- Monday Momentum
- Posts
- The New Cold War
The New Cold War
How America, China, and the UAE are fighting for AI supremacy with three radically different strategies
Justin Wright
July 28, 2025 • Est. Reading Time: 7 minutes
Happy Monday!
While AI companies compete for market share, a more consequential battle is quietly unfolding: the race between nations to define how artificial intelligence shapes their future. Three distinct strategies have emerged, each revealing fundamentally different philosophies about government's role in the AI revolution.
The White House just unveiled America's AI Action Plan, partnering with Anthropic to build critical AI infrastructure requiring 50 gigawatts of power by 2028. Meanwhile, the UAE made ChatGPT Plus free for all 10 million residents, positioning AI as a public utility. And as we explored last week, China's open-source strategy is democratizing advanced AI capabilities globally.
Three nations, three radically different approaches to the same existential question: how do you win the AI race?
The global AI competition has crystallized into three distinct national strategies: America's infrastructure-first approach focusing on massive computational buildouts, the UAE's universal access model treating AI as public service, and China's open-source strategy prioritizing ecosystem velocity. Each reflects fundamentally different philosophies about technology governance and reveals that AI leadership may depend less on raw computing power and more on smart distribution and human capital development.
The Three Paths to AI Supremacy
For decades, the assumption was simple: whoever builds the biggest, most powerful AI systems wins. But the race for AI dominance isn't just about computing power, it's about thoughtfully implemented strategy.
While Sam Altman testified to Congress about America's need for massive infrastructure investments, the UAE quietly demonstrated that you can achieve AI parity through smart partnerships rather than raw buildouts. The Emirates committed an estimated $200 million annually to provide ChatGPT Plus access to all residents, instantly creating the world's most AI-literate population without building a single data center.
Meanwhile, America's AI Action Plan reveals the scale of infrastructure required to maintain technological leadership: 50 gigawatts of computing capacity, federal land leases, and regulatory streamlining that would make a highway project look simple. Anthropic's accompanying report laid out the technical requirements: 5-gigawatt data centers for single AI training runs, transmission line partnerships, and priority grid access.
As we explored last week, China's open-source strategy continues accelerating, with companies releasing breakthrough models under permissive licenses that prioritize ecosystem velocity over proprietary advantage.
The Meta Trend: AI as National Infrastructure
The emergence of national AI strategies represents a key deviation away from viewing AI as just another technology sector. Treating it as critical infrastructure like electricity, highways, or the internet seems to be the modern playbook. Countries are no longer asking whether AI will transform their economies, but rather who will control that transformation and how benefits will be distributed.
This reflects a growing recognition that AI advantage isn't just about having the best models or most compute. It's about creating sustainable ecosystems that align technological capability with national values, economic priorities, and geopolitical positioning.
Pattern Recognition: Three Models of AI Governance
Pattern #1: The American Infrastructure Approach
The U.S. strategy focuses on removing regulatory barriers while building massive physical infrastructure. Anthropic's report calls for 50 gigawatts of AI computing capacity by 2028. This approach emphasizes private-sector leadership with government facilitation. Key pillars include accelerating permitting for data centers, leveraging federal lands for AI infrastructure, and creating categorical exemptions under environmental review processes. The strategy explicitly aims to "remove red tape and onerous regulation" while ensuring America maintains technological dominance.
Pattern #2: The UAE's Universal Access Model
The UAE's partnership with OpenAI treats AI as a public service, providing ChatGPT Plus (normally $20/month) free to all residents. This goes beyond affordability and ensures universal AI literacy in an increasingly AI-driven world.
The initiative pairs consumer access with infrastructure investment: the Stargate UAE project will build a 1-gigawatt AI computing cluster in Abu Dhabi. The UAE is essentially nationalizing AI access, turning GPT-4 into a public utility while building sovereign AI capabilities.
Pattern #3: China's Open Source Strategy
As we analyzed last week, China's approach emphasizes open collaboration and rapid iteration. Companies like Moonshot AI, Tencent, and Baidu are releasing state-of-the-art models under permissive licenses, creating compound innovation effects impossible in closed environments.
This strategy sacrifices proprietary advantage for ecosystem velocity. By open-sourcing breakthrough models like Kimi-K2 and Hunyuan-A13B, Chinese companies accelerate global adoption of their architectural innovations while building influence through technological leadership.
Contrarian Take: The Infrastructure vs. Access Paradox
The conventional wisdom suggests that massive infrastructure investment automatically translates to AI leadership. But the UAE's approach reveals a different path: strategic access partnerships can deliver advanced AI capabilities faster and cheaper than building from scratch.
While the U.S. focuses on $50 billion infrastructure buildouts and China invests in open-source development, the UAE spent an estimated $200 million annually (10 million residents × $20/month) to instantly provide its entire population with frontier AI access. This approach delivers immediate productivity gains while the data centers are still being constructed.
The UAE's strategy exposes a critical insight: in the short term, smart distribution beats raw production capacity. By partnering with existing AI leaders rather than competing with them, smaller nations can leapfrog traditional development timelines and focus on application rather than foundation-building.
The Bigger Picture: Competing Visions of the AI Future
These three strategies represent fundamentally different philosophies about how AI should be developed and distributed:
The American model prioritizes private innovation with government infrastructure support, betting that removing barriers will unleash market forces to drive breakthrough technologies.
The UAE model treats AI as public infrastructure, ensuring universal access while building sovereign capabilities through strategic partnerships.
The Chinese model emphasizes collaborative development and rapid iteration, using openness as a weapon to accelerate global ecosystem development.
Each approach reflects different strengths and constraints. The U.S. leverages its venture capital ecosystem and regulatory flexibility. The UAE uses oil wealth and geographic positioning. China deploys its manufacturing scale and state coordination.
The winning strategy may ultimately be the one that best aligns technological development with human capital development. The UAE's universal access approach ensures its workforce becomes AI-native. China's open-source strategy builds global influence and rapid iteration cycles. America's infrastructure focus aims to maintain technological leadership through computational advantage.
In motion,
Justin Wright
If small nations can achieve AI parity through strategic partnerships, while large nations compete on infrastructure and open-source development, does traditional economic scale still matter in the AI era?
