If you believe Anthropic CEO Dario Amodei, India can see exponentially high economic growth rates of 20–25% if it deploys AI effectively. India recorded such mid-20s GDP growth only once, and that too post-pandemic due to low base effect. And, he knows that such a number “sounds absurd”.
However, Amodei still paints a “very bullish picture” that such growth can happen.
In a fireside chat at the India AI Impact Summit, alongside Nandan Nilekani, Co-founder of Infosys and the architect of India’s digital public infrastructure, he argued that India may be uniquely positioned to turn frontier AI breakthroughs into a large-scale economic shift.
Diffusion is Key
Amodei, in his essay Machines of Loving Grace, once wrote about AI creating a “country of geniuses in a data centre”. At the summit, he invoked the idea again, asserting that model capabilities are advancing at an extraordinary speed.
He added that systems are getting better at software engineering, mathematics, and even early-stage biomedical reasoning. “We’re on this very fast exponential,” he said.
But capability does not automatically translate into impact.
“There’s a duality,” Amodei explained. Even if AI development froze today, its economic impact could still be far larger.
Enterprises face friction in adopting new systems. Integration, workflow redesign, compliance, and trust-building take time. The lag between invention and diffusion remains significant, particularly in the developing world.
Nilekani agreed. Foundation models are essential, he said, but technological diffusion is “a different ballgame.” Technology alone does not scale transformation. Institutions, policy, guardrails, negotiation with incumbents, and public trust are equally critical.
Speaking from his own experience with digital public infrastructure, including the Unified Payments Interface, which now handles billions of transactions every month, Nikenani said, “Diffusion is a technique. It’s both an art and a science.”
“We’ll have to look at diffusion pathways to take this to everyone, and I think India will lead on that. That’s why I’ve always been saying that India should focus on becoming the use case capital of the world,” he added.
He warned that if AI fails to deliver a visible public benefit, backlash is inevitable. “The resentment of the blue-collar worker led to the train wreck of globalisation. The resentment of the white-collar worker is going to lead to the train wreck of AI.”
He added that there may be dire consequences if AI creates deepfakes or raises the power bill due to the steep energy demands of data centres. Amodei also agreed that in India, particularly, there are concerns around economic displacement.
For Nilekani, the only sustainable path forward is large-scale, tangible use cases. Farmers should earn more. Children should learn better. Citizens need to access healthcare and public services in their own languages.
Why India is Different
Amodei said India stands out for its developer energy and institutional readiness. Usage of Anthropic’s Claude models for coding and technical tasks has grown rapidly, reflecting what he described as “brimming excitement” among developers in India.
Amodei also mentioned Anthropic’s partnership with the EkStep foundation, which will help the company get access to India’s digital public infrastructure.
He pointed to India’s scale as a strategic advantage. With a vast population and growing research ecosystem, the country could become a powerful engine for AI-driven medical breakthroughs.
Referring to Google DeepMind’s AlphaFold, an AI model that predicts proteins’ 3D structure, Amodei said the world needs “50 improvements like that.” India, he suggested, could help accelerate them.
But scientific breakthroughs alone are not enough. Translating them into a broad societal impact requires institutional pathways.
Nilekani announced a new initiative, 100 Diffusion Pathways by 2030, aimed at developing playbooks that combine technical, regulatory, and institutional lessons for deploying AI at scale. The coalition includes EkStep Foundation, UNDP, People+AI, Anthropic, Google, and the Gates Foundation.
Nilekani framed the stakes bluntly. “Right now in AI, there’s a race to the top and a race to the bottom,” he said. “And the race to the bottom is faster than the race to the top.”
The answer, he argued, is faster diffusion. AI must move beyond data centres and elite enterprises to work at scale in agriculture, education, healthcare and small businesses.
With its digital public infrastructure legacy and population scale, India may become the most consequential laboratory for that transition.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.