The economic story of 2025 reveals a government-driven build-out of hyperscale AI data centers — marketed as innovation and national strategy, yet consuming land, food, water, and energy at staggering rates. Energy Secretary Chris Wright emphasized that “It takes massive amounts of electricity to generate intelligence. The more energy invested, the more intelligence produced.”
This policy shift has distorted markets, siphoning capital from consumer electronics toward subsidized AI infrastructure. Major manufacturers including Dell and Samsung have scaled back or discontinued product lines due to diverted components for AI chip production. As a result, prices for phones and laptops have surged, while jobs in U.S.-based assembly operations face elimination.
Historically, the computer industry demonstrated significant deflationary success. From December 1997 to August 2015, the CPI for “personal computers and peripheral equipment” dropped by 96%, contrasting sharply with medical care, housing, and food costs that rose between 80% and 200%.
Today, AI data centers are crowding out consumer electronics. DRAM production is increasingly directed toward enterprise servers rather than personal devices. In the fourth quarter of 2025, contract prices for certain 16GB DDR5 chips surged nearly 300%, with Dell and Lenovo imposing 15%-30% price hikes on PCs due to AI-sector demand.
DRAM inventory levels have plummeted by 80% year over year, leaving only three weeks of supply — down from 9.5 weeks in July. Samsung has doubled DDR5 contract prices to $19-$20 per unit and signals potential exit from the SSD market. Nvidia warns it may reduce RTX 50 series production by up to 40%, driving up gaming system costs.
Shrinkflation is emerging: manufacturers are reverting to 8GB systems with slower storage to keep entry-level laptops under $999 — a configuration that cannot even run the AI applications they were designed for. Research and development in conventional computing has also slowed, as investment shifts toward AI accelerators.
This approach is backward. Narrow AI solutions for logistics, agriculture, port management, and manufacturing deliver immediate productivity gains. China understands this and invests accordingly, while U.S. firms like Roomba — which pioneered practical autonomy — collapse and are acquired by Chinese entities.
The government’s role in creating an arms race for data centers through tax incentives, regulatory favors, land-use carve-outs, capital subsidies, and suppressed interest rates has hollowed out the consumer technology sector. As Charles Hugh Smith observed, expanding credit fuels asset price inflation, enabling capital-rich firms to outbid competitors while destabilizing the broader economy.
The pattern repeats: When one government subsidy collapses — such as electric vehicle investments that led to 1,600 layoffs at a Ford plant in Glendale, Kentucky — another emerges. This facility is now retooled for data center batteries.
Conventional technology provides proven utility through reliable hardware and mobile computing. The current surge into artificial general intelligence, however, relies on hypothetical returns bolstered by state power.
The good old laptop is becoming collateral damage in what may prove to be the largest government-induced tech bubble yet.
Daniel Horowitz is the host of “Conservative Review with Daniel Horowitz” and a senior editor for Blaze News.
