Satya Nadella: Leading through the AI Revolution

0:00We sometimes shortchange human ambition, human agency's ability to deal with unbelievable new technology that comes along. Once every 10 years, once every 100 years, once every millennium, whatever. Like even the most magical technology has been used only to help humans achieve bigger and greater things. How I educate myself, how I get information, how I conduct commerce, all of these things are going to be changed. That's Satya Nadella, Microsoft's CEO since 2014, and someone I've had the privilege of working for for over a decade. As Microsoft celebrated its 50th anniversary, we sat down to discuss the path to becoming the company's third CEO, the surprising parallels between poetry and programming, and perhaps most important of all, the future of AI.

0:53My conversation with Satya Nadella, up next on Tools and Weapons. Well, Satya, thank you. It's great to sit down. Of course, we sit down all the time, usually not with microphones in front of us. But this is part three of a trilogy, obviously, the conversation with Bill, with Steve, now with you. And I thought maybe we'd talk a little bit about your roots, as I did with Bill and Steve, and then really dive into the Microsoft of today and your approach to being CEO. But let's start with where you began growing up in India. I think one of your first passions was cricket. And I think that molded not just your approach to the sport, but a little bit to teamwork and the like.

1:36Tell us where that came from. It's interesting, Brad. When I look back at it, I was, you know, quite honestly privileged in the sense that I had two parents who gave me the room in some sense to truly go and explore what I wanted to explore. whether it was the great Russian sort of literature or playing cricket in the dusty Deccan plateau. And I do think about team sports in general. You know, I distinctly think about all the lessons I learned. Right. One of the things that in particular that stands out for me is what leaders do to instill in people they lead confidence. In fact, I was having a very ordinary day in the field.

2:30And there was this guy who was my school captain who then sort of changes me over and then takes a wicket and then gives me the ball back to ball. And then I go on to have perhaps my best ever performance. And so I asked him afterwards as to why did you do that? And, you know, he looks at me and he says, because I need you for the season, not just for the match. I didn't want to break your confidence. And I'm thinking, God, this is a high school captain. With that level of enlightened leadership skills. And so that's always stood out for me. And so there's lots of things that all of us who, you know, grew up playing any sport, I think, learn in school that I think come back to us in different forms as we go on into our professional lives.

3:14But sport has been a major, major part of, was a major part of me growing up. And I still think about it. Well, I think you were exposed to so many interesting things growing up. And one, obviously, was sports. But another was, I think, your dad gave you a computer. That's right. When did that happen? Yeah. So my dad luckily had a chance to work in one of the UN organizations. And so he was an Indian civil servant. And so when he was working in, you know, in the UN, somehow he had both the monetary capacity and the access to a Sinclair Z80. And the thing I remember is it's the first piece of, I don't know, technology, tool, whatever that I've ever used that I felt reacted to what I wanted to get done.

4:07It is the most malleable material. I mean, these are words that I picked up later, but it is fascinating. And it had an impression, but I was not one of those people who said, oh, my God, that means computer science is my calling and high tech is my calling. I was still more interested in cricket and doing the things I wanted to do at that time. But it stuck somehow that this is a tool that can change the world. And that, I think, is what sort of later on formed into more of a real thesis for me. The other thing I think is interesting is, obviously, when you had a computer and you started to code, you were able to, as you say, take words, make them malleable, and get the computer to respond.

4:50But you also developed a real love of poetry as well growing up, another exercise in using words. Did that come from your mother? Yeah. So my mom was a professor of Sanskrit. And she had this very cool phrase in Sanskrit where it sort of paraphrased, translating it in English, she said, you know, if you're a child of a scholar, you are an idiot. And so she would always say this to me and say, you know, God, I wish you knew a little more of literature or enjoyed literature. And she was a professor of Sanskrit drama. And so she had a very aesthetic sense of, you know, what beauty is. And so she was the one who introduced in me.

5:38And in fact, my dad and my mother would talk about Sanskrit drama. My father also had deep interest. He was an economist. But he had deep interest. And then late in the night, I would see them talking about some stuff about Sanskrit drama and what it meant and what are those lines and what have you. And so even growing up, I thought, why are they talking about stuff that is not even a living language? And I had sort of slightly disdain because, one, I was not good at it. But I must say it had a lasting impression because in some sense, you know, as you learn to code, you realize coding is like another form of compression.

