A growing movement wants AI out of schools

TEXT START: America's two largest school districts both imposed sweeping AI bans this past week, hitting the brakes on a technology that's rapidly making its way into classrooms around the country.

The Dissection

This is not fundamentally about whether AI improves education. It is about institutions discovering that cognitive work can no longer be kept inside their controlled enclosure.

The bans are local lag defenses: they preserve unaided writing, teacher authority, and conventional grading long enough to postpone redesigning assessment and credentials. They slow visible disruption. They do not alter the underlying capability shift. Mixed research on tutoring and critical thinking is a pedagogical complication, not a refutation of the Discontinuity Thesis.

The Core Fallacy

The text treats AI as a classroom variable administrators can simply switch off. Under P1 and P2, that assumption fails. Students can access models beyond school control, and institutions cannot coordinate a durable human-only cognitive domain at scale.

The ban protects the measurement ritual, not the economic substrate. It may preserve the appearance of unaided achievement while AI makes cognitive output cheaper and more abundant elsewhere. It does not restore the mass employment-to-wage-to-consumption circuit.

Hidden Assumptions

• Enforcement can remain broad and durable despite access outside school.
• Human-only essays remain a meaningful proxy for economically valuable ability.
• Critical thinking is preserved primarily by banning assistance rather than redesigning tasks and verification.
• Credentials will continue to ration opportunity through pre-AI forms of assessment.
• Students can afford to postpone AI competence without losing strategic ground.
• The educational question is more important than ownership, control, and economic necessity.
• A temporary brake can become a permanent human-only settlement.

Social Function

Primary classification: transition management. Secondary classifications: copium and partial truth.

The article accurately identifies possible learning harms and legitimizes short-term containment. But it converts a structural rupture into a manageable school-policy dispute. “Protect critical thinking” becomes respectable language for postponing harder questions: who owns the systems, what credentials mean when output is cheap, and what students are being trained to do when cognition is automated.

The ban is a tourniquet on a severed pipeline. It may limit specific misuse, but it cannot regenerate labor’s economic necessity.

The Verdict

The movement can delay AI use in classrooms; it cannot evict AI from the economy or restore the postwar wage-consumption circuit. This is institutional recoil, not strategic reversal.

If the bans persist, they risk producing a two-tier system: students practicing obsolete unaided outputs inside school while better-connected students acquire the tools governing post-transition work. Under the Discontinuity Thesis, the schools are defending the pre-AI examination machine. That is not survival. It is hospice care.