I build AI tools professionally here in San Francisco, where I use American models from companies like OpenAI, Anthropic, and Meta every single day. I take for granted that these companies generally respect user privacy, don’t automatically report user activity to the government, and won’t help track political dissidents.
But that’s not a given everywhere in the world. China is building AI incredibly fast and at massive scale, with fundamentally different values baked directly into their systems from the beginning.
Whoever wins this race will determine what AI looks like for the next 50 years globally. I genuinely care about America winning this competition, and not out of blind patriotism or nationalism, but because I’ve personally seen what happens when authoritarian governments get control of powerful technology.
Table of contents
Open Table of contents
What’s Actually at Stake
This isn’t about American exceptionalism or jingoistic nationalism. This is fundamentally about values that get embedded into the technology itself.
American-developed AI operates under free speech protections enshrined in the Constitution, privacy laws that are imperfect but exist and improve over time, independent courts that can challenge government overreach, public accountability through democratic processes, and a strong tradition of open research that benefits everyone.
Chinese-developed AI operates under pervasive internet censorship controlled by the state, mass surveillance systems monitoring all citizens constantly, a government that makes critics simply disappear without due process, mandated cooperation with state intelligence agencies, and tightly controlled model outputs that reinforce party ideology.
If Chinese AI becomes the dominant global standard, those authoritarian values get exported worldwide to every country that adopts the technology. The stakes here are about fundamental values and human freedom, not about GDP growth or national pride.
China’s Actual Advantages
China isn’t losing this race by any means, and on some dimensions they’re actually ahead of the United States right now.
Data access at unprecedented scale: China has 1.4 billion people with severely limited privacy protections, giving them access to training data at a scale that American companies simply can’t match due to our stronger privacy laws and regulations.
Centralized coordination without bureaucratic delays: When the Chinese government declares AI a national priority, that means massive funding flows immediately without endless congressional debates or budget fights, just pure resources deployed at scale.
Regulatory speed advantages: China faces substantially less regulation, public scrutiny, and ethics review processes, allowing them to move faster from research to deployment without the friction democratic societies impose.
Deep talent pool: China graduates more STEM students annually than the US, and many of the absolute best AI researchers in the world are ethnically Chinese working either in China or abroad.
Ignoring or downplaying these very real advantages is genuinely stupid. We need to be coldly realistic about the competitive landscape.
Where America Still Leads
Innovation culture that rewards bold risk-taking: Silicon Valley’s unique culture rewards entrepreneurs who take big risks, and the ability to fail spectacularly, restart immediately, and try again creates breakthrough innovations that centralized planning systems fundamentally can’t replicate.
Open research culture accelerates collective progress: American research gets published openly, debated publicly by peers, challenged rigorously by critics, and improved collaboratively by the global community, which dramatically accelerates the pace of progress.
Immigration as competitive advantage: The best global talent from every country still desperately wants to come to America and build here, so we should make immigration dramatically easier rather than pointlessly harder, especially for STEM talent.
Venture capital funding at massive scale: American VC funding for AI research absolutely dwarfs China’s venture ecosystem, with billions of dollars available for genuinely open-ended research without guaranteed commercial applications.
Rule of law protections for builders: You can build companies in America without constant fear of arbitrary government nationalization or forced technology transfer to state-owned competitors.
These advantages genuinely matter and give America a real edge. But they’re absolutely not guaranteed to last forever, and we can definitely lose them through bad policy choices.
What Losing Looks Like
If America loses this race, global technology standards will increasingly reflect Chinese authoritarian values rather than democratic principles. Countries around the world will adopt Chinese AI systems because they’re cheaper, technically advanced, and politically aligned with authoritarian governments’ desires for control. Censorship capabilities, mass surveillance infrastructure, and state control mechanisms will be baked directly into the foundational technology layer.
American technology companies will find themselves systematically locked out of global markets. China already aggressively blocks US tech companies from operating inside China, and if they achieve true AI dominance that means American companies lose access to enormous markets worldwide as countries align with Chinese technical standards.
