Airbnb CEO on supporting OpenAI's Sam Altman： 'You learn a lot about people in a crisis'

You came out here in part to sit in a conversation with one of your great friends, Sam Altman, from OpenAI. And one of the things I thought was fascinating was you went pretty deep about what took place during what was a dark period for him and the work that you were doing behind the scenes to try to help him. What was really, for you though, the lesson of the whole thing when you were in it? I think that there's an old saying, the absence of information is filled with dirt. And I think that everyone was assuming the worst of everyone else. And I think when you really unpeel it, I think people had much better intentions. And I think there was just a lot of fear. Things were going so quickly. And as I uncovered what was going on, I realized there was mostly a lot of misunderstandings. A lot of they didn't have full information so people would assume the worst. And the other thing I learned was you learn a lot about people in a crisis. And they reveal themselves. And some people, I think, weren't the best. But I actually, and I say this as Sam's friend, but I also say this objectively, I was surprised, in a good way, at how he was not in self-preservation mode. When people are attacked, they typically go to themselves. And Sam was focused not on himself, but open AI. And that made me wanna fight for him. What was the worst moment of the whole thing, do you remember? The worst moment was when we thought that, okay, this is the worst moment. The worst moment was, it was probably 48 hours in. The board had led us to believe that he was gonna get reinstated. And right before midnight, we get information that Emmett Shear is now the CEO. In which case, it just seemed like the table, the page had turned. And Sam and Greg said, we'll go to Microsoft. And I said, you gotta fight. This is crazy, you gotta fight. And he said, we wanna do what's best for open AI. We wanna do what's best for the employees. And if what's best is for them to turn the page, we'll turn the page. And I thought that was the end of it. The miraculous thing was, it turns out, Emmett Shear is also a friend of mine. And that opened a new line of communication. And there were many other things that happened, but I think that allowed us just to have an open dialogue. So how involved are you now with open AI? I'm involved to the extent that Sam needs advice, and I try to help him in areas. Anything that's not core technology, if he has a question, I'm involved. So what are those things? How to product, design, maybe marketing, advertising, communications, that kind of stuff. Things that are maybe above the layer of the stack of technology. What do you do with the technology? How much do you think about the safety issues and also all the headlines that we keep reading about this debate that's clearly happening internally at that company? I definitely think about it, but I'm, to be clear, not involved in the company officially, so I am not inside the inner mechanisms. But I'm mostly focusing on giving him advice on how do you manage an organization? How do you make sure you instill trust? How much transparency is the right amount versus too much transparency? There's a lot of considerations. How do you tell a story to the public? How do you bring society along on what is no doubt a technological revolution? And we should be transparent about things that can go wrong, but we also gotta be careful about a lot of overly existential pontification when we don't really know what's gonna happen. And to also tell a story about, these are the practical five or 10 things this technology can do for our lives right now. So it's a lot of judgment. You went through your own fire, baptism by fire. Was there a moment that you look back on that you think is either a parallel or something you learned from that? The thing I learned is you learn a lot, I said you learn a lot about people in a crisis, but the other thing I learned is you only survive because certain people have your back and believe in you. And when the crisis happened, my co-founders, my board, totally had my back. If they didn't have my back, I might not have been as decisive. And if I wasn't as decisive, we probably wouldn't have come out of it the way we did. And so that lesson for me is when I saw it going, Sam said people need to have his back. And his company had his back, but people didn't quite know how to support him because no one was saying anything. And I just said, we're gonna lean in, we have your back. It was remarkable, by the way, how much the employees, for the most part, supported him. I think that tells you something. I think that tells you that if there's only one piece of data you should know, it's that the employees had the most information. And in that moment, I think they rose up and they said, this is what we wanna do. We'd rather be with Sam somewhere else than with OpenAI without him. Let's talk about Airbnb, but let's talk about in the context of AI. You can't go anywhere here without talking about AI. When are we gonna see AI built in in a real way into applications in Airbnb, but also applications across the board? It's 100%, great question. Over the next two years. So I think one of the observations I've made is ChatGPD came out basically like 18, 19 months ago. And I think we all had this feeling like there's a revolution and it's gonna happen. And funny enough, you pick up your phone, look at every app in your home screen and they're almost exactly the same. So AI hasn't actually changed our daily life. It's captured imagination, a little bit of our fear, but it hasn't changed our app. It turns out it's taken a little bit longer, maybe not exactly self-driving cars, but that last 10% is where all the gains are, right? If you get 90% of the tasks done and it's wrong, that last 10%, it's actually terrible. So I think next year you're gonna see a bit of a step forward and I think within 2026, you'll see a giant leap. And I think apps across the board, two to three years from now, 2026, 2027, they're gonna look really different.