Hello and welcome. Artificial intelligence is progressing at breakneck speed, and the critical question of our times really is: are societies, people, the public at large, governments, companies ready for the changes that AI is bringing upon us? Joining me on NDTV at this time is a very special guest, the world's first and, at this moment, only Chief AI Readiness Officer. Lila Ibrahim is the Chief AI Readiness Officer at Google DeepMind, and it is an absolute delight to have you with us at NDTV. Thank you.
I want to start by asking you about being the Chief AI Readiness Officer. People have heard different kinds of portfolios, some very interesting titles and names, but this is the first I've come across — a Chief AI Readiness Officer. Explain to everyone watching at this moment: what does that really mean? What do you do?
Well, thank you. It's a great question because, as to your point, we get a chance to define what this title means, and if we do it right, perhaps other organizations will follow. If we take a step back, our mission at DeepMind is really to build AI responsibly to benefit humanity. We're at a point in history where AI is mature enough to have impact here and now, and also new enough to be able to shape the future. We want to make sure that the future is built with and for society to have the positive impact that we aspire to have.
So really, the first part of my job — the past nine years since I started at Google — mine was really thinking about how to build the technology and how to build the organization. My role was the founding Chief Operating Officer. Now, as I said, technology has advanced enough, but what we started to see was that we wanted to make sure that the communities were really ready. Readiness isn't something that you import into an organization, into a country, into a community. So we have to spend more time externally. We've spent all this time internally readying our own organization. How do we take those learnings now and work with communities? Now, that could be everything from teachers to farmers. It means getting their inputs and their needs — what are the problems, their concerns, the opportunities? How do we think about partnering such that we can also bring that back into how we develop the next generation of AI, so that it happens with and for communities.
Your job is to see whether community, society, the public at large are ready for AI technologies. And from your lens, Lila, I wonder who looks least prepared and who looks most prepared? Because when we read some of the comments coming from within AI companies, I wonder: is anybody prepared at all? So, a) deal with that, and b) who looks least prepared for what's coming.
I think we're all on a journey of learning together, and I think that's the most important thing. Also, we have to continuously update our models — and when I say model, I mean our mental model of thinking. I'll give you an example. Maybe I thought some of my use cases for AI were quite revolutionary, quite novel, six months ago, but now the technology — it's like I need to make sure that I'm always developing my skill set, my unique abilities alongside AI as a tool. And one of the things we've seen is teachers in some countries being much more advanced in how they're thinking about using AI as a tool, whereas in other areas the teachers almost have a fear of using it. That could even be within the same state, within the same community. What we're trying to do is ensure that it doesn't happen so jacked. What we want to do is make sure that we start bridging the opportunity so people can make use of it.
Perhaps one area I can give you a specific example is in the area of scientists.
DeepMind worked on a technology called AlphaFold.
So it was the world's first Nobel Prize that was awarded for the application of AI, and it was awarded because we were able to get the technological breakthrough of predicting the 3D structure of a protein. Now, a protein is a basic building block of life, and if you can predict how it folds, you can understand its function. If it misfolds, you can understand diseases. So we were able to use AI to help unlock our understanding of 250 million proteins in the world and publish it in a database freely available to researchers and scientists worldwide. Now, the technology was just the first step. We had to think about how to actually make it accessible to scientists. And just because they're scientists, it doesn't mean that they know how to use AI as a tool. So we had to think about how to ready the scientific community. Now, that meant access to the database. It meant the skill set, because maybe they didn't have access to the tools beforehand to spend the time trying to understand the protein. Today we have over 4 million researchers in 190 countries using this technology, including 180,000 right here in India working on solutions to challenges facing the country here — whether it's regarding how to interact with crop disease and crops, to how to think about proteins to understand some diseases so we can come up with better therapeutics. It's a great example of where there may be both a technological breakthrough, but it's only as good as getting it into the hands of the people who understand the local challenges and can work on the local solutions. It's not any one company. It's not any one country. It really takes this collaboration.
So India is apparently the fourth largest user of AlphaFold technologies in the world at the moment. When you look at how ready India is, first just as a nation-state versus some of the other countries that you've traveled to, where do we come out in your readiness index in terms of being ready or not ready for AI?
It's been really exciting to be back in India. We were mentioning, you know, I've been coming for nearly 30 years of my career, and it's been extraordinary to see the technological progress that happens here and how much the companies and the citizens really care for the broader India culture — whether it's languages and making sure the technology is accessible, to thinking about very local use cases such as what's happening in farming. What I've really been impressed with so far during my visit is also to see not only what is happening here for India, but what we've been able to learn from and think about the implications — like maybe India first, but really thinking about the global implications. Some of our work in agriculture, our Anthro program, which is working with 600,000 farmers here in India, we've been able to export that technology now to 11 different countries around the world, with ambitions to take it even broader. I think with the diversity of the country itself, along with a technological mindset and focus on real-world problems for real people, we're able to get extraordinary progress here and to think about how we are not only benefiting India and collaboration, but how can we use this as a blueprint for the world.
