Taxing AI to Help Workers Sounds Good, But Public Deserves More - Bloomberg Law

URL SCAN: Taxing AI to Help Workers Sounds Good, But Public Deserves More - Bloomberg Law
FIRST LINE: Rep. Greg Casar’s (D-Texas) proposed AI Tax and Work Protection Act has a solid premise: Firms capturing gains from AI-driven labor force reductions should carry some of its social costs.

THE DISSECTION

The article wins the narrow tax-design argument and loses the systemic war. It correctly identifies tokens as a fraudulent proxy for displacement: token volume does not reveal whether AI replaced workers, augmented them, or had no labor effect. Equity is a more coherent way to capture economic gains without pretending the Treasury can price billions of unstable computations.

But the article converts a civilizational rupture into a revenue-engineering problem. Under the Discontinuity Thesis, AI is not merely generating an externality like cigarettes or carbon. It is attacking the wage-to-consumption circuit that makes post-WWII capitalism function. Public equity can collect rents from the machine. It cannot make human labor economically necessary again.

The proposal shifts government from taxing the meter to owning part of the franchise. That is structurally smarter. It remains a transfer mechanism, however—not a restoration of productive participation. A public shareholder can receive dividends while the majority remains excluded from production, bargaining power, and economic necessity.

THE CORE FALLACY

The text treats labor displacement as a cost surrounding capitalism rather than a process that can break capitalism’s load-bearing mechanism. If P1, P2, and P3 hold—AI dominates cognitive work, human institutions cannot preserve human-only economic domains, and mass productive participation collapses—then taxation redistributes the surviving surplus. It does not restore the lost wage system.

Equity is a better proxy for AI-generated gains than token usage. It is not a solution to obsolescence. The article also fails to distinguish ownership from control. A passive sovereign wealth fund is a dividend pipe. Durable voting rights over compute, energy, logistics, maintenance, and deployment would create genuine strategic leverage. The text gestures toward public ownership without confronting whether the public would actually control anything.

HIDDEN ASSUMPTIONS

• AI value will remain concentrated in identifiable, taxable firms rather than migrate across chips, energy, cloud infrastructure, private entities, jurisdictions, and supply chains.
• Share prices and reported profits will reliably reflect durable AI rents rather than speculation, accounting choices, or concealed value extraction.
• The state can impose equity assessments before firms restructure, dilute ownership, relocate, or capture the regulator.
• A public investment vehicle will serve the population rather than become another asset-management layer controlled by financial and political elites.
• Transfers can preserve social stability after labor loses its productive function.
• The institutional state will remain competent and legitimate enough to administer the scheme during the very collapse it is meant to manage.
• The central harm is lost income, when the deeper loss is the majority’s disappearance from economically necessary activity.

SOCIAL FUNCTION

Primary classification: partial truth and transition management. Secondary classification: ideological anesthetic.

The article offers a valid correction to a badly designed tax and a plausible mechanism for capturing AI rents. It also makes dispossession sound like participation: the public receives an upside from the system that has made its labor unnecessary. That may preserve consumption and delay political rupture. It does not preserve the old social contract.

This is carcass management with a respectable legal vocabulary. The state is being prepared to distribute pieces of the automated economy after the employment economy has begun to die.

THE VERDICT

The article is right about the meter and wrong about the fire. Public equity is superior to token taxation and could finance transfers, but transfers are not productive participation. Under DT logic, it is a lag defense—possibly useful for cushioning collapse and delaying revolt, incapable of reversing system death.

The decisive question is not whether Treasury taxes tokens or owns stock. It is who controls the AI capital stack when human labor loses market necessity. Without real control rights, public equity is managed obsolescence with a dividend attached.