Can America Still Out-Innovate China?

An Asia Times piece recently argued that the U.S. is unlikely to keep outpacing Chinese innovation. Looking at the evidence, it’s hard to say they’re completely wrong.

For years, American experts predicted China would stay stuck as a copycat because it lacked free markets, strong patent protection, and democratic debate. Those predictions aged poorly. China now leads or is highly competitive in electric vehicles, batteries, drones, solar, 5G hardware, and industrial robotics. It didn’t get there by copying Silicon Valley. It got there by treating manufacturing itself as the main engine of innovation and by using the state as a big coordinator rather than just a referee.

America still does some things better. We lead in the most advanced AI models, chip design, software platforms, and venture capital. The biggest breakthroughs in frontier AI still tend to come from U.S. labs. That edge is real.

But the old phrase “Designed in Cupertino, manufactured in China” still captures the problem. The U.S. is excellent at inventing and designing. China is excellent at the harder, less glamorous work of turning those designs into refined, affordable products at massive scale. When most of the actual making happens elsewhere, you lose the daily learning that comes from production. China never stopped treating that learning as valuable.

That loss of manufacturing depth is now showing up in national security. The U.S. struggles to build its own navy ships on time and in meaningful numbers, while China turns them out faster and in far greater quantity. A country that cannot produce the physical tools of power at scale eventually finds its strategic options limited, no matter how advanced its software or designs are.

Part of the difference also comes from how the two governments are built. China’s senior leadership has long included a high share of people with engineering and technical backgrounds, and the system allows for longer stretches of policy continuity. Once they decide something is strategic, they can keep pushing for years. U.S. politics is dominated by lawyers and business people who change with every election cycle. Priorities shift, funding gets interrupted, and long-term industrial efforts are harder to sustain.

American innovation is also heavily shaped by what capital markets will fund in the short term. China can stick with strategic technologies for years even when the quarterly numbers look bad. That patience has paid off in several key industries.

Does this mean America is doomed to fall behind? Not necessarily. We still have big advantages in talent, basic research, and the ability to attract ambitious people from everywhere. But those advantages erode if we keep treating manufacturing as someone else’s problem, if regulation and energy costs keep making it hard to build things at scale here, and if political turnover keeps interrupting any attempt at sustained industrial strategy.

The article’s core point holds up better than the old “China can’t innovate” consensus did: different systems produce different kinds of innovation. Right now, China’s system is better at turning technology into widespread, practical dominance in a growing number of fields. America’s system is still better at the very frontier of certain technologies.

The open question is whether leading at the frontier is enough if the other side keeps winning the race to deploy and scale. Looking at batteries, EVs, drones, industrial robots, and now even naval shipbuilding, the answer so far has not been encouraging for the U.S. The Cupertino half of the equation is still strong. The manufacturing and long-term coordination half is the part that keeps slipping.

AI Got Really Smart… and Almost Nothing Changed

AI just solved a math problem that humans couldn’t crack for 87 years. It also figured out some tough quantum crypto stuff. The best mathematicians on the planet are now using AI to help them understand the answers.

This is the worst AI will ever be at this kind of thing. It’s only going to get better.

And yet… look around. Daily life is pretty much the same. People still drive to work, sit at computers, scroll their phones, and collect paychecks. Jobs haven’t disappeared in some massive wave. The economy is growing okay, but not exploding.

A lot of smart people expected the world to transform the second machines got smarter than us. Sci-fi always shows an “intelligence explosion” where AI becomes god-like overnight and rewrites reality. That didn’t happen.

Some folks are starting to wonder if intelligence itself has limits — like there’s only so smart you can get before extra brainpower stops helping much. Making an AI smarter might be less like building a taller tower and more like polishing a ball that’s already pretty round. At some point the gains get smaller.

Even if pure brainpower has a ceiling, AI can still change a lot of things for three practical reasons:

1. You can make more of it.
Humans are limited by birth rates. AI isn’t. You can just build more data centers and more robots. Suddenly one person can have a whole army of smart machines helping them.

2. It can notice the tiny stuff humans miss.
A lot of valuable know-how in companies is just a bunch of little tricks workers do without even realizing it. AI that watches everything can spot those tricks, test improvements fast, and spread them. That alone could speed up how quickly average companies catch up to the best ones.

3. It can handle messy patterns we can’t.
Some things in the world are too complicated for any single human brain (or even a group of them) to fully grasp. AI might still be able to use those patterns usefully even if we never really understand them.

So the big payoff from AI may not come from it being a genius that outthinks every human. It may come from pairing human-style thinking with the things computers already do better: speed, perfect memory, and handling huge amounts of data without getting tired.

The intelligence explosion already happened. It just didn’t look like the movies.