AI's Power Crisis: A Finance Chief's Bold Move to Build the Future

AUSTIN, TX – August 18, 2026

On the surface, the announcement was standard corporate procedure: Poolside Infrastructure Company (PIC), a new developer in the tech ecosystem, has appointed Justin Lagasse as its Chief Financial Officer. But to read this as a simple executive shuffle is to miss the tectonic shift happening beneath the foundations of the entire artificial intelligence industry. Lagasse’s move from a $50 billion energy conglomerate to a nascent infrastructure startup isn't just a career change; it's a clear signal that the primary constraint on AI's future is no longer code, but kilowatts.

The race to build the next generation of frontier AI models has created an unquenchable thirst for electricity. PIC's stated mission is to support this demand, and with Lagasse, it has recruited a leader whose entire career has been steeped in the world of large-scale power and capital markets. This isn't just about balancing books; it's about financing the very power plants needed to keep the lights on for an industry that promises to reshape our world.

The Gigawatt Challenge

For years, the bottleneck in AI was computational power—the chips themselves. Now, as chip supply stabilizes, the true barrier is the energy required to run them at scale. A single large AI data center can consume as much electricity as a mid-sized city, and the industry's appetite is growing exponentially. A recent University of Houston study projected that data centers could increase Texas's statewide electricity consumption by up to 300% by 2035, creating a potential annual shortfall of 40 gigawatts.

This is the reality that gives weight to Lagasse’s own words on his departure from his former role. "The opportunity in front of us at PIC is generational and narrow," he stated. "The constraint on AI innovation and growth is the ability to firm power capacity at gigawatt scale, on an AI timeline." His phrasing is precise and revealing. "Firm power" refers to the constant, reliable energy that AI training requires, something intermittent sources can't guarantee. "AI timeline" means immediately—not in the five to ten years it can take to navigate the bureaucratic and physical logjams of connecting to a public power grid.

From a Utility Giant to an AI Upstart

To understand the gravity of PIC's strategy, one must look at who Justin Lagasse is. He is not a Silicon Valley CFO accustomed to software margins and venture capital rounds. He most recently served as CFO and Controller of Avangrid, an energy titan with a $50 billion asset base and a sprawling portfolio of regulated utilities and one of the nation's largest renewable energy businesses. At Avangrid, he executed billions in financing agreements and, tellingly, navigated the company's complex delisting from the New York Stock Exchange in 2025.

This background is not a coincidence; it’s a prerequisite. Building the infrastructure for AI is a game of staggering capital expenditure, and Lagasse's experience is in securing and managing the massive sums required to build and operate power infrastructure. His hiring is a direct appeal to investors, signaling that PIC understands the financial and logistical realities of its ambitious goals. As PIC CEO Robert Bonar noted, "With Justin leading and overseeing the company's financial operations, PIC is well equipped to deliver and help power frontier AI models." In essence, PIC has hired a specialist in building the engine, not just the car.

Project Horizon: A Vertically Integrated Bet in West Texas

PIC's plan to deliver on this promise is Project Horizon, a 400-megawatt data center campus taking shape on a 559-acre footprint in the vast expanse of unincorporated Pecos County, Texas. The choice of location is a strategic masterstroke. West Texas offers not only affordable land but, more importantly, proximity to the Permian Basin's abundant natural gas. This allows PIC to pursue a strategy of vertical integration, building its own "behind-the-meter" power generation to directly fuel its data centers.

This approach effectively creates a private grid, bypassing the congested and slow-moving interconnection queues of ERCOT, the Texas grid operator. While earlier reports hinted at a more expansive vision for Project Horizon, the company's current focus is on delivering this foundational 400 MW campus—a significant undertaking in its own right that reflects a pragmatic response to the immense challenges of such a build-out. It's crucial to distinguish Poolside Infrastructure Company, the 2026-founded developer led by Bonar, from a similarly named AI software firm. PIC's sole focus is this critical, capital-intensive infrastructure layer.

PIC is not alone in seeing West Texas as the answer. Energy and tech giants, from Amazon to Chevron, are also staking claims in the region, planning multi-gigawatt facilities powered by the same logic. The race is on to create these energy islands capable of sustaining the next wave of AI development.

The New Blueprint for Technological Supremacy

The hiring of an energy veteran like Lagasse by an AI infrastructure startup like PIC is not an anomaly; it's the new template. The future of AI hinges on this very fusion of expertise: big energy finance, complex infrastructure development, and cutting-edge technology. Companies that can master this convergence will hold a decisive advantage.

The vertical integration model offers speed, control, and cost certainty in a world where public infrastructure cannot keep pace with technological ambition. It is a pragmatic, if audacious, solution to the market's most pressing bottleneck. This strategy accelerates AI's advance, but it also concentrates immense power—both electrical and economic—into private hands, raising profound long-term questions about resource management and environmental stewardship in regions like West Texas. For now, it appears to be the only path forward for an industry whose thirst for energy is just beginning.