The Real Story Behind AI’s Energy Crunch
Energy access has been the story of humanity's evolution for centuries now
Why This Macro Dive Matters
Energy affects everything we do: how we travel, how we get goods, how we produce, and even how we grow food. If you have not really considered the importance of energy, or how it has been the single greatest accelerant to society and economic progress, now is the time to pay attention. I believe we are entering a period where energy is going to become the focus of nearly every major conversation, simply because it is a critical ingredient to everything, and today’s demand is starting to outpace our capacity.
The Mid-Atlantic Energy Price Spike: A Wake-Up Call
Electric bills in Maryland, New Jersey, and Pennsylvania are spiking. In some places, rates are up by double digits, leaving both households and businesses wondering what’s going on. It’s easy to blame the usual suspects: rising fuel costs, inflation, even the weather. But this time, there’s a new player on the grid: artificial intelligence.
Behind the scenes, massive new AI data centers are being built across the Mid-Atlantic. They aren’t just another server farm. They are power-hungry behemoths designed to train and run the next wave of generative AI models. This year, PJM, the electric grid operator for the region, saw capacity auction prices surge by more than 800%. The result? Consumers are on the hook for nearly $15 billion in capacity costs for 2025/26, up dramatically from just $2.2 billion the year before. The main culprit, according to energy analysts and local officials, is the rapid buildout of AI infrastructure.
This isn’t a temporary blip or a local problem. It is a preview of the new reality for all of us as AI adoption accelerates. So, what’s really going on and why does it matter? Here are five things everyone should know about AI and the coming energy crunch.
1. AI Is Driving Electricity Prices Higher—Fast
Those Mid-Atlantic price spikes? They are just the beginning. The demand for electricity from AI data centers is rising so quickly that it is straining local grids, forcing utilities to pay higher prices for capacity and, ultimately, passing those costs on to consumers.
What is different about this moment is the scale and speed. When a company builds a next-gen AI data center, it might draw as much power as a small city. Multiply that by dozens of projects cropping up nationwide, and you have a recipe for double-digit price increases that hit regular folks, not just big tech companies.
2. AI Data Centers Are on Track to Double U.S. Power Consumption
AI is not just another application running in the cloud. Training large models and running them (sometimes millions of times per day) takes an enormous amount of electricity. Nationally, experts predict that U.S. data center power consumption could more than double by 2035, reaching 78 gigawatts. That is the equivalent of adding dozens of new cities’ worth of demand to the grid in just over a decade.
Some estimates suggest that AI data centers alone could eventually account for up to 21% of global energy demand. Between the hardware, the cooling systems, and the non-stop computations, AI is changing the entire landscape of energy consumption.
3. There’s No Simple Solution: Nuclear Helps, but It’s Not Enough
So how do we meet this runaway demand? There is no silver bullet. Yes, I personally think new-age nuclear power is a huge part of the answer, but even if we greenlight every proposed plant today, it will take years (if not decades) for them to come online.
Renewables like solar and wind are growing fast and can provide low-cost, clean power. They require massive investments in battery storage and grid upgrades to handle variable supply. Meanwhile, there are promising advances in energy efficiency. New AI chips use less power, smarter software can optimize data centers, and some AI models are designed to run "leaner." But these improvements cannot erase the sheer scale of what is being added.
In summary, meeting AI’s energy appetite will take a combination of nuclear, renewables, storage, efficiency, and serious investment in our grid infrastructure.
4. The Link Between Energy and Economic Growth Is Changing, but Not Gone
For decades, economic growth and energy use rose together. In the U.S. and other wealthy countries, we have started to "decouple," using less energy for each dollar of GDP, thanks to efficiency, cleaner generation, and a shift to a more service-based economy.
Since 1990, U.S. energy per dollar of GDP has dropped by 46%. Globally, energy intensity keeps falling by about 1% per year. But here is the kicker: even as we have gotten more efficient, electricity demand is starting to rise again because of AI. In emerging markets, economic growth still relies heavily on more energy access.
In other words, we can make progress on efficiency, but the relationship between energy and economic prosperity is still alive and well, especially as new technologies like AI ramp up demand.
5. This Will Shape Our Economic Future, and We Need to Start Talking About It Now
Most people have no idea how much AI is poised to reshape the energy conversation or how that could affect everything from utility bills to the pace of innovation and economic growth. If we do not start planning for this now, energy could become the bottleneck that slows down not just AI, but the broader economy.
There is no single, easy answer. We need all hands on deck: faster deployment of renewables, a real look at nuclear, smarter grids, more efficient technology, and a willingness to have tough, honest conversations about trade-offs.
Final Thought: Do Not Let Energy Become the Story After the Fact
Energy is about to become a front-page story again, thanks in no small part to AI. If you are in business, technology, or just paying an electric bill, now is the time to start paying attention and asking the right questions about where our power will come from in the AI age.
For anyone who wants to dig deeper, I have included a few links below. If this resonates, please share with others who care about the intersection of technology and the real world. The conversation is just beginning, and the stakes could not be higher.
Further Reading: