The global AI race: who will define the next era?

The global AI race: who will define the next era?

Aurora Insights Founder and CEO Todd James joined Fast Company Editor-in-Chief Brendan Vaughan and Bot Auto CEO Xiaodi Hou at the Fast Company Impact Council Annual Meeting for a discussion on which global players are best positioned to define the AI era, and what it will actually take to win.

His view: the race has moved past models alone. The next era will be shaped by countries and companies that can build the infrastructure, supply chain resilience, and operating capacity to convert AI into durable economic advantage.

U.S. vs. China: two different bets

The U.S. advantage remains market dynamism. Capital, startups, frontier labs, hyperscale infrastructure, universities, and enterprise demand all move toward what works. That system is difficult to replicate.

But the model gap has closed quickly. Stanford’s 2026 AI Index shows the top U.S. model leading the top Chinese model by just 2.7 percentage points as of March 2026. Private investment still heavily favors the U.S., but those figures do not fully capture state-directed capital flowing through Chinese guidance funds and industrial policy.

China’s strength is not government funding alone. It is manufacturing depth, industrial deployment at scale, and coordinated physical buildout. The next phase of AI will not live only in software. It will show up in robotics, autonomous systems, manufacturing, logistics, energy, and the physical operating layer of the economy.

Europe and the caution trap

Governance matters. Trust matters. But trust is necessary, not sufficient.

Europe’s risk is not that it wrote rules. The risk is that it may define the rules without also building scaled platforms, scaled capital formation, and scaled adoption. Institutions often reward caution because people are rarely criticized for slowing down. But in a technology cycle this important, excessive caution has a real cost: slower company formation, weaker industrial competitiveness, and less influence over the systems others end up deploying.

The lesson is not that governance is bad. It is that governance has to enable responsible deployment, not become a substitute for building.

Supply chain fragility: resilience is not decoupling

For decades, supply chains were optimized for cost efficiency. Resilience was underpriced. The AI race forces a different accounting: what is the risk-adjusted cost of losing access to something strategically critical?

The answer is visible now. The IEA’s 2025 Global Critical Minerals Outlook found that China is the dominant refiner for 19 of 20 important strategic minerals, with an average market share of roughly 70 percent. The U.S. once led global rare earth production through Mountain Pass in California, which supplied the majority of global rare earths from the mid-1960s through the mid-1990s. Mining was suspended in 2002, and rebuilding domestic capability is a long-term effort, not a policy announcement.

Total decoupling is unrealistic. But selective resilience in critical layers is not optional. Advanced chips, compute infrastructure, power, rare earth processing, and key hardware components all matter. The goal is not supply chain isolation. The goal is supply chain resilience.

The next era

The next era will be defined by countries and companies that can balance three things: speed, resilience, and trust. Speed without resilience creates fragility. Resilience without speed creates stagnation. Trust without adoption becomes theater.

The winners will build quickly, deploy responsibly, and convert AI into real economic performance.
‍

Sources: Stanford HAI 2026 AI Index; IEA Global Critical Minerals Outlook 2025; MP Materials; USGS historical production records.

‍