Mission AImpossible

India has built an LLM, has good startups & upcoming data centres. But that’s not enough 

Big things happen when people put their heads together. In Feb 1974, a handful of Taiwanese engineers and officials met for a breakfast meeting. Taiwan’s economy depended on labour-intensive factories at the time, but the breakfast group saw competition from neighbouring countries ahead. They decided it was time to go high-tech. Four years later, Taiwan was not only making semiconductors, but was also third in terms of digital watch exports. It’s Feb again, and world’s AI captains, along with armies of delegates, are gathered in Delhi. At India AI Impact Summit 2026, there will be many opportunities to put heads together. But what does India want from these meetings? 

There’s the official manifesto that says AI should develop keeping people, planet and progress first. A laudable aim, but far removed from reality at present. As to people, talk over the past few days has dwelt on the coming jobocalypse – annihilation of entry- and middle-rung white-collar jobs. Regarding planet, AI’s water and energy demands are well-documented. Progress under these conditions is questionable. So India cannot realise its vision if all it brings to the table are high moral principles. What it needs is the underpinnings of AI: its own research, data centres, foundational models, and popular applications. 

True, US seems to have an unassailable lead in AI today. But so was the case with semiconductors, and rare earths, at various points last century. It gets off the blocks fast with a combination of wealth, cutting-edge research, and first-mover advantage. That doesn’t mean the race is over. And AI, especially, is still in its infancy. While US firms hog headlines, India has quietly built BharatGen, its own LLM that supports 22 languages. States like Andhra have drawn multi-billion data centre commitments. IIT Madras has emerged as a cradle of startups, drawing over $2bn in VC investments. But it’s time to shift gears now.

As the accompanying article says, GOI and the private sector must invest more in R&D. It doesn’t cost as much as the infra buildout, or the long-term operational costs, but is the bedrock of true aspiration. Because AI infra requires tens of billions of dollars, we will need foreign capital, but good projects shouldn’t be held up because of funding gaps. Govt must play enabler, sitting down with researchers and industry to spot obstacles, and remove them. Taking care not to entangle them in red tape. We have a vision, let’s put heads together to execute it.

END OF ARTICLE