AI Industry Driven to Build, Spend, Repeat

Microsoft plans to triple its data center power from around 12 gigawatts today to more than 38 gigawatts by 2032, making it one of the largest consumers of electricity in the world. This growth is part of a broader trend in the tech industry, with many companies expanding their data center capacity to meet increasing demand for AI computing.

Oracle reported a significant increase in revenue and cloud infrastructure revenue, more than doubling its quarterly earnings. The company spent $28.5 billion on capital projects, but only burned about $5 billion of its own cash because customers prepaid for the services. This shift towards prepayments is a key part of Oracle's strategy to reduce debt and reliance on its balance sheet.

SpaceX is also expanding its data center capacity, renting out AI capacity by the billion a month while reorganizing how it builds its infrastructure. These moves are all part of what the author calls 'business as usual' in the AI industry, with companies scrambling for data center capacity and spending more to get more.