AI Is Making Economists Rethink the Story of Automation. | Social Sciences and Humanities

TEXT START: This article examines the changing views of economists regarding the impact of artificial intelligence (AI) on employment.

The Dissection

The article documents economists retreating from the comforting automation narrative: technology does not reliably create compensating jobs, and digital systems have already hollowed out middle-wage work while widening inequality. But it stops at diagnosis. Its real function is to convert a structural rupture into a policy-management problem—better institutions, worker participation, education, and newly created work.

The Core Fallacy

It treats the employment system as repairable after AI severs the productivity-to-wages link. Under the Discontinuity Thesis, P1 makes cognitive labor cheaper and more scalable; P2 prevents institutions from preserving human-only work at competitive scale; P3 removes the majority from economically necessary production. Redistribution may preserve consumption, but it cannot restore productive participation or make displaced labor economically indispensable.

The article also assumes that “new work opportunities” will emerge in sufficient volume and quality. That is historical extrapolation applied to a discontinuity. Earlier technologies displaced tasks while leaving new labor markets to expand. AI attacks the coordination and cognitive functions that generated much of the replacement work itself.

Hidden Assumptions

• AI’s gains can be distributed fairly without changing ownership and control of AI capital.
• Education can outrun systems that continuously reduce the value of educated cognitive labor.
• Policymakers can coordinate a stable human-only economic domain despite competitive pressure to automate.
• Worker participation in AI deployment can alter the underlying cost curve.
• More jobs are inherently preferable to fewer jobs, even when the new roles are politically manufactured or economically parasitic.
• Inequality is the central danger, rather than the collapse of mass productive necessity.
• Productivity gains will remain compatible with the post-WWII wage-consumption circuit.

Social Function

Primarily partial truth, transition management, and ideological anesthetic. The article is more honest than the old automation lullaby because it concedes that technology can reduce middle-wage demand and increase inequality. It becomes anesthetic when it presents education, institutional reform, and worker voice as mechanisms capable of reversing the underlying transition. That preserves the legitimacy of the existing order while avoiding its terminal implication: the mass worker may become economically surplus even if society remains materially wealthy.

The Verdict

The article correctly identifies the corpse’s first unmistakable symptoms but refuses to name the cause of death. It sees inequality and labor-market damage, yet frames them as defects to be corrected rather than consequences of AI-driven productive participation collapse. Its policy remedies may delay social death or redistribute the carcass; they do not restore the mass employment system. Under DT logic, this is a sophisticated early-warning document trapped inside a reformist fantasy: accurate observation, incomplete mechanics, and a final recommendation to negotiate with the machine that is eliminating the bargaining position.