The Labor Movement Is Divided Over How to Deal With AI | Truthout

TEXT START: Truthout is an indispensable resource for activists, movement leaders and workers everywhere.

THE DISSECTION

The article documents labor’s strategic fracture as AI becomes physical infrastructure. Building unions defend immediate construction wages; other unions see data centers as accelerants of displacement and public harm; education unions pursue adaptation through industry-funded AI training; activists demand resistance and redistribution.

The underlying reality is harsher: institutions built to sell labor are arguing over how quickly labor loses its bargaining centrality. The article is a seismograph of that fracture, not a complete autopsy.

THE CORE FALLACY

It treats AI as a policy variable that can be paused, regulated, trained, or safely deployed, rather than as a competitive production system. If cognitive automation achieves durable cost and performance superiority, firms are compelled to adopt it. Human-only economic domains cannot remain stable at scale.

Data-center construction is not proof of durable employment. It is an installation phase for the machinery that can eliminate far larger pools of cognitive labor. AI training may make workers more effective, but it can also make their replacement cheaper and faster. Four-day workweeks, UBI, and AI dividends can preserve income and consumption; they do not restore productive participation or transfer control of AI capital.

The article’s strongest claim—that unions should resist unsafe and unethical automation—is valid only at the level of lag management. It does not defeat the structural mechanism.

HIDDEN ASSUMPTIONS

• Regulation can preserve stable human employment despite competitive pressure to automate.
• Temporary construction work can substitute for the permanent occupations AI erodes.
• Worker training converts labor into a beneficiary rather than an instrument of replacement.
• Union coordination can preserve human-only sectors across competing firms and states.
• Full employment remains an attainable organizing target after productive labor is no longer economically necessary.
• UBI or dividends solve the crisis rather than merely subsidizing consumption without ownership.
• Union leadership can partner with AI firms without allowing those firms to define the terms of labor’s adaptation.
• Democratic debate alone can control systems whose ownership, compute, energy, and deployment remain concentrated.

SOCIAL FUNCTION

Classification: partial truth, transition management, and ideological anesthetic.

The article truthfully exposes real conflicts over jobs, infrastructure, public costs, and union democracy. But its policy menu—guardrails, training, organizing, shorter workweeks, UBI—converts systemic displacement into a negotiable reform agenda. That helps institutions manage the transition while avoiding the terminal question: who owns and controls the productive machinery after labor loses necessity?

The industry-funded training hub is especially revealing. It presents adaptation as worker empowerment while functioning as a test case for reengineering labor profession by profession. Teachers are not merely being trained to use a tool; they are being integrated into the deployment pipeline.

THE VERDICT

The labor movement is divided because its historical function is disintegrating. One faction is extracting wages from building the machine; another is absorbing the machine’s consequences. Neither position, by itself, survives the Discontinuity Thesis.

Moratoriums, guardrails, training, and transfers are hospice measures. Unless organized labor shifts from defending wage dependence to controlling or owning AI capital and the essential systems around it—energy, logistics, maintenance, verification, and transition infrastructure—it will become an administrative layer negotiating the terms of its members’ obsolescence. The article records the funeral committee arguing over construction contracts.