While governments fight over territory, oil, shipping lanes and military influence, another global competition is accelerating at extraordinary speed.
There are no tanks crossing borders.
No warships exchanging fire.
No armies occupying cities.
But the stakes could eventually be just as important.
Artificial intelligence.
Help keep this independent voice alive and uncensored -> Buy us a Coffee <-
The United States and China are competing for technological leadership while governments around the world are confronting a problem that becomes harder to ignore with every new generation of AI:
The technology is advancing faster than the international rules designed to control it.
Singapore brought that warning directly to the United Nations General Assembly, arguing that AI capabilities are moving ahead of humanity’s understanding of the risks and calling for international safeguards that ensure human beings ultimately remain in control.
That’s not science fiction anymore.
That’s diplomacy.
And it raises a question governments may eventually wish they had answered much earlier:
Who controls the machines when the machines become powerful enough to matter?
Singapore Just Put AI Safety on the Global Agenda
Singapore Foreign Minister Vivian Balakrishnan used his country’s address at the UN General Assembly to push for stronger international cooperation on artificial intelligence.
His argument was straightforward.
Stopping AI development is no longer realistic.
The commercial incentives are enormous. Competition between major powers is already underway. Companies are investing heavily in increasingly capable systems.
So if development isn’t going to stop, safeguards have to keep up.
Singapore proposed exploring something potentially significant: a UN Framework Convention on AI Safeguards.
The idea would be to establish broader international principles and institutions for governing increasingly powerful AI systems.
The comparison with other global governance structures is worth paying attention to.
Humanity has spent decades building international systems around technologies and weapons capable of causing enormous harm.
Artificial intelligence is developing on a dramatically shorter timeline.
The Genie Is Already Out of the Bottle
That’s essentially the dilemma governments now face.
AI isn’t a hypothetical technology waiting to be invented.
It’s already here.
Businesses are using it.
Governments are using it.
Millions of ordinary people are using it.
AI systems are writing software, generating images, analyzing enormous quantities of information, assisting scientific research and increasingly interacting with other digital systems.
The question therefore isn’t whether artificial intelligence should exist.
That debate is already outdated.
The real question is what happens as those systems become more capable.
What should an AI system be allowed to control?
What decisions should always require human authorization?
What happens if autonomous systems behave in unexpected ways?
Who is responsible when something goes wrong?
And perhaps most importantly:
Which decisions should machines never be allowed to make?
Would You Give AI the Nuclear Button?
Singapore’s foreign minister used a deliberately provocative example at the United Nations.
Would anyone want the nuclear button controlled by artificial intelligence?
The answer illustrates the larger problem.
Society already accepts automation in countless areas.
Algorithms recommend what we watch.
Software determines routes.
Computers execute financial transactions.
AI increasingly assists people in making complicated decisions.
But there is an enormous difference between software recommending a movie and an autonomous system making a decision involving human life, military force or national security.
Where exactly should that line be drawn?
Governments don’t yet have a universal answer.
Technology isn’t waiting for them to find one.
Meanwhile, America and China Are Talking
Something else happened that deserves attention.
The United States and China have agreed to establish a dialogue focused on artificial intelligence.
That’s important precisely because the two countries are also major technological competitors.
Washington and Beijing aren’t suddenly abandoning that competition.
They remain rivals across trade, technology, national security and geopolitical influence.
But both sides appear to recognize that AI creates risks where communication between competitors could become necessary.
Senior American and Chinese officials have discussed AI safety, emergency communications and ways of notifying one another about serious AI-related incidents.
Another round of talks is expected.
That doesn’t mean the United States and China suddenly agree on how artificial intelligence should be developed or governed.
They don’t need to.
Communication itself matters when two powerful competitors are developing technology that neither side completely controls internationally.
We’ve Seen This Problem Before
There is an uncomfortable historical comparison.
Nuclear weapons transformed international politics because humanity suddenly possessed technology capable of causing destruction on a scale previously unimaginable.
Governments eventually developed treaties, communication systems, inspection regimes, doctrines and rules intended to reduce the possibility of catastrophe.
Those systems weren’t perfect.
But they recognized something fundamental.
When technology becomes powerful enough, misunderstanding becomes dangerous.
Artificial intelligence is obviously not identical to nuclear weapons.
The technologies are fundamentally different, and pretending otherwise would oversimplify the issue.
But they share one important characteristic.
Both raise questions where the consequences of getting the rules wrong could become extremely serious.
There’s one major difference.
The nuclear age gave governments decades to construct much of today’s international architecture.
AI development is moving at digital speed.
Governments Are Writing the Rulebook While the Game Is Already Being Played
That’s what makes the current moment unusual.
Normally society tries to establish rules before deploying dangerous systems widely.
Artificial intelligence has largely reversed that process.
