Microsoft CEO Satya Nadella on the OpenAI Debacle
28:14 Share

On September 28th, the Global Citizen Festival will gather thousands of people who took action to end extreme poverty. Join Post Malone, Doja Cat, Lisa, Jelly Roll, and Raul Alejandro as they take the stage with world leaders and activists to defeat poverty, defend the planet, and demand equity. Download the Global Citizen app today and earn your spot at the festival. Learn more at globalcitizen.org.com.

On September 28th, the Global Citizen Festival will gather thousands of people who took action to end extreme poverty. Join Post Malone, Doja Cat, Lisa, Jelly Roll, and Raul Alejandro as they take the stage with world leaders and activists to defeat poverty, defend the planet, and demand equity. Download the Global Citizen app today and earn your spot at the festival. Learn more at globalcitizen.org slash bots. It's on!

Hi, this is Kara Swisher with a big GET interview. This time with Microsoft CEO Satya Nadella, the man at the center of the issues around OpenAI. It's a special drop episode of On with Kara Swisher.

Thank you for talking to me, Satya. I'm very excited to talk to you. It's great to be with you, Cara. So I've got a lot of things to get to. So I want you to sort of give me an overview of where you are right now and what's happening, because this has been quite a story for technology. I'd like to start with how you found out this was happening. I would say, you know, to your fundamental point, Cara,

I am exactly where I was on Friday morning, which is on Friday morning, if you'd come and ask me where I was and what was happening, I would say, hey, look, I feel fantastic about...

Microsoft and our ability to confidently go innovate in AI as it was in full display at our customer conference from Silicon to co-pilots and everything that we were doing and our commitment to our customers as far as the innovation there. And second, our, you know,

partnership with OpenAI and that we are with SAM. Those would be the three things I would have said on Friday, and that's exactly what I would say today. And so on Friday, I get a call saying that the OpenAI board had a decision where they were making a CEO transition. And we issued a statement right after that, which said they had appointed Mira, whom I had known for a long time. Obviously, she was very much part of the OpenAI team. And so we said, yeah, we're going to continue to work with OpenAI.

And so that's kind of what all happened. And so there's no real change, at least as far as I'm concerned. Yeah, I get that. I get that. But-

They didn't consult you, correct? Yeah. I mean, you know, we were mostly working with Sam and the management team and the for-profit entity, and we didn't have any relationship with the nonprofit board, which has the governance of this entity. That's correct. Right. And so did you feel like you should have been consulted? No.

At the very least, right, of course, the nonprofit boards have their own commitment to their mission. But for, you know, given who, you know, given all of our investment in open AI, it's not even the money and the capital. I mean, here's a simple way to think about this. You know, Sam chose Microsoft once. Sam chose Microsoft twice.

And someone's got to think about why. You know, there is no open AI without sort of Microsoft leaning in in a deep way to partner with this company on their mission. We love their mission. We even love their independence. We have no issues with any of it. We just want, and look, I've always, and I know in Silicon Valley, people talk about, you know, who is sort of, you know,

getting ahead of the other. I believe in partnerships. In fact, one of the most understated things is great partnerships can create lots of enterprise value. Sure. And so I deeply always have remained committed. And yes, they should have, you know, as a partner, I think it does deserve you to be, you know, consulted on big decisions. Right. Yeah. I think I probably found out almost when you did, which I rather

That's fine with me, but not with you necessarily. So your first statement on Friday was pretty diplomatic saying, and we're going to get to the bigger picture of where AI is going to in a second. But as you saw at Microsoft Ignite this week, we're continuing to rapidly innovate in this era of AI. And then you went on to say, we have a long-term agreement with OpenAI with full access to everything we need to deliver our innovation agenda and exciting product roadmap and remain committed to our partnership and to Mira and the team. So,

Mira, then they started negotiations. Were you, you were part of that from what I understand of getting Sam back because Mira had expressed interest in bringing them back. I think she moved up to that CEO position, but was already working to bring Greg and Sam back to the job. Yeah. I mean, overall, again, all through the weekend as, um,

I quite frankly didn't engage with the OpenAI board. The only person I know there is Adam D'Angelo. And I was consistent. Same thing I said, look, we really are very committed to OpenAI. We are very committed to Mira as the interim CEO. And we