6:16And poetry is also compression, right? Which is you are able to take pretty long forms of content and compress it into two lines of poetry that captures the essence of it. That's what even code does. And so, yes, poetry is another thing. I'm not good at it. I'm not like I think I'm given more credit for my love for poetry than I really deserve. So I got inspired around poetry and how poetry can be perhaps a great way to sort of, you know, understand the world and express how you feel that just like Code does.

6:55I want to fast forward all the way to 2013. Steve Ballmer decides to step down. People were asked if they wanted to raise their hand and apply to become Microsoft CEO. You were not one of the first to raise your hand. Tell us about the thought process. Yeah, the first thing I'd say, Brad, it's kind of fascinating when we look back at it. At 2013, when Steve sort of said, oh, I'm going to retire. It is not like I didn't grow up in a Microsoft where at least I didn't, I don't think at least most of us didn't have, I clearly didn't have a concept call that is going to be a Microsoft after Bill and Steve.

7:35Right. It is not a night thought, right? I started like, these guys are here, they're going to be here forever and I'll work for you forever. That was the concept I had. So in 2013, it was a bit of a game changer when, oh, wow, Steve's, you know, going to retire. And it was quite frankly unclear as to exactly what happens next. And it was not even, And I remember, like, I think it was August of 2013, I think was when it was decided. And then I didn't even know what, you know, that you needed to raise your hand or there was going to be a process. It was not like even a thing that I thought of. There's a board and the board now has to go through a process of getting the next CEO.

8:15And I had just taken two years prior a new job, which I was fully occupied with. And I thought it was like, that's the one thing. I remember in 92 joining the company and feeling like, wow, this is the best job. And if I retired doing that job, that'd be fantastic. It was not like I was looking for anything else. And then the rest of all of what I did at Microsoft was I felt like, yeah, this is the job I want to be doing. And so it was not like I'm waiting. I always say this, I was never waiting for my next job to do my work. Just as growing up, you wove together different themes, different aspects of what became important to you in life.

8:53Let me start with the thing that I think people thought of you first, culture. You talked about the importance of it. Remind us a little bit of what brought you to this focus on culture and why even that first week as CEO, you made that a big part of what you asked the senior leaders at the company to think about. Yeah, I mean, there are two things I really think, Brad, I came to very quickly, and it sort of came naturally, was mission and culture. And the reason I came at it because I felt that as the first non-founder, I mean, like, you know, technically Steve is not founder, but man, like he had founder status as a company, right?

9:37And so to me, if you're going to step into the shoes of two unbelievably legendary sort of CEOs, then you better sort of start with re-articulating the two things that I felt mattered, which is what's that sense of purpose for the company? And so that was one pillar. And then the other pillar was this culture pillar. And to me, the culture pillar was important because in tech business, technologies will come and go, platform shifts will happen. You need to reinterpret your mission with a changing landscape. And that can only be done if you have a culture that allows you to reinterpret, right? Because if you feel like, I don't need, it's, you know, scripture from about, and I'm never going to have to reinterpret it, then that's going to be tough.

10:33Because so you needed a learning culture. And so that's why we were, you know, we were lucky enough that I had, you know, I had a chance, thanks to Anu, my wife, who introduced me to this Carol Dweck's work around growth mindset, more in the context of children's growth and what have you. And it stuck with me, right? I'd read it, I think, in 2012 or 2011. And so that always was like, wow. And I said, maybe let's bring this on the company and say, you know, this growth mindset as a cultural sort of meme can be a great handle to talk about learning culture. And so that was why I thought, like, it was important for us to talk about mission and culture.

11:13And you grounded everyone. I mean, it really did, I think, unite the senior leadership team. It united and unified the employee base. And, you know, it's sort of a classic of what people think about as sort of more soft skills. But you didn't shy away from making hard decisions. If we go back to 2014, by the late spring, early summer, there was this project that had just almost this nondescript label, Project Violet. But one of the first things you did was say, we're going to have to restructure some. We're going to have to cut some things. We're going to even have to eliminate some jobs so we can invest in the things that will be fundamental to the company's future.

