SK Group Chairman Chey Tae-won Warns that Market Optimism Over SK Hynix’s Potential Over $100 Billion Operating Profit This Year Could Turn Into a $100 Billion Loss. He Emphasizes that Artificial Intelligence Is in a Transformation Phase Devouring Everything, and the Volatility Caused by Technological Change Has Reached Unprecedented Levels.
Chey Tae-won stated on the 20th in Washington D.C. at the “Trans-Pacific Dialogue” event that the uncertainty of the AI era has rendered traditional planning cycles meaningless, “A one-year plan is meaningless; the changes between the beginning and end of the year are too great.” He pointed out that companies with ample funds and resources will dominate the AI competition, but even industry leaders find it difficult to determine whether they have the right solutions.
This statement highlights the double-edged sword facing the semiconductor industry amid the AI wave. Despite SK Hynix achieving its best-ever performance last year driven by surging demand for high-bandwidth memory, with a market value soaring to 690 trillion won, Chey Tae-won’s warning reveals the risk that accelerated technological iterations could instantly change market dynamics, posing a significant challenge to the semiconductor industry, which relies on long-term investments.
Chey Tae-won also emphasized that the energy demand of AI data centers is becoming a societal challenge, with the infrastructure for AI data centers in the U.S. alone requiring a $5 trillion investment. He stated that the cooperation among Korea, the U.S., and Japan at this turning point will determine future order.
Profit Expectations Doubling from $50 Billion to $100 Billion
SK Hynix’s performance outlook has changed dramatically in just a few months. Chey Tae-won revealed that market forecasts in December last year predicted over $50 billion in operating profit this year, but the latest expectations have surpassed $100 billion. This figure is approximately 144.85 trillion won.
This surge in expectations is based on SK Hynix’s strong performance last year, with sales of about 79 trillion won and operating profit of 47 trillion won, both hitting record highs. The growth in high-bandwidth memory demand and robust sales of general DRAM were the main drivers, and the company’s market value, along with Samsung Electronics, helped push the Korean KOSPI index into the 5,000-point era.
However, Chey Tae-won remains highly cautious about this seemingly positive outlook. He said, “New technology may be a solution, but it can also make everything vanish,” emphasizing that even the most advanced technological advantages can quickly be lost amid industry upheaval driven by AI.
Market “Distortion” Phenomenon
Chey Tae-won pointed out that the construction of AI infrastructure is causing structural distortions in the semiconductor market. He said companies need to produce “monster chips” like HBM using the most advanced technology, which can bring truly huge profits.
However, abnormal phenomena are occurring in the market. Due to AI infrastructure consuming large amounts of memory supply, the profit margins of general memory are surpassing those of higher-tech HBM products. Chey Tae-won called this phenomenon “distortion,” reflecting that AI demand’s impact on the entire semiconductor supply chain has exceeded normal market rules.
This distortion highlights the strategic dilemma faced by semiconductor companies: on one hand, they need continuous R&D investment in cutting-edge technology; on the other hand, rapid changes in market structure can make the value of technological advantages unpredictable.
Energy Bottlenecks as the Biggest Challenge for AI Development
Chey Tae-won views energy supply as one of the most severe challenges in the AI era. He said, “AI is almost consuming all the energy and electricity we need,” and if the electricity demand is not properly addressed, it will become a huge societal challenge.
Specific data underscores the seriousness of the issue. Chey Tae-won revealed that building a data center requires about $50 billion, and the capacity demand for AI data centers in the U.S. is about 100 gigawatts, with infrastructure investments alone needing $5 trillion, not including energy costs.
He admitted, “I’m not sure if we have a solution,” and stated that the current situation has become one where each data center needs to be matched with a 1-gigawatt power plant. This close coupling of energy and computing power far exceeds traditional industry planning capabilities.
SK’s Integrated Data Center and Power Plant Plans
In response to energy challenges, Chey Tae-won disclosed that SK is preparing a new solution to build AI data centers together with power plants. He said that even after building data centers, many technologies can be developed to make their use more efficient, and the company will focus resources on R&D to enhance competitiveness.
This strategy aligns with the energy infrastructure development focus promoted by the Trump administration. When asked about U.S. investment plans in the energy sector, Chey Tae-won said, “We have been doing that all along,” but emphasized that AI changes occur within months, whereas energy projects typically take five years, making the two rhythms severely mismatched.
Regarding the Supreme Court ruling that declared Trump’s tariffs illegal, Chey Tae-won said he would comment after reviewing the judgment.
Korea-U.S.-Japan Cooperation Will Decide Future Order
Chey Tae-won emphasized that the current transformation is not just a simple challenge but a structural reality that determines everyone’s survival. This forum, launched in 2021, gathers current and former senior officials, scholars, think tanks, and business leaders from Korea, the U.S., and Japan to discuss international issues.
He stated that how Korea, the U.S., and Japan cooperate at this turning point will determine the future order. This statement reflects the increasing importance of technological alliances and supply chain collaboration amid intensifying competition in AI and the semiconductor industry.
Chey Tae-won’s warning provides a sober perspective for the market: despite the enormous opportunities AI presents, the rapid pace of technological iteration, market volatility, and infrastructure bottlenecks mean even industry leaders face unprecedented uncertainty.
