lies in skepticism about return on investment (ROI). Some point out that, despite massive investments, actual profits are only about one-sixth of expectations.
Typically, when bubble theories gain traction, the market undergoes a correction. This leads to reduced over-investment, a slowdown in infrastructure spending such as graphics processing units (GPU) and data centers, and the elimination of projects with no tangible results. Many countries and companies become hesitant to invest. This was the case during the dot-com bubble in the late 1990s, when the bursting bubble caused countless internet companies to collapse.
However, there is a twist. The optical communication infrastructure and database technologies built during the dot-com era became fertile ground for the growth of Google LLC, Amazon.com, Inc., and Facebook after the bubble burst. In this way, bubbles can be part of an industry’s maturation process. Every time a new technology emerges, excessive expectations and investments are followed by a correction phase, which stabilizes the industry. This process reins in overheated markets, weeds out baseless speculation, and helps distinguish genuine innovation. In that sense, bubble theories can be both poison and medicine.
There is another trap that fuels misunderstanding and confusion: 'AI fear marketing.' It is no exaggeration to say that the advent of AI has split the industrial landscape into AI and non-AI sectors. As investment pours into AI, previously promising sectors like the bioindustry are now experiencing a funding drought. This concentration of investment in AI can be seen as a factor fueling the AI bubble debate.
Interestingly, fear marketing actually increases publicity for AI. Warnings about 'AI concentration' breed anxiety that 'we might be falling behind in the AI race,' prompting every industry to adopt an 'AI+X' strategy. What was once a warning about market risk has become a catalyst for expanding AI investment.
Distinguishing between global trends and South Korea’s unique circumstances is also helpful in understanding the bubble debate. The global AI market must grow for products and services to sell well. However, the ability to develop and market these successfully depends on each country’s industrial structure and corporate competitiveness. South Korea possesses world-class competitiveness in industries such as manufacturing, logistics, automobiles, and semiconductors, where AI adoption is particularly challenging. This means the country has a solid industrial base that directly impacts productivity and quality. The perception of the AI bubble varies from country to country. Conversely, South Korea’s industrial competitiveness is also at a crossroads. With the People’s Republic of China (PRC) overtaking in some sectors, a breakthrough in innovation is urgently needed. Maintaining the status quo with existing industrial policies will only preserve the current state. Whether it is AI or any other technology, a major momentum is needed to dramatically enhance industrial competitiveness.
The pain of bubble debates lies in the decision-making paralysis they cause. There are two simple ways to respond to the AI bubble theory. If market risks seem high, it is reasonable to go along with the bubble narrative. Conversely, if there is potential to create real market value, one can work to establish the dominance of AI. The essence of the bubble debate is not rejection but adjustment. During correction phases, the easiest choices are project and workforce reductions. What is clear is that a handful of companies that seized the golden opportunity during these periods of adjustment have emerged as leaders in future markets.
jjack3@fnnews.com Reporter