TRANSCRIPT

We cannot know how fundamental and transformative generative AI will be in the future in defining international relations, the political systems, inter-Korean relations, and the future prospect of unification.

The first question is, while we try to build a very much coherent and developed sovereign state according to the modern way of organizing international relations, we have, in part, failed because we are divided. We are an incomplete sovereign state, and we have to reunify the Korean Peninsula.

Q. What led to a divided Korean Peninsula?

Why did we end up with this kind of situation?

Because we failed to develop and meet the civilizational standard from the 19th century, up to the beginning of the 19th century, China was the first nation-state in the world in terms of GDP. But because of the failure of the Industrial Revolution, all the states in non-Western regions were destined to fall into colonies or semi-colonies. After that, we have been trying to build our own successful nation-states in these regions, but we are still struggling.

Looking back on history, we can say that the development of technology and meeting the civilization standards are very important. But before we complete the modern type of nation-state, there is another trend coming from the future, or probably already in the present, which is the Fourth Industrial Revolution.

So if we think about the future trends or challenges, we have to deal with this kind of new revolution. To do that, we have to review the relationship between technological development and changes in state form in terms of knowledge production.

In the second part, I divided the development of the state into three stages: from traditional state to network-knowledge state, which basically correspond to the era of Third Industrial Revolution and the digitization of technology.

Q. What will future state development look like?

In the foreseeable future, the next type of state will be the AI intelligent state.

So we have built, like China and Japan, a very traditional form of a state, which is called a “Tianxia state. We had this traditional state and then we tried to transform ourselves into an early modern type of state, which is the “modern state.” That modern state was based on a very effective control of knowledge, such as statistical knowledge about its own population. Western European countries, from the 17th century, tried to build a knowledge state, which corresponds to the early modern type of technological development. After a long historical turn, we had a so-called informational turn in the late 20th century, which corresponds to the period of post-Cold War. Back then, we had a lot of jargons such as soft power, knowledge power, and network power.

Based on the success of modern state, now we try to build a knowledge-network state. There are so many information and knowledge in the information space such as Internet and digitized ways of data. So there was, in a broader sense, a digital revolution from the late 20th century.

All this was combined with a very modern form of state. Max Weber and Michel Foucault tried to have an idea of rationalization of state bureaucracy and expanding the state power into a very micro type of biopower.

Now, states have a very precise knowledge of its population and try to exchange information and knowledge with other states, building a global public sphere in which there are very lively conversations and exchanges between states or civil societies in different states.

Q. How will AI change states?

In a broader form, the development of AI, or generative AI, is inside that digital revolution, but it's much more fundamental in the sense that AI power is enabling technology, or meta-technology, that augments the effectiveness of other technologies. Some theorists compare AI with the invention of electricity, or even fire, so it's very much a meta-enabling technology. So we don't know how transformative of that technology will be.

And AI has gone through several stages: first the computational, computer stage and then artificial intelligence and now we have from the end of 2022, generative AI, which is really transformative.

They produce knowledge. They try to be, not just automated, but also very autonomous in building a new type of decision making process.

Maybe in the future, we will have artificial intelligence that humans cannot control. So we are in the middle of a very rapidly developing roadmap of having a new type of artificial intelligence which will have grave impact on our society. It's very hard to predict the impact this generative artificial intelligence have on our society.

Q. What impact could AI have on the military and economy?

We are just beginning to explore a various type of causation here, and then if you look at many sizes and military size there are so many discussions going on; for example, AI-enabled nuclear command and control. How should we have an AI-based arms control? For example, if you have an AI-enabled weapon system, but you don't have AI system, can you deter that type of AI-based conventional nuclear attack?

So there should be AI to AI deterrence. If you fail to develop your own AI version of military power, that will be a big problem. In international relations and also inter-Korean relations, if we try to equip ourselves with AI-based weapons so we can monitor the DMZ with AI surveillance system, we can cope with the North Korean aggression based on AI surveillance system, or we can even monitor the human rights situation in North Korea. If North Korea data is digitized, then we have a better idea to use that data to make full use of those data to monitor the situations in North Korea, or to cope with hackers. We can raise hackers to do that.

In economic terms, it has a lot of profitable future. So if we develop our economy based on AI industrial model, the economic gap between inter-Korean relations will widen, so there will be more economic differences between North and South Korea in economic terms as well.

Now with the assistance of knowledge management and data analysis, we can make a much more precise decision, we can have a more precise resource allocation, we can be more transparent. Government will raise accountability more, so in a good sense, we can have a better, much intelligent state and public administration in the future. We all use navigation when we are driving, so if there is a navigation in decision making, from the governmental point of view, AI-powered decision support systems, then we can make a more precise decision which is more farsighted.

Q. How will AI development affect the Korean Peninsula?

In doing so, we can have a more precise simulation, for example, if there is a unification between the two Koreas, what kind of problems will arise? What will be the cost of unification by having this kind of simulation with improved AI technologies? So those models, those analytics will give us quite an edge in doing this state governance and administration.

Last part, we have two models, liberal democracy model of developing generative AI and authoritarian, or even autocratic, model of developing AI. There are downsides of authoritarianism in developing AI and North Korea has been developing AI technology from 1998. We think that North Korea is backward in developing the emerging technology but they clearly have their own concerns and ideas about how to develop their own version of AI. So, they cannot have original technology of having an AI software, or even the semiconductors they can develop, for example, customized way of intervening in South Korean electoral process by developing their own AI based software, which will be very powerful. So if there is an open source AI and a way of using that, then North Korea cannot be excluded in using those technologies.

South Korea also has developed a lot of AI technologies in many areas. So, in the future, authoritarian and autocratic governments have their own strong point. They can process all these developing procedures based on the state decision but they have some deficiencies in doing their own version of innovation because they cannot use the full power from the civil society.

So we don't know how power balance between North and South Korea will change in the future.

South Korea has much more merits in developing AI technology, developing knowledge based state but North Korea may have their own way of developing AI technology, or using AI technology.

In the future, when we think about the situations for reunification, that might change very radically, we will have different ideas about each other, we will have very different identities.

There will be a big difference in state capability to monitor the other part and there might be some military and economic balance of power, which is changing based on the development of AI.

So how to adapt to this fast change determines the future of our prospects for unification.

■ Chaesung CHUN is the Chair of the EAI National Security Research Center and a Professor at the Department of Political Science and International Relations at Seoul National University.

■ Typeset by: Jisoo Park, Research Associate
    For inquiries: 02 2277 1683 (ext. 208) | jspark@eai.or.kr