Letter #156: Bret Taylor and Clay Bavor (2024)
Salesforce Co-CEO & Facebook CTO and Head of Google Labs // Cofounders of Sierra Technologies | AI Startup Sierra Comes Out of Stealth
Hi there! Welcome to A Letter a Day. If you want to know more about this newsletter, see "The Archive.” At a high level, you can expect to receive a memo/essay or speech/presentation transcript from an investor, founder, or entrepreneur (IFO) each edition. More here. If you find yourself interested in any of these IFOs and wanting to learn more, shoot me a DM or email and I’m happy to point you to more or similar resources.
If you like this piece, please consider tapping the ❤️ above or subscribing below! It helps me understand which types of letters you like best and helps me choose which ones to share in the future. Thank you!
Earlier this week, Bret Taylor and Clay Bavor publicly launched their new startup, Sierra Technologies. After a year of stealth, they announced a new round of funding and spoke to Jon Fortt, revealing what Sierra is, the simple thesis behind the company, raising $110mn (from Benchmark and Sequoia), meeting at Google, Clay’s experiences at Google, Bret joining the board of OpenAI, the need for AI infrastructure, their relationship with customers, in-person vs remote work, science vs engineering, the future of artificial intelligence, what success looks like for Sierra, and their hopes for the future.
Bret is one of the cofounders of Sierra and has one of the most illustrious careers in Silicon Valley. He started his career at Google in 2003 (pre-IPO), where he famously rewrote Google Maps in a single weekend (this story is definitely worth a read). He left Google in 2007 to become an entrepreneur in residence at Benchmark Capital. That same year, Bret founded and served as CEO of FriendFeed, a social network best known for inventing the “Like” button. Just two years later, in 2009, Facebook acquired them for $50mn. Bret stayed on at Facebook, and within a year, was Facebook’s CTO. But after two years as Facebook’s CTO and Facebook’s IPO in 2012, Bret jumped back into the startup world with Quip, a Google Docs competitor. After five years, in 2017, Salesforce acquired Quip for $750mn. Once again, Bret stayed on and was quickly promoted. He started as Chief Product Officer in 2017 before being promoted to President and Chief Operating Officer in 2019 and Co-CEO in 2021. He left Salesforce in 2023 to launch Sierra. Besides his day jobs, Bret joined the board of Twitter in 2016 and became Chairman in 2021, where he oversaw the company’s sale to Elon Musk. After Sam Altman was (briefly) ousted from OpenAI, Bret joined the reconstituted board of OpenAI as Chairman. He also serves on the board of Shopify.
Clay Bavor is a cofounder of Sierra. Clay started his career at Google in 2005 (post-IPO), where he first met and worked with Bret. Whereas Bret went off into the world of startups, Clay stayed at Google for 18 years, where he worked on Search and many of Google’s advertising products as he worked his way up from an associate product manager to the VP of Product Management, where he lead product management and design for Gmail, Google Drive, Google Docs, and Google Apps for Enterprise (now Workspace). While VP of Product Management, Clay started and led Google’s AR/VR effort and several related projects, including Project Starline and Google Lens. In 2021, they rebranded and consolidated to Google Labs, which focuses on forward-looking bets for Google. In 2023, he left Google to start Sierra with Bret.
I hope you enjoy this conversation as much as I do!
[Transcript and any errors are mine.]
Related Resources
Company
Meta Podcast (2 hours) [Free]
Meta Deep Dive (Full report: 166 pages)
Meta Business History (~18 pages)
Letter #77: Mark Zuckerberg (2012)
Meta Investors
Peter Thiel Compilation (536 pages)
Marc Andreessen Compilation (1386 pages)
Salesforce
OpenAI
Benchmark
Letters
Compilations
Bill Gurley Compilation (652 pages)
Mitch Lasky Compilation (474 pages)
Sarah Tavel Compilation (300 pages)
Sequoia
Letters
Compilations
Michael Moritz Compilation (642 pages)
Don Valentine Compilation (347 pages)
Transcript
John Fortt: All right. Welcome to a Fortt Knox update. And I'm calling it that even though I haven't spoken to Clay before. Brett and I have done the hour long one on one. Welcome to a Fortt Knox update. Brett Taylor, Clay Bavor here. It's coming--the startup is coming out of stealth. Brett, you and I talked about this some months ago--I don't even remember how long ago. Before you were chairing the board of OpenAI, before some other things happened. You told me you were working on an AI startup. And here it is. So Clay, nice to meet you. Guys, tell me about Sierra. What is it? And how did you raise $110mn for it?