We want to make sure that whatever Sam does, he'll definitely do with us. And you should know that. And we would be happy for Sam to come back if that's what you choose to. And I was very clear that in any which way that we would be supportive of Sam and Greg and all the team. Because the thing that we didn't want is the team to get splintered.

and the mission to get jeopardized. And our partners and customers, whether they're using OpenAI or Microsoft, but the point is we were very confident in our own ability. We have all the IP rights and all the capability. I mean, look, if tomorrow if OpenAI disappeared, I don't want any customer of ours to be worried about it, quite honestly, because we have all of the rights to continue the innovation, not just to serve the product,

But we can, you know, go and just do what we were doing in partnership ourselves. And so we have the people, we have the compute, we have the data, we have everything. And so, but at the same time, I'm committed to OpenAI partnership and that's kind of what I expressed to them. And where does that stand? Because you have invested, is it 13 billion or 12 billion in the company?

It's a complicated thing. Yeah, I get it. We have a significant investment. It sort of comes in the form of not just dollars, but it comes in the form of compute and what have you. And that gives us significant rights. It's expensive.

as I said. And also this thing, it's not a hands-off, right? We are in there, we are below them, above them, around them. We do the kernel optimizations, we build tools, we build the infrastructure. So that's why I think a lot of some of the industry analysts are saying, oh, wow, there is this, it's really a joint project between Microsoft and OpenAI. And,

And the reality is we are, as I said, very self-sufficient in all of this. - I get that too, but still you want to know when something like this is happening, 'cause they need to deliver the partnership, correct? - Correct. - I mean, you invest thinking that they can deliver what they promised on their side. - Absolutely, absolutely.

Was that a worry for you in terms of if they putting themselves in this? It was clearly a crisis. I mean, let's be honest. I mean, this was unprecedented. I've covered Silicon Valley. I've known you for decades. I've never seen anything like this. I mean, yeah, it's of course. I mean, anytime you have...

you know, significant turnover of a place, there's always risk. And, you know, there's a key person risk and then the team risk. And I mean, at the end of the day, we can talk a lot about technology and other things, but it's about human capital of any organization. And so, yeah, I mean, that's why, I mean, to me, that's where I go back to what,

My message over the weekend is, you know, I like to keep things simple. I start with, hey, what is the thing that we were doing on Friday morning? We were trying to build great products and great technology and do real research around everything from AI safety and alignment to the next generation models. I want to continue to do that on behalf of our customers and partners. And what's the best way to do that? Have the people.

And so let us optimize- - Exactly, walk out the door every day. Your assets walk out the door every day. - And optimize for the people because, and then, you know, obviously the boards have to do what they have to do around their governance.

But I don't know. I'm not a part of their board. So all I said is, hey, whatever it is that you're doing, just make sure that you don't compromise the mission of the organization in which we have invested and the people who are behind that that we bet on. And so that's it. I'm sure that was disturbing coming from you. But talk about Sam Altman. He talked about key people. He is the one that got the investment, who has the relationship with you. As you said, you don't know this board.

Um, talk a little bit about what you thought if he was, you've hired him now to create this new advanced research division. Can you talk a little more in detail about what that means?

Yeah, I mean, to me, you know, Sam is someone I've known for a very long time. And quite frankly, in the last, I would say, four and a half, five years that we've worked together closely now in this context, we've obviously gotten to know each other very deeply, not just as people who know as acquaintances, but as deep professional partners. And so I have great admiration for his vision, his ability to bring teams together and push together.

And so that's kind of what I want him to do, irrespective of what happens right now. I just wanted him to make sure that he's able to continue. And he's a very mission-driven guy. He wants to sort of continue to push this advanced AI industry

that he and Greg and team were doing. And so quite frankly, right now also, the way we will do this is that team will come and they'll join Microsoft and we'll continue to work with OpenAI. So nothing changes. After all, we had many engineers from Microsoft who are also working with OpenAI

you know, on Friday. And so we'll continue to do the same. Well, I don't want to, I don't want to jump ahead of myself, but more than 600 now or more of the employees want to leave at the company. So there's not much AI to work, open AI to work for if that happens. Yeah. I mean, this is where, you know, this is where I leave it to individual employees to decide and their board to decide what happens. But I, as I said, this is,

We will definitely have a place for all AI talent to come here and move forward on the mission. And we will be supportive of whoever remains even at OpenAI or whatever. Like, I mean, to me, I kind of look at it and say, hey, all three things have to equate for me. At the end of the day, I'm doing things that are meaningful for my customers. And I'm going to continue to partner with OpenAI. And I'll continue to make sure that Sam and team are supported.

