The most honest sentence in the recent Fox News coverage of Trump’s AI data center push was buried in the middle: “Most Americans oppose having a data center in their community.” The rest was political theater. Jobs. Tax revenue. Energy independence. The same script every infrastructure booster has used since the interstate highway system. But this time, the underlying asset is not concrete and steel. It is compute. And compute, unlike a road, does not wear out with use. It becomes obsolete. The code doesn’t care about the ribbon cutting.
I measure risk in gas units, not in hope. And what I see in the current political alignment around AI infrastructure is not a rational market signal. It is a structural pre-mortem playing out in slow motion. The Trump administration’s public stance—welcome AI data centers, give them tax breaks, fast-track permits—is a policy signal that will reshape the geography of compute for the next decade. But the narrative being sold to local governments is built on a foundation of unverified assumptions. The jobs are temporary. The tax revenue is often cannibalized by incentives. The electricity demand is real, but the grid is not ready. And the public opposition is not noise; it is data.
Let me be clear: I am not arguing against AI data centers. I am arguing against the fantasy that they are a pure public good. The same reasoning applies to Bitcoin mining facilities, which the same political forces have treated with suspicion. The difference is not technical. It is narrative. AI is the new infrastructure darling. Crypto is still the outlaw. The fork was inevitable; the error was optional.
Context: The Infrastructure Hype Cycle
AI data centers are the new oil wells. They consume vast amounts of electricity, require specialized cooling, and generate enormous heat—both literal and metaphorical. The industry has been growing at a compound annual rate of over 20% since 2020, driven by the insatiable appetite of large language models and generative AI. But the growth has been concentrated in a handful of markets: Northern Virginia, Silicon Valley, Phoenix, Dallas. The rest of the country is being asked to catch up.
Trump’s statement, reported by Fox News, is a clear political endorsement of this expansion. He argued that local governments should welcome data centers because they bring “tremendous amounts of money and tax revenue” and create jobs in construction. He even acknowledged that the AI industry needs “PR help” given the public resistance. This is not a technical analysis. It is a political pitch. And it is being delivered at a time when the US is facing a structural electricity shortage. The North American Electric Reliability Corporation (NERC) has warned that the grid could face capacity shortfalls as early as 2027, driven in part by data center growth. The AI data center boom is not just an economic opportunity; it is a stress test on the entire energy infrastructure.
Based on my due diligence experience analyzing mining facilities and Layer2 tokenomics, I have seen this pattern before. A new technology arrives. The politicians promise jobs. The local governments offer incentives. The construction crews arrive. The jobs are largely temporary. The tax incentives erode the base. The environmental costs are externalized. And then the next cycle begins. The only difference this time is the scale. AI data centers can consume 100 megawatts to 1 gigawatt each. That is the equivalent of a small city. The political support is trying to smooth the path, but the physics of power generation and transmission do not care about political endorsements.
Core: The Systematic Teardown
Let me dissect the three claims in the Trump statement and test them against the data.
First, jobs. The claim that AI data centers create significant employment is technically true in the short term. Construction jobs are real. But the steady-state operational workforce for a large data center is remarkably small. A 200-megawatt facility might employ 30 to 50 people on-site after construction. The rest of the economic activity is indirect: equipment suppliers, maintenance contractors, and the power utility. The multiplier effect is real but modest. Compare this to a traditional manufacturing plant of equivalent capital investment, which might employ hundreds or thousands. The jobs narrative is not fraudulent; it is misleading. The code doesn’t require a union.
Second, tax revenue. Local governments are often so eager to attract data centers that they provide significant property tax abatements and sales tax exemptions. The net fiscal benefit is often negative in the early years, and the long-term benefit depends on the facility staying competitive. AI data centers have a shorter useful life than traditional industrial facilities. The GPU clusters are replaced every three to five years. The building shell may last 20 years, but the compute equipment becomes obsolete much faster. The tax revenue from the equipment is often depreciated rapidly. The net present value of the tax stream is lower than the headlines suggest. Chaos is just data waiting to be compiled.
