Reed Hastings on Disruption Culture and AI Blackstone CEO Conference

Well, good morning everyone. So today we are joined by a leader who has transformed multiple industries, often actually before many of us even realized that they were changing. So, Reid co-founded Netflix in 1997, CEO for 25 years, and guided it into a global entertainment powerhouse with over 300 million subscribers worldwide. But his impact goes far beyond entertainment, founded Pure Software, took it public, serves on the boards of Bloomberg and Anthropic, where you're helping shape the future of information and the future of AI, which we'll touch on a little bit. And he's even building a new community at Powder Mountain, which I'm very excited about. And through it all, he's remained a committed education philanthropist, leading the California State Board of Education, supporting organizations like KIPP, City Fund, and the Charter School Growth Fund. So I'm very excited to welcome founder, builder, and visionary who keeps pushing boundaries. Reid, thank you so much for joining us today. What a pleasure. So, Reid, you know, we talked a lot yesterday and our CEOs today are all wrestling with this idea of disruption. You know, whether that's from competitors, market disruption, trade policy, artificial intelligence, you've lived it, right? You were in an industry over the last 20 years that has seen no shortage of disruption. You've even, you know, self-disrupted your own business. So, you know, you went public at $15 a share in 2002, just announced a 10-for-1 stock split at $1,100 a share. Talk about that a little bit, right? Like, what were some of the pivotal leadership decisions you made, some of the hardest calls as you navigated such a transformative sector? You know, I loved Hewlett-Packard, I loved Sun Microsystems, Wang, and they all died. And so in Silicon Valley, you really, it's very personal, you become a student of like, how do companies with good hard-working people die? And mostly it's this subtle thing about disruption. So if a competitor has a product that your customers prefer, you know how to deal with that. You've got to up your game and you fight that, and that's kind of normal, non-disruptive. And then if the competitors are serving a segment that's a little different than yours, you tend to ignore it, you know, it's not in our core market. And what you have to think is, what is the capabilities going to be in 10 years of that competitive approach? And it comes down to this strange exponential phenomena that we have in semiconductors and then really in software. And most of the world, let's say an automobile, the other competitors get better at a steady rate, and so you can kind of watch them, you know, move and, you know, it happens in a long time. And in software and hardware, you get these accelerations, basically, a low market disruption that then just explodes into your market and then very hard to adapt. So you always need to be watching for that and anticipating. And in our case, it's a classic thing. We were mailing DVDs at Netflix, and they took two or three days to get there. And Blockbuster was a well-run consumer-focused company. They did big roll-ups. They were the leader of the industry. John Antioco was very thoughtful. He could be in this room, and you'd see him as a thoughtful peer. But they'd only ever experienced kind of linear phenomenon that way. And they looked, and they went to their customers and said, how do you feel about renting by mail? You get good selection, but it takes two or three days to get to me. And most of their customers were like, two or three days to get to me? Forget it. And so they were like, not a big deal, Silicon Valley, waste of time. And they put it to the side for a number of years. And then we kept growing, and then we refined it. We opened up mailing distribution, nothing fancy, but we opened up 50 distribution centers, changed it to overnight delivery, got better in stocks and these things. And then suddenly, it was a mainstream proposition for at least maybe a third of their renters. And then it became mainstream for two thirds. And then them being retail, you only have to take 20% or 30% of the business and you tip them from profitable to unprofitable. They start closing stores, and then you get the secondary tipping effect. So even in something as simple as DVDs by mail, you can get these disruptions by not anticipating what this solution could be in five years, which is the core thing. And a lot of people like Andy Grove's phrase, only the paranoid survive. And in some sense, that's good, like, you know, always be aware what's going on. But in another sense, the paranoid are delusional, right? They see a lot of things that don't exist. And that undercounts what we as operators have to do. So like all of us have to sort through which things are really going to become material big challenges in which are noise. So it's sort of like playing chess. And in chess, you have to think, you know, five, six, seven moves ahead. But our human brains can't do five, six, seven moves ahead on every path. And so