South Korea's AI Leap Forward

This is a bigger gamble than the Park Chung-hee era.

June 29, 2026, Seoul. President Lee Jae-myung bowed almost ninety degrees in front of the national cameras to two entrepreneurs. He called them "national heroes"—Lee Jae-yong, Chairman of Samsung Electronics, and Chey Tae-won, Chairman of SK Group, both of whom had been imprisoned for legal cases: Lee Jae-yong served 207 days for the "Choi Soon-sil gate" scandal, and Chey Tae-won was sentenced to four years for embezzlement. Today, South Korea has entrusted them with the most important industrial transformation of the next decade.

The name of that press conference was grand—"National Report on the Republic of Korea's Three Major Leap Forward Projects." Samsung and SK Group immediately promised to invest a combined 4,800 trillion won (about 21 trillion RMB) over the next decade, betting on artificial intelligence, semiconductors, and energy infrastructure. Combined with the special law passed by the National Assembly for a $350 billion investment in the U.S., South Korea is essentially betting its entire industrial future.

Seeing the term "Great Leap Forward," many people feel a jolt of recognition. Because half a century ago, South Korea already had one "Great Leap Forward"—Park Chung-hee's heavy and chemical industry declaration. The two leaps, fifty years apart, use the same underlying logic: mobilizing national resources, concentrating them, and betting on an industry track deemed a national lifeline. This logic has written both the glorious chapter of the "Miracle on the Han River" and the painful lesson of the 1997 financial crisis in Korean history. This AI gamble may be an excellent window to observe the "state capitalism" playbook itself.

The Success Paradigm of the National System: Government Selects, Chaebol Act

On January 12, 1973, Park Chung-hee announced the "Heavy and Chemical Industry Declaration" at a New Year's press conference, designating ten industries including steel, petrochemicals, automobiles, machinery, shipbuilding, and electronics as national strategic priorities, and established a Heavy and Chemical Industry Promotion Committee chaired by the Prime Minister. To encourage companies to take risks and invest, the government drastically overhauled the banking system: controlling the central bank, holding stakes in private banks, providing full credit guarantees to key chaebols, and setting interest rates lower than the inflation rate. Samsung, Hyundai, SK, Hanwha, and Lotte all completed their transition from light to heavy industry under this policy tilt.

The effectiveness of this approach is undeniable today. When Park Chung-hee came to power in 1961, South Korea's per capita GDP was only $94, making it a poorer agricultural country than many African nations; by the time he was assassinated in 1979, per capita GDP had risen to $1,772. Landmark companies like POSCO (steel) and Hyundai Heavy Industries were established. South Korea became the world's second-largest shipbuilder after Japan, and the automotive industry achieved domestic production and began mass exports during this period.

In just over a decade, enabling a resource-poor, small-market latecomer to join the ranks of industrialized nations is a rare success story in global economic history. South Korea's world-class competitiveness today in semiconductors, automobiles, shipbuilding, and display panels is all built on the industrial foundation laid during this period.

Today's script framework is strikingly similar. The Lee Jae-myung administration named this round of investment the "Three Major Super Projects"—semiconductors, AI data centers, and Physical AI—all linked by "Project Trinity." The Ministry of Trade, Industry and Energy, the Ministry of Science and ICT, the Ministry of Environment, and the Ministry of Land, Infrastructure and Transport all appeared, promising full support in power supply, water supply, and land. Deputy Prime Minister and Minister of Economy and Finance Koo Yun-cheol put it bluntly: Without the government's strong support commitment, companies might have to choose to invest overseas.

The difference lies in the source of the chips. The money chaebols poured in back then was mainly bank loans guaranteed by the government, relying on "national credit first, corporate follow-up." This time, Samsung and SK's investment confidence comes from real operating profits—Samsung Electronics' Q1 2026 operating profit surged 756% year-over-year, and SK hynix's 2026 projected profit is expected to reach 239 trillion won. Both are "national selection, chaebol expedition"—one is an industrial leap driven by national credit, the other is an industrial bet driven by market profits. Both embody the "Korean model" of government strategic guidance combined with corporate execution capability, only this time, the companies' own cash generation ability is much stronger than before.

Concentrated Dividends: One Supply Chain Supports Half the Nation's Exports

During the Park Chung-hee era, the heavy and chemical industrialization had a relatively decentralized industrial layout: steel went to POSCO, shipbuilding to Hyundai Heavy Industries, automobiles to Hyundai, and chemicals were divided among multiple companies. This AI wave is more focused—Samsung Electronics and SK hynix stand at the core of the global supply chain with a single product: High Bandwidth Memory (HBM).

