Uncategorized

all around the Bharat Mandapam. So once again, thank you very much for your time this afternoon and for choosing us to have a conversation with. To start off, I would like to introduce John, who will make some opening comments for the World Bank.

So thank you very much, Jeanette. It’s a great pleasure to be here speaking to all of you this afternoon. Over the past week, we’ve heard from a lot of world leaders, tech leaders, experts from across many, many countries about how AI is fundamentally reshaping our world, presenting not just a technological shift but a structural transformation with profound implications for economies and societies everywhere. For emerging markets and developing economies, as for all economies, AI could be a game changer. So sorry, that probably helps. I thought the mics were on. So, you know, for all countries, but especially for emerging markets and developing economies, AI can be a game changer, a unique opportunity to leapfrog longstanding development challenges.

It offers clear opportunities to enhance growth and productivity. We recently did some work in South Asia at the World Bank Group to see what sort of impact AI was having on jobs in the region, and we found that approximately 15 or 16 percent of jobs here have strong complementarity with AI. AI enables people in those jobs to expand their skills and their effectiveness in delivering the products and services that they are trying to provide. It also helps, you know, very, very diverse groups of people in many, many different sectors of the economy. It helps farmers to identify pests on their crops. It helps farmers to identify pests on their crops, diseases in their crops, and also how to address them.

It helps farmers to identify pests on their crops, diseases in their crops, and also how to address them. It helps nurses to identify the ailments and illnesses that their patients may be suffering, particularly the ones that they’re not very familiar with, but that they can research using appropriate AI applications. It helps financial institutions to understand better the ability of borrowers to take on loans, which, of course, expands the ability of the borrower to expand his or her business. So there’s clearly enormous potential for AI to fill skill gaps in the areas that I mentioned, also in education, in health care services, to detect patterns, to generate forecasts, to guide the allocation of public resources, and so on.

Of course, at the same time, on the flip side, AI also creates a number of challenges. One of them is there will be some job losses, particularly sort of entry -level jobs that are very much knowledge or document -based, performing relatively rote work that can be taken over by automation. And we’re actually seeing this in the World Bank Group. We went and looked at the number – the types of jobs that we are advertising these days compared to a couple of years ago, and what we found is that that layer, sort of at the bottom of the professional classes inside the bank group, there’s just fewer of those types of jobs being advertised in the World Bank Group today than there were a few years ago.

At the same time, you know, particularly for developing economies and emerging markets, many of them are going to struggle to harness the potential that AI offers because of very basic issues around the foundations for effective AI use. They may not have reliable electricity. We can start with that very basic one. They may not have an internet backbone that’s sufficiently strong. People in these countries may not have very, very basic skills of literacy and numeracy that enable them to work effectively with higher end devices. They may need to use very, very basic devices, not even smartphones, and rely on voice communication, asking a question and hearing a response. So there may be struggles of that kind in developing countries and emerging markets.

And I’m not even talking about all the governance and regulatory safeguards that can also come into play. So the question, of course, is how can emerging economies, developing markets, harness the potential of AI and avoid the pitfalls? And for us in the World Bank group, we’ve been very, very focused on focused recently on basically small AI. Small AI meaning practical, affordable, locally relevant AI that addresses specific problems and also works where connectivity, data, skills, infrastructure are fairly limited. And this is extremely important in countries like India where all of those conditions can apply. And yet there’s tremendous potential for people to expand their, to grow their productivity if they have timely access to information of the right kind in their local language tailored to their specific circumstances.

So that’s what we are trying to do in South Asia today and across the globe actually. And this is really about some of the examples that I mentioned earlier, having bespoke… applications that help farmers to do very basic investigation of the types of issues that they’re facing using their phone to analyze what’s going on to identify it to find out how to address it even to find out who within their local area in their market space can help them by providing the tools or the products that are necessary to address whatever they’re running into so India of course is a very strong example of what’s possible India has been a leading country in digital innovation for quite some time after the United States and China it has the largest if you like digital universe you in the in the world today it’s got some very good foundations there’s the the digital identity program as well as the digital payment platform that currently exists.

There are lots of Indian firms that are innovating in AI, including in the small AI applications that I’ve been talking about. And the governments of India have an objective of ensuring that there is AI for all. So they are very, very aware of the challenges that need to be overcome to make AI accessible to a very, very broad spectrum of the population and not just the very rich that, to some extent, need assistance the least, right? It’s the poorer parts of the country that benefit the most because they will be leveraging a tool that they are not very familiar with and have not been using that much in the past. So we’re working in India.

We’re working in a lot of different states, Uttar Pradesh, Maharashtra, Kerala, Haryana, Telgana. these different aspects working with governments to work on the foundational elements, interoperability, making sure that the accessibility is possible, that programs can run offline as it were so that people who aren’t able to get online all the time can benefit and so on. And then we’re also working with private sector investors who are developing apps. I mean we’re not actually developing many apps ourselves. That’s not really in our comparative advantage. Our comparative advantage as the World Bank Group is to do the more advisory work, make sure that the backbone information that’s embedded in the application is reliable and trustworthy because of course that’s critical for ensuring successful uptake.

But we are helping governments to create. We are helping governments to create the space that enables experimentation in AI sandbox to develop the different applications that people in this incredibly creative country are coming up with to help people get on with their work and become more productive. So I think it’s important to recognize that if we’re going to make effective use of this tool, we need both a public -facing effort to address the standards and the other issues, the interoperability and so on that I mentioned before, but also a private -sector -facing effort because it’s the private sector that’s actually generating, creating most of these applications that are working, particularly in the small AI area.

We’re doing a little bit on bigger AI. There’s obviously a connection between the two. Big AI can, through computational power, generate new knowledge that can help us to do things that we haven’t done so well in the past much, much better. But for… There are countries like India translating that. into small AI will also be very, very important for uptake. So I’m looking forward to hearing from all the distinguished speakers in this panel about their thoughts on what’s happening today in this sector. So thank you very much.

Thank you very much, John. John spoke about, of course, the use cases for AI, and on the other side of the spectrum we have the large language models, we have the foundational AI. But no matter where you sit on the spectrum, no matter where your interests lie, AI, innovation never disperses and never diffuses equally. Today on this panel, I hope to unpack what determines whether AI narrows the development gap or whether it widens the development gap. Especially we are looking to talk about the real world. What should policymakers in the real world think about and keep at the top of their mind as they go ahead preparing policies considering AI? Before I start, just setting the stage.

To a man, to a woman, everybody I spoke with who’s attended the first AI summit to today, this is, I think, the fourth AI summit being held. The first one was held in the UK. And without exception, all of them made it a point to tell me how the first session was full of fear. It was, oh, my God, AI is this terrible technology which is going to steal all our jobs, make us redundant. And when they come to India, they see the hope that technology and AI brings. And that’s the spirit of the discussion this afternoon, to figure out how can we balance both of those extremes, hope and concern, and go ahead in a pragmatic, policy -first way to prepare for the real world.

So if I could start with you, Ufuk, how do you think about AI? And especially, where do you see areas of creative destruction? To foster the innovation that we need.

Thank you very much. And so, of course, creative destruction is an important driver of economic growth in the long run. So that’s why, you know, it’s an interesting question how AI will affect creative destruction in general. Of course, we are at a very early phase of AI, and it’s a GPT. And typically, you know, when GPTs are emerging, there’s a huge surge of new businesses. And this should not be misleading. I think the main question we should be asking ourselves is what will happen to the creative destruction in the future? How does the future look like in terms of creative destruction? And I’m a macroeconomist, so that’s why I like to look at this with a, you know, bird’s eye view.

And I would like to, you know, separate advanced economies from emerging or developing economies. So when it comes to advanced economies, there, again, we need to split the issue into two layers. One, the foundational layer. and the other one is the application layer. When we look at the application layer, it’s great. You know, the entry barriers are low. Small businesses can do what only large businesses could do in the past, and, you know, they can do their accounting, marketing. You know, there are so many opportunities now. The entry barrier is low. As a result, this suggests that, you know, this is going to be more, you know, friendly for creative destruction on the application. But then there’s also the foundation layer, and I think that’s exactly where the bottleneck is.

When we look at the foundation layer, the entry barrier is really, really high, and, you know, the compute is very compute -heavy. It’s very data -heavy. It’s very talent -heavy. So as a result, you know, this market, at least this layer, is very concentration -prone. Of course, it’s very early. But, you know, normally we have to be concerned about the foundational layer and how things will pan out because this is the upstream to the application layer, which is downstream to foundation layer. So that’s why whatever will happen at the foundational layer will potentially spill over to application layer two. So that’s why I think we need to look at early indicators. But, you know, in the interest of time, I don’t want to go into the empirical evidence yet.

Maybe we can come back in the second layer. When we look at the developing countries, so I think, you know, I agree with Johannes. You know, I think AI is creating fantastic opportunities. So that’s why I think it’s really important to understand the opportunities as well as the risks for developing countries. And together with the World Bank, we are working on the world development. Report 2026, which is going to be on AI and development. And these are exactly the issues that we are focusing on. But I think before we go into those details, we should ask ourselves one major question. Why was there no entrepreneurship and dynamism before the AI revolution in emerging economies? Why was, you know, when we looked at the firm’s life cycle, for instance, why was it not up or out?

Why was it not, you know, very competition friendly? Why did the best predictor of firm size in emerging economies or developing economies was the size of the family and or the number of male children? These are still lingering issues and AI is not, you know, will not bring magic unless we understand and fix the business environment in these economies. You know, AI will just create new tools. But at the end of the day, we need to make sure that the business friendly environment is there for entrepreneurs to come and exercise their ideas

Ufuk, that’s a very interesting leaping of point, the real world. And the intention of this panel is to get exactly there. So if I may turn to you, quite literally turn to you, Michael, and ask you about the real world. You’re obviously doing a lot of work on the ground. Where do you see the potential for AI to spur gains? And are there any really transformative breakthrough areas that you’re looking at right now?

Yes. Thank you. Thanks very much. You know, I don’t want to minimize the existence of forces that may widen gaps. I think that if policymakers, primarily at the national level, but also in multilateral development banks, take appropriate actions and make appropriate investments, then I think AI has the potential to substantially narrow some of the gaps. And, you know, I think the… which policy actions to take can be informed by thinking through relevant market failures and relevant government failures. Let me give a concrete example or two. So private firms have incentives to develop and improve applications of AI that can generate profits. But there are some very important applications of AI for public goods, for example, that will not attract commercial investment to measure it with their needs.

And that’s an area where I think governments and multilateral development banks can play an important role. And I think some of this very much echoes what you were saying about small models, but also I’ll mention the link between the two. So an obvious example where I think India has been a leader for the world is in the development of digital identity. You know, this is… will enable, as Ufuk was saying, this enables a lot of work by individual entrepreneurs, a lot of other applications. So that’s a huge success, and I think multilateral development banks together with India can help bring that to many other countries. Let me take another example, one that’s not as well -known, but picks up on your comment about farmers.

So one thing that’s critical for farmers, they have to make a bunch of decisions that are weather -dependent. You know, when do you plant, for example? What varieties do you use? A drought -resistant variety, another variety. That, most farmers don’t have access to state -of -the -art weather forecasts around the world. I’m not talking about one country. In low – and middle -income countries, they don’t have access to that. Now, there’s a huge advance. We tend to think of large language models, but obviously AI is pushing science forward, and that includes in weather forecasting. There’s really a revolution driven by AI. But weather forecasts are non -rival. They’re largely non -excludable. They’re the classic definition of a public good.

So there’s a strong rationale for national governments, in some cases supported by multilateral development banks, to make investments in producing and disseminating AI weather forecasts. Again here, India is a leader. So if you, India in particular, in particular, India’s, the Indian government distributed forecasts to AI weather forecasts to 38 million farmers last year. And the evidence suggests that farmers, both from India, from this particular case, that in areas, I’ll say a little bit about last year’s monsoon, it came early in Kerala and southern India, but then there was an unexpected delay in the progression. The AI forecasts got that right, that was the only source of information that reached farmers with that. In the areas, we did a survey above that line, and farmers are responding, and they transplant more, they use hybrid seeds more.

Evidence from around the world is consistent with this. Farmers respond to these AI weather forecasts. So I think that’s one example, but many others, and happy to discuss them in education and traffic enforcement and elsewhere.

Michael, your answer should be read the book. Okay. We’ve spoken about the use cases of India, but setting up digital IDs, of course, is a sovereign decision. It’s something India could do unilaterally. When it comes to the large language models, that’s not reality. The large language models are concentrated in the US, in China now with DeepSeek. Anu, in a world where you largely have the rules being set by the two large powers, the US and China, arguably, there’s of course the EU as well, and you’ve done a lot of work on that. Who sets the AI rules for the Global South? Is there even the possibility for the Global South to talk about sovereignty?

So I think the Global South has the same kind of incentive for their own AI sovereignty, including then regulatory sovereignty, to design the rules that better work for their economies, for their societies, for what the public interest in these jurisdictions calls for. But regulating AI is really difficult even for very established bureaucracies. You need to be able to make sure that it is an innovation -friendly, and yet you at the same time need to be careful in managing the risks for individuals and societies. So even very established regulators like the European Union have found it one of the most challenging tasks to come up with the AI Act. So there’s probably something to be learned from these jurisdictions that have gone ahead and done the kind of thinking that had then resulted into some of those regulatory frameworks that we have now in place.

So if you think about the choices that India has when it looks around, one of them is to think about, okay, how does the EU go about this? The EU follows what I would call a rights -driven approach to regulation. So what is really characterizing this, the first horizontal binding, so economy -wide regulation that the Europeans enacted, it is a regulation that seeks to protect the fundamental rights of individuals, the democratic structures of the society, and that also seeks to ensure a greater distribution of the benefits from AI revolution. So the European approach is very conscious that it wants to also share some of the benefits so they don’t all go to the large developers of these models, but individual use as society at large.

smaller companies benefit from AI as well. So there’s something I think the Europeans can teach in terms of that regulatory approach in addition to maybe then some details of how that regulation in the end was constructed. But just one word, India is a formidable economy that doesn’t need to take a template and plug it into the economy as such. I think India is in a very good position to take the lessons that serves its needs yet make the kind of local modification and variations that are more reflecting the distinct priorities of this country.

Anu, before I turn to Iqbal, a quick follow -up question to you. As India makes its own rules, where does the trade -off lie between regulation and innovation?

So this is very interesting because often I am based in the U.S., but I’m initially from Europe, and these two jurisdictions are described as the U.S. develops technologies and the Europeans regulate those technologies. many ways does India want the innovation path or the regulation path? And I think there are many votes who would go for innovation. But I really would like to debunk this myth that to me it’s a false choice to say that the reason we don’t see these large language models being developed in Europe is not because there’s a GDPR, the General Data Protection Regulation. It’s not because there is AI Act. So the reason there is a perceived innovation gap between the United States and Europe is, I think, four things.

So first, there is no digital single market in Europe. It’s very hard for these AI companies to scale across 27 distinct markets. Second, there’s no deep, robust capital markets union. 5 % of the global venture capital is in Europe, over 50 % in the United States. That explains why the U.S. has been able to take much greater steps in developing AI technologies. Third, there are legal frameworks and cultural attitudes to risk -taking. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone.

You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. You are not alone. I wouldn’t encourage you to replicate that because it’s very hard to innovate on the frontier of technological innovation because sometimes you fail. But you need to be then given the second chance.

And the fourth, I think, the sort of foundational pillar of the robust U.S. tech ecosystem is that the U.S. has been spectacularly successful in harnessing the global talent that has chosen to come to the U.S., including many Indian data scientists, engineers, who think that U.S. is the place where they can start their companies, scale their companies, fund their companies, U.S. universities can attract them. So the idea that choosing to follow… Or imitate aspects of the European rights protective regulation would come at the cost of innovation, we need to understand better what drives the technological innovation and whether regulation should

Thank you, Anu. Iqbal, turning to you. You’re working in an area of the world, South Asia, where what is regulation? What is enforcement? At the risk of sounding like a provocateur, it’s the Wild West a little bit. And therefore, we talk a lot in our part of the world about small AI, about targeted AI. My question to you is that what should policymakers keep in mind when designing AI -enabled interventions, especially when it comes to small AI and the targeted use cases?

vulnerable public schools all the way from 11th to becoming the second best performing state in just a matter of two or three years. Phenomenal results, right? But then you start saying, let’s unpack this. What was this thing doing? The first thing that they find out is that a lot of people are like, oh, does this mean that I don’t need teachers anymore? No, you still need the teachers. What it replaces is the road task of the teacher having to correct spelling mistakes, calling you to the room and saying, hey, you forgot your comma, you forgot to capitalize. Instead, AI takes care of all of that. And now the teacher can sit with you in the free time and say, how did you set up the structure of this essay?

Did you think about this analytically or not? And that’s the first insight that comes from evaluation. It frees up the teacher time. Everything that we do in the field ends up adding to teacher’s time, adding to the nurse’s time, adding to the Anganwadi worker’s time. Very few teachers do that. Free up time. So if your AI application can free up the time of the health frontline workers, first of all, that’s a winner. The second thing that is really important here was that this is a demand -driven thing, right? Like there was a demand by the kids to improve their essays. There was a demand by the teachers to free up their time. But most importantly, there was a demand by the school districts to show progress.

So I think those is kind of a great example of how everything comes together if you think about it ahead of time.

Ladies and gentlemen, a topper of India’s notoriously difficult civil services exam. So take Iqbal more seriously than you would as just a normal.

Thank you. I thought that was history now.

It’s never history in India, Iqbal. Michael, turning to you, almost as equal in accomplishment by winning a Nobel. What risks should multilaterals like the World Bank keep in mind? Or let me rephrase that actually. Is there a risk that multilaterals are moving too slowly relative to the technology?

I think there certainly is. As I noted before, there are certain areas where the private sector is going to move, but there are other areas where they’re not going to move quickly, and it’s going to be very important for governments and for multilateral development banks and for philanthropy to move. I think there are a number of approaches to this. One way is by encouraging innovation by setting up institutions like innovation funds, particularly evidence -based, to echo Iqbal, I think evidence -based innovation funds. So I’ll give you one example of something that I’m involved in. Development Innovation Ventures, that was initially set up in the U.S. government, but it’s now been relaunched independently. It has tiered funding, so there’s initially very small… grants to pilot new ideas.

Then there’s somewhat larger grants to rigorously test them as Iqbal emphasized and then for those that are most successful there’s funds to help transition them to scale up. I think why is that important? Well that’s important because if we’re thinking about the services that public services and there are other sectors where this is needed but there’s probably going to be insufficient competition. Private developers are going to come up with innovations but then there if they have to sell them to the government they’re facing a monopsonistic buyer. They’re not going to probably not going to get rich doing that. So some support to generate more in that market, generate more entrance in that market, well I think is very important.

It’ll also mean that prices will go down and quality will go up when the government does that thing. Does that. Let me, I’ll just again let me give a example of the potential of how you know we we tend to focus on certain examples time after time here let me give another another example that is you know something that I doubt many people here are thinking of when they think of AI you know one of the things that you know traffic safety and we’ve all been exposed to traffic in the past few days you know traffic is a real problem interfering with urbanization which may drive growth there are a lot of deaths from from traffic a lot of citizens around the world have very difficult and painful experiences with traffic enforcement well you know you can have automated traffic cameras that have the opportunity to improve improve traffic outcomes but also improve people’s perception of fairness in government India’s moving in this let me mention another thing that within traffic safety that’s being done Microsoft Research India developed a program called the India Research Program and it’s a program that’s been developed by the government and it’s a program called HAB that is for driver’s licenses and that it automatically uses AI to test are that are the drivers until they actually pass in their exams they when this was introduced it’s been introduced I believe in 56 sites across India hundreds of thousands of people have taken tests this way we took a leaf from a false book we followed up the we’ve got information from Ola on ratings on and the number of drivers who were rated as driving unsafely that went down 20 to 30 percent where hams had been installed so you know that’s something that was developed not by Microsoft’s main business but by Microsoft research we can just create some support for more ideas like that to be developed to be rigorously tested that can benefit India can benefit the whole world we are we are running out of time probably this is this is one place in in India where time is really respected and we have to end in time.

So I had a list of wonderful questions, but if I could now move to a space where we are really giving shorter answers and quick answers and the deeply, deeply interesting ones about who’s winning and who’s losing. Michael, if I could start with you, actually. We’ve seen many promising technologies fail to live up to their promise. How should we think when we are evaluating AI interventions? How should we think about it? What should be the metrics that we use? Okay. First, model evaluation. So AI companies typically do that part. How good is the model output for specific tasks? You know, forecasting the weather. Does it do a good job? Does it match your local language well?

Second, user impact. Here, I think there’s a role both for sort of initial pilots akin to a medical efficacy trial. If you put the work into trying it, does it lead to improvements and outcomes for the users? Second… scalability and usage at scale that’s more like an effectiveness trial in medicine that it’s important to think not just about the tech but also about the human systems are the teachers actually going to use the product I think is it is an example how can you get the teachers to use the product and then the fourth area is continuous improvement you want a system that improves the underlying models so I think in procurement we might want to think about requiring continuous a B test publicity about what the what the impact usages and impact is and perhaps even thinking about requiring open access as part of the procurement package

thank you Michael. Iqbal, I want to flip that question to you where do you see where do you see hype in the promises of AI that you don’t think will play out

I think hype is natural because the technology is exciting. It’s a general -purpose technology. It’s evolving so quickly. The marginal cost of deployment for the next users is very low. It’s multimodal. Today you are doing it in text. Tomorrow you’re doing it in video. Day after tomorrow you’re doing it on audio. Everybody who has a smartphone has it. So I can understand the hype, right, like where it is coming from. But I think what we really need to do is separate the hype from the reality on the ground. And the reality on the ground is that many of these technologies are not having the final impact that we are having. And I see kind of two, you know, like once again my job at J -PAL always, you know, sitting at the top is like to say not worry about one professor’s evaluation or one researcher’s evaluation, but say when I connect all these dots, what am I seeing?

And I’m seeing two patterns. One is about trust in technology, and the second part is about the reality of the policy world. Let me elaborate quickly on both. Trust in technology. There are studies which found that even if you give doctors and frontline health care workers access to diagnostic tools, including radiology, tools, using AI, AI enabled prediction of the diseases, oftentimes it doesn’t lead to an improvement in results. And when you try and unpack that, even though this technology worked even better than the human intervention in the lab, right? So some of these diagnostic things can work, have better predictability in the lab, but in the field, they end up decreasing, not only is their efficiency lower, but it lowers the efficiency of the doctors, because we have not trained them enough important.

And the second thing is the enabling mechanism, the world around us. We just assume that just because the technology works, even if it works in the field, the rest of the system will adapt to it. No, you have to adapt the system to the rest of the world. So this example quickly comes from India, where, you know, we have a with one particular state government, we try to improve the collection of value added taxes, it’s called GST in India, there is a whole worry about bogus firms that are created to get these GST or value added tax thing. The machine learning algorithm is able to increase the probability of predicting a bogus firm from 38 % to 55 % in one shot at a very, very low cost.

When it came time to scale up this program by the government, they refused to scale it up because you think about it, you have taken away the discretion of the human to decide whether they should raid Michael’s firm or they should raid Iqbal’s firm. That is power. And if you haven’t thought through that point, what is the point of technology?

I won’t terrify anyone in the room by asking why they didn’t want to scale up this tech. But talking about weeding out the bad actors, talking about firm -level decisions, moving on to UFOOC, does the firm -level evidence show productivity gains diffusing evenly across?

So just going back quickly to the question of the firm. In the earlier model that I highlighted, I think it’s important to understand what’s happening at the upstream. so that we can then understand where things will be going in the future. And the evidence there, the early signs, is a bit worrying. So first of all, when we look at, for instance, the dynamism or market concentration in the U.S., market concentration has been increasing since 1980 but in an accelerating way after 2000. So that’s the first set of evidence. The second set of evidence comes from how innovative resources are allocated across firms. And when we look at the inventors who are creating the creative destruction and technologies, there’s a massive shift towards market incumbents.

And when I say incumbents, those firms that have more than 1 ,000 employees. In around 2000, 50 % of employees used to work for incumbent firms in just 10 years. That shifted. To more than 60%. A massive reallocation of innovative resources. And the final piece of evidence, and we are going to release this study next week, we looked at the universities, how AI is impacting universities, and we look at the AI publishing scientists. And AI publishing scientists in academia, the top 1%, used to make around $300 ,000 in 2000. It went up to $390 ,000 over two decades. Similar people in industry used to make around $550 ,000. Now it went up to $2 million.

And there has been two breakpoints. One of them was in 2012. The other one was in 2017. Of course, image processing and then the foundational model revolution in 2017. The more worrying part about this, which brings me back to the foundational model side of things, is that this created a massive out -migration from academia to industry.

And after 2017 especially, B2B. When the compute and infrastructure became so important. And then we saw the rise of AI. The target or the destination is large incumbent information companies, which again highlights where things are going in terms of the concentration. And the worrying part also is that when people are moving to industry from academia, their publication record goes down by 50%. They start patenting by 600 % more after they move, which means that we are moving from open science to more protected science. Now, spillover is extremely important for creative destruction, for the future of innovation. So that’s why, and if we will keep the foundational layer contestable, I think that the fundamental players there will be universities.

And keeping universities in a healthy way is extremely important, but there is very little discussion on this, which I think before it gets too late. Because once you start buttoning the wrong button, and then the rest will follow wrong as well. So that’s why I think we have to have this frank conversation early on in the game, otherwise it might… too late.

Ufuk, what you spoke about boils down to something Iqbal mentioned as well, power. Because power still makes decisions in this world today. So Anu, before I move to the final section of this panel, if I could ask you if the finance minister of a developing country let’s say India, comes to you and asks you, Anu, how should I think? What would you tell her?

So today if you think about how much political power but also geopolitical power is shaping our conversations around AI it is something where I think each country is now pushed towards greater techno -nationalism, techno -protectionism AI sovereignty has become almost a sort of uniformly goal for everyone. But I would remind even when encountering players like the United States and China that nobody in today’s world will be completely sovereign when it comes to AI space. If I just take one layer of the AI stack as an example. What is now driving a lot of the global AI race is this idea that we want to do frontier AI we want to have these powerful foundation models.

That means you need to have a lot of computers. You can’t have a lot of compute unless you have access to the high -end semiconductors. The U.S. is well positioned there. It is hosting companies like NVIDIA. The U.S. leads in the design of semiconductors. But who is manufacturing them? We really need to think about the role of Taiwan there. But then the Europeans have ASML in the Netherlands that leads in the high -end manufacturing with the equipment needed for manufacturing. But that is dependent on chemicals where Japan is leading. And the entire supply chain relies on raw materials from China. So ultimately, all these choke points can in principle be weaponized, but that is not ultimately a sustainable strategy.

Even President Trump had to walk back some of the export controls to China because Chinese were saying, okay, then the raw materials are not coming your way. So there are the potential ways to weaponize these interdependencies that ultimately make us all poorer. So as a finance minister of India, when approaching other middle powers, the great powers,

Easily said than done. Our final, final section is, of course, the rapid fire round. We all love this in this room. In one sentence, in one sentence, if I could ask all of you, and Johannes, you’re not getting away easily, you’re going to answer this as well. So in one. if I could ask you, we’re sitting in New Delhi 2035. Could you predict one development outcome that will have dramatically improved with the use of AI and one risk we’ll regret not addressing now? I guess you already know my second answer.

I think the concentration, the future of market concentration is something that we should be concerned about and we might regret not having discussed this sufficiently in 10 years. On what will change in a positive direction, clearly health care and education, I think. It’s a no -brainer.

Anu?

So first of all, it’s so inspiring to hear all the use case examples, whether we talk about traffic or agriculture or education, because I often talk about the risks and the downsides, so it’s a really good reminder. I’m personally very excited, especially what happens in the education space but also in the health space. In terms of the risks, I think one thing that we are not paying attention to, and what I would even call a systemic risk, is the idea that many worry about AI getting almost too smart. But I am more worried about us getting dumber as a humanity. There is a temptation to start skipping steps, outsourcing your thinking and your creativity to these models.

And as an educator, when I think about how I will teach my students to use generative AI to enhance but not substitute their capabilities, we will just make a tremendous mistake if we just forewent that hard work, that beautiful moment of thinking hard problems and creating and investing in our own capabilities. And all that just cannot be so outsourced, because otherwise we don’t even know what kind of questions we should be asking the AI going forward.

Michael.

I agree that there is huge potential in health. and education. I think we’ll see big improvements in that, but the risk is that the public sector won’t adopt these, and therefore the poor won’t have access to them. And that’s because the public sector, as Iqbal indicated, the government systems and the government workers may not adapt to use these. There’s also risks of copycat regulation that are over -focused on certain problems that other countries may be worrying about, but might not be relevant for emerging economies. And then final risk is that the procurement systems are just set up in such a way that we don’t get sufficient competition, we get lock -in, and then we just don’t wind up with good quality.

Thank you, Michael. The buzzer’s down, but I’ll take a risk and quickly run through the other.

Yes. I think I am much more optimistic about the government actually adopting this thing. Whether it is when you call 100, your call is going to get answered very quickly. The PCR van is going to be at your house much faster. The hospitals are able to be able to link your health record. So I think the government sector productivity is going to improve leapfrogs. The biggest risk, I think, is definitely the labor market. If there was a dial where I could slow down the adaptation and give time to the labor market to catch up, that’s my biggest worry. You talked about entry -level jobs. An entry -level coding job might be an entry -level job in the United States.

It’s the aspirational job that created Gurgaon’s and Noida’s and Mohali’s of this country. And those people are going to be running out of jobs very, quickly. And I think the labor market, whether it is ESI, Provident Fund, Gratuity, we are piling on and making it harder and harder to hire labor. when, on the other hand, capital is not taxed. We are giving incentives to people to use AI, and we are taxing them through provident fund and labor market regulations to hire labor. And I think that, for me, is the biggest risk, actually.

So I think that for the first time in human history, we may actually have the tools available to enable us to target poverty reduction, poverty elimination initiatives on individuals. And that could be tremendously transforming. But at the same time, I do worry that we will not get the governance right or we won’t be able to make that governance sufficiently robust to prevent abuses.

Thank you very much to all of our panelists and to you for your time and attention once again. I had the very rare fortune of being able to peek into Michael’s screen while he was speaking, and I saw all the messy human notes. Our panelists are definitely not outsourcing their thinking anytime soon, and thank God for that. Thank you, ladies and gentlemen

sci -fi movies that we grew up watching and what it primarily also reminds me of is in specific terms the avengers right the avengers are the superheroes and they’re trying to you know save the world and decide how one can do that and they all have very different strengths so i was wondering that if all our panelists were superheroes who would they be introducing our panelists i have our first avenger captain america principled steady under pressure obsessed with doing the right thing even when it’s unpopular professor seth is exactly that and reminds me of the lens that he brings in he studies how societies hold together how coordination succeeds or fails and why systems need shared values as much as intelligence next we have spider -man spider -man strength isn’t brute force it’s his ability to navigate through complex webs adapt quickly, and see connections that others miss.

Professor Nirav thinks the same way. At the University of Bristol, his work focuses on multi -Asian systems because societies like Spider -Man are all about networks. Andhra Vasudev reminds me of Captain Marvel, operating at scale, moving across institutions, pushing boundaries. Through her NGO service, she uses AI to amplify citizen voices and reshape how power flows between governments and people. And of course we have Iron Man, Iron Man who is obsessed with execution, iteration, and making ideas work in the real world. Mr. Bal is our Iron Man, focused on execution, scale, and impact in the real economy. He leads the McKinsey Digital and McKinsey Analytics practices in India. Last but not the least, no team is complete without Bruce Banner.

Deeply aware of the challenges that we face, of AI’s raw power and focused on how to control it before it controls us, Professor Mantunath’s work reminds us that intelligence at scale can cause damage if we don’t fully understand its consequences. My name is Janhavi and today I’m embodying Jarvis, except for being the one answering the questions, I’m the voice asking them. Every Avengers story has a Thanos. The real question is whether AI becomes our ally or the great snap that we didn’t see coming. So when we talk about AI for collective good, we’re not just talking about smarter apps, we’re talking about systems that influence how people live, work and participate in society. Before we start, I would request all my panelists to just stand up for a quick photo op.

So, quick show of hands from the audience. How many of you feel that technology today is only with those who have power or resources or information, that technology has been reserved for the elite few? Do we have a show of hands in the house by any chance? Okay, clearly we don’t really have an opinion as such over here. But moving on, Professor Seth, when we look at society, you know, governments, markets or online platforms, we often assume that problems exist because we don’t have enough intelligence or data. Your work suggests something a little bit deeper, that perhaps failures come from how decisions interact at scale. From a systems perspective, Do you think our biggest societal problems are intelligence problems

Thanks a lot. So it’s great to be here in India. I think this topic is extremely relevant to both the UK where I’m working and India. And I think the answer is that coordination is intelligence in this situation that we’re interested in. So I guess we’re used to situations now where we interact with an AI as an individual. One person asks the AI a question and gets one answer. But really there’s the potential for us to develop AI systems that are designed to support a whole population at once. So a population of people that are affected by a flood, a population of people that all are coping with the same. disease or medical condition, a population of people that are all trying to get taxes to and from a summit.

So instead of AI answering individual questions, AI can help coordinate those people, share intelligence, share their knowledge, and achieve better outcomes. And I think that’s quite a different way of framing AI than many of the systems that we’re hearing about and requires different technologies and different ways of delivering that to people, different ways of engaging with populations. So I think that’s something that can only really be achieved by partnerships between researchers and companies and not -for -profit organizations and governments and requires probably interventions in the way that we promote AI rather than letting the sort of path of least resistance develop AI commercial tools. I think there are opportunities to really engage with the idea of making AI for populations.

Wonderful. Professor Nirav, you’re also from the University of Bristol and your work focuses on multi -agent systems where basically intelligence emerges from all these entities interacting with one another. What kind of social problems are best suited for these multi -agent approaches?

Thanks, Chandni. Good question. And I think partly Seth already answered what multi -agents could do. So all problems that we’re thinking over here are in terms of, if you’re understanding those problems, they are socio -technical in nature. So there are social entities including people, organizations which interact. All of us also use some technical tools. These could be intelligent agents. These could be applications, softwares that we use. And all of these combined together, help us. So all problems… include or all domains are socio -technical in nature. Multi -agent inherently can encapsulate socio -technical systems. So that is how I would look at it. If you’re talking about, say, ride sharing, for instance, or hailing a ride, current system could be optimizing only for me, right?

And then what we end up with, we could end up with local maxima. So if we are optimizing for each one of us, we are doing a local optimal for each of us. But we may not be doing a global optimal. And global optimal would map to social welfare. What does social welfare mean? Does it mean just maximizing experience for everybody? Or are we meaning satisfactory experience? So I think any problem that we think about, say, epidemic, pandemic prevention, making sure. That is. are located properly, all of that would be multi -agent in nature.

Interesting. Professor Seth, do you have anything that you’d like to add on to that?

So, yeah, I think we’ve heard, I think, a little bit from some AI leaders about a next wave of AI that will be agentic, where we won’t be just interacting with ChatGPT as a monolith. We will be interacting with an agent that has purposive aims and is helping us to achieve tasks. And it might do that by communicating with other agents. Whenever we interact with AI, we would be, in fact, interacting with a population of AIs that are sending each other information, that are tasking each other with different jobs to do. And actually, it might not be clear whether one of those agents is artificial or a person. And so, if we enter into that sort of world, I think we have to really understand whether those agents are going to be able to do that.

I think those agents are interacting with each other. in a way that is likely to advantage the community of users because the amount of resources that will be consumed by these population of agents and the potential for them to interact in ways that have unforeseen consequences for other people are going to ramify. When we do that manually, really we can only hold so many interactions with other people at once, and so we’re limited in the scale. You know, one request does not create this kind of cascade of other requests in the system. But as we move to artificial systems, that scaling will rapidly increase and potentially one trivial request by me asking a computer to make a picture of a dog riding a skateboard could create a whole kind of wave of different agentic interactions that consume loads of resource and also, depending on what I’ve asked for, disadvantage other people.

So embedding some kind of social responsibility… into those agents, some appreciation for how their behavior impacts other agents in the system, I think is going to be imperative. Otherwise, we end up with systems that create conflict and contestation for resources.

Interesting. Whenever I’m on Instagram or Facebook, and let’s say if I’m talking to my friends, I’m really thinking about buying this Dyson or a particular product, it’s always weird to me how the next time I open the app, it’s almost like the app has heard me, and I start seeing the ads for those exact things, even if I’ve not searched it. I’ve just talked about it to someone. Has anybody here also experienced the same thing, a show of hands quickly, where you feel that maybe the choices that we make, are they really our choices, or are we being nudged by algo somewhere? So, Professor Manjana, your work focuses so much on recommendation systems, and we often hear that these algos are just tools.

Perhaps your research suggests that they actively shape what people see, buy, believe. How much of human behavior today is genuinely chosen by us and how much is subtly nudged by these algorithms?

Yeah, recommendation systems and the way they shape many of our feelings and our attitudes and our habits has essentially been a significant concern for me for a while. One of the things that you have to think about when you look at recommendation systems is that they’re essentially learning agents. So they want to learn your preferences, your likes, your dislikes, etc. And when they’re trying to do that learning, they do things. They’re trying to sort of give you options, different kinds of options, and then see how you react. So there is the first way in which the interaction between you and the learning system happens. The learning algorithm. So corresponding to our recommendations. So according to our recommendations.

system happens is this, they are showing you a variety of things and the way you react. And then your reaction is usually captured in some kind of utility function, something that the algorithm believes is positive for whoever is designing that algorithm. Now, what exactly is that utility function essentially determines what gets recommended to you in the future and what the system learns about you. Now, there is no such thing as the right utility function and every organization will figure out what they want for themselves. And if you just look at some of the, we have actually done several models, mathematical models on this and show that if I, depending on the kind of learning algorithm that I have and I am assuming a benign recommendation system here, depending on the kind of learning algorithm that I use.

Where I start off with a set of preferences. by the end of the day or over a certain time horizon, my preferences can be dramatically different. So there is a certain nudge that is steadily pushed by these algorithms and in which direction the nudge is pushed depends on the kind of algorithms they use and the kind of what we call utility functions that they use. So what exactly are they trying to optimize for themselves? And if you look at various analysis of many of these, especially Facebook algorithms, there is a very famous book that came out recently by somebody called Sarah Wynn Williams who was an insider. You can see the impact of what that had on some sections of some society elsewhere when the whole recommendation system went berserk.

So there is definitely a huge impact on the population’s preferences by the recommendation systems. And if you want to sort of give a quick understanding of that, recommendation systems essentially are advertisements. The difference between a standard… and this advertisement definitely shape our preferences. If you see something more often, you will start thinking about it and so on. The difference between, at least in my opinion, the difference between the advertising advertisements that you see on the street and the advertisement corresponding to a recommendation engine is that you are significantly more receptive. You are looking to do something. And when you are trying to look for something to do, if the recommendation pushes you in a certain direction, you are naturally going to go there.

So the impact of recommendation systems on the population’s preference, in my opinion, is spectacularly large.

Wow. That’s quite a lot to actually digest and hear from. I really wonder how much my personality is my own at this point. Antara, so from your work in civic engagement, when AI enters governance, is it to primarily help citizens be heard or is it helping governments manage complexity? And where do… Where do citizens struggle the most when technology becomes the interface between them and the government?

Thank you for that question. Just want to make sure that everyone can hear me. Thank you. Some problems like on -stage mics, AI cannot solve. No, but I just wanted to, of course, next year, correct. Thank you for that lovely question and lovely being here with all of you today. Jandi, to your point, I think AI currently is being used in both use cases. It’s allowing us to engage with citizens who perhaps have little or limited knowledge about law and policy and to be able to help them clarify doubts, for them to be able to air out their grievances, for them to actually be able to understand the frameworks of policy and law that govern their lives.

But in addition to that, it is also being used in a very large way for optimization. In a country of India’s size and diversity, I think the only other ways to perhaps not tackle circumstances, not an important governance does. So better than that is to actually build strong and robust frameworks for how governance can utilize AI, which is put out in a manner which is transparent, accessible, and one that actually has certain equity built in, which is really what the panel is also discussing today. And once you have that, to know that these optimization solutions can perhaps be built by AI rather than citizen -led. So at CIVIS, we’ve actually been working on gathering a lot more public feedback on draft laws and policies using AI.

And again, we see optimization in both ends, but very, very mindful of the fact that the frameworks that govern that level of optimization are what needs to be designed before perhaps even we race to the next model.

Got it. Can you share some examples of the kind of laws that have been impacted or the kind of work that you’ve done? Have you worked with different state governments? Governments where citizens of that particular state have been able to engage with the government about a certain law? or practice has been happening. Thank you.

Absolutely. So I’ll share one example from recent work with the government of Maharashtra that Civis led. The government of Maharashtra actually undertook a very ambitious mission of trying to understand how the next 22 years of the state can be governed by citizens’ voice. Now, this is something which is honestly quite remarkable on their part. What Civis was able to do is that we built out a very easy -to -use chatbot, wherein you could send in a voice note, you could send in any text messages, or you could even, we had people send in drawings, letters that they had personally written to the Chief Minister and other things. Civis aggregated all of that feedback. So that was almost 3 .8 lakh citizen responses from 37 districts across Maharashtra.

And that was aggregated, sorted through, and then shared with the government as well. The Vixit Maharashtra report, as it’s called, is now publicly available at the government of Maharashtra has put it. out on their own website as well. But in addition to that, what’s been really interesting about it is that they have said that every law that’s going to come out in the state for the next coming years has to, in some way, factor in what citizens are saying about that problem area or that district for where the law is being made. And you can only do that if you’re able to actually engage at scale. And I think that’s the beauty of what that entire project showed.

Absolutely. Professor, how do you feel about the government in terms of what approach should they be taking when it comes to AI and technology?

Yeah. One of the fears that I have when the government gets involved in technology development is that they want to start controlling the direction. They want to tell what to be done at a very micromanaging kind of level. And I recently had an article on Tuesday, I think it was in the Financial Express. There was an op -ed where we talked about me and a colleague of mine. We talked about, you know, looking, we looked at history. kind of successful and spectacularly unsuccessful involvements of the government when they wanted to direct technology. So I’ll just give you two quick examples. So in India, about 40 years ago, there was something called CDOT. So that developed some spectacular technology when it was left alone.

The government started to direct it and micromanage the flow of technology. Many of you probably don’t even know CDOT. They don’t even come to IIT Bombay campus, for example, for recruitment. That’s just one example. If you look at Japan, just to give you another badly successful story, many of you are too young to know about something called the fifth generation computing systems that they wanted to start off in. The AI boom that we see today was originally planned to be launched in Japan in the 1980s. There was a huge project that the government wanted to micromanage, develop native hardware for AI and everybody thought they would be successful. It was a spectacular failure. The failure essentially stemmed from the fact that the government was directing everything.

Governments are generalists. People who run governments are generalists. They are brilliant people. They know society. They understand administration. But they don’t understand technology. Especially a technology that is moving too damn fast has a very large surface area and they cannot control it. They cannot control that. So it is best that they just enable and let others, let the people on the ground, people with a track record and people who want to take risks manage them. They should be enablers. They should also be monitors. Monitors nudging it in a certain direction making sure bad things don’t happen. But that’s a very hard task. So the biggest role that the government should have is just enable and step away.

Just to give you one positive example the NPCI in India is a spectacular example of where the government started something and let the private sector and sort of technologies handle that. In the US many of you may be familiar with the internet. It was exactly that. It was just a vision that somebody had and said let’s build this. and the technology is built. That’s the way I would think the government should handle it, but we’ll have to see how that goes.

So just a quick question for the audience. You guys can shout the answers out loud. What emotions come to your mind when we think about AI? Are we feeling excitement? Are we feeling anxiety? Are we feeling FOMO? What are we feeling, guys? Curiosity. Dangerous, somebody said. What else? Definitely opportunity. Opportunity. The man over there? Confusion. Confusion. Anything else? Responsibility. Responsibility, fantastic. Great. So Mr. Bhai, this question is for you. There’s a lot of anxiety, a little bit of excitement as well about maybe AI replacing jobs, especially in India’s tech and services sector. From your experience working with different companies, where is AI genuinely replacing humans? and where is it actually creating new forms of value and roles?

Yeah, that’s a great question. Thank you. So I think the, let me try and give you the very brief answer, because I could talk about this for a long time. But there is a lot of focus on AI being used to replace humans in particular operations. So, you know, when you have an AI taking a call center call, that’s the simplest example of that. And what, and, you know, the math, the way it works is that, you know, if you’re spending 100 rupees on something, you can save 40 % of that roughly by replacing it with AI, with the current economics of the way it works. And obviously, if you’re in a high -cost geography, you can save more.

In a country like even in India, you can save that much. What we have found, though, is that most of the cases where you do this simple replacement of a human with AI, that’s not the case. cost reduction doesn’t really sustain. There’s a famous example of Klarna in Europe where they brought back a lot of the costs called center costs because they had to bring back some of the senior customer support people because a lot of the conversations were not going well and they were losing customer satisfaction. The same thing with IT, you can replace a lot of developers with this, but then people will come back with more projects and there’ll be more things to be done.

The real value unlock, which is sustaining, is actually when you get AI to do something which humans can’t do or are not able to do because it’s so time consuming and so difficult. For instance, a genuinely personalized customer engagement engine using the kind of recommendation system that he was talking about, which actually engages in a personalized way with every customer that I have as a company, for instance, or every entity that any organization is dealing with. That genuinely has value. It creates huge value unlock. So like for instance, I mean if I spend 2 -3 % of my revenue on say customer support and even if I save 40 % on that, I’m saving like 0 .8 % or 1%. But if I can generate even just 10 % more revenue from existing customers with hardly any marketing cost and I make 30 -40 % margin on that, I’m getting 3 -4 % more to the bottom line.

So that is a huge, it’s like almost 5x of what you can save. So the value unlock is very large and that’s sustainable because you’re really getting AI to do, no human being is going to sit and figure out exactly for millions of customers exactly what is the kind of personal message to send because the amount of experimentation you have to do and the kind of connections you have to draw between individuals and similarities and so on, which the recommendation engines are based on are impossible to do humanly. And that’s where the biggest value unlocks are at least that I’m seeing and those are sustainable and they’re actually even applicable in high cost geographies. It’s just that unfortunately a lot of the initial focus of the innovation has been on this just save, you know, do the easy stuff, right?

Have an AI agent replace a human agent. But that’s not the real power of where what AI can bring. So hopefully we’ll see a lot more of that type of innovation going forward as well.

Right. I think I see a lot of students here today. What kind of backgrounds do you all come from? Hands up if you’re from STEM at all? STEM backgrounds? Okay. Anybody from business, humanities, arts? Okay. So I read this LinkedIn post. I’m not sure whether it’s a great post or not. Apparently it’s going to be a little tough for STEM students to, you know, get into this world of AI because they could be replaced a lot easier. What kind of measures, businesses or like degrees does one, should one essentially come from to sustain in this world of AI, do you think? What should the next five years look like?

Yeah, I think there is some near -term potential impact on jobs and particularly on entry level coding jobs and so on. But honestly, there’s nothing which tells us that there is a, firstly, nobody knows exactly how the math is going to work. So between new work that people do for AI enabling versus the old work that may get more efficient because of AI enabled coding and so on, will we next see an increase or decrease of employment? Nobody actually knows. There are many, many forecasts and so on done by economists much more qualified than me. But what one can see is certainly that the enterprise adoption of AI has not really happened. So right now, the impact has not really happened of all of this.

So you’re seeing some initial hit on maybe, okay, this year I have promised I’m going to use AI and reduce my budget by a certain amount, so I’ll stop hiring. That’s the kind of, I would say, almost knee -jerk impact that you’re seeing right now. What eventually plays out will be… A mix of, okay, I will do the work more efficiently and use… a lot more automation, but now I have a lot more things to do as well. So I would say that students in general, actually forget just STEM, students in general need to be focusing a lot on how I can use AI to do the best possible thing that I can do in my field and in every possible field.

So whether I’m studying marketing or if I’m studying science degree or if I’m studying any form of the humanities, you know, there is a lot of journalism. If I’m, you know, whoever I am, right, there’s so many things that I can actually be doing with AI to do my, to do things which I was not humanly possible earlier. And that’s really what the students should be equipping themselves with. And then, you know, potentially innovating and also creating things, you know, around that, but also personally equipping themselves to actually leverage AI. And I think there are lots of examples of how that can play out and will serve people really well.

Absolutely. We are now going to get into a quick rapid fire round and then I want to open up the floor for audience questions. So the only rule here is I want short answers only. No explanations. We only have 10 seconds to answer. So I am going to start off by putting Antara on the spot. Does AI in governance shift power towards citizens or towards institutions today?

I want to say citizens because it allows for a lot more information asymmetry to be addressed which is where a lot of the power gaps come up today.

Professor Manjunath, are algorithms today more likely to reduce bias or hide bias better?

Hide bias. No, the options don’t look right to me. What would you put as the options then? The bias will start increasing. I think they are not trained. I don’t expect training to get better. I think it will be better in the immediate future, maybe much later. But I also want to disagree with what Antara said.

I’ll come back to you for that one. Professor Seth, what worries you more, AI being used with bad intent or AI being used widely without anyone fully understanding its consequences?

Well, they’re both terrible, aren’t they? I think people will always use technologies with bad intent and it can only really be addressed if a large number of people understand that technology and can then resist it. So I think the second is more important. Uplifting the public’s understanding of AI and kind of engagement with AI properly will protect us against malign uses of AI because we will be able to spot them.

Got it. Professor Nirav, what’s harder to design, ethical individuals or ethical systems?

I think that becomes tricky. Like what do we mean by ethical, right? So ethical individuals, if you’re combining ethical individuals and we say individuals combined together is a system, then ethical individuals.

Mr. Bal, in India, will AI mostly replace jobs, reshape jobs or polarize jobs?

Reshape.

That’s a very quick answer. You win the rapid fire answer. Right. Professor Saad, where does AI struggle more today, with people or with systems?

I mean, I think it struggles with people, but we don’t notice because it resembles the kind of natural language. When I say AI, I’m talking about something like chat GPT. So I think there’s a disguised problem with people there because… Those AIs, they don’t really mean what they say, they don’t really understand what they say, but it seems very strongly that they do. So I think that’s the problem. But what’s coming is AI embedded in all of our systems and then that will create its own set of problems as well.

Mr. Bal, who benefits more from AI today, companies or employees? AI mostly replace jobs, reshape jobs, or polarize jobs?

Reshape.

That’s a very quick answer. You win the rapid fire answer. Right. Professor Seth, where does AI struggle more today, with people or with systems?

I mean, I think it struggles with people, but we don’t notice because it resembles the kind of natural language. When I say AI, I’m talking about something like chat GPT. So I think there’s a disguised problem with people there because those AIs, they don’t really mean what they say. They don’t really understand what they say, but it seems very strongly that they do. So I think that’s the problem. But what’s coming is AI embedded in all of our systems, and then that will create its own set of problems as well.

Mr. Bahl, who benefits more from AI? Today, companies or employees?

I would say that right now, no one is benefiting from AI. But if I were to bet, it will be companies who will benefit first. And then employees will benefit. And the whole idea of having sessions like this is that we can get the employees to learn what we talked about, right? Students equipping themselves right from college. Absolutely.

Andra, for AI used in public systems, what matters more, transparency or effectiveness?

Transparency, off the bat. It’s the only way that we can actually design AI for public systems. It has to be at the front and center of all of our efforts.

Got it. Before we get into the last question for the entire panel, I do want to get your answer to Andra’s statement. If that’s fine. The question that I had asked. I’d ask, does AI in governance shift power to a citizen or to an institution?

Absolutely to the institutions. They have the money. to invest and discover what’s going on. There is no way citizens can beat that so easily. It requires a different whatever. I’m not allowed to say anything.

My last question for all the panelists before we open the floor for audience questions. If we get AI right, what is one everyday improvement people in this room would actually feel within the next five years?

So I think something that connects, there’s a thread that runs through this or there’s supposed to be and I think one thing that AI could give us is a greater sense that we are properly connected with each other and learning from each other. So the possibility for AI to break down barriers between people because of language and expertise and distance I think is huge. So the kind of traditional collective interaction intelligence that we’re used to where we put an X in a box when we vote for someone. It’s very, very simple, right? We can’t write an essay like the users of Antara’s system and send an essay to the government about what we want because there’s so many people, we can’t read all of those essays.

But AI can enable that kind of rich interactions. It’s an example of one of the things that Kush is talking about, that AI delivers something that is impossible for humans to do. It doesn’t just replace something that humans are already doing. So a future in which we all feel like we have a voice and AI is helping us mediate between each other, I think is something that is technically possible. There’s a whole bunch of political and social barriers to prevent that from happening. But I think five years is a timeline during which we could see the starts of those sorts of systems.

I can talk about what I’d like to see if we get AI right. We talk a lot about institutions, we talk about companies, we talk about individuals. But not enough talk happens specifically about small businesses. India is a country of self -employed people and small enterprise. I think there are about 150 million self -employed people. If each of those people could somehow earn 600 rupees more because of AI, and I’ll talk about how, that’s a unicorn. So 600 rupees more of allocation for each of these 150 million people is not, I mean there’s a lot of large numbers in India, but it’s true, right? It’s a unicorn. So I think we think of the next 50 unicorns. We may not think of like 50 companies worth a billion dollars, but we may think of 50 innovations that puts 600 rupees more in the pockets of 150 million people.

And how does one do that? I mean if you look at all the important things all of us use today, ride hailing, e -commerce, this restaurant ordering, food ordering, right? All of these created by… On institution, they make an app and then they do spend money on marketing and so on. Today, you have AI systems that are incredibly low cost. You know, 50 cab drivers can get organized. There’s an AI agent can do the scheduling and whatever. You have a WhatsApp chat with them and you can just find the driver, right? There’s no reason why we can’t have innovation like this. Very low cost. The price, the cost of the tokens can be funded in that ride.

It can be, right? That’s all that there is to run it. It’s an autonomous system which just runs off publicly available infrastructure. I think that, to me, is the real unlock that we can see. And those same systems can then serve anyone in the world. So you can do this for taxi drivers. You can do this for lawyers. And those lawyers can then serve anyone anywhere in the world. So I think that’s the real, real unlock that we are waiting for. These systems are very low cost to build. They can be built by anybody. They can be self -built by people. And it just takes a few groups, a group of a few of these self -employed people to get together.

And then, you know, suddenly this can go viral. So I would love to see that type of innovation coming. Rather than necessarily, you know, the stuff that we know for the companies we’ll do or the things that we’ll all play around with on our LLMs on ourselves.

Great. Antara?

Thank you. I think building on what Seth and Mr. Baral just said, there’s two things that I see happening. One is the disaggregation of systems and a lot of decentralized control mechanisms, right? When that happens, you have very fragmented channels to actually engage with institutions, to Seth’s point about building collective and new ways of collective intelligence. What I want to see happening for all of us in the room is greater access and connectivity to public institutions, which actually fuels us to get easier access to entitlements and benefits that the state is supposed to provide to us. If AI can get that right, if we can solve for that, I think there is a long and a big argument to be made about that being the sort of rising tide that lifts all boats.

Building on to what people have been talking and last on Antara’s point, thinking about collectives, right? So we can build systems which work for individuals, but how do we make sure that those individuals could be, like each individual have different preferences. How do we take into account different people’s preferences? How do we aggregate people’s preferences and then come up with a collective decision? If you are coming up with a collective decision, how does that decision affect various other people? How do we explain that decision to other people that, hey, we have taken into account your preferences in this particular way? So we need to get that part of AI right to make sure that people have a buy -in, people trust the system that we are designing.

So that is what I would want to see and I’m thinking that we are moving forward with that. We are thinking about fairness. We are thinking about, transparency. we are thinking about accountability and so on and so forth

yeah I can probably say what I already see the homeworks that my students submit are perfect the essays are spectacularly written the presentations are beautiful the only hope that I have is that they actually understand what they say so if that happens I will be very happy I think the output is perfect the understanding behind that output I hope will get better and better that’s my my wish for

I’m going to open up the floor for audience questions

my question is sir I want to understand what kind of impact AI will be having on management consultants and the business

I have no idea I have no idea really it’s very hard to say every industry is going to evolve Obviously, management consultants like everybody else are using AI for every possible thing that they can do with it. So they’re also trying to become more efficient, more productive with it. We don’t know what that means in terms of reshaping of the business. If you look at past tech innovations, which have also had a very big impact on productivity in many sectors, it’s not that entire sectors have disappeared or things have got, but things have got reshaped significantly. That has happened a lot. So I think the job that consultants do, like today when we do research, you don’t wait for one week for somebody to go and find things from everywhere.

It comes in a few minutes. Unfortunately, I find that a lot of the output, I have also seen a lot of the output, like Professor Manjunath said. I find two issues right now with the current versions of the AI. When it writes, it has no soul. So it’s correct, but it has no soul. And when it prepares a presentation or a piece of communication, it’s not inspiring. So it is correct, but it’s not inspiring. So I think there is a, so the consultants will spend more time on actually communicating in a way that’s inspiring while the desk, you know, the basic desk work will be done for you. So you spend time doing more, I would say, human tasks.

And that’s going to happen actually in a lot of other, in a lot of service jobs, right? You’re going to do, you’re going to spend time doing what humans are truly supposed to do and are really good at, which the AI models are not able to do.

Okay, thanks. So my question is for everyone. I have a younger cousin who is in high school and her entire life is on chat GPT at this point. So she shares everything, relationship issues, family issues, and it knows more about her than I do. And I kind of worry when I see the younger generation getting on these AI platforms. So what is your take on this, like, impact? What is the impact of this technology on young minds?

So. I share your concern I have slightly older kids I think we have to trust that we’ve been through these technological shifts before so my parents when they looked at me watching television had similar worries about they told me that my eyes would become square because I watched too much television so actually my generation became much more sophisticated consumers of television and were much more savvy about TV ads than my parents’ generation so I think we have to listen to our children about the way that they’re using these technologies they’re natives in this new world I’m calibrated for a world where AI doesn’t work where AI is not rolled out across the whole world so I’m the wrong person really to ask about how AI is going to change people we should ask young people how they’re using it and engage with them before they start to use their AI in a way that we don’t understand in secret

I have a funny answer and a short answer. But I think that one, I think the real danger actually is not with the chat GPTs of the world, but with the earlier addictive systems like the Instagrams of the world, right? Because they are genuinely playing on our brain’s dopamine circuits and are genuinely addictive and can therefore be harmful. I think with chat GPT, I think the only thing I would say is, I think it makes one actually question where we are as individuals, as parents, as family, that our children prefer to communicate to a relatively soulless communication device which answers everything like an American therapist textbook would, right? That they prefer to talk to that than to us.

It shows what a distance we have created. With each other, right? And that may be a good reminder to us as individuals around the task that we have to do in to rebuild bonds with each other.

I think on a very similar note actually to what Kusha just said, I think there have been studies from Youth Ki Awaaz and a number of other global youth -based organizations which have been looking at why exactly we turn to AI. And the phraseology is very interesting there because it indicates that turning to AI is something that you can also turn away from. I think the questions really come up where exactly what was just mentioned about understanding what are the kinds of tactile family bonds, what are the kinds of lived experience -based interactions that we can keep having with the younger generation to show that AI is a part of their life, but it’s not the only part of their life.

And I think that’s maybe my hypothesis on where we’re headed there.

I have a quick follow -up and you can connect with the previous question also. Many countries right now are trying to ban the new AI. Clearly there is evidence it is harmful in the course it’s coming. You mentioned Instagram or any other. AI is an amplifier. So unless we design, whether it’s regulation or whether it’s guardrails or whatever, what is our hope and what is the hope for a society not to get amplified harm than what they have already experienced, especially for the generation? Shall we start with you, sir?

Well, I think that’s basically what I wanted to say was to, the countries of Spain and Australia are two examples of where severe restrictions have been put on social media companies to at least give access to children. And that’s an interesting experiment. One has to see what’s going on. What will happen because it’s not an easy thing to do. I mean, I think technologically it’s not easy. Legally, I’m sure there are a lot of loopholes in all of this. We have to see how that evolves and potentially apply a similar. similar kind of guardrails with respect to AI. That’s the view, at least that’s the view that I have on that matter.

No, it has to start somewhere. I mean, this exactly goes to my point that I made earlier. Generalists in government cannot handle the space at which technology can move. You cannot put guardrails on that at the beginning. The moment you know something is happening, you have to get into the act as quickly as possible. Somebody is making an attempt. So let’s understand what’s going on. Maybe it’s, I mean, exactly what goes on is, what will happen is something that we have to see. I mean, what was interesting, at least in that attempt, was that the way in which the social media companies reacted to both the Australian and the Spanish ban. Okay, so to me, the most interesting part was they all said it was too fast, they’ve not thought about it.

things through. And then I remembered what Facebook’s slogan was, move fast and break things. They are allowed to move fast, but the legal system is not allowed to experiment. That seemed like an interesting contradiction for me to study.

Relatedly, so the first AI summit in London was very closed, right? Politicians and the leaders of big tech firms. And the idea that a couple of years later governments would actually be legislating in ways that limited in this case social media companies is very good news. After London, you could imagine that regulatory capture had happened, right? Governments were not going to be able to resist these big companies and their multinational power. So those first couple of steps of regulating social media for under -16s, even if it doesn’t quite work, even if it’s not exactly right, it at least is a step of introducing regulations and it will make AI companies… at least aware that that is a possibility.

Because they have to take that responsibility, I think.

Professor Nirav, do you have any other input on that as well?

I think I agree to the points that have been made. I think there could be different ways to think about a blanket ban, for instance. If you try to restrict something, people may not… They can have more curiosity in terms of why is it something which is getting banned. So we have to be thinking about that as well. But there is a step. There will have to be some regulations that should come into place. What those regulations would be, we need to be thinking about that. I think a lot of times the worry is people keep scrolling. And then the way the algorithms work, Professor Majunath knows better, but recommender systems would put you in a rabbit hole.

And you keep going into one direction. There could be echo chambers that could get informed. So the younger population is more vulnerable there, and that is where possibly a ban or restricted access helps. We have to be thinking about how can we, say in YouTube, there is YouTube Kids, and they only see kids’ content, but then there are malicious actors who would post some content which is targeted towards kids, but it is not actually kids’ content. There could be somebody could come up with a new social media platform for kids. I am not very sure what it would look like, but there would be new technology that would come, but that needs some guardrails to be put into place.

What kind of guardrails? Research and the legislation will have to be thinking about it.

Sure. I think we have time for one last question. Can we give it to somebody at the back? Yeah. The jean jacket. Yeah. Go for it. Can we pass the mic at the back, please?

So, definitely AI has enabled in the education and medical domain. But do we think that it has influenced, reached or violated the concept of the developers as well? There are singers who no longer exist. We are getting to hear those songs in the new generation. The ones who are alive, they definitely have a way to improve. But those who are not going to exist, it’s a breach of concept that, of course, it is falling under the domain of ethical AI. But just wanted to know your thoughts.

Is there someone that’s directed route for this question or is it open for all? Ethical. Okay, we can just, whoever would like to take that.

So, I think it’s a completely legitimate concern. Okay. And it’s difficult to understand where we go from here because the cat is already out of the bag, right? The models are already trained on everyone’s data without our consent. And how do we put that back in the box? I’m not sure that we can. I think, so there are currently legal cases that are going through the courts about the IP claims of musicians and artists, and it will be very interesting to see what law courts decide about that. I do think the kind of systems I’m interested in are systems that are built on consent. So a population of people that all have diabetes who sign up for an app that will track their disease, and then they gain by being part of a community where information is being shared to help people manage their diabetes.

So that’s a much more consenting model. It’s not about stealing people’s writing and art and music from the Internet, but that activity is already underway, and I don’t see a way of really putting it back in the box.

Let’s do one last question.

Yeah, I guess it’s not the… the topic of education and the internet is strong and all of those things. One thing that we have observed is that instant feedback even by AI tools in education especially, students do not go through the whole process, the step by step process of foundation. So if your, let’s say your courses work in a way or the tools work in a way that they are step by step trying to make learn the person, make learn the student instead of giving instant gratification with the output. So one thing, the question is like this that has any of the professors in the panel been approached for this kind of a thing for modeling of the education process or process of getting educated or learning especially.

And the other thing that would you, can we see a collaboration in that regard where we can try to create a regulatory thing for us or a guidelines that how AI tools should be constructed for imparting education in a step by step so that that is structured with gratification. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you.

Yeah, the short answer to honor his, whatever, never mind. I didn’t get that right. So, yeah, the short answer is no, nobody is thinking along those lines. And handling AI in a classroom has been quite painful. And to give you one example, there was an example in which I asked somebody to do something, essentially write a certain program to perform a certain task. I gave the data. The student, because the student went to chat GP to understand what the question was about, created her own data to do and did not know how to use the data that I was giving. So the point you are making is extremely valid. if you want to think about legislation or any other guardrails or anything like that, I’m up to discuss those with you offline.

Give a very brief answer today. More generally, I think every university is struggling with that question. And I’m hoping that there are lots of bright people and we will start to see some answers. But it’s not easy.

Well, a big thank you to all the panelists here. And a big thank you to all the audience members as well for being such great and engaging people. We have a token of appreciation from the University of Bristol side for all the panelists. From all of our sides. From somewhere. Thank you very much. Thank you. Thank you. Thank you. Thank you.

Good afternoon, everyone. On behalf of Software Technology Parks of India, I extend a very warm welcome to all the dignitaries on Dias and the entire audience to today’s session on Scaling Innovation, Building a Robust AI Startup Ecosystem. I am Shelly Sharma, Deputy Director, STPI, and it is my privilege to host this session.

Good afternoon, everyone. I am Vani Kapoor, Manager, STPI, your co -host for the session. May I now begin by respectfully welcoming our guests. Our distinguished dignitaries on the Dias. Our chief guest for today, Ms. Neerja Shekhar, IAS Director General, National Productivity Council. Sri Arvind Kumar sir, Director General, STPI Sri Rakesh Dubey sir, Director, Startup and Innovation, STPI Sri Bala MS, CEO, Strat Infinity and Ms. Geetika Dayal, Director General, Thai Delhi NCR and all other senior officials, ecosystem partners, startup founders and delegates present here today. We are truly honored by your presence. Today’s session brings together government, industry and the startup ecosystem to deliberate on building a future -ready AI innovation landscape while also celebrating startups that have demonstrated measurable impact across revenue, employment and business.

Government, innovation and inclusion. without further ado may I now invite Sri Rakesh Dubia sir director startup and innovation to kindly deliver the opening address sir please

incubators, accelerators, even state governments, academia, everyone can come to this platform and find the resources that they need here. This platform also serves as a repository of various government policies that come from time to time. It also serves as a platform where contests of not just STPI, but any incubator anywhere in India or even the world can host their contest, get their application invited, get the results published after screening and evaluation, and further handhold that startup’s entire life cycle online. This portal is, I think, one of its kind portal, not just in India, but across the world. It is a very valuable thing, and we are adding more and more features to it as time go.

For example, we have added features like a product marketplace as well as a hiring hub on it, using which a startup wanting a niche management, can post its requirement and individuals can apply against it. An individual wanting to look for a niche job can post his resume here and probably a startup can pick it up. It also has a feature called product marketplace in which any startup can post its product for anyone to see. And if any viewer finds interest in it, the two can interact together via this platform. That being said, the STPI is always looking for doing more and more things to support the innovation and startups across India as well as the world.

And we will be welcome to hear any thoughts from you. And there are many experts lined up. I am sure you will hear many more learnings from them also. With that, I thank you everyone and hope to see you. Thank you very much.

Thank you so much, sir, for setting the context so beautifully and highlighting STPI’s growing national impact. now may I request the technical team to play the short audio video presentation titled STPI Startup Ecosystem Drive Impact STPI Startup Thank you. Thank you. Thank you. Thank you, team. That gives us a powerful snapshot of how innovation is translating into real outcome across the country. Now to share insights from the industry and global capability center perspective, may I now invite Shree Bala, MSCO, Strat Infinity. Please come.

Very good afternoon. Namaste. Dr. DG, DIG, LPC, DG, hi. Good friend here, Rakesh. And everyone, very good afternoon. Thank you for the opportunity. Scaling innovation, building a robust AI, the perspective from GCC is going to… be phenomenal. We are not just living through the AI wave. We are rather living in the AI restructuring of global economy. If you look at the overall, if you look at the AI contribution, about 15 .7 trillion dollars by 2030 globally, whereas close to about 5 trillion dollars are going to be in productivity. That is an enormous opportunity. If you look at India by 2030, almost there is going to be 3 ,500 plus GCC, which is about 1 ,900 today now, and going to contribute about close to 150 billion dollar of software exports, and it is going to be 3 .5 million employees dedicatedly working for global capability center.

That is the very high level statistics for 2030. These are not just employment statistics, my dear friend, but this more like enterprise grade innovation, innovation, infrastructure at a national scale. Right. That said, if you look at AI leadership globally, most conversation comes on three things. Right. First thing is the model that we are building. Second thing comes the power, the compute power. The third thing may come under the funding perspective. Nothing wrong about it. All these three things are important. But in my experience working with a global organization, the scale is not determined by the AI you build. The scale is determined by the way how your AI gets integrated to the global organization. That’s where today the fundamental gap is all about.

And if you look at the real competitive advantage, in fact, the experimentation is really abundant today. But institutionalization is really limited. So that is where the real challenge and gap comes from. And the GCC component steps in this point. This is an inflection point in my view. If you look at the transformation. GCC has about from 1900 plus GCC are there in the country today and it was those days it was more looked at as a cost center or a labor arbitrage center but today it is more for the engineering centers, R &D centers if you look at more than 50 % of the or 40 % of the India’s GCC is more on R &D today the emerging technologies like AI, cyber security product development lot of new things have been developed out of the global organization now India is considered to be a digital talent center for the global organization which is not tactical but it’s truly very very strategical my dear friends.

If you look at the startup ecosystem that’s where the main thing comes from the venture capital, the investors STPI and lot of government organizations played a phenomenal role in making sure the grants are given which is very good very very important that accelerates the AI innovation, lot of fundings have come but capital alone cannot solve this friction that’s very very important one capital cannot solve the friction in fact when you look at global surveys in the recent AI study report that shows majority of the enterprises are piloting AI but only minority have scaled across the business units and the gap is not the technology problem it’s not a technology gap it is an operational readiness it is an organizational readiness that’s the biggest gap it’s not the technological capability so having said that if you look at any AI tools when you get into any organization the startups comes from across India you need to pass on through the risk you need to pass on through the compliance security the fitment the global workflow design lot of challenges that you come across again that is where the GCC component comes in because working with a global organization the enterprise organizations are totally different you can’t even think of one said we say India has lot of startups and we are phenomenally doing well at the same time the market access to the global organizations are really big question mark that’s where the GCC component comes in why the GCC matters for the startup it matters a lot here is the thing right so today what is needed for AI startup companies right you need a real data set you need a real infrastructure capability you need to have the enterprise validation so who is going to give that who is going to trust your model and put it in their system that’s the biggest question mark again that is where the GCC comes in where it is going to be the bridge between the startup ecosystem and the enterprise organization because you are going to work within the ecosystem within the infrastructure capabilities of the GCC that gives the confidence the enterprises to try you test you work with you and this is something important right today sorry yeah so this is something very important co -creation model so there were days where the startups were looked as vendors or suppliers but today the co -creation model is very very powerful in fact my personal experience from stat infinity for as you see 24 plus COEs are then STP we’ve been working with the fin blue we’ve been working with the ICOE where the global capability centers work with us through the STPI.

They’re able to identify a phenomenal startup and startup scaling for the global organization. So the co -creation model is the only model in our context how the startups, AI startups can get into the global capability centers because they provide the control sandbox, they provide a domain expertise, they provide a production grade environment gateways, pathways which basically helps to reduce the pilot to production cycle which globally remains, which is the basic bottleneck of any AI startup is that pilot to production. That can be solved by the co -creation model my dear friends. And coming back to that global why India is unique today, right? The economic multiplier effect. See, anything that comes into the GCC, the ecosystem grows, the value chain grows, the skill development happens, right?

And it generates a lot of revenue. Of course, the software exports increases. So a lot of possibilities, not just having one employee that indirectly helps many people to grow that. So that is where India is unique and the GCC ecosystem is going to make a phenomenal impact. And institutionalization the model. What must happen next is if you look at the global comparison again India plays a phenomenal role in terms of the GCC density the talent and also the local ecosystem connect even with the things like STPI working on the GCC policy so lot of such ecosystem connects truly helps the global capability centers to adapt the AI ecosystem and make it. Why the strategic partnership is very important.

What must happen next is something very important. The co -creation platforms have to be formed. So especially the organizations like STPI are the real right organizations to build this co -creation platform and build an enterprise sandbox which is already there in the COEs but it has to be nurtured for the GCC perspective. one of the large US multinational banks have done the FinBlue in Chennai COE which has gained the phenomenal success making sure the FinTech COE under the SCPI have factored to the global ecosystem global large multinational bank has benefited out of that that’s where and joint IP framework that’s another important thing which is still under discussion but definitely it’s getting into a better space with that said I just wanted to submit upon one aspect which is the broader reflection right today startups create innovation velocity whereas enterprise creates scale and between these two the global capability centers are the pathways for building between the innovation velocity and the enterprise scale because that helps you to get the pathway to navigate the challenges that you get in the global enterprise ecosystem and be working in the GCC’s that helps to get your product or take your service, use it locally in the environment of global ecosystem, get the acceptance even if it doesn’t work, it fails fast, nothing is going to harm the GCC or the global organization.

That’s where the opportunity through the GCCs are truly evolving to work for your AI products to the larger ecosystem. Thank you so much. Jai Hind, Jai Bharat.

Thank you, sir, for your valuable industry perspective and for highlighting the role of GCCs in nurturing startups. Next, may I invite Ms. Geetika Dayal, DG Thai Delhi NCR, to address the audience on the entrepreneurial and startup ecosystem.

My warm greetings to the dignitaries on the dial. Thank you so much, Arvindji, for this opportunity. And to you and to your entire team for, I know, countless hours of effort that have gone into this massive exercise of putting this program together from which all of us are benefiting. A very good afternoon, friends. We are gathered at probably one of the most important AI policy and innovation programs or platforms that our country has seen. And this summit truly represents the national ambition at its strongest and very best. But it is the dialogue and this session that we are doing here today that will help realize this ambition, help execute this ambition. All the discussions that have gone on for the last few days around AI policy, national strategy, global competitiveness, etc.

They must translate into real support for startup founders who are building AI products. And that translation only happens through the kind of ecosystems that we build together. I think India’s landscape is expanding very rapidly, making it among the top three global AI ecosystems. But it’s not numbers that actually build scale. It will be these kind of programs and innovation ecosystems that come together to make that happen. At Thai Delhi NCR, we’ve been at the forefront of mentoring and accelerating, working closely with startup founders. And over the last few years, we’ve worked with many of them to help them bridge the gap from innovation to market readiness. What we learned is that some of the areas where startups struggle are areas around business capability, around market access, around access to patient capital.

And therefore, our approach focuses on primarily these levers. which provides deep mentorship from entrepreneurs who’ve already scaled globally, market access with enterprises and GCCs, as you just heard, investor access through various funding stages, and structured capability building for our founder. STPI, of course, over the last, you know, more than two decades, 25 years and more, has done such remarkable work and played a great role in actually creating infrastructure and incubation support, a strong policy alignment, a massive pan -India presence and regulatory and institutional strength. And therefore, together, we really create the complete support stack for founders. And this has been demonstrated by the success of some of our key initiatives around Deep Ahead or Samark, etc., which really helped us to create a strong policy alignment.

And it really shows, it proves that collaboration does multiply outcomes as we work together on it. also I think there are certain strengths that India has which are raw ingredients for what we are all working towards now some of that is world class technical talent that comes from our premier institutions, cost effective innovation which is probably around 30 -40 % lower than operational costs in Silicon Valley strong public digital infrastructure as well as policy momentum to India AI mission and what we are seeing now so we must use all of that and bridge the gap around access to data, compute, infrastructure etc. but the collaboration that we are really excited about is the one with STPI which demonstrates how complementary skills when they come together can create real impact.

I think what we feel is that there are five structural pillars that are needed for scaling innovation. This is around knowledge and capability building, resource access market validation, funding access as well as well as of course ethical and responsible AI which our Prime Minister has been talking about and I think these kind of ecosystem collaborations and organizations they act as trust bridges which reduces the friction between government, startups, corporates and investors so I think as we move ahead we are very keen to work and to see how we can move out of the format of isolated programs and come together to create coordinated strategy there are certain immediate priorities that we can definitely work on this would be around expanding joint accelerators scaling up Samarth which has been going on so beautifully many more corporate challenge programs export readiness and perhaps AI benchmarking reports etc.

What we’d love to see is how AI startup ecosystems thrive, not by competition, but by collaboration. And as you’ve seen over here, to build a robust ecosystem when STPI, TI, GCCs, government, corporates, etc., come together with a shared vision, scale becomes inevitable. And that is what we are all here for, scaling innovation. Today, when we sign an MOU with STPI, it amplifies the impact and relevance, and it is a great pleasure and a great matter of privilege and pride for TI to work with STPI as a key enabler and a partner. So our very best wishes to all of you. As Rakeshji mentioned, these are times of great change, probably something that our generation has been very fortunate to see where we were and what we are heading towards.

And for all of us, to play a small role in what the years ahead will bring is really a humbling experience. It’s a great opportunity to be here. And my congratulations to all of you and my thank you for having all of us together here. Thank you so much.

Thank you, ma ‘am, for sharing Ty’s remarkable journey and continued commitment to entrepreneurs. May I now invite Ms. Nija Shekhar, Director General, National Productivity Council, for her special address. Ma

‘am, please. Good afternoon to you all. It’s a delight to be here at the AI Impact Summit. And specifically in this session, being hosted by the Software Technology Program, Park of India, where they have invited the National Productivity Council, who I represent, Ty, and other partners together, the GCC partners together. we are all talking about scaling AI innovation through the startup system. My warm greetings to everyone, to our ecosystem partners, industrial leaders, GCCs, mentors, investors and the startup founders who are also here as we work together on our next growth journey of innovation and AI impact. Anchored in the seven chakras of human capital, I am talking about the event that is anchored in the seven chakras of human capital, inclusion for social empowerment, safe and trusted AI, science, resilience, innovation and efficiency, democratizing AI resources and AI for economics.

Economic development and social growth. and the Three Sutras of People, Planet and Progress. This summit is focusing very effectively on a development -oriented framework for artificial intelligence. Today’s special session, where we are discussing the national imperative of scaling AI innovation, we will exchange a memorandum of understanding with STPI, NPC, and STPI have planned and pledged to work together for scaling this AI innovation, support the AI startup ecosystem in the country, and also to bring together innovation and collaboration. Because we know this is… This is not the era of competition. It’s an era of collaboration, where we have to put our energies together. and focus on areas that impact the population for good. This diffusion at scale across sectors, value chains, MSMEs, clusters, and public services.

This is what we are looking at. NPC, the National Productivity Council, works for productivity in the entire economic sector in the country. So we work on the total factor productivity, labor, land, infrastructure, capital. We support these areas and bring every player, make every player a part of the larger growth in the Indian economy. And manufacturing is a major focus area. Services, of course, but manufacturing. Because we know that that is where the employment is. And that is where. And the maximum exports also are growing. going to grow in the future. And of course, it is also going to maximize the GDP in the country. We’ve seen in this Expo area, we’ve seen many small AI applications, many of which are from the startups, working on areas of agriculture, health, some very, very interesting innovations on health and education, which is something very, very dear to all of us.

In areas like textiles, pharmaceuticals, etc. The question now is, how do we reliably move from ideas to impact and be meaningful to the society under the overall theme of welfare of all and happiness of all? Let me offer a crisp three -part framework for startups and ecosystem builders, which is trust, testbeds, and and traction. Trust is the entry ticket. If customers can’t trust our AI, they will not adopt it, on a scale not at least. So trust brings privacy and cyber security by design, transparency, accountability. It also means operational reliability and responsible governance. Testbeds. This will bridge the promise and proof. Startups need real world sandboxes. Labs, testing environments, applications, reference architectures, etc. And traction is what turns pilots into scale.

Not just a demo, but actual implementation. So we feel that the STPI will bring the ecosystem together and play a pivotal national role by bringing the industry a connected innovation landscape through their entire setup of innovation hubs, platforms, structured programs, centers of excellence across the country and their digital enablement frameworks. They have very successfully over a period of time connected mentors, labs, startups and resolved the challenges facing the startups leading them to larger markets. So their ecosystem of shifting from incubation to scalable infrastructure has seen very good days and we are going to see much much more success in the future. NPC’s role is to strengthen infrastructure. In the adoption spine of this ecosystem. productivity, quality, capability and industry alignment.

In the AI era, productivity is not just efficiency in land, labour and capital, but also reliability, repeatability, safety, security and responsible performance. Everything at scale. Startups scale faster when they can demonstrate measurable outcomes. Better output, better quality, fewer defects, faster service delivery, better customer experience, end -to -end experience and success. NPC supports this outcomes -driven pathway through benchmarking. We are very good at creating models, frameworks, assessment assessments and assessments, evaluations and providing a platform for the industries, MSMAs or the sector wise platforms are very well developed by NPC. This is an area many of you would have maybe if you are not associated with NPC you would know many people who worked with NPC, moved out in the economy, in the consultancy sector, in the evaluation sector and worked through benchmarking, capacity building and spreading of the productivity culture.

That’s why the partnership that we are looking at with STPI, between STPI and NPC is very timely, very strategic and we feel it will accelerate responsible digital transformation. And EA adoption. especially for MSMEs, clusters, industry ecosystems and RAI startups. We are really looking forward to a partnership where we can bring in more productivity into the ecosystem and today’s summit is a context that provides us and asks us and exhorts us to reorient our energies towards a more productive AI system that is scalable, that supports the AI startups and also has a very productive

Thank you, ma ‘am, for your inspiring words and for reinforcing the importance of productivity and capability development. Now, I would sincerely request our DG sir, Sri Arvind Kumarji, to kindly enlighten the audience with the keynote address.

Hello, namaskar. Good afternoon. I think this, when such session happens with Expo, it’s very difficult to have the attendance. Since morning, I am fighting for this only, who is speaking, kindly ensure that attendance is there. But here, there is no problem with attendance. So, organizer, thank you very much. I think you did a wonderful job. The Expo is going on. still we have the full attendance, people are standing also there. Neerja ma ‘am, other dignities on the dais. I think there are a lot of other business pending, some felicitations and all right. So just those who are not familiar with STPI, two minutes about STPI, two minutes about the subject and then I will end.

So STPI has 70 centers across the country where we provide incubation to all small IT companies. These centers are generally in tier 2, tier 3 cities. So we have 62 centers in tier 2, tier 3 cities. Apart from the 70 centers, we have 24 centers of entrepreneurship which are domain specific wherein we provide at least 60 degree support to start -up. We nurture them. We provide some seed fund to them. we provide global reach to them, we provide market access to them and of course incubation to them. So this is what the STPI is doing when it comes to startup and all. Other things we are also doing like BAPT, network security, data centers, cloud services in PPP partners. So lot of things are doing just for that STPI domain.

Now as far as this topic is scaling innovation is concerned, so I mean there is a big change when there is no concept of startup in the country. We used to call it MSMEs and those MSMEs are generally meant for supporting the big companies, especially PSUs. They create something, then merge with either PSUs or they provide some product which is a product for something as input to PSUs or the big organization. Now this change of the startup and with support by the government to the startup has changed the whole landscape. Now startup can themselves scale their product. This is the change which you can see in last 5 to 6 years. And if you really want to scale up your innovation, then what is actually required for the startup, that is that product or that innovation should be safe and trusted.

Unless it is not trusted, nobody is going to use it and it is not going to scale up. Now how to make this trusted and safe, especially in the AI era? That means you have to make your product which is responsible and the product which also ethical. You have to make sure that the product is safe and trusted. only then people will have trust in it only then it can be scaled. Now people are generally confused between two words responsible and ethical. Though these two words are interconnected but different. When I say responsible or when I say ethical, though it is part of all five big parameters like we say it should be accountable, it should be secure, there has to be privacy, it has to be fairness.

These are five words we use. But difference just by examples, when you say something ethical, that means like as a CEO of the company because you are owner of the startup, whether you are taking care of environment or not, when you are producing your product. that is a part of ethical or when you are going to product whether you are taking care of the jobs creation or not this is a larger part this is your ethical attitude what I am going to do with the product when it comes to responsible responsible means fairness which means whether it is a not biased towards anything whether it’s not biased towards country whether it’s biased towards male female gender caste religion then it is a fair product and when I say responsibility it also means somebody should be accountable accountability is the very important part of the responsibility which means that suppose there is a car hit somebody in the road and the car is a driverless car now who is accountable for that accident whether the person who has purchased the car that is a person who has created that car name of the company or somebody who develops algorithm or the this is even the large language model which has been used by that very rappers.

So this is accountability. So unless you are not able to make something which is responsible ethical and therefore safe and trusted, it can’t be skilled. So all startups are there. They must ensure that whatever they are going to create today everybody is using UPI because they have able to create the trust among us. Lot of things came in this country. But now today biometric attendance or biometric identity has become scalable because they were able to create that this product is trusted, this product is safe. So any product which you are going to create whether the product is related to anything if you really want to innovate which is a very good thing and if this country has that opportunity, scale up anything we have population of 1 .4 billion and here scalability is very important.

and therefore if you really want to scale your product, you really want to scale your innovation, that must be safe and trusted. Thank you. Thank you very much.

Thank you so much, sir, for always encouraging, enlightening and guiding us throughout the journey. Now we begin with the MOU exchange ceremony. The first MOU between STPI and National Productivity Council May I request Sri Ashok Gupta, Sir, Director, STPI Gurugram to please come on the dais and Sri Nikhil Panchabai, Director, NPC to please come on the dais and exchange the MOU. I would also request DG Sir and DG Ma ‘am to grace the audience. Sanjay Sir, please come and grace the dais, please, Sir. Can we have a round of applause for Shri Sanjay Gupta sir, our Senior Director, STPI Thank you so much Sir, please be on the stage sir Shri Ashok Gupta sir, if you would like to The next demo you exchange is between STPI and Thai Delhi NCR For that may I request Shri Sanjay Gupta sir, Senior Director, STPI and Ms.

Geetika Dayal to please come forward and exchange the MOUs please Thank you Thank you. Thank you.

So we now come to one of the most awaited segments, the startup felicitation ceremony. Today, we recognize startups supported under STPI ecosystem for excellence across revenue, funding, employment, women participation, innovation, and AI -led impact. I would like to request our Honor Dignitaries, DG Sir STPI and Nirja Shekhar Ma ‘am, Director General, National Productivity Council to kindly come forward to present the certificate and trophy to our startups. I request these startups to kindly come on the stage as per the name announced. So the first name is, may I invite Phoenix Marine Exports and Solutions Private Limited to come on the stage. They are being recognized under the category highest revenue up to 25 CR revenue and highest impact based on revenue tier 2 and tier 3 reason.

May I request DG STPI and DG NPC to please present the certificate and trophy. Once again, a big round of applause for their outstanding contribution. Now, may I invite Vimeo Consulting Private Limited to please come on the stage. They are being recognized for highest funding raised up to 25 CR revenue category. Heartiest congratulations on your fundraising success. A big round of applause. A louder round of applause, please. Now, may I invite Swada Agri Private Limited to the stage. Thank you, Judy. They are being felicitated for highest employment generation up to 25 CR revenue category. Congratulations for generating valuable employment. A big round of applause. Thank you. invite Strangify Technologies Pvt. Ltd. to please come on the stage.

They are being recognized for highest number of women employment up to 25 CR revenue category. Well done for empowering women in the workforce. A big round of applause. A louder round of applause for women participation. Now, our next startup is Suhora Technologies Pvt. Ltd. May I invite Suhora Technologies Pvt. Ltd. to the stage. They are being recognized for highest revenue up to 50 CR revenue category. Congratulations on your outstanding business performance. A big round of applause. Suhora Technologies Pvt. Ltd. Now I invite Puvation Technology Solutions Private Limited. They are being felicitated for highest funding raised up to 50 CR revenue category. Applause for your impressive funding milestone. A big round of applause. Now I invite our next startup, Sequera Tech IT Solutions Private Limited to come on the stage.

They are being recognized under multiple categories. so the categories are highest employment up to 50 CR revenue category highest women employment up to 50 CR revenue category highest AI based impact based on revenue a special recognition for excellence across multiple dimensions a big round of applause now I invite our next startup Atmik Bharat Industries Pvt. Ltd. to the stage they are being recognized for highest impact based on beneficiaries congratulations for touching countless lives a big round of applause May I invite Mobile Pay E -Commerce Private Limited. They are being validated for highest impact based on beneficiaries as a second position. Well done for your meaningful outreach. A big round of applause. Now I invite the another startup Devnagri AI Private Limited to please come on the stage.

They are being recognized for highest AI based impact based on revenue as a second position. Congratulations on leveraging AI for impact. A big round of applause. Thank you. Thank you so much, sir, DG, sir, for attending us. Now I invite our next startup, Dactrosel Healthcare and Research Private Limited. They are being recognized for most innovative startup. Applause for Breakthrough Healthcare Innovation. A big round of applause. Now I invite our next startup, EZO5 Solutions Private Limited. Please come on the stage. They are being felicitated as most promising innovation. Please, please A big round of applause Thank you Now I invite our next startup Connector Foods Private Limited. Please come on the stage for a beautiful couple.

They are being recognized as most innovative startup as a second position. Well done for creative excellence. A big round of applause. Finally, our last startup, may I invite Fuse Ledge Innovations Private Limited. They are being recognized as most promising innovation, second position. Congratulations on your forward -looking journey. A big round of applause. A big round of applause for all our felicitated startups. Your innovation, resilience and contribution to India’s digital economy truly inspire us all. May I request our dignitaries to kindly resume their seats on the dais. We will now invite our selected startups to briefly share their journey with us. So may I invite Fuselage Innovations, Private Limited, to kindly come on the

Hi everyone, my name is Devika Chandrasekaran. I’m the co -founder of Useless Innovations. It’s truly an honor to stand on a stage today being felicitated by STPI. This moment feels very special because we started our journey with STPI in our early days. Back in 2021, we participated in a program called Scout 2021. At that time, we were building our prototype. The support we received through the program was not just a funding, it’s a validation. That recognition gave us the confidence to push forward. We’re going to do it. Today, Fuselage Innovations manufactures drones in agriculture, defence, disaster management applications We are working with more than 10 ,000 farmers across India helping them to improve productivity, efficiency through drone technology We are also contributing to defence, disaster management and maritime operations serving critical national needs Last month, we were deeply honoured to receive National Startup Award and we got the opportunity to present our journey in front of our Honourable Prime Minister, Narendra Modi Sir I would like to sincerely thank STBI and everyone involved in the journey to believe a startup like us The ecosystem, the encouragement and the early trust that make a huge difference in our journey Thank you so much

Thank you for sharing inspiring story. Now may I invite Dr. Rosals to kindly come on the stage and share your startup journey with us. Dr.

Good evening, everyone. My name is Dr. Soumya, and I’m really glad to be a part of this prolific platform today. Just very quickly, I’d like to walk you through what we build. So at TectoCell, we build AI -powered diagnostic solutions at the intersection of radiology and artificial intelligence. And DNA sequencing, while addressing the huge havoc of drug resistance and robust clinical trials panning across India, facilitated by the software technology parks of India. We’ve been able to sort of exceptionally benchmark our accuracy, clinical accuracy, that sort of amplifies the reliability of our products. And the continued commitment of Software Technology Parks of India to sort of help us navigate through our regulatory compliances, get global collaborations, and also sort of get data acquisition, which is sort of machine readable, is extremely noteworthy.

And this unique foundation sort of puts us in a very good position, in a very strengthful position to now sort of scale this globally, building from India for the world. So I’m very grateful for this. Thank you.

Thank you. Lots of applause. Thank you so much for sharing your story and journey with us. Now I invite Sequeira Tech IT Solutions Private Limited to come on the stage and share your startup journey with us.

Hi everyone. Good evening to everyone. So my name is Arita Dalal. I’m heading this region for North with SecureTech. I have been in this organization from the last 11 years. But during this journey, we have a lot of interaction with STPI as an organization. They are one of the nurturing body which has done a lot of collaboration in the industries as well. They are one of the bodies which has given us an opportunity to talk to the investors as well. And there are various industry connect as well that is being established by the organization. And we are very sincerely thankful to the entire organization and the team of STPI as well. Just to give you a brief about.

SecureTech is a cyber security organization in VR. Our mantra is to simplify security. We are touching, we are securing the security for the large enterprise organization, with size organization across the industry, whether it is pharma, banking finance organizations, or even the small organizations which are currently establishing the digital landscape in the country, while they are being regulated by large RBI and CBI. So, in nutshell, we are providing them all the frameworks, security parameters, and the solutions as well, so that they can be powered, they can be enabled, and they can secure their infrastructure platforms and the data that they are processing for the countries or for the users that they are providing services. So, whether it is a startup organization or even a large infrastructure organization, we are securing.

We are providing them end -to -end. Thank you. Thanks, everyone.

Thank you. Now, I invite… Caneboard Solutions Private Limited to come on the stage and share your journey with us.

helping us sharpen our positioning as a deep tech company. Most importantly, STPI’s recognition has strengthened our credibility with customers, investors, and government stakeholders. We are very happy and very honored to be here today, and we thank you so much to STPI and everybody who is present here today.

Thank you so very much. May I invite now EZO5 to kindly come on the stage and share your startup journey with us.

Hi, everyone. Good afternoon. I’m Noor Fatma, co -founder of EZO5 Solutions.

Hi, I’m Meenal Gupta, founder of EZO5 Solutions.

at EZO5 we have built an AI powered platform Imagix AI that does precision treatment planning for oncology cases and so in the in the startup journey there was a time for us when we one and a half years back when we had just two months of cash flow with us we were thinking a lot what to do and that is when STPI came to our rescue and it helped us raise money and there has been no looking back since then so in the past three years where we have been incorporated we have processed around one million scans we have in the last three months we have scanned around 50 ,000 chest XAs where we have flagged around 4 ,000 cases of TB cut the transmission by short we have flagged six cases of lung cancer where the intervention was still possible so We have prepared 1000 radiotherapy plants in the last three months and we have cut short the treatment planning and start from around one month to a week.

So that is the impact we are making to the support of the whole ecosystem and STPI.

And proudly I say that the impact that we have brought, even our Prime Minister Mr. Narendra Modi was interested and he invited us to discuss our solution in IMC. And just day before yesterday, we have gone global because Bill Gates showed interest in our solution and he invited us in Microsoft to show our solution and he was discussing how he can help us. Thank you.

So now we are going from local to global serving the whole world. Thank you.

Yeah, thank you. Thank you to all the founders for sharing such inspiring stories. So, we now proceed with presentations of mementos to our esteemed dignitaries. To begin with, may I request Shri Ashok Gupta, Sir, Director, STPI, Gurugram to kindly come on the stage. Sir will present the memento to Nirja Shekhar ma ‘am, Director General, NPC. A big round of applause. Thank you so much, Sir. And thank you so much, ma ‘am. Next, may I request Shri Atul Kumar Singh Sir, Additional Director, STPI to kindly come on the stage and present the memento to Shri Bala M .S. A big round of applause. May I now request Shri Praveen Kumar Sir, Joint Director, STPI to kindly come on the stage and present the memento to Geetika Dayal Ma ‘am.

A big round of applause. May I also request Shri Atul Kumar Singh Sir, Additional Director, STPI to kindly come on the stage and present the memento to Geetika Dayal Ma ‘am. Shri Praveen Kumar sir, Joint Director, STPI to kindly present the memento to Shri Rakesh Dubey sir, Director, Startups and Innovation, STPI. Thank you sir. A big round of applause. Now, I would like to request Shri Praveen Kumar sir, Joint Director, STPI to present

the formal vote of thanks. Expected dignitaries, speakers, startup founders, innovators and ladies and gentlemen. On behalf of Software Technology Parks of India, it is our true privilege to thank each one of you for making this session focused, meaningful and definitely forward. looking. Nirja Sekhar ma ‘am, thank you for your thoughtful reflections on productivity and growth. Your perspective adds depth and direction both to our collective mission ma ‘am. We are truly encouraged to have your presence. Thank you so much. We are grateful for it. Sri Rakesh Dube sir, thank you for your profound support which has been both guiding and grounding sir. Your constant encouragement and hands -on involvement in shaping the entire session together has helped us immensely sir.

My sincere appreciation to Geetika Dayal ma ‘am from Thai Daily NCR for your continued partnership and reinforcing the importance of collaborative startup ecosystem building ma ‘am. Thank you. Thank you, Mr. Bala, for bringing up a sharp industry lens and pragmatic approach that startups can directly relate to as they scale. So your thoughts on the GCC is definitely going to help them all. To the startups, all the startups who were felicitated today, congratulations. Your achievements demonstrate that innovation from India, including Tier 1 and Tier 2, is both scalable and globally relevant. To all the founders who shared their journey, thank you for your candor and inspiration. Your stories remind us why platforms like STPI matter. And before I conclude, I sincerely appreciate.

My organizing team and every colleague who worked diligently. behind the scene to ensure the session came seamlessly. With that, I once again thank all of you and the dignitaries and I request dignitaries and startups to come forward and have a group chat. Thank you. Thank you again.

I request to all the felicitated startups to kindly come on the stage and have the group photograph with all the dignitaries on the dais. Thank you. Thank you. I also request the other directors as well to please come on stage and join us for the group photographs. Yes, Kavita ma ‘am, please come on the stage. I also request Kishori ma ‘am to please join us for the group photograph. Thank you. Thank you. Once again, thank

I give this example because I’m fairly confident that when you look it up and when you try it yourself it will work. And I know it will work, by the way. That is, it will fail rather. On the current versions of ChatGPT, it will not fail, by the way. In the next generation, I do some stuff with Google for example, it won’t fail in the next generation of Gemini anymore. Because they’re putting a lot of effort into fixing this one error. They haven’t fixed the underlying problem. They saw some presentations of people like me pointing this stuff out so they’ve just put a band -aid on top. Now we can’t run life on band -aids.

Band -aids is what? Band -aids is students mugging up one answer before the exam so they get the marks for it. That’s not real learning, by definition. The problem is that we’ve built this system which is our attempt to have general software. And we don’t quite know how to do it. We don’t quite know how to handle it. So let me clarify what I… I’m going to say something incredibly stupid and then I’ll bring it into place. We were talking about this not too long ago. A long time ago you had machines that could do one thing. A hammer is a hammer, a car is a car, a door is a door. You can’t use one as the other.

I’m saying something that sounds incredibly stupid, but think about it. Why don’t you need two separate computers, one to run Excel and one to run PowerPoint? How come both run on the same machine? This is not obvious at all. We’re just used to it, so it seems obvious, but it wasn’t obvious. In fact, the first few computation machines that were made, if you go back, look at all of this Vannevar Bush and even before that Charles Babbage, all of these names one reads in history books or whatever, you’ll see they had differential analyzers and this, that and the other. Oh, this machine, it can add. That machine, it can solve differential equations. This other machine.

It can fit curves. This other machine. It can. do this mapping task. This idea that you could have one machine which could do everything was completely ridiculous because there’s only one thing in the universe that we know of that can do that and it’s the human brain. The human brain is a singular object that can retrain itself to play billiards, to arrange chairs in a room, to present, to drive a car, it can do all of these things. So due to a bunch of very clever people like Alan Turing and co, we figured out that wait a minute, we can have one computer, we can build this one machine. I mean think of it just from a manufacturing point of view, like jackpot.

We can build one machine and it can do all the things. All we need to do is we need to have different software, one for each task. So we’ll have one software for Excel, one software for PowerPoint and the same physical machine will be able to run both. So we built general hardware. And that worked for decades and the fact that we had general hardware to the computation and information revolution. Now for the first time, instead of just having general hardware, that is one machine that can run all software, we have general software, which is you don’t need PowerPoint and Excel separately. You can have one software which you tell it what to do and it will do the job of PowerPoint and you tell it something else and it will do the job of Excel also.

That’s what we are trying to build with AI at the end of the day. Going from general software to general hardware. And as we know, this edit, this ability that we got when we built a general purpose machine, before you needed to spend all this money and build separate machines for every task, and the moment you had a single machine that could do all things, that led to an absolutely massive change. It was a massive revolution. Now that you have general software coming in, right? Once we learn how to do that, think of how the world is going to change. Software companies which used to be, there’s a very interesting graph that I really should have put here, which is if you’re manufacturing something, there’s a burn rate.

So you have an increase in the amount of money you have to invest in your company initially. And then if you manufacture 10 cars, you have a certain amount of money you need to invest. If you want to increase the number of cars you manufacture, I’m talking toy cars, I’m not rich enough to manufacture real cars. But as you increase the number of toy cars you’re manufacturing, your costs go up and sort of linear. And there are bumps every time you do a new round of R &D or something. Software companies don’t do that, right? Software companies, you have this huge expense at the beginning to build everything up. And then once you have that, your burn rate is relatively low.

Selling 50 ,000 units of a software and selling it for $1 ,000. Selling 2 ,000 ,000 units of a software. isn’t going to make a material change in the amount of money you’re investing every day. That entire economy is now going to be gone because you don’t need that kind of investment in software anymore. And this has led to multiple real economies collapsing. So I’ll give you two examples just off the top of my head. Web design companies. There were thousands and thousands of them all over India. You know, a group of college students get together, they say, look, we’ll build websites for people. And these were all micro and medium industries, maybe employing anywhere from 10 to 50 people. That economics is just gone.

We all learned when we were small, right? What is the definition of economics? Economics is the study of the allocation of resources under conditions of scarcity. What if it’s not scarce? There’s no economics affair, despite the fact, again, that we’re in Delhi. There’s no econ affair. But similarly now, econ of maybe writing novels is gone. on, right? You saw what happened with C dance recently, just 24 hours ago. The movie industry is worried. Who’s why should people invest in making movies? If I can write, you know what? I want a movie like Sherlock Holmes, but I want Salman Khan to be the main character and I want me to be the side character and I want this to be the story maker to our movie.

I press enter movies done, right? If that comes to pass, then that, that entire economics is just gone, right? We have seen, these are me talking about the future. Let’s talk about right now, right now at this very moment, a large portion of the internet is collapsing because what used to happen is a large portion of the internet used to run on ads, right? So if I have a recipe website, what do I do? I put some ads on it. You visit my website to read my recipe for blueberry cupcakes or whatever, and you get that ad displayed to you. Okay. And you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get that ad displayed to you and you get I get some money from like the Google, you’ve seen that at the bottom of pages and so on, right?

So I get some money off of that. The problem is that now who’s going to come to my stupid website? They’ll just ask ChatGPT or Gemini for that and they’ll get it and nobody’s going to come to my website. Generally speaking, if you got your search engine optimization correct and you were on the first page of Google, your click rate was one in six, okay? This was the official statistic. That is, let’s say I am a top blueberry cupcake chef. I don’t, that’s definitely not a thing, but let’s say I am. I’m very proud of my blueberry cupcakes. And I’ve made my website and everyone agrees it’s a great site, so when you search for blueberry cupcake recipes, let’s say I’m one of the top 10 and so I would normally have, because people don’t just click one link, they usually go to two or three, I would have a one in six chance of getting clicked and I would make some money off of it.

That number in the past year has gone from one in six to one in seven. in 1500. This means that this is multiple orders of magnitude. So all of these websites that ChatGPT and Gemini and DeepSeek and all of these people, they got the data from these websites only. But now no one will go to these websites and they’re all dying. This is even true of open source tools. So Tailwind, which is a major CSS platform, had to let go of a lot of its engineers because what’s happening is these tools have eaten all the open source code and then people are no longer going to the open source libraries to get it. They’re just saying make me this thing that does that and do it.

Of course there are positive sides. There are non -technical people who can now just say things to the system and it will build them a nice little app, which is great. But simultaneously we are destroying much of the infrastructure and much of the information landscape that made this possible. In the first place. So we’re going to do this. So we have to be exceedingly careful about that. Let me sort of poke on that last sentence that I said. And I think that’s a really important thing when we talk about correctness, trustworthiness, and all of this, right? Which is, in many ways, you know, we had machine learning before 2020 also, right? We were doing classification. We were doing all sorts of clever things.

What really changed with ChatGPT was that anyone could use it. It was the genius of the interface. You had the simple chatbot. You didn’t need to program anymore. You could just say things and it would do them, right? And it is this ease of that interface which changed everything about how we interact with these powerful AI systems. But there is an inherent danger in that. What is the danger? Well, we didn’t build, you know, computers. Languages, all their brackets and, you know, weird expressions. We didn’t do that for fun. Okay? We could have had computer, if we could have written computer programs in English and have them run, we would have stuck with that only.

Why create all of these complicated looking languages where if I miss a semicolon, my computer is going to turn into a peacock, right? We did it that way because our normal language is too ambiguous. There are too many ways in which we say things where we assume you already know what I’m talking about. It’s too easy to miscommunicate, right? The teacher told the student that he was going to the fair. Who’s going to the fair, the teacher or the student? This is obviously a very stupid example, but we have thousands and thousands of ambiguities in our language which make it exceedingly difficult to understand what the other person even wants. That’s why we had computer languages in the first place, to disambiguate.

Now we are saying, no need. I will just give the problem description. This general purpose. Software is just good. going to basically custom solve it. Think about how useful your instructions are. This is deadly, right? We have literally got stories about this, right? About how easy or hard our instructions are. We have cautionary tales about the genies and monkeys for storylines, right? Yeah, yeah, yeah. We’ll switch at 15, don’t worry. Yeah, so when we get to those storylines, we hear that someone says, I want to be the richest person in the world or I want to be the most beautiful person in the world. And what happens immediately after that is it kills everyone else. And it says, I have technically, correctly satisfied your query.

Everything you said, I have done. And so when we give a query, we want the machine to basically align with my expectations. That’s

what alignment is. That’s what alignment, that term means, right? That we wanted to align with my expectations of how this stupid thing is going to act. That leads us to the following conundrum. That I have the system, it’s going to do certain things. I worry that it may do certain bad things. How do I define what is the risk of it getting into this bad thing and doing this bad thing? Do we have a clear way to define risk in our context? And for that, I’ll hand over to Anirban. Alright, I can take the clicker. So, I will keep it slightly brief and I’m going to skip over some slides in the interest of time.

We have looked at different aspects. Three of us are at Ashoka, we work together on different aspects of risk. Safety, risks, harm reduction, risk management, safety, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk management, risk and trying to quantify them and understand them.

This is just the map of India part. I’ll get back to India as a question. India is a big nation, as we all know. But there’s a lot of technology, and we have a tendency to solve our questions of scale using a lot of technology. That naturally introduces many challenges. You’re on the fifth day of the summit. I don’t need to tell this to you. All of you have seen different examples of how empowering this technology could be and why it’s important to be a bit skeptical about its deployment, because that could introduce new kinds of risks. But what is a risk? It’s hard to quantify that and define it. Risks and harms would mean different things in different contexts.

Our goal as a team was to understand and try to make sense of these risks. So hard to define. One definition that we’ve chosen is that the probability of an undesirable outcome characterized by two things. The two things are its likelihood and its severity. I think the airplane example is a good example of that. The two things are its likelihood and its severity. This example is just soon up. Okay, it’s coming back. But basically, airplanes are unsafe, all of you know that. Most of you also take airplanes. It’s because the probability of something happening is lower, that’s what likelihood comes in. But airplanes are dangerous, that’s why we like watching aircraft investigations, because of severity.

Those two are just oversimplifying what I mean here. These definitions also need to be grounded in context, context such as where you’re deploying these systems, so education, healthcare, some of the many areas that have been discussed in many of these panels and discussions across different halls here. I’m going to keep it brief. But these risks go beyond hype. There are real, real challenges and the real costs that everyone has to pay when such systems are deployed at scale, without taking risks into account. Some cut off, but one example is from Air Canada. There are many such examples. These are examples of real people suffering. loss of life, loss of liberty, loss of money and property because of AI safety risks.

So we have taken a life cycle view of AI safety risks and tried to create a taxonomy. It’s a comprehensive taxonomy of 37 different kinds of risks. We have launched it earlier today and it’s now available online. I’m just going to give you a brief overview of the kind of work we have done towards that. Again, here are some examples of what is a risk in our definition or not. So what is not a risk is physical destruction of infrastructure. It is an AI related risk but we are not talking about that. Our scope is very limited. There are many global frameworks that talk about these kind of things. You have some coming from Singapore in Asia, we have Europe, we have the US.

But they do not take into account the main challenges that we see in India. India has scale, India has linguistic diversity, but India also has a lot of different things. India also has certain problems like low network connectivity. If you, for example, are deploying AI in a space which is safety critical, but you lose network and someone’s life depends on it, then it could be another kind of challenge that has to be uniquely defined in India. We see that many of these challenges are not covered in international repositories and risk databases like these. So what they have is what we call contextual blindness, where they are not realizing the social challenges and the socio -technological challenges.

India, again, as you know, deploys large amounts of technology. We have larger technology

systems than any country in the world. UPI, EVM, Sadahar are just simple examples of that. The safety risk database that we have launched, it’s in partnership with AICSTEP Foundation, and it’s called ASTRA. It’s AI Safety, Trust, and Risk Assessments. We’ve tried to create a fun acronym that is easy to remember. ASTRA is now formally launched. Some of us worked on it. Ananya, who’s in the audience, is also one of the contributors. AI. It is a seven -step process. And maybe, Anurban, you could just quickly walk… through this process and how Astra was built. Yeah, hi everyone. So both Devayan and Alok did a good job summarizing the overall work. So these are a bit of technical details.

I’ll probably skip most of it. Basically what is there to understand is that this, if you think about it simply, it’s basically a risk, it’s a database of risks, right? But they are contextualized in the Indian context heavily, right? So one formula fits all kind of a narrative does not work in AI safety. This is what our claim is and this is in line with many researchers, right? Many prominent researchers. So what we started with was resource identification and here’s what our work differs from many of the global frameworks that people have built. So when it comes to resource identification, we had to actually do bottom -up research of how and where exactly these risks occur in the Indian context, right?

We have primarily education and financial as of now. but we started an exhaustive study of how exactly these risks manifest across sectors, right? And the final step of this is a comprehensive risk taxonomy and ontology, right? So taxonomy is basically categories and subcategories of risks which you will find probably in many global frameworks but what is there in our database is an illustrative set of use cases, right? Where you have a use case, a risk use case which you can go and click if you are in the financial lending sector, right? You can go click and see what kind of risk has happened in the Indian context exactly related to our language, our caste, our whatever kind of variables we care about, right?

So these are some of the basic steps through which we have worked on building Astra. So there are two parts to it very briefly. So one is the causal taxonomy. So one is we also tell you through this database at which stage the risk has occurred. So it can occur during development. for example bias in AI we all know about it it happens because of probably bias training data that is one of the sources right so it happens during development deployment let’s say you take an AI system which was built in the US and you implement that or deploy that in an Indian solution setup where most of the people speak in Marathi right this is a deployment problem right so it manifests in deployment and usage I take the AI system it was never meant to disseminate disinformation but I did that as the user I actually manipulated it so that’s in usage and then there are stakeholders is the AI system primarily responsible for the error or risk or is it that it happened because of a deliberate end -user kind of an action it also tells you about whether this risk is intentional or unintentional again in no way do we that this database is in any way exhaustive or foolproof right it is currently you know advancing it more and more expanding it to other sectors but But the target is to also tell you about these granularities around risk.

Because risk is not just one term like Alok explained, Debaian explained, right? You also have to look at what is the intent behind it. So there are two main categories of risk. And this is the part that we struggled the most about. By the way, this Astra, it’s currently available on archive. And you can probably go and read this paper. And you can also take a look at the database whose link is present in that paper. But this work took us almost six months. And again, Ananya, if you could wave. So Ananya is a primary contributor of this risk database that we formed. And so we categorized after looking at the type of risk. There are social risks which are easily quantifiable, which you can easily observe.

For example, linguistic bias. An AI system trained in English does not answer Hindi queries that well. So this is a typical risk which comes under social, right? Frontier risks are risks which are very, very difficult to observe, right? There are risks that we know. Could occur. tomorrow AI could replace jobs we all know about it but how do you quantify it I mean in many of these risks have haven’t even occurred in the Indian context you know about it because from some remote Western translation you could translate it we know there’s a gut feeling that it might go wrong but we don’t we can’t quantify them very easily these are the kinds of risks which come under frontier so there are some examples here I’m not going to the details in the interest of time but there is bias and exclusion toxicity risk categories right and then in frontier you have mostly around power seeking an AI system going rogue I’ll just quickly cite an example right I’m not naming the firm but there’s this news on a trading firm which applied an AI system to go do quick trading according to market variables right the AI system performed very well initially and then without the consent of the firm and because they were not monitoring it properly it went rogue it started doing transactions which were extremely lossy you know it was a risk category and then in frontier you have mostly around power seeking and AI system going rogue I’ll just quickly cite an example right I’m not naming the firm but there’s this this news on a trading firm which applied an AI system to go do quick trading according to market variables right the AI system performed very well initially and then without the consent of the firm and because they were not monitoring it properly it went rogue it started doing transactions which were extremely lossy and not just that in a very high volume it started doing that right so this is the example typical example of power seeking now in India Well, there might be some examples abound, but then do you really know whether this kind of risk can be easily quantified?

We don’t know what will happen. We’ll probably deploy and we’ll have to watch. So those are the kind of risks that we have listed in frontier risk. One quick example is also human -computer interaction, right? So we all know, I mean, sorry, there’s a student sitting here, but I’m going to say this, but in most universities, okay, students are using AI and we know that that leads to cognitive decline and lack of critical thinking. But again, how do you quantify it, right? It’s very difficult. So these are frontier risks, right? I’m not going to the details of this. You all know about caste bias, linguistic bias of AI systems, hallucination we all know about, right?

Incorrect outputs by AI and then infrastructure exclusion. So this is one critical example and this came up from a discussion with the XTREP team that let’s say there’s an AI system that you deploy and a farmer is trying to use it. In many regions of India, there are connectivity issues, right? There is an internet connectivity issue and the entire app starts loading, loading and buffering. It doesn’t work, right? Now this is a typical example of infrastructure exclusion. So again, remember the stage of error manifestation is the deployment. Chat GPT or any open AI for that matter will not care about this. It’s not their job, it’s our job. When we are deploying it in context, it’s our job to take into consideration that our connectivity might be poor.

So this is a typical example of some examples of social risks. So this is one reason why these social risks manifest at this level. As you go higher and higher models, they have more persuasive power so they can manipulate you. Frontier risks I already spoke about. I’ll quickly move on to mitigation. The one quick point I want to make about mitigation is it’s an extremely challenging task. So while the database is the first step, as per our AI safety risk framework of Astra, mitigation as they buy and adequately pointed out, is the hardest task. That we have at hand. So these mitigation measures are often not effective. They are very context specific. and there are certain kinds of mitigation measures that also lead to loss of utility.

So we have to be super careful about that, right? You put a very strong mitigation measure but then that leads to lack of utility on the user’s front. That is not a very good mitigation measure contextually speaking. So according to this work, what we want to carry forward is we want to empirically ground these risks going forward. What is the probability of risks really? And finally, we are also trying to include more and more domains. Currently it’s on education and financial lending. We want to expand it very soon to agriculture and many more