Uncategorized

accessibility, language accessibility and language inclusivity. We are a country of 1 .4 billion people. More importantly, a country of 1 .4 billion voices. We all think differently, we all speak differently, and we all dream differently. But whenever we go online, everything is available only in one language. Majorly English.

To break the language barrier that exists in our country. And we have different solutions and different integrations that we have. One of them is Pashni translation plugin, which is already sitting on top of more than 500 websites, if I’m not wrong, the exact number. And we are enabling people, we are enabling citizens of India who are essentially not being able to understand in English and Hindi because most of the digital content that you see, primarily the website, maximum you’ll see is a website which is sitting in a state. The default language would be there or English primarily. But then what about rest of the languages? Imagine a scenario that I’m someone from north and I’m living there in Maharashtra.

Mostly you will see the content in Marathi or English. But then what about having the same content? I don’t know English. But I really want to understand what is there. And I want to convert it, the different policies at the state level, different guidelines, different content, maybe creative content, etc. You need to know in my language. So, Bhajani Translation Plugin is one of the engineered solution using all the models that you might already be aware of. 350 plus models from our platform. We have this solution as Peksa Swati.

So, as Shailendra Pal mentioned, last year a farmer wanted to apply for the PM Kisan Samman Nidhi. It’s basically a very simple form that the farmer has to fill. But the form was in English. The farmer literally had to travel 40 kilometers only to find somebody who can actually help him out filling the form. This is the language divide. This is the barrier that we are trying to avoid. Eliminate. 800 plus million people. are not fluent in English. And 95 % of the content which is available, it is in English. This is where Bhashni comes into picture. The National Language Translation Mission of India. We are trying to transcend the language barrier. We are creating a unified multilingual layer for India’s digital ecosystem.

We are not just providing language as a feature. We are providing language as an infrastructure. We are encouraging language as the foundation for digital inclusion. Next slide, please. So, like sir introduced, the Bhashni Translation Plugin. It’s a powerful product through which you can have any website being translated into multiple languages, being accessible to all the people in the last mile. And this happens in matter of minutes. Not days. Or months. Or just minutes. This is the power of the product that we are talking about. And you don’t have to rebuild the entire website. You don’t have to redesign it. There is no back -end overhaul. Just one liner, very lightweight, simple code that you can copy and paste onto the website and you will have your website speaking multiple Indian languages.

This is how accessibility is made effortless, inclusion is made scalable, and the last mile reach is made real. So I just want anybody to see. Anybody who can copy and paste. Like we don’t need a developer or a person who knows JavaScript or the entire back -end. just somebody knows copy and paste and we’ll see how with the help of that you can have the entire website multilingual. So anybody who would like to do that? Yes, sir, please.

Maybe, you want to open a website first and show what exactly VashuCast is.

So this is the Vashni’s website and here is the plugin that has been integrated on the website. This plugin will help us have the entire website available in all 22 Indian languages. All right, so while we just give a quick glimpse of what Bhajani translation plugin is, it is basically a very lightweight utility, though. you find it very simple but the content that we have on this website primarily is in English and there are other challenges that you that we would like to discuss later on as how this translation plugin brings in though it looks very easy just you clicked on a button and then you do a translation all together but then we’ll discuss more about what are the different challenges we come across not from the fact the engineering side of it but on the language side of it how we cater and have this challenge taken care so this is just a plugin we just wanted to tell you this is how it works but you know if you go back to English then and then you know we will just talk about what you wanted to we’ll continue with that so I just wanted to have a quick demo of how you can integrate this plug -in onto the website so I think some if yes you can come we’ll just see how with the help of just the knowledge of copy and paste we can have the entire code implemented and you’ll have the entire website translated into multiple different languages.

For the purpose of this demo we had created this dummy website and the code for this website is here. So this is the code that none of us would most of us would not understand. And I would like to request sir to just copy and paste the plug -in code that we have. So we want to tell that this website content is only in English and you want to add multi -lingual flavor to it using Bhashini. You can integrate the solution that we have on the top of

That’s what you’ve meant.

Yes. So if you can sir just copy and paste this code. The code which is written here. Yes.

Anywhere here.

If you can just add a hyphen between translation and plug -in.

Yes.

Can you go back to the website? Refresh it. So you can see that the plug -in is added. And we can now have this website available in all 22 Indian Schedule languages. So that’s the power of this code. We’ve taken care of everything that is happening at the back -end. and you just have to copy and paste the code that we’ve created for you. It’s as simple as that.

So Swati, so let’s say I’ve embedded this particular thing on this particular website. Now it is available. There is the icon. What about if I go to next pages, right? Will the system understand that there’s a link in the I chose and I go to any page? It will reflect Hindi or I have to select every time I go to any page as my language, which I chosen as Hindi.

So you don’t have to apply this code on every page. The pages of the website will automatically understand that the multilingual feature has to be embedded on all the pages. So if you move on to any other page of the website. So this was just a dummy website that we had created. Let me. Go to Bajni translation plugin. in Bhajani’s website. So if you go to any of the pages, the plugin will remain there. And you will have the multilingual feature added on all the pages of the website, not just the home page. So let’s go back to the slides now. So like we just demonstrated, the code that we have for the plugin that we are talking about is a one -liner, very lightweight, simple integrated, simply integrated code, which you can use to have your website available in all 22 Indian Schedule languages.

It is DBM compliant and framework agnostic. So if you have your website, in different, made in different languages, it’s irrespective of that, the code will be applied to your website and you can use the same code.

So Swati, can you just give some light on what is DBM compliant as how the website is DBM compliant? If I, let’s say I have a government website and I want to include the Bhasini translation plugin onto it, what is this DBM compliant that you talked about?

So these are the compliances mentioned in the digital brand identity management compliance book that is available. So for everybody to have an accessible website, the DBM compliance have to be followed. And we have the DBM compliant code with us wherein all the accessibility features like, you know, that happens in the backend, you know, for, any person who is a visually special person who wants to access the website. is able to do that with the help of the technical integrities that we’ve incorporated into the plug -in code that we have. So this is a glimpse of the impact that we’ve already created. We have approximately more than 400 plus websites that are already integrated with Pashni translation plug -in.

From those websites, we get approximately 24 million plus inferences. And we’ve created 1 .5 million plus glossaries. So glossary is something that I will take at a later section during the session only. But just for a short description, glossary enhances the translation in such a manner that the end citizen who is actually consuming the content from the website is able to understand the content. And also, these are the 22 Indian scheduled languages in which the plug -in is available. available. Next slide, please. So while we were creating the plugin, we had to create something that, you know, one size fits all product and which is something very difficult to create because everybody has different requirements and to cater to all those requirements, we had to make one product that can simply be accessed by everybody.

So these are some of the use cases that I will be discussing that our plugin has the capability to resolve to. The first one is that generally what happens in, you know, a product like this, you translate, you know, from English to the target language. But here in our plugin, what we’ve done is that even if your website is, let’s say, created in a language other than English, let’s say Marathi. That can also be translated to the targeted language directly. So you don’t have to first translate the website to English and then move on to the targeted language. You can have the source and target language as per your requirement. So that’s how we’ve not, you know, you don’t have to get into the bridge of creating English as an intermediary to move from one language to another.

Next slide, please. Okay, so when I talk about a website, there are different sections of the website. And not all these sections would you want to translate. For example, the calendar, if there’s a calendar, you would not want it to be translated into, you know, the target language. Including email IDs and, you know, there are certain sections that a lot of people didn’t want to be translated. So there is one class that you can embed that is the skip translation class. Embedding that will help you. Navigate to the, navigate the sections that you don’t want to be translated. So, that’s also one feature that we have with our plugin. Next slide, please. Okay, so, you know, you saw the plugin, right?

There were languages listed in a certain manner in the plugin. So, what happens is at, you know, many regional places, we want the plugin to have the regional languages on top. So, for example, after English, people don’t want to go alphabetically like Assamese, Bengali. They would want their regional language. In this case, they wanted Hindi to come in the, you know, to change the order of the languages that are appearing. And that is also possible. So, if you want your regional language to come on top, you can have that with our plugin. So, you know, majorly what we say is that we would want to… We want to display our website in a certain language.

So for example, if you created the website in let’s say English, but you would want all the users to have the language to be displayed as Hindi first. And probably then they can navigate to their own targeted language. So even if your website, the source language of your website is English, you can, there is a possibility of adding the parameter which can have the source language as Hindi or Marathi or Punjabi as the user requires for all your websites. Next slide please. Okay, so what if your… Your website has the, you know, has been created in two languages. So for example, you’ve created your website in English and Marathi also. So that was the use case that we had with finance department Maharashtra.

So they did not want translation to happen in the Marathi language and the English language, though their source language of, you know, so basically the source language of the website was English and Marathi. So if you want to skip translation for different languages also, you can do that. So in this case, what happens is that the user selects a language. If the language of the source is selected, let’s say, you know, English or Marathi, it will go redirected to the English or the Marathi page of the website. And if the user has selected any other language, it will move on to the normal process of translating it into the target language. Next slide, please. So, you know, sometimes we have portals also.

Yes. So, you know, because we would want to have websites available in all 22 Indian scheduled languages so that we try and reach out to the maximum people. But if that is your use case wherein you would want just three or four languages to be displayed for every user to be seen, you can have that also. So the drop -down will only display four languages in that case? Yes. But it’s always encouraged to have all the languages so that everybody, you know, who’s accessing your website can have the website available. Thank you, then. So, talking about this use case, what happens is that in most of the cases, we also have portals. And in portals, we have forms or, you know, we basically ask input from the user who is using the portal.

So if they apply Bhajani translation plugin and they, you know, move on from one language to another, it will reload the entire page. If it reloads the entire page, whatever the user has filled in, like their details, their name, their email IDs, all that information was lost. So what we did to capture this was that now plugin can also have the portals without the reload picture. So if you don’t want the plugin to reload every time a user selects a language from the drop -down, you can have that. Next slide, please. So this was a very interesting use case. You know, you can see this is how the website was displayed. So the source language of the website is English.

But like we can see, after every English, below every English word. there is a different language. So Haryana written in English, then Haryana written in Hindi. Puducherry written in English and some other language. So here this was use case of handling mixed languages. So what we did here was that whenever the plugin sees that the source language of the plugin is different from what characters it is getting, like here in Haryana, it is getting Hindi characters also, it will skip this translation automatically. So you would not have to skip it at your end. We’ve done it and we’ve created it, we’ve designed the plugin in such a manner that if the source language of the website is, you know, if the contents going for the translation are different from the source language of the website, it will automatically skip the translation.

Next slide, please. So… With certain use cases, what happened was… that there was a lot of dynamic content on the website. So, static content can easily be translated. Like, it is also difficult, but it’s not as difficult as handling the dynamic content. But for certain, like for State Bank of India and for MyBharat Hotel, the dynamic content was changing so rapidly that it was making too many API calls and the response time was getting delayed. So, what we did there was that we intelligently had the code running in such a manner that the dynamic content was, the translation of dynamic content was handled in batches. And that’s how the, you know, API calls, the increased API calls reduced and the response time was stabilized.

Next slide, please. Okay, so now… We all can, you know, navigate to the website. select the target language on the website and have the website available in the target language. But what if somebody cannot navigate, cannot select a language from the drop -down? We also, with Rail Madad, you know, if you go to the Rail Madad’s website, there is a mic button. So you just say out your language. So for example, if you say out Gujarati, the entire website will turn into Gujarati. So that’s the capability of it. Next slide, please. Okay, so this is a very recent use case that we’ve handled. So like you can see here, there is the MSD website. And there is also another domain name, which is Hindi, which is in Hindi.

So what the client wanted was that, you know, once the user selects Hindi as the drop -down, the translation happens, but it also redirects to the… Hindi domain of the website. So that mapping of which language to which domain, that is also something that we have done at our end and you can have URL redirection also. Next slide please. Okay, so what happens, so let me just ask, I hope everybody here understands Hindi, right? What is the translation of home in Hindi? Ghar, Ghray, that’s right, right? But the home tab on the website, if it is getting translated to Ghar, it’s not the correct translation. It should be translated to Mukhya Prash. So these kind of use cases wherein the translation which is being given by the model is correct but you would want a specific different translation for a specific word or phrases that can also be handled through glossary.

So, the way that we have done this is that we have website. So, we have a lot of information in the information in the Just now, after we complete this, next slide please. So, this is the future roadmap for plugin that we have. We have expanded it to 36 languages, 36 more Indian languages. So, you can go to Vashni Pavilion which is right here in this hall only. We have a demo of the plugin which is available in 36 languages. We are also incorporating the 35 international languages. We have done that for certain use cases which are displayed here today at Bhaat Mandapam. Secondly, we will be talking about glossary but the glossary in, you know, traditionally the glossaries were sent to us through emails and there was a process to, you know, process the glossaries and then ingest it.

But now, we are also planning to get it automated wherein, you can just simply upload the glossary from your onboarding portal. and third, we are also adding the accessibility bar to the plugin. So if you want to have text -to -speech also integrated or screen reader also integrated with the plugin that we just showed, that is also something that we are going to do in some time. So technology for dignity, Bajni Translation plugin would help. It is a powerful tool that will empower you to actually disseminate whatever information you want to, to actually reach the last mile. Moving on to the next segment, which is the glossary. So, you know, we all of us here, we would have some application, some website developed for…

the ease of the user. We would want a person, a student who is registering for a form who can actually do we would want the person to do it in their own preferred language. We would want a farmer to listen to the schemes that are available for him in his preferred language. We would want an Angadwari worker to have the schemes that are available for her told to her in her own language. So that is all what we are working for. We are working for inclusivity and we are working for accessibility. Next slide please. So while we do that we also add Bhashni’s layer to all our solutions or websites to have the actual information reach the last mile.

But generally what happens is that you know we get a remark that the translation is not correct. It is wrong. And after doing analysis with most of our customers we realize that the audience, that the users who are trying to actually use our product, they are not looking for accurate translations. They are looking for understanding the content, the intent of the content which is there on the solution on the website that they have created. And this is not the result, you know, for this we don’t have to focus on getting the accuracy of the translation. We actually have to focus on the context of the translation, use case of the translation, domain of the translation.

