[10th EAI Academy] Lecture 6: AI, Semiconductors, and Other Advanced Technologies in Future World Politics

YouTube Link: https://www.youtube.com/watch?v=R0nuhmF45OM

■ Bae Young-ja_Professor, Konkuk University.

Video Transcript

Today, I will be lecturing on technology and the future of world politics. In a way, this entire lecture is about the future of world politics for the next generation. But predicting the future is a formidable challenge, especially when we do not fully understand the present. It is also a particularly important question for you. I see that most of you are in your 20s and 30s. What will the world look like when you reach my age, in your early 60s? While it is a fascinating question, we likely will not know the answer until that time comes. We often use the year 2050 as a reference point for such discussions. In any case, while we cannot know the future with certainty, we can explore it from a particular perspective. The goal is to consider what possibilities might emerge. Today, we will focus specifically on how technological change will shape the future of world politics. So, to begin the lecture, the first

story I want to share is a quote you may have heard: "The future is already here – it's just not very evenly distributed." The novelist who said this is William Gibson. He is the person who first coined the term 'cyberspace,' a word you all use frequently today. Cyberspace is a new realm that emerged with the creation of the internet. Before that, could people living in the Joseon Dynasty have known about cyberspace? No matter how brilliant they were, the concept of cyberspace simply could not have existed in their minds.

It is a new space created by this technology, but this was before that technology became widespread. I attended university in the 1980s, and during my undergraduate years, I was part of the generation that submitted reports and assignments handwritten. I believe we were probably the last such generation. We wrote by hand. This was a time when computers were not yet common. However, when I went to study abroad in the early 1990s and finished my dissertation in the late 1990s, the internet grew explosively during that period. So naturally, not only did I write my dissertation on a computer, but I also used the internet for research. From the mid-to-late 1980s through the 1990s, the term 'cyberspace' became widely used. But surprisingly, William Gibson had already used it in his 1984 novel *Neuromancer*. I later found this out and read the novel, though I did not find the story itself particularly interesting.

But he called this machine-mediated space 'cyberspace,' and there are many truly remarkable writers like him. It seems that writers and artists are far superior at capturing and expressing the future much earlier. People like us in academia, who work with academic language, seem to have limitations in terms of imagination. Novelists, artists, and musicians appear to grasp these trends much more quickly. In any case, the future is already here, but what exactly is it that points to this future? Not everything that exists in the present will survive into the future. Some things will endure, and others will fade away. So, among the things that have emerged now, which ones will be the ones that survive?

Technological Development and Changes in World Politics

And as we consider how the world will change as a result, the first thought that comes to mind—and I have structured today's talk around the format of technology and world politics theory, technology and changes in world politics, U.S.-China competition, and technology and the future of world politics—is to ask how technology changed world politics in the past. Since social science cannot conduct experiments, history serves as its laboratory. We say the internet is changing the world today, so we can ask: which technologies changed the world in the past? The most frequently mentioned example is the internet, which I just discussed, and before that,

before that, there was the invention of electricity, and even before that, technologies of the Industrial Revolution like the steam engine, cotton textiles, and the spinning jenny changed the world. How did they change it? How did the Industrial Revolution change the world? It made Great Britain an incredibly wealthy and powerful nation, allowing it to become what was truly the "empire on which the sun never sets"—though its status has since declined. You know that Canada and Australia are still Commonwealth nations, right? They are independent countries but belong to the Commonwealth of Nations. These legacies remain. So, it made one nation immensely powerful. This leads to the question: which nation is artificial intelligence making powerful today? Which country will benefit the most from AI? We can ask these questions. In addition to these major trends, there have been many other technologies that changed the course of history, such as malaria treatments, nuclear weapons, semiconductors, and artificial intelligence.

For example, let's take the malaria treatment I mentioned, quinine. I do not know if any of you have been to Africa, but if you go, you need to get vaccinated. There is some talk that malaria could even appear in Korea due to global warming, but you still need shots to go there. Before the availability of such treatments, many Europeans traveled to Africa by ship. However, they found it difficult to settle. Why? Because they all died.

They would contract endemic diseases. While the indigenous people survived, Europeans would get bitten by a mosquito and die. Therefore, they could only eke out an existence in small areas along the coastline; they could not go further inland to establish colonies. But with the advent of this malaria treatment, people could take it, venture deeper into Africa, settle, build colonies, and go as far down as South Africa. This is the kind of change it created. So, as I said before, this technology could either make a single nation immensely powerful or

it could lead to phenomena like imperialism and the colonization by Western powers, which are major international political events. You might think people went because they had power, guns, and swords. But the technological infrastructure that enabled them to go played a crucial role. For example, the invention of ships that could travel into the African interior made the colonial domination of Africa by white Europeans possible. So, if you think about it a little, you realize that technology has significantly changed world politics throughout history. It is likely changing it a great deal now as well, but what is that technology? Now, if we look at the hegemonic powers in world politics, China was actually dominant before the West. China swept the world in the 14th and 15th centuries, during the Yuan and Ming dynasties.

China was dominant, but then the Industrial Revolution happened, the West rose, and Western powers colonized Africa and South America. This is the history that unfolded. However, if we look at the empires throughout history, from Alexander's empire to the Roman, Persian, and Chinese empires, there have been these great powers. For instance, the Mongols, who founded the Yuan dynasty, established the largest land empire in history. But was their empire built on superior technology compared to others? Not necessarily. Not necessarily.

Historically, that was the case. So when did the game change? It was after the modern era began. If you look at the nations we call hegemonic powers since then, first there was Great Britain. And which country became the hegemon after Britain? Germany and the United States competed, and after World War I and World War II, hegemony eventually passed to the U.S. So, just as the U.S. and Germany once challenged British hegemony, we are now in an era of American hegemony. The U.S. has been the dominant power for about 70 to 80 years, since the end of World War II. Is this order about to change? If I ask this question, which country comes to your mind first? Naturally, you think of China.

You think of China, but let me tell you one surprising thing. In the 1980s and 90s, when I was a student like you, working on my master's and Ph.D., I wrote my master's thesis in 1989 on the topic of 'A Critique of the Theory of American Hegemonic Decline.' In the late 1980s, the academic debate in international relations about the decline of U.S. hegemony was already very heated. It was such a significant debate that I chose it for my master's thesis. At that time, who was the country threatening American hegemony?

Surprisingly, China was still sleeping at that time. It was still dormant on the world stage. So which country were Americans making such a fuss about, claiming their hegemony was under threat? Surprisingly, it was Japan. Japan. At the time, Japanese products like Toyota cars and Sony electronics were everywhere. When I was in school, you carry around iPhones or Galaxys, but back then, the Sony Walkman was the ultimate fashion item for trend-setting university students. Products like these from Panasonic, Sony, and auto-makers swept the globe. They began to penetrate the U.S. market in a major way. So,

Americans were in a panic, saying the American era was over and the era of the Asia-Pacific was dawning, with Japan leading the way. That debate took place about 30 years ago. Looking back now, what happened? Did Japan succeed in its hegemonic challenge? It did not. Today, prices in Japan are so low that it is a top travel destination for you, is it not? Because it is cheaper than Korea. This shows how difficult it is to accurately distinguish between what is a genuine trend and what is an optical illusion at the time. Back then, all the top-tier international relations scholars

who looked at Japan and made those predictions—I do not know what they are doing now, but they should have all quit their jobs. They made such wildly inaccurate claims. Japan's trajectory was broken in the 1990s as it entered its 'Lost Decade.' After all the talk in the mid-to-late 1980s that 'Japan will become the next hegemon,' its rise just collapsed. So, the question is, will China follow the same path? We talk about China's rise and its challenge, but will it also end with people saying, 'It's not ready, it can't make it. How can it succeed with its aging population and that political system?' Or is the other view correct: 'China is different; its scale is unlike Japan's'? This is a simple but very important question for seeing the future. When you try to understand the future of world politics, you must ask: Will China's hegemonic challenge succeed or not? Will American hegemony continue? Will the U.S. still be the hegemonic power in 30 or 40 years? Or will it become a fallen hegemon, like Great Britain?

