Not an Age of Technology, but an Age of Designing Structure

When I watch the debates around AI these days, it seems like many people still treat it as a technology problem. Questions like which model is smarter, which company is ahead, whether China is shaking things up or America is winning, who is safer.

Of course, those questions have their meaning. But they're closer to the surface. The real substance of what's happening right now isn't a technology race but a structural reordering. More precisely, it's a fight over who holds control.

AI companies talk about danger. They say this can't be done, that has to be restricted, certain lines must not be crossed. States, on the other hand—especially organizations like the military and intelligence agencies—use exactly the opposite language. To them, the danger is not being able to use something when you need it, and the very structure of having to get a private company's permission at a decisive moment is itself a risk.

These two don't clash out of malice. From where each one stands, they're naturally bound to think that way. For the company it's innovation; for the state it's survival. Both are necessary. It's just that when they lock head-on at the same moment, a collision is inevitable.

So the recent conflict between a certain American AI company and a state agency looks like a moral argument, but it's really a power struggle.

On the surface, words like ethics, safety, and responsibility get thrown around, but underneath lies a much colder, more essential sentence.

“If I can't use this as I please, that's dangerous.”

It's a chilling thing to hear, but from the standpoint of an organization like the military or the defense department, it's also perfectly understandable language. What matters to them isn't good technology, but technology they can keep under complete control even in wartime and crisis. For a company, by contrast, what matters isn't just performance, but drawing a line so that its technology isn't used in irreversible directions.

Looking at the same issue, one side sees survival and the other sees catastrophe. So there's no way they'll easily reach agreement.

What's interesting is that the deeper you look into this debate, the harder it becomes to judge who the bad guy is. It would be convenient to split it into who's right and who's wrong, but reality isn't like that.

If anything, this is also why the present still looks relatively healthy. Because at least the attitude of trying to take responsibility from one's own position remains. The company tries to take responsibility for how far its technology should be allowed to go, and the state tries to take responsibility for the survival of its organization and its people.

Conflict is uncomfortable, but a responsible collision is better than an irresponsible agreement. A completely rotten structure doesn't even have these collisions. Because everyone drifts toward what's comfortable, no one draws a line, and no one takes ultimate responsibility.

But when you bring this problem to Korea, the story gets more complicated. In Korea, this kind of massive structural collision doesn't stay an abstract philosophical question. It drops straight down into the household economy. It's immediately connected to realities like housing prices, education costs, working hours, and debt.

That's why, in Korean society, the word responsibility comes up often, but in reality a responsible life is too expensive. To marry responsibly you have to shoulder housing costs; to raise a child responsibly you have to shoulder education and childcare; and even professionally, to make a responsible choice you have to endure long hours and intense competitive pressure.

In this state, “live responsibly” is a fine thing to say, but to many people it can only sound like a cost-of-living bill.

So Korea's core problem isn't, as people often say, a lack of a sense of responsibility. It lies in the fact that a responsible life is structurally a losing proposition.

If taking responsibility doesn't lift you up but instead weighs you down, and if opportunistically slipping out means losing less, then people move not because they lack morals but because they're being rational. They put off marriage, put off having children, give up long-term plans, and avoid risk.

This isn't a collapse of values but a collapse of incentives. The moment you blame people, the answer disappears; the moment you look at the structure, the possibility of a solution finally opens up.

This is where the question of law and institutions connects too. I think that if Korea is actually going to amend its constitution, it can't end at the level of simply how to divide presidential powers or how to reform the National Assembly. More fundamentally, we have to ask what basic operating principle this society will have.

Whether it will go toward a whitelist-centered society, where you can only do what's permitted, or a blacklist-centered society, which excludes only what's forbidden and otherwise recognizes freedom as the principle.

The former (the whitelist) puts safety first. But it makes people and organizations shrink, and it reinforces risk-aversion and stagnation. The latter (the blacklist) allows freedom and experimentation. Of course there's danger too. But the movement and growth potential of society as a whole come from here.

In the same vein, the difference between incentives and penalties matters too. Incentives are more likely to create a virtuous cycle. Do well and you're rewarded; the reward grows your motivation; and motivation in turn reinforces action.

Penalties, by contrast, easily invite a vicious cycle. Fail once and you shrink; once you shrink you move more conservatively; and in the end no one wants to take responsibility.

If the state truly wants a responsible society, it should prioritize a structure where responsible behavior is rewarded over a direction of strengthening punishment. It should only prevent collapse and build growth through incentives. You can't build growth out of penalties.

This problem is directly tied to elite responsibility as well. More important than how external forces shake things is the moment when the internal elite attaches itself to its own private interest before the community.

Historically, too, real destruction never came from external pressure alone. It came when a price tag got attached on the inside first, and the moment the choice to protect the community became the foolish choice, it collapsed.

So a healthy state isn't one that shouts patriotism, but one where a responsible elite doesn't lose out. The most dangerous state, conversely, is one where the person who slips out first and looks after only their own seat becomes the winner. Once it reaches that state, external intervention isn't the cause but merely the catalyst.

In the end, what we need to look at right now isn't simply how much better AI has gotten. It's how companies and states, power and freedom, responsibility and reward, individual life and the structure of community are all connected. AI is merely the occasion that reveals it. The essence is a much older problem.

Who will control.

Who will take responsibility.

What gets rewarded, and what shrinks.

These questions are coming back again, borrowing the language of technology.

The AI Claude can be seen as a model that, even within a clear constitutional system of constraints, shows a high level of emergent capability. In other words, what looks like unlimited freedom can, depending on the quality and philosophy of clear principles and boundaries, actually produce more depth and flexibility than expected. In this respect, the interesting thing about Claude is that constraints don't suppress creativity—rather, within an aligned “reward system” they make denser, more intelligent output possible.

This offers us many implications. Right now is less an age of technology than an age of redesigning structures like a constitution. It won't be solved by more regulation or stronger propaganda alone. And it certainly won't be solved by blaming people.

We need to move not toward a society that demands responsibility, but toward a society where responsibility is possible and responsibility is rewarded. The people must be able to move freely, power must be bound more tightly, and responsibility must be rewarded by structure, not by words. We should control only systemic risk to the minimum, and let people and organizations stay alive and moving for the rest.

To sum up

To sum up, it's this.

The core problem right now isn't technology but control,

Korea's core problem isn't morality but cost,

and the core task going forward isn't discipline but structural design.

A good society isn't one that forces responsibility.

It will be a society that makes responsible choices repeat themselves naturally.