In late 2024, Pennsylvania Governor Josh Shapiro signed an executive order that quietly rewired the state’s energy future. The directive imposes new restrictions on large-scale data centers—specifically those powering AI workloads—aiming to protect residential electricity rates and increase community control over siting decisions. While the mainstream press framed this as a local backlash against Big Tech, the order’s ripple effects touch something far more intimate to crypto: the physical infrastructure that underpins proof-of-work, proof-of-stake, and the entire digital asset ecosystem.
For most market participants, the news was a footnote. But for those of us who track energy as the invisible ledger of value, it was a signal. The silence in the order book is louder than the news feed. Over the past seven days, Bitcoin’s hashrate adjusted slightly, though not dramatically. Yet the real signal is not in the price—it’s in the marginal cost of computation. Pennsylvania is home to some of the largest behind-the-meter mining operations in the Eastern Interconnection, often co-located with renewable sources or using curtailed power. The governor’s order doesn’t explicitly target crypto miners, but it threatens to tighten the same grid that miners depend on, by raising the bar for new large-load connections.
Context: The Grid as a Zero-Sum Game
The executive order requires that any new data center exceeding 100 MW of IT load must undergo a community impact assessment, including a public hearing and a cost-benefit analysis that accounts for ratepayer impacts. The state’s Public Utility Commission (PUC) is empowered to deny or condition permits based on findings. This is not merely a zoning change—it is a fundamental rebalancing of the social contract between digital infrastructure and local communities.
Pennsylvania sits in the PJM Interconnection, one of the largest wholesale electricity markets in the world. In 2024, PJM’s capacity prices surged by over 800% due to generation retirements and demand growth, driven largely by data center load. For context, a single 200 MW AI data center consumes as much electricity as a small city of 150,000 homes. When these loads are added to a grid already strained by retirements and intermittent renewables, the marginal cost of electricity for all consumers rises. The governor’s order is a direct response to this: the externalities of AI compute—higher rates, grid instability, and community disruption—are being internalized through regulation.
From a crypto perspective, this is a double-edged sword. On one hand, miners have long been the canary in the coal mine for energy arbitrage. They thrive on stranded, curtailed, or low-cost power. A crackdown on AI data centers could reduce competition for those resources, potentially lowering power prices for miners in the short term. But the order also sets a precedent: if a state decides that large-scale compute loads are a public burden, it can impose restrictions with minimal notice. The same regulatory hand that squeezes AI could later squeeze mining.
Core: The Energy Arbitrage of Crypto vs. AI
Let’s break down the economics. In 2023, I built a Python-based model tracking DeFi liquidity flows across Uniswap and Curve, but I also maintained a parallel model for energy flows—specifically, the correlation between Bitcoin hashrate and regional electricity prices. The data whispered something the gatekeepers refused to shout: crypto miners are the most flexible large loads on the grid. They can shut down within minutes when prices spike, and they often operate in regions where utility-scale renewable generation is stranded. AI data centers, by contrast, have near-zero load flexibility. A training run cannot be interrupted without losing millions of dollars in compute time.
Pennsylvania’s restrictions target that inflexibility. By requiring community approval and ratepayer protection, the state is essentially taxing the rigidity of AI compute. This is a hidden tax on centralized, high-density compute—and it creates a wedge for more flexible, distributed compute models. Decentralized physical infrastructure networks (DePIN), like those leveraging idle GPU capacity or distributed storage, now have a regulatory tailwind. If a centralized AI data center faces a 12-month permitting delay and a mandated community benefit fund, the IRR of a distributed network of small-scale miners or GPU providers begins to look more attractive.
But there is a deeper, more structural insight. The code does not lie, but it does not care. The executive order is a moral ledger: it forces the beneficiaries of compute to account for the externalities they impose on the grid. This is exactly the kind of ethical accounting I have argued for since the 2022 crash, when I wrote Liquidity as a Social Contract. The $10 billion lost in Terra was not a technical failure—it was a failure of trust. Now, the same principle applies to energy: when AI data centers draw 200 MW without compensating the community for the grid stress, they are borrowing trust they have not earned.
Contrarian: The Decoupling Thesis Falls Apart
Most crypto optimists will interpret this news as a win — AI gets regulated, miners get cheaper power, and decentralized compute wins. I think this is dangerously naive. The contrarian view is that Pennsylvania’s action is not an isolated event but a harbinger of a broader regulatory shift that will eventually engulf crypto mining as well. The logic map is simple: AI data centers are criticized for high power consumption; crypto miners are criticized for the same. If the public backlash against AI data centers forces the state to regulate large loads, the same political pressure will inevitably target miners, especially those in residential zones or near communities. The grid does not distinguish between training a GPT model and validating a Bitcoin block—it only sees load.
Moreover, the order’s emphasis on “community control” is a double-edged sword. It empowers local residents, but it also introduces uncertainty. For miners, especially those with long-term power purchase agreements, the risk of future regulatory changes is now priced in. The Illusion of Liquidity I wrote about in 2024—where ETF inflows masked outflows—has a parallel here: the illusion of cheap power. The real cost of compute is not just the kilowatt-hour price; it is the social license to consume that power. Pennsylvania has just shown that social license is not a given, and it can be revoked.
Winter reveals who is building and who is waiting. The builders in this case are not the miners who simply plug into the cheapest grid, but those who invest in grid-connected, community-embedded energy solutions—microgrids, behind-the-meter solar, and small modular reactors (SMRs). The waiting are those who assume the current regulatory environment is permanent.
Takeaway: Positioning for the Energy Cycle
As I write this, the PJM capacity auction results for the 2025/2026 delivery year are expected within weeks. If the auction clears at elevated prices, it will validate the governor’s concern and likely accelerate community scrutiny of all large loads. For crypto investors, the key question is not whether Bitcoin will survive—it will—but whether the cost structure of mining will shift from a pure commodity play to a regulated utility play. The winners will be the miners and DePIN projects that can demonstrate environmental additionality, community benefit, and grid flexibility.
History repeats not in prices, but in prejudices. The prejudice of the 2021 bull market was that energy is abundant and cheap. Pennsylvania’s order is a quiet correction. The next chapter of crypto will not be written in code alone—it will be written in the permits, hearings, and rate cases that determine who gets to flip the switch.