Historical Context of the Competition for Technological Hegemony

Britain still has the aura of a ruined aristocrat, but it is ruined nonetheless. Will the U.S. follow that path? You have heard lectures from professors who talked about the United States and those who talked about China. You have also heard about Japan. So today, while my topic is closely related to these national-level discussions, I want to focus a bit more on technology. The theory for viewing technology and world politics, as I mentioned a moment ago, is that the country with the world's leading technology becomes the hegemonic power of its time. This common-sense idea was developed into formal theories like hegemonic transition theory and power transition theory by scholars such as Robert Gilpin, A.F.K. Organski, and George Modelski. All of them are

first-rate international relations scholars. They created these theories. In simple terms, the theory is this. George Modelski, in particular, identified cycles in the economy. The 'K-wave' here is not the Korean Wave, or Hallyu. It refers to the Kondratiev waves discussed in economics—cycles of technological innovation. At certain times, a wave of technological innovation is triggered. During the Industrial Revolution, it was cotton textiles and the steam engine. Then, during the Second Industrial Revolution, railways and steel began to be laid down. Next, with the rollout of electricity by Edison and Tesla, the United States began to rise. In fact, Britain led until the railway era, after which the U.S. rose. Then came automobiles and petrochemicals, which further fueled America's rise, until around the 1980s when, as I mentioned earlier,

the internet began to be widely used and the electronics industry emerged. At this time, Americans thought Japan might lead this wave, but after a brief moment in the spotlight, the leadership shifted back to the United States with the rise of Silicon Valley. This was the era from Reagan to George H.W. Bush and Clinton. The period I studied in the U.S. was the Clinton era in the 1990s. This was a very fortunate time; the U.S. at that time, and Silicon Valley, were very lucky. Technological innovation had been underway since the 1970s, but it truly blossomed in the 1990s. Similarly, even with all the current hype around AI, it has not fully blossomed yet. It needs to be more deeply integrated into industry, and we need to see things like autonomous vehicles actually emerge.

It takes time for a technology to emerge, often 10 or 20 years. That is why Silicon Valley's technological innovation brought a complete boom to the U.S. in the 1990s. It was not because Clinton was a great president; rather, the country was lucky in many ways. Because the U.S. was booming at the time, it could offer plenty of scholarships to foreign students. So, my generation was able to study in relative comfort. Now, it is probably difficult for foreign students to get scholarships unless they are exceptional. Because the universities themselves are facing financial difficulties, they are all tightening their budgets. Back then, money was flowing. If AI could spark another revival and bring the U.S. into a boom like that again, we might see a different story. That would be another very important point to watch when looking at the future.

Because it takes time for a technology to emerge, it can take 10 to 20 years. The technological innovation from Silicon Valley in the 1990s brought a full-fledged boom to the United States. This was not because President Bill Clinton was particularly effective, but rather because various factors aligned fortunately. At the time, the U.S. was booming, so it provided many scholarships to foreign international students. My generation was able to study with relative ease, but the situation is different now. Universities are implementing austerity measures, making it difficult to offer scholarships to foreign students. Back then, money was circulating widely. If artificial intelligence (AI) were to spark a similar revival and lead the U.S. into a boom like that of the 1990s, we could face a new phase. This will be a very important point to watch when forecasting the future.

Will AI bring a new boom like in the 1990s and revitalize the United States? Just as the U.S. revived in the 1990s—recovering from the trade deficits of the Reagan administration and getting back on a growth trajectory—can it do so again? The reason the Trump administration implemented tariff policies is because of America's massive trade deficit, fiscal deficit, and debt. Could this situation be reversed by AI-driven growth, or will this time be different?

Just because it happened in the past does not guarantee it will happen now. History does not repeat itself. Will the next-generation information and communication technologies currently driving technological change—such as AI, quantum technology, clean energy, and biotech—help the U.S. revive? Or will China catch up, or will a third country, perhaps Korea, rise? This question offers another lens through which to think about the future. The most prominent technology today is undoubtedly AI. When I was studying in the 1990s, the most prominent technology was the personal computer, and I had some experience studying the CPUs and semiconductor memory chips that played a crucial role in PCs.

AI and Changes in the Military Sphere

At the time, I started just for fun, but I had no idea the technology would become such a boom. Having set foot in that field has allowed me to remain active to this day. As AI has regained prominence, there was much talk about semiconductors for a while, but now AI has become more important, and at its core are semiconductors. We will look at how AI is changing world politics, and the changes in the military sphere are particularly significant. Professor Kim Yang-gyo will cover this part in more detail in the next lecture. Today, as a preview, I will talk about autonomous weapon systems, such as drones, and cognitive warfare. Cognitive warfare refers to intervention in shaping public opinion during an election process. For example, there have been suspicions of Russian interference in U.S. elections, and in Korea, there have also been allegations of Chinese interference in public opinion through hacking or comment manipulation.

These suspicions are not entirely baseless. While it is difficult to precisely measure their impact on election outcomes, there can certainly be attempts to interfere in presidential elections, which have a significant impact on a country's future and policies. However, it is difficult to conclude that such interference directly affects election results. Nevertheless, it is a clear fact that these things are happening constantly. What is important is to filter out this information and, as voters, elect a good president for the future of Korea. Similarly, the United States must also choose its president properly from its own national perspective.

Cognitive warfare, that is, information warfare like fake news, affects our thinking. This is because our brains are wired to react to such information. AI-based data synthesis, information warfare, and command and control systems have also played a crucial role in the Ukraine-Russia war. For example, when Ukraine initially faced communication difficulties due to Russian attacks on its internet infrastructure, it was able to secure internet access through Elon Musk's Starlink. AI technologies like drones are significantly changing the nature of warfare.

Contrary to initial expectations that Ukraine would be unilaterally defeated by Russia, it has been able to withstand the war thanks to the continuous support of AI drones and infrastructure systems. As such, AI is playing a crucial role in changing the game in warfare.

Israel is also one of the countries actively utilizing AI for military purposes. Intelligence agencies like Mossad collect and use information on Hamas operatives to track communications, card transactions, and employment data, and to carry out drone attacks. AI systems such as Lavender are used for tracking and attacks based on this information on Hamas operatives.

It is known that the misfire rate of these AI systems is about 15%, which has drawn criticism that it can lead to harm to innocent people other than the target. While the article highlighted the 15% misfire rate as a problem, others might view it from a different perspective, noting the low misfire rate of 15%. Additionally, the fact that AI systems can harm people other than the target is pointed out as a problem. These technologies will become increasingly sophisticated, and intelligence-gathering capabilities are being used in warfare, as seen in the case of tracking Osama bin Laden.

Palantir's information system utilizes not only card and communication data but also biometric information. In the case of Osama bin Laden, based on information that his family was in a hideout, they collected blood under the pretext of giving children vaccinations, analyzed the DNA, and confirmed his presence. This astonishing intelligence-gathering capability is being used in warfare.

