AI 革命の代償：労働なき成長と人間の犠牲

At the turn of the millennium, India became the global hub for medical transcription. With fiber optic cables laid in the 1990s, American hospitals found it cheap and efficient to beam voice files across the ocean. Indian workers, fluent in English and medical terminology, turned dictations into polished reports overnight. The time-zone gap became an advantage: a surgeon in Boston could dictate before bed, and wake up to flawless notes typed in Bengaluru.

0:00/9:03

By the early 2000s, the industry seemed secure. Companies built training academies, invested in IT systems and promised young recruits that this was a career with a future.

But when I met Aakash (name changed) in June 2024, that promise had already unravelled. At 25, he was clinging to a short-term contract in a Bengaluru firm. He had joined a year earlier, after months of training. During recruitment, he asked the CEO whether AI posed a threat. ‘At least five years away,’ came the reassurance.

Yet, by March 2024, as the financial year closed, American clients began to vanish. Contracts were cancelled or shifted to automated platforms. Hundreds of jobs disappeared almost overnight. The cafeteria that once buzzed at 3 a.m. emptied out. Hiring froze; the recruitment team itself was laid off. New employees were brought in under “conditional retention training,” only to be dismissed weeks later (an accounting trick to show inflated headcounts to prospective clients) and workers lived in limbo. Economists have a phrase for jobs like Aakash’s: high exposure, low complementarity. Which means that these are roles where machines can easily take over, rather than raising the human’s capacity to work.

In Manila, tens of thousands of Filipino transcriptionists have already been displaced. In Nairobi, call center operators now compete with chatbots for contracts that once fed families. In Colombia, customer support roles vanish into generative systems. The pattern is the same: codified knowledge that once built middle-class livelihoods is being swallowed whole by automation. Even the people training the systems feel the blade at their neck. Contractors working on products like Gemini and AI Overviews describe a familiar dread: they are annotating their own obsolescence.

Unlike earlier waves of mechanization that chipped away at manual labour, these systems target the cognitive core of professional life. The scale of this change is huge.

Even jobs that once seemed safe because of their creativity or complexity are now at risk: bioengineers at 84%, mathematicians at 80% and editors at 72%. What once felt untouchable, the world of judgement, invention and artistry, is now being taken over by algorithms.

The middle-class dream that powered India’s outsourcing boom is rapidly unravelling.

In finance and IT, the tide is rising the fastest. Reports once drafted by junior analysts, code once written by entry-level developers and risk models once tweaked by human hands are all being re-channelled through machine logic.

If current trajectories hold, the world could soon see the automation of a quarter of all work, erasing the equivalent of 300 million full-time jobs. The labor map stands to be completely recalibrated.

In September 1945, New York City froze as more than 15,000 elevator operators, doormen and porters walked off the job. The Empire State Building, the Chrysler Building and the countless towers that symbolized American modernity were reduced to stranded shells.

Businesses stalled, government offices slowed and even the federal treasury bled millions a day in uncollected taxes. Mailrooms were piled high with undelivered correspondence. Railways struggled to function as station lifts stood idle.

What startled many was the revelation of dependence. Within a few years, manufacturers redesigned the technology: emergency phones, stop buttons, alarm systems, automated doors. By 1950, Otis had installed the first fully automated elevators. By the 1970s, the operators had vanished.

If the lift operators’ strike showed how technology erases professions one by one, the story of AI is how entire classes of work can be hollowed out at once. The benefits are unevenly shared, and the costs, which are immense, fall hardest on the young, the clerical, the feminized and the Global South.

For every AI ethicist or prompt engineer role created in San Francisco, dozens of livelihoods vanish in Manila, Johannesburg or Mumbai.

In India, for example, the promise of IT as a secure middle-class ladder is crumbling. Between 2022 and April 2024, the Indian tech sector shed more than 500,000 jobs, with 425,000 layoffs in 2023 alone. The former CEO of HCL Technologies has warned that as much as 70% of IT roles could disappear. In fact, net hiring across India’s top five IT companies in the first nine months of 2025 was just 17. The middle-class dream that powered India’s outsourcing boom is rapidly unravelling.

This is not just an Indian story. Globally, the technology sector continued deep workforce restructuring in 2025, with independent trackers reporting roughly 122,500 layoffs across 257 companies, and broader analyses indicating that over 244,000 tech workers were cut worldwide, often with efficiency gains from automation and AI cited as contributing factors.

Behind every statistic is a story like that of Kamlesh Kamtekar, a graphic designer from Mumbai who retrained in 3D animation only to find the market shrinking under the weight of generative tools. His viral LinkedIn post described how he was forced to trade in his design career for driving an autorickshaw. Kamlesh’s journey is the lived expression of a broader divide: AI marches forward and leaves people behind.

In high-income economies, about 60% of jobs show significant generative AI exposure; in low-income economies, the figure is closer to 26%. On the surface, this looks like a cushion, but in truth, it means fewer opportunities to leverage new tools while still being hammered by falling wages and displaced industries. An International Monetary Fund (IMF) paper called this the Great Divergence: capital flows uphill to the economies best able to deploy robots and AI, while developing countries see temporary GDP declines and long-term terms-of-trade losses. Even in advanced economies where AI increases productivity, the benefits may not translate into broad prosperity because employment falls. Economists at the OECD note that generative AI could add up to 6.4% to GDP in advanced economies while displacing millions of workers. “Growth without work,” as they put it, is simply precarity disguised as efficiency.

For much of the 20th century, GDP growth and job creation moved in rough tandem—more output meant more work. That relationship has fractured. Now, the economy grows; the labor market does not follow. Economists call this phenomenon “productivity without participation,” where output expands and corporate margins widen, but the gains do not translate into either employment or wages for the majority.

A world where things get more “efficient,” but people lose dignity, and wages stagnate even as the economy grows.

A 2026 study by Anthropic’s researchers, measuring actual AI usage patterns against Bureau of Labor Statistics employment projections, found that occupations with higher observed AI exposure are projected to grow less through 2034. More striking still, the workers most at risk are not the low-skilled or poorly educated. They are disproportionately female, more educated and higher-paid, the very people who were told their qualifications would protect them.

In five of six countries studied, women are at a higher risk of displacement than men. The one exception is India, where the heavy presence of women in agriculture (a sector with minimal AI integration) masks deeper vulnerabilities elsewhere.

In effect, automation is creating three kinds of workers: those who can use AI to amplify their labor output, those who are erased by it, those who are excluded from it altogether. Advanced economies wrestle with the first; emerging economies face the doubled burden of the second and third. In consequence: a world where things get more “efficient,” but people lose dignity, and wages stagnate even as the economy grows.

The IMF’s AI Preparedness Index, spanning 174 countries, quantifies this divide: advanced economies score far higher than emerging and low-income ones, especially on digital infrastructure and human capital. For every AI ethicist or prompt engineer role created in San Francisco, dozens of livelihoods vanish in Manila, Johannesburg or

Mumbai.