AI is creating lakhs of low-paying jobs in Gurugram, Noida. It is India's new BPO - ThePrint

TEXT START: Play the video.

The Dissection

The article exposes the hidden labor beneath AI: low-paid Indian workers perform repetitive annotation under surveillance, strict targets, and disposable employment conditions. It correctly shows that AI is not autonomous magic. But its dominant frame—“India’s new BPO” and a “short window of economic opportunity”—turns temporary scaffolding into a supposed employment strategy.

These workers are not building a durable new labor market. They are feeding the machine that will reduce the need for their own labor. The article documents the contradiction, then partially disguises it as development.

The Core Fallacy

The central error is treating job creation during AI’s construction phase as evidence that AI preserves mass employment.

Annotation is valuable precisely because it converts human judgment into training data. Once models improve, annotation is compressed through automation, synthetic data, active learning, and smaller human verification layers. The worker is not a permanent complement to the system; he is bootstrapping it.

The BPO analogy is structurally backward. BPO commercialised repetitive human work. AI annotation industrialises humans for the purpose of removing humans from repetitive work. Calling the task “high value” does not change its wage, substitutability, or trajectory. A job can be essential to building a system and still be terminal as a career category.

Hidden Assumptions

• That annotation demand will expand faster than the technology’s ability to automate annotation.
• That low-skilled workers will climb into durable technical roles, despite the article describing work requiring basic English and computer literacy.
• That cultural and linguistic representation creates bargaining power for the people supplying the data. Representation without ownership is extraction with better symbolism.
• That the BPO boom provides a valid model for the future, ignoring that AI’s competitive objective is to eliminate the need for human repetition.
• That “human-in-the-loop” means humans remain economically indispensable. In practice, it can mean humans are temporarily retained as a cheap error-correction layer.
• That market growth and workforce growth translate into worker prosperity. The article itself shows the opposite: expanding output, stagnant wages, surveillance, and weak employment security.
• That clients will preserve large human workforces instead of consolidating suppliers and demanding cheaper, faster, more automated pipelines.
• That creating work is equivalent to preserving productive participation. Under the Discontinuity Thesis, it is not.

Social Function

Classification: partial truth, transition management, and ideological anesthetic.

The partial truth is real: AI currently depends on invisible human labor, and Indian workers are supplying it cheaply. The transition-management function is also clear: present the coming displacement as a fresh BPO opportunity and encourage another generation to enter the pipeline.

The anesthetic lies in the language of opportunity, inclusion, and Indian representation. Those things may improve models, but they do not alter who owns the AI capital, who captures the surplus, or who becomes redundant once the training pipeline matures. The article shows the workers training their replacements, but stops short of treating that as the governing economic fact.

The Verdict

This is not AI creating a durable mass-employment solution. It is AI hiring its temporary scaffolding in Noida and Gurugram.

The workers are already experiencing social death: low wages, disciplinary surveillance, weak security, and no meaningful control over the system they make possible. Mechanical death follows as annotation becomes more automated and the human layer contracts. The “new BPO” is therefore not a new engine of mass prosperity. It is a waiting room between the old wage circuit and its replacement.