Compared to past wars, the United States and China are striving to secure an AI advantage. The U.S. is promoting initiatives like the Replicator Initiative and Project Maven, while China is responding with concepts like 'intelligentized warfare.' The military is currently the most active field of AI application, as it can determine the outcome of a war.

AI and Economic Growth and the Job Market

AI is being used more extensively than one can imagine in the U.S. Department of Defense and the Chinese military sector. Discussions are underway regarding issues of accountability. In the economic sphere, the main concerns are how much AI will improve productivity and how the job market will change. Economic growth is expected to be driven by AI.

Goldman Sachs anticipates AI-driven growth, but it is important that this growth is distributed equitably. AI is the key engine that will drive growth at a time when growth in other sectors is difficult. In terms of productivity innovation, it is important to feel that you are becoming a more productive person by using AI. You must use AI to become smarter or to acquire the skills society demands.

You must strive to achieve what you want by utilizing AI. For example, AI can be used to improve foreign language skills or acquire knowledge. I currently subscribe to and use AI tools like Claude and Gemini. I use Claude for research and Gemini for daily tasks. I am working with these two AI assistants.

While AI serves as an excellent assistant, users must also make an effort to utilize it effectively. AI requires clear instructions and a well-structured environment; otherwise, it can produce nonsensical results. Even when creating a PowerPoint presentation, there are times when AI upgrades it in an unexpected direction, requiring precise prompts and continuous interaction. This process can sometimes feel tiring. I find myself comparing the results with and without AI, contemplating the pros and cons of its use.

To use AI effectively, the user must also change. It is a process of adapting oneself to make AI work well. Productivity gains do not just happen automatically. Furthermore, changes in the job market are another important issue. Since AI can perform mid-skill jobs and entry-level tasks well, the relative influence of experts whose knowledge is accumulated through long experience and time may diminish. It is difficult for AI to replicate the deep expertise of someone who has worked in a field for 20 years, but the current younger generation may find themselves competing with AI before they have had a chance to build such experience.

AI can analyze vast amounts of data, processing far more information than a human can. This capability can be overwhelming for those in the early stages of their careers. Although the content itself may not be profound, the output generated by AI can appear sophisticated and voluminous. This can be intimidating for collaborators or those just starting their careers. From a company's perspective, using AI may be judged as more efficient.

Reskilling and Individual Adaptation in the AI Era

In this situation, 'reskilling' is crucial. In the past, a university major could sustain a lifelong career, but that is no longer possible. A university major can be an asset, but it does not guarantee a job for life. Therefore, one must constantly acquire new skills and adapt to change.

Reskilling is a lifelong process of skill acquisition. Instead of finding it tiresome, one should embrace it with a sense of challenge. You need to develop your own unique strengths that AI cannot replicate and obtain high-value information that cannot be acquired by asking AI important questions through human relationships. For example, personal information about a specific individual or investment information is difficult for AI to access.

AI can provide general information about a specific person like Professor Bae Young-ja, but it cannot provide personal details such as her personal relationships, food preferences, favorite restaurants, or high-value information like her stock investment status. This kind of information can only be obtained through direct interaction with people. Therefore, there is no need to be too intimidated by AI.

AI excels at processing and presenting the most standard and general information, but it has difficulty finding the deep, messy information that comes from human relationships—the information that is truly important. This information can only be obtained through relationships between people, so there is no need to be overly intimidated by AI.

Widening Gaps and the Global Divide in the AI Era

The more I use AI, the more I think not of being overwhelmed, but of how I must make really good use of it. In any case, it is now transforming the military and the economy. Various comparisons are being discussed in relation to past new technology investments like electricity and the internet. It probably needs no explanation that disparities are widening because of artificial intelligence. The widening gap within a country, and between countries, is a very important issue in international politics. If we look at the period of the Industrial Revolution,

the gap between developed and developing countries was not that large. Internally, the gap between aristocrats and commoners would have been enormous, but at the national level, there was little difference. So when did the gap start to widen? This light blue line is the core, and the light green is the periphery. They started to pull ahead from the 1940s. In other words, the gap between developed and developing countries is also a modern phenomenon.

In the old days, the world was not divided into rich and poor places. Everyone was poor, except for some classes in some advanced countries who were very well-off. The current situation is that the middle class has grown. The biggest beneficiaries of globalization since the Industrial Revolution are the middle classes in countries like Korea. They live a life like past aristocrats. They use clean toilets and drink sanitary water. Statistics from before the Industrial Revolution show an average life expectancy of 25 years. This does not mean people did not live long, but that the infant mortality rate was very high.

Babies did not survive. Just one generation above mine, if you had ten children, you were lucky if half of them survived. Many died young from things like measles. In my generation, it is very exceptional for a sibling to have died. In your generation, it is almost unheard of, although there are still very exceptional cases where a mother has only one or two children. In the past, if you had ten children, maybe three or four would survive. These phenomena began to change after the Industrial Revolution. If I had to pick the single most important factor among the many that have changed human life, from the perspective of someone who studies political economy, it is clearly the Industrial Revolution. This is because many changes, such as the liberation of ordinary people from hunger and improvements in sanitary water and health conditions, began with industrialization. In our country,

when was the Industrial Revolution? The 1970s, from the mid-to-late 1960s through the 1970s and 1980s, was our period of industrialization. During that time, environmental improvements occurred explosively. The generation before and the generation after are completely different. I was born in 1964, so the early 1960s generation received the most benefits from that explosive growth. That system of benefits should have continued to expand for you to enjoy as well, but unfortunately, it did not, and now your generation and the ones below you

are likely to have a harder life than your parents. In my generation's mentality, it was a given that we would always live in a more advanced world than our grandmothers', mothers', and fathers' generations. We never once considered living in a world that had regressed. But you may have to start thinking that way. Your childhood may have been the most affluent time of your life, a time of carefree existence. This is also true in terms of the climate and environment. The weather is becoming increasingly extreme. Environmentally and in many other ways, the world is becoming more uncertain rather than better. We cannot definitively say it has gotten worse. There are still parts that are improving, but it is becoming uncertain. High volatility demands a strong ability to cope. You have to survive the bad times and accumulate a lot during the good times. We are now living through such a period.

This is the global AI divide. There are strong signs that the gap is very likely to widen even more explosively in the AI era. This is an international political problem. There is a limit to the disparity a system can sustain. If it becomes too large, the system is bound to collapse. Therefore, even for the sake of system maintenance, the gap must be reduced, and social mobility must be smooth. This is not about altruism but a condition for the system's own survival. Yet this condition is becoming increasingly precarious. Some argue that imperialistic phenomena will intensify due to algorithms, while a UN report suggests it could open a window of opportunity.

AI and the Potential of Developing Countries

For example, in the era of wired communication, developing countries faced many difficulties, such as having to lay cables under the desert to build a communication network. But with wireless communication, they only need to install base stations, so developing countries benefit more. This is called 'leapfrogging.' In the case of AI, if it is used to address endemic diseases in Africa or improve agricultural productivity, Africa could be transformed. Some people even say that the future of humanity is in Africa. This is because the conditions for economic growth, such as population growth, are uniquely progressing in places like Southeast Asia and Africa, which were previously underdeveloped. These places have lacked money and high levels of education until now, but if AI is introduced, it could solve many problems without much complexity, revitalizing their economies and leading to a dramatic increase in productivity.

There is also talk that these regions will become important sources of dynamism and growth for the world economy. Whenever I hear such stories, I think, 'I really hope so,' but I am not sure if it will actually happen. In any case, regarding who will reap the most benefits from AI, there is an argument for developing countries, but is that really the case? They are already reaping some benefits, but I have an uneasy feeling that the big tech companies are taking almost everything, and the rest are fighting over the remaining 10%. In any case, this debate is ongoing.