we have to figure out which paths are likely not takes a lot of judgment to become an expert chess player. And if you're wrong about which path and the other guy checkmates you. So this is a lot like our art in business, which is we have to always be sorting through which threats can become material over five years, and which are very unlikely to. And so our challenge is much more like being a chess grandmaster than it is as simple as let's be paranoid and then we'll live forever. So that is the challenge around disruption. Talk a little bit about the decision to disrupt your own business and to go from DVD by mail to streaming. Frankly, before streaming was what everyone is now accustomed to. That was actually surprisingly straightforward. Mostly when you build a business, you fall in love with that business doing what it does. And when we started with DVD, we always knew it was a bank shot to get to streaming. And so we never were in love with DVD. In fact, we bemoaned it, you know, the plastic would crack, we're dealing with a post office, crazy things to simulate the digital distribution network. So when we could finally actually deliver over the internet, it was hallelujah. So we had a little bit of an advantage of not hanging on in the romantic way to the business that made us in that. And then it was, you know, set of big challenges to do that and to become profitable in it, right? It's kind of easy to be in a new business. It's not easy to make money on it. And so that took a number of years. Understood. One more question on this, before we move to artificial intelligence. You know, at Blackstone, we talk a lot about our culture and the importance of culture driving, compounding at scale. And, you know, your culture deck back in 2009, was it? I think actually went quite viral. And you have a very innovative culture at Netflix. Can you talk a little bit about how that has helped you navigate a lot of this disruption and stay ahead of the innovation curve? You know, some people say culture eats strategy for lunch. I'm like, why are you putting them opposed? You want great strategy, that's super important, and great culture. And they're both critical. For us, it was really understanding that we wanted to be performance oriented. And to do that, we had to break out of the family metaphor. So most companies throughout history have been family companies. Most countries have been family countries, kingdoms, you know, from time immemorial. And that's the natural state, okay? And then you've got this thin layer of corporate. And so people really think, and then many leaders say, we're like a family at work. But the truth is, you're really like a professional sports team. You want to assemble the set of people that are going to win the championship. And we're all working as hard as we all, we're all in this room because you want to win the championship in your field and make a big contribution. And so we were early on saying team, not family, which is, in a family, you really want people to be nice, and you want your kids to be nice and polite, and you want them to be loyal. But at work, we want people to be very honest, okay? And we're not trying to get them to be un-nice, but you've got to put nice aside as the primary and say honesty and constructive confrontation is the positive thing. So it's creating a value system that's around teams, and therefore you change out, you know, everyone knows that professional sports, you've got to play for your position, you know, every season, maybe every game. And that's challenging, but that's the key metaphor. And so then we built a whole infrastructure around a set of values and things around that, so that we would be judged by our employees on, were we a good team, rather than were we family-like, which we definitely were not. So the theme of this conference has been artificial intelligence, and with your, you know, seat as one of the board members of Anthropic, Netflix being one of the first AI native businesses, what's your view on superintelligence, right? We get views that this is two years out, some are 10 years out. Where are you on this? Well, I've been very frustrated by AI for 40 years. So 40 years ago, I got a degree in AI at Stanford in that first wave, and the first wave of AI was expert systems, and we didn't know it at the time, but it turned out that that's the sun goes around the earth school of AI. Like, the theory was just raw, okay? We were all about reasoning, not about learning. And neural nets was around, but an obscure little thing. And then the field of AI or expert systems died. We all went on to other things like Netflix, and that's been super fun, but then to see neural nets really rise on the back of GPUs and become, you know, incredibly impressive has been so exciting. But it definitely presents a lot of challenges. I'm pretty convinced that AI is going to keep getting better and better and better. There's sort of two theories. One is it'll be like Moore's law, okay? And intelligence is just going to go like that. And it spikes because eventually the AI starts to be