I’ve spent countless hours tracking the AI arms race between the US and China, and honestly, it feels like we’re all stuck in a loop. It’s always about tariffs, export bans, who has the biggest supercomputer… it’s exhausting, and it feels like we’re missing the forest for the trees.
Then Jensen Huang, the leather-jacket-wearing CEO of Nvidia (you know, the company literally building the engine for this entire revolution), stepped up and said what I think a lot of us have been feeling. He basically said the US is playing the wrong game. The real battle for AI supremacy isn’t about hardware or regulations.
It’s a war for people. Specifically, for developers.
And right now, America is at risk of losing. Badly.
The Real Battlefield: It’s Not What You Think
We get so obsessed with the physical chips. The H100s, the B200s. We see them as weapons, as assets to be hoarded and denied to our rivals. But that’s a Cold War mindset for a digital-age problem.
Huang’s point is devastatingly simple: The most powerful chip in the world is just an expensive paperweight if no one is building anything on it. The magic isn’t in the silicon itself; it’s in the millions of brilliant minds who use that silicon to create applications, build models, and solve problems we haven’t even thought of yet.
Think of it this way: You can have the most advanced F1 engine ever designed, but if all the best drivers, mechanics, and race strategists are working for the competition, you’re still going to lose every single race. The platform is only as powerful as the ecosystem built on top of it.
And here’s the kicker, the statistic that should be blared on every news channel:
Huang says, “50% of the world’s AI developers are in China.”
Read that again. Half.
While the US has been focused on building a fortress, it has effectively locked itself out from the largest pool of AI talent on the planet.
⚙️ The Game-Changing Strategy: The ‘American Tech Stack’
This is where Huang’s thinking gets really interesting. He’s not just pointing out a problem; he’s offering a completely different strategy. He argues that the US goal shouldn’t be restriction, but influence.
The mission? Make the “American tech stack” the undeniable global standard.
What’s a tech stack? It’s just the set of tools and platforms you use to build something. For AI, the American stack is an absolute powerhouse:
- Hardware Layer: Nvidia’s GPUs (the undisputed kings).
- Software Layer: Nvidia’s CUDA platform (the programming model that unlocks the GPUs).
- Cloud Layer: Amazon’s AWS, Google Cloud, and Microsoft Azure.
- Model Layer: Foundation models from OpenAI, Anthropic, Google, etc.
Huang’s vision is to make this stack so dominant, so easy to use, and so powerful that every developer on Earth—whether they’re in a garage in Lagos, a university in São Paulo, or a tech park in Shanghai—chooses to build on it by default.
He made an awesome comparison:
“Just as the American dollar is the global standard.”
That’s the playbook. Not walls, but ubiquity. You don’t maintain power by hoarding the dollar; you maintain power because everyone else needs to use it for trade. The goal is to make the American AI stack just as indispensable.
✨ Why the Current US Strategy Is a Failure
This is where Huang gets brutally honest. He says Washington’s export controls, designed to slow China down, have completely backfired. Ouch.
His logic is solid. Instead of kneecapping China’s progress, the restrictions lit a fire under them.
- 📌 It created urgency: The export controls served as a national rallying cry. China realized it couldn’t depend on foreign tech and poured immense resources, government support, and energy into building its own alternatives from the ground up.
- 📌 It sparked innovation: We’re already seeing the results. Chinese companies are developing credible homegrown AI models like DeepSeek and are making huge strides in their own chip designs. The US essentially forced its biggest competitor to learn how to sprint on its own.
Huang put it perfectly at a conference in Taipei:
“The export control gave them the spirit, the energy, and the government support to accelerate their development.”
In trying to play defense, the US inadvertently created a stronger, more resilient opponent.
🚀 The Real Winning Playbook
So, if the current strategy is broken, what’s the right move? Huang lays out a clear, actionable path forward that is all about playing offense, an offense of attraction and influence.
Here’s what it looks like:
- Open the Floodgates, Don’t Build Dams: The core principle is to get your technology everywhere. The more developers who have access to the American tech stack, the more entrenched it becomes. Stop trying to restrict access and start focusing on expanding your footprint. Let them use your tools. Let them build on your platform. Their success becomes your success.
- Win the People: This is the heart of it. Make the American stack the most loved, best-documented, and most powerful platform for developers. This isn’t just about China. It’s about winning over the rising tide of talent in Africa, Latin America, and Southeast Asia. Be the platform that empowers them, and you win the future.
- Think Like a Platform, Not a Fortress: This is the crucial mindset shift. Fortress America is an outdated idea in a connected world. The great tech victories of our time, such as Windows vs. Mac or iOS vs. Android, were platform wars. They were won by attracting the most developers to build the most apps. The winner wasn’t the one with the highest walls; it was the one with the biggest, most vibrant city inside those walls.
And look, Nvidia is already putting its money where its mouth is. The recent news that they’ll resume selling their H20 chips to China isn’t a sign of weakness. It’s a strategic move to get the Nvidia platform back in front of that 50% developer base. Huang is playing the long game of influence, not the short game of restriction.
This is a fundamentally different, and frankly, more intelligent way to think about global competition in the 21st century. Leadership isn’t about who you can block; it’s about who you can convince to join you. The US needs to stop worrying about building a higher fence and start focusing on setting a bigger table.
- The Talent Imperative: A key part of the AI race is the competition for talent. Nvidia’s CEO notes that with roughly 50% of the world’s AI researchers based in China, their participation is critical for any technology platform that hopes to achieve global dominance.
- The Counter-Productivity of Controls: Jensen Huang has argued that US export controls on advanced AI chips are a “failure.” He believes they inadvertently accelerate China’s indigenous AI development and empower local competitors like Huawei, potentially creating a strong, independent tech ecosystem that rivals the US.
- Nvidia’s Balancing Act: To navigate these geopolitical tensions, Nvidia has developed export-compliant chips, like the H20, specifically for the Chinese market. The company recently received approval to resume selling these chips in China, signaling a strategic effort to balance US policy with the need to engage a massive market.
- The Geopolitical Landscape: The competition involves high-stakes diplomacy. Huang has met with both US policymakers and visited Beijing to discuss market access and AI development. This highlights the ongoing tension between national security concerns, which drive restrictions, and the argument for open collaboration to maintain a globally integrated tech industry.