The Current State and Outlook of U.S.-China Technology Competition

I will stop here for a moment. We have discussed the general discourse on technology, the general view on the relationship between world politics and technological change, and how artificial intelligence is evolving. We looked back at the past and examined theories to understand the future of world politics, and we also saw how AI is actually changing the military and the economy. Now, if we think again about the idea that the future is already here, will the United States and China still be competing over technology in about 30 years? You are probably aware that they are competing fiercely right now. Even if you are not interested in world politics, you know that the U.S. and China are striving to gain an edge in AI. You are likely familiar with the scenario where the U.S. created AI first, but China is rapidly catching up by imitating it.

How should we view China's pursuit? You might be wondering if its influence is truly growing to the point of surpassing the United States, or what the situation is, and how long this fight will continue. By the time you reach my age, will the winner be decided, with one country, either the U.S. or China, becoming the hegemon, and the other becoming a fallen old power like Britain? That is why we need to look more closely at the U.S.-China conflict over AI. We need to see exactly what they are fighting over and what the nature of that fight is. Looking from a distance is not enough. We need to look closely. I am not an expert in AI either, but I did take apart a PC back when I used one in the past.

I bought a cheap one, so I had to know what the components were. So I took it apart and so on, but with AI these days, the technology is changing so fast that that level of technical knowledge is not enough. At some point, I found myself reading more articles about AI technology than about international politics. I do not know it well, but I have to learn. Because it is an unfamiliar field, I end up reading more about it. I get to meet many AI technicians through various opportunities, and they are the ones who need people like me.

Through my activities on the National Strategy Committee on Artificial Intelligence and other bodies, I came to understand how AI is situated in the overall international political landscape. I also had many opportunities to hear from experts with specialized knowledge. So I am diligently accumulating information on various aspects of this issue. If we look at when AI became a focus and why, the period between 2012 and 2015 was the trigger point. The reason 2015 is significant is because of

China's Technological Rise and the U.S. Response

the "Made in China 2025" initiative. China has pursued reform and opening-up since Deng Xiaoping came to power in the late 1970s. After Mao Zedong's death and the Cultural Revolution, China's economy was in ruins, and it had become a forgotten entity on the world stage. Through its communization in 1949, the Korean War in the 1950s, and the Cultural Revolution in the 1960s, it had degenerated into a very poor, underdeveloped country. However, with Deng Xiaoping's rise in the late 1970s, he advocated for embracing capitalist economics to achieve growth, famously using the "black cat, white cat" theory. To this end, four eastern provinces were opened up to allow Western capital investment and economic development.

The capital that flowed in was almost entirely from Hong Kong or overseas Chinese sources. They built factories for their convenience, and economic growth began. When I was studying for my master's degree in the 1990s, China was not a major topic of instruction. My professors did not consider China a significant concern in world politics. Instead, they taught about Japan.

China did not experience explosive growth immediately after starting its reform and opening-up. It was very slow at first. Then, upon joining the WTO in 2001, its growth became explosive. This was a policy of the Clinton administration. Clinton created the WTO to liberalize world trade and judged it would be beneficial for China to join. Although there was debate within the U.S. about China's accession, the Clinton administration predicted that as China grew, a middle class would emerge. This middle class would then become interested in democratization and the environment, putting pressure on the government. The thinking was that China would then adopt democratic institutions and maintain good relations with the United States. Today, that thinking is heavily criticized as being completely wrong. Some argue that the Clinton administration's decision to let China into the WTO was a mistake, and that China should have been isolated to slow its economic growth. After joining the WTO, China's growth rate surged to 10%. With adjustments to tariffs and other factors, China became the world's manufacturing hub, a feat made possible through access to markets in the U.S. and other Western countries. For the decade from 2001 until Xi Jinping came to power in 2012-2013, China continued its steep growth trajectory.

Growth cannot be sustained indefinitely. It might be possible for a decade, but not for 20 or 30 years. Around 2012-2013, the limits of China's growth model, based on cheap labor, began to appear. The Chinese economy was still growing, but it could no longer exceed the 10% growth rates of the past. It started to fall to the 8% range. While this is still a tremendous growth rate, the rate itself was declining. This began to create a sense of crisis within China. "What will happen to China if this decline continues?" Consequently, voices within China called for moving to the next stage. The argument was that China needed to produce more sophisticated goods based on capital, knowledge, and technology, rather than cheap labor. To make products at the level of those made by South Korea or Japan, technology is essential. Because China still lacks this technology, "Made in China 2025" is an ambitious economic policy for China to become a major producer in advanced manufacturing by 2025. The plan is to domestically produce about 70% of the machinery it currently imports.

This is "Made in China 2025." Announced in 2015, this plan was prepared for one to two years after Xi Jinping took power in 2013. The goal is for China to become a major producer in advanced manufacturing by 2025. It is a very ambitious economic policy that aims to domestically produce about 70% of the machinery China currently imports.

China has set the goal of becoming a major producer in advanced manufacturing by 2025. This ambitious economic policy aims to domestically produce about 70% of the machinery it currently imports. This shows China's determination to move away from a growth model dependent on cheap labor and toward producing high-value-added goods based on capital, knowledge, and technology.

To make products at the level of those made by South Korea or Japan, technology is necessary. Because China still lacks this technology, "Made in China 2025" is an ambitious economic policy for China to become a major producer in advanced manufacturing by 2025. The plan is to domestically produce about 70% of the machinery it currently imports.

U.S. Technology Regulations on China and Their Background

When China announced this, the U.S. president was Barack Obama. President Obama, also a Democrat, had a very positive view of China's growth, believing that as China grew, it would eventually become a democracy. Therefore, his administration initially did not impose any significant checks on China. The Obama administration was no different, though there was a sense of crisis regarding the semiconductor sector. However, the Obama administration finished its term without taking any major actions. Subsequently, starting in 2015, China began pouring massive funds into advanced manufacturing, semiconductors, and other areas. It started to concentrate the money it had earned into these sectors. As the sense of crisis grew, the Trump administration came into office in 2017. At the time, President Obama

seemed to think that China would become a democratic country. Had a Democratic candidate like Hillary Clinton been elected instead of Trump, the situation might have been different. Trump was inaugurated in January 2017. Just before his inauguration, in January 2017, a report on restoring America's semiconductor competitiveness was prepared and handed over to the next administration. The hope was that a Democratic government would take it up and move it forward. However, this plan fell through, and Trump took office in 2017. In his inaugural address, Trump pointed not to technology issues but to the severe trade deficit. As a business operator, he found the deficit unacceptable. A business needs to make money, and continuously running a deficit is a problem.

Looking at where the deficit was coming from, it was mainly the trade deficit with China. This was due to the large influx of low-cost Chinese products into the U.S. market. From a business perspective, the deficit needed to be resolved quickly, but at the national level, this deficit was not a simple issue. The availability of cheap Chinese products helped maintain the standard of living for the American middle class, who could buy inexpensive goods at places like Walmart or dollar stores. When I went to the U.S. in 1991, I was surprised by how many disposable products they used. And the quality of those disposables was very good.

Things like paper towels and party cups. I saw people use these cups once at a party and then throw them away. Having grown up in a time of scarcity, we were very used to a lifestyle of conserving resources: we had to save things, wear clothes sparingly, always turn off lights in empty rooms, and finish all the side dishes in our lunch boxes. For us, it was natural to eat all the food in front of us, wear clothes for a long time, and always conserve electricity and water. But when I went to the U.S.,

I noticed that even supermarkets kept their lights on all night. When I asked why, I was told it was more economical to keep the lights on for security reasons. It was better to keep the lights on and monitor the store than to have a break-in. I also noticed that my American friends often threw away clothes after wearing them just a few times. American washing machines and dryers are particularly harsh on fabrics, especially the dryers. So, seeing them wear something once or twice and then discard it was very difficult for me to understand.

