Inside the Race to Build AI Infrastructure w/ Sheldon Kimber

2025-11-25

Intersect is solving AI's fundamental constraint: gigawatt-scale power delivered outside traditional grid infrastructure. The company builds behind-the-meter renewable and gas generation paired directly with data centers in locations like West Texas, bypassing the Balkanized regulatory system that makes grid expansion nearly impossible. Currently constructing $9 billion in assets across multiple sites and breaking ground on $10 billion more in 2025, Intersect represents the emerging infrastructure layer that neither traditional utilities nor...

Topics Discussed:

Why the US grid's Balkanization across state, federal, and regional system operators makes it structurally unfixable
The business model shift from utility PPAs to operating in deregulated commodity energy markets
Intersect's hybrid model: flexible reciprocating engines and aeroderivative turbines paired with renewables
Why AI reasoning models and larger context windows guarantee sustained electricity demand growth despite efficiency gains
The capital formation playbook: hundreds of thousands for early development, billions in non-dilutive project finance post-contract
How single gigawatt data centers translate to $50-60 billion all-in infrastructure when accounting for power, shell, turnkey, and chips
The AI infrastructure stack battle: Neo clouds pushing down, power companies pushing up, and who owns the constraint

GTM Lessons For B2B Founders:

Design around regulatory arbitrage, not reform: Intersect's entire business model targets unregulated spaces. Sheldon rejected the conventional utility PPA path, stating: "We built a business model that is not go right at something." Their behind-the-meter, off-grid approach operates outside state public utility commissions, FERC jurisdiction overlaps, and ISO/RTO coordination failures. When entering regulated markets, map the jurisdictional gaps and build your product to live in those spaces rather than fighting for reform that won't come.
Bet on technological disruption of calcified systems: Sheldon explicitly compared the grid to the Bell System, which wasn't reformed but rendered irrelevant by cellular, fiber, and digital infrastructure. He observed: "Nobody inside the regulated telecom industry said, let's fix this. Brand new technologies just completely gutted it from the inside out." When facing entrenched infrastructure with misaligned incentives across fragmented stakeholders, identify the technological bypass rather than incremental improvements. The constraint that makes reform impossible also creates the asymmetric opportunity.
Optimize for project scale, not team expansion: Intersect maintains radical efficiency by building only gigawatt-scale projects with lean teams. Sheldon's framework: "It takes the same number of people in contracts to build a gigawatt plant as it does to build a hundred megawatt plant, so you might as well build the bigger one." This isn't generic "do more with less"—it's identifying which costs are fixed regardless of project size (legal, interconnection applications, permitting) versus variable (construction materials). B2B founders should analyze their own cost structure to find similar leverage points where deal size scales independent of resource requirements.
Lock major contracts before infrastructure deployment: Intersect's capital strategy starts with hundreds of thousands for land options and interconnection applications, uses early reputation to secure anchor contracts, then raises billions in non-dilutive project financing against those contracts. Sheldon described it as: "Making enterprise software where your first sale was to 10 Fortune 500s and they took everything you could possibly build for the next five years." The contract becomes the collateral for low-cost debt financing. For capital-intensive B2B models, structure early sales to de-risk the bulk of your capital deployment.
Identify where constraints migrate in your value chain: Sheldon analyzed the AI infrastructure stack from dirt to chips and concluded: "For the next decade, there'll be a constraint in the physical infrastructure. It'll need cloud providers that have access and expertise to build infrastructure and achieve capital formation around infrastructure." Value accumulates at bottlenecks. Neo clouds like CoreWeave started at the model layer and are pushing down; Intersect started at power and is pushing up. Position your product where the constraint is moving, not where it currently sits. This requires continuously modeling your industry's supply chain physics.
Let unit economics override industry narratives: While energy policy debates favor nuclear and central-scale combined cycle gas plants, Intersect's math shows their hybrid approach—flexible reciprocating engines and aeroderivative turbines with renewables and batteries—delivers power at a fraction of the cost. Sheldon noted alternatives run "$120-150 a megawatt hour versus most people paying $30-40." He emphasized: "It's just math and it's not hard math." Strip away conference circuit consensus and regulatory capture. Model actual delivered unit economics, then let superior economics compound into strategic advantage while competitors chase policy-driven narratives.

Sponsors:

Front Lines — We help B2B tech companies launch, manage, and grow podcasts that drive demand, awareness, and thought leadership. www.FrontLines.io

The Global Talent Co. — We help tech startups find, vet, hire, pay, and retain amazing marketing talent that costs 50-70% less than the US & Europe. www.GlobalTalent.co

Don't Miss: New Podcast Series — How I Hire

Senior GTM leaders share the tactical hiring frameworks they use to build winning revenue teams. Hosted by Andy Mowat, who scaled 4 unicorns from $10M to $100M+ ARR and launched Whispered to help executives find their next role.

Subscribe here:

https://open.spotify.com/show/53yCHlPfLSMFimtv0riPyM

Comments (3)