Brookfield's Flatt: AI Race Is Slowing as Builders Fall Behind
TL;DR
- Brookfield CEO Bruce Flatt told the firm's annual investor day the AI race is already slowing because developers cannot build what customers demand.
- Flatt said the industry 'cannot build enough power' or enough compute, framing the bottleneck as physical rather than a capital shortage.
- Brookfield estimates AI infrastructure will need more than $7 trillion of capital over the next decade, and cast a slower rollout as useful spending discipline.
Bruce Flatt's message to Brookfield's annual investor day, per Bloomberg: the AI race is already decelerating, and not because anyone chose to slow down.
"We as an industry can't build enough," the Brookfield chief said. "We can't even build a fraction of what everyone thinks they need."
Asked whether too much capital was chasing AI infrastructure, Flatt turned the question around. "We cannot build enough power. We cannot build enough compute." The constraint he described is physical, not financial. Power first, then compute.
Brookfield has still put a headline figure on the demand: the firm estimates AI infrastructure will need more than $7 trillion of capital over the next decade. On the supply side of its own book, The Logic reported an additional US$2 billion from Nvidia going into a Brookfield AI infrastructure fund, alongside Flatt's line that his team is being "really, really choosy" about which data-centre deals it takes.
A slower rollout, in Flatt's telling, could benefit the sector by tempering the spending. The industry, he added, is still "just scratching the surface" of what productivity gains AI can unlock. It is a striking pitch from a firm that has just anchored its next decade around that same $7 trillion number.
Originally reported by bloomberg.com
Read the original article →Original headline: Brookfield's Bruce Flatt Tells Investor Day AI Race Is Slowing Because Builders Can't Keep Up