A CEO trying to reindustrialize America says blue-collar pay is headed for ‘massive hyperinflation’ and kids should skip college to become welders

Hadrian CEO Chris Power warned that AI could automate many white-collar roles within as soon as 18 months while driving strong demand for blue-collar workers. BLS data: median welder pay was $51,000 in 2024 versus a ~$60,000 median for all U.S. workers, with welder employment projected to grow 2% through 2034 while electricians are projected to grow 9% through 2034. AI data-center investment is running at roughly $700 billion this year, construction workers at these sites earn about $81,100 on average, and there is cited demand for ~250,000 shipbuilders; Hadrian is simultaneously automating welding in its defense contracts despite reporting welder shortages.

Analysis

AI-driven infrastructure buildouts create a localized, high-beta bid for on-site trades (electricians, crane/rigging crews, heavy equipment operators) because labor is costly to mobilize quickly and training pipelines are shallow. Expect wage discovery to show up first in regional labor markets and subcontractor margins (3–12 months) before broad national statistics move; this compresses margins for firms that can’t pass through higher site labor costs and inflates margins for staffing and specialty contractor winners.

Automation of traditional blue-collar tasks is the countervailing force and is already migrating from discrete cells (robotic welding) to full-line autonomy; however, adoption is non-linear because certification cycles, capital intensity, and low-volume complexity (defense, shipbuilding) slow replacement. Net effect: demand shifts from headcount to higher-skilled maintenance/robotics technicians and to robotics OEMs/software vendors over a 12–36 month horizon, capping pure wage inflation for commodity trades while boosting spending on sensors, power distribution and materials (copper, conduit).

Key tail risks and catalysts — hyperscaler/datagroup capex pauses, defense procurement schedule slips, or sudden changes to skilled-immigration and training subsidies — can reverse the trade within quarters. Monitor near-term indicators: regional overtime hours for construction trades, order backlogs and lead times at robotics and tooling vendors, and margin moves at subcontractors; these will signal whether this is a transient labor squeeze or a durable reallocation of labor demand.

Key Decisions for Investors

• Long NVDA (or 6–9 month call spread) to play sustained AI infra spend: buy NVDA 2026 Jan 1,000/1,200 call spread funded by selling 1,300 calls (6–12 month horizon). R/R: directional upside if GPU demand persists; risk: inventory correction and guidance resets — cap position size to 2–3% notional.
• Buy ABB (NYSE: ABB) 12–24 month LEAPs to capture industrial-robotics exposure as blue-collar tasks automate. R/R: convex upside from multi-year automation tail; risk: slower adoption in low-volume or defense segments and margin pressure from competition — hedge with 10–20% cash buffer.
• Pair trade to benefit from localized labor tightness: long FAST (FAST) + long HD (HD) for 6–12 months to capture tools/materials volume and pricing power; hedge with a modest short position in discretionary retail ETF (XRT) to offset consumer cyclicality exposure. R/R: asymmetric upside if construction/installation activity remains elevated; risk: macro slowdown reduces DIY and non-infra construction.
• Portfolio hedge: buy 9–12 month puts on a broad AI hardware name or the semiconductor index (e.g., SOX ETF) sized to 20–30% of AI longs to protect against an abrupt capex pause. R/R: preserves gains in a capex drawdown; cost is premium erosion if buildout continues.