Authoritarian governments everywhere will become dramatically stronger and more resilient. AI-powered surveillance systems, sophisticated propaganda generation, and algorithmic population control will help dictatorships maintain power more effectively. Protests will get crushed faster through predictive policing, and dissidents will be found more easily through omnipresent surveillance networks.
America will lose crucial geopolitical leverage on the world stage. Whoever controls foundational AI infrastructure wields enormous geopolitical power through technical dependency. If China wins decisively, American global influence inevitably shrinks as other nations become dependent on Chinese technology.
This isn’t some hypothetical distant future scenario. China exports surveillance technology to authoritarian governments globally right now today, and AI capabilities will make these systems 10x more powerful and pervasive.
What Winning Requires
Winning this race means American AI systems and values set the dominant global standard that others follow.
Massive research investment at moonshot scale: We need to fund fundamental AI research like we funded the space race, with sustained investment in pure open research through universities, national laboratories, and independent researchers without demanding immediate commercial applications.
Comprehensive immigration reform to attract talent: Roughly half of all AI researchers working in America today were born abroad in other countries, so we desperately need to make immigration dramatically easier with automatic green cards for PhD graduates and stopping our current policy of shooting ourselves in the foot by pushing talent away.
Critical compute infrastructure investment: We need substantial investment in domestic chip manufacturing, massive data centers, and energy infrastructure at unprecedented scale since China is investing extraordinarily heavily in all three areas and we absolutely must match them.
Smart balanced regulation: We need regulation sophisticated enough to prevent genuine harms without accidentally killing beneficial innovation, carefully balancing legitimate ethics concerns with avoiding a mass flight of innovation to less-regulated offshore locations.
Deep allied cooperation across democracies: America can’t win this race alone competing solo against China’s centralized system, so we need tight cooperation with Europe, Japan, South Korea, and Taiwan to share research openly, coordinate technical standards, and diversify vulnerable supply chains.
Why I Care
I build products professionally and use AI tools daily to ship faster and build better. I desperately want AI that makes people more capable and empowered, not more controlled and surveilled.
If China wins this race decisively, then AI systems come bundled with built-in censorship, pervasive surveillance, and mandatory government backdoors as standard features. I’ve traveled to China personally, seen the internet censorship in action firsthand, and talked extensively with Chinese developers who can’t access half the development tools and resources I use every day without thinking about it.
That’s absolutely not the future I want for myself or anyone else. America isn’t perfect by any means, tech companies here do plenty of shady things, and our government makes terrible mistakes regularly.
But I’d genuinely rather build in America’s messy and imperfect freedom than in China’s efficient and optimized authoritarian control.
This Isn’t About Hating China
I’m absolutely not anti-Chinese as people or culture. I deeply respect Chinese culture, history, and people, many of my closest colleagues are Chinese, and some of the smartest AI researchers I personally know come from China.
This is entirely about systems, values, and what kind of AI shapes the next century of human civilization. AI built in democracies with public debate, meaningful oversight, and protection of individual rights differs fundamentally from AI built under authoritarian control. I strongly want the democratic version to win because I believe it produces better outcomes for humanity.
What You Can Do
Build if you’re technical: If you have technical skills, actively build AI tools and contribute to open source projects that make American AI better and more capable.
Learn regardless of background: Make serious effort to understand AI technology rather than either fearing it blindly or ignoring it completely, so you can make genuinely informed decisions.
Vote strategically: Support politicians who take AI competition seriously, who will fund fundamental research adequately, and who will reform our broken immigration system.
Speak up publicly: Actively call out surveillance and authoritarian control when you see it, and publicly support good ethical research that advances human capability.
One individual person can’t win this race alone, but millions of people building, learning, and advocating together absolutely can.
The Bottom Line
America genuinely needs to win this AI race. Not because Americans are somehow better people than anyone else, but because American democratic values will produce better AI for humanity than authoritarian government control.
I want AI that augments human capability without constantly surveilling behavior, that helps organize peaceful protests rather than crushing them with force, that produces open debatable improvable research rather than state secrets, and that creates code that isn’t subject to government approval before shipping.
That world requires America and our democratic allies leading AI development globally. We need to win this race, not for blind nationalism or national pride, but for genuine human freedom.