So when we realize that, we understood the concept of glossary and that’s how glossary was formed. Next slide please. Now you all would be, you know, waiting for, to understand what glossary is all about. So glossary saw… It involves two kinds of use cases. One is the post -translation that I just told you before. That, you know, home being translated to ghar in Hindi is absolutely right. But home being translated to home tab being translated to ghar is probably not correct. So post translation wherein you would want home to appear as Mukhya Prasht on the tab, home tab, that is something that we cater through with glossary. The second use case is like in the example, there is a bridge called Vakil Saab Bridge in Gujarat.

So Vakil Saab Bridge if translated to English would become something like Lawyer Bridge or something. We wouldn’t want that. Vakil Saab Bridge is our coined terminology and we would want it to retain its identity. We would want Vakil Saab Bridge to be written as Vakil Saab Bridge only in English. And this is the use case of transliteration. So these two kind of use cases are solved through glossary. What we do is we create. We create these glossaries with our customers and we ingest it to the customer’s specific API. Next slide, please. So, you know, like I told you the meaning of glossary, all of us here have different glossaries. Like, you know, the science domain glossaries are different.

Gen Z has a different glossary altogether. You know, any region would have a specific kind of a glossary. So all of these glossaries have to be created with us. And, you know, the customers who created those glossaries have got the translation, which are accepted by the end user, which are understood by the end user. Like you can see, Ministry of Panchayati Raj gave us 15 lakh words of the Panchayat. Survey of India has given us 16 lakh words. So if we create glossaries together, we can have the translation barrier completely eliminated. So I will now walk you through certain. Use cases wherein we faced problems with our customers, but they were not. translation issues, they were actually issues that could have been easily resolved through glossary.

So if you can read this sentence here. So this use case, you know, this problem was reported to us by Ministry of Home Affairs where Honourable Home Minister Sir’s profile was not reflecting correctly. So this was the English sentence. Okay. And this was the translation that we were getting. So if anybody can tell me what is the problem here? So because of this full stop, Srimati full stop, SMT full stop, what the model thought was that the sentence has ended here. And that is why the formation of the sentence is entirely incorrect. But the solution was very simple. What we had to do was just add SMT dot to the glossary or just remove the dot from the SMT.

And we could have the correct output. So it’s as simple as that. It’s not the translations problem. It’s the understanding of glossary problem. Next slide, please. So, okay. Can anybody tell me what’s the difference between this and this puzzle? To these two puzzle pieces, what is the difference? Yes. So one of them has a hyphen and the other one does not have a hyphen. So when we received the glossary from MSME, there was a hyphen in between PMS and dashboard. but actually on the website it was displayed without the hyphen. So glossary is that sensitive. If you give me PMS hyphen dashboard, it will only recognize that and translate that. But if there is no hyphen, it will not recognize that and it will not give you the translated output which you have given us in glossary.

So that’s the, and again, here there was a singular and plural problem. So street vendors was mentioned in the glossary sheet that we received. But actually street vendor was mentioned on the website. So if there is a singular and plural difference in the glossary sheet that you are giving to us and what is actually reflected on the website or your solution, it will create a difference. So this is one thing that you can also do through glossary. So, you know, we received. A requirement wherein they wanted, you know, the animal husbandry department wanted that the entire sentence should not be translated. the, you know, abbreviations should be skipped from translation. So if you just give me this sentence and this sentence as glossary pairs in English and Hindi, this can easily be achieved.

Next slide, please. Okay, so in one of the glossaries that we received was authorized officer, so they wanted us to write authorized officer as newt adhikari. But actually, newt adhikari means appointed officer. So this is also something that we have to be careful of. Because, you know, for the end user, so there are two kind of users that we have in this case, the English users and the Hindi users. So the English user would read it as authorized officer, but since we have added glossary and changed it to, newt adhikari in Hindi, the Hindi user would read it, would understand it as appointed officer. So we have to be very careful while drafting the glossaries.

Next slide please. Okay. So if I can just ask what is the full form of BN? Normally what do we consider as the full form of BN? Billion, right? We would not consider BN as battalion. But in BSF’s case, this was a huge problem. So BN for BSF means battalion and not billion. The entire context changes. So for BSF, we have created glossaries for all the abbreviations. So it is always suggested that you know whatever abbreviations that you are displaying on your website or your solution, just give it to us as glossary so that the correct one can be displayed. So okay. So, So can you tell me if PS to Minister being translated as Maanenya Vastra Mantri Ji ki Iji Sacheev, is there a difference or like what would be the problem here?

Fine, let me tell you. So this is also correct, this is also correct. But this is not the actual translation of PS to Minister. If we want to have Maanenya written in the Hindi translation of it, we should always have it in the English version of it also. Glossaries are supposed to be equal. They are supposed to be equally weighted. You cannot expect the model to add or delete words as their own. So basically what we did here was we went back to the customer and said, that if you want to add Maanenya, you can add Maanenya to the text. at the output, please add respected or honourable in the input. Only then it will be balanced out.

Next slide, please. So this is one request from our end only. We receive a lot of glossaries that are redundant in nature for us. By that I mean that, you know, for example, we received employment and skill development as the glossary terminology and the translation that we are getting in Hindi. That was the actual output of the model also. So in this case, if you are giving us the glossary, which is actually the output of the model, you are only creating redundancy. So if you can just avoid that and give us translations, give us the post -translations or transliterations that are not recognized by the model, that would be handy. Next slide please. Next slide.

So in the end I would just like to say that language is not just words, it is identity. Let us prepare India’s languages for the future of AI and let us create glossaries, let us have multilingual AI in, multilingual layer in all your solutions so that actually the end user is benefited, actually there is digital inclusion, accessibility, inclusivity. Thank you. Any question?

Can you hear me? So I was saying like the translation thing which you were showing, is it, I know like this has been sponsored by the government and stuff, so can it be also used for commercial purpose like for private or private? Public entity? Can they also use that in their websites also?

So, you know, we have different kind of collaborations with us. So for that collaboration, there is a different agreement altogether that is being created. If you want to know more about it, just go to the Bhajani Pavilion. We have stakeholders who are handling the startups, the private organizations also, and they can help you there.

And one more thing which I wanted to know. So like you were showing for the websites, it was by default we can choose the default language, right? So can it be also extended? Like let’s say in some use cases, we could have someone who is logging in from Delhi. They would want to see it in Hindi and someone who is coming from Maharashtra. So can it change the default languages? Can it change from the region perspective?

So that’s an interesting use case. From what I’ve understood, you want different regions to have websites opened in different default languages. As per my knowledge, I don’t see a technical challenge to it. But again, we will have to look at the use case at our end and see if this can be deferred. It’s a very use case. This is a very interesting use case. We’ll look at it. Thank you.

Hi. Hi. So we all are aware that we have multilingual languages. And apparently, they have been trained on a lot of words also according to their domain knowledge. So if we have glossaries, how do we ensure that each and every glossary according to the domain is maintained and then trained or fine -tuned?

So glossaries are customized. For example, somebody from Ministry of Home Affairs would not want the glossary of, let’s say, CSI. Right. Right. you know the domains are different the contexts are different so glossaries are ingested only are customized and are ingested to the client itself they are not we have general glossaries also that can be applied to all but since glossary does not have the you know one glossary fits all type of a solution so we customize it for a client and then ingest it on to that client solution itself not to other clients or other environments.

Thanks So the glossaries you have do you have them do you use them to fine tune your models or is it just available as documents to infer while using?

So we do that we try to fine tune the models as well but we there are lot of things that we have to look at it look around while doing that because you know we have to classify them into different domains and then apply fine tuning models for the domain space. So it’s a long process, but we do that. Okay. Thank you. If there are any other questions, I will be available at the Bhashni Pavilion here also. And I would request everybody to please come visit us, explore our solutions, explore our services. And thank you so much for being a lovely audience. Thank you.

That was really an interesting session by CEO Cisco, highlighting the agentic AI, the role of agentic AI, as well as the physical AI and the current scenario. And also the last line was really an assuring line saying that the future will not be built by AI, but by humans who can confidently put AI to use. Well, ladies and gentlemen, moving on. Now it’s my honor to introduce a leader who’s been at the forefront of shaping the future of wireless technology and intelligent computing. Mr. Cristiano Amon is the president and chief executive officer of Qualcomm, a company that has defined and continues to redefine the global compute connectivity and AI landscape. And well, AI doesn’t just live in the cloud, it runs in your pocket, in your car, in the factory floors.

And Mr. Amon is leading Qualcomm’s push to bring powerful AI processing to the edge. enabling billions of devices to think locally and act intelligently. Ladies and gentlemen, it’s my pleasure to invite Mr. Amon, President and CEO of Qualcomm, to the stage. Please give a round of applause.

Good afternoon, everyone. Very, very happy and privileged to be here. I’m incredibly excited and energized about what’s happening here in India with AI and I think what’s happening with AI in general. What I’d like to talk to you today is about the next chapter of AI. And this is something that’s very near and dear to Qualcomm. We’ve been talking about this because I think we’re really entering now the next phase of AI. As AI gets developed, it’s going to be part of everything that we do. And especially… the interaction that we have with computers and with digital… So intelligent is now shifting for something that we kind of started and we all experience going to, you know, a chat box and asking questions into something that is going to be all around us and everywhere all the time, especially with the devices.

I actually love the presentation right before from my friend Jitu from Cisco when he talked about the traffic change from chat box to agents. And this is important. You know, I’ve been often talking about this, how we should be thinking about AI in a much broader sense. And it’s easier for a company like Qualcomm to talk about this because we build a lot of the chips that go into devices where the humans are. So as you create AI in the data center and you train and create those models, all this data. And you deploy this, you’re starting to see that this gets utilized in different ways. One fundamental thing that AI is doing for us.

it is changing the human computer interface because we don’t have to now learn how to use a computer if you know i’ve been uh often talking about this in different presentations we learn how to use an s2 keyboard and we still use that on a laptop then we use like to touch a screen but now the ai understands what we see what we hear what we say what we write so in itself it’s changing computers it’s changing the devices we interact with and uh it’s becoming a pervasive technology that is going to be everywhere and i think that’s the mission i think of qualcomm when i think about uh what we’re going to do is the same way that what we did with mobile communications and the creation of of the computer that fits in the palm of your hand is the ability to take that intelligence everywhere so we’re going to be creating a number of important shifts in the industry and i want to start talking about the mobile industry we may have had the privilege as a company to be part of every single transition of wireless technologies and let’s talk today I’m going to talk about the next one that is coming as well and what we saw with the transition of wireless technology that fundamentally at every generation of wireless you saw big shifts not only in devices and companies and because of the transition especially for example when you went to the ability to have a phone that you carry with you all the way to connect the phone or the internet all of a sudden that phone became a computer and it started to drive the future of the internet like a country like India that leapfrogged I think the internet and went straight to the mobile internet and that’s going to be true again when you think about AI for example in the mobile ecosystem AI is constantly changing and it’s changing and it’s changing and it’s changing and it’s changing going to fundamentally change how we think about the mobile device All of you today, and me included, I think we look at our smartphone, our inseparable device, most of our digital life is.

And the smartphone today is at the center of everything that we do. But now that’s going to get replaced by an agent. Now, when you think about the entire value chain that got created, for example, for the mobile industry, there’s an enormous amount of value on things like OSs and application stores. And that becomes like the platform when you’re going to develop an application that you’re going to do different things into the platform. An agent that now understands human intentions because, you know, you just need to tell him what you want. Or he’s going to see what you see and make a decision for you, assuming you will authorize it. it. When that happens, that’s where the value is because then the agent is free.

It can go to the internet and do things. It can go to your phone and do things. And you’re no longer bound by constructs of your hardware or your apps in the application. So as a result, we expect the AI is going to have a fundamental shift in the mobile industry where the agent is going to be at the very center. And as the agent is at the very center, everything surrounds the agent. You can access the agent from your mobile phone, but you can also access the agent from your glasses or for a pendant or for anything that you wear. And I think we’re going to look at the mobile ecosystem right now, not only as a single device experience, but you’re going to connect to agents across multiple types of devices.

And I think that’s incredibly exciting. And that’s not only unique to what you’re going to see in consumers. That’s going to happen also with things, because you can also have create AI that’s going to get trained on different things. on physical signals, like physical AI, on sensor data, and you’re going to deploy that in every computer. So what’s exciting about AI, it’s going to very quickly evolve for something you go to a browser and you ask a question. And I think, as my colleague from Cisco said, it’s got train and all the public available data on the Internet. You’re now going to go to a different type of AI experience that’s going to be the fundamental software that is going to run in all the devices around us and how you’re going to have interaction with the devices.

So I also want to basically, you know, as we think about this future, I just want to give you an example. What we saw across the industry is workloads or use cases have shifted. Devices didn’t go anywhere, but their workloads shifted. We used to do a lot of things in the early days of the Internet on your laptop. And forget. For example, e -commerce, you will do it on your laptop. Now, most of the e -commerce in the world is done on a phone. Tomorrow, or it could be like as early as, you know, within the end of this year, as you start to see the proliferation of glasses. If you have a glass that has agents, is connected to the Internet, has camera on those smart glasses, the glass see what you see.

You can just look at something and say, I’d like to buy this. What is, you know, can you check this? For example, check this on Flipkart. Just buy it for me. I’d like to buy this. Integration of payment system. You got a bill, say, pay this, notify me when I’m done, and so forth. So I think we’re going to see this fundamental change of devices. But that’s also going to be true about the revolution that’s happening in robotics and the revolution that exactly happened on industrials. So that’s an incredible opportunity. And we have been incredible. Incredibly focused as a company to basically drive that future of computing. There’s also a big debate, which I believe is the wrong way to look into that, which is about cloud and edge.

There’s a lot of debate about, oh, this is going to be running on the cloud. This is going to be running on the edge. And actually, it does not matter. Think about your device today. Your smartphone today has incredible amount of processing power, and there’s a number of different things that run in your smartphone. If you put it on airplane mode, you probably don’t use it. You just put it back and wait until you get connectivity again. It’s the most cloud -connected device because those things work as a one. And you’re going to have now intelligence that’s going to be incredibly distributed across the cloud, across the near edge, the network in itself, in and on device.

