People is the final frontier of US-China competition

[This post was originally shared on

The Great Divorce

, a newsletter of mine dedicated to documenting US-China competition]

For the past few days, I have been watching with sheer amazement at how the likes of Elon Musk argue with the MAGA base on the virtues of H1B visas. The civil war between tech-bro MAGA and real MAGA has finally erupted.

On the one hand, I totally agree with Elon that immigration of high-skilled labor is the single most important factor for the US to keep its competitive edge over China right now. On the other hand, I am keenly aware that telling many Americans in their face that they are not good enough, that they are not smart enough, and that they are just too dumb to be employable, is a political dynamite that will explode one day or another. Politics simply won’t work this way.

People - the final frontier of US-China competition

First, we need to understand the central role the people factor plays in the US-China competition, a.k.a the Great Divorce.

Right now, America actually does not have so much leverage on China. Sure, the US can impose tariffs on Chinese goods, but there are clear limits to how much they can impose. After all, the US is but one market that China exports to, and US companies (Apple, Tesla, etc) also reap massive profits from working with China.

Sure, the US can impose export controls on key technologies. Small yard, high-fence. However, the recent model of DeepSeek, a potent but fantastically cheap AI model has demonstrated that those export controls not only fail to blunt progress, but may actually force Chinese people to be smarter.

But there is one edge that the US can expect to enjoy for the foreseeable future: its ability to create truly ground-breaking, zero-to-one innovation.

In the case of AI, this is the ability to create OpenAI-level, ChatGPT-level innovation in the first place.

It is no surprise, in a veiled response to DeepSeek, Sam Altman tweeted:

it is (relatively) easy to copy something that you know works.

it is extremely hard to do something new, risky, and difficult when you don't know if it will work.

individual researchers rightly get a lot of glory for that when they do it! it's the coolest thing in the world.

it's also extremely hard to rally a big talented research team to charge a new hill in the fog together.

this is key to driving progress forward.

I think his “a new hill in the fog” analogy is exactly right.

Of the whole world, only America has this kind of spirit in large quantity.

Note that quantity is important here. Every society has geniuses, but to mobilize geniuses together to create unimaginable wonders requires critical mass.

How does America achieve this?

First, it has a highly failure-tolerant culture. In America, you are okay to fail. You are okay to dream of things like sending rockets to the sky as a private company, even if it had never been tried before by non-state actors.

You are okay to bet on the value of large-language models, even if they have never been proven to work before.

You are okay to test out crazy ideas, and there will be still enough people backing your crazy ideas.

In China, on the other hand, we don’t have that kind of culture. Even if you may have some crazy ideas, you have to persuade many, many people - investors, employees, regulators, media - that your crazy ideas actually can work, and all of them will doubt it, because you can’t prove, simply by talking, that it will actually work. They will tell you, why do you waste your time? Why can’t you work on something more practical? Without their backing, you can never make your ideas into reality. Therefore, most of the “0-1 innovators” in China will be caught in this chicken-and-egg problem.

Things become much easier when, and only when, a first variant has already been proven to work. Be it EV, private space exploration, or ChatGPT, when the US companies prove that some new concept works, the Chinese companies would scramble to catch up, scale, iterate, and can usually be much more efficient than the original variant. China has an overabundance of “1-to-100” innovators, so long as that “0-1” has been proven to work. We act exceptionally well when the yardstick is hanging right there, just like how well we perform at schools when the scores are hanging right there. This is what we are good at (This is by the way why I think export controls will never be able to blunt China’s progress. Scaling on existing innovation is China’s comparative advantage.)

America’s second weapon is its ability to attract the best talents from around the world, who are mostly attracted by this failure-tolerant culture, who in turn will also help to contribute to this culture.

Here, China has significant disadvantages. Language barriers and cultural differences already make China nearly impossible to attract most non-Chinese talents, while it also keeps losing many of its own talents to the US. So while the US is attracting the best from all over the world, China is only able to attract some top talents from China. This smaller pool makes it difficult for China to achieve the “critical mass” of top talents that is necessary for the most cutting-edge research.

But our relative lack of top talents is more than made up for by the abundance of engineers, who enable China to always quickly learn from the new frontiers of innovation and quickly build processes that can provide scale to technologies.

In such light, America’s best strategy is clear:

1. Do not count on holding back China’s technological progress, because it will never work. Learning from the best is our top skill.

2. Maintain the failure-tolerant culture and keep attracting the smartest people from all over the world, so that the US can always keep a 6-18 month lead time over China.

This is why I think Elon Musk is totally right here. He already sees the truth. It’s also interesting that Trump chose to side with Elon in this case, showing that Trump is actually smarter than he appears to be (and also that since he has already won the re-election, he could give a smaller F about his voters.)

However, this is a very inconvenient truth, and I don’t think the MAGA base will ever be able to stomach it. Steve Bannon is already threatening to “rip off” Musk’s face. I also consider the exploding Cybertruck in front of the Trump Hotel to be a warning from this base.

If (and I think it’s very likely) the US slides more and more into this kind of anti-immigrant culture, it might lose its only remaining edge over China.

The question is, is China able to seize on this opportunity? In our century’s competition of “who is less worse比烂”, the jury will still be out on this one for a long time.