In the AI economy, the ‘weirdness premium’ will set you apart. Lean into it, says expert on tech change economics

Brandeis economist Benjamin Shiller argues in his new book that AI will struggle to replace workers in niche, data-scarce professions, giving rise to a “weirdness premium” for unusual skills. He highlights Elon Musk’s claim that roughly 10 billion miles of driving data (about 750,000 years of human driving) are needed for safe unsupervised self-driving and cites Waymo’s failures with kangaroos as an example of edge cases outside training data. Goldman Sachs’ estimate that 300 million U.S. and European jobs could be susceptible to AI-driven change underscores the broader labor-market risk, but Shiller’s thesis implies sustained human comparative advantages in specialized roles—an important consideration for staffing, human-capital valuation, and sectoral exposure to automation.

Analysis

Market structure: AI’s diffusion favors data-rich, generalizable platforms and counterintuitively raises pricing power for niche human specialists whose skills are data-sparse. Expect wage and margin expansion for boutique engineering, aviation-safety, specialized healthcare, and classified-defense contractors over 1–5 years as supply remains inelastic; commoditized BPO and routine white-collar roles face downward price pressure and potential consolidation.

Risk assessment: Tail risks include a rapid algorithmic breakthrough that collapses the data requirement (months–1 year), regulatory bans on certain AI uses (0–18 months), or a high-profile autonomous-vehicle accident triggering tighter rules (immediate–6 months). Hidden dependencies: company access to proprietary edge-case datasets and human-in-the-loop IP will drive moats; failure to control data pipelines is a silent bankruptcy risk for AI plays.

Trade implications: Favor small, high-margin specialist operators and incumbents with proprietary fleet/data (positive for TSLA’s data moat) while underweight generic staffing/BPO and low-margin logistic plays. Use long-dated options to capture multi-quarter re-rating of specialists and protective hedges for broad AI names through put spreads; rotate into Healthcare, Aerospace/Defense, and Cybersecurity over the next 3–12 months.

Contrarian angles: Consensus overweights “scale AI” platforms and underestimates pay-up for weird skills — mispricing exists in small caps with embedded human expertise. Historical parallels: past automation waves re-rated niche specialists (1990s IT consultants). Unintended consequence: higher specialist wages could be inflationary and compress corporate margins outside tech over 2–4 years.

Key Decisions for Investors

• Establish a 2–3% portfolio long in TSLA via 12-month LEAP calls (buy Jan 2027 calls ~30% OTM) to capture Tesla’s fleet-data moat if self-driving data remains a competitive barrier; size as a directional, skewed upside bet, cap position at 3% of NAV.
• Allocate 1.5–2% long to HEICO Corp (HEI) or AAR Corp (AIR) as boutique aerospace/parts providers with pricing power; buy shares or 9–12 month 10–15% OTM call spreads to limit downside while capturing margin expansion from specialist demand.
• Establish a relative-value pair: long 1.5% HEI (or AIR) vs short 1.5% ManpowerGroup (MAN) via short-dated (3–6 month) call overwrites on MAN or buying 3–6 month puts if MAN rallies >10%; thesis: specialist wage premium vs commoditized staffing squeeze.
• Use options hedges in AI platform names: buy 3–6 month put spreads on broad AI/mega-cap software ETFs (or GOOG/MSFT if preferred) sized to cover 5–8% portfolio exposure; unwind if regulatory clarity emerges within 60–120 days.