Curiosity, Agency, and AI with Bret Taylor [Glue Guys, EP.57]

Most of the time it's gradual, but you know, there are some professions, like for example, being a professional athlete where, you know, somewhere around the age of 30, you have to gear shift into something completely different. It's much more dramatic for most people's careers, it's much more subtle. But I think it is one of the hardest things for people to do to grow, which is to let go of their identity to grow. And I think it is by far the reason people sort of plateau in their careers.

I'm very happy, Brett, to see you. You know, I was thinking about how to talk about you today and I was going to describe all the things you've accomplished in your life. But I guess I'll suffice to say you have a resume that's very inadequacy feeling inducing. You know, there's just been a lot of stuff that's been done, you know, starting with co creating Google Maps or co founding Google Maps, you know, being the CTO of Facebook when it went public, being the chairman of OpenAI, being the co CEO of Salesforce, and now co founding Sierra with my dear friend Clay Bevore. And I don't even think that covered it. But I can tell you the thing that actually most made me feel inadequate was the first time we really hung out. You described a part of your business education and you made an analogy to. It was like learning the west coast offense under Mike Shanahan. And I'll tell you why. I kind of.

I think I said Bill Walsh. I think I said Bill Walsh.

You probably said Bill Walsh. Great. Well, I was, I was actually kind of pissed when you said it because in my mind I'd been like, well, this guy has accomplished a lot in technology, but he's obviously a nerd who has no idea of anything in sports. And then when you gave like a pretty good sports reference, I was actually kind of mad because I couldn't even have my one point of superiority anymore.

Ravio, it makes you feel better you still have that against Clay, who I don't think even knows the rules to most sports.

The problem with Clay is he might not know a lot about sports, but he's built much more like an athlete than anyone I know. It's amazing.

He's more healthy than all of us combined.

Yes, yes. But anyway. Okay, so here's my question that I wanted to get started with. I actually don't know anything about how you grew up. I don't know. I know you grew up in California, but I don't know if you have siblings. I don't know anything about you as a kid. And so I'm curious, like, other than being a Niner fan, you know, what else do we need to know about how you grew up?

So born in Oakland, California. That's where my sports allegiances were. I don't think there are any sports teams left in Oakland other than the. The roots, the soccer team. So we've got so. And then my mom worked for Chevron, so we moved around a little bit. I ended up in Marietta, Georgia, for middle school. That's how I actually developed my love of football. When you get off the tarmac in Atlanta, they just put shoulder pads on you as you're a teenage boy. So played for the Blackwell Bears. And then what position did you play? I played tight end and defensive end. And then, you know, when you're in eighth grade, you typically play for the high school team, so they can kind of scout you. So I played for Spraberry High School there for the eighth grade team. And then we moved back to California. Yeah, I have a sister, older sister, kind of Irish twin. She's one year older than me and very close to her. We would always like, you know, edit each other's papers and have the good rivalries siblings have. Of course, she did better me than this. I'll get better than her at this, that type of thing. Still very close. She's a public school teacher in Rockland, California right now. Yeah, I ended up going to Stanford largely because of her, actually. I had taken a couple classes on the east coast and just kind of wanted to get away from California a little bit. But she went to Stanford and then going to a few parties on Stanford campus as a high school senior. I was like, I want to go here. Probably would have experienced that in most places, but I followed her to Stanford and that's how I ended up there.

That's awesome. Okay, cool. I didn't even know that you had an older sister, and I didn't know she was a public school teacher. That makes a lot of sense. Okay. One of my favorite things that you say in all seriousness is there's a lot of people in the Valley who I think identify themselves as, I'm an engineer, I'm a salesperson, I am a C level exec, I am a startup person, I'm a big company person. Right. You are all of those things. Right. You have not stayed in one lane. And you described one time you said, I don't think that you should let an ossified version of your past Self limit your future self. I think that might be the words that you used. And I loved it. I loved everything about it. I want you to explain that idea and I also want you to tell us why you've come to that.

Yeah. So I've told this story many times about Sheryl Sandberg, but it's previously guest. Was that.

Yep.

She's a previous Glue Guy guest.

Oh, really? Oh, that's awesome.

Blue gal. True Blue gal.

Yeah, she's. She's remarkable. And in particular, I think one of the. She's remarkable in many ways, but the number of people who have stories like the one I'm about to tell is a true testament to her character. And. And her now deceased husband Dave was exactly the same way. Just, you know, they're. They just have a network of people who. I remember actually going to Dave's funeral and watching people give speeches and realizing my relationship with him, he had with like 100 people and just being like, gosh, what a. I mean, what a remarkable person. But going back to Cheryl. So I was the CTO of Facebook. I had started that role with. A CTO can mean a few different things in Silicon Valley. Can mean kind of, you know, kind of a very senior individual contributor type who's kind of architect of the system, or me and I run engineering, and it's a fairly large operational role. I made this transition from the former to the latter, so I had very few people reporting to me, and then I had over a thousand people reporting to me. And it was the largest job I'd ever had by a mile. And I was super grateful to Mark Zucker for taking a bet on me, but I was sort of fumbling in all the ways you'd expect. You know, it's. It's, you know, just when you. You first have a job that big and. And candidly, you know, Facebook was a pretty young company, so there wasn't a lot of, you know, folks to copy around you other than Cheryl. So I was actually. The. Part of the story I haven't told a lot is I was actually about to go see Steve Jobs. We were on the cusp of forging a partnership to integrate Facebook with a version of iOS. And it was. I saw him a couple times in that process, but he was, you know, definitely near the end of his life. He was pretty frail at the time, but I got the presentation for what we were going to show him ahead of that, and it was just total shit. And so I was in a conference room just redo. I'm not Going to show a crappy presentation to Steve. It's just, you know, he's like one of my heroes, like everyone else in tech. And Cheryl saw me kind of, you know, just rewrite in a presentation. And it wasn't just that, but she started asking me questions and I was sort of somewhat griping about the quality of the work that the team was doing in some ways. And I think she was really, really got hard on me and like, what are you doing to fix it? And why aren't you holding to your team, to higher standards? Do you have the right leaders in place? And I didn't have great answers to all those questions. And she kind of tore me a new one a little bit. I don't think she would characterize it that way. It felt that way to. And I sort of went through all the stages of grief over the next like 12 hours. So like, you know, that evening the presentation went fine, by the way. And you know, and I'm. You first get defensive and they're like, ah, it's, you know, she doesn't know what she doesn't know everything going on. And then, you know, and then I was like, oh man, I'm a really bad person. Oh, I'm gonna get fired. I'm like inadequate for this role. And then you sort of get to what is the last stage? Acceptance or something like that. So. And I woke up the next morning and I was like, man, she's 100% right. And everything she said. But the thing that she didn't know, just because she didn't know me that well, was sort of the root cause, which is I was approaching every day thinking, here's who I am and what I do well, and I'm going to do that every day and sort of try to form the company around those skills and around my conception of what I'm good at. And that was being an engineer, sort of being technically and sort of product wise, really proficient. And I would sort of push out things I didn't feel like were core to who I was, like recruiting or you know, sort of like organization design or all these other things that just. Or business development stuff that I just didn't care about. And I was realizing that actually my job wasn't just like my job wasn't to just be good at the things I was good at. My job was to make fun of Facebook successful. And I was. Had a really narrow interpretation of that. And so I decided at that point I was like, I'm going to have a. In this job as CTO of Facebook. I'm going to have a much more expansive view of what that means and I'm going to wake up every morning and think what is the most high leverage, high impact thing I can do today to achieve whatever I wanted? You know, we were trying to achieve. Usually it was a big product initiative and I got better at recruiting, I got better at management, I got better at everything just because I would spend time on it. And then the counterintuitive thing for me anyway has actually really enjoyed it. It turns out what I thought what drove me was this technical thing. What actually drove me is having an impact of being successful and you get this positive feedback loop on it. And I was like, I'm a lot better at my job than I was before. And then it became self fulfilling in a lot of ways. Just by having that open mindedness about my own identity, I would approach every job that way. And then after that I ended up in technical roles at Salesforce. After the QUIP acquisition I became coo which was a very traditional operating role. I became co CEO which was sort of like, I guess everything, you know, in between. And then I, you know, after that gear shifted down to starting the company again. And as you know as well as anyone, right when you start a company your main job is to build something. So I'm, you know, open up cursor and you know, get to work. And I could make those gear shifts pretty naturally because that morning that was what was the most important thing. And it's been really liberating for me. And I always give this advice to people, particularly founders because I think a lot of founders end up kind of single issue voters. You know, often you end up a founder because you have a superpower of some sort. If you're a great AI researcher, you're a great AI researcher. If you're a great product person, which is very typical, you have a great insight on the market, great intuition about product. You fast forward two years in and the problem might be you have a bad head of sales or you need to open a new office in Japan and that's the most important thing in your business. And if you always approach it through the lens of your superpower, that's not going to be the thing that actually makes your company successful. And I think it's very hard for people to do this and so I'll stop talking. But I think it's really, I mean it's interesting because most of the time it's gradual but you know, there are some professions, like for example Being a professional athlete where, you know, somewhere around the age of 30, you have to gear shift into something completely different. It's much more dramatic for most people's careers. It's much more subtle. But I think it is one of the hardest things for people to do to grow, which is to let go of their identity to grow. And I think it is by far the, I feel the reason people sort of plateau in their careers. And it's quite subconscious and kind of insidious because often it runs counter to a lot of the advice you've gotten or the feedback you've gotten up to that point.

Well, where did you learn that? Because that's an amazing skill and that's a theme that comes up often on blue guys. But like, did you just wake up one day and say, look, I got to be real about who I am and what my trajectory is and I have to reinvent myself in a new way? Like, did you learn that from a mentor? Like, where does that, where does that come from?

I was scared straight by Cheryl. I mean, that was it. But and the thing was, you know, because I wasn't great at all these other things, I had to have the self awareness of where I had gaps in it. But I do feel that if you are self aware of your strengths, you can. What is the. Good artists cheat, great artists steal, whatever. I can't remember copy. Great artists steal. I just butchered that quote. But you can find people are great at those things and learn from them. But the most important thing is to work on the most important thing. That's the thing that is in any job, Even if you're 75% as good as the best person. And recruiting, if recruiting is the most important thing to do, spend time on recruiting. You know, you get the at bats. And so the self awareness just came from the being sort of scared straight by Cheryl and then the positive feedback loop that came from doing it and then realizing, wow, this is, this is the right way to, you know, form a career. In my mind machine.

Can I, can I give you a hypothesis on what I think for Brett, you know, also as root cause there? And Brett, you tell me if you agree, right? But I, when we started the podcast, we found this Bill Russell quote that was like amazing, which is, you know, ultimate winner, right? And the quote was my ego demands, comma, for myself, comma, the success of my team. And it was like this incredible juxtaposition of like, look, we have to win. Like that is the thing that drives me. And so what's really Interesting is, I think, to your point, if you just sort of have your identity, it's like, I want to be good at the thing I'm doing. Right. But if you expand that of like, the team I'm on needs to win and all of a sudden I have to like, figure out not the most important thing to make me feel good that day, but the most important thing to make my team win, all of a sudden, like, you're doing all the stuff you're saying. And I do observe you to be like maniacally competitive. And I say that as an extreme compliment, you know, But I think if you really do broaden this idea of like, the thing that matters is us winning no matter what, you know, I think you kind of go and do it. And Shane, it reminds me of you, you know, of the stuff that you would do on the court. It's like, I don't know if I know how to do that yet. I don't care. But that's what we have to do to win. And so I don't know, Brad, I don't know how that, like, you're right.

I mean, it definitely comes from. You have to. If you're really obsessed with a goal, you're willing to give up a lot to get there. And that's certainly part of it. If you're apathetic about where you want to go, it's, you know, a lot of these things are pretty hard and uncomfortable. So it's hard to find, you know, the wherewithal to do it. But I'm, I am very competitive. That's absolutely true.

It's a lesson that I, or it's a question I get from young people especially. And in our world that is, is very curated via social media. A lot of especially young people think they have to be perfect in, in a certain way. And so your argument is like, do it all shore up your weaknesses on the way up. And I mean, that's my, that's my advice to young people. But it's a really hard lesson, especially for young people to, to embrace and they feel that they have to be so excellent in one thing and that's what's going to separate them. Like someone's out there searching for, oh, I need someone who's just an amazing coder or amazing engineer versus having an all in one game.

Can I take this into AI for a moment to give a hot take? I'm not sure it's true, but my intuition is here. I think AI will produce an era that benefits generalists. Patrick Collison has talked a lot about science having plateaued in terms of real step change, progress. And one of the theories that I've, I've read about, I don't read. He reads more than I can physically read, so I haven't read everything he posts. But you know, really this idea that scientists are increasingly incentivized to go really specialized, you know, to, to actually, you know, have your thesis, get through the PhD program or go through a postdoc program, you have to get really deep but narrow. And that's the way science has kind of gone. And a lot of historical scientific breakthroughs came from sort of cross pollination of ideas. And if you look at the heyday of science of you know, sort of like early 20th century physics as an example, it was a lot of different disciplines and sort of combined. And then, you know, similarly, even with technology, you know, it's hard to be, you know, like a deep systems engineer who understands distributed systems and be good at UI design. And so to produce things of great value you often, you have to find a bunch of specialists and you have product managers or whatever bringing those people, people together. But many the issue with that is there's just a lot of gatekeeping. You know, it's like so to make something really great technically with a great user interface, you need multiple people, which means you need funding, which means you need to convince people like Ravi that it's a good idea. And what if it's not obviously a good idea, which is actually many of the greatest businesses of all time have been not obvious and now you're like, wow, that back engineer can vibe code a ui, that front end engineer perhaps have an AI agent help with the database, you know, how to connect it to DynamoDB correctly, you know, perhaps a scientist who is actually understands a lot of different domains well enough to kind of navigate, you know, sort of AI to help them, you know, poke at things, can actually integrate, you know, that deep expertise because deep extraverties is becoming more commoditized with AI. And I wonder if we will be in an era where, you know, a great filmmaker, a great scientist, a great product thinker can do more individually. And I believe that's actually really important for invention because I think so much of invention is someone who's like mildly unreasonably obsessed with an idea that not everyone sees but them. And I actually think that AI can be a great unlock there. So Shane, not only from a personal growth standpoint, I actually do believe people undervalue trying to be as great as possible at all things, not just the one area of specialization. But I think AI will start to benefit generalists in a way that I believe will really benefit humanity.

I would agree. And I think the underlying trait that will bring that out is curiosity. What are you willing to learn that you don't know today? And it's going to be the most curious people, just like they have throughout the history of humankind. But it is only unlocked if you say, that's interesting, I need to learn more.

Absolutely. And it's interesting because AI is an amplifier of agency and curiosity. So real you now have this roughly super intelligent thing in your pocket to fulfill that curiosity, to be a foil in those areas. And it's interesting there's these studies that I don't know how, I don't think many of them are peer reviewed yet. So I don't know how true they are. Just about AI can sort of make you lazy, you know, and in some ways AI is this ultimate tool that if you want to cheat on