11:54How did you build the support for that? Yeah, I mean, the main challenge in that time frame we faced were we were, you know, quite honestly, behind on two fronts. We knew that the world was going to be shaped by both the mobile transformation and the cloud transformation. and we were in both and we were weak in both. And so we needed to kind of decide what is it that we are going to go concentrate and win. And if we had a clear path to winning in both, then that would be wonderful. If not, we had to sort of make some choices. And so that was what led to that Project Violet and about saying, okay, we have a real shot.

12:45Like if I, you know, I always say that if somebody had to give me, hey, pick one, right, cloud or mobile, I would always pick cloud because I know that the cloud will outlast any one device as an endpoint. And it was going to be the control plane for what is going to be all experiences going forward. And so that was the intuition. And this is not like a thing that just dawned on me on one day. is something that I sort of learned throughout my Microsoft career. That was, in fact, very much part of even my CEO interview as to what was my worldview. And so it was sort of like I had a deep sense of, okay, what's the direction for the company?

13:27Given our structural position, given where technology is going, what's the winning hand? And in there, we had to make some tough choices. And there were a couple of new bets that you started to place, building on Steve Ballmer's advice, you know, be bold, but be right. One of the first that I think surprised people was saying, we're going to bring Office to the iPad. It was an important moment for us. I mean, because the Microsoft I grew up in, in fact, I was talking to Bill even the other day when, you know, after all, Office was born on the Mac first before even Windows happened. And so it's not like that alien to Microsoft's culture or history, which is we conceived, as Bill says, it's a software factory.

14:11And his idea was, hey, I'm just going to build lots of software for all platforms, and that's what we are. And so it was not that hard for us to say, okay, let us make sure that the office franchise is vibrant on any new form factor in which it can be used effectively. And iPad definitely was one of those things. And so I'm glad we did it. But it was, you know, I think a decision that helped, to your point, communicate more internally and externally our strategy than a real big decision that I made, quite frankly. It was sort of, I think, a great communication tool.

14:58One of the things you've been talking a lot about is the evolution of work, whether it's writing software or being an organization with a manager or being a manager. We're migrating to a world where people are going to get more help from AI itself, not just as co-pilots, but agents, this agentic world. And we need to embrace that ourselves as a company. What do you see as the biggest challenges in persuading people who create code for a living to rely on code in these new ways? Yeah, it's a fascinating thing, right? In fact, one of the first products that even for me helped, I'd say, solidify why this phase of AI is going to be a real platform shift was GitHub Copilot, right?

15:53It started with this codex model in its early form with just code completions, right? I mean, software engineers are pretty skeptical people like all engineers are. And, you know, no one thought that this thing would work and be any good. And it started working. And the interesting thing is it went from being a joke to being standard issue in like months. And so and now, of course, we've now gone all the way to even, you know, co-pilot with agents and co-pilots with autonomous agents. It's just unbelievable. You can't think of software development without sort of AI being part of it. It's kind of like I say, I would never be employable at Microsoft.

16:30but for the red squiggly in Word, because I can't spell. And so it's kind of like becoming like that when it comes to software tools. But the interesting point you make, though, is I go back to our mission of empowering every person in every organization on the planet to achieve more. I think that's what AI is about. I think that people have different notions of AI, but my fundamental view still is, even with all of the advances, not taking away from it, is it's still a tool. I fundamentally believe that we sometimes shortchange human ambition, human agency's ability to deal with unbelievable new technology that comes along once every 10 years, once every 100 years, once every millennium, whatever.

17:20Like even the most magical technology has been used only to help humans achieve bigger and greater things. In that context, if you go even narrow, I think we conflate sometimes what is knowledge work with knowledge worker, right? So, for example, if somebody had said in whatever, early 1980s, everybody in the world is going to be typing, people would have thought, it's crazy. Hey, we have a typist pool. Why the heck does everybody need to type, right? But guess what? That's what PCs did and phones have taken that to the next level where if somebody came from Mars and saw what I'm doing, they would say, oh, what, this guy is a typist, right?