And it’s all going to work similar. There are going to be things that you’re going to be able to do on the device because they’re… They require an instant response or require unique context, unique information that is relevant to you. Something is going to do on the cloud and they’re both going to be growing and it’s going to be transforming how we think about computers. So I like to provide the simple, I think, a description. Let’s say we are all using agents and you’re going to pick the agents that you like and the agents to be useful. It needs to be fast. It needs to be relevant for you. Let’s say, go back to the example I provided on the glasses.

And you have those smart glasses and you’re walking around and you have a camera. Then all of a sudden you see somebody and you ask this glass, like it’s your friend next to you and say, who is this person? And you want to get a response. This is so and so. Or you’re going to say, can you translate this for me? What is this? Can you pay this for me? You want to, this thing has to be similar. Similar is no friction. So certain things are going to be done on your device and the thing’s going to be on the cloud. It’s going to be completely transparent to you. But the interesting thing is those agents, for them to be very useful, they needed to be contextually aware of what is relevant to you.

So over time, the agent I’m going to be using, the agent you’re going to be using, they need to be relevant to me. So you’re going to have a lot of things that are going to be being processed and understood about you. So much so that I believe that in the end game, I think it was said in the prior presentation from Cisco that all this available data that is publicly on the Internet that you train models, it’s a fraction of the data that is going to be generated. If you have, for example, a glass of a camera that sees everything that you see, try to annotate the image, get information about the image and the context, reads what you read.

And so forth, that is an incredible amount of data, and that’s going to be providing a lot of important context for those models that are going to be relevant to you. That is the future, and it’s an incredible transformation. It’s going to transform every industry. No industry is immune to this. And I think what we’re doing at Qualcomm is really creating the future hardware and software that will help enable this future across all the devices. We’re a very unique semiconductor company. I think we’re probably one of the few companies that can be working on chips from sub -2 milliwatts to a smart earbud that you’re going to wear all the way to now 2 ,000 watts per chip on the data center.

But I think that’s the incredible future that AI is going to transform every single computer. And the agents are going to be at the center of the experience. It’s going to replace a lot of the OSs and applications. And that is the new future of technology, including the future of mobility. And that’s why we’re incredibly excited about this. And with that, I want to talk about something that is happening, which is about the next generation of wireless technologies. I would like to provide an example from the past. When you think about telecom networks, and I think we’re probably one of the, you know, American telecom companies that really focus on the evolution of cellular technology.

When you think about the evolution of this sector, when this all started, it was about providing a telephone. I think all of us was an incredible thing. You have a twisted copper pair to get to your home. You pick up. You get a dial tone. You dial. And eventually, you could dial. Anybody in the world of a telephone. Even how cellular started was about making sure all of us had the ability to carry a telephone. That was 2G, that you can call everyone. That’s different today. Now you have a very high performance broadband network for data. Voice is just one application in the many applications that you do with the network. It fundamentally changed the nature of the infrastructure.

The equipment was different. The use case is different. We’re heading to the next big transformation of the telecom sector. So 6G is going to provide an evolution of connectivity, faster speed, lower latency, higher coverage. But that’s not the story. That’s just a piece of the story. It’s just continue to improve the connectivity. The biggest part of 6G is AI, like I said before, is now going to come to the telecom network. And that becomes a large scale. 6G. AI network that is processing and get trained on all of the signals that happens at the network and providing new capabilities. One of the biggest features of 6G is the network, is the sensing network at scale.

I’m going to give an example. The network not only will provide a connectivity between your device and the Internet, but will sense everything that’s around you. We’ll use techniques that you see today in autonomous driving cars, like radars, as an example, to detect your environment. It’s going to provide a map of everything that is happening at scale. And you’re going to have completely different type of services for different industries. It will provide context for your agents. Very important. And the network will have that role. It will provide traffic management systems and some of the use cases that are going to be part of full self -driving cars. It will do drone detection and manage the traffic control.

Off the economy, there’s going to be an aerial in the wide area network and much more. because AI is also going to the network. It’s going to be one of the biggest transitions I think we have, as big as going from voice to data, and it’s all going to be part of this future of AI. And I just want to now make another parallel, I think, to the presentation from my colleague from Cisco. It puts a fine point on the network that needs to be built, the capability of the infrastructure, the security and trust, but that is an incredible future with technology. And as I get to the end of the presentation, I want to highlight that India has an incredible opportunity with this transformation.

We have seen that those big shifts in technology creates opportunity, change players. It changed, I think, the role of different countries as they provide globally. It’s a global scale for the technology, and that’s an incredible opportunity for India. I look of what happened in mobile in India, and one of the largest data consumption per user in mobile devices in the world is in India. The whole Internet is mobile. When you think about the potential and all of the things that I just discussed about how AI is going to change everything, creates new device, new experiences, new services, that becomes a massive opportunity. And when I look at the ambitions that were set by the AI Summit, I’m going to provide just some examples.

Those are just examples. It can be very broader, but I just want to connect with some of the ambitions of the Summit. There is a process of jumping into a large -scale industrialization. India is becoming a global manufacturing hub as well. And with AI, you… You go from the very beginning. with smart manufacturing and automation with incredible change that is happening in this sector enabled by those technologies. Same thing with smart cities, the ability to continue to evolve the infrastructure, the ability to use AI to increase the scale, the reach, the access for healthcare. How you change education. Those are incredibly powerful learning tools. The ability to actually use some of those technologies to empower people with information and you’re going to have an ongoing learning experience.

Think about those agents with you all the time answering questions, telling you how to do things, especially when you think of context, for example, of those new devices such as smart glasses. And it can fundamentally change industries, for example, such as agriculture. Right. Right. Right. Just a few examples of the potential of connecting this technology with everything, I think, that is going on in India. It’s an incredible and exciting future enabled by AI. And really, it’s about meeting the ambition of democratizing this technology for everyone and actually have an important role in increasing the global welfare. And, you know, as a company that has always been focused on enabling our partners and other industries to innovate, I think the history of Qualcomm, we never believe is the job of one company to be responsible for all the innovation.

It’s really to enable many industries and partner. We’re incredibly excited to play a very small part on this mission. Thank you very much for the opportunity to talk with all of you and

Well, ladies and gentlemen, now it’s my pleasure to invite our next speaker, Mr. Giordano Albertazzi, who is the chief executive officer of Vertiv, a global company that provides digital infrastructure solutions for data centers, communication networks. Under his leadership, Vertiv is advancing its role as a global industry leader by accelerating innovation, strengthening technology leadership, and enabling the digital infrastructure that powers critical applications worldwide. Ladies and gentlemen, please welcome Mr. Giordano Albertazzi.

Thank you very much. The clicker? Oh, yeah, here. Better with the clicker. Good afternoon, everyone. And it’s absolutely a pleasure and an honor being on this stage where so many distinguished presenters. In the last two days, I’ve had the opportunity to talk about AI. An astonishing thing to me is that the majority of the AI conversations, as it should be, are about what AI can do. Very interesting presentation, just finished, tells about all the beautiful things that AI can do and particularly what AI can do here in India. But when we talk AI, we also talk about data centers. But let me go then to the physical part of AI, not just what AI can do for us.

There is an important, very important physical part of AI that sometimes is overlooked. And it shouldn’t, because it’s that physical part that makes AI actually possible. So I’ll talk about the physical part today. I will talk about the power. The cooling, the data center infrastructure. Vertiv and myself, with Vertiv, have been in the industry for decades. Well, Vertiv longer than me. It used to be part of Emerson, Emerson Electric, and we are almost 10 years as an independent company now publicly traded in New York. But what we do is really make sure that that physical part is provided with the best technology that supports the continuous evolution of the AI IT stack as those rapidly, almost exponentially, and I’m talking almost exponentially from a mathematical standpoint, evolve.

And it’s no easy task, a task that we would do very well because we know the space a lot. We have a lot of innovation. But there are several dimensions. to this. One is the extreme densification. Now, we all know what GPUs are. Probably two years ago, majority of people didn’t have any clue about what a GPU is. But now, GPU, NVIDIA is absolutely central to everything, all the conversation about AI. Well, that phenomenal evolution from a technology standpoint is changing the DNA of a data center. What used to be a rack with IT inside 10, 15, 20 kilowatt per rack is rapidly becoming more dense and with more power and heat to dissipate in it. This is going to 30, 50, 150 kilowatts per rack all the way in the possible future.

One megawatt per rack. That’s a lot of power in a single rack. The design of a data center is changing dramatically. As this design changes, of course, also the technology that supports it needs to change. But let me go back to AI, artificial intelligence. Let me go back to, and let me draw a parallel. Human intelligence. Human intelligence happen in the brain. But the brain doesn’t survive without a body. What we are, what we do at Vertiv, make that body, provide that technology for that body so that the brain can function, and that brain is the IT stack. But not only that brain can function, but also can produce intelligence. And that’s what an AI does.

That’s what an AI factory, an AI data center is doing. But just like the body, historically, data centers and data center engineering was viewed as disparate systems coming together. Now, we cannot think about a human body or anybody as individual parts, a chiller, a liquid cooling unit, an interruptible power supply, or whatever else in the powertrain or thermal chain you can think of. Everything needs to be orchestrated. Everything needs to be interoperable. Everything must be thought of as one thing. And that’s what we do in a world that is extraordinarily challenging, but it’s a challenge that we, of course, respond to very successfully, challenging in terms of time of deployment and in terms of scale of deployment.

Okay. Data centers need to be. developed faster and faster and are becoming bigger and bigger. You heard that. It is about data centers. India is certainly privileged from an AI standpoint also because there is a lot of power available that can be harnessed for more and more powerful and larger data centers. Now, as that happens, again, if you think, go back to my analogy of the body, you think about a system, you think about everything that is the body of artificial intelligence, then it is about changing the way we build that body from one piece at a time with a lot of activity happening on site, laborious, hard from a quality standpoint, to most integrated at factory level and deployment.

As NVIDIA continues to lead the world in terms of technology, in terms of IT stack, but also in terms of thought process for the infrastructure. And it’s something that we do a lot together. Well, then it is not just about the infrastructure and the speed and the size and the scale. It’s also about optimizing the infrastructure with reference designs that exactly target that type of application. So we, of course, are thrilled and always honored to partner with NVIDIA in this adventure and venture and lead the market in this respect. So here you have an example of what we call one vertigo, one core. There’s an example of a fully pre -engineered, defined data center. But when we talk about the body, the body of AI, the data center, then let’s talk very simply.

We talk about three fundamental, fundamental elements of that body. One is the powertrain. So everything that goes from the grid, if you will, for your utility, takes that power all the way to the chip. That power infrastructure is changing, is evolving as the power density changes. And the current architectures are migrating towards, over time, what is an 800 -volt DC power infrastructure. I’m going technical on you. Some of you I know are very technical, so I’m not afraid about that. But I will not go deep. So everything you see on the left side of this is exactly a representation of that powertrain. So bring the energy, take that energy to the chip. Then the chip and all the electronic components in a server generate heat.

And that heat can be very dense. And require very, very advanced… cooling mechanisms. And that’s the beginning of what we like to call a thermal chain that starts, and it’s what you see on the right side of this chart from the chip all the way to the heat extraction, the heat then rejection, or even more importantly, and more extensively so, is the heat reuse. So this is the system, the fundamental systems of this body. But again, it’s not just about the components of the system, it’s how the entire system works. And more and more, we see that when we think about the AI IT infrastructure, what used to be thought of as a server at a time is becoming really an AI, AI.

pod, an AI unit at a time. The unit of compute is no more the server. It is the pod. Unit of compute is not even the pod. It’s the entire data center operating as one single computer. A unit of compute that can go all the way to gigawatts. So it is about making sure, and I believe we do it very well, that I say uniquely well, but of course I root for ourselves. It is about making that infrastructure available at scale and in a very easy modular to deploy fashion. And that’s what we do. So a repeatable converged infrastructure major. So we have a lot of building blocks that can go from a 12 .5 megawatts all the way to gigawatts, all the way to gigawatts.

So clearly it’s not just about building that infrastructure, but that infrastructure over time needs to be, like we like to say, future resilient. Some people, like myself, have been in the industry of data center for quite some time, and it’s fascinating the speed at which things happen. And this speed is also enabled by new solutions that make prefabricated and very fast to deploy part of data centers that used to be very, very laborious. Take a data. It’s empty when the building is new. You have to fill it with power, with cooling, with cables. You have to put the racks. very laborious and time consuming. Time to token is of the essence. A prefabrication, for example, with what we do with Verti Smart Run, reduces the time to deploy almost 85%, almost an order of magnitude.

So the industry is changing, not only in scale and in density, but also in the way things are done and deployed. Let me take a different angle now and focus on India. India clearly central to the AI evolution revolution and central certainly in terms of the infrastructure that is being built and the infrastructure that will be built in the future and in the coming years. This infrastructure and the speed at which this infrastructure will be built, of course, will depend, as I was saying, by the ability of the likes of Vertiv, but certainly Vertiv, given our prominent position also in India, to really enable this at scale and at speed in the ways that I explained.

So Vertiv in India has a long tradition. We’ve been here for decades. We have what I believe is an awesome team and awesome partnerships. And now this forum, these sessions, these few days convinced me even more of the importance of India as a place. A place to invest. And invest we will. We are expanding our capacity and will continue to expand capacity. We see India certainly as an extremely promising market. as a hub for AI, not only for India, but globally. So it has got the power availability. Certainly India has got the right demographics. So I couldn’t be more excited about the business in India. I couldn’t be more excited about what we’re doing in India and what our partners are doing in India.

So with that, I’m extremely optimistic. I’m a big optimist about what AI will bring, as we heard. And with that, thank you very much. Thank you.