In any case, that American consumer culture was possible because cheap goods from China were supplied to the United States. That's why Americans sometimes do challenges like boycotting Chinese products, but it's practically impossible. In a way, American life is sustained by China's supply of inexpensive goods. However, the Trump administration focused on the massive trade deficit caused by China. Many people around him fueled this sentiment.

As a result, calls to sanction China grew louder. The Trump administration did not initially approach this as a technology issue but aimed to resolve the trade deficit. In the process of trying to reduce the trade deficit, the focus eventually shifted to the technology sector. At first, they tried to block exports by imposing tariffs, but they saw China developing its technology by making huge investments in areas like semiconductors. And it turned out that most of this technology was copied from the United States. This was the problem for the Trump administration: alongside the trade deficit issue was the fact that China was stealing American technology.

The issue stemmed from the belief that China should pay a fair price for technology instead of taking it for free. This led to the start of export controls. These began in 2017 and 2018. In 2018, in particular, they focused on a company called Huawei. At the time, Huawei was a representative Chinese company that was growing very rapidly. Huawei was originally a telecommunications equipment manufacturer, but it had begun to surpass established giants like Nokia and Ericsson, capturing over 30% of the global market. It was also expanding into the mobile phone business, growing to the point where it was challenging Samsung and Apple. However, it was determined that the products Huawei was making were copies of American or South Korean technology.

So, to sanction Huawei, it was placed on a blacklist, and U.S. companies were banned from supplying it with components. However, a year later, in 2019, Huawei continued to grow. In response, the U.S. government reinvestigated why Huawei was still growing despite the export controls. The investigation revealed that Huawei was using components from Taiwan, South Korea, and Europe instead of American ones.

The U.S. government determined that companies in Taiwan, South Korea, and Europe should also restrict their dealings with Huawei. However, the U.S. government could not directly prohibit a South Korean company like Samsung from doing business with China, as this was an issue between nations. Therefore, the Trump administration introduced a new regulation called the Foreign Direct Product Rule. This rule stipulated that products from foreign companies that used 20% or more of U.S. components or software would be treated the same as products from U.S. companies. This rule did not exist before; the Trump administration created it specifically to sanction Huawei. Through this regulation, they sought to prevent South Korean, Taiwanese, and European companies from doing business with Huawei.

As a result, Huawei, unable to use components from South Korea, Taiwan, and Europe, faced difficulties in mobile phone production. It became difficult to make high-end smartphones that require the latest chips. This was because the companies that produce the latest chips were in South Korea or Taiwan, not the United States. While the U.S. could not subdue Huawei with its domestic regulations alone, once it included South Korean and Taiwanese companies in the scope of the regulations, Huawei was hit hard. Consequently, Huawei withdrew from the mobile phone business for a while, but it is now resuming its operations.

The Biden Administration's Technology Strategy on China

This is because China has developed its own domestically produced chips that, while not top-of-the-line, are serviceable. The Trump administration tried to pressure China through measures like the Foreign Direct Product Rule. Then, the Biden administration took office. President Biden was known to be pro-China, having visited the country several times and met with President Xi Jinping. Therefore, it was predicted that the Trump administration's regulations on China would be relaxed. However, contrary to these predictions, the Biden administration began to contain China more sophisticatedly and forcefully. While the Trump administration focused on the trade deficit and technology theft, the Biden administration took a broader view to identify the root causes of the problem.

The Biden administration began to ponder why the U.S. had fallen behind South Korea and Taiwan in the advanced technology race and how long South Korea and Taiwan would continue to follow America's lead. It identified the reasons why the U.S. lacked state-of-the-art chip manufacturing technology and began investing massive funds into American companies to address this. Companies like Intel and Microsoft received investments, but unlike Microsoft, which is growing in the memory semiconductor sector, Intel did not achieve significant results. In response, the Biden administration requested that Samsung Electronics and TSMC establish factories in the United States.

This was because it was judged that U.S. companies would find it difficult to develop advanced chip production capabilities on their own. However, Samsung Electronics and TSMC were reluctant to build factories in the U.S. This was due to high labor costs, difficulties in securing skilled personnel, and the challenges of maintaining equipment that operates 24 hours a day. In Taiwan or South Korea, there is a culture where employees can be called in to resolve problems even at night in case of an emergency, but there were concerns that productivity could decline in the U.S., where such a culture is not established. Additionally, the fact that production costs in the U.S. are over 30% higher than in Taiwan or South Korea was a burden. Nevertheless, due to the U.S. government's coercive request, Samsung Electronics and TSMC began to consider establishing factories in the U.S. In return, the U.S. government proposed providing subsidies.

These subsidies were provided through the CHIPS Act. However, the situation changed when the Trump administration returned to power. Former President Trump expressed a negative stance on providing subsidies to foreign companies. Both Samsung Electronics and TSMC have already made significant investments, but the payment of subsidies has become uncertain. The Trump administration, championing "America First," has a strong tendency to prioritize its own interests over cooperation with allies. In contrast, the Biden administration judged that cooperation with allies is essential to contain China and worked to strengthen these relationships.

Changes in the Trump Administration's China Policy and the Role of Corporations

However, with the re-emergence of former President Trump, the cooperative relationship with allies is shaking. To overcome his current political crisis, former President Trump is again playing the North Korea card, seeking a blatant escape route by flaunting his friendship with Kim Jong Un or advocating for the reduction of U.S.-South Korea joint military exercises. In this process, South Korea is being used. Former President Trump is very transactional and has a strong tendency to prioritize his own interests. In this situation, figures from U.S. companies, particularly Nvidia's Jensen Huang, are reporting to President Trump that it is difficult to stop China's technological rise. They argue that China has already secured and is using a great deal of technology.

Meanwhile, Samsung and TSMC are building factories in Giheung and Hsinchu, Taiwan, respectively. The U.S. is concerned they might sell their products to China. Although this has been blocked, another concern is that Taiwan could be invaded by China, or that South Korea's relationship with North Korea is uncertain, which could halt semiconductor production. If that were to happen, U.S. companies would bear the full brunt of the losses. Apple, for instance, would be unable to produce iPhones, facing a major blow. Therefore, the U.S. decided it had to attract these companies to its own soil. Companies like Samsung and TSMC did not want to build factories in the U.S., where labor costs are higher and securing a workforce is more difficult. Semiconductor companies need to run their equipment 24/7. In Taiwan or South Korea, there is a culture of immediately rushing to fix equipment that breaks down, even at night. In the U.S., the attitude is more like, "I'll come in tomorrow morning and handle it." Consequently, productivity drops significantly.

The U.S. has long had a strong work-life balance culture, making this system a poor fit. That's why building a factory in the U.S. costs 30% more. Nevertheless, the U.S. almost coercively demanded the construction of factories. In response, the companies asked, "We don't really want to do this, so what can you do for us?" The U.S. government then offered to provide subsidies.

Accordingly, the CHIPS Act was created to provide subsidies. The Biden administration had agreed to pay the subsidies, but the situation changed when Trump became president. Trump's position was, "Why should we give our tax money as subsidies to foreign companies?" Both Samsung and TSMC have made large-scale investments in the U.S., but the payment of subsidies is not fully guaranteed, and they have only received a portion. The policy change stemmed from the intention to secure semiconductor production capabilities within U.S. territory. Furthermore, Trump, chanting "America First," did not prioritize allies.