18:01Because that's all I do. And so to me, that is what is going to be the transformation with AI, which is we are going to go to the next level of abstraction where what I want done, like when I use researcher every day today, right? Preparing for meetings, you and I had this chance, right? It's magical. It's like having the smartest analyst that is there with you 24 by 7, that you can ask the hardest questions and they come back having thought about it, quite frankly, not just an instant thing, with some great reports, which sometimes will have some mistakes, which will give you better insight. That ability to sort of work with your AI agents, to be more prepared, have more sense of accomplishment and more agency in your job, that's the future of knowledge work.

18:47In fact, I kind of described to Bill and I talk a lot about this, which is one of my favorite tools is Excel, right? When I first had the ability to say, God, I can think of numbers in rows and columns, it was a mind-blowing way to change things. And that's the way I think it's like when PowerPoint, Word, Excel, all became standard tools everybody could use to create information, express themselves, share information, I think AI is going to be the next big 10xer on that. I do think you're so right. I mean, I found even just frankly preparing for this, once it becomes part of your life, I was using Researcher this morning on three or four different projects.

19:31You suddenly step back from time to time and say, what did I do before this? I think there were just more questions that went unanswered every day. Absolutely. Absolutely. And so, but we still are at a point where the percentage of the world's people, even who are connected to the Internet, who are experiencing this, it's a minority compared to the percentage of people, the billions of people who will now need to experience this as we go forward. How do we reach the world? And I know both of us are big fans of Jeffrey Ding and his sort of theory of diffusion and what have you, because that's what I fundamentally believe, right?

20:11Thanks to sort of the rails, right, that exist, whether it's the mobile phone or the cloud infrastructure, it is possible. Because one of the other demos for me, Brad, which shaped at least my worldview was watching a rural Indian farmer use a WhatsApp bot that was built on top of GPD35 by an Indian developer, which translated that farmer's local language and need for access to some subsidy program and give that farmer real agency to be able to do things using that phone and that bot and that AI. that was built in the West Coast of the United States. And it was mind-blowing to me, which sort of the rate at which things were diffusing and also what it meant that it was meeting the needs of someone where they are at versus expecting them to do multiple things before they can even use it, right?

21:06And so if you say that's what is possible, then I think the key thing is to think about diffusion broadly. That means skilling broadly, right? It's not just about technology being accessible, getting people to be equipped to be able to use the technology with whatever skills they need, right? One skill could be just get over the fear of using something. And that's where learning by doing is perhaps one of the most important things that we can do is to get broad issuance of all of this technology around the world, get people to use it, then have the agency to change their work, their work artifacts and workflow.

21:47How I educate myself, how I get information, how I conduct commerce, all of these things are going to be changed. For us, it's never, never about celebrating tech for tech's sake. It's about celebrating what we can do with technology to create magical experiences that make a real difference in our countries, in our companies, in our communities. Maybe that's a good place to bring this conversation to a close because it almost brings us back full circle. Your life growing up was about these different themes, these different disciplines, ways of thinking that you wove together. That's been a part of Microsoft since the company was founded.

22:30It's been such an important part of your part of this trilogy, the three CEOs. Let me take you back to poetry. As we look to the future, as we think about AI as this new tool, is there any inspiration that you draw from, say, one of your favorite poems that frames where we might come? I'm thinking about a lot because in the context of our sort of 50th year and, you know, this sense of a new beginning, if you will, with this AI age and even this conversation around what does it mean to build for the future? I think that the two lines that are the best lines for anybody who ever wants to build anything is from Rilke, where he talks about, you know, letting the future enter us and changing itself in us before it becomes real.

23:22That's the most beautiful lines of poetry that capture, I think, the essence of what anybody who's a builder wants to achieve. Well, and I think there's something poetic, if I may say, about that. Because I think it keeps us grounded. It keeps us broad-minded. And I think it gives us inspiration to build things that are truly great tools for people. So thank you. Thanks for the conversation. Thank you so much, Brad. It was such a pleasure. Look forward to what comes next. That's right. Thank you.