Let's finish with Chugmat. So what is the advanced AI team? What does that mean? And how are you going to fund it? Where's it going to be? What is that? Where's it going to be? It's going to be like, you know, think of it like how we organize GitHub or what have you. It'll be, you know, a real team inside of Microsoft. One of the things we have learned, Cara, and I mentioned this, I think in one of the, whatever you call them now, XS, I think, the,

the idea that, you know, we learned everything from having created GitHub, LinkedIn, Mojang studios, how to have, you know, innovators and builders who come to Microsoft and then continue to have an identity, a culture, a spirit to do things. And so therefore that'll be, that's kind of how I look at it. So having an advanced, I mean like Microsoft research, I mean, think about this, right? Microsoft research has a strong heritage and,

of doing some of the foundational works. In fact, some of that was even in display, right? As we talk about large models, one of the leaders of what we call small language models is Microsoft. And some of the stuff that we announced at Ignite. So I kind of felt like, hey, great. If Sam and Greg and crew can come, we can now create another specialized unit, very focused on what we were doing as next generation. We were already talking about what comes in 25, what comes in 30. What kind of funding does this get?

The beauty of anything inside Microsoft is I don't have to talk to you about it. But I'd like you to tell me. That is the beauty of it. It's just the small numbers that get talked about. Well, you might have some billions laying around if everybody leaves OpenAI. The silliness of it all is the little acquisitions we make and the little investments we make get more press.

than all of the money, which is the organic investment we make, but that's the world and that's okay. I know you're doing a lot of research, but I think you, you know, a lot of people and I am among them think you just possibly got this company for free. This, you know, if all these employees go over, were you surprised by how many employees were backing Sam? Yeah, I mean, I mean, I think speaks to, you know, Sam and his mission driven approach to building teams, which obviously I admire a lot.

And, you know, I think that, you know, if, you know, if that is the eventuality, then great. But, you know, I don't think of this as, again, like, hey, somehow we are trying to use this to get something for free, as you described. I mean, we were happy on Friday morning. Yeah, I get it. So I have no interest. I think, let me say, you're in a better position. You're in a better position. You are.

I was in a good position, Cara. This is where I think most people sort of overestimate, underestimate, like there is nothing that has really changed.

On Friday, I could have done exactly what I'm going to do today. I think that, yes, it's sort of better copy, but there's no real change. In fact, on Friday, I was more productive than the weekend. Let me put it that way. Yeah, of course, of course. So this new unit, it'll sit within Microsoft. It could have quite a lot of the OpenAI employees who have been promised jobs. Is that correct? Yes.

That's correct. I mean, we said, yeah, anyone who wants to join us, because one of the things I want to be, let's face it, right? There's a lot of competition for AI talent. And I just want to make sure in the AI talent wants to go where there is significant compute and resources and data and all of those assets that we have. And so therefore, I just wanted to make sure that in a world where people have choices, if they're going to leave OpenAI, that Microsoft was a choice for them.

Right. You know, Mark Benioff's trying to hire. He just sent me a tweet. He's trying to grab some of those people. I'm sure everybody is. Do you know, as you're working, but you continue, you will continue to work with OpenAIM no matter what happens. Do you know the interim, the newest interim CEO, Emmett Shearer? Yeah, I have, you know, I had not known him before, but I met him for the first time last night.

And I've had a couple of subsequent conversations and I relate essentially the same message that I just told you, Cara, which is we are committed to our customers and the innovation and them and Sam. And he knows that and, you know, and he's good with all of that. And that's why I even put out the statement yesterday, last night, when I first came to know that he was going to be the new interim CEO. Yeah.