Companies innovate.
Consumers adopt the technology.
Businesses integrate it.
Governments react.
Regulators debate.
International institutions hold meetings.
And meanwhile, another generation of AI arrives.
Then another.
The rulebook is being written while the technology itself continues changing.
Imagine trying to write traffic laws while cars are becoming twice as fast every few months.
That’s roughly the governance challenge policymakers are beginning to describe.
The US-China Competition Makes Slowing Down Harder
There is another problem.
Even if one country wanted to slow development dramatically, would its competitors do the same?
That’s where national security collides with AI safety.
If Washington believes Beijing is racing ahead, pressure grows for America to accelerate.
If Beijing believes Washington is gaining a decisive technological advantage, China faces the same incentive.
Companies face similar pressures.
Nobody wants to be the organization that voluntarily slows down while a competitor develops something dramatically more powerful.
That creates a classic race dynamic.
Everyone may understand the risks.
Nobody wants to finish second.
AI Isn’t Just Another Technology Industry
It’s tempting to think of artificial intelligence as another version of smartphones, social media or the internet.
That may underestimate what governments are actually competing over.
Advanced AI could affect economic productivity.
Scientific research.
Cybersecurity.
Intelligence gathering.
Military planning.
Manufacturing.
Education.
Medicine.
Finance.
Information warfare.
And potentially the development of future technologies that don’t yet exist.
Whoever leads in AI may gain advantages across many other sectors simultaneously.
That’s why the competition matters far beyond Silicon Valley.
Artificial intelligence is becoming part of national power.
What Happens When AI Meets the Military?
This is where the debate becomes especially serious.
Militaries have always adopted technologies that provide strategic advantages.
Radar.
Satellites.
GPS.
Drones.
Cyber capabilities.
Artificial intelligence will be no different.
AI can potentially help analyze intelligence, identify patterns, operate autonomous systems and process information far faster than human beings.
But speed introduces another problem.
Humans are slow.
Machines aren’t.
In a crisis, governments could eventually face pressure to let automated systems make increasingly rapid decisions because human authorization takes too long.
That’s precisely where the principle Singapore emphasized becomes critical:
Humans must ultimately remain in control.
Because once machines begin making irreversible decisions independently, accountability becomes much harder.
The Race Isn’t Only Between Governments
There’s another unusual feature of the AI era.
Some of the organizations developing the world’s most advanced systems aren’t governments.
They’re companies.
Private technology firms now control computing infrastructure, models, research teams and technical capabilities that governments themselves may depend upon.
That creates a governance problem unlike many previous technological competitions.
International treaties traditionally regulate countries.
But what happens when the fastest innovation comes from corporations operating across borders?
Who establishes the safety standards?
Governments?
Companies?
The United Nations?
Independent scientists?
Or some combination of all four?
Nobody has completely answered that yet.
The World Doesn’t Have Unlimited Time
None of this means artificial intelligence is destined to become dangerous.
AI could produce extraordinary benefits.
It may accelerate scientific discoveries, improve medicine, increase productivity, make education more accessible and solve problems that humans struggle to manage today.
That’s exactly why stopping development entirely isn’t realistic.
The potential benefits are simply too large.
But enormous benefits and enormous risks can exist simultaneously.
That’s why Singapore’s warning deserves attention.
The goal shouldn’t necessarily be stopping the engine.
It should be making sure the brakes work.
We Built Nuclear Rules After Seeing What Nuclear Weapons Could Do
Humanity already knows what nuclear weapons can do.
That terrible knowledge shaped generations of international policy.
Artificial intelligence presents a different challenge.
We’re trying to build the rules before we fully understand what the most advanced versions of the technology will eventually be capable of doing.
That requires governments to regulate something whose future capabilities remain uncertain.
Move too slowly and safeguards may fall behind.
Move too aggressively and governments risk restricting technology that could produce enormous benefits.
Do nothing and private competition may effectively write the rules by default.
None of those choices is simple.
This May Become One of the Defining Competitions of the Century
The geopolitical map is already crowded.
Ukraine.
Iran.
Hormuz.
Taiwan.
The South China Sea.
The Red Sea.
Governments are dealing with traditional struggles involving territory, military power, resources and trade routes.
But underneath those visible conflicts, another competition is accelerating.
It doesn’t require soldiers.
It doesn’t require aircraft carriers.
It doesn’t even require borders.
It requires computing power, data, energy, scientists and increasingly sophisticated algorithms.
And unlike traditional geopolitical rivalries, the battlefield exists almost everywhere simultaneously.
We spent generations developing rules around nuclear weapons.
Artificial intelligence may not give us generations.
The technology is developing now.
The competition is happening now.
And the world is still debating what the rules should be.
That may ultimately be the most important part of the story.
About the Author