I was reading that DeepMind is now going to partner with Indian farmers to try and improve the quality of jowar. Can you explain how that works and how AI can improve the life of Indian farmers?
Yeah, this is another great example of where we think AI can have impact to accelerate scientific discovery to benefit society at large. We're delighted to announce a partnership with ICRISAT, which is actually headquartered in Hyderabad. Now, together we're working on a very important technology to help advance the area of sorghum. Sorghum is jowar, if I've got that right. It provides a food source, fuel, and also feed for livestock for over 500 million people. It's significant and it's critical for the livelihood of so many people. But we're facing challenges. We have droughts. We have disease increasing, and it's just getting more and more complex to think about how to provide this sustainably. One of the biggest challenges with sorghum is that it's a neglected crop. It's an orphaned crop, what they call it. There hasn't been that much research on it, and the research that has happened is largely defined by Western sorghum, Western crops, not what's indigenous to here in India or say Africa. Now, with ICRISAT, we're excited to think about how do we provide an open-source toolset for the crop itself, such that we can develop more disease-resistant, drought-resistant crops to help the sustainability of the food supply chain. We'll be doing this in partnership with, as I mentioned, not just the technology and the open-source platform that will be developed with AI tools, but also how can the industry, the agriculture innovation, happen on top of that, as well as how do we build the capacity building to be able to continue these advancements. Maybe something that would take a decade plus of work to do, we can now shrink. We like to call this science at digital speed.
And I think it's another example of where we're really trying to partner locally to address local problems, and yet they have large-scale implications. And if we could do this for sorghum, imagine what we could do with other crops and other grains.
So how long is this likely to take, and how do you ensure that whatever technological advancements happen with the seeds are actually then made available to farmers so they can roll it out as fast as possible?
You know, right now we're meeting every other week with crop breeders. So we're understanding what their needs are and what their challenges are, so that we're actually designing that in from the beginning. It's part of our responsible approach. We don't think you wait until the end to think about the farmers and the crop breeders and the farmers. You have to bring their inputs in from the beginning. This is exactly what I mean by readiness. So this is, you know, it's hard to — we're in the research stages, but I think one of the important things to realize here is that we are partnering with India from the research, not coming with a solution, but defining the research from the beginning within the scope of what needs to happen here, so that throughout the process, by the time the database is ready, we'll be able to provide this catalog to the crop breeders. And of course, some of this is actually you have to then develop the seeds and plant and learn from that. But the fact that it's happening here together is really critical. Why I'm so passionate and excited about this — maybe I can share a little bit of context. My mom actually was in agriculture research, and she was part of the 2009 World Food Prize, and what they were working on was drought-resistant sorghum.
And so what I feel extraordinarily privileged to be here to announce this together with ICRISAT, because I've seen the — I grew up in a house where sorghum was decoration in our house because my mom was always talking about it. So to be able to see the advancements now that AI might be able to do — I mean, that's been 20 years almost since that progress, and yet, you know, how do we make sure that we're providing a more sustainable food line? We don't have any more time to waste on something so critical.
Very interesting. The other thing that DeepMind thinks very deeply about is education, and at this moment there are very conflicting reports on whether AI is enabling education — whether students are going deeper, engaging with topics — or is it just becoming a tool to rip answers off, complete assignments faster than is required without learning anything? Where do you come out on that, given that you're exposed to this from all angles all the time?
Yeah. Well, and learning is something that is so important, I think, to so many people, right? I wouldn't be here today if I hadn't had a solid education, even if my mom hadn't had a solid education. So we really think: how can AI be the first tool, the first technology that actually helps close the gap on things such as personalized learning to help us throughout our lives? And within DeepMind, I think so many people feel this connection to lifelong learning that we said, you know, we're known for our scientific discovery and advancing science. What if we treated learning like a first-class science problem? How would we go about it? So about three years ago, we started working with taking a Gemini model and working with educators and pedagogical experts around the world to understand the science of learning. We tested and trained and tested and had a lot of deliberations around this, and then last year Gemini had advanced enough that we were able to infuse that learning science back into Gemini. We created a new feature called Guided Learning, and rather than give you an answer, it takes you step by step to help really break down a problem so you can understand the context. And what we found is — we're in the process of testing this — and in Sierra Leone in Africa, we found that students in an eight-week trial period, with the support of the teacher using Guided Learning, were able to advance in the subject of mathematics, middle school mathematics, a year and a half just within eight weeks. Now, why is that? Our understanding is that students then let go of the fear of asking the question. If you get a problem wrong, the AI tutor can help you understand what might have gone wrong. So it can be more personalized — you and I probably learn quite differently. So the AI being able to meet me where I'm at in my learning. Keeping the teacher in the classroom — the teacher in the classroom is the magic. They can see where a student might be struggling and give a little support. So what the teacher found was they were able to actually help more of the students in a very big classroom. In Northern Ireland, they have crowded classrooms, teachers are overworked, students with different learning abilities. We did a pilot program with Gemini as well there. Teachers had the agency to say how they were using AI, and they reported back to us that they saved 10 hours per week per teacher. Can you imagine saving 10 hours for a teacher? What that means they can do? They told us they could be a better father because they got time with their family in the evening, or they were able to drink a cup of tea when it was warm still, or help the student who was having a really tough day. So I think these are where we think that if we can use AI as a tool to again complement the individuals, the teachers and the students in a classroom environment, we have the potential to really unlock personalized learning — something that, you know, I've been passionate about for 25 years and we've been still working on, and I believe AI might finally be the tool to help us get there.
That's one way of doing it, Guided Learning, which you just described. But the other is the easy temptation of just typing into Gemini or ChatGPT whatever assignment that you've been given and getting the answer. Do you think about how to prevent the misuse of AI for education? I'll come to the rest later, because if they do it in the step-by-step fashion that you described in Northern Ireland or Sierra Leone, you come out stronger and better. And if you don't, then you're just being very, very superficial in your learning.
Yeah. Well, so we have to think both about the here and now and the challenges and the opportunities, as well as the future. Right now, we're talking about: are students really learning what they need to learn? That's the question that we should be asking. And how are we testing for it? And are our reward systems set up for it? I would love it if we really focused on our kids learning, not are they just passing a test. Because I think there's a lot of rote memorization, and we all know that what we studied in school is sometimes quite different than what is needed in the workforce. You can get an answer quickly, right? But that's the least interesting thing about AI. I think it's really about how do you encourage responsible use for the students. So I think this is where we count on the teacher being in the classroom in a supportive environment to teach the students how to use the technology as a tool, not as a quick replacement.
Lila, there have been a lot of comments in the last few weeks about whether AI is driving humanity towards extinction — alarmingly coming from within AI companies. As somebody who's right in the middle of that potential storm, what are you making of all these AI researchers coming out and saying they're very worried that their research potentially can lead to all of us being extinct at some point in time in the future?
I joined DeepMind because I really care about responsibility, and the first thing I did as Chief Operating Officer was: how do you operationalize responsibility in a research environment? I had hoped that, you know, my 35-year-plus career — that if I could bring all of my learnings from the semiconductor industry, venture capital, Coursera, and education into this moment in AI to make sure that we're not treating responsibility as a bolt-on, or safety as a bolt-on, that we're having the tough conversations up front of when we're deciding to have a specific area of research, whether it's AlphaFold, our work on weather, our work with farmers, that we're bringing those voices in so that we can develop the technology solutions better and together. That's the real way to have the technology defined from the beginning, to have the governance systems throughout, and then to think about how to be responsible in deployment of the technology. So I think responsibility and safety has to have a spectrum. It's a lot of hard work. I showed up at every single responsibility meeting that my team had when I was Chief Operating Officer because it's that important. I think the responsibility has to come from the top, and it's not an easy thing to do. You have to continuously be learning. I'm pleased we're all talking about both the risks, because I think it helps us to make sure that there's accountability in the system to help make sure that AI gets developed for and with humanity. And we can't have those conversations without thinking about how to also invest in the opportunities. So I think these two go hand in hand.
But what do you think of the question about AI driving humanity to extinction? I mean, I read somewhere that you said the chances are not zero, but not zero relative to people saying it's up to 10%. Where do you come out on that?
Well, as an engineer, I look at things very clearly, right? Are there risks in any technology? Yes. And it's our responsibility to actually openly talk about that. Some of the work that we've done that I'm really proud of are things like thinking about the Frontier Safety Framework — of saying what could go wrong and being very transparent in that, in the different levels of what could go wrong, doing the ethics research ahead of time to make sure that we have both the policy and the evaluations to be able to test with the partners externally. I think this is the only way that we're going to develop the technology safely. And I really just want to make sure that we're having these conversations. So I'm really appreciating that you're asking these questions. I think a lot of times, like Hollywood gave us this vision of what could go wrong, and it's almost created a way for us — despite living in different countries, maybe having different cultural contexts — that we can have a shared context for being able to discuss these types of topics. And that's the only way that we're going to prevent what could or could not happen and shape the technology in a way that is actually meaningful for us.