Thank you so much, Mr. El -Battazi, for your impactful address and also for…

Data centers have, up until now, been usually constructed in one of two ways. Traditional data center build follows a sequential process, materials and equipment arriving individually on site, with the build progressing from the ground up. Alternatively, prefabricated modular construction can offer many advantages, such as quicker deployments and risk reduction. Vertiv offers many solutions in this space. However, in the age of increasing IT loads powered by artificial intelligence, there’s another option that combines the advantages of both. Vertiv OneCore. Vertiv power and thermal infrastructure building blocks are inserted into a brand new Vertiv -supplied steel building shell, or an existing building. Infrastructure building blocks are made in controlled factory environments and tested before construction. The system is also equipped with a new, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient, more efficient,

Very exciting sessions. I’ll just wait. Guys. So we are meeting for a tremendously important subject. And this has been a great summit. I know you’re all energized, inspired, excited and exhausted at the same time. And we will get a moment when this subject becomes a room of 5 ,000. So that’s what we’re going to work towards. But you’re here today. We’re delighted to have an absolutely tremendous panel with us today. I’m deeply honored, flown across from the U .S., from Europe, from Singapore and so forth. And we have a lot of material to cover. I should say that the triple challenge that we are dealing with in this panel is perhaps the most important challenge any of us will face in our lives.

Which is to promote development on the one side. While dealing with the creation of a sustainable planet. and in terms of climate change, your self -selecting group, you’re all here with us and there’s a reason for that. You already know about climate change. You already know about AI. Is that, as you know, we are not necessarily winning the battle on climate as yet and so we need to deal with both mitigation and adaptation and this panel will address both of those two things. We have very little time in the panel so I am going to speed along but that’s a good metaphor for the very little time we have to do something about climate change.

So we’re in action mode. It’s a call for collaboration. We’re not going to be, I apologize to our speakers and panelists for a number of things. One, this is not a real panel. We’re not going to be having discussions. This is just boom, boom, boom, talking to you about what everyone’s doing. Secondly, there’s going to be a kind of switcheroo moment when some other speakers come up and some of us are replaced up here. Apologies for that. It’s just the intensity of the panel and for all those things I apologize in advance but I don’t apologize. for the incredible quality of our panelists today. These are amazing people. And I would just end by saying that this is not a normal session.

This is an invitation for radical action -oriented collaboration with all of you. On that basis, let me begin by talking a little bit about a summit we held last year in London and the background to it and an organization that some of us are very deeply involved with and almost everyone here is a friend of called the Green Artificial Intelligence Learning Network. You will immediately note it has the cutesy acronym of GRAIL, like the Holy Grail. And what we’re trying to do is really see what the synergy is between the development agenda and the climate agenda through the application of AI. I’m going to speed through this. We’re going to then move to Professor David Sandelow, who actually anchored our summit last year, which is the first major global summit on the application of AI, to climate change.

and has very kindly flown in from Colombia. I’m going to ask our speakers one more favor, which is instead of my coming up and introducing all of you, if you don’t mind introducing yourselves, that will speed us along the way. So let’s go through this. So as you all know, and perhaps Professor Sandler will talk about it, the IPCC gave us a target, 43 % decarbonization from 2019 levels to 2030 levels. We were meant to reduce GHG emissions by that amount. In 2021, some of us at COP26 in Glasgow had a meeting to look at the likelihood of that happening. And we came to the conclusion that the likelihood was very low. And therefore, traditional approaches to climate mitigation and adaptation needed to be enhanced with new solutions.

And we thought, what was J -curving as fast as climate change was J -curving? And the only thing… There was nothing we could think of. There was nothing we could think of. There was nothing we could think of. was the application of AI, this great new suite of technologies, including, of course, quantum and all the other things that go with it. And we started to talk to people, and we talked to a whole bunch of people in the AI community, a whole bunch of people in the industrial and power, automotive, all the different sectors that produce emissions. And we said, are you talking to each other? And shockingly, people were not talking to each other.

There’s very little going on with some honorable exceptions at Google. Very few people were really in the AI community focusing on downstream issues around climate change. And similarly, the big industrial domains were not really focused on the use of AI for decarbonization and economic value creation. So with that lens, think about this session as throwing one J curve against another J curve. Can we throw the crazy increase in AI technology represented by this great summit against the world’s greatest challenge? That’s the purpose of the Grail organization, which is a not -for -profit based in London. It’s a vast terrain. We don’t have time to cover it all. It’s obviously mitigation. It’s obviously adaptation. We have to hit both.

And within that, there are endless taxonomies of all the wonderful things that AI can do. And, of course, you’ll be worried about the increased Gs from data centers, but that’s not the primary focus of our session today. That has been quantified. The Grantham Institute did a quantification last year of 0 .5 to 1 .4 gigatons of extra GHGs from data centers, but that’s from every kind of utilization against the potential benefit of 3 .5 to 5 .4 gigatons being sucked out of extra GHG emissions. So there’s clearly a very strong balance towards what AI can do to helping the planet in its shift towards a clean and green economy. Grail, what’s Grail? Grail is an attempt to create… …to create a collaborative network.

of great academic institutions, commercial institutions, AI companies, industrial companies, philanthropic institutions, private sector sustainability networks like WBCSD, bringing them all together with governments to try and create massive collaboration. In the next slide at the bottom, you see that same group of institutions. Bottom right, the ideas and deal flow. Going back into Grail, bottom left, the fact that this becomes a collaborative community to get all these solutions scaling at speed and at the top, then getting that deal flow funded through grants, through government programs, through venture capital, corporate funds, but to move the agenda to real solutions at massive scale as quickly as we possibly can. All of this led to a summit that occurred that I mentioned earlier last year.

Sean, will you keep me real on the time? Thank you. Okay. And that led to… 200 people. 115 organizations, including all the organizations represented here today, 60 speakers, and we looked at AI for power, AI for building materials, AI for everything you could think of vertically and horizontally, looking at the issues of materials innovation, looking at the issues of value chains, looking at carbon markets, and so forth. What has happened after the summit? Three things. One, we’ve created an online collaborative platform, and we invite all of you to join it, to co -create those solutions that can make a difference. Second, we’ve started to engage with governments around the world. Imagine a summit like this that was focused, yes, on development, but with a central climate focus as well.

How amazing would that be? And most importantly, we’re focusing on taxonomies that lead to massive calls for action from the innovation community. So we started to work on taxonomies for a variety of sectors, the energy sector, the built environment, materials innovation, and we worked with groups of AI experts and power experts and figured out what the big wins were, what are the big opportunities for companies to create economic value while at the same time massively decarbonizing. And this was an intellectual process, including many experts, some of whom are here today, that led to this astonishing work and identified the big win -win opportunities for economic value creation and decarbonization. On the bottom right, the teams from academia, industry, industry associations, a variety of other people and countries, eight country teams as well were involved in various ways.

So where’s this all going? Well, thank you to McKinsey for your very kind collaboration to after we had done all that work saying, hey, we want to help and kicking in and working with us to further refine those offerings and look at cost benefits and cost curves and all sorts of things. Delighted about it. And then there are, apart from working on the power sector, to look at generally what we can do, apart from working on the built environment, generally what we can do, apart from looking at materials innovation, generally what we can do to accelerate solutions for decarbonization through AI. We have two big partnerships that are emerging. One, and I want to slow down here.

Okay, 250 companies are part of the World Business Council, their network. That represents, in scope one, two, and three, 26 % of World GHGs. They’ve realized that that’s mainly in supply chains who are going into a partnership, so is McKinsey, so are other partners, to look at what are the AI opportunities to take startup and scale -ups into these decarbonization opportunities at massive scale with the 250 largest companies in the world representing 24 % of world revenues and 26 % of World GHGs. Finally, working with coalitions of energy companies, and Nalin, you’ll hear, and we’ll hear from later on, and we’re deeply partnered with Nalin. on this and others on this. How can we take this into accelerating? For example, UNESA has 71 energy companies, 750 gigawatts of clean power.

They want to go to 1 ,500 gigawatts of clean power by the end of the decade. How can we help them with AI? It’s a very practical lens. We invite you to join us and be part of this. On that note, I would like to briefly to invite to introduce Professor David Sandlow. David, I’m not going to go through your very distinguished background to take the whole panel to do that, except to say that you have worked in every different field, most importantly in the past in very senior positions in government, but now, of course, you’ve been the luminary on AI solutions on climate. That is the worst introduction you’ve ever got. I’m sorry about that, just in the interest of time, but I’m…

Really very honored that you’ve flown all the way to come here, and I’m handing them over to you. to you.

Thank you so much. Uday. Uday, thank you. Your energy, your enthusiasm, your passion, they’re all infectious. And your intellect is remarkable. What you’re driving forward in this area is world changing. You are not just an inspiration, you’re a gravitational force that are pulling people together to work on this, so thank you so much. Your, what you did in London was remarkable. What you’re doing here is incredible and I’m looking forward to being part of what you’re doing in the future. So I’ve, it’s been my privilege to lead some teams that have been working on these issues over the course of the past couple of years and I’m going to talk about one of the projects that we’ve done.

It looks like the slides are not there. There’s a certain, turning on the screen. There it goes. I will say that while we wait, I’ll say that I really like the metaphor that you had, Uday, about two hockey sticks. And this is just a remarkable convergence of two of the most important trends that are happening in human history right now. One of them is, alas, the increase in greenhouse gas emissions, which is happening at such an astonishing pace. But the second is the exponential growth of the capacities of artificial intelligence. What’s driving me is we need to find a way to make sure that that second trend, artificial intelligence, helps to solve the first problem.

And that’s the study that we did, which we brought together a team of 25 experts. Just wonderful people. One of them was . Song Lee, the head of the IP, last head of the Intergovernmental Panel on Climate Change, and some top experts. And the question we asked was, it was very simple, how do you use artificial intelligence to reduce greenhouse gas emissions? There it is. It came up on the screen. Thank you so much. I appreciate it. And so it’s a very simple question. How do you use AI to reduce emissions of greenhouse gases? We came up with 17 chapters. We wanted to do more than just provide analysis. We wanted to provide actionable ideas for what to do.

So every chapter has recommendations. You can find a print version available on Amazon, and free downloads of the entire volume, including chapter -by -chapter versions, are available at these websites. I want to thank the government of Japan, including NITO and MEDI, for supporting this work. They’ve been very important supporters of work on AI and clean energy more broadly. Oh, I’m going to promote my podcast later. But I have a podcast that’s talking about this topic as well. But so here’s our – Here’s our – Here’s the table of contents for our work. We talked – We have introductions to both AI and climate change in this volume. One of the things we’re trying to do is target this both to experts and to people who are beginners in this topic.

And, you know, Uday talked about bringing together different communities. One of our basic conclusions was we need to bring together experts in climate change and experts in AI. And there are a lot of people who know a lot about climate change but don’t know a lot about AI. A lot of people who know a lot about AI, they don’t know about climate change. So we decided to have primers on each of those topics. And then we talk about eight different sectors and a number of cross -cutting topics. So we have five key takeaways. This was an interesting exercise with all of our authors taking 300 pages and trying to distill it into five key takeaways. But here’s what we came up with.

The first one, I mean, this is a kind of bottom line, but it’s important. AI does have significant potential to contribute to reductions in greenhouse gas emissions. And we categorized it with two categories. One is climate change. It’s incremental gains such as just improving efficiency. It’s output itself. And then we have the other category, which is the environmental impact. solar farms, building energy efficiency. There are lots of incremental gains that can be made, but also transformational gains. In particular, new tech, new materials and other things. We looked at whether greenhouse gas emissions are causing increases in, or greenhouse gas emissions are increasing as a result of computing operations. We decided, based on the available evidence, that the best estimate is less than 1 percent and maybe much less than 1 percent of greenhouse gas emissions are currently coming from AI.

That tracks with the Grantham study that Uday talked about. That tracks with what the IEA has said as well. The main barriers to AI’s impact in reducing greenhouse gas emissions are a lack of data and a lack of trained personnel. There’s other barriers as well, but obviously you need data. A lot of places we don’t have the data for this purpose and you need people. Trust is essential. People aren’t going to use AI unless they trust it. And then every organization with a role in climate change mitigation should consider opportunities for AI. And we need AI to contribute to its work. I think as AI grows in the public… consciousness at summits like this that’s becoming less and less of a kind of radical recommendation.

But it’s just so important. I think if you’re working in climate mitigation, you need to have a team dedicated to AI and how AI can help. So I’ll just run through quickly because we only have a little bit of time, some of our chapters. We have a chapter on the introduction to AI, and if you’re a climate person that doesn’t know a lot about AI, this might be helpful to you. And one of the things we did is we broke down AI capabilities into four basic categories at a very high level. The first thing AI can do is detect patterns. And how can that be helpful in climate change mitigation? Well, one example is detecting methane emissions from satellite data.

You know, some of you probably know this, but I mean, we know much, much more today than we did 10 years ago about methane emissions. And that has helped us dramatically to begin to reduce methane emissions. That’s entirely dependent upon the optical sensing process. And we’ve been able to do that over the last 10 years. So, you know, we’ve been able to do that over the last 10 years. And we’ve been able to do that over the last 10 years. So, you know, we’ve been able to do that over the last 10 years. And we’ve been able to do that over the last 10 years. And we’ve been able to do that over the last 10 years. So, you know, we’ve been able to do that over the last 10 years.

And we’ve been able to do that over the last 10 years. impact so far. AI can also predict, such as weather patterns at solar and wind farms. It can optimize, such as power flows on transmission lines. And it can simulate, such as battery chemistry action. So I think for me, in fact, I’m teaching a course at Columbia right now where we’re emphasizing this framework of detecting, predicting, optimizing, and simulating. And those are, broadly speaking, the capabilities that AI brings to the table. A lot to say about climate change, but just for those who aren’t paying attention, atmospheric concentrations of heat -trapping gases are now higher than any time in human history. In fact, higher than any time in the past three million years.

And July 22nd, 2024, was the warmest day ever recorded. 2024 was the warmest year ever recorded by far. And the warmest 11 years ever recorded were in the last 11 years. So we are living in an era of climate change. We do deep dives into a number of different sectors. I’m just going to talk about a few of them. Power sector is… is maybe the most important just because it’s already 28 % of greenhouse gas emissions, and our strategy for reaching decarbonization requires us to electrify lots of things. So we need to grow the power sector and decarbonize the sector at the same time. I don’t think we’re going to be able to do that without AI tools. AI is already helping decarbonize the power sector, optimizing location of generation transmission, increasing output at solar farms, but it can do much more.

Dynamic line rating is optimal power flow analyses. But to do this, we need standardized data. We need trained personnel. The utility business model is a challenge. So this is a really important area that requires a lot of attention and work. Oh, and a final point, the last bullet on this slide. Using AI in real -time operations can cause real security and safety risks. So we need to be very careful about generative AI in context. So even as we look to deploy AI to help reduce greenhouse gas emissions, we need to be very attentive to these risks. risks. I kind of find it amazing how few people pay attention sometimes to food systems and climate change, that 30 percent or more of greenhouse gases are in some way related to the food system, and the food system has, is threatened by climate change.

AI can do a lot to improve both mitigation and resilience in the food system. We just, there’s a few examples here, integrating data from soil sensors to create fertilizer management plans, creating virtual farms. There’s lots of things that can be done here. But coming back to this issue of lack of data, it’s a huge problem, especially in the Global South. So the efforts to build up a digital public infrastructure that are happening here in India are so important in this regard. I’m going to go quickly here. We look at buildings where there’s tremendous potential. I think materials innovation is one of the most important areas. And, you know, 150 years ago, when Thomas Edison invented the modern light bulb, he literally spent a year, he spent a year, he spent a year, he spent a year, he spent a year, he spent a year, he spent a year, he spent a year, running electricity through dozens, I think hundreds of different filaments to figure out how much light and heat would be produced.

So today, we can simulate a million of those interactions in a second. And there’s already tremendous advances in the pace of innovation in battery chemistry and some other areas using AI tools. And for me, this is one of the most promising areas in terms of transformational gains in reducing greenhouse gas emissions. Extreme weather response is extremely important from a resilient standpoint, and we don’t have a lot of time to get into it, but I think that the AI ML -enabled forecasting is transformational because it’s so much cheaper, for example. I mean, at really 1 ,000x the cost, 1 ,000x less the cost, we can run AI ML weather prediction tools and make a big difference on extreme weather response.

We have findings and recommendations throughout this report. You can see it here, again. We just did a new report in the same series on sustainable data centers. And our main message is there are that with this data center construction boom happening now, this is the time to be paying attention to data and sustainability. We are investing right now in multi -decade assets. We need to be paying attention to this. Smart siting is a key. And finally, here’s a plug for my podcast. It started about a year ago. I’ve had some great guests, Jensen Huang, Dami Lola, Ogunbi, the head of sustainable energy, for all Jennifer Granholm, the U .S. Energy Secretary under Biden. Listen, as they say, available on all major podcast platforms.

Uday, once again, thank

I feel horrified we’ve got speakers of this caliber and so little time. So thank you so much for your leadership. May I invite both of you to speak? You can speak from here if you prefer. We’ve got two great leaders from Google. Obviously, you know, in the sphere. of corporate AI leadership on climate, there is no one that parallels all of you and we look forward to hearing from you your thoughts. Thank you.

Thank you very much for hosting this. I don’t know if that’s a privilege or that’s pressure when you start with that sentence about the leadership position Google has. I just want a quick question. Raise of hands, how many of you have used Google today for either maps or searching something? Thank you. So you know who we are. This is my car. I’m Vrushali Gaud. I’ll introduce myself and then Spencer you can answer that. I lead Google’s in a nutshell decarbonization water and circularity strategy for the company. Essentially what that means is I’m responsible for quite a few things that you had in your slide that we should be doing and a good way to introduce myself is also I like getting things done and so I feel like my inner calling around this is we’ve had a lot of conversations we’ve had a lot of playbooks and research and things But it’s almost like, how do you act on it?

How do you execute on it? How do you start delivering the outcome that I think we all are looking for? So that’s the kind of space that I come from. And a privilege to be at Google, who allows us to kind of expand that space. So the reason I asked you all to show your hands, most of you know Google as a search or map and similar pieces, information source. One of the other things Google is now, I think, as a company, is a full -stack company. And when I say full -stack, that means the search and the information is a top layer of it. But underneath that sits the entire physical infrastructure that drives that.

And so that’s data centers. That’s the way you operate that. That’s the networks that feed into all of the applications. And so when we look at climate, and my title actually is Global Director of Climate Operations. So I say that out of humility because when we look at climate, we’re trying to put it across our operations the best we can. Thank you. So we, and good examples of that, I’ll start with data centers. The big topic right now, how do we operationalize them? Where do we cite them? The location, what impact it has on the community, what impact it has on the infrastructure there. Citing is a big part of it. Access to clean energy is something we’re looking at, and pretty much we have a carbon -free energy goal.

So I think for us, if you look at climate, a big portion of climate is emissions. How do we impact emissions? It’s from electricity. What do we do with electricity? Shift to clean energy. Or renewables. And so that’s the spectrum that we look at. And so a lot of our investments are in carbon -free energy and how we think about it. And it also is more not just take from the grid or expect the government or, you know, sort of the infrastructure to get you there, but how do we invest and bring more clean energy to the grid? I think that’s a big piece of, I think, what companies can do at the speed at which we are all moving is, how do we take these sort of bigger picture systems problems and embrace them and solve them?

So one is big, one is small. is generation of clean electricity, and the other is grid, and how do you solve the grid problems? So that’s the infrastructure of AI. Then using AI, I think that’s going to some of the other things, Professor, you were saying is, we look at how we could use AI to drive our operations more efficiently. It’s very boring pieces. It’s not really shiny superstar things, but a lot of the impact comes just in general. I look at water taps, and I remember the amount of leakages we have on water taps, the amount of electricity wires that are not connected, just the inefficient use of resources is a big one, and how can we use AI to sort of optimize, whether it’s optimizing within the use of our chips, optimizing the grid, optimizing which applications run from where.

That’s a big part of our strategy. And then the third piece is, what do you use AI, and how do you use it for climate? Now, clearly, our business is information and search, but which means we also have access to a lot of data. And so one of the ways we consider, as what you can do in AI is, how do you use these large data sets? A, find a way to open source them, encourage different use of them, but also incubate certain initiatives that can help to show the light to others. So Earth AI is a big one in which we you’ve got satellite images, you’ve got weather data, you’ve got all of these big chunks of information that we can put out there.

And then there’s an application layer, which I think is of interest to you in terms of resiliency or mitigation. So one of the things, you know, which you probably haven’t heard of as much is Flood Hub. So we have a lot of information put out there as to flood risks of different region, which then other companies can use for whatever products they’re launching, whether it’s insurance, whether it’s real estate, fire sat, wildfire risks. How do you do prediction around it? Something all utilities companies, especially from in California, where I’m based, is we’re very passionate about using that for prediction. I can go on about the list of sort of what data can be used and how it can be leveraged.

The thing I think I’m going to go back to the, you know, crux of what you had in this is. is we’re in the timeframe of two hockey sticks. One is the impact on emissions, and I completely appreciate that the tech companies, hyperscalers, data centers are at a scale contributing to it, which we want to obviously help mitigate or replace with clean electricity as much as we can. And then the other is, how do you use the innovation curve on this? And I think we’ve just scratched the surface. And there’s, of course, like, you know, there’ll be trials and errors, but the surface around how do we democratize data, how do we encourage innovation, and how do we scale it very quickly?

Because I think those are the three, the trifecta of how do you drive this change? And so one of the ways I’ll end with saying I’m super proud of what we’ve done this week, trying to bridge those two gaps is we’re working with the Principal Scientific Advisory of the Government of India to launch a Google Center of Climate Tech. We call it Climate Tech because it does, those are the two hockey sticks that you’re trying to get in, right? The tech scale and the climate impact. and we are our goal is to encourage academic research but research that is actionable so five pilots first of all kinds and how you can scale um and there’s a lot of uh you know focus already on electricity so we are trying to do the non -electricity pieces in that which is around low carbon steel low carbon materials built environments um low carbon sustainable aviation fuel and then the biggest one we don’t talk we talk about what i think is a big lever across everything is green skills you need to embed this sort of a thinking which is green climate first across every domain and how can we encourage that in in india and especially the tier two cities so super excited about those two hockey sticks and how we’ve as a company can bridge those gaps

Intensity of what is produced in this part of the world actually is really important globally. But what’s really distinctive about APAC is actually the third major topic, which is livelihoods. As I mentioned, this is the part of the world which has a lot of developmental ambitions, and livelihoods are key. So my colleague touched on, Vushali touched on Chapter 3, power systems. Actually, I would like to touch on agriculture and food systems, which is your Chapter 4. So agriculture and other land use is actually the largest employment sector in the Asia -Pacific. I mean, I believe in India it’s about 46 % of jobs. And for the region, it’s the largest sector, about the same as the next two sectors added together, which are actually manufacturing and wholesale and retail trade.

Add those together, you get the same number of jobs as in agriculture. Now, over 80%. 80 % of farms around the world, and especially in India and the rest of the global south, are smallholder farms. farmers. And that creates an issue because a lot of the technology for agriculture is developed for large commercial farms, satellite imagery, et cetera. So this is one example I’d just like to delve into in terms of what Google is doing to contribute to the data, the digital public goods that Rushali spoke of. So if you want to use satellite imagery and actually understand agriculture so you can do things with it, you need to find the boundaries of your individual farms. That’s your individual unit and often less than two hectares, if not smaller.

And so you can do that with people poring over maps or satellite imagery, but that’s not scalable. But this is a really interesting problem for AI. And so for those of you who’d like to know more about this, there’s actually an exhibit at the Expo at the Google Pavilion. But this is what we call agricultural landscape understanding and agricultural monitoring and event detection. So we’ve trained AI to actually digitally enhance the environment. The field boundary. and you can say, well, that’s interesting because you can zoom into India and look at the Indo -Gangetic plain and see all the field boundaries. We’ve also trained the model to distinguish what crops are being grown through multispectral imagery.

And with that, we can detect events like tillage, sowing, harvest, et cetera. And all this data is now available. It is part of the Krishi DSS. So it’s contributing to the digital public infrastructure of the Indian government through the Ministry of Agriculture, through state governments, for example, like that of Telangana and the ADEX system. And what this does is it allows NGOs, government bodies, et cetera, to actually give advice to farmers because you now understand what’s going on on the ground, which is a critical driver for mitigation benefits, but also adaptation as actually the planting and the growing of crops. So we’re seeing best practices for planting. and what to plant is actually changing over time with climate change.

So do find out more at the pavilion. But one thing that I like to double -click on, as we say, is actually the innovation part of it. This digital public infrastructure is only helpful if it can be really used. And it’s not just governments and NGOs. It’s also startups. They’re innovating and finding new ways of using this information. So companies like Carbon Farm, they’re in France. They’re using this data. Varaha, which is a social startup, entrepreneurship. But also, Wadwani AI is another startup that we are supporting as well in terms of driving innovation in the agricultural space. So this is really all going to be accelerated through the use of AI, and we’re very excited to contribute to that.

Thanks.

Wonderful. You can see – I’m going to just grab one of these. Thank you. So, hello. Yeah, you can see that Google represents the convergence of the two themes we were talking about. And I think you have a wonderful web. At least I’ve seen access to materials about your sustainability strategy online. So if people want to know more, I’m sure they can go there.

Yeah, I’ll make a plug. Website, sustainabilitygoogle .com. It has all of our information. And the expo booth has all of our information. So thank you very, very much.

Now, we are inevitably vastly behind schedule as we are with climate change. However, we’re going to keep going with great focus. And we’re going to turn to the energy and power sector. Now, this is a bit embarrassing because we have to do a little bit of a switchover of people. But we don’t have time to put up the new names here. So we’re just going to announce them and listen with great attention. So we have two fantastic speakers from the energy sector, which, as you know, is one, if not the most important sector. Do you want to come closer, Nalim? And we can just be together here. I would ask, obviously the decarbonization of the energy sector is absolutely critical without that nothing happens so I’m going to hand over straight away first to you if I may Nalin, to set the stage a little bit and what you’re trying to do with Climate Collective and UNESA and then Dan more specifically to what you’re up to so over to both of you and obviously introduce yourself sorry I haven’t done it for you

no, thank you I understand we are short on time so I’ll keep very brief I’m Nalin Agarwal, one of the founding partners of the Climate Collective I think we’ll have the slides up soon great, so today I’m just going to talk about very quickly a program that we’ve been running for 6 years and where we partnered with Graylon to really drive decarbonization and grid modernization starting with India but across the global south so if I can move on who’s operating that? let me go there to you I’ll do it. Okay, just quick snapshots. We are an ESO enterprise support organization, largest in the global south, about 1 ,500 startups supported. Key partnerships, UNESA is a key one. I don’t want to spend too much time here, but that’s what we’re going to spend some time on as well.

We do a lot of work in AI, in power but beyond. So next week we’re doing the Delhi Climate Innovation Week. In fact, Google is a sponsor and partner there, and of course, Grail is as well. But happy to chat about this later. Here’s what we’re trying to do. I think what’s happening is that a lot of the challenges on renewables are being solved. They will be solved. I think one of the increasing recognitions is that the grid is a key bottleneck now, and we need to really work on grid transformation. That includes both decarbonization and modernization. So that’s what we’re working towards. So we work with utilities. There’s about 22 of those that we’ve worked with so far.

Work on a problem statement approach. Get startups to apply. Select startups. Get them to create business cases and pilot plans and eventually lead to pilots, right? So there’s 22 utilities that have participated and have led to about 20 pilots, a subset of which have become large deployments, right? So it’s a very unique program in the global south, actually. It’s the only one, right? High conversion ratio, so about 30 % of the pilots that have been proposed have come in. Key partners, I mean, 22 utilities and all the people that are working in power sector reform are part of this program. I mean, again, I won’t spend too much time, but there’s a lot of this information available online. All the startups that are vetted, ready to deploy, are available for utilities to engage with.

we have a bunch of case studies also but the key point is this we are now developing this along with Grail into a global AI for power innovation platform which has three components the open innovation program which is electron wipe on the top there is the knowledge hub which is basically a peer sharing platform where we do convenings co -located at COPS at climate weeks etc and then there is an online solution database of pre -wetted solutions. I’ll stop there and hand it over to Dan

Thanks. Thanks Nalan I’m going to stand up too because I like to stand up and talk so my name is Dan Travers I’m from Open Climate Fix we are a startup doing AI for grid. I’m going to dive a little bit into the grid area which has been talked about a bit in order to get to net zero we need to electrify we need to green the grid and we need to electrify everything the grid of the past had Usually in each country there was tens of generators and the grid operator would know those people on a first -name basis and they would ring them up and tell them when to turn up and down.

We’ve now got millions of generators with solar panels, wind turbines everywhere. The grid of the past had variability from just demand. We’ve now got variability from demand and the wind speed and the clouds, right, so three sources of variability. The grid of the past had a normal demand that we understood well. We’ve now got data centres, we’ve got EVs, we’ve got batteries, we’ve got AC, so the demand is changing shape incredibly. How are you possibly going to address this balancing of this grid with a bunch of people in a room, right? You need AI solutions. You need a highly digital grid. You need something which can schedule and marshal all of these assets in a digital at sort of AI speed.

So… That’s really important. And why is it important… It’s important because if we don’t do it, we’ll have blackouts, and if we don’t do it, we’ll have costs increasing because the way that grid operators are currently dealing with this challenge is they’re actually scheduling a lot of backup generation. It’s usually gas -fired generation. It’s very expensive, so bills are going up. And if you look around what’s happening now, there’s a push back against the green revolution. If we don’t address these problems, we’re going to have a democratic pushback, and we will have a reversal. So AI solutions can really help us in fighting the battle for hearts and minds as well as the actual physical battle.

So myself, I came from sort of banking tech space. Jack, my co -founder, came from Google DeepMind, who the name keeps coming up. We both saw there was a big gap between the amazing tech that was available in some of these industries and grid operators and the electricity industry, which is by nature very risk -averse. It has to be worried about things failing all the time. So we saw the gap between those two, and we formed Open Climate Fix to really try and address that gap. to bridge that, to take sort of moonshot ideas and actually build a rocket ship that is going to fly to the moon and actually implement something and give data to researchers.

The company’s non -profit, we’re open source, and that’s about the scaling, which I think is a key part of the title of this talk. So we’ve built the best solar forecast in the UK, we think, by about 20 % or 30%, like quite a long way. We now want to take that, we are starting to take that to India. We’re working with Adani, we’re working with Rajasthan Grid Operator, and with a combination of open source plus commercial sort of expansion, we see the AI tools as super transferable across grids. So I’m really excited that we can take tools from one grid and apply them to all the grids in the world and use AI to solve climate change.

Thank you.

Thank you so much. You can imagine if we had more time, we would have had a panel on the built environment, a panel on industrial decarbon, a panel on transportation. We don’t have the time. But thank you for that fantastic presentation. A couple of interventions. Now we turn to the last segment. We have three very distinguished institutions with us, all involved at the strategic level. with Grail and with this process. And I’ll start, Ankur, with you at McKinsey, who have been close partners.

Thanks a lot. Another race against time. So what would I like to… I should say that Sean went out of the room and negotiated five minutes more for us in the room. So… Okay. So, while the slides come up, it’s firstly a privilege to be here. And thank you for the opportunity for McKinsey to be part of the journey that you are leading, Uday. And thank you all for being here and shaping this in your own special way, at your own scale. I’m Ankur. I’m a partner based out of our India office. I lead Quantum Black in India, which is our AI team. And I work across sectors, because that’s really… a lot of fun.

And part of my work has been in energy. Part of my work has been in the built environment. Thank you. but I’m representing the team which is quite global that has had the privilege to work with the GRAIL effort. So I’d like to just talk about how this little effort with us is shaping the larger movement that GRAIL represents. So everybody’s talking about the impact of AI, so I’m not going to talk more about that, but the promise of AI, let’s just be clear. I think the way large global efforts have sort of found shape is to focus around a few challenges. So one of the big pieces that the GRAIL work has been about is shaping these four challenges and articulating them.

They’re about operational improvement in our current way of working. Big consulting words, strategic intelligence and foresight, basically better planning. Okay, build things better. Transformation, innovation. So can we do new things that don’t exist? That will help the future. And the last one is autonomous operations, which is essentially do you do current operations in a very different way? Use drones instead of people to go see how the wiring is in a large electric plant. and create more impact. Several of the examples you heard about will fit into this across energy -built environment materials and this can keep expanding food systems perhaps. Now, there’s a huge amount of work going on in just collecting the knowledge on each of those challenges.

Then you think about those fields of play, the energy -built environment. Within that, there are stakeholders. So for each stakeholder, what’s relevant? And then for each stakeholder, let’s say if you talk about system operators here as an example, there’s network planning is a domain to think about. Asset management is a domain to think about. Delivery is a domain to think about. Field force execution. Think of this as you’re now bringing in the language of the industry into this knowledge base so that if someone manages a power plant, they’ll be like, okay, what’s my library of things I need to look at? Tomorrow, that can then connect it to people who are innovating or providing these solutions.

One important gap in the middle is, okay, how valuable is it? each of these ideas when it comes to cost, when it comes to emissions. And the work’s not yet ready to be unveiled, but we are quite privileged to work with the Grail team and, of course, global experts to start to now quantify, both in terms of economic impact, but also in terms of direct emissions impact, what each of these applications could be worth. Because then our scarce resources and limited time can be focused on the most important problems. And I think that’s what’s coming up ahead, and I look forward to all of you pushing the boundary further, and it’s a privilege to be part of this.

Thank you.

Okay, I will kick off. So as a metaphor for the climate, we’re drastically running out of time, and I can see a clock ticking down in front of me. So I’m Rob. I’m from University College London. 200 years ago, University College London was founded with a… a purpose to drive… change to be impactful and to create useful knowledge. That’s really important for the climate because we no longer have the ability to let knowledge sit on the shelf when it comes to climate. So in 2026 at UCL, the way that we bring our community together is through what we call the Grand Challenges. These are a self -funded, cross -university way of tackling problems that are too complex for any one discipline.

Climate crisis at UCL sits alongside challenges like mental health and well -being and data -empowered societies, and they’re found in all 11 of UCL’s faculties, from engineering to health and arts and humanities. So where does AI come into this? Well, AI at UCL is not seen as a single discipline, but as an enabling layer embedded across the entire institution. It builds on our heritage in AI. We’ve got three Nobel Prizes, we’re the birthplace of Google DeepMind, we’ve got several Nobel Prizes, we’ve got the Nobel Prize for the companies all at unicorn valuations. Four quick examples at UCL at the moment. Starting at home we use our own campus as a living lab we’ve got sensor data from across our estate that forecasts energy demand and detects unusual patterns across UCL’s buildings and we turn that into insights for practical intervention.

Second example we’ve got our spin -out Carbon Re which uses deep reinforcement learning and digital twin optimization to cut fuel use and emissions in energy intensive processes like cement production. Third example is a partnership UCL Center for sustainability and real tech innovation is created in partnership with PGM real estate it links computer science to the built environment and accelerates AI enabled sustainability in the real estate it drives impact on the environment but also value for the real estate investors. And then we’ve got our digital innovation center which uses deep reinforcement learning and digital twin optimization to cut fuel use and emissions in energy intensive processes like cement production. Third example is a partnership UCL’s center for sustainability and real tech innovation is created in partnership with PGM real estate it links computer science to the built environment and accelerates AI enabled sustainability in the real estate it drives impact on the environment but also value for the real estate investors.

And fourth UCL Grand Challenges has supported an inclusive and AI tool that transforms satellite and drone imagery into accessible web -based sea ice classification that’s being used to support safer travel for Inuit communities. Aviation is another frontier for us. It’s a grand challenge in its own right. And in there, we are looking at short -term and long -term interventions. AI is used to create short -term interventions that drive down its impact on the climate, while engineering is undertaking long -term technology transform in electrification and hydrogen propulsion. And finally, for UCL, convening really matters. In April 2025, as Uday mentioned, UCL, along with GRAIL, hosted our International Summit on AI Solutions for Climate Change, exploring sectors like energy and the built environment, and moving from discussions and pilots to deployment and impact.

I’ll finish with a quick call to action, which is that the grand challenges created by the U .S. government have been the greatest challenge for us. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past.

We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of challenges in the past. We’ve had a lot of

Cheers, and we’re properly into Alex Ferguson overtime now. So hopefully not with the climate change, so I’ll try and leave a little bit of time for Uday. So I’m Adam Sobey, I’m from the Alan Turing Institute. This is the UK’s National AI Institute. We focus on five missions across environment, which is focused on environmental forecasting and climate change, on sustainability, on defence and security, on health, and on foundational research. And as the Director for Sustainability, obviously I think that’s the most important mission, and that’s why I’m here. But we believe that the time for action is now. is literally on fire. We saw fires in the US which have been linked heavily to climate change.

We are seeing droughts in India which is affecting the food and people’s lives. We’re seeing pollution in Southeast Asia which is affecting health. We cannot wait for new fuels for the energy transition to occur. We need to do something immediately starting today. And we believe that AI can play that role. We know this because we’ve, as a part of our institute, have applied AI and data science to shipping and reduced emissions by 18%. We have done this in buildings where we’ve improved HVAC optimisation to reduce emissions by 42%. And we’ve created an underground urban farm that works entirely off renewable energies in the UK, allowing us to grow crops without using any CO2. However, I think we’ve done some relatively impressive things for a relatively small institute.

We’ve done some really impressive things for a relatively small institute. We can’t do this alone. we realised that this is a global problem and the Sustainability Missions chief funder is Lloyd’s Register Foundation which is a global charity heavily focused on the global south and so we think that it’s really important that we work together both within the UK and outside of the UK to solve these problems and that’s why we’re really pleased to be part of Braille to look for global solutions to global problems so thank you very much

It’s a tribute to all our speakers that they managed to put extraordinary quality into this ridiculously short time frame. I’ll just end on three words. One word is that we have come through our work together to find hundreds of examples of opportunities where businesses for example can save money or increase revenues, improve their economic value while at the same time massively improving their emissions profiles on the mitigation side. On the adaptation side… to your points. There are many examples from Google to all of your institutions where these technologies are already being deployed to save lives at a big scale. And the last point I’d make, apart from I’d ask you to thank our speakers with a big round of applause, is first to say that again and again you’ve heard one theme coming out of this group, which is radical collaboration.

Work with us to make the difference that we all believe and know can be made through the application of AI solutions to climate change. So maybe we could give our speakers a round of applause. Thank you very, very much.

to be with us, so thank you. We are here because we believe in AI’s transformative potential, and I’m certain you’ve heard a great deal about it over the past few days. Today, this session is about something deeper. The digital divide is rapidly becoming a compute divide. AI today is not constrained by imagination. It is constrained by infrastructure, by who has access to GPUs, to cloud capacity, to scalable compute. And that divide will determine who shapes the future of AI. Democratization in this context is not about catching up. It is about expanding who gets to lead. It is about ensuring that the next generation of AI breakthroughs are not concentrated in a handful of geographies, but are shaped by diverse talent, languages, and lived realities across the world.

And here, India is not waiting for permission. India is not waiting for permission. India is showing that it can be done differently. Through the India AI mission and through the compute capacity plan, mobilizing more than 38 ,000 GPUs as public infrastructure, India is building one of the most ambitious public interest compute ecosystems anywhere in the world. This is not incremental reform. This is infrastructure at scale. This is sovereign capability combined with openness. India is demonstrating that public interest AI infrastructure can be built in the Global South by the Global South and for the Global South. And this leadership matters because equitable access to compute is not just about hardware. It sits alongside access to data, open source models, talent, and institutional capacity.

India is proving that you can design AI ecosystems that are both globally competitive and globally competitive. And locally grounded. At the Rockefeller Foundation, we believe this moment requires moving from diagnosis to action. Philanthropy’s role is to be catalytic, to reduce risk, unlock capital, and convene unlikely partnerships that accelerate progress. Over the next hour, our discussion will unfold in three acts. First, what exactly are we democratizing? That’s an important question. Second, how do South -South partnerships and catalytic financing accelerate progress? And third, what concrete commitments can we land this year? If India’s example shows us anything, it is this. Democratization is not theoretical. It is operational. It is scalable. And it is already underway. The question now is how we accelerate it together.

Before we begin, let me take a moment to acknowledge a few leaders in the room. Shri Abhishek Singh who unfortunately has been pulled into another meeting but his leadership has been amazing his steadfast partnership and support has been something that I am extremely grateful for his vision of guiding this important work with clarity has been just spectacular Dr. Sarabgarg we are honored by your presence you have been in sessions since this morning and thank you for your leadership it’s truly a privilege to have you with us today my colleague Andrew Sweet who has joined us from across the world one of the sharpest lines of the Rockefeller Foundation and truly a force for good thank you for being with us today and supporting this conversation and of course I want to also thank Charu who has been working endlessly and very hard to kind of get us to this place thank you Charu for your leadership Martin, Vilas, Sean and Dr.

Shikoh thank you for lending yourself your voice and expertise to today’s discussion Your perspectives will help ground this dialogue in both ambition and action, and I know all of you are action -oriented folks, so we’re going to have something really cool come out from here. And last but certainly not least, our partners at Kalpa Impact, Sushant, Anish, Jennifer, thank you for being extraordinary collaborators and for helping shape today’s session. It is now my pleasure to hand it over to Dr. Gorf. Please, over to you, sir, or maybe I’ll hand it over. Okay.

Thank you. Thank you, Deepali. When I mentioned the report, I fumbled the name, so I’ll go again. Opening up computational resources for new AI futures, new AI world is possible. And this is something that the team has worked really hard over the last few months. And today is an opportunity when we release a working version of that report and invite inputs, feedback, comments, and suggestions, which we will work through over the next few months. This research helped us think through and work with the Democratizing AI Resources Working Group under the leadership of Dr. Saurabh Garg. And he’s here. So it’s a pleasure and a privilege for us to invite him. And the other panelists to release this report.

Thank you. Thank you. Thank you. Thank you. opening up computational resources or in fact all resources that are necessary for development of AI in public interest and for real world impact. I could think of no better person than Dr. Saurav Garg under whose leadership I think we have come a long way in not just the intellectual thinking but as he will tell you in terms of operationalizing how we can bring this to life for billions in the global south and also the other countries in the world. Dr. Saurabh Garg, please for your keynote.

Thank you and colleagues panelists great to be here and great to see a large kind of attendance that we have seen over the past few days in the AI summit and And there were seven working groups set up under the AI Summit umbrella. And one of them was on democratizing AI resources. I had the privilege to chair that group along with Kenya and Egypt. So I’ll obviously talk a bit on that. But before that, just to say that I think all of us are of the opinion that AI will definitely transform the world. I think the question is whether that transformation would be equitable, would be inclusive and aligned with public interest. And I think that’s really the issue which concerns a lot of people.

The AI Summit itself was built around three guiding sources. Sutras, the people, planet and progress. And therefore, the concept being that AI ultimately must serve human welfare. advance sustainable development and enable shared prosperity. I think these would be key background in the way these sutras were developed. And obviously, democratizing many of these resources would be key to that. During our working group discussions, we had the opportunity to talk to a large number of countries, people from academia, civil society, and other international organizations. And I think one consistent message was that most countries are not really seeking only access to AI, but also seeking agency in AI. And I think that’s key. And how the AI systems need to reflect each country’s own development priorities, languages, and social contexts.

From these discussions, there were six foundational pillars that we had to address. And we thought need to form the backbone of the collective roadmap for the future. computer capability collaboration connectivity compliance and context and I’ll just briefly speak on each one of these a bit compute no doubt is today’s defining barrier the access to GPUs accelerators high -performance clusters is a major issue for all AI ecosystems but the issue is how it can be made distributable affordable and reliable across and not concentrated in a few geographies and this would no doubt require us to look at whether compute can become a shared infrastructure in future or kind of a which supports public interest innovation and to the extent that we are focusing on innovation how that part can be a public interest infrastructure secondly infrastructure structure would not be sufficient there is a widening skills gap.

So how we can consider capability diffusion focusing on joint research, shared standards, open platforms and mutual learning. What needs to be done for this responsible deployment is so that we can link innovators to compute resources and citizens to trustworthy AI enabled services. Equally important would be governance. The governance framework needs to be robust enough to build trust, yet flexible enough to adapt to diverse social and cultural contexts. Open source and maybe modular AI stacks would help in enabling localization without creating dependency. So looking at some of these issues, on what mechanisms can be done to facilitate accessible and affordable computing resources by improving utilization rates and reducing transaction costs and also to lower barriers for access regardless of geography.

The working group looked at how this can be taken forward through a collaborative platform designed to expand shared access to compute data in partnerships. And the platform has been termed as Maitri, which is friendship in Hindi. Maitri, M -A -I -T -R -I, standing for Multi -Stakeholder AI for Trusted and Resilient Infrastructure, to be developed as a digital public good that countries can adopt, customize, and build upon. And obviously, it is a non -binding, voluntary, modular approach. depending on the context of each country, what kind of compute and what kind of methods can be used to have accessible, at least for innovators and researchers looking at data sets that can be put out, which are take care of the national laws and national protocols in place and look at models.

So what which are open source and which can be placed. So this this we envisage would help to at least ensure that portions of AI are a global public good, because we we we are focusing on innovation and research out here. And this would go beyond just a focus on hardware and platforms, but also in skills, institutions and governance capacity. I would just like to mention one area. The other is that the technology is a very important part of the development of the technology. but how perhaps it might proceed in future is also the fact that while infrastructure or compute seems to be the biggest constraint going forward as of now, that’s perhaps also based on the present models requiring large amounts of compute capacity and energy.

Going forward, would models retain this system of algorithms that they have, or would there be obviously small domain -specific niche models? I think yesterday there was a very nice remark made by Vishal Sikka, who mentioned that unlike when we talk of compute infrastructure, we are talking in terms of gigawatts, nothing less than that of whatever. But when you talk of a human being, you talk in terms of only 2 ,000 calories requiring a human being to sustain a computer. Which is not more than a 100 -watt bulb. for a day. So are we missing something out here and I think that’s a very important point that he made yesterday and that’s why the focus I think we need to have much more on the models and that itself might solve a lot of the areas that we are and when we’re talking of democratizing AI perhaps that’s the path forward.

So I’ll stop here and thank you all. Thank you for this opportunity.

We now transition to the panel discussion and may I request Andrew Sweet, VP at the Rockefeller Foundation who is the moderator for the panel. Please join us here on the stage. May I request other panelists, Dr. Shikogitao, Martin Martin Tisney Vilas to join us on stage. Yes and Sean sorry Sorry Andrew over to you

thank you Dr. Garg for those inspiring remarks and for the framing insight and perspective that you bring to this conversation all of the many conversations that you’ve had throughout the course of the week so we’re excited to continue and deepen the conversation today and very excited that we have five of the world’s brightest minds to discuss this topic these are all people that have been in the AI arena for decades, this is not new to them and all people that have deep regional expertise and global perspectives so very excited for this conversation today. We don’t have a lot of time we have about 25 or 30 minutes for the conversation so we’re going to dig in, we’re not going to have a number of speeches, Dr.

Garg’s speech will be the only speech that you’ve heard today but we’ll have a short series of provocations with actionable ideas for how we can move this agenda forward And so hopefully this conversation can be, you know, informal, back -and -forth banter. I think we’ll have one round of questions, but it would be great if we could kind of feed off of each other’s questions and energy because I know we all have a lot to say here on the panel and a lot of expertise to share. So I’ll briefly introduce the panelists, then we’ll dig in. You’ve already met Dr. Garg. He’s the Secretary of the Ministry of Statistics and Program Implementation for the Government of India.

He has been instrumental in shaping India’s AI governance and previously led the technology stack for the Transformative Adhar Initiative. We have Martin Tisné, founder of Current AI and public interest envoy for France’s AI Action Summit. Martin has spent 15 years building multi -stakeholder initiatives like the Open Government Partnership that we talked about earlier today to govern technology based on democratic values. We have Vilas Dhar , president of the Patrick J. McGovern Foundation. Vilas serves on the UN Secretary General’s High -Level Advisory Board on AI and leads one of the world’s largest philanthropic movements to AI for public purpose. my friend Dr. Shikoh Gitau founder and CEO of Kala AI a visionary from Kenya. She established Safaricom Alpha and has been a leading voice in ensuring that digital transformation in Africa solves real problems in education, healthcare and agriculture and finally, Shaun Seow CEO of Philanthropy Asia Alliance.

Sean is working to catalyze collaborative philanthropy across Asia, leveraging deep expertise from his time at Temasek and is CEO of Mediacorp so we’ll continue the conversation first question will go to Dr. Gerg India has launched the India AI mission with a target of 38 ,000 GPUs if we view compute as a public utility, much as we do with water and electricity what is the governance model that India is envisioning and should compute access be rationed or priced differently for public interest applications

so I would say that the focus is not on rationing but on intelligent prioritization I think that’s going to be the focus, that the impute capacity is an enabling platform, and as I mentioned, as a digital public good, at least that’s where innovation and research is going. So that we focus, and I think that’s where a lot of the philanthropic organizations would have a large role to play, given that their focus is also on ensuring that AI benefits all. So with that focus in view, how governments, philanthropic organizations, and the private sector can collaborate to ensure that affordable compute capacities are accessible to all. I think that’s the models that we are looking at, and that will ensure experimentation going forward.

Thanks, Dr. Garg. Martin, I’ll go over to you. Through current… AI and the Paris Charter, you’ve convened governments to discuss public interest AI. How do we move nations from being consumers to genuine co -creators? And quickly, you’ve also spoken about this looming data bottleneck. What do we do to unlock data sets for training without compromising privacy?

Okay, two big questions. Thank you. So, as you mentioned, we launched Current AI last year. We’ll be launching just this afternoon our first product, which is an open hardware product looking at linguistic diversity. I think I’ll be a little bit provocative to maybe start our session. I think compute is critical for obvious reasons. I think that from a financial, from an innovation, and from a sovereignty perspective, it is also possible to overplay it. I’ll tell you what my worry is, and I’d love to know what the panel thinks. I do have a worry that we could end up in a few years’ time in a world where we succeed in having compute capacity in inverted commas, in a number of countries, including in the global south, but where effectively the data centers are not used.

We’ve been talking to colleagues around the world. You do also have data centers that are effectively kind of white elephants and that are not used anywhere close to full capacity. And so I think for countries to be able to exercise sovereignty, they need to have contextual AI. They need to have contextual data in their languages with all of the diversity and the incredible richness that typifies their cultures available in order to create contextual localized AI that actually serves outcomes that people care about. And so while the compute piece is important, I think it’s one part of the issue. We need to talk about the data piece and we need to talk about the second part is the open source one.

So briefly, I think throughout the event, people talk about open source AI, that it’s a really good thing, that we’re all pro it. I think we also need to talk about how, from a philanthropic perspective, we resource the open source ecosystem. The reality of open source software is it’s mostly the top tier of open software is funded by large companies that are using it, right? Linux is partly funded effectively by volunteers working for SpaceX that are using it. There’s a bottom tier of dependencies in open source that are run on a shoestring, you know, by a few like critical, amazing people working overnight as volunteers. And there’s very few organizations, one of them, which is a part of the current AI roost, which looks at robust open source trust and safety, that are funding those critical dependencies.

So I think that for states across the world, in the global south and the north, to really be able to exercise sovereignty, and I’d love to talk about this a bit before, but I don’t want to hog the mic, we need to talk about compute, but also we need to be realistic about what the compute is going to be used for. So I think the data piece and the open source piece are really important. I think I’ve probably run out of time to talk about the data bottleneck.

Go for it.

Well, so the number… There are people in the room I’ve worked with for a long time on this issue. Velas, you’re one of them. Sushant, you’re another. I won’t name check everyone. I think it’s fantastic that there’s been so much innovation in compute and we’ve seen such change over the past 10 years. In contrast, I think it’s a complete tragedy that we haven’t seen as much, anywhere near as much innovation when it comes to data and specifically the ability for people to be able to share personal data in ways that both respect privacy and contribute to outcomes. And that’s effectively it. I think we need a huge amount more resources and thought, both when it comes to the technical side of the issue, and here, other than the side, I think that partly it’s solved, but enterprise users of AI have access to these kind of technical safeguards in a way that private users don’t.

And there’s a story that we can talk about if we have time. And then on the governance side, so for example, Velas, you and I have talked for a long time about different, and now there’s different forms of data stewardships, whether data trusts or others. To the day… I haven’t seen one that really scales to the level that we would want to see it scale. to and that I think we need a lot more resources, a lot more thinking there’s been work done but if we could harness even 20 % of the sort of like brain capacity of the world that’s going into compute right now I think we would be in a very different place. Thank you.

Excellent, thanks Martin. Actually Vilas I’ll go next to you because I think this reminds me a little bit about a recent article you wrote about the Indian Premier League as a model for how India builds world class institutions I re -read it this week in preparation for this conversation is there a similar IPL playbook for public interest compute or is the window for building these public institutions closing as commercial consolidation accelerates?

Well I can’t think of a more controversial topic to spend our time here in this conversation than cricket. It’s been a good week all around but I think many of the people in this room probably know. Before I start I just want to say Dr. Garg I want to acknowledge in particular your leadership on this work. I spend a lot of time with senior decision makers across governments and the conversations that we have had have really given me great hope for the combination of technological confidence but also an understanding of what this means across an ecosystem. And so I want to acknowledge your leadership in particular. Thank you. Look, this question around the IPL I think is great, right?

I mean, let’s not torture the analogy and take something really fun and then try to, like, tie it to AI. But here’s what I’ll say about it. I think in many ways what we need is a new institutional framework that goes from the elites who are participating in their own places to something that feels deeply participatory. And I think around compute infrastructure in particular, we are stuck in a model where we keep reengaging and renovating old concepts and try to describe a new world. I will tell you sovereignty has been the buzzy word of the moment, right? Everybody wants to talk about sovereignty and diffusion. Sovereignty as a Westphalian concept that goes back a few centuries tries to take the idea that ownership of pieces of silicon somehow magically results in outcomes and impact that transform lives.

Now, there are logical links. And, of course, there are codependents. competencies, but to simply say that we will site compute in a particular geography and so figure out a way to disconnect ourselves from the interdependence of the 21st century doesn’t really bring us to a good outcome. I’ll tell you the second part of this, AI diffusion. If you haven’t heard this already, every tech CEO here, this has been the buzzword of the moment. I spent some time yesterday with the prime minister and a number of tech CEOs who wanted to talk about their investments in India. Those investments in many ways followed the playbook of the PR press release. They were, we’re going to build a new data center, we’re going to invest in a new compute capacity.

But when you dig deep and you ask the next question, who will this really benefit? What value does this create for public impact and outcome? How does putting a large number of servers in a particular place result in that community finding an economic uplift, a benefit in economic opportunity, a sense of dignity? The conversation sometimes falls. flat. So AI diffusion to me in its core concept, the idea that you hyper concentrate technological capacity, compute data, and somehow the rest of the society benefits sounds a little too much like something that as an American I know too intimately as trickle -down economics. The idea that if we made the rich as rich as possible somehow the benefits would filter down to everybody else and it would work.

AI diffusion is a passive concept. It starts on the premise that we build technological capacity for a few and somehow it works out for everybody else. But there’s an alternate model and it ties directly to this report that’s been issued today and the work that we’ve been talking about. For AI to benefit everyone requires a direct and active impact. It requires us to step in and say what are the institutions we have to build that actually physically and metaphorically transform the idea of compute infrastructure to be something that everybody can use. It requires us to build the institutional layers and the capacity that lets a community that’s trying to solve a local problem know that compute isn’t the thing that holds them back.

rather the conceptualization of the problem the aggregation of the full stack of resource sets as Martin described that include compute that include data governance mechanisms that include the political agency of communities to participate and let us then turn that into that final app solution infrastructural development that actually leads to the outcome we’re solving for in many ways I think this is the great role of the institutions that are represented here on the stage and in this room for philanthropies to transform the capital landscape in a way that says great entrepreneurs and leaders like my dear friend Chico here and so many here in India that are building open source public access AI stacks don’t have to worry about the resource constraints of the private capital markets that they know that they can access governmental and substantive structural resources that let them build the tools that they want to and know that they have equitable access to the markets as well as a matter both of policy and as a product that they can go out and get to consumers and creators that they can provide a service that lets them people use it at scale.

And the last part of this, and I have to say this, is this doesn’t happen, as we’ve discussed in the private market, but it also doesn’t happen exclusively by going to frontline nonprofits and saying, now you’re supposed to be the builders. It requires us to innovate a new institutional set of intermediaries. I think of groups like Culpa Impact, which I think is an incredible example of a combination of technical sophistication, policy impact, support for government, that actually sits at the layer that connects these different elements and lets us build on top of it. I think this is the work that’s ahead. If we really think about pragmatic outcomes to this conversation, Andrew, I think one of the questions we might ask is, what are the institutions we need to build in the next 12 months that connect the dots around all of these different pieces and support this transformation at scale?

Dr. Shikha, I came across a recent article that you put out there saying that for the West, AI is a matter of efficiency, but for you, it’s a matter of life or death. You’ve been a champion for AI access. You were very active in this summit. You were very active in the Kigali Summit. We were together at the launch of the first ever AI factory for Africa together in April in Kigali. You’ve also talked about global tech companies. If they want African data, they should provide compute infrastructure in return. How do we formalize these reciprocal agreements, and what does a true India -Africa partnership look like that doesn’t just reciprocate global North -South models, similar to what Vilas was just talking about?

Thank you very much for having me. It’s always fun to listen to everyone here on this. I was hoping somebody was going to preempt some of the work that I was going to talk about, but lucky for me, I have some stuff to talk about. Thank you, Vilas. So when we talk about compute, it’s this amorphous thing. In fact, we launched an AI research lab in Nairobi, and we have some GPUs there. And one of the key things was like a demo showing up what a GPU is. And our PS was like, oh, my God, this is what a GPU is, because he’s never seen a GPU. And then I made sure, like, every time I’m speaking, I’m asking, how many of you actually have seen a GPU, not on the Internet, touched one?

Maybe five people. And this is everywhere. Every single room that we’re talking about compute, we ask the same question, have you ever seen a GPU? And so right now, five to ten people. So it’s this thing that people talk about. We need GPUs, we need compute, we need all of these things. And for us, it is very important, as an African continent, we had, like, our research. Colo came for the Global South a few days ago. Same question, how many of you have seen a GPU? about 10 people had never touched a GPU. How many of you need compute? Everybody raises their hand. But what does that actually mean? In fact, one of the panels, the starting point was like, when it comes to compute, we all need Jesus.

And I thought, how do we quantify this? So we, and I think we have already spoken to Calpa about this. We’re working, I think, on the same time about a framework. So we just released a compute demand index. Because we realized every time we speak about compute, people have ideas, they have thoughts, they have proposals. They don’t have the numbers to say that. We need GPUs. How many? We need megawatts. And the gigawatt, megawatt conversation, what does a gigawatt of compute actually mean? So we went ahead and said, for Africa, we need to, every time we’re having conversations with these governments, this is actually what you need. But you actually need to put money into it.

So we, our first index was, did it demand? And the second one is, is your country ready for this? which we are calling AI Investment Readiness Index. So I’ll give you some numbers. Africa needs 2 .5 million hours of GPU hours a year, 7 .5 million for the next three years to be able to start computing well. This is for training as well as research. That is something that I can work with. So when I come to India and say I need 2 .5 million GPU hours a year, how many of them can you give me? And we have this conversation with the UNDP in Italy, and they said, oh, we have 1 .5 million GPU hours that we can donate. We have 1 .1 million more to go.

Cassava is saying we are putting 2 ,000 GPUs. How many hours of GPUs with those GPUs, hours, not actual physical, how many hours of GPUs without those 2 ,000 GPUs actually provide for the continent? So we need to be. We have to start being very practical rather than being arbitrary on what we. want. Of these 7 million GPU hours we need in the next three years, Africa only has 5 % of that. So we are doing the math. We only have 125 ,000 of these GPU hours a year, which is like times three of that for the next three years. So you’re solving, when I go to villas, I’m saying I need these GPU hours. It’s very practical as I’ll say I can be able to do half a million GPU hours.

So it’s not just going with an arbitrary number. I need GPU hours to be able to put this. And for us, that is important. But for me, it is the conversation about investment, and that’s the conversation that we asked. How do we have this South -South collaboration? How does we have this India collaboration? How does it actually look like? There’s the paradox. Everybody, as he said, as Martin said, everybody wants sovereignty. Everyone wants to talk about diffusion of AI. But what does that actually mean? Do we actually need it? So I’ll give you two examples of my two favorite countries. Hopefully none of them is here, actually in Nigeria. So there’s something you’re calling the Nigerian paradox.

Nigeria is the number one country in the computer demand index. Why? Nigeria is doing very well. It’s 110 million Internet users, a huge population. They’re doing very well in terms of, like, e -commerce, financial services. So they’re up there when it comes to why they need compute. And we’ve seen this in India. India is very high there as well when you’re doing the same exact thing. But what about investment readiness? And investment readiness is are they able or capable of running a compute facility? Do they have power? Do they have the talent? And I love, I think, what Mateen and the minister spoke about, what we think about. When you think about compute, you don’t think about just GPUs.

It’s a whole stack of things. It’s talent. It’s governance. It’s all these things. And you’re thinking about. When you think about investment readiness when it comes to compute, you have to look about all these things. Because I can give you GPUs, as he said, and I’ve worked in digital transformation for the last 20 years, and I work for the AFDB as a digital transformation lead, and we’ll buy computers and go three years later, and they have never been powered at all. And that’s the case that is going to be with GPUs, because you’re going to give countries these GPUs. If they don’t have talent, they don’t have the power to run it, they don’t have the data sets, they don’t have models, they don’t have use cases to build on top of it, you’re wasting that money.

And that’s where the investment readiness comes in. So we’re talking to countries, and we’ve had this conversation with African countries, is there’s no point of investing all your dollars in putting a compute facility. Get your talent ready. Get your data sets ready. Have strong use cases that people can back. Then, with all of that, can we then define what are the demands? Can we make money? you need. You don’t need a gigawatt Kenya, you do not need a gigawatt of compute to be able to run. Maybe you need a 200 megawatt facility and that’s where we want. So coming back to the question, how do we interact with India? This is our demand. Burundi might need 50 megawatts of GPU.

Can India be able to facilitate that? But it’s not just about facilitating the GPU, it’s what is the GPU in service of solving for health, education, agriculture. And when you have clear use cases, then the GPU demand becomes an obvious ask. And I think bridging that and convincing governments especially of bridging that gap is what we need to be able to do. And then the governance framework actually comes to play. Thank you. I know that’s a lot.

That’s great. Thank you. Thank you, Dr. Shikha. We’ll go through Sean, and then I want to keep it kind of informal for the remaining ten minutes after Sean speaks, so any reactions to any of the comments, and then we can do a lightning round if we have time, but if we don’t have time, that’s fine as well. Sean, over to you. The Philanthropy Asia Alliance brings together 80 members and partners to address Asia’s interconnected challenges through collaborative philanthropy. Is there an opportunity for Asia’s philanthropic networks to coordinate shared compute and infrastructure, pulling resources from places like India, Indonesia, and other nations, rather than competing, and what would unlock that collaboration?

Thanks, Andrew. The advantage of coming last is that I could say I agree with all of them. Actually, I’m going to add to the much maligned word called compute. Maybe we could end the panel right away. I’m going to join Martin in actually agreeing that compute is actually a bit overrated. The ownership… of compute… So when you think about the stack, I’m going to add another way to frame the conversation, Jensen Huang’s AI stack. When you think about energy, hardware, compute models and applications and the top layer applications is really what will drive and value capture for the economics as well as the impact, social impact. Really the stumbling block is probably energy at a bottom level.

And thankfully for many countries in Asia, the costs have been driven down because of the abundance of hydro, solar and wind. Then when you think about the next layer of hardware, I mean that’s obviously dominated by China, Chinese and American players. And when you think about the compute level, I understand why we fuss over compute because the Americans own 75 % of the GPU cluster performance. The Chinese 15%, Europeans maybe about 4%. and the rest of us are only like 0 .1%. I think even India is just 1 % of that. But I think the issue is actually deeper than just the ownership. I mean, if you think about what it needs to get the work done, it’s more access. So the question you’ve posed me about sharing of compute, for example, between Indonesia and India, I live in Southeast Asia, and that’s why Indonesia is like a couple of hours away from where I live.

And we know the situation in Indonesia quite intimately. There’s data residency requirements, and that’s why there’s a build -out of data centers. Think also of the physical limitations of actually the latency of sharing compute between India and Indonesia. For example, 10 ,000 kilometers apart, when you think about the latency of what, 50 to 100 milliseconds, it’s just not going to work for the sharing of compute between Indonesia. and India. Attractive as the idea is, it doesn’t work. I think they’re just physical limitations, data sovereignty, privacy issues that prevent that from happening. So I just want to look at the positive side of what’s happening. When you think about the cost of compute is coming down, when you think about the emergence of new clouds, GPU for a service, I think these developments are actually going to be good for the unleashing of AI, for social impact, for economic capture.

So the way you can think about it is, how do we then make it a little bit more accessible for startups, for impact organizations? Maybe the way to think about it is really, how do you think about aggregating demand so that you can actually negotiate with the new cloud providers and get a cheaper pricing? How do you then think about philanthropy coming in? And to subsidize some of the compute costs. and I think I kind of agree more with the observation that you really need to go beyond just infrastructure you need to think about the ecosystem you’re building I think the skills gap in Asia is actually huge and that could be really what’s stopping us from really optimizing maximizing the power of AI in what we want to do Is that too long?

No, that’s perfect I’m not sure if anybody wants to react to any of that Martin, I see you scribbling furiously maybe first reaction to you

No, I am scribbling. I’m scribbling because I’m thinking about your points I’m thinking about the points of the panel and I’m thinking about the term sovereignty So my scribbles are to your point about the Westphalian concept of sovereignty that’s about the ability to make law within your territory and it’s a very global north concept and it’s a notion of territory which has physicality and I’m just what I was scribbling was the physicality of the territory, it’s like we’re very focused on the physicality as you were saying on the GPUs of the bricks and mortar, so we’re going to be okay because we’re going to be sovereign on this data centre, the data centre it’s on my territory and what got me thinking is other concepts of sovereignty such as when I was spending a lot of work working on data calamities and data stewardship, thinking about indigenous data sovereignty which is a different type of concept it’s a more relational concept than a territorial concept, right?

It’s about a pre -existing an inherent authority a relational authority over that which makes up a people, and so when we were studying for example indigenous data sovereignty in the Maori context in New Zealand the Maori community any data that in any way involves Maori, the Maori community … legacy is part of the Maori community so I think that there’s something here in thinking about a very in some ways a quite rigid approach to sovereignty which is about control as mentioned and one which is more about agency and which is more relational so that’s what the panel has got me thinking and I’ve been doing some writing and thinking with colleagues and friends around the notion not of a sort of like a controlled national stack but a global open resilient collaborative stack and that’s not one at all and just I’ll finish with that, that doesn’t mean that like all the data is open and anything goes and anyone can extract your personal data and you’re back in a sort of Zuboff you know surveillance capitalism world it’s one where it’s a question of choice and agency and the what you wish to exercise authority over and how.

That’s my scribbles. Thank you.

As you can tell, when we get on a panel with people you love and respect, the conversation just flows. So I want to build off this point and a little bit of what you said, Chico. I want to take a different tack to this question of agency, which is if I had asked any development leader in the world 10 years ago, if you could have your dream of an extra gigawatt of energy capacity in your country, what would you do with it? I can’t imagine that any of them would have said, well, I want to use it to run a bunch of computation on things that may or may not have short -term economic value for my country.

Andrew, your organization has been incredible around the world at building capacity and grids in power production, in ensuring that people can use power for development. And yet somehow, I think for many of us, we are surrounded by conversations where now the question has become, how many megawatts and gigawatts can you put into compute for AI? It is a fundamental challenge when you think about what are our priorities in development. Again, going back to core principles of human rights and dignity and participation in the world, to say governments who have limited capacity should now all of a sudden be focused on this topic. It brings us again from this question and this shift. In many ways, I acknowledge that the traditional conversation around compute is one of breaking over -dependence on the American AI stack, on other international players that are coming in.

But the response to over -dependence isn’t internal dependence, it’s interdependence. It’s saying if there are places that have incredible capacity and even potential to drive, as Shiko said, the availability of compute hours, how do we build interconnectedness that lets that be a mutual value exchange? Not merely, again, clients who have to go to another country and say, please give or let us buy compute, but rather the products of that compute are going to be to build the infrastructure that you can then use in your own country. To allow for centers of excellence that allow for local capacity and local competence to drive what gets built and to allow that to be the new tokens of international trade in a way that leads to a much more connected and shared prosperity rather than descending back to that 200 -year -old concept of how do we make sure that we’re competitive in an adversarial frame.

I recognize that what I’ve just shared with you is maybe not where the dominant private sector conversation is. And to those who would oppose it, the primary critique is, well, that sounds quite naive. And yet we’ve seen it happen. We’re seeing it in the few areas of hope in the multilateral system where we’re actually finding that technology governance is something that brings everybody to the table, that lets people engage in meaningful shared outcomes. We’re seeing the seeds of it. The question is whether we’re going to let them sort of die out in the sun or if we’re actually going to water them, invest in them, in order to grow.

Great. Dr. Garg, any final insights?

I know there’s a little time, but just one thing I would say that perhaps we need to spend a bit more time going forward on the frameworks that will help ensure public interest frameworks looking at things beyond compute, looking at models, looking at talent, looking at data, how that can be shared and interoperable and in a manner which takes care of public interest. So I’ll just stop it out there.

Well, thank you. Thank you to the Indian government. thanks to our partners at CalPA for putting this together especially for the authors this is now officially out there I think you have until March 31st to read the copies are available out there you have until March 31st to review the document and submit your reactions thank you to the panelists, really appreciate it enjoy the rest of the summit, thank you I think the NDIA team wants to hand over some souvenirs from the panel Thank you Thank you. Thank you. you you Thank you. Thank you.