Bret Taylor: Well John, it's so great to reconnect. Sierra, we're creating the conversational AI platform for businesses. Our thesis is really simple. We think that conversational AI is perhaps the most important consumer technology trend in a generation. And now if you're a consumer brand, the most frictionless experience isn't installing an app, it's not clicking a link, it's just having a conversation. And every brand is going to need to make their own conversational AI. And that's what we're trying to do it at Sierra, is to enable every company to create their own AI. We've had just an incredibly fun time in stealth mode, as you said. We've been partnering with incredible brands like WeightWatchers, and Sonos, and OluKai to create their own AI. And now we're excited to bring this technology to the world.
John Fortt: So, Clay, give me an example of how a company like Sonos would be able to use its own conversational AI in a way that makes sense, especially because I've got a Sonos at home, have for years, it's linking up to multiple streaming services, maybe Spotify, Amazon Prime, Apple Music, etc. If you're looking for a song, building a playlist, you've got to deal with those services. So how does having conversation with Sonos help the experience--AI driven with all those different services?
Clay Bavor: Sure. First of all, John, thanks for having us. I'm so glad to be here. We have built Sierra to enable our customers to create customer facing AIs that can handle a wide variety of tasks. And so for instance, with Sonos, we're working with them to create a conversational experience for device setup and troubleshooting so that if a listener is trying to configure new speakers, add a device to their system, they can just have a conversation and ask questions just as they would to a person. And our other design partners, and early customers like WeightWatchers are using their virtual agents to support customers in a wide variety of tasks, from changing subscription plans to getting advice on points and meals and more, with great success early on.
John Fortt: So, about the raise and about the vision--you guys met at Google. And, as Bret, you and I have discussed, your rise through Google coincided with the rise, at first, of Web 2.0, which, for those who aren't familiar, the Gmail experience was revolutionary because the user interface was a lot more flexible than you could get on other browser based mail clients. Google Maps blew people's minds because before was Mapquest and it was very static and here it was flexible and you can zoom and you can layer in different experiences. Is AI--I know it does more than this--but does it change the user experience in a way that's parallel to the way Web 2.0 and those--in a lot of cases, Java and JavaScript based approaches revolutionized the web experience?
Bret Taylor: John, it's a great question. And I would argue it's actually even more significant than Web 2.0. We think that the impact of conversational AI on consumer experiences will be on par with the internet. Fundamentally, in 1995, you needed a website, if you wanted your business to exist digitally. Maybe in 2005 you needed a profile page. Maybe in 2015 you needed a mobile app. Well now, every company in the world needs an AI agent. Because your customers just want to have a conversation with you. And what's so interesting, if you saw, ChatGPT was the fastest consumer service in history to get to 100mn consumers, it's because we're building in all of these previous technology trends. And we think that the adoption of conversational AI will dwarf them all. But the challenge for businesses is it's really, really hard. You shouldn't have to have a PhD in artificial intelligence to make your business succeed in this new world of conversational AI. And that's why we created our company. We want to enable every business, whether you're a consumer brand that has a lot of technologists on your firm, or consumer brand with none, we want you to enable your business to succeed in this new world of conversational AI.
John Fortt: And Clay, I know what Bret has been working on since he left Google, we talked about that--the startups, and Salesforce, etc. You were working with Google Labs. Tell me about the experiences that you continue to have there, and why you decided to work on this?
