Conversation with Bloomberg's John Fraa.
Nandan, it must be very nice to have the prime minister as your PR man.
No, no, I think he's very kind and very gracious. I'm grateful to him.
Well, we are so happy to have you today. I mean, so many of the transformations of modern India over the last 30 years that we heard about this morning from Prime Minister Modi and from Mike, you have had a role in. So for those of you — the very few of you who don't know — Nandan was co-founder of Infosys and he also helped build Aadhaar, the world's largest biometric identity system, which now covers roughly 1.3 billion people. And he played a central role in creating the digital infrastructure that made India's UPI payment system possible, bringing hundreds of millions of people into the digital economy. So very few people combine your understanding of how to create a truly global technology business with the know-how of how technology manifests itself in people's lives. So you are the perfect guest for our first morning of Bloomberg New Economy to help us think about how, I think, AI can be harnessed to help India become a developed economy by 2047. So thank you again. We can't really start any conversation about economic growth today without talking about AI. So let's first start, I think, by talking a little bit about AI sovereignty and India's role in the global AI story. You know, AI is often represented as a tussle, a battle for supremacy between the US and China, with everyone else fighting for scraps. But is that the right way to look at it?
Well, I have a slightly different perspective. I think everybody is saying that AI is a general-purpose technology. They compare it to fire, to, you know, the printing press, industrial revolution, whatever. But the thing about general-purpose technologies is nobody can own or control it. You don't think of the whole world's electricity being supplied by one company. You don't think of the world's books by one printer, and so on. So AI, in my view, is not a winner-take-all economy. It's going to have a huge number of players, a huge number of producers, and we are seeing that. Every week there's a new model being announced. So the whole thing is moving across the world — in India, in the US, in China. So the action, really, in a general-purpose technology is how it is used, how the technology is diffused in a society, how does it make businesses dramatically more productive, how does it make people in an economy do better. And I think in that part of AI diffusing and making it work for people at scale, India will be the leader, and that's where I think the action will shift. So today we're in the production phase — you know, who's making the chips and the data centers and all that — but that's all supply side. Ultimately, this has to work for people's benefit, and I think India will, precisely because we have the experience in implementing sophisticated technology at scale for the last several years.
I mean, I want to talk a little bit about the use cases and how India can help develop, you know, these practical examples of how AI can change people's lives. But I just want to stay a little bit, for a little bit, on the sort of the geopolitical or the strategic part of it. Is there not, though, a risk in your position that India becomes strategically dependent on the US and China, you know, if they're building the sort of the railroads?
Sure, but these are not choke points because ultimately, if you take a longer view — take a five- to seven-year view — there will be hundreds and hundreds of thousands of models. Many of them will be open source. Today we're already seeing very large open source models. In India itself, we are building — companies like Sarvam and BharatGPT are building very large models which would have been state-of-the-art maybe two years back. So I think there'll be more and more. You cannot bottle this technology. Similarly, on the chip situation also, I think you'll see a lot more chips coming. So ultimately, as I said, this is not a winner-take-all choke-point technology. This is something which is going to become the environment. And so, while we of course have to be careful about what happens, but I think by and large, if I take a long view, I don't see how anybody can create a choke point here.
So, if India becomes the crucible in many ways for these use cases, the testing ground for these use cases, can you give us some examples of where you're most excited?
I think, you know, the prime minister mentioned one example, which is what Amul is doing, providing a very, very capable chatbot for veterinarian advice to the cow owners, the dairy farmers in Gujarat — you know, thousands, a few hundred thousand of them. Or what we are doing in agriculture, where all farmers now are getting a very sophisticated source of information: access to market information, pricing information, weather information, soil information, whatever. Or what India is doing in language. India, as you know, has several hundred dialects and so on. And now, if you really want to make computers inclusive, a person should be able to talk to their computer in a language of their choice. And that's happening here as we speak. So I think all these are examples of how AI will be used in an empowering way, as the prime minister mentioned, so that millions of people benefit. And that experience we have got from our prior build-out of large-scale technology.
So I want to talk a little bit about the Indian economy and how technology can be used to build it more from here. However, over the last 30 years, you know, the Indian growth model — so much of it has been based on the growth of companies such as yourselves, Infosys and others. But yesterday we heard the managing director of the IMF, Kristalina Georgieva, talking about AI being a slow-moving tsunami wiping out demand for many human workers by taking over routine, repetitive tasks. And this is a narrative that's also been applied to India, has been applied to the software services industry here. You told me today you think this argument is hogwash.
No, I think this is all disaster marketing to suit the interests of large capital. The fact of the matter is that the requirement for companies like Infosys or others is actually more than ever before, because the transformation of enterprise is a very complex activity, and AI requires you to fundamentally reimagine the way you do business. And it's a root-and-branch surgery that is required. And this, in fact, even more requires firms like ours to show companies how to go about making those transformations, how do you reimagine the future. And I think we are going to see only more work coming our way. Of course, individual companies will have to get their act together. They'll have to transform themselves, reskill their workers, build new capabilities. All that is true, but this will only mean more things to be done. I mean, the capacity of AI technology is far ahead of its deployment. If you think about the cloud — the cloud was invented 20 years back. It's still not that ubiquitous. So technology diffusion in companies takes years and has to deal with complex issues of legacy, of change management, of power structure and so on. So I think there will be lots more to do. It'll in fact be more complex to do and therefore more rewarding.
And yet at the same time, if you look at the shares of companies like Infosys, TCS, others are all down 33% this year. Is there a narrative problem here? Do these companies need to explain themselves?
The markets will have their own view on this. I'm speaking as someone who's been closely involved with the tech change. If you look at the ID system, Aadhaar, which I was involved with, we designed it 18 years back. At that time it was obvious to us that this was going to work, but it took us seven, eight years to get to a billion people. So, you know, technology diffusion, penetration takes time. It deals with human beings. It deals with change management. It deals with anxiety and fear. So I think I'm convinced that there will be a big opportunity for transformation, which will require people to execute.
And where do you see most of those new jobs coming? I mean, the classical work of coding is definitely becoming automated. You have a lot of these coding agents from all the companies. Now you have open-source coding agents. So coding as an activity will become much more productive.
But the business transformation is not just about coding. I mean, in any technology project, technology is only 30%. 70% is governance, change management, transformation, guardrails. How do you make sure your AI bots don't hallucinate? How do you make sure your agents don't go rogue? These are the things you have to worry about if you want to make it work in the real world. You have a lot of other things to do. And that's why the new jobs will be in understanding client requirements, in collaborating, in getting people to work together. So there's a whole host of things to do.
And why is it, do you think, so — India, and we've heard this morning how India is incredibly good at using technology to solve specific problems, the UPI, biometric IDs, etc., etc. — but other problems here remain very intractable, and we're going to get on to education and unemployment. Why do you think that is?
No, I think in fact that's why I'm bullish on AI diffusion and AI application, because some of the most intractable issues we have faced — education, for example — I think definitely with AI we can make a big dent in literacy and numeracy for millions of Indians. Language — speaking language to the computer — will allow people to get access to the latest information, whether it's on agriculture or anything else. We'll have agents making life simple for small businesses so they can do more things. And I think India is already a very strong country for small business, and all the reforms that the prime minister spoke about, GST and all these things, have made small businesses more attractive to grow. And that will be where the job creation will come, because a small company of three to five people — you cannot really automate those jobs, because they do very complex things. In the morning they'll pick up the stuff, in the afternoon they'll stack the shelf, in the evening they'll deliver. These are not AI. They're actually AI-proof jobs. So once India unleashes its small business economy, millions of small companies will hire a couple of people each, and that's how the jobs will happen. So we'll see a new way of job creation enabled by open networks and AI, which is a paradigm we are not seeing. Therefore we can't visualize it.
So you're presenting a very optimistic view of where jobs in the future are going to come from. And yet, you know, you just need to look around Delhi today and read the news out of Delhi over the last week and over the last few months to see that India does have a serious — we know that India has a very serious youth unemployment problem. I think youth unemployment in the cities, based on official data, is 20%. Could well be a lot higher. Isn't it true that if a country like India, if you want to be a developed country, you have to bring all cohorts of society with you? And we saw that with China, we saw that with Singapore, many other countries presenting a vision to rally the entire society around. Looking at India from the outside, it seems like that's not quite happening here.
Well, I think India, like every part of the world, has to create enough jobs for its young people. I believe it's better placed to do so because it is also a very enterprising country. We have millions and millions of entrepreneurs. Those entrepreneurs have not had access to markets, access to capital, access to knowledge, access to agents to make their lives easier. And therefore, if we can do that, that's where the momentum will come from. So some of this has not happened. So we have to make it happen and make small businesses more productive. But I do believe that they will be the engine of job creation for the future.
