India less exposed to AI job risks than others: Goldman - ET Enterprise AI

TEXT START: The services sector may face some impacts due to AI, but key areas like construction and retail remain unaffected for now.

THE DISSECTION

This is a reassurance memo disguised as labor-market analysis. It takes current task composition, current adoption rates, and a five-year window, then inflates them into a national conclusion. It acknowledges substitution in call centers, IT, postal, and telecommunications, while using construction and retail’s present exposure as a shield against the larger question: what happens when AI lowers the cost of coordination, administration, supervision, and service delivery across the economy? It also substitutes aggregate productivity gains for the employment and ownership question.

THE CORE FALLACY

The text confuses delayed exposure with structural immunity. Physical or mechanical work is a lag defense, not a permanent moat. AI can first automate the cognitive layer surrounding those jobs and later enable more machine-based coordination. A large workforce does not protect labor; it creates a larger reservoir of people competing for fewer economically necessary tasks.

Most importantly, productivity gains do not automatically accrue to displaced workers. They can increase output while reducing labor demand and concentrating control in AI owners. “Benefits outweigh job losses” is a GDP statement masquerading as a human-survival statement.

HIDDEN ASSUMPTIONS

• Gradual rollout can be controlled despite competitive pressure to cut costs.
• Construction and retail remain outside the automation frontier.
• Service substitution stays limited to call centers and the few sectors named.
• Productivity gains become wages and new jobs rather than owner returns and labor displacement.
• A five-year balance predicts the long-run outcome.
• Workers displaced from services can move into physical roles without wages collapsing.
• Aggregate national gains compensate the specific people made redundant.
• Current non-impact persists as AI systems, robotics, and coordination tools mature.

SOCIAL FUNCTION

Partial truth functioning as ideological anesthetic and transition management. The article’s facts—current concentration of labor in physical tasks and current service-sector exposure—are real lag conditions. Its social function is to turn those lags into reassurance, make gradual deployment sound benign, and postpone the ownership and distribution question. It is also elite self-exoneration: if productivity rises, the casualties can be labeled an acceptable technical adjustment.

THE VERDICT

India is less exposed early, not exempt. Its labor composition may delay P3, but it does not defeat P1 or P2. The article measures how slowly the guillotine descends, not whether the blade exists. If cognitive automation achieves durable superiority and institutions cannot preserve human-only economic domains, India’s physical-heavy workforce becomes a transition reservoir, not a sanctuary. The Goldman thesis is a lag forecast inflated into a survival thesis—and structurally false.