As investors race to identify the next winners in artificial intelligence, George Rockett, founder of Data Center Dynamics and the Yotta conference, argues that the biggest opportunities may not lie with AI model developers themselves, but with the companies supplying the infrastructure required to make AI possible.
Speaking with Adam Shapiro on AInvest’s Capital & Power, Rockett described the AI boom as a modern-day “picks and shovels” story, echoing the California Gold Rush analogy often used by infrastructure investors. “When you see headlines of trillions of dollars being spent, they're being spent on picks and shovels,” Rockett said. “They're being spent on generators and thousands of kilometers of cabling and fiber optics and roofs and buildings and concrete."
His comments come as spending on AI infrastructure reaches unprecedented levels. Alphabet, Amazon, Meta and Microsoft are expected to collectively invest roughly $650 billion in AI-related infrastructure in 2026, a sharp increase from approximately $410 billion in 2025. The companies are building out their AI infrastructure, including data centers, chips and networking equipment. Goldman Sachs has projected that AI-related capital expenditures expenditures could reach $1.1 trillion by 2027, with an upside scenario approaching $1.4 trillion.
Rockett believes many investors remain focused on highly visible AI companies while overlooking the broader supply chain supporting the buildout. The smaller companies that supply that AI construction boom attend Yotta 2026.
“You've got people making chillers, you've got people making batteries,” he said. “Just at the event in Las Vegas, we've got already 250 exhibitors, people that make items.” He added that many suppliers are reporting “100% growth year on year” as data center construction accelerates.
Among the publicly traded beneficiaries, Rockett highlighted power and cooling suppliers such as Vertiv and Schneider Electric, describing them as major participants in the infrastructure supply chain supporting hyperscale AI deployments.
The spending surge is already producing new financing structures. Reuters reported that AI startup Anthropic is pursuing data-center leases and seeking financial backing from Google to support its growing infrastructure needs and that Amazon recently secured a $17.5 billion loan facility to help fund its expanding AI infrastructure investments.
Rockett argues that these developments illustrate a broader shift in how investors should think about AI.
“If you're serious about investing in the picks and shovels, everything that sits behind AI infrastructure, this is an event that you should be at,” he said. “You can actually meet the people that are involved in this. You can listen to them and you can make up your own mind with primary evidence.”
Looking ahead, Rockett said the next wave of capital could become even larger if anticipated public offerings materialize.
“We've got OpenAI, we've got Anthropic and we've got SpaceX coming up,” he said. “That liquidity coming into the market for them as public companies, that's gonna rain money down on this infrastructure level.”
For investors searching for AI exposure beyond software and semiconductor giants, Rockett’s message was clear: the most consequential opportunities may be found in the physical infrastructure powering the AI economy.