self-improving. So you get it to a certain level, and then it programs the hardware for itself, it programs the software for itself, and, you know, it really gets unbelievable. The other one is intelligence like cancer has evolved over millions of years. And so the war on cancer has been very frustrating and slow because cancer has so many different ideologies, so many different mechanisms. So we cure one part, there's another part that really is hard to address. And it's been incredibly trench warfare and not Moore's law like at all in our therapies and life outcomes. So intelligence has also been developed over a very long time. Think about spatial reasoning and emotions and all these things. So it may turn out that AI is a lot harder to get to real human level intelligence than we thought. That's the kind of like cancer theory versus the like Moore's law theory. And we simply don't know right now. But whether it takes 50 years in the second thesis or five years in the Moore's law thesis, it's kind of a drop in the bucket, okay? Basically somewhere between five and 50 years, it's really going to be better than all of us at nearly everything. And that's traumatic for society in a deep way. The only good news here in terms of that trauma is software engineering will be first. So AI is going to eat its own. And either software employment will decline a lot over the next five years, and you will all be able to watch it. In which case, you know, lawyers and accountants and investment managers are next. Or because it's lower cost and higher productivity, we'll write a lot more software. And so that's another possibility that we get the productivity bonanza. In which case, we can put aside the great warnings of the white collar collapse. But software engineering, again, is essentially the canary in the white collar mine. And so that's something just to continue to watch and see what happens. And, you know, over the last 10 years, radiology has been the field that I thought would get wiped out. Because of the image processing, it's very good for computers to do. And the computers are better. But partially because of regulatory, partially for human reasons, the typical radiologist today does five times more scans than they did 20 years ago, okay? Because they're using the AI, but it's going through them. And their field has not gotten devastated in terms of income and employment, okay? And we humans are willing to do five times more scans because the costs have come down. And so that would be this elastic, positive reaction that I hope we see in the rest of the economy. Because then it means incredible productivity and outcome from the AI. So you're in the more optimistic camp about productivity gains, implications on employment. I'm in the profound uncertainty camp. So I would say it's 20 percent that we're going to have significant white collar unemployment and then have to deal with the political mess that that will create. And maybe 80 percent that it's the positive elasticity. I'm not confident of either one. I think we have a lot to learn. And I think we need to be able to talk about the negative, not obsessively, but to inhibit it. But it's something in the order of 20 percent. Can we talk a little bit about the impact of AI on content and content creation? So we at Blackstone, we ran a commercial, it's a Eureka ad, you guys may have seen it, Picks and Shovels. You know, the team actually on our eighth floor replicated that with AI at a fraction of the cost. And do you see a world where AI is driving a lot of the scripts? Or do you think human audiences are going to need that human touch in order to be able to connect with it? Sure. Well, you'll automate the video creation part for sure in the ad. But think of all the work that goes in up front of that, which is we're making this ad that's going to make us feel like we're Picks. Everyone knows the story that the Picks and Shovels companies did well in the gold mining. And we're going to make that our metaphor as Blackstone. So there's a lot of creativity that goes into that big judgment call. And then, yes, making frames of video is pretty easy. But you really, you've automated a minor part of the total picture. And in the past, people thought, you know, we used to, for filmed entertainment, take it on cellulite film and then go chemical process it and then edit it by hand. And everyone thought, oh, digital filmmaking is going to massively reduce costs. And it didn't. What happened is we said, let's do more takes. And let's do fancier editing and jump cuts and fast things. And let's make entertainment better. And so that's another case where the economic benefits were not low cost. They were better output. And I'm going to guess your advertising budget is not going to shrink. What you're going to do is more creativity to do even better ads. Let's do three versions and test them, those kinds of things. So if you're looking at it primarily cost savings, I would say that's the unlikely path as opposed to, again, more productivity oriented. Reid, thank you so much for the time. Awesome. What a pleasure.