A lead is not the same as a better life
If you answered the questions, you made a guess about what winning means and what it might cost you. If you skipped them, consider this: the country with the most users could also have the cheapest tools. Neither fact tells us whether people are freer or better off.
The opening choices mix several things. Some measure popularity. Some measure resources. Some measure what the technology can do. Others measure power. We should not treat them as one finish line.
There is no automatic line from China having the cheapest model to your job disappearing. There is no automatic line from America having the strongest model to humans winning.
Three questions matter: Who builds the most capable AI? Who has authority over it? Who gets the benefits and bears the costs? A country can lead on the first and fail its people on the other two.
The story I grew up with
I am an American. I was brainwashed to think communism was bad, China was bad, America was good, and capitalism was good. That story was simple and comforting. It was not enough.
That is not an argument that China is good or America is bad. It is a reason to examine what I was taught instead of using a national label as the answer.
In late September, in Lisbon, I met a Chinese woman named Jenni. She was a consultant and a single mother with a very good job. She was traveling with her seven-year-old daughter, June. June was charming, smart, and remarkably mature.
We talked for about an hour. Jenni mostly uses DeepSeek. She said it was good, and she said many people use Claude because they think it is better. She knew the American systems too.
We talked about freedom, suppression of truth, and corruption. She did not tell me China was perfect. When we talked about business, money, and connections, I said America has that too. She agreed.
That conversation did not prove our governments are the same. It did not prove that China should run the world. It reminded me that a person is not a government. Jenni was living, working, raising June, and choosing the tools she thought worked best.
An ordinary Chinese person having a better life is not an American defeat. An ordinary American having a better life is not a Chinese defeat. That is where I want this argument to start.
The strongest reason to worry about China
A reader can agree with that and still worry about China’s government gaining more power. The concern might be censorship, military advantage, or a government having more influence over the rules everyone’s technology follows.
Those are serious questions. A cheap tool can be useful while the organizations behind it still need to be examined. Meeting Jenni does not answer a question about government power.
But we need to explain the connection. Which kind of AI lead creates which power? How could that power affect someone’s work, speech, privacy, or safety? What evidence supports the concern?
“China wins” is too vague. “This system gives a government a new way to silence people” is a claim we can investigate. I want the same scrutiny when the government is American, or the power belongs to a company.
A dinner is not a safety guarantee
On September 24, Donald Trump hosted Xi Jinping and Peng Liyuan at the White House. A friendly state visit can coexist with serious rivalry. Hosting a dinner does not, by itself, prove that warnings about China are dishonest.
The more useful question is what each leader promises, and whether policy makes that promise real.
Listen to the words. Then look for proof.
During the September 24 remarks, Xi said AI should remain under human control and serve people’s well-being. He also said competition should be healthy and kept within bounds.
Those words matter. They are not proof of what his government will do. I would want to see who can challenge decisions, who can stop harmful uses, and what happens when a powerful institution breaks the rules.
America has to pass the same test
America’s AI Action Plan calls for leadership in AI. It also proposes research into understanding AI systems, keeping control of them, and making them more reliable. It would be wrong to say America talks only about winning while China talks only about control.
The question is whether safeguards can hold when leaders and companies feel pressure to move faster. A promise about safety needs evidence, independent checks, and consequences when it is broken.
What does losing control mean?
On October 5, 2026, Jacob Coxon, a former OpenAI and Anthropic researcher, warned New York City lawmakers about losing control of AI.
There are two different concerns. People could use AI to control other people. Or an advanced AI system could behave in ways that the people responsible cannot reliably direct or stop. We should discuss both without confusing them.
Imagine an AI system helps decide whether you get a job. You cannot find out why it rejected you, and nobody has the authority or willingness to review the decision. That is a concrete failure of human control. It does not require a machine to become an independent ruler.
The broader fear is that future systems could become too capable or unpredictable for the protections we have. That is a risk to investigate, not a certainty I can establish here.
Competition can make the problem harder. If each builder believes a rival will release first, waiting to test a system may feel like losing. That gives people a reason to cut corners. It does not prove everyone will, or that disaster is inevitable.
What I find unacceptable is treating speed as the answer before showing how people will keep control.
Your view
Which concern deserves more attention right now?
These risks can overlap. Choosing one does not mean the other is harmless.
Optional. You can read on without answering.
More abundance does not settle who benefits
In a November 30, 2025 interview with Nikhil Kamath, Elon Musk predicted that AI and robots could make work optional and eventually make money less important. That is a prediction, not a future we have already secured.
Even if machines can make far more goods and services, we still have to decide who owns them, who gets access, and what happens to people during the change. More production does not, by itself, answer those questions.
That is why the old argument about capitalism, communism, and socialism cannot be settled by saying a machine will make everything plentiful. The practical questions remain: who has power, who shares in the gains, and who can object?
Humans must decide means something practical
I am not betting that human judgment will always be better than machine judgment. A machine might make an excellent recommendation. That does not give it the right to decide what every person must accept.
People affected by an important decision should know when AI is involved, understand the reason for the decision, and have a way to challenge it. Someone must be responsible for correcting harm. Where a system cannot be safely controlled, someone must have the authority and ability to limit or stop it.
People will disagree about the rules. “All of us” cannot mean every person can block every decision. It should mean public decisions can be checked by people with authority to enforce the rules, affected people have a voice, and basic rights do not disappear because a majority or a powerful machine finds them inconvenient.
No president, company founder, or machine should get to turn a personal vision into a future everyone else has to accept. The rules should limit powerful people as well as powerful machines.
What winning should mean
Countries are not going to vanish. National security matters. So does the ability to build useful technology. But national leadership and human well-being are not automatically the same thing.
My test for winning is a better life people can actually experience: useful tools they can afford, a fair share of the benefits, protected freedoms, and a real way to challenge decisions that affect them.
So, maybe China should win at making a useful tool. Maybe America should. Neither earns the right to control everybody else by getting there first.
Before accepting a claim that someone is winning AI, ask three questions: What improved for ordinary people? Who can challenge harmful decisions? Who can stop a dangerous use?
If the answers are missing, the claim is unfinished. The winner should be humans.