We have the most powerful technology in history, and we’re using it to stop hiring young people

U.S. employers announced 573,195 job cuts through September, with AI cited in 120,136; Stanford economists found employment among 22- to 25-year-olds in the most AI-exposed occupations was about 19% below its counterfactual pace as of June, mainly because of reduced hiring. The commentary also highlights household cost pressures and estimates that data centers drove 63% of PJM’s 2025/2026 price increase, or about $9.3 billion. The author argues for requiring data centers to fund grid costs and tying public AI incentives to entry-level hiring, while directing AI investment toward utility, water and other public infrastructure.

Analysis

The investable signal is not a near-term collapse in AI demand; it is a possible repricing of who bears the marginal cost of power, grid capacity and local infrastructure. If regulators shift those costs to large loads, data-center project economics weaken at the margin, while regulated utilities with approved cost recovery may see better capex visibility. That is not automatically bullish for utilities: affordability backlash can delay rate cases, and cost allocation may differ by jurisdiction. The article’s policy proposals are advocacy, not enacted nationwide rules; near-term political attention should not be mistaken for a uniform change in tariffs.

Over 1–3 months, watch state tariff proceedings, PJM capacity outcomes and data-center project cancellations or delays. Over 6–18 months, entry-level hiring conditions and documented utility-bill impacts could strengthen the case for mandates, but water-leak and grid-management deployments are unlikely to offset data-center demand quickly. Suppliers such as Xylem and Itron could benefit if public procurement accelerates, though project awards and realized savings—not pilot results—are the validation.

Contrarian view: AI-related job attribution and infrastructure-cost estimates do not establish that AI is the principal cause of broad household pressure, or that proposed regulation will pass. The regulatory risk is real but uneven; a broad short of AI beneficiaries is not supported by this commentary alone. Falsify the cost-shifting thesis if large-load tariffs remain limited, project commitments hold, and utilities continue recovering incremental costs without material consumer backlash.

Key Decisions for Investors

• No broad AI short on this article alone. Treat the near-term impact as a policy-risk watch item; avoid paying for an assumed nationwide cost-shift before state rules and utility filings confirm it.
• Conditional pair for the next 1–3 months: favor regulated utilities with explicit large-load cost recovery over data-center developers whose projects depend on customer-friendly interconnection terms. Size only after checking jurisdictional exposure, rate-case recovery and project cancellation protections; cut the view if tariffs remain permissive or utility recovery is denied.
• Watch Xylem and Itron as potential 6–18 month beneficiaries of utility leak-detection and grid-management budgets, not as immediate buys. Require disclosed contract wins, funded procurement and independently verified savings; pilot evidence alone is insufficient.
• Track state legislation, PJM capacity prices, utility bills and large-load project cancellations as catalysts. Reassess if proposed rules fail to advance, data-center load forecasts are revised down, or ratepayer affordability concerns block utility cost recovery.