[The Editors' Verdict] Sorting Out Winners and Losers in AI... Are Companies Prepared? View original image

At the end of last year, I attended a year-end gathering hosted by a private economic research institute that publishes an annual outlook on the Korean economy. Amidst well-wishes exchanged among attendees, an economist sitting at the same table abruptly raised a question: "What do you think about the artificial intelligence (AI) bubble?" As an economic advisor to the president, he added, "I am concerned that the AI bubble might burst in the new year."


He explained that his mention of the AI bubble was due to the fact that, despite massive investments by companies, profitability has been slow to improve. He questioned whether such a situation—like pouring water into a bottomless pot—could be sustainable. The chairman of the board at a domestic brokerage firm, also present at the table, agreed, saying, "Considering the potential impact on the capital markets, this is not an issue that can be ignored." Another economist linked the topic to his own research group, saying, "We're considering how to combine AI and finance, but the bubble issue could become a significant variable."


The AI bubble theory is hardly a new topic. It resurfaces every time performance reports are released by related companies. A prime example is OpenAI. Despite having sparked a global AI revolution three years ago and generating enormous revenue, OpenAI has yet to turn a profit. According to U.S. IT media outlet The Information and others, OpenAI's estimated revenue last year was about $13 billion. However, with massive amounts of capital required for model training, there are no signs of losses improving. There are even predictions that OpenAI's cumulative losses could reach $115 billion by 2029, with occasional rumors of bankruptcy. OpenAI's partner Oracle also fueled the AI bubble debate when it reported quarterly results earlier last month that fell short of market expectations.


So far, the AI bubble controversy has been little more than a "storm in a teacup." There are no signs of trouble on the investment front yet. OpenAI has been attracting investors from around the world, dismissing concerns about a bubble. Not only did it secure $40 billion (about 58 trillion won) from Japan's SoftBank, but it also received multi-billion-dollar investments from Disney. MGX, a UAE-based tech investment firm specializing in AI and semiconductors, also invested in OpenAI. The demand for semiconductors, essential to realizing AI capabilities, is at an all-time high. The performance of memory companies such as Nvidia, Samsung Electronics, and SK hynix has been extraordinary. Jensen Huang, CEO of Nvidia, even declared 2024 as "the first year of a decade-long AI boom." Moreover, considering the fierce competition between the United States and China for AI supremacy, some believe that it will not be easy to halt the inflow of capital.


However, investment without guaranteed returns cannot continue indefinitely. With the full-scale launch of physical AI products and other tangible solutions, market watchers project that this year will be the first to truly test AI's profitability. There is talk of a market shakeout, with selection beginning to favor companies that either have strong financial resources or possess highly efficient AI models.



Among Korean business leaders and CEOs, AI was by far the most frequently mentioned word in New Year's addresses this year, reflecting high expectations for industrial innovation driven by AI. However, what preparations are being made for the silent threats like the AI bubble? If the AI bubble bursts, not only semiconductors but also infrastructure such as power and data centers will be affected, dealing a blow to all of Korea's key industries. Given that the president's economic advisor has raised this issue, it appears that the government will pay close attention to this matter throughout the year.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing