Thank you, ladies and gentlemen. Welcome to the India Today AI Summit 2026, being held parallelly to the AI Summit Impact that the government of India has done, that's brought together thought leaders and visionaries from across the world. And I am delighted this morning to start the session with Hemant Taneja, CEO of General Catalyst, Pashchim Vihar boy as he reminded me, now of course in the Bay Area, and someone whose company, as a global venture capital firm, has invested in more than 800 businesses, including crucially in the area of applied AI. And we're going to focus on applied AI in a moment, but first, welcome him back to your roots in a way. But I want to start by asking the big questions that's dominated a lot of the conversation globally, in Davos more recently: Is AI still to be seen as a bubble, or is it the future that's already here? Let's settle this once and for all.
Well, as I always say, bubbles are very good for us. I think what bubbles do is they mobilize talent and capital behind important problems, important technological areas. So I think in a lot of ways, if you think about it with that perspective, we are in an AI bubble, but in the sense that there's so much investment going into it. There are some areas where there'll be some money lost, but on the other side of what we're going through now, in a few years, you'll see we'll build some incredible companies and we will change the world with it.
The reason I asked you that is because surely there must come a time that beyond the hype, companies like you all will demand profits. That it's not about only AI foundational models, it's about creating products that will give profit. Are we therefore entering this post-hype phase now where people are getting real and saying, look, you're going to have to show us products where AI will actually deliver hard revenues and profits?
So look, I'll separate the AI investments into two categories. One is there's continued investment into research, and the whole idea is that the token prediction models, the language models, they will only take us so far from an intelligence perspective and that you'll need fundamentally new breakthroughs. So there's a lot of capital going into that particular category, which is hard to assess when is that going to become a business, but the promise of it is why capital keeps going into those projects. But for most of it, which to me is about the applied AI layer, I really don't think that we're looking for proof points and that they're not showing up. I would actually argue 2025 was the year that diffusion started, and the whole investment case of these hundreds of billions of dollars being invested in the compute infrastructure is all about: will you actually transpose labor spend into productivity, into AI productivity? And you started to see evidence of that in enterprises in meaningful ways late last year. So I do think you're going to start seeing the profit margins come as a result as well, because now you're doing really high-value work with AI.
Give me examples of that, because applied AI itself, the specific areas where you believe that AI will now turn into revenue, into a legitimate revenue-profit model. You yourself are investing with existing companies to make them AI-ready. Is that the way for the future, or is it about ensuring that you create new products in specific industries? And if so, which are those specific industries where applied AI will be transformational?
So look, AI is a transformative force and the biggest value is about applying it to enterprises. So the way I think about is the value getting created or not is: look at every department and think about what is AI enabling to transform those departments in a business. So if you follow the news, you would say over the last few months you're seeing a lot in the programming department, in the coding department, with Anthropic and the cloud models. The continued momentum in the advancement of that technology is really stunning. You've gone from can these models help you write faster to they're actually generating pieces of software for you. Right? I think that's a pretty profound implication. So rather than is this a software tool and therefore a software budget for an engineer, you're actually getting work done that teams of engineers would do in a matter of days and weeks as opposed to months and years. That's high value. So I think that if you think about what's the margin in that, you could actually charge for those things a lot more. Imagine there's a version of that going on in customer support and customer success broadly. There's a version of that going on in sales and marketing. Every department is going through that kind of a transformation.
But if I were to specifically hone down areas where you believe applied AI will be truly transformative, to give you examples, healthcare in India is seen as a huge opportunity out there. Education is seen as a... do you look at those areas, or do you look at existing companies and see how can I transform those companies?
So I think let me think about the altitude of what you just said. Step one is to transform the enterprises, right? If you can make them AI-native, then they get unleashed to do all of what you're saying. If an enterprise becomes AI-native, then in its own industry it can go from being operationally efficient to having an offering of abundance for their customers. So if you can actually create an AI-native hospital that has not only nurses that are taking care of patients in person but also agents that are on the phone that effectively cost, you know, a deflationary way essentially to zero over time, you would have a free nurse that takes care of everybody, right? So the transformation is only going to happen when you take your enterprise and transform and then start to imagine how your offerings should be fundamentally different. And I would sort of use the framework around abundance to take care of your customers and constituents.
The reason I also asked you is because India's experience has been we produce a large, strong class of engineers, a large pool of engineers, but we are producing fewer AI product companies. And let's be honest, I mean the Chinese and the Americans have been way ahead when it comes to innovation at the moment and enterprise. So what are we missing here? Are we missing capital, risk appetite, or product thinking? If we are to also join this engine of applied AI, what does India need to do?
I'm very optimistic on India. I would say India is behind when it comes to development of infrastructure and models. I think that's going to be a hard place to go compete and try to be a leader. But the way we think about our investment landscape, we say there's the infrastructure and models, but then there is the application of AI. And every single company that is doing really well in the US at the applied AI layer, you know, what they have as a core skill set, they know how to engage with customers in what's now called forward-deployed engineering capabilities. India's IT industry has matured actually into that. The workforces that you have here are actually pretty good at it. They just need to be applied towards going to these customers with the AI toolset now versus your typical digital transformation toolset and go solve their problems with AI and create these next-generation companies. I do think that's going to get unlocked and create some global companies that start right here with that kind of a mindset.
So do we piggyback in a way on what the Americans have done or what the Chinese have done and become the sort of back, you know, famously 25 years ago we became the outsourcing capital over time, is that the model that we take? Or do we encourage Indians to innovate and compete in the product space?
Look, I think the definition of products is actually changing in terms of what it means in AI. You have such fast-moving technologies that it's hard to understand what the capabilities of the models themselves are. So what do you really build on top? And so the products that get built are almost thought about in the context of the problems you're solving for your customers versus you have a minimum viable product that you can go sell over and over again. So when you sort of grapple with that and understand these are the capabilities of AI, it doesn't need to be a discrete product. It needs to understand the problems and solve them structurally. The whole innovation landscape changes, and the talent base here is actually well-positioned to go do that and create these next-generation companies for different industries that are using the AI toolset.
But when you, as a global venture capital firm, look at Indian companies, what are the kind of elements that you're looking at when you decide I am going to invest in this specific company? When you're looking at applied AI from your perspective, what kind of... are you looking at companies which will essentially scale up in the shortest possible time? And if you can give me an example.
I think a couple of things. One is when it comes to India in particular, because AI is deflationary, how to bring capabilities in healthcare, in education, to the masses is actually going to be a really interesting opportunity. If you can solve it for a billion people with all the complexities that lie here in India, you are going to be well-positioned to solve those problems globally. So I think that's a place where, just like India did with UPI and Aadhaar, the innovations that leapfrog the world, I think you're going to see some innovations that come out that way. The second thing is, you know, can you unleash the technology talent here, give them access to these tools, make the use of AI pervasive, and then go build globally, much like India has started to do in the SaaS world. I just think that the paradigms change. You know, SaaS is not where you're going to create the most valuable companies because AI can do that work for you. Well then, let's use AI and go solve problems for customers with that new technology platform versus the older metaphors.
Okay, let me for a moment just show you, since India Today had an AI anchor that we introduced on air a few years ago, started with weather. There's a question that AI Rajdeep has for you, and I want you to look at that question and answer it for me. Take a look.
Thank you, Rajdeep. When I see you, it seems I'm looking at the mirror. But first, my question to Mr. Taneja is: advocates of AI claim that AI will lead to a second industrial revolution. Do you agree? Thank you.
I do. Look, I think the key important thing with AI as we look at building the new businesses, as we work with entrepreneurs everywhere in the world, is you have to think from first principles. The way you build products is different. The way you want to build your own companies is different. The way you engage with customers is different. And so it is a next phase and it is a next revolution in the context of what AI can do. And imagine it actually making a difference in the physical world as well, which is, you know, we don't feel it as well today, but in the next decade we'll be talking about that as well. So I think it's pretty profound what's about to come in what this technology enables for us.
