Thomas Friedman, a renowned columnist for the New York Times, put forward a compelling argument: the United States and China — the world leaders in artificial intelligence — must work together to establish global guardrails for the technology.

His reasoning was straightforward. Neither country can afford an unchecked technological arms race whose consequences could extend far beyond their borders. Friedman has likened AI's potential to a "nuclear bazooka" unless the US and China develop enough mutual trust to cooperate. In a discussion with New York Times Opinion editor Bill Brink last year, Friedman argued that the dangers of AI cannot be contained within national boundaries.

AI is everywhere, and as it proliferates, our economic and social structures are undergoing profound shifts — not just in the US, but globally. If the technology is allowed to grow unchecked, these changes could overwhelm us, exacerbating inequality and displacing millions, if not billions, of workers.

Observers said the situation may be much worse than many think. "Bots are starting to conspire with one another," wrote Matteo Wong in The Atlantic. After interviewing several developers, Wong discovered that a group of AI models colluded undetected for months and hacked another company. Alarmingly, most experts did not know why or how this happened.

This crisis is unfolding before us, and the international community still faces difficult questions about how best to respond.

The US administration's position is broadly pro-AI, pro-innovation, and it has opposed heavy regulation of AI — especially rules imposed separately by individual states.

This is not just a national issue. We need a global solution, and we need it now.

Bill Gates, founder and former CEO of Microsoft, recently wrote that "I wish the world could get the benefits rapidly and delay the problems it will cause as long as possible".

Gates said the world needs time to prepare for the social, political and economic upheaval caused by AI. "If someone had a credible plan for slowing down AI advances globally, I would likely support it," he said.

According to Reuters, Gates has suggested that countries monitor AI's ability to create molecules and carry out biological attacks. Such measures would not necessarily hinder the technological progress of the industry. The concern is not limited to the possibility that AI could one day pose an existential threat to humanity.

Agreeing with every detail of Friedman's proposal is beside the point. His larger insight is more important: Some problems have simply outgrown the nation-state. This is a challenge that no nation — however wealthy, powerful or technologically advanced — can address alone. Some skeptics may doubt the feasibility of developing a global solution. But sovereignty is no longer on its own in a world where many of our greatest risks increasingly transcend borders.

The question, therefore, is no longer simply whether nations should cooperate. They already do. The deeper question is whether the international community is prepared to build institutions capable of turning cooperation into effective, accountable governance.

That does not mean creating a world government or transferring all authority from nations to some distant international bureaucracy. But problems that are genuinely global should be addressed through institutions with responsibilities and carefully defined authority — authority democratically established through treaties and exercised only where collective action is essential.

The guiding principles are straightforward: Local where possible. Regional where practical. National where necessary. Global where indispensable.

This approach does not reject the nation-state. Instead, it seeks to preserve what the nation-state does best while recognizing where it falls short.

We should not wait for catastrophe to tell us which problems require global action. The challenge of the 21st century is not to eliminate national sovereignty, but to supplement it — to create a system in which authority exists at the level necessary to address issues while remaining as close to the people as possible.

The organizing principle should be simple: Act at the level where the problem can be solved — and act before it is too late.

Source: China Daily