It took India 250 years to catch a break after its textile artisans lost out to new technology in the 18th century. How long will it take for the country to find the next growth frontier in the age of artificial intelligence?

By Andy Mukherjee The upheaval of the past came in waves. First was Hargreaves’ jenny. Drawing, twisting and winding 16 or more threads at a time, it outproduced the spinners making some of the world’s finest yarn. Then handlooms gave way to power looms. (British protectionism didn’t help, either.) A century later, Mahatma Gandhi turned the symbol of wounded pride — the ancient spinning wheel working a lonely spindle — into an emblem of India’s anti-colonial struggle.

Expect a similar story to play out as AI blunts India’s edge in modern craftsmanship: the maintenance of the world’s business software. Autonomous AI agents are already shrinking the value that coders in Bengaluru and Hyderabad bring to global enterprises. Technology services are a $250 billion export engine supporting 6 million jobs. Now local politicians are wooing Western cloud computing companies with liberal land grants and subsidized solar energy, which could potentially make India a go-to source for cheap tokens (a measure of AI computing power). But data center investments will create few new jobs.

Prime Minister Narendra Modi’s vision to make India a developed economy by 2047 required 9%-plus growth in per capita dollar income even before coding tools such as Anthropic PBC’s Claude and OpenAI’s o3 arrived. With global businesses like Starbucks seeking to become their own software companies by replacing companywide systems from International Business Machines, Microsoft and Oracle—applications Indian programmers specialize in maintaining — that ambition looks more daunting than ever.

The angry youth that clashed with police on the streets of New Delhi this summer were not putting up a Luddite resistance to AI; they were merelying their frustration with the state’s failure to ensure the fairness of university entrance exams. Still, the fraying link between education and good jobs is a palpable subtext for the self-proclaimed “cockroach” movement, named after a perceived put-down by India’s chief justice of some unemployed youth. (The justice said he was misunderstood.) The stakes for a Generation Z population of almost 400 million are immense. If nominal per capita dollar income keeps growing at the past decade’s 5.5% rate, India will be old before it’s rich.

To join the league of wealthier nations, a country must create jobs, especially better-paying ones, in export-oriented sectors.

But despite committing billions of dollars in taxpayer-funded incentives over the past five years, India’s share of global merchandise exports is stuck at less than 2%. Vietnam is pulling ahead in manufacturing. What should New Delhi do differently?

Policymakers are fixated on industrial policy, or using the government to steer investment into favored sectors. Trouble is, nearly all countries — developed and developing — have exactly the same idea nowadays. What’s more, both the government and the entrepreneurial class in India are ill-suited for the mandate.

Local businesses like to corner fiscal incentives, such as those for producing electronics and smartphones. But assembling iPhones with imported components creates jobs that pay 18,000 rupees ($188) a month, which, as former Reserve Bank of India Governor Raghuram Rajan noted recently, is not all that different from what a chauffeur or household helper makes in a large city such as Chennai.

India’s public finances are too precarious for that model, and its business leaders are too risk-averse. A permanently undervalued currency can’t be a strategy either. The West won’t enable the rise of another billion-person economy after being caught asleep at the wheel as China went from a supplier of inferior goods to a challenger in modern technologies.

Instead of trying to build an entire supply chain by coaxing a domestic private sector that has no technology of its own, India could rely on the existing R&D centers of global leaders. Large tech companies Advanced Micro Devices, Intel, Nvidia and Qualcomm employ thousands of engineers in India for cutting-edge R&D. Give them tax breaks for investing in underfunded public universities. Let them set up advanced testing, simulation and verification labs — not just within elite institutions but also the broader network of engineering schools. They can produce all the scientists and researchers they will need, and more. The surplus talent will drive a startup boom where the founders’ ambitions go beyond 10-minute food delivery.

A similar logic applies to data centers. Token factories using local land and power could contribute, say, 5% of their high-bandwidth memory chips toward an existing national research cloud, accessible to universities and research centers. Any resulting scientific breakthroughs could finally spur local industry to invest in real R&D.

Multinational companies, however, will do the heavy lifting. When I visited the John F. Welch Technology Center in Bengaluru two years ago, engineers were collaborating with colleagues at GE Aerospace’s Niskayuna, New York, operation on a novel hybrid-electric jet engine platform—an effort validated this July in the world’s first high-altitude hybrid-electric flight. Multinationals already employ 2.4 million professionals in India, generating nearly $100 billion in annual revenue. The real policy question is how to supercharge this sum to $1 trillion. Embedding local research directly in the gear and services global companies sell worldwide will be far more remunerative than getting paid for lines of code.

That is why young Indians care so deeply about fairness in university entrance exams. They can see the power of a good education even in the age of AI, and they don’t want it denied to them. With a substantial lift to the quantity and quality of the 3 million undergraduate and graduate degrees India mints annually in science, technology, engineering and medicine, the nation can conceivably expand its frontier cutting-edge talent pool by tenfold over a decade.

Each of these R&D professionals, earning globally competitive wages, can generate the badly needed tax resources for public employment and urban infrastructure. Better-paid teachers, doctors and frontline health workers will start filling out the middle of India’s employment pyramid, where income and wealth are far too concentrated at the top. Even as it boosts the quality of civic services and expands career opportunities for women and socially disadvantaged caste groups, the state must loosen the grip of paper-shuffling bureaucrats holding the economy back.

But while this vanguard of future growth needs to be nurtured, those who aren’t at the same level of skills need to be pulled out of a low-productivity rut. More than 15 million Indians work as handloom weavers, knitters and tailors, very often in solo establishments. An additional 5 million are employed by mom-and-pop food-processing operations. When they get even a little bit of scale and capital, the value added per worker jumps from about $4 a day to $6 or even $8. If India wants to emulate Taiwan’s success in manufacturing, then it must also follow its path and build up its small businesses first.

India’s window for this transition won’t stay open forever. As AI agents proliferate and take over more tasks, there will still be demand for human engineers to supervise the code they write along the way. But those specialist positions won’t follow the logic of labor-cost arbitrage that made outsourcing such a lucrative trade. The jobs can come up just as easily in Nashville as Bengaluru. And if it’s the former, Indian techies will be at a disadvantage because Washington’s stance on immigration has also hardened.

In 1804, a few decades after the spinning jenny, Joseph Jacquard invented a mechanical loom to control the weaving of complex fabric patterns using a sequence of punched cards. Those cards contained the seeds of binary code — the zeros and ones that would eventually give birth to the modern computer. Between the factory automation that impoverished India’s artisans and the computer revolution that finally gave the country a way back in, a quarter of a millennium was lost.

Your $30 Jeans Are Running Out of Cheap Power: David Fickling How to Fight the Data Center Power Grab: Pedro Sanchez Quiet Quitting Hits India. AI Will Make It Worse: Andy Mukherjee