The numbers are straightforward. The global HBM market is almost split three ways between Samsung, SK hynix, and Micron. SK hynix alone holds over 50% of the global HBM market share, is NVIDIA's exclusive core supplier, and has orders booked through 2029 and 2030. Samsung Group's 2024 sales accounted for about 13% of South Korea's GDP. In June 2026, SK hynix's market capitalization surpassed Samsung Electronics, topping the Korean stock market for the first time since 2000 that Samsung lost the "No. 1 Korean stock" position.

This concentration brings immediate macroeconomic benefits. In June 2026, South Korea's exports exceeded $100 billion in a single month for the first time, setting a historical record, up 60.4% year-over-year. Semiconductor exports alone reached $48M, nearly tripling year-over-year.

Behind this is South Korean companies' precise positioning in the AI infrastructure wave: Samsung is jointly developing TPU chips with Google; Samsung and SK hynix are collaborating with OpenAI and Oracle on the "Stargate" supercomputing project—during the most intense years of the global AI computing race, South Korea has become one of the most obvious beneficiaries of this wave, leveraging its technological accumulation in memory chips.

This is also the confidence behind such a heavy bet in this investment plan: it's not blind risk-taking, but amplifying on an already proven competitive track.

Of course, high concentration also means the economy is more dependent on a few industries and a few companies. Professor Kim Kwang-seok of Hanyang University's Department of Economics has observed that besides semiconductors, other Korean export sectors are relatively weak. That's why the Lee Jae-myung administration specifically included plans to break the monopoly of the capital region and spread production capacity to areas like Jeolla Province, attempting to allow the concentrated industrial dividends to radiate more widely.

The Flow of Wealth: From Gangnam New Town to Dongtan Apartments

The industrial miracles driven by national power eventually must land in ordinary people's wallets and living spaces. Interestingly, the path of wealth flowing into real estate is also highly similar in these two "Great Leaps."

During the Miracle on the Han River, heavy and chemical industrialization genuinely improved ordinary people's living standards—a 1978 Korean newspaper survey showed that white-collar families with over ten years of employment all had TVs, 90% had refrigerators, over 60% had washing machines, and over 40% had purchased pianos, which was quite impressive among developing countries at the time. But another part of the wealth settled into real estate. To disperse the overly concentrated population and industries in Seoul, the government led the development of the Gangnam New Town. The result, however, was telling: urbanization was too fast, and the new town failed to lower overall housing prices; instead, Gangnam itself grew into the most expensive富人区 in Seoul, laying the earliest seeds for Korea's long-term housing price anxiety.

This time, the transmission of wealth to real estate is faster and more direct. The total stock compensation paid by Korea's 100 largest companies from January to May this year amounted to 2.28 trillion won, 3.3 times that of the same period last year. Performance bonuses for Samsung and SK hynix employees quickly translated into purchasing power: exemplified by the "semiconductor corridor" around chip factories, a 84-square-meter apartment in the Dongtan area of Gyeonggi Province jumped from 998 million won to 1.12 billion won in two weeks; apartment prices in Seoul had risen for 72 consecutive weeks as of mid-June. High-end consumption such as imported cars and luxury watches also surged, with many dealers reporting that nine out of ten customers looking at cars were semiconductor company employees.

The improvement in living standards from this wave of wealth effects is real, but the distribution is uneven. Kim Yong-beom, Head of the Policy Office of the Presidential Office, once mentioned a noteworthy set of numbers: In Q1 2026, South Korea's real GDP growth rate was 3.8%, while the real Gross Domestic Income (GDI) growth rate reached 13.2%—the huge gap between the two indicates that the purchasing power increase from rising chip prices far exceeded the growth in actual output, and this benefit is currently concentrated in a few industries and households, not yet fully permeating the broader economy.

Historical experience shows that this type of excess liquidity often eventually flows into the real estate market. This is, to some extent, a modern version of the same logic as the Gangnam story fifty years ago: the first landing point for the wealth created by industrial prosperity is often an apartment in a prime location.

Faced with rapidly rising housing prices, the Bank of Korea chose to keep its policy rate unchanged at 2.5% and maintained a hawkish stance— This also illustrates from another angle the delicate balance policymakers must strike between "supporting industrial prosperity" and "controlling asset bubbles": tightening monetary policy too much could disproportionately burden small and medium-sized enterprises and ordinary borrowers who haven't yet shared the dividends; loosening too much could further inflate already high housing prices.

The Dual Faces of National Power: Success and Failure Both Stem from It

Any industrial strategy driven by national power inherently has its dividends and risks as two sides of the same coin—the experience of the Park Chung-hee era makes this clear.

In the early stage of heavy and chemical industrialization, government-guaranteed low-interest loans helped chaebols cross funding thresholds that were difficult to overcome by market forces alone, spawning a group of internationally competitive large enterprises. This was arguably the only realistic path for latecomer countries to achieve industrial catch-up at the time. But as this mechanism operated into its later stage, cumulative effects began to show: By 1997, the average debt-to-equity ratio of the top 30 chaebols exceeded 500%, with some individual chaebols surpassing 1000%. When the crisis hit, the fragility of the high-leverage structure was exposed. Over 30k companies went bankrupt, the second-largest chaebol, Daewoo Group, collapsed, and the economic growth rate plummeted from over 8% to -5.13%. This crisis later drove reforms in four major areas—corporate, financial, labor-management, and public institutions—and put the Korean economy back on a more regulated track after the pain.

This AI Great Leap Forward starts from a different point than back then—Samsung and SK are deploying their own operating cash flow this time, not debt supported by credit leverage. This means the risk-bearing entity is clearer. If the industry experiences a downturn, the hit will first be absorbed by the companies' own balance sheets, rather than directly transmitting to the entire banking system. This is a clear improvement in Korea's industrial policy compared to fifty years ago.

But several structural challenges inherent in the "national power" model itself have not disappeared with time; they have merely changed form:

Whether the scale is sustainable. The total investment commitment of the two groups is equivalent to a significant portion of Korea's annual GDP. If the U.S. investment agreement and others are included, the scale is even larger. Whether such a massive investment commitment can be sustained over the next five to ten years, or even twenty, depends on a combination of factors like the economic cycle and corporate profitability.

The laws of industry cycles still apply. The memory chip industry has historically had clear cyclicality: demand rises, prices increase, industry-wide expansion, overcapacity, prices fall, and the cycle repeats. SK hynix itself has already proactively slowed the expansion pace of HBM4 in Q2 2026, shifting some capacity back to general DRAM, preparing in advance for the potential concentrated capacity release around 2028. This proactive posture, to some extent, indicates that companies have a clear awareness of cyclical risks, and it is also a more prudent aspect of this investment compared to the past.

The interconnectedness from deep supply chain integration. South Korea's AI strategy this time chose to deeply embed itself in the U.S.-centered AI ecosystem. This has given South Korea an irreplaceable position in the Sino-U.S. tech competition landscape, but also introduces new interconnected risks—once global supply chain dynamics or industrial policies between major powers change, Korean companies at the core node of the supply chain will feel the external environment changes more directly.

The distribution rhythm of industrial dividends usually lags behind the pace of industrial development itself. For every 1 billion won of value created by the Korean semiconductor industry, the number of jobs generated is only one-fifth of the industry average. SK hynix employees' year-end bonuses once reached 3,264% of base salary. But this wage flexibility is currently concentrated among employees of a few core companies.

This is not unique to South Korea—high-tech, capital-intensive industries generally exhibit this. How to spread dividends more evenly is a common challenge for economies betting on a single strategic industry in the mid-to-late stage. Kim Yong-beom of the Presidential Office has proposed using excess tax revenue from the AI boom for broader public welfare—the emergence of such discussions shows that Korean society is already thinking ahead, rather than waiting for a crisis to strike before remedying.

History Does Not Simply Repeat, But the Nature of the Challenges Is Universal

When placing these two "Great Leaps" side by side, the biggest commonality is not specific numbers but the logic itself: a resource-limited economy chooses to concentrate national power and resources into a single track deemed a future lifeline, entrusting the nation's industrial fate to a few super-enterprises with execution capability. The advantage of this model lies in high decision-making efficiency and rapid resource allocation, helping latecomer economies achieve catch-up or even surpass during critical windows. The rise of South Korea's semiconductor, shipbuilding, and automotive industries are all fruits of this logic.

Park Chung-hee relied on administrative power and forced allocation of bank credit; Lee Jae-myung relies on voluntary reinvestment of corporate excess profits—from a "blood transfusion" model to a "self-sustaining and repaying" model, this round of investment has a smaller risk exposure and stronger resilience than before. But self-sustaining ability also has cyclical ceilings: Samsung and SK can invest 21 trillion won in the future only because HBM demand has experienced a historic explosion in recent years. Once the global AI infrastructure boom cools and memory chip prices correct—as has happened multiple times historically—whether this ten-year blueprint, propped up by current excess profits, can proceed as planned will require time to answer.

This may be the true meaning of "Success and failure both stem from it" in the context of industrial policy: the mechanism capable of concentrating resources to accomplish great things is itself a double-edged sword. During the climbing phase, it can create speed unattainable by spontaneous market forces; in the mid-to-late stage, the institutional arrangements supporting prosperity also need to be adjusted in step with the times to digest structural issues like leverage, concentration, and distribution fairness. The lesson of 1997 is not that heavy and chemical industrialization was wrong, but that the credit mechanism failed to transition in time from "expansion during the catch-up phase" to "regulation during the maturity phase."