Risk Disclaimer and Liability Waiver
Market risks exist; investment should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their particular circumstances. Invest at your own risk.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Chairman SK, Choi Tae-yoon, warns: AI is consuming everything, and this year's hundred-billion-dollar profits could instantly turn into massive losses
SK Group Chairman Chey Tae-won Warns that Market Optimism Over SK Hynix’s Potential Over $100 Billion Operating Profit This Year Could Turn Into a $100 Billion Loss. He Emphasizes that Artificial Intelligence Is in a Transformation Phase Devouring Everything, and the Volatility Caused by Technological Change Has Reached Unprecedented Levels.
Chey Tae-won stated on the 20th in Washington D.C. at the “Trans-Pacific Dialogue” event that the uncertainty of the AI era has rendered traditional planning cycles meaningless, “A one-year plan is meaningless; the changes between the beginning and end of the year are too great.” He pointed out that companies with ample funds and resources will dominate the AI competition, but even industry leaders find it difficult to determine whether they have the right solutions.
This statement highlights the double-edged sword facing the semiconductor industry amid the AI wave. Despite SK Hynix achieving its best-ever performance last year driven by surging demand for high-bandwidth memory, with a market value soaring to 690 trillion won, Chey Tae-won’s warning reveals the risk that accelerated technological iterations could instantly change market dynamics, posing a significant challenge to the semiconductor industry, which relies on long-term investments.
Chey Tae-won also emphasized that the energy demand of AI data centers is becoming a societal challenge, with the infrastructure for AI data centers in the U.S. alone requiring a $5 trillion investment. He stated that the cooperation among Korea, the U.S., and Japan at this turning point will determine future order.
Profit Expectations Doubling from $50 Billion to $100 Billion
SK Hynix’s performance outlook has changed dramatically in just a few months. Chey Tae-won revealed that market forecasts in December last year predicted over $50 billion in operating profit this year, but the latest expectations have surpassed $100 billion. This figure is approximately 144.85 trillion won.
This surge in expectations is based on SK Hynix’s strong performance last year, with sales of about 79 trillion won and operating profit of 47 trillion won, both hitting record highs. The growth in high-bandwidth memory demand and robust sales of general DRAM were the main drivers, and the company’s market value, along with Samsung Electronics, helped push the Korean KOSPI index into the 5,000-point era.
However, Chey Tae-won remains highly cautious about this seemingly positive outlook. He said, “New technology may be a solution, but it can also make everything vanish,” emphasizing that even the most advanced technological advantages can quickly be lost amid industry upheaval driven by AI.
Market “Distortion” Phenomenon
Chey Tae-won pointed out that the construction of AI infrastructure is causing structural distortions in the semiconductor market. He said companies need to produce “monster chips” like HBM using the most advanced technology, which can bring truly huge profits.
However, abnormal phenomena are occurring in the market. Due to AI infrastructure consuming large amounts of memory supply, the profit margins of general memory are surpassing those of higher-tech HBM products. Chey Tae-won called this phenomenon “distortion,” reflecting that AI demand’s impact on the entire semiconductor supply chain has exceeded normal market rules.
This distortion highlights the strategic dilemma faced by semiconductor companies: on one hand, they need continuous R&D investment in cutting-edge technology; on the other hand, rapid changes in market structure can make the value of technological advantages unpredictable.
Energy Bottlenecks as the Biggest Challenge for AI Development
Chey Tae-won views energy supply as one of the most severe challenges in the AI era. He said, “AI is almost consuming all the energy and electricity we need,” and if the electricity demand is not properly addressed, it will become a huge societal challenge.
Specific data underscores the seriousness of the issue. Chey Tae-won revealed that building a data center requires about $50 billion, and the capacity demand for AI data centers in the U.S. is about 100 gigawatts, with infrastructure investments alone needing $5 trillion, not including energy costs.
He admitted, “I’m not sure if we have a solution,” and stated that the current situation has become one where each data center needs to be matched with a 1-gigawatt power plant. This close coupling of energy and computing power far exceeds traditional industry planning capabilities.
SK’s Integrated Data Center and Power Plant Plans
In response to energy challenges, Chey Tae-won disclosed that SK is preparing a new solution to build AI data centers together with power plants. He said that even after building data centers, many technologies can be developed to make their use more efficient, and the company will focus resources on R&D to enhance competitiveness.
This strategy aligns with the energy infrastructure development focus promoted by the Trump administration. When asked about U.S. investment plans in the energy sector, Chey Tae-won said, “We have been doing that all along,” but emphasized that AI changes occur within months, whereas energy projects typically take five years, making the two rhythms severely mismatched.
Regarding the Supreme Court ruling that declared Trump’s tariffs illegal, Chey Tae-won said he would comment after reviewing the judgment.
Korea-U.S.-Japan Cooperation Will Decide Future Order
Chey Tae-won emphasized that the current transformation is not just a simple challenge but a structural reality that determines everyone’s survival. This forum, launched in 2021, gathers current and former senior officials, scholars, think tanks, and business leaders from Korea, the U.S., and Japan to discuss international issues.
He stated that how Korea, the U.S., and Japan cooperate at this turning point will determine the future order. This statement reflects the increasing importance of technological alliances and supply chain collaboration amid intensifying competition in AI and the semiconductor industry.
Chey Tae-won’s warning provides a sober perspective for the market: despite the enormous opportunities AI presents, the rapid pace of technological iteration, market volatility, and infrastructure bottlenecks mean even industry leaders face unprecedented uncertainty.
Risk Disclaimer and Liability Waiver
Market risks exist; investment should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their particular circumstances. Invest at your own risk.