Sridhar Vembu says AI job losses need new work, not universal basic income - India Today

TEXT START: Sridhar Vembu has argued against universal basic income as AI threatens intellectual work.

The Dissection

This article relocates the post-AI labor problem instead of solving it. It accepts that AI may hollow out software, mathematics, and other intellectual jobs, then tries to preserve a human-centered wage narrative by reclassifying care, ecological restoration, culture, and religious practice as new work. It correctly surfaces ownership, but supplies no mechanism converting near-laborless output into durable pay for billions. Its repeated sections add volume, not evidence.

The Core Fallacy

The article confuses social value with market demand and payment with productive necessity. If AI performs economically necessary cognitive work at lower cost, paying humans for nature, nurture, culture, or scripture requires transfers, patronage, or protected domains. That is not a self-sustaining replacement for wages. It is UBI or a dividend wearing a job-shaped costume.

The text also treats anti-monopoly policy as a switch governments can simply flip, ignoring P2: institutions cannot reliably preserve human-only economic domains at scale against competitive pressure.

Hidden Assumptions

• AI-generated abundance will be broadly distributed rather than captured by its owners and controllers.
• Governments can disperse ownership despite concentration, competition, and institutional inertia.
• Billions of displaced workers can be assigned scarce, measurable, well-paid roles.
• Care, restoration, art, ritual, and sport can support mass incomes without subsidy.
• Human economic necessity is required for dignity; the article never separates meaning from production.
• Socially valuable activities will automatically become remunerated activities.
• The transition will remain orderly long enough for these new payment systems to be built.

Social Function

Partial truth, transition management, and ideological anesthetic. The article correctly identifies the exposure of intellectual labor and the importance of ownership. Its anesthetic is the promise that displaced people can remain workers in culturally respectable forms. That avoids the harder admission: after P1 and P3, broad human employment may be socially desired but no longer structurally required. It makes post-work transfers politically acceptable by insisting they be called earnings.

The Verdict

Vembu identifies the ownership problem but misidentifies the route back to life. Distributed AI ownership may create Sovereigns and durable Servitors; it cannot manufacture billions of economically necessary human jobs. Nature, nurture, culture, and scripture are transition niches unless society pays for them through dividends, public provision, or protected patronage.

The article therefore does not refute UBI. It proposes UBI’s transfer mechanism with a moralized work wrapper. Under the Discontinuity Thesis, the post-WWII wage-consumption circuit still dies. The decisive question is who controls the AI surplus and who receives the carcass—not whether ceremonial human activity can resurrect mass employment.