Nearly 197 million Indian workers in jobs vulnerable to AI automation, finds Great Lakes ...

TEXT START: Nearly 197 million Indian workers are in jobs with a high share of routine tasks that could eventually be automated, while around 46 million work in occupations where artificial intelligence (AI) could improve productivity, according to a new study.

The Dissection

The article maps exposure, substitution risk, and regional differences, then channels the findings into a policy-management narrative: teach AI, expand access to tools, retrain vulnerable workers, and act early. Its strongest contribution is measurement. Its weakness is that it treats the destruction of labor demand as a workforce-preparation problem.

The 197 million figure is the blast radius. The 46 million complementary workers are the smaller group positioned to capture productivity gains. The article acknowledges this asymmetry but does not follow it to its structural conclusion: productivity gains can increase output while reducing the number of humans required to produce it.

The Core Fallacy

The central error is confusing task-level adaptability with continued economic necessity. Training may help some workers move into AI-complementary roles, but it cannot create enough scarce, indispensable human labor if AI steadily performs more cognitive work at lower cost.

The article also treats uneven adoption, deployment costs, and machine limitations as meaningful protection. Under the Discontinuity Thesis, these are lag mechanisms, not reversals. Once cost and performance advantages become durable, firms have a competitive incentive to automate. The transition may be slow, but the direction is not negotiable.

The report implicitly assumes that productivity gains will circulate back into employment through new tasks, higher demand, or better wages. That is the exact wage-to-consumption circuit AI threatens to sever. A more skilled surplus population remains surplus if the system no longer needs its labor.

Hidden Assumptions

• AI adoption will remain partial rather than expanding from tasks into whole occupations.
• Human skills can remain scarce enough to preserve broad wage-based participation.
• Retraining can move workers into the small class of complementary roles faster than automation expands.
• Firms will distribute AI-created gains through employment instead of concentrating them among owners and controllers of AI capital.
• State policy can coordinate a stable human-only economic domain at national scale.
• Measured “AI exposure” is merely a risk indicator rather than an early measure of productive displacement.
• Women’s relatively higher share of complementary work will translate into broad gains, despite the article’s own finding that only about 18 per cent of women aged 20–29 are in paid employment.

Social Function

This is partial truth wrapped in transition management and ideological anesthetic. The exposure data are not copium; they identify a real and unusually large vulnerability. The anesthetic begins when education and training are presented as if they can preserve mass productive participation.

The article performs a useful administrative function for institutions: acknowledge the incoming disruption without admitting that the labor market may lose its central role. “Act early” becomes a substitute for answering the harder question—what happens when early action still cannot manufacture enough indispensable human work?

The Verdict

The report maps the corpse before declaring the patient treatable. Its numbers support the Discontinuity Thesis more strongly than its recommendations do: hundreds of millions of workers occupy roles exposed to routine-task automation, while the immediately complementary group is far smaller. Skills policy may create a temporary layer of better-positioned servitors, but it cannot preserve the post-WWII employment–wage–consumption system once AI makes human labor broadly optional.