Teachers need help with AI. A union is offering training – with $23m in funding from big tech

TEXT START: Darius Saczuk, a high school teacher, views artificial intelligence as his enemy.

The Dissection

This article is not primarily about teacher training. It is about converting an AI labor threat into an acceptable classroom reform. Big tech supplies $23m; the union supplies trust, access, and peer-to-peer distribution. The union becomes a legitimacy wrapper around products whose vendors profit from adoption.

The training has tactical value: teachers can identify hallucinations, address privacy risks, redesign assignments, and use AI for limited accessibility or administrative tasks. That is adaptation, not proof that teaching survives. The claim that AI cannot replace teachers conflicts with the article’s own evidence that AI executives envision virtual tutors and replicated teacher functions.

The Core Fallacy

The text confuses knowing how to use AI with remaining economically necessary. A teacher can become more competent at supervising AI-mediated instruction while the system needs fewer teachers. Augmentation is frequently the measurement phase before substitution.

AI-generated lessons, personalized practice, student-data analysis, tutoring, and routine feedback attack the repetitive cognitive work that supports current staffing levels. Human judgment, safeguarding, and social authority may delay replacement, but they do not automatically make the majority of teachers indispensable.

The union also mistakes voice for power. Teachers may influence implementation details while capital owners retain the models, infrastructure, data, and product roadmap. That is consultation inside someone else’s machine.

Hidden Assumptions

• AI will remain a tool teachers control rather than the basis of a cheaper instructional operating model.
• Human judgment can be made economically indispensable at mass scale, rather than merely pedagogically desirable.
• Vendor promises about safety, privacy, and standards can survive competitive pressure and be enforced without teacher ownership of the underlying systems.
• Product-neutral training is possible when every dollar comes from firms that profit from usage and lock-in.
• The absence of district guidance means adoption is inevitable, rather than a political opening to restrict or ban uses.
• Better individualized instruction will not produce surveillance, deskilling, or tighter managerial control.
• Schools will preserve current staffing after AI lowers the cost of instruction, assessment, tutoring, and administration.
• Demand for training proves legitimacy. More likely, it proves workers recognize the threat and are trying to remain useful inside it.

Social Function

Primary classification: transition management, layered with ideological anesthetic and elite self-exoneration.

This is not pure copium. Teachers face a real immediate problem, and practical training can reduce harm. But the political function is to make displacement appear participatory: workers are invited to shape the tool while ownership and strategic control remain with the companies building it.

The $23m buys access, trust, feedback, and a softer landing for adoption. It does not buy employment security. The union can call this empowerment; the companies can call it partnership. The transfer of cognitive production from labor to capital proceeds regardless.

The Verdict

This is a transition memo disguised as professional development. Teachers are being trained to operate the system that is learning how to operate without them.

The immediate survivors will be indispensable servitors: verifiers, disciplinarians, relationship managers, special-needs specialists, and liability buffers. The mass teaching role remains fragile because its most automatable functions are being converted into software while the institution meant to defend the profession accepts corporate funding to supervise the conversion.

Under the Discontinuity Thesis, the $23m is not a moat. It is hospice funding with a branding budget.