Clay Bavor: Yeah, first of all, I'm so grateful for my experience at Google. It's an incredible company. And what Bret and I saw, with recent advances in AI, was really a sea change in technology that we think will enable some fundamentally new and better experiences. And when there is a sea change and technology like that, it affords new opportunities for younger companies in particular, to explore the space, to take some risks. And that was very compelling. And then there was also just the element of building a company with a friend of mine-- Bret and I have known each other for almost 20 years--that was quite special. And so that is what--the opportunity I went to.
Bret Taylor: John, in fact, my wife interviewed Clay at Google, if you want some dirt on him.
John Fortt: Wow, um, I don't even know where to go with that. There's a good quip or joke in there somewhere, but I don't know. I'll--get back to me on that. So, I want to, Bret, go back with you to joining the board of OpenAI. Because that happened after a moment that was, some have argued, disruptive and maybe even transformational, for the AI space, in the AI conversation. I'm talking about back in November, when Sam Altman was pushed out of OpenAI by the previous board, and then brought back when the board reconstituted. I believe the first major change to the board that was announced was that you would be joining. In that series of days, weeks, after the initial disruption, I remember having Connie Loizos, the editor of TechCrunch, on with me on CNBC. And she described that as an SVB, Silicon Valley Bank level moment, for the AI ecosystem. Because people sort of realized we can't have all of our eggs in just one AI model basket, even though OpenAI and ChatGPT were getting so much attention. How did you think about that moment, from the Sierra perspective? And what are you building on top of, model wise, philosophically, as you build Sierra, maybe based on that?
Bret Taylor: Well, I'll start with just the personal. I, like so many others in the technology space, was just really worried about OpenAI that weekend. Clay and I left our jobs to work in this space, because we're so passionate about the potential of this technology. And you have this incredible mission-driven nonprofit, and it felt for a moment like it might not exist. So I chose to take on the role of Chairman in large part just out of a sense of OpenAI's mission being so important for the world. And at Sierra, we're building a solution for businesses, so we actually build on a variety of different models. And in fact, if you chat with the OluKai agent to exchange your shoes, there's probably four or five different AI models that are invoked to decide what to do at that point. So like so many others, we're a part of this ecosystem. And what I would say is, it's pretty amazing to see the innovation in what Reid Hoffman would call these frontier models, these models that require incredible amounts of capital expenditure, probably the best researchers in the world. And I think Sierra is a benefactor of all that research and development. And as I said, when I think about it from Sierra's perspective is, I really think of frontier models as pieces of infrastructure that are becoming fundamental to software, really trying to package up a solution so that businesses can actually solve real world problems with that infrastructure without being experts in the field.
John Fortt: Okay. And Clay, another piece of infrastructure that's very important to the AI ecosystem is accelerators, the chips involved. And right now, as reflected in the stock price of Nvidia, the stock rise of Supermicro, and others, there are fewer accelerators, there's less hardware available to fuel this then people would like. How does that influence what you need to raise money to spend it on in order to build Sierra, and where do you see that heading?
Clay Bavor: It's certainly something that we think about. We're somewhat insulated from it, however, by the fact that we use these frontier models built by the likes of OpenAI and others to run the core pieces of our infrastructure. And so we're able to scale through them. And also, as Bret said, our architecture is built in such a way that we can be flexible across which models we use, and use a number of them. And so, I would say we're one step removed from it, but it is always on our mind.
John Fortt: So, you're a step removed from it. Are your customers as removed from it for the same reasons?
Bret Taylor: Well, John, it's a great question. And I think if you look at just the build out in semiconductors, you mentioned Nvidia as a supplier, the build out in all the infrastructure as a service platforms to provide this technology, no one's completely insulated from it. There's definitely a shortage in the computation required for both the training and inference in these AI models. At the end of the day, though, we've really chosen an architecture that really tries to maintain as much flexibility as possible so that we can insulate our customers from the day to day disruptions that this infrastructure shortage represents.
