AI Runs on Electricity. Electricity Runs on Reality 📱

Why It Matters

The split between deregulated and cost‑plus electricity markets dictates price dynamics that directly affect AI‑driven load growth, making market‑structure awareness vital for investors seeking stable returns in the energy sector.

Key Takeaways

• U.S. electricity market split: 42% deregulated, 58% regulated cost‑plus.
• Demand growth now ~1.5‑2.5% annually, driven by data centers and reshoring.
• Competitive states see 30‑35% wholesale price rise; regulated states flat.
• Deregulation created profit incentives but also price volatility for allocators.
• Understanding market structure is crucial for family offices investing in energy assets.

Summary

The interview explores how artificial intelligence’s soaring electricity demand intersects with the United States’ bifurcated power market. Roughly 42% of demand resides in deregulated states where generation competes in wholesale markets, while the remaining 58% operates under a traditional cost‑plus, vertically integrated model.

Jim Murchie explains that after two decades of flat demand, electricity usage is now rising 1.5‑2.5% per year, propelled by new data‑center construction and a reshoring of manufacturing. In competitive states, this modest demand uptick has already pushed wholesale prices up 30‑35%, whereas regulated states have seen prices remain essentially flat because fixed‑cost structures dampen price movements.

A striking example cited is AES’s investor deck, which shows that adding data centers can actually lower residential rates in a cost‑plus system by expanding the denominator of total kilowatt‑hours. Conversely, in a market‑driven environment, higher demand translates directly into higher wholesale prices, a nuance that most public commentary overlooks.

For allocators and family offices, grasping the distinction between these market regimes is essential. It determines where risk‑adjusted returns can be captured, how hedging strategies should be structured, and which policy shifts could reshape the profitability of energy‑linked investments.

Original Description

AI Runs on Electricity. Electricity Runs on Reality

The AI trade is being priced like a software story.

That is the mistake.

AI is not only a model, a chip, or a data center. It is a claim on the physical economy. It requires firm power, transmission, gas supply, cooling, land, transformers, turbines, permits, regulators, pipelines, and a grid that was not built for this load profile.

For two decades, the United States treated flat electricity demand as normal. Investors built assumptions around it. Regulators built rate structures around it. Utilities operated inside it.

Now that regime is ending.

AI data centers are creating industrial-scale load growth. Nuclear is back because reliability is back. Natural gas is unavoidable because time matters. Pipelines are being revalued from “old energy” to strategic infrastructure. Batteries are useful, but not magic. Carbon credits are being exposed as accounting in a world that needs electrons. And Hormuz reminds us that energy security is not a slogan. It is geography, shipping, fuel, storage, insurance, military power, and political will.

This Tuesday, June 2nd at 4pm ET, live across my socials, I sit down with Jim Murchie, Co-Founder, Co-Portfolio Manager and CEO of Energy Income Partners.

This conversation matters for serious investors because the AI boom is creating a new capital stack beneath the digital economy.

The questions are no longer theoretical:

• Is the binding constraint on AI now electricity, not chips?

• How much projected data-center demand is real, and how much is phantom load sitting in multiple interconnection queues?

• If hyperscalers require massive transmission upgrades, who pays: Big Tech, the utility, the ratepayer, or the investor?

• Are residential customers about to subsidize the AI economy through regulated utility bills?

• Which markets are better positioned for load growth: vertically integrated regulated systems or competitive power markets that price scarcity?

• Can solar and batteries truly support the AI buildout, or does the math break when the grid needs firm power across time, weather, and seasonality?

• Is natural gas the only scalable bridge for AI load this decade?

• Are hyperscalers quietly becoming energy companies while publicly marketing decarbonization?

• Is nuclear now AI policy, not climate policy?

• Are pipelines being mispriced as declining fossil assets when they may be the physical layer beneath AI, LNG, power reliability, and geopolitics?

• Are carbon credits real decarbonization, or reputation management?

• What would oil at $100, $125, or $150 do to capital allocation, inflation, private markets, and family office portfolios?

• What does Hormuz reveal about the fragility of global energy assumptions?

Why should billionaire families and long-duration investors care?

Because this is not just an energy conversation.

It is an asset allocation conversation.

If electricity demand is entering a new growth regime, then the market is mispricing duration, infrastructure, utility rate base growth, midstream cash flows, turbine scarcity, nuclear optionality, and the value of dispatchable generation.

If AI load is overstated, some parts of the power complex will be overbuilt.

If AI load is understated, reliability becomes the most valuable product in the economy.

Either way, the winners will not be the investors who repeat the AI narrative. The winners will be the investors who understand the constraint behind the narrative.

Software scales in the abstract.

Electricity scales through steel, copper, concrete, fuel, permits, balance sheets, and time.

That is where the real opportunity sits.

Live across Angelo’s socials.

Tuesday, June 2nd @ 4pm ET.

The Membership That Replaces Everything Else: Most observers study billionaire families. Angelo operate inside their command centers.

https://www.angelorobles.com/membership

#AngeloRobles #FamilyOffice #AI #Oil #Iran #JimMurchie