Artificial intelligence: Embrace optimism, but calculate with a cool head
Translated from Korean, summarized and contextualized by DistantNews.
At a glance
- South Korea is poised to benefit significantly from the AI boom due to its manufacturing strengths in semiconductors, but faces challenges in rapidly scaling up infrastructure.
- The government and corporations are pushing for large-scale projects, including semiconductor factories and data centers, potentially increasing reliance on nuclear and LNG power and shortening environmental reviews.
- Concerns are rising about the sustainability of massive AI investments, potential financial risks from over-investment, and the need for a more cautious approach to infrastructure development and resource allocation.
South Korea stands to gain substantially from the artificial intelligence revolution, leveraging its established strengths in semiconductor manufacturing. However, the nation faces a critical juncture, balancing ambitious expansion plans with the potential risks of rapid, large-scale infrastructure development.
The decision to focus on information technology infrastructure in the mid-1990s, with the slogan 'Let's lead in information technology even if we are late in industrialization,' made South Korea an outstanding student in the knowledge economy.
The government and major corporations, including Samsung and SK, are accelerating plans for massive projects. President Lee Jae-myung has announced three mega-projects totaling 4,755 trillion won, emphasizing speed in constructing semiconductor complexes and data centers. This push involves potentially increasing reliance on nuclear and liquefied natural gas (LNG) power plants and streamlining environmental impact assessments to meet the projected surge in electricity demand, which could reach 15% of current domestic usage by 2040 for data centers alone.
The government is prepared to push through any obstacles like a 'battle tank' to build semiconductor factories and data centers and supply them with electricity and water.
This optimistic outlook is being challenged by emerging concerns about the feasibility and sustainability of such rapid growth. Globally, the construction of large data centers is facing delays due to financial and physical constraints. The astronomical investments in AI, exceeding $1 trillion by major tech firms like Google, Amazon, and Microsoft, have strained corporate finances, leading to increased borrowing costs. Furthermore, power grid bottlenecks, lengthy waiting periods for grid connections (5-10 years), and growing public opposition to data centers due to concerns over electricity, water, and noise pollution are complicating site selection.
The premise of this optimism is the assumption that massive data center investments for artificial intelligence services will continue. However, that premise is currently being shaken.
The article warns of potential financial contagion, drawing parallels to historical bubbles like the U.S. subprime mortgage crisis. The current AI investment frenzy, coupled with significant off-balance-sheet debt among AI companies and complex financing structures, could trigger a credit crunch or even a wider financial crisis. The author urges a more cautious approach, advocating for a thorough re-evaluation of mega-projects and a more meticulous planning process rather than a headlong rush. The "blitzkrieg" strategy, while effective in initial advances, risks disconnection from supporting forces, highlighting the need for a balanced and realistic assessment of risks and capabilities.
The pattern of over-investment through debt and off-balance-sheet financing, driven by a winner-take-all mentality, without profitability to match, has always ended in 'investment suspension and recession.'
Originally published by Hankyoreh in Korean. Translated, summarized, and contextualized by our editorial team with added local perspective. Read our editorial standards.