AI and the US Economic Transition: Redefining Productivity, Labor, and Policy

Core Conclusion

The US economy has entered an AI-driven transition that will structurally lift productivity growth, reshape labor demand, and force a reassessment of the neutral interest rate. The macro playbook from 2010-2020—lowflation, falling rates, abundant low-skill labor—is obsolete. This is not a cyclical uptick; it is a regime change in the economy’s supply-side capacity, with direct consequences for capital allocation. The primary investable conclusion: overweight productivity-enabling technology and traditional sector leaders deploying AI for efficiency; underweight labor-intensive, low-value-add business models vulnerable to disintermediation.

What the Market May Be Underestimating

Consensus acknowledges AI as a technology theme but misprices its macro transmission. The focus remains on semiconductor capex and hyperscaler earnings; the underappreciated channel is AI’s diffusion into services, which constitutes 70% of US GDP. Early adopters report 15-20% output efficiency gains (Morgan Stanley internal survey), and multi-factor productivity growth hit 1.8% annualized in 2024-2025, above the 2010s average. If sustained, this implies potential GDP growth above 2.5% without generating proportional inflationary pressure.

A second mispricing: AI is not a pure labor-displacement shock. JOLTS data shows tech-related vacancies materially exceed those in other sectors. The shock is one of skill demand rotation, not aggregate employment collapse. This means the economy can absorb productivity gains via output expansion rather than through payroll contraction, defusing the most bearish growth scenarios.

Evidence Chain

First, the productivity signal is already visible in national accounts. Multi-factor productivity—the cleanest measure of technology’s contribution to output beyond capital and labor inputs—has accelerated from its post-GFC doldrums. This correlates temporally with enterprise AI adoption and points to a structural, not purely cyclical, tailwind.

Second, high-frequency labor market data confirms structural churn, not destruction. The ratio of unfilled tech roles to unemployed tech workers remains elevated. Historical analogs—electrification, early computing—show net employment gains with long and variable implementation lags. The adjustment mechanism is occupational composition, not aggregate unemployment.

Third, the policy feedback loop is changing. If trend growth rises and real neutral rates drift higher, the Federal Reserve’s terminal rate for this cycle—and the floor for the next—will be higher than pre-2020 benchmarks. This lengthens the duration of the restrictive rate environment relative to consensus expectations embedded in bond markets.

Key Divergences and Risks

The primary risk is a capital misallocation cycle: overinvestment in undifferentiated AI infrastructure that fails to translate into end-user productivity. This would strand assets and depress returns in the very sectors currently priced for secular growth. A second risk is distributional: if skill-biased technical change concentrates gains among high-wage cognitive workers while displacing mid-tier service employment, inequality widens and aggregate demand may underperform even with supply-side gains. Third, policy lag—in education, retraining, and regional adjustment transfers—could amplify labor market friction and generate political backlash that constrains technology deployment.

Investment and Asset Allocation Implications

The macro regime favors equities over duration-sensitive fixed income, particularly in sectors where AI adoption improves unit economics: software, capital goods, financial services, and healthcare IT. Within credit, differentiate between productivity beneficiaries and secularly challenged labor-exposed names. The re-rating of neutral rates means the path back to a 2% handle on 10-year Treasuries is blocked absent a severe recession; this anchors a volatility floor on long-duration assets. Real assets that benefit from higher nominal growth with constrained commodity supply—select energy, power infrastructure—offer diversification against this rate backdrop.

Appendix: Key Data Points

Cross-Sector AI Adoption and Productivity Lift (Survey-Based Estimates)

Employment Projections by Skill Level (2025-2030)