It's time to clean house — AI should replace some humans - The American Bazaar

TEXT START: The number in Anthropic’s latest economic scenarios that deserves our attention is 11.9 percent.

THE DISSECTION

This article performs a controlled rhetorical pivot. It begins by admitting the real threat: AI can make output rise while knowledge workers become economically unnecessary. Then it relocates the moral and economic question from displaced people to inefficiency: if a job can be eliminated, the worker is invited to prove the job deserved to exist. Bureaucratic waste and government incompetence are real, but they are used as a safe sacrificial target. The text turns an ownership-and-distribution crisis into a process-optimization exercise.

Its actual function is to give institutional and managerial permission for AI headcount reduction while preserving the comforting fiction that displaced labor will be redirected into things that matter. It never identifies who owns the systems, who receives the surplus, or how demand survives when wages are removed.

THE CORE FALLACY

The central error is confusing the elimination of waste with the preservation of mass viability. Removing pointless paperwork is beneficial at the task level. It does not create replacement employment at the system level. The same AI that audits bureaucracy also automates the supposedly valuable cognitive work around it.

Under P1, AI’s advantage spreads across cognitive tasks, not merely useless ones. Under P2, institutions cannot maintain stable human-only domains simply because work is meaningful or socially desirable. Under P3, the majority can lose access to economically necessary labor while the economy grows. Redirect human effort is therefore not a mechanism; it is a slogan. A worker is not saved because society can imagine a better use for them. They are saved only if someone still needs to pay for that use.

HIDDEN ASSUMPTIONS

• Productivity gains will be distributed widely enough to preserve purchasing power. The article never addresses ownership, bargaining power, dividends, or transfers.
• New and meaningful work will appear in sufficient volume. It supplies no demand mechanism and ignores the difference between socially useful activity and market-paid labor.
• Retraining or transition can absorb a 17.9 percent knowledge-worker unemployment scenario. That is asserted by implication, not demonstrated.
• Human final responsibility will preserve human labor. In practice, a small number of accountable overseers can supervise systems that replace the preparatory labor of thousands.
• Government inefficiency is the main problem. It is a convenient visible target, but AI’s reach extends through private professional, managerial, analytical, and creative work.
• AI auditing will be deployed for public value rather than primarily for cost reduction, control, and surplus capture.
• Human dignity can be recovered after the wage function collapses. The article names dignity as an end while offering only productivity as the instrument.

SOCIAL FUNCTION

Classification: partial truth, transition management, ideological anesthetic, and elite self-exoneration.

The partial truth is substantial: many jobs exist to move information through obsolete procedures, repair institutional errors, or protect organizational layers. But the article weaponizes that truth. It teaches readers to distinguish the valuable displaced worker from the supposedly useless one before the same logic reaches their own occupation. This is how a broad labor shock is made politically digestible: first cleanse the bureaucracy, then call the remaining casualties unfortunate but necessary.

Its government focus also narrows the indictment. The crisis is not merely that agencies are inefficient; it is that AI can sever the employment-to-consumption circuit across the economy. By framing the answer as better auditing and better redirection, the article lets owners of AI capital avoid the central question: who controls the surplus created by the labor they replace?

THE VERDICT

This is a polished permission slip for automation disguised as institutional reform. It correctly sees that AI will expose waste and raise output while reducing labor demand, but it refuses to follow that logic to its terminal conclusion. A larger economy with fewer economically necessary humans is not a repaired post-WWII system. It is a more productive ownership regime with a displaced majority.

The article does not solve the discontinuity. It sanitizes it. Its prescription—eliminate what no longer creates value and redirect people toward what matters—contains no ownership pathway, no demand-preservation mechanism, and no answer to productive participation collapse. Under the Discontinuity Thesis, the text is not a survival blueprint. It is transition management for the people operating the machine that is removing the need for them.