I get that, but you were more excited for the SAM acquisition, I would say. I'm going to call it the SAM acquisition just to annoy you. Is he in this capacity at Microsoft? Could this change if the board leaves and he goes back? Are you open to that possibility with more Microsoft employees?

bored presence? Oh, yeah. I mean, look, I mean, that's not obviously my... So if something happens where they're bored and people decide that, you know, they want to get back, as I said, to some kind of a state. And one thing I'll be very, very clear is we're never...

going to get back into a situation where we get surprised like this ever again. And so therefore, that's done. And so therefore, if anything, we are much more resolute that we cannot have our customers and so on feel like there is any, you know, going to be surprises. So therefore, yeah, we'll definitely take care of all of the governance issues and anything else

And as I said, we have all the rights. So therefore, we will make sure that if we go back to operating like on Friday, we will make sure that we are very, very clear that the governance gets fixed in a way that we really have.

more surety and guarantee that we don't have surprises. That's all. Don't have surprises. That mean a board observer or board seat? We'll sort of cross that if that happens, you know, but I think that we will, we'll need to make sure that we, our interests are solid. And it's not like I felt that on Friday morning, somehow our interests were not solid, but we've learned a lesson or two and we'll make sure that we will double down on. And as I said, I think the key thing, I think at least for our,

you know, our customers is to know that this is where I look at it and say, is like, if even any, everything disappeared, we have everything we need in order to just continue to keep innovating without a blip. And that I think is the one thing that is not understood by the world. You know, we wouldn't be Microsoft at, in 2023 with the capability up and down the stack, if we didn't have that. And it's silly for people to keep thinking about like, Hey, one company with little technology is going to make all the difference. We are a lot bigger. We love the partnership.

but we are not single-threaded. We'll be back in a minute. This episode is brought to you by Shopify.

It is over.

Were you worried about any of the allegations the board made, including Ilya Sutskever, who has now signed the letter saying...

and also noted he deeply regrets his participation in the board's action. But there were some allegations that were, I couldn't find out what they were. It was unspecific. Did you have any worries about that relating to Sam? I did not hear anything. For me, first of all, Ilya is an amazing scientist. First of all, he was another core person at OpenAI that I admired a lot and I still admire for all the unbelievable work

he has done. And, you know, I was surprised. I mean, this is all I'm also just reading. I don't know anything about his opinions about what happened or not and so on. But nevertheless, I've not heard anything about

And if the question is about safety, oh my God, that matters. I know to Sam, it matters to OpenAI. In fact, one of the things that I think brought Microsoft and OpenAI together was, hey, look, we're going to build AI that can broadly be distributed so that people can take advantage of it with products like GitHub Copilot or ChatGPT or...

co-pilot in M365 or what have you. And at the same time, we think about alignment, long-term, I mean, both the immediate term safety and guardrails and the long-term safety research. And these are all, we are allocating a lot of- You had no worries about Sam Altman that the board had raised. I, you know, I'm not aware of any of those things. And even in my conversations with them, none of them were raised to me.

Yeah, a lot of them said there was no Malfi. It certainly left things unclear. That was my issue with it. It was so non-transparent and unclear if you're making those kind of allegations, calling someone a liar, essentially. And your hope is that you do this thing with Sam and also OpenAI stays a company or he moves back to OpenAI and then you stay with more observation.

at least observation of the board, et cetera. I mean, fundamentally, I want us to just continue to

to serve our customers with a roadmap that I'm very confident about and committed to. And so that's the number one thing and perhaps the only thing. And in order to do that, on Friday morning, we were partnering with OpenAI and SAM. And today we are going to be working with SAM and OpenAI. Same thing. Will that be the case if this board stays? I don't think they will. That's my...

That's my opinion. I mean, to me, it's not my decision. So therefore, I really don't...

have an opinion to express on that. It's for them and, you know, others, constituents who have any say on that. But, you know, to me, I was comfortable partnering with OpenAI with their board and I'm happy to sort of keep working with OpenAI and they'll, you know, they'll have to sort out their board and what have you. Okay. I have one more specific question. Does this, would SAM become a

I'm not trying to replace you, Satya, but a potential CEO of Microsoft in this case. This is where your future is going. You know, let's just say that it's fantastic to have lots of capable people who can become CEOs of Microsoft. Right. And do you think about that now? Have you been there how long? More than, oh gosh, a long time. I have been here for 32 years and I'm having the time of my life. As CEO, is it 12 years? No.

