AI data centers face power shortages, leading to a reliance on off-grid natural gas, but financial misalignment stalls projects.
José Ortega identifies the financial architecture of partnerships as the main barrier to launching off-grid AI data centers.
A cultural shift towards collaboration is needed to align interests across power, infrastructure, and hardware sectors.
Ortega argues that relational innovation, not just technological, is essential for scaling off-grid AI data centers.
The U.S. tech industry must overcome internal financial hurdles to compete with state-sponsored AI initiatives globally.
The biggest bottleneck to actually getting these projects off the ground is aligning on the cap table with the different groups that now need to work together.
The AI boom has triggered a compute arms race, and it runs on electricity. As power demands surge, national grids are straining, forcing a search for alternatives. Off-grid natural gas, once a backup, is now a leading candidate to power new AI data centers. But across the U.S., these projects are stalling before they start. The issue isn’t fuel or tech. It’s the deal itself.
The insight comes from José Ortega, an energy and tech innovator whose career has placed him at the center of this collision. As a General Partner at FO Permian Partners, a firm that builds off-grid power sites for digital infrastructure, and with a background that spans both traditional energy and disruptive tech, Ortega has a unique perspective on the friction between these two worlds. According to him, the roadblock isn’t in the oilfield or the server farm; it’s in the financial architecture of the partnerships themselves.
The cap table bottleneck: “The biggest bottleneck to actually getting these projects off the ground is aligning on the cap table with the different groups that now need to work together,” Ortega states. An off-grid AI data center isn’t a single entity, but a fragile alliance between three distinct value chains: the power generators, the digital infrastructure builders, and the GPU hardware providers. Getting them to agree on a shared financial structure is where most deals die.
Anatomy of a deal: “By definition, a healthy cap table is where all the stakeholders feel that their return on capital is being met, and there’s alignment across all three value chains: the power generation, the hosting facility, and the hardware itself,” Ortega explains. But he stresses that the mechanics are more human than financial. It’s about ensuring every partner feels secure.
“A healthy cap table means the power provider is comfortable with their PPA, the co-location partner trusts the deal structure, and the hardware investor knows their costs make sense for delivering compute under their cloud contracts—so tech companies can actually build the products we’re all counting on,” says Ortega.
You have certain countries that have one cap table. The government just spends, they do what they need to do, and they get computation in line. We really need to get out of our own way as independent free capitalists and work together to unlock that value and grow.
A cultural chasm: Achieving that comfort requires a “cultural shift” that has proven difficult for industries accustomed to operating in silos. Ortega points to the ingrained mindset of traditional data center operators as a key hurdle. “If you’re a co-locator, you’re not thinking about partnering with new power generators or independent producers. You’re used to simply buying power off the grid to run your business.” This old-world thinking stands in stark contrast to the collaborative models pioneered by companies like Crusoe Energy, who has successfully bridged the gap to build major AI infrastructure.
Dealmakers to the rescue: This challenge reframes the very definition of innovation in the AI era. While the industry obsesses over the next-generation chip or algorithm, Ortega argues the most critical breakthrough required is relational. “Really getting everyone to play together and getting everyone to underwrite the value together is what is keeping the off-grid opportunity from really scaling and advancing.” The heroes of this next phase may not be just engineers, but expert dealmakers who can build these complex, multi-party alliances.
A national call to arms: Ortega’s vision is rooted in a pragmatic, balanced perspective. He is quick to credit the existing system, noting, “The way that the grid balances power and enables demand response is a very healthy thing. You want to take care of cities.” But for the specific, immense needs of AI, he argues a parallel path is a national imperative.
The internal struggle to align business interests becomes a critical security issue when compared to the efficiency of state-sponsored rivals. “You have certain countries that have one cap table. The government just spends, they do what they need to do, and they get computation in line,” he warns. “We really need to get out of our own way as independent free capitalists and work together to unlock that value and grow.” Without that shift, Ortega believes the American tech industry is squandering a massive opportunity. “I feel like they’re dropping the ball.”
