CapitalStack: Introducing the AI Lab
A weekly series on the AI tools and workflows reshaping real estate capital raising — starting with how to build an investor database in hours, not weeks.
AI, power constraints, hyperscalers, and urban inference are rewriting the rules
One of the country’s most experienced data-center developers says he likes to open his interviews with LPs and other real estate professionals looking to pivot to his sector with a deceptively simple question:
“What’s a data center?”
He pauses. “I can tell you: Nobody knows.”
A touch of exaggeration? Maybe. But some confusion is understandable in a sector that in recent years has grown faster, and morphed more completely—in scale, governmental policies, underwriting strategies, even form factors—than any other corner of commercial real estate. In the span of a decade or so, data centers have gone from a niche industrial subset to the physical backbone of AI and national economic strategy. And yet many real estate investors are still on the sidelines, or approaching this asset class with frameworks borrowed from office, industrial, or basic infrastructure.
Those frameworks no longer apply.
Today, data centers are a splintered, high-velocity ecosystem shaped by the towering needs of AI and cloud computing, defined by power constraints and increasingly complex geopolitics. On one end of the scale are the hyperscale campuses: multi-hundred acre compounds, often in rural or remote suburban markets, devoted to training AI models or cloud computing. On the other: sub-50MW inference-driven AI sites proliferating in more populated areas in and around cities..
Rapid developments have created real opportunities for investors—but also risks and the need to learn new models to approach capital markets. So we decided to pull together a new playbook for investors: the data center market shifts they need to know. We reached out to experts on the operational and advisory sides to gain some insight into the state of play in this asset class:
• Daniel English, Managing Partner at Legacy Investing, a veteran data center developer/operator whose projects span hyperscale campuses and urban inference hubs
• Ivo van Breukelen, Co-Founder & Managing Partner at The Proptech Connection, a global advisory firm that tracks data centers and digital-infrastructure trends
In this letter, we’ll cover:
One of the fastest-growing parts of the built environment now behaves more like civic infrastructure than traditional real estate. AI workloads, industrial digitization, and surging compute demand have pushed data centers into the same category as ports and power plants—assets that set a region’s economic ceiling.
More cities now treat them that way. Planning departments evaluate grid impact, sequencing, and long-term public value before they look at land use. As van Breukelen of The Proptech Connection notes, “data centers are now deeply coupled with electricity grids,” and that coupling dictates siting, permitting, and growth.
Operators have watched the shift up close. “They were always considered a niche asset class, like a marina or an RV campground,” says Daniel English of Legacy Investing, who has been in the data-center business for more than 15 years. “They were never considered core real estate or core infrastructure. Obviously, that shifted.”
The implication for investors: underwriting a data center now means underwriting a region’s energy governance, policy direction, and digital ambitions. Compute capacity has become civic capacity, and investor models need to start there.
Gigantic hyperscale campuses, usually located in low-density exurban or rural areas, still dominate headlines and the public imagination, but a significant restructuring and expansion of the asset class is now happening inside and around cities too.
Covering the future of real estate and the people creating it