No, no, no. I think I'll be coming up on 10 as CEO in February. Right, right. But you think of those things. Obviously, you're part of a big company. Not enough CEOs certainly do. But is that something on your mind of where the company's going? Yeah, I mean, look, at this point, as I said, the thing that I'm most focused on, look, is what is it that Microsoft is doing? And am I able to contribute to build the organizational and institutional strength of this company? Yeah.

And having fun doing so. And I'm absolutely doing that. And then when I think about organizational and institutional strength is what you just said. Like at the end of the day, I'm grounded in the fact that I was growing up in Microsoft. I was not sitting around thinking, oh yeah, someday Steve will retire and I'll become CEO. I was happy in my job and continue to work on it.

And then- Gosh, the rest of us were thinking that, but go ahead, keep going. But to some degree, that's kind of what I think is gonna happen. But both people inside the company and people who are close to Microsoft and who are in our orbit, so to speak. And so, yes, I'm thrilled to be partnering broadly, getting more people interested in becoming CEOs of Microsoft, let's say.

Okay, so all this puts us spotless. I don't want to talk the broader issues around AI because one of the things that set up that I had reported early on was this misalignment between the nonprofit board and Sam's efforts and this split between those who see the AI world, between those who see the business opportunity and a little bit of hype and others worried that,

about the dangers of moving too fast. And de-sell is a word that gets thrown around. What do you think about this? Where do you stand? Because it seems like this binary idea, and of course you read Marc Andreessen's

whatever you want to call it, about, you know, if you're either with us or against us, what I thought was simple-minded, but you don't have to comment. But how do you look at it? Because it seems like that's the split happening. Does there need to be that? Or is there this sort of, it's going to be great. And also there are dangers. How do you see it? The interesting thing is the word you use, the split. I mean, the way, at least the optimism I have for perhaps,

this generation of technology is the fact that this debate exists, right? I actually welcome this, right? Which is, hey, look, at the end of the day, you know, as Brad Smith would call it, every piece, every technology ever invented was both a tool and a weapon. And the question is, how can you really make sure that the use of the tools gets propagated and the weapons not so much? And so therefore, us to be able to make sure that

we are having that conversation right up front versus sort of dealing with the unmitigated consequences subsequently, it's got to be welcome. So I have been, I don't think of this as a split. I mean, like I think about also this, there's a timeline, right? There is the existential issues in the long run if you lose control of AI and that requires real thought and real safety research, real alignment research that you have to commit to.

And then the real world harms today, right? Whether it's election interference, deep fakes, bias, bioterrorism, the amount of work we did, whether it's for Bing chat or chat GPT or co-pilot to make sure that harmful content doesn't get created.

So I feel that we have sort of got to work all ends of it on a daily basis. It's not just not talking about it, learning the engineering process, having these principles guide the engineering process, and then having oversight and safety boards and with real teeth. And that's all something that we constituted even with OpenAI. So-

So I am very committed to responsible, safe AI development. So, and at the same time, driving the benefits of it, right? I mean, you know, to me, when I see that rural farmer in India who's able to use something that was produced in the West Coast to the United States to have more agency in their life,

That's magic. When I see sort of, you know, somewhere in the Middle East, a kid can have a personalized tutor because of what we built. That's magic. And I want that benefit. And at the same time, let's make sure that all the harms that people talk about and we take them seriously.

I know you have to go. I have just one more quick question. So the India-Australia cricket thing. So I'm so sorry for you there, but are you thinking of buying Australia now? That would be like buying OpenAI. None of this can be accomplished, but we can partner well with OpenAI and enjoy Aussies play cricket. All right. Thank you so much, Satya. Thank you for the time. I really appreciate it. Thank you so much. Thank you, Cara. All right. Bye. Bye.

Today's show was produced by the ON team, Naeem Araza, Christian Castro-Rossell, Kateri Yochum, and Megan Burney, and the Pivot team, Lara Nauman, Zoe Marcus, and Taylor Griffin. Special thanks to Andrea Lopez-Cruzado, Brandon McFarlane engineered this episode, and our theme music is by Trackademics. If you're already following the show, you get to pull off the biggest coup in tech. If not, your company will go down in flames. Go wherever you listen to podcasts, search for ON with Kara Swisher, and hit follow.

Thanks for listening to On with Kara Swisher from New York Magazine, the Vox Media Podcast Network, and us. We'll be back on Monday after Thanksgiving with more.