Art Of Boring

In this episode, we explore the growing electricity demands of data centers stemming from artificial intelligence (AI) with Chris Silvestre, Analyst on the U.S. Equity Team. He shares insights from a recent research trip visiting data centers in Northern Virginia—the data center capital of the world—and discusses challenges around meeting demands given constraints of land availability, energy generation and transmission, and the significantly increased power demands required to develop large language models.   Key points from this episode:   • Electricity demand in the U.S. has been relatively flat for a decade but is expected to increase 25% by 2050 due to electrification, ESG goals, and data center expansion to meet AI demand. • For the same footprint and the same data center, you need 10 times more power for AI workloads than you did for the old-style data centers. • Data centers cluster organically in hubs, such as Northern Virginia, to minimize data transfer costs, but this clustering strains power infra