The U Lab Brief 17 | Google, SpaceX and The Compute Race

Google just signed a deal to pay SpaceX $920 million per month for AI compute. Google is already one of the largest owners of AI infrastructure in the world. Yet it is still renting massive amounts of external compute capacity. That tells us something important. The bottleneck in AI is no longer intelligence. It is capacity. For decades, technology companies competed for users. Today, even the largest technology companies are competing for compute. That is a fundamental shift. When a company with Google’s scale chooses to rent capacity rather than wait to build it, speed becomes more valuable than ownership. Google isn’t buying a model. It’s buying time. And that may be the defining economics of the AI era. The winners may not be the companies with the best models alone. They may be the companies that can secure compute capacity exactly when demand exceeds supply. In previous technology cycles, firms competed for customers. In the AI cycle, they are increasingly competing for capacity. I’m Hurratul, and this is The U Lab Daily Brief on venture capital, technology, and the future of innovation.