The Robot Tax Is Back. Here's the Part Nobody Wants to Admit. | Journal

TEXT START: From 18th-century loom taxes to Bill Gates to the IMF, the real fight has never been about machines.

The Dissection

The article is performing a controlled downgrade of an existential rupture into a tax-design problem. It correctly identifies several early symptoms: AI is narrowing entry-level careers, concentrating market power, reducing labor’s share, and creating a training gap that retraining cannot quickly close.

Then it retreats into fiscal engineering. Tax excess profits, reduce payroll taxes, fund training, subsidize human-heavy services, and preserve consumer demand. This is transition management for a system whose productive core is being removed. The article treats AI as a distributional shock inside capitalism rather than a mechanism that can sever the employment–wage–consumption circuit itself.

Its strongest practical point is that taxing owners’ rents is more coherent than taxing machines as legal subjects. But that is a tax-policy improvement, not a solution to the Discontinuity Thesis.

The Core Fallacy

The article assumes that if the surplus is redistributed, the old economic order remains viable. Under DT logic, that is false.

A tax on AI profits can transfer purchasing power. It cannot restore the economic necessity of millions of workers. Transfers may preserve consumption, but they do not restore productive participation, bargaining power, status, or ownership. The system becomes a managed distribution regime feeding people from capital they no longer control.

The article’s claim that “tools are neutral” is also defective. AI is not merely a neutral tool whose returns policymakers may distribute differently. Under competitive pressure, firms are compelled to substitute AI wherever it is cheaper, faster, and more scalable. The substitution is structural. Tax policy can slow it, redirect it, or fund its aftermath. It cannot repeal the incentive.

The proposed refuge in service industries and human-machine collaboration is another lag defense masquerading as a destination. If cognitive automation becomes dominant and coordination cannot preserve large human-only economic domains, those sectors are temporary absorption zones, not permanent employment foundations.

Hidden Assumptions

• Governments can identify and tax global AI rents before firms shift profits, restructure ownership, or move across jurisdictions.
• Large-scale transfers can preserve social stability without restoring mass employment.
• Training can convert displaced workers into scarce AI-related roles, despite the article’s own evidence that declining occupations vastly outnumber expanding ones.
• Service industries supposedly resistant to AI will remain resistant long enough to absorb the displaced majority.
• Human-machine collaboration will create enough economically necessary work rather than simply making fewer humans more productive.
• Capital owners will accept redistribution that materially threatens their control of AI assets.
• National tax systems can manage a globally scalable technology whose marginal replication cost is extremely low.
• Consumption can remain the central stabilizer even after productive participation collapses.
• The current institutional state can execute a complex transition without capture, delay, or political fracture.
• The AI shock is mainly a question of distribution rather than ownership and control.
• “Normal” profits can be separated cleanly from AI-created rents in industries where AI changes the entire competitive baseline.

The article does acknowledge correlation problems and the limits of retraining. That caution is real. It simply stops before following its own evidence to the terminal implication.

Social Function

Primary classification: transition management and ideological anesthetic, with a genuine partial truth.

The text tells policymakers and capital owners that the crisis can be contained by rebuilding the tax relationship between labor and capital. That is useful as a delay mechanism and potentially useful for financing transfers. It is also politically convenient: it avoids the more dangerous question of who owns and controls the AI systems generating the surplus.

Its language converts displacement into a manageable fiscal imbalance. “Tax the rents” sounds muscular, but without ownership reform it means the Sovereigns retain the productive machinery while the state negotiates over a portion of the output. Workers may receive compensation for losing their economic function. They do not regain that function.

The article is therefore not pure copium. It accurately identifies rent concentration and the failure of robot-specific taxation. But it uses a correct short-term diagnosis to imply a false long-term cure.

The Verdict

This is hospice accounting for mass employment.

Taxing AI rents is more rational than taxing robots and may buy time, fund transfers, and blunt the first wave of concentration. It does not reverse P1, P2, or P3. It cannot make human labor indispensable again, cannot guarantee enough replacement jobs, and cannot stop ownership of productive intelligence from concentrating.

The machine does not need to be taxed. Its owners need to be understood as the new economic sovereigns. Until that reality is confronted, robot-tax debate is a dispute over how to distribute the proceeds of obsolescence while pretending obsolescence itself remains negotiable.