John Fortt: Okay, and finally, what's next, you're announcing this $110 million raise. To what degree--actually, let me not say "finally," because I asked you about this before we actually started rolling. And I asked about the structure of the company--you guys are sitting in CNBC San Francisco Bureau. But it's 2024. Commercial real estate is expensive. I was wondering--is the company based in San Francisco? Is it based anywhere? And you said, Actually, you're an in person company, based in San Francisco.
Clay Bavor: John, we are proudly based in San Francisco. And we are an opinionatedly in-person company. Bret and I felt from the earliest days of the company, building a new company, a culture and everything that comes with that, that it was really important to be together in person. And so we are. Our employees love it. We in fact have lunch together every day as a team, it's one of the things that I think we both really look forward to in a given day. And so we're very much in person.
John Fortt: Either one of you, tell me more about that. Some people would believe that technologies like those that Google has given us, that Salesforce has given us, that artificial intelligence gives us, that they drive down the need for as much in-person human interaction, because we can get information in different ways, we can have conversations across thousands of miles, the way that we are now, what is it about building a start up at the stage where you are where you have that strong feeling?
Bret Taylor: Well, John, I think there's really a difference between science and engineering. And sometimes, when you're building software, you know how to do it. And as an old boss used to call, it's a simple matter of programming, you just need to get it done. What we're in with AI right now is true science. We're building new capabilities that, really, fundamentally didn't exist before. We have a lot of researchers on our staff, we have an old Princeton professor who is leading our research team right now. And we're really inventing new ways for companies to interact with their customers. And how AI agents are built is a fundamentally new concept, a new technology. And it's not one of those things that we felt you could do just over Slack. We really felt like that congregation around the proverbial water cooler, the whiteboard, or the lunch table, is really necessary when you're doing pure invention. And it's also why Clay and I started a startup. If you look at the history of technology, a lot of new companies are born in these technology waves. And it's not because incumbents don't have advantages. It's just, in these moments, where the technology is so new, the value of agility, the value of having these unique customer insights, starts to benefit startups a bit more. So it's a fun time to be a startup, and it's a fun time to be in person.
John Fortt: All right. And finally--now actually, finally--what's next, stage-wise? Now that you've raised this funding is going to go mostly toward hiring more of those humans who are going to gather for lunch there in San Francisco isn't going to go into infrastructure, which you've just said you--because of the way the company is structured, you don't necessarily have to overspend on--what's the next few months hold strategically?
Bret Taylor: Well, John, if we're here with you a year from now, we'd like to not just be telling the stories of Sonos and WeightWatchers and Sirius XM, and OluKai. We'd like to tell you that we went to every major consumer brand in the world, and we helped them build their own AI agent, and helped them succeed in this new world of conversational AI. So we look at that money--it's really a means to an end of really driving more success for our customers. And as you know, AI is one of those areas where it's easy to make a demo and hard to succeed in practice. We're really proud of the success we've driven with our customers today. And I hope we're here with you a year from now, sitting side by side with some amazing brands and really driving more success than any other company in this space.
If you got this far and you liked this piece, please consider tapping the ❤️ above, subscribing, or sharing this letter! It helps me understand which types of letters you like best and helps me choose which ones to share in the future. Thank you!
Wrap-up
If you’ve got any thoughts, questions, or feedback, please drop me a line - I would love to chat! You can find me on twitter at @kevg1412 or my email at kevin@12mv2.com.
If you're a fan of business or technology in general, please check out some of my other projects!
Speedwell Research — Comprehensive research on great public companies including Constellation Software, Floor & Decor, Meta (Facebook) and interesting new frameworks like the Consumer’s Hierarchy of Preferences.
Cloud Valley — Beautifully written, in-depth biographies that explore the defining moments, investments, and life decisions of investing, business, and tech legends like Dan Loeb, Bob Iger, Steve Jurvetson, and Cyan Banister.
DJY Research — Comprehensive research on publicly-traded Asian companies like Alibaba, Tencent, Nintendo, Sea Limited (FREE SAMPLE), Coupang (FREE SAMPLE), and more.
Compilations — “An international treasure”.
Memos — A selection of some of my favorite investor memos.
Bookshelves — Collection of recommended booklists.