Realizing that containing China cannot be done by the U.S. alone but requires cooperation with allies, the previous administration made great efforts to persuade them. However, with Trump becoming president again, cooperation with allies has been pushed to the back burner. To escape the predicament he is in, Trump is once again playing the North Korea card. He is rehashing old policies, such as flaunting his friendship with North Korea and reducing U.S.-South Korea joint military exercises, which is a very transparent escape attempt. Unfortunately, South Korea is becoming the victim. Trump is very transactional and focused on self-promotion. Consequently, figures from U.S. companies, particularly Nvidia's Jensen Huang, have reported to President Trump that "just because we block them, it doesn't mean China won't buy semiconductors; they are already purchasing and using them through underground channels."

If China's domestic production capabilities strengthen to the point that the U.S. loses the market, it would be very advantageous for the U.S. if Chinese companies utilize software ecosystems like Nvidia's CUDA. However, moves to exclude China from such ecosystems are not desirable. Therefore, the argument was made that the Chinese market should be opened for the benefit of U.S. companies. This meant that the export of the latest chips, such as the H200, which the Biden administration had blocked, should be allowed. Former President Trump likely found this argument persuasive. Since 20-30% or more of Nvidia's revenue came from China, a decline in sales due to the regulations was inevitable.

This caused a significant drop in the revenue of U.S. companies, and voices calling for its recovery grew louder. Initially, there was a sentiment that companies should participate in export regulations for the sake of national interest. However, while they could endure one or two years of regulations, as the decline in revenue continued and they faced the prospect of losing the Chinese market entirely, companies began to push back. Former President Trump also viewed this issue from a transactional perspective.

Ultimately, the policy shifted to an individual transaction licensing system, allowing exports on a case-by-case basis when the end-use was clearly defined. However, after the summit meeting, as these restrictions were eased, President Xi Jinping encouraged the use of domestic semiconductors instead of American products, effectively banning the purchase of the latest chips like the H200. Chinese companies now find it difficult to openly purchase the latest chips.

Of course, roundabout imports—bypassing third countries through the Asian market or black market, such as importing through Malaysia and then bringing them into China—are still happening. However, these roundabout imports have not increased significantly. Despite the export permits being issued, China's move to strengthen its own technological capabilities continues, and it is showing confidence that it can achieve domestic production without American products.

The Core of Advanced Semiconductor Manufacturing and China's Challenge

Currently, the biggest obstacle for China's semiconductor industry is the EUV (extreme ultraviolet) lithography equipment produced by ASML, which is essential for manufacturing advanced chips. This equipment is known as the most complex and sophisticated machine in the world and is indispensable for manufacturing chips at the 2nm process node and below. While TSMC and Samsung Electronics are major customers, Chinese companies like SMIC also tried to purchase this equipment, but their purchases were restricted by the Trump administration's export regulations. ASML initially resisted but eventually succumbed to U.S. pressure exerted through the Dutch government.

Ultimately, ASML was banned from exporting EUV equipment to China and could only trade in the lower-tier DUV equipment. China is trying to use DUV equipment in a way similar to EUV by connecting two machines, but it is significantly inferior in terms of speed and performance. Therefore, the advanced lithography equipment technology held by ASML has a decisive impact on China's semiconductor and AI chip development.

Even if ASML were to sell advanced equipment to China, it would not immediately lead to an improvement in chip production capabilities without accumulated operational know-how. Samsung Electronics and TSMC have decades of accumulated know-how, which China lacks. Merely acquiring the equipment does not mean that complex, advanced chips can be produced immediately; it would even require ASML's technical support staff to be stationed on-site. While acquiring the equipment alone would not allow China's technological capabilities to surge in the short term, if it were to secure the equipment, it could narrow the technology gap significantly within 5 to 10 years. For now, ASML's equipment sales are restricted.

For China to manufacture cutting-edge semiconductor chips, it needs a combination of ASML's equipment and the decades of know-how accumulated by Samsung Electronics and TSMC. Currently, it faces difficulties because these conditions are not met. However, while not yet at the 2nm process, it has already caught up significantly in technology at the 28nm process level. Although there is a high technological barrier to reaching the 2nm process, China has the potential to overcome it. Currently, the core of advanced semiconductor manufacturing technology lies in the Netherlands, South Korea, and Taiwan, and the direct influence of the United States is limited.

China's AI Technology Development and Competitiveness

In response, China has created a large-scale fund for its own semiconductor production. China is leveraging its abundant financial resources and human capital to foster its semiconductor industry. It is utilizing the capital accumulated from trade surpluses and the top talent from its 1.3 billion population. Although there are difficulties in catching up with technology in the short term, it is steadily advancing through the production of chips at the 28nm level, a mature process technology, and projects like the Big Fund Phase 3 in 2024. However, it is still classified as a second-tier country in the most advanced technology sectors. In the AI field, achievements like DeepSeek, announced in January 2025, are emerging.

Chinese companies competing with leading firms like OpenAI, Google, and Meta—such as Baidu, Tencent, Alibaba, and Huawei—are showing remarkable achievements in AI model development. In particular, DeepSeek surprised U.S. companies by introducing a cheap and efficient AI model without using H200 chips. This was a major shock because Chinese companies achieved a level of performance that U.S. companies could only reach with massive capital investment, but at a much lower cost. This achievement shows that China's AI technology is better than expected and has raised questions about the effectiveness of U.S. export control policies.

Furthermore, Huawei is continuously demonstrating its technological prowess by releasing high-performance processors like its self-developed Ascend chip. In July of this year, a world artificial intelligence conference was held, hosted by President Xi Jinping, where he proposed the construction of a Digital Silk Road and an AI Silk Road, particularly targeting developing countries. This shows China's intention to provide its AI technology and infrastructure cheaply and efficiently.

Chinese models are released in an open-source format, allowing for cost-effective use. This contrasts with the technology of U.S. companies like OpenAI or Anthropic, which is difficult to release publicly due to their business models. While China releases its models and encourages their use, U.S. companies keep their technology proprietary. This open approach is a factor that is even leading Silicon Valley companies to use Chinese models.

In South Korea, there are also discussions about the potential use of models like DeepSeek. Even if not applied to personal computers, there is a demand for small, efficient models to achieve specific purposes, and China is leading in these areas. This presents an approach different from the American style of AI development, which is centered on large-scale investment and vast computing power.

China's rise in the AI field coincides with the spread of AI awareness in East Asia after the 2016 Go match between AlphaGo and Lee Sedol 9-dan. After the "AlphaGo shock," AI technology advanced further with the emergence of OpenAI in 2023. While AlphaGo demonstrated capabilities limited to Go, OpenAI presented the possibility of artificial general intelligence (AGI) capable of performing various tasks such as drawing, composing music, and learning, delivering a second shock.

From 2016 to 2024, the United States held an overwhelming lead in AI technology, but starting in 2025, achievements like China's DeepSeek began to create cracks in this dominance. While the U.S. is still in the lead, it suggests that China is also securing competitiveness through its own technological development.

Following DeepSeek, various other Chinese AI models have emerged, such as Moonshot AI's Kimi 3 and Alibaba's Qwen. Kimi 3, in particular, was unveiled at the World Artificial Intelligence Conference in Shanghai and was evaluated as not being inferior in performance to top-tier U.S. models. The fact that China developed such technology amidst U.S. export controls is noteworthy, but there is also controversy surrounding it.

