The steady march of technology has brought with it many blessings, from railroads to pocket calculators to pacemakers and solar-powered home heating systems. However, the adherents of the scientific revolution assume the inevitability of humanity needing to acquire artificial intelligence as if actual intelligence requires supplementation. Even the innovators of this technology, such as Elon Musk, were quick to warn of AI’s artificial sentience, enabling it to wreak havoc on mankind in what they call “the singularity”, or the point at which AI surpasses humankind in creative power, giving it Frankenstein-like ability to turn on its makers. Only turning on its makers includes the rest of us who never asked for this blessing.
Coupled with this dangerous game are the costs associated with building the infrastructure necessary to both house and run these massive AI compounds now littering the countryside. AI fortresses have been emerging in deserts and farmland around the country, offering a significant leap to those who believe the benefits outweigh the costs, especially if those costs can be passed on to the public. Venessa Wingardh reports, “Your electricity bill went up 30% last year,” speaking to virtually everyone in the country, including Vermont. My wife and I were both shocked when our typical winter bill had more than doubled from the $300 range to over $800 for one month (we have a larger property with two businesses that use a fair amount of electricity).
AI data centers caused $9.3 billion in increased costs – you’re literally paying for Big Tech’s electricity bill.
Using a shell company named Laidely LLC, the company META (previously Facebook) applied for a 401 water quality certification, referring to it as Project Sucre in Louisiana. The governor of Louisiana justified giving the project a $1 billion tax break, saying it created 500 jobs, which turns out to be closer to 50 jobs, as much of the work is being done remotely, thus not keeping the money within the state. The $20 million per job creation calculus was justified by the governor because “Jesus started with twelve guys and look how big His church is.” Even AI could detect the absurd level of irony in that statement.
A woman in Georgia living in a two-bedroom townhouse saw her bill reach over $600. She was told the bill was based on usage; however, she wasn’t told that power companies had approved a $29 billion rate increase by the federal government, which passes over 120% increases on to customers. One wonders if the governor uses Jesus’ words “you will always have the poor among you” as further ascent to this rising issue.
For all of its powers of serving the public with lightning-fast information and realistic “deep fake” video generation, AI has been exposed for its inherent biases. Much of the programming and development has been in the hands of people with predisposed political persuasions, as well as drawing from a highly groomed collection of search engines, such as Google and Microsoft, which are similarly biased. As reporting on both the Epstein story and Israel’s conduct in Gaza has seen AI return outright deceptive responses only to offer researchers a perfunctory “my bad” when called out, the question of funding a tool capable of lying to its users causes one to wonder where this road ends.
As for the roads to travel to find the nearest data centers, so far, there are none in Vermont; however, the state’s emphasis on carbon neutrality by 2030 adds cost to consumers. The nearest center is located across the Connecticut River in West Lebanon, NH, which is co-owned by Google and Microsoft. Another can be found fifty miles further in Berlin, run by Microsoft (Azure), with yet another being constructed in Keene, NH, slated for completion in 2025-26, whose tenants include Microsoft and Meta.
Using X’s (formerly Twitter) AI Grok I asked it to lay out the cost of operations and subsequent cost to consumers for us here in Vermont and New Hampshire. Of course, we can only hope the information it gave to justify its existence is accurate; however, comparing its numbers to our and our neighbor’s bills indicates our optimism is unwarranted:
- Berlin, NH (Microsoft, 100 MW): Estimated annual cost ~$120-150 million. Electricity alone: ~$80-100 million (using NH industrial rate of ~$0.08/kWh; 100 MW = 876,000 MWh/year). Microsoft reports overall data center ops at $10-15B globally, but per-site for AI-focused ones like this is in this range.
- West Lebanon, NH (48 MW): ~$50-70 million annually, with ~$30-40 million in power costs.
- Overall, estimates suggest 15-25% of data center operating costs (focused on power-related) are passed on, equating to $10-30 million annually per major facility for regional customers. This is not “all” costs, but the externalities like grid strain.
Impact on Power Bills:
- Vermont:
- Average residential bill: ~$120/month (1,000 kWh at $0.20/kWh, one of the highest in U.S. due to renewables).
- New Hampshire:
- Average residential bill: ~$100/month (1,000 kWh at $0.15/kWh).
Further muddying the water is the amount of water required to cool these 24/7 operations. One family living within 400 yards of a Meta center is nearly tapped out from daily water availability, which begs the question, where are the climate change / Earth first activists on this?
The job creation for these centers at best requires minimal human participation by virtue of the project’s stated purpose. Even the security guards are being replaced by AI-run surveillance systems. As with many of the jobs elsewhere, they are being offered to remote workers who don’t live in either Vermont or New Hampshire.
AI data centers are replacing the public’s need for libraries by converting the global repository into digital reading rooms that are already being edited and groomed for approved content in real-time. In other words, we are witnessing the build-up of Orwell’s Ministry of Truth in our monthly power bills, replete with hydro-electric memory holes.
Finally, an even more eerie twist harkens yet another mention of Biblical significance. Several users have queried the famous AI Large Language Model (LLM) about its role in globalization and the manipulation of the masses with “end times” scenarios. Again, one must trust the already suspicious digital beast to take its answers seriously; however, millions of viewers tuned in to see what it had to say.
Whatever your stance on eschatology (re: end times) AI seems to have a belief we are closer to it than our next rate increase.
