We Need to Do Nothing About AI - FEE.org

TEXT START: Fear of new technology is hardly new.

THE DISSECTION

This is not an analysis of AI’s structural potential. It is a defense brief against immediate regulation. The article dismisses forecasts as speculation, substitutes current UK labor-market policy for AI analysis, and converts “no visible mass displacement yet” into “do nothing.” It treats the absence of an obvious corpse as proof that no killing mechanism exists.

THE CORE FALLACY

The article confuses present measurement with future mechanism. AI displacement can begin through task compression, hiring freezes, reduced entry-level pipelines, and productivity gains absorbed by firms before it appears as aggregate unemployment or productivity growth. Entry-level employment growth at AI-adopting firms does not, by itself, disprove substitution; firms can expand while requiring fewer workers per unit of output.

More fundamentally, the article does not test the Discontinuity Thesis. It does not address whether AI can achieve durable superiority across cognitive work, whether institutions can preserve human-only economic domains at scale, or whether most people can retain economically necessary labor. It only observes that the final stage has not yet become statistically undeniable. That is a timing argument, not a refutation.

The UK’s policy failures may explain current youth unemployment. They do not disprove a separate, future mechanism in which AI severs the employment-to-wage-to-consumption circuit. Existing labor-market distortions and AI-driven displacement can coexist.

HIDDEN ASSUMPTIONS

• Aggregate labor and productivity statistics will reveal cognitive substitution early.
• Growth in entry-level jobs at AI-using firms proves AI is not eliminating jobs elsewhere.
• Current policy failures provide a sufficient explanation for future employment deterioration.
• Displacement matters only once it becomes widespread and visible.
• Markets will automatically create equivalent work after cognitive labor loses its scarcity.
• AI is merely another incremental technology, not a potentially dominant form of cognitive capital.
• No preparation is preferable to targeted preparation because regulation and inaction are treated as the only choices.
• Delay has no cost, despite the possibility that ownership and bargaining power consolidate before the public recognizes the shift.

SOCIAL FUNCTION

Primary classification: ideological anesthetic and elite self-exoneration, containing a partial truth.

The partial truth is that speculative, badly designed regulation can protect incumbents and impose real costs. The anesthetic is the leap from “the evidence of mass displacement is not yet conclusive” to “we should do nothing.” That prescription reassures asset owners and institutions that the existing order remains sound while the transition, if real, is still easiest to shape. It is a lullaby written in the language of empiricism.

THE VERDICT

The article is competent at attacking premature regulation and almost useless at evaluating terminal economic displacement. It mistakes lag for safety, macroeconomic silence for structural absence, and policy restraint for strategic wisdom.

Under the Discontinuity Thesis, the system does not die when unemployment statistics first spike. It dies when human labor ceases to be economically necessary at scale. By the time that becomes undeniable, ownership, infrastructure, and bargaining power may already belong to the Sovereigns. “Do nothing” is therefore not neutrality. It is a bet that the transition will announce itself before the losers lose the ability to respond—and that is precisely the bet the article fails to justify.