Third, public opposition. Trump acknowledged that most Americans are against data centers in their neighborhoods. This is not a minor detail. It is the single most important risk factor. The opposition is not irrational. It is based on local concerns about noise, water consumption, visual impact, and the strain on local infrastructure. The AI industry has not done a good job of addressing these concerns. The “PR help” that Trump mentioned is a euphemism for a systematic failure of public engagement. The industry wants the benefits of political support without the accountability of community consent. That is not sustainable.
What does this mean for blockchain? The parallel is direct. Crypto mining facilities faced the same opposition. They were marketed as economic development tools. They were opposed by environmental groups. They were subject to regulatory uncertainty. The difference is that AI data centers have a stronger political champion. But the underlying structural dynamics are identical. The grid is the same grid. The water is the same water. The local communities are the same communities. The AI industry is getting the political cover that crypto never got. That is not a sign of superior technology. It is a sign of superior narrative framing.
Contrarian: What the Bulls Got Right
I am not a contrarian for the sake of being contrary. But the narrative around AI data centers is so one-sided that the counterarguments deserve a fair hearing. The bulls are right that AI data centers will create real economic activity. They will drive demand for electrical equipment, transformers, cooling systems, and construction materials. They will create jobs in engineering, logistics, and operations. They will generate tax revenue, even if the net benefit is smaller than advertised. And they will accelerate the modernization of the grid. The push for AI data centers may force utilities to invest in transmission upgrades, renewable integration, and demand response programs that benefit all customers. That is a genuine positive externality.
Moreover, the political support from Trump is not purely symbolic. It could lead to federal-level policies that streamline permitting, allocate funding for grid upgrades, and provide tax incentives for energy-efficient data centers. If that happens, the entire compute-intensive sector—including crypto mining—could benefit. The infrastructure is fungible. A grid that can support a 500-megawatt AI data center can also support a 500-megawatt Bitcoin mining farm. The regulatory environment is the gatekeeper, not the physics.
The bulls are also right that the public opposition is not insurmountable. The NIMBY effect is real, but it can be managed with transparent communication, community benefits agreements, and siting in less sensitive areas. The AI industry has the resources to do this. The question is whether they will. The historical record suggests that industries with strong political backing tend to overestimate their social license and underestimate the backlash. The Terra Luna collapse was a textbook example of a community that believed its own narrative until the math stopped working.
Takeaway: The Fork Was Inevitable; The Error Was Optional
The political endorsement of AI data centers is a structural shift. It is a recognition that compute is infrastructure, not just software. That is a good thing in the abstract. But the current implementation is flawed. The jobs narrative is inflated. The tax revenue is uncertain. The grid is not ready. The public opposition is ignored. The AI industry is being given a pass that crypto never got, not because of superior technology, but because of superior political alignment. That asymmetry is a risk. It creates a bubble of political capital that can burst when the reality of local opposition, grid constraints, and fiscal costs catches up.
I measure risk in gas units, not in hope. The gas units here are the megawatts that the grid cannot supply, the community meetings that turn hostile, the tax incentives that fail to deliver ROI. The AI data center boom is real. But the political narrative is a sugar coating on a bitter pill. The industry needs to address the fundamentals: grid capacity, water consumption, community engagement, and long-term fiscal sustainability. If it does not, the crash will be as predictable as the Terra death spiral. The fork was inevitable. The error is optional.
For the blockchain community, the lesson is clear: political support is a double-edged sword. It can accelerate adoption, but it can also create a false sense of security. The code doesn’t care about the president’s endorsement. The grid doesn’t care about the press release. The local community doesn’t care about the GDP multiplier. They care about their property values, their water bills, and their quality of life. The AI industry is about to learn that lesson the hard way. And when it does, the crypto industry, having been through the same cycle, will be watching. Chaos is just data waiting to be compiled.