Pick any conversation on AI in India, one project keeps coming up again and again: UPI. Every major stakeholder — across government and industry — seems to be interested in replicating the success story of the payments interface. Praveen Gopal Krishnan argues that Sarvam AI — the startup that is a VC darling and government favourite at once — is at the heart of this narrative.

Here is the problem. 

The UPI moment came to India in very specific circumstances. We did catch “lightning in a bottle” as Rohin Dharmakumar puts it. 

But it is unlikely to happen again.

The AI supply chain is fundamentally different. There are many moving parts and a range of use cases to be solved for. A state-led centralised approach — with a main character around which everything pivots — is not the answer.

What are the alternatives? Competing AI companies. The government enabling innovation in a diffused manner. Research incentives. Encouraging competition so that Indian AI companies are at par with the best in the world.

There is precedent to this. Brady Ng, our resident China expert and tech nerd, gives us a history lesson.

And if you have been wondering where the foundational model is, Sarvam AI is expected to unveil one soon. We will cover that ground when that happens. Until then, tune in for a round up of India’s AI strategy and its main company.

------
Zeus, the mascot of Zero Shot, was generated using AI. Everything else is made by humans, just like all articles, columns, newsletters, and other podcasts created by The Ken.

This episode was produced by Vidhatri Rao and edited by Rajiv CN. Share your comments, critiques and suggestions with us at zeroshot@the-ken.com. Or write in just to say hi. We would love to hear from you!

------

Additional Resources: 

Sarvam AI’s Rs 10,000 crore pivot

India’s ‘UPI Moment’ in AI has Arrived

For AI, India can build on the Aadhaar-UPI model

Voice AI is India's next UPI moment: Infosys chairman Nandan Nilekani

The Economic Survey

PhonePe dominates payments but loses money. Now what?

AI is not UPI: Why going by the UPI model risks stalling progress on artificial intelligence