The Grid Isn't Running Out of Power. It's Running Out of Time.
For years, the conversation around America's electrical grid has revolved around one question:
How do we build more power?
The Wall Street Journal recently framed the challenge in stark terms. Utilities across the United States are preparing to invest roughly $1.4 trillion over the next five years as electricity demand accelerates, driven by artificial intelligence, hyperscale data centers, manufacturing, and electrification. At the same time, the cost of building new generation, substations, transformers and transmission infrastructure continues to rise, while permitting and equipment lead times stretch ever longer.
The numbers are staggering. But perhaps more important than the numbers is the question they don't ask.
What if we're solving the wrong problem?
Everyone Is Building Supply
Nearly every major investment announced today follows the same pattern.
Build another gas plant.
Build another substation.
Build another transmission corridor.
Build another battery installation.
Build another peaker plant.
Each of these investments has merit. Each is also expensive. Each takes years.
Meanwhile, electricity demand is arriving now. AI doesn't wait for permitting. Data centers don't pause while transformers are manufactured. Communities don't become more patient because transmission projects require another environmental review.
Infrastructure is becoming a race against time.
There Is Another Source of Capacity
Imagine if the fastest megawatt isn't one that's generated. Imagine it's one that no longer needs to be consumed. Every large commercial building already contains something remarkable.
Thermal mass.
Concrete.
Steel.
Water.
HVAC systems.
Building automation.
Occupancy patterns.
All of these represent stored flexibility that has largely been ignored. Instead of treating commercial buildings solely as energy consumers, what if they became active participants in balancing the grid? Not emergency demand response. Not simply energy efficiency. Something much more valuable.
Dispatchable electrical capacity. The next megawatt may already exist. It simply hasn't been orchestrated.
Buildings Could Become Infrastructure
For decades we've thought about infrastructure as physical assets.
Power plants. Transmission lines. Substations. Transformers.
But digital infrastructure has changed that thinking.
Software now controls logistics. Software controls financial markets. Software controls transportation. Why shouldn't software help create electrical capacity?
Instead of asking how quickly another substation can be built, perhaps we should also ask:
How many megawatts could be unlocked across existing commercial buildings within the next thirty days?
That question deserves far more attention than it's receiving today.
The AI Economy Changes Everything
Artificial intelligence isn't simply another large customer of electricity. It fundamentally changes the economics of time.
Every month a new data center waits for interconnection represents delayed revenue, delayed innovation, and delayed economic development. The fastest infrastructure project may not involve pouring concrete at all. It may involve orchestrating the buildings already connected to the grid.
A Different Way to Think
The energy industry has historically viewed efficiency as a way to lower utility bills. That certainly remains important. But perhaps efficiency is becoming something larger.
Perhaps it is becoming infrastructure.
Infrastructure that can be financed. Measured. Aggregated. Dispatched. Valued. Traded.
Integrated alongside traditional grid assets. That possibility is becoming increasingly difficult to ignore.
Parkwood Perspective
We believe the next decade will not simply be defined by building more electrical infrastructure.
It will be defined by making existing infrastructure dramatically more intelligent. The companies that succeed won't necessarily generate the most electricity. They will help the grid use what already exists more effectively. That's a very different conversation.
And we think it's only beginning.