This episode features an interview between Tucker Carlson and Clayton Morris, host of Redacted News, examining the rapid, largely unreported buildout of AI data centers across the United States and its immediate consequences for communities, resources, and human autonomy.
The Scale and Speed of the Data Center Buildout Is Far Beyond Public Awareness
Morris argues the current AI boom is not a bubble but the very beginning of a massive infrastructure wave; thousands of new data centers are projected, doubling the existing count of roughly 4,800.
Nvidia and other firms are now taking on debt to fund construction, shifting from self-funding to a credit-driven expansion backed by Wall Street.
Amazon Web Services has sold out its compute capacity through 2030, signaling demand that far exceeds current supply.
The buildout is not creating lasting local employment; construction crews rotate through sites for ~18 months, and operational data centers employ only ~50 people each.
Ancillary industries — equipment rental (United Rentals), portable sanitation, generator suppliers — are scaling rapidly to service this transient construction model.
Water Consumption Is Already Causing Shortages and Conflict in Arid Regions
Most data centers use evaporative cooling, consuming hundreds of millions of gallons annually; a single facility can use 300–500 million gallons, while an average U.S. home uses ~200–300 gallons per day.
In the desert Southwest (Colorado, Nevada, Utah, Wyoming), ranchers are being denied water permits while data centers receive carve-outs to tap already-stressed aquifers.
Aurora, Colorado exemplifies the disparity: residents face smart-meter monitoring and fines for irrigation use, while the massive QTS data center nearby consumes vast quantities without comparable public oversight.
New “dry cooling” technologies (e.g., Vertiv) reduce water use but dramatically increase electricity demand, shifting rather than solving the resource problem.
Electricity Costs Are Rising Nationwide Before Data Centers Even Arrive
Utility auction markets price in projected future demand from data centers, driving up per-kilowatt-hour rates 12–14% in areas without any local facilities yet.
In Homer City, Pennsylvania, a shuttered coal plant is being converted to natural gas to power a data center complex; locals see no power benefit, only higher bills and warnings of coming blackouts.
Three Mile Island is being restarted to supply data centers, not residential grids.
Amazon and others buy defunct power plants, convert them to natural gas, and keep all output for their own facilities — communities bear grid strain without receiving energy.
The Federal Government Is Actively Overriding Local Opposition
Early Trump executive orders created a DOJ “AI Litigation Task Force” that sues communities attempting to regulate or block data centers via zoning, environmental, or noise ordinances.
The task force targets local rules on diesel generator particulates, water usage, and other externalities, framing resistance as an obstacle to national security.
National security designation enables eminent domain seizures for data center land (e.g., in Georgia) and bypasses EPA permitting for construction startup.
New York State has banned new data center builds; Texas (Gov. Abbott) imposed a one-year moratorium pending efficiency audits — but Project Stargate in Abilene, championed by Trump, proceeds regardless.
National Security Framing Merges Corporate Infrastructure With Surveillance State
The FCC now mandates 65% of humanoid robot components be U.S.-made (rising to 80% by 2030), forcing supply chain onshoring for Tesla Optimus and similar platforms.
Amazon builds hardened, multi-layer secure facilities for CIA, Pentagon, and Palantir — often hidden inside ordinary government buildings (VA hospitals, IRS offices) — then hands operational control to intelligence agencies.
Flock Safety license-plate readers and other surveillance feeds flow through AWS; the government wants all AI training and inference on domestic, controlled hardware to exclude Chinese models.
This infrastructure doubles as a domestic monitoring grid, not just a military asset.
AI Is Being Embedded Into Every Software Layer, Often Without User Consent or Control
Zoom, once a video tool, now positions itself as an “AI productivity company” — transcribing meetings, taking notes, and storing conversation data on AWS infrastructure.
Morris asserts user data control is illusory: deletion requests are honored superficially, but data persists in backups and reappears under subpoena.
The “AI you don’t see” — efficiency layers in logistics, manufacturing, retail — is already generating measurable revenue gains (e.g., 15% revenue lift at an outdoor furniture firm via AI-driven merchandising; $129M scrap reduction at a metal processor).
These early wins suggest a coming wave of AI-driven margin expansion across the economy, reinforcing the buildout incentive.
Cognitive Offloading Threatens Human Reasoning, Especially in Children
The “Google effect” (2011) showed search engines reduced memory retention; AI accelerates this by synthesizing, organizing, and writing for students.
Studies find children produce high-quality AI-assisted work but fail retention tests — they never engaged in the cognitive labor of research, synthesis, or argument construction.
Morris worries this erodes the ability to reason, debate, write, and think independently — core human capacities.
His own son is pursuing electrical trades rather than white-collar paths (accounting, piloting) he believes will be automated.
The Orientation of Technologists Suggests Humans Now Serve the Machine
Carlson and Morris observe that AI proponents rarely argue the technology improves human flourishing; instead they frame it as an inevitable imperative — “we have an obligation to serve technology.”
The internet’s 30-year record: less peace, stability, wisdom, or knowledge — yet society is tripling down with AI.
Resource extraction (water, power, land), rising costs, surveillance, and cognitive atrophy are treated as acceptable externalities.
A counter-trend may be emerging: young people rejecting smartphones, seeking analog connection (board games, card games, phone-free spaces), and valuing human-generated content over AI “slop.”
The Endgame May Be Control Infrastructure Repurposed as Carceral Infrastructure
Morris speculates that when compute demand plateaus, the massive, fortified, remote data center campuses — built of cement and steel — will convert to for-profit prisons, a model the Trump administration supports.
This would complete a cycle: public resources subsidize private buildout, federal power crushes dissent, and the resulting physical plant becomes a tool of population management.
The robotics layer (humanoid battlefield robots, domestic automation) adds a physical enforcement dimension to the digital control grid.
No credible “kill switch” or democratic off-ramp exists; the military is years ahead in operationalizing autonomous systems.
