“Data centers should be a way to make electricity cheaper, not a way to make electricity more expensive”
Robert Boswall with Jordan Nanos (@RobertBoswall @JordanNanos) / SemiAnalysis Something Else Weekly — Ep. 026
The conventional story about data centers and energy is one-directional: AI demand grows, grids strain, prices rise. That framing is accepted as gravity.
But on a recent episode of SemiAnalysis’s Something Else Weekly, Robert Boswall and Jordan Nanos push back hard on that assumption — and the pushback is worth sitting with.
⚡ The Structural Tension
Data centers are large, predictable, controllable loads. In theory, that profile is exactly what a grid operator wants — a customer you can schedule, curtail, and integrate efficiently. The question Boswall and Nanos are raising: why isn’t that theoretical advantage showing up in the bill?
“Data centers should be a way to make electricity cheaper, not a way to make electricity more expensive.”
📐 The Structural Read — FourWeekMBA Analysis
This is a Map of AI moment at the infrastructure layer. The AI stack runs deep — chips, data centers, power — and decisions made at the grid-planning level ripple upward into every modelcost and every inference price. If grid economics are miscalibrated at the base, the distortion compounds.
The Boswall-Nanos argument implies a design failure, not a demand failure. Big load is not inherently expensive load. The question is whether procurement, interconnection, and rate structures are built to capture that distinction — or whether blunt modeling is erasing it.
💡 Why Ratepayers Are in the Frame
The episode title references PJM — one of the largest grid operators in the world — and a modeling error allegedly costing ratepayers billions. The implication in Boswall and Nanos’s framing: the people paying residential electricity bills may be absorbing costs that could have been structured differently. That is a political and economic problem, not just a technical one.
🔍 The Opportunity Cost Read
If data centers were properly integrated as flexible, dispatchable demand, they could theoretically smooth grid peaks and reduce marginal cost pressure. The gap between that potential and the current outcome is where the policy and regulatory argument lives. Boswall and Nanos are pointing at that gap — and it’s a large one.
The Bottom Line
The AI infrastructure build-out doesn’t have to be a zero-sum trade between compute growth and affordable electricity — but making it additive requires grid planning, rate design, and regulatory frameworks to actually keep up. Boswall and Nanos are arguing that right now, they aren’t. That’s a solvable problem. The question is whether it gets solved before the bill lands.
Clip via Robert Boswall with Jordan Nanos (@RobertBoswall @JordanNanos) / SemiAnalysis — Something Else Weekly — Ep. 026 – PJM’s $12B Modeling Mistake Is Hitting Ratepayers Again (Datacenter, Energy).
This is editorial analysis by FourWeekMBA, not investment advice. The views expressed in the quoted material represent the speakers’ argument as articulated in that episode.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