Policy Implications: If We Must Go All In Nationally, What Lessons Need to Be Applied?

Since the path of using national power to bet on strategic industries will likely be chosen repeatedly by more economies, a more practical question is: Drawing on the experiences and lessons from these two Korean "Great Leaps," what should be done to maximize the chances of keeping the "success" and minimizing the probability of "failure"? Based on the clues from the previous sections, at least four directions deserve attention.

First, where the money comes from, the risks go. The lesson from the Park Chung-hee era is direct: government-guaranteed credit made companies bold to invest, but the risks eventually settled in the banking system. Once the economic cycle reversed, systemic risk was almost indigestible. This time, Samsung and SK are funding from their own operating profits, which is already a more prudent arrangement. However, if future investment scales exceed their ability to generate cash flow, and they revert to massive debt reliance, the same hidden dangers will re-emerge. Meanwhile, the laws of industry cycles have never disappeared—the ups and downs of the memory chip industry have been repeatedly proven. SK hynix's proactive slowdown of HBM4 expansion and shifting some capacity back to general DRAM is essentially leaving a buffer for the potential overcapacity around 2028. The sooner funding shifts from "credit dependence" to "profit self-sustenance," and the sooner capacity planning reserves buffers for cyclical reversals, the stronger the resilience.

Second, don't put all eggs in one basket—domestically and internationally. In the heavy and chemical industrialization era, steel, shipbuilding, automobiles, and chemicals developed separately, allowing the overall economy room to maneuver when one sector encountered a cold spell. This time, the bet is more concentrated on the semiconductor and HBM chain. The Lee Jae-myung administration's investment blueprint includes multiple directions like physical AI and energy infrastructure, and spreading capacity to areas like Jeolla Province—this is the right direction. Meanwhile, this AI strategy chose to deeply embed itself in the U.S.-centered AI ecosystem, which gives South Korea an irreplaceable position, but also means that once global supply chains or major powers' industrial policies adjust, South Korea will feel external shocks more directly. Diversification of industry portfolios, coupled with moderate diversification of market and technology routes, is essentially the same risk hedging logic—both are necessary.

Third, start discussing distribution while prosperity is still ongoing, don't wait for problems to emerge. The number of jobs created per unit of output in the Korean semiconductor industry is far below the industry average, which naturally makes industrial dividends tend to concentrate in a few core companies and a few engineer groups. Discussions by Kim Yong-beom about using excess tax for public welfare, and Samsung's union negotiations on profit sharing, are all addressing distribution issues while the boom is still ongoing—this is much cheaper than waiting until a crisis or social discontent accumulates to a tipping point.

Fourth, industrial policy and housing regulation must be discussed at the same table. Dongtan apartments skyrocketing in two weeks, Seoul housing prices rising for 72 consecutive weeks—this shows that liquidity from industrial prosperity naturally flows into real estate, the easiest reservoir. The Bank of Korea's dilemma between inflation pressure and asset bubbles highlights an easily overlooked principle: while using national power to develop strategic industries, monetary policy and housing regulation must keep pace. Otherwise, industrial gains can easily be diluted by housing bubbles, even pushing up labor and living costs for the industries themselves.

These four points are all different aspects of the same principle: The power of a national system is almost irreplaceable during the industrial climbing phase, but whether this power can carry all the way to the end depends on whether the designers have prepared for the day when the boom recedes, before the boom truly arrives.

Half a century ago, South Korea used a "Heavy and Chemical Industry Declaration" to bring an agricultural country onto the fast track of industrialization, creating the "Miracle on the Han River," but also experiencing the profound crisis of 1997 in the latter half of that process. Today, South Korea is using another "Three Major Super Projects" to bet its national industrial fate, even more concentrated, on the semiconductor and AI supply chain.

This time, the "萧何" (the factor driving success) supporting the investment is the companies' own accumulated operating profits, not government-guaranteed bank credit— This itself is a genuine institutional evolution and a result of Korea's industrial policy reforms over the past thirty years.

Whether this ambition can successfully run through the next decade, the answer may not lie in the size of the investment itself. But rather, whether South Korea can, before this round of prosperity truly peaks, address the issues of leverage, industry cycles, and distribution fairness more calmly than it did fifty years ago. This is not only a test for South Korea but also a sample worth observing for all economies attempting the "national power to develop strategic industries" path.

Source: Chen Lili Chen

Risk Warning and Disclaimer

Market risk exists; invest with caution. This article does not constitute personal investment advice and does not consider the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment based on this is at your own risk.