A positive view highly evaluates the speed of China's AI technology development, its efficiency, price competitiveness, and ecosystem-building capabilities. On the other hand, a critical view raises suspicions of "imitation," alleging that Chinese models were developed by copying the data and technology of U.S. models, and points out that they still have limitations in terms of performance.

Watermarking technology to trace the data sources of AI models is becoming increasingly important. Companies like Anthropic are attempting to apply watermarks to their model outputs to determine how much of their technology is being used in Chinese models. Despite these technical efforts, the fact that China has developed advanced AI models with its own technology is very significant and shows that China's technological capabilities should not be underestimated. It is also noteworthy that a significant portion of U.S. IT companies rely on talent of Chinese descent.

The core workforce in the U.S. IT industry is largely composed of foreign-born individuals, including those of Chinese, Indian, and Middle Eastern descent, while the number of native-born American personnel is relatively small. Recently, with the strengthening of regulations against China, it appears that talent of Indian descent is filling a significant portion of the vacancies left by Chinese personnel. This suggests that the U.S. IT industry is heavily dependent on attracting foreign talent.

The United States is regulating the influx of Chinese talent and strengthening investigations into students and researchers of Chinese descent studying at American universities. This has led to some Chinese professors returning to China. This situation could make it difficult for the U.S. to maintain its talent pool and could affect the competitiveness of its IT industry.

Questions are being raised. In the field of AI, the U.S. has set a goal to create the latest and best technology by seizing the AGI frontier. This is the Genesis Mission, similar to the Manhattan Project used for nuclear weapons development, a policy aimed at strengthening scientific competitiveness by integrating all of America's scientific data to build a platform. If the U.S. is pursuing a policy of dominating cutting-edge technology, it is showing its intention to share it only with its allies. In contrast, China's keyword is "AI Plus."

China's position is that it is acceptable not to have the absolute best technology. Instead, it is focusing on how to use artificial intelligence to improve manufacturing productivity and solve educational problems. In other words, a practical level of AI utilization is sufficient. While the U.S. concentrates on capturing the top technology, China, while also valuing top technology, intends to focus on technologies that can be directly applied to daily life. It is developing models that can be used in various fields such as consumption, public welfare, governance, and industry, aiming to use them first and pursuing a policy of disseminating the technology in cooperation with countries in the Global South.

This Chinese policy is wiser and more likely to produce practical results. While competition for cutting-edge technology is important, what ultimately matters is how technology is used to increase productivity. China is concentrating on developing practical technologies with the confidence that "it doesn't have to be first-class technology." Of course, it is also making efforts in that direction, but there are difficulties. The semiconductor and AI ecosystems are partially shared but also partially decoupling. The area the U.S. will never concede is cutting-edge chips, and China has already secured 28nm process technology. Furthermore, it is important that China has secured key minerals essential for semiconductor production. Semiconductor production requires special minerals, and a significant portion of these minerals are processed in China.

Although these minerals are also found in the United States, environmental regulations are strict due to the severe pollution caused by mining and processing. As a result, the cost of mineral production in the U.S. is more than 10 times higher than in China. In contrast, China prioritizes economic benefits, often ignoring environmental regulations to focus on production.

China produces minerals in this way and supplies them to the rest of the world. Documentaries and other sources reveal the severe environmental problems at mineral mining and processing sites. Because China controls key minerals, it is difficult for the U.S. to effectively counter a Chinese offensive that uses them as a weapon. Key minerals are an important lever for China, a weapon that contrasts with the cutting-edge logic chips or technologies like Nvidia's that the U.S. possesses. In the field of AI open-source models, the U.S. is developing frontier technologies, but China is also developing open-weight models, and research cooperation in academic and talent exchanges is also being restricted.

The Decoupling of Technology and Capital and the Mitos Incident

There have also been changes in capital investment. In the past, U.S. funds made large-scale investments in the development of China's semiconductor and AI sectors. This was not to aid China's technological advancement but was driven by the expectation of investment returns. However, concerned about the accompanying transfer of technology, the United States is now restricting investment in China by American companies through regulations. Specifically, it is prohibiting investment in China's AI sector through outbound investment regulations. In this way, capital investment is also being decoupled, leading to a phenomenon of technological and capital separation. This is sometimes referred to as "selective asymmetric decoupling." A particularly noteworthy event in 2023 has been the Anthropic Mythos incident.

In March of this year, Anthropic developed a model called Mythos, which outperformed its previous model, Claude. This new model was unveiled via a memo, and its performance was outstanding. However, when the Mythos model was downloaded and used, a vulnerability was discovered that allowed it to penetrate and attempt to hack other systems.

The model possessed a capability similar to a zero-day vulnerability, allowing it to penetrate systems with existing security measures, identify weaknesses, and execute hacks. Consequently, developers grew concerned that if the model were released, it could be exploited for malicious hacking attacks. They felt threatened by the possibility that they themselves could be hacked. Therefore, they withheld the model's release and took measures to prevent a collapse of security frameworks.

In the wake of this incident, "Project Glass Wing" was launched. This project granted access exclusively to major tech companies and 50 partner institutions to study the security vulnerabilities of the Mythos model and develop security patches to address them. It was a precautionary measure against the security threats that could arise if the model were to be publicly released.

South Korean companies also participated in this project, contributing to the development of security patches. As a result of this incident, Anthropic strengthened its cooperative relationship with the U.S. Department of Defense. Previously, the relationship between Anthropic and the Pentagon had been poor, but the Mythos incident made cooperation unavoidable. While this is what has been reported in the media, the event served as a turning point that redefined the relationship between Anthropic and the U.S. Department of Defense.

It is questionable whether they focused solely on building their own defensive systems. The possibility that this model was used to hack the cyber systems of other countries, such as China and Russia, cannot be ruled out. Hacking countries with limited internet access like North Korea would be difficult, but countries where internet use is widespread, such as South Korea or the United States, could be vulnerable. This asymmetry exists.

China, too, would have recognized this fact and, in great alarm, strengthened its security measures. It is highly likely that government agencies entered a state of emergency after this incident. Hacking and being hacked is nothing new, but this case is significant because it marks the emergence of a technology that can demonstrate hacking capabilities so openly. The importance of this incident suggests that regulations must move beyond hardware chips to prevent AI models like Anthropic's Mythos from being released carelessly.

Furthermore, it is highly significant that this situation was led not by the U.S. Department of Defense but by a private company, Anthropic. For a private entity to take the lead on an issue directly tied to national security, such as cybersecurity, carries great meaning. It also highlights the absence of norms to regulate AI technology, which is innovating at a rapid pace. Although this incident was reported sporadically in the media, awareness of Mythos's potential began to spread in March, leading to active discussions from late April. By June, as the Mythos model began to be more widely distributed, related articles started to appear. In reality, however, intense discussions must have been taking place behind the scenes since March. This event is assessed as one that will fundamentally change the paradigm of cybersecurity in the AI era, especially since even more advanced technologies are expected to emerge after Mythos.

Cybersecurity and the Role of Private Companies in the AI Era

This is like a battle between sword and shield. In cybersecurity, as weapons become more powerful, shields must also become stronger, leading to an endless arms race to create weapons that can pierce those shields. The problem is that this race is now led by private companies, with governments playing a supporting role. This shift is a fascinating development. Technological competition has become a crucial element in world politics.

The Complexity of Technological Competition and State-Corporate Interests

In the past, technological competition was led by states, but today, the critical nature of dual-use technologies and the role of corporations have become paramount. The interests of corporations and the state do not always align, which further complicates the situation, just as the interests of Samsung Electronics do not necessarily align with those of the South Korean government. The current U.S.-China technological competition has many similarities to the period between World War I and World War II.

When I received my Ph.D. in 1998 and presented a paper at the Korean Political Science Association, people found my topic strange and showed little interest. Initially, I couldn't find a university position, so I worked for four years in international cooperation at the Science and Technology Policy Institute (STEPI). After publishing more papers, I moved to Konkuk University in 2004, but the field was still not considered important. So, I continued my research alone.

The Persistence and Importance of Technological Competition

About a decade after I began my research, the field suddenly started gaining attention following the Trump administration. Starting in 2017, I was inundated with requests for lectures from all over. Although I couldn't call myself an expert, it was the result of having studied the topic steadily for ten years. Now, I am so tired of it that I wish the technological competition would subside a little. The war in Ukraine and the Hamas crisis occurred, but the tech competition has not diminished. We must now accept the fact that this competition will not disappear. Technological competition has become a permanent fixture of the global stage.

Crisis and Uncertainty in World Politics in the AI Era

Technological competition has become critically important in world politics, and corporations have emerged as major actors on the global stage. The status of companies has been elevated, with presidents attending meetings accompanied by corporate leaders. Furthermore, there is a deepening crisis of values, norms, democracy, and governance. This is a global phenomenon, not limited to South Korea. Democracy is threatened by problems like fake news, while challenges arise from the absence of norms and growing inequality, such as the AI divide. These factors are interacting to increase uncertainty and volatility in world politics. The question, then, is how long can the United States check China's technological pursuit?

Reports from within the United States seem to indicate that the U.S. itself acknowledges it cannot stop China's technological pursuit forever. Rather, the intention appears to be to buy time—delaying China for five or ten years to prepare. It does not seem to believe it can stay ahead of China indefinitely. While America's technological superiority is expected to continue for the time being, the pace of China's semiconductor technology development is extremely fast, despite U.S. containment efforts. Technologies like Kilmi 3 and Deep-K are expected to continue emerging, and these achievements are rapidly narrowing the technology gap. The U.S. is still ahead, but the gap is constantly closing. Therefore, the U.S.-China conflict will only intensify. Because the two countries have different goals, points of conflict may arise even as both push toward frontier technologies. The U.S. is focused on frontier technologies, while China is focused on application, and we will have to see what changes this difference brings.

In the past, technological advancements like the development of rockets or nuclear weapons were led by states, but today, the critical nature of dual-use technologies and the role of corporations have become paramount. The interests of corporations and the state do not always align, which further complicates the situation, just as the interests of Samsung Electronics do not necessarily align with those of the South Korean government. The current U.S.-China technological competition has many similarities to the period between World War I and World War II.

Historically, there was a time when Germany and the United States were challenging British hegemony. At the time, the British regarded the U.S. as a vulgar country and looked down on it. Americans themselves saw their country as merely nouveau riche compared to the lofty and superior culture of Europe. This led to many marriages between the daughters of wealthy Americans and fallen British aristocrats, a theme explored in many novels from that period.

They needed each other. The British aristocracy had prestige but no money, while the American nouveau riche had plenty of money but lacked social refinement. So, the two came together in marriage. It may be hard to understand now, but at the time, America was an unsophisticated country. How do we think of China today? Frankly, we do not yet see China as a sophisticated or culturally advanced country. But this is what changes.

China's Changing Status and the Importance of the U.S.-China Relationship

Perhaps our generation will be the only one that looks down on China. The history of the 20th century was an exceptional period of weakness for China. In the future, China's status will change dramatically. Whether or not it becomes a hegemon, it is bound to establish itself more firmly at the center of the world and exert an overwhelming influence on us. If China's history for the past 200 years, since the Opium Wars, was a period of humiliation, then China is now in the early stages of normalization, returning to its original place. We know very well how China treated us in the past. That is why the United States is a crucial card for checking China. While we also face the task of maintaining very good relations with China,

When I was at the research institute in the late 1990s, I met with high-level Chinese officials and heads of research institutes. But now, China's sense of pride has grown so much that people at the professor level cannot meet such figures; our status has changed to the point where we are only met by lower-level officials.

Questions About the Sustainability of China's Growth

The level of people I can meet in the international community is gradually declining. This is because China is such a dynamically growing country. The problem, however, is whether this dynamic growth can be sustained. Historically, there are no cases of dictatorial or totalitarian states maintaining market dynamism and growing for a long period. This has always involved a process of democratization; our country, too, went through a democratization crisis in the 1980s and achieved growth through pro-market deregulation. But in China, the state still controls the market. In this context, China is conducting a world-historical experiment to see if the vitality of economic growth can be maintained. We cannot definitively say this experiment will fail. We have to wait and see, as China may create its own unique model. But what is happening now is that,

the population is aging, and China's growth momentum is also slowing rapidly. If the period from 2001 to 2012 was one of explosive growth, we are now a decade past that. Technologically, it is still growing, but as economic bubbles burst, China is not what it used to be. Of course, a growth rate of 4-6% is still very high compared to other countries, but these problems exist. Therefore, we must look closely at the factors that cast doubt on its sustained, dynamic growth. What is important is that as the U.S.-China conflict intensifies, the costs become higher.

The Costs of Deglobalization and the Need for U.S.-China Cooperation

The advantage of globalization was that we could buy things cheaply. But now, with tariffs and other measures, prices are rising significantly. Rising import prices are a major cost of the unraveling of globalization. The world is becoming bifurcated with decoupling, and alliances are becoming important. Furthermore, issues like climate change, artificial intelligence, and pandemics are not matters of U.S.-China conflict but problems where the U.S. and China must join forces. Joint efforts on these issues are not materializing because the logic of competition takes precedence, hindering the shared efforts needed at the level of humanity and civilization. So, when professors are asked who will be the winner,

the U.S. or China? I jokingly reply that at this rate, both the U.S. and China will be ruined, and artificial intelligence will be the ultimate winner, leaving only machines behind. In any case, cooperation is failing to materialize in areas where U.S.-China collaboration is essential. Will we face catastrophe through intense competition, or will a global community emerge to respond to crises like climate change and artificial intelligence? This is a flight of fancy; the reality is a situation where competition and cooperation coexist. Don't you feel the grave threat?

Common Threats to Humanity and the Role of Middle Power Diplomacy

The same goes for climate change. There is likely no one who does not feel the threat. However, that does not mean the U.S. and China will suddenly step up to cooperate. We must look at both aspects simultaneously. The threat is real, but nations do not abandon their self-interest. In this state, only very partial cooperation is taking place. The minimum norm we should share is to 'avoid catastrophe.' We must do something to prevent climate change from destroying human civilization or artificial intelligence from conquering humanity. U.S.-China cooperation is crucial for this, but there may also be a part for us to play. This is why we talk about middle power diplomacy. It's not as if things will change just because we say something to the U.S. or China, but we are in a situation where we have no choice but to find and do our part. The UN is also making various efforts, but it has not been very effective.

The Challenges of Technological Competition and Cooperation in the AI Era

Artificial intelligence technology will be a crucial factor in determining the future course of world politics. This competition will unfold fiercely, but technological superiority alone does not determine hegemony. The competition is not merely about technology but about its application—specifically, who can better leverage it to improve productivity and create systems for a better society. Nevertheless, technological development has not fundamentally changed the basic framework of international politics, which is characterized by an anarchic structure and power competition. While intense competition is underway, we must also work diligently to cooperate on the challenging aspects. The message is that we, as middle powers, must make diplomatic efforts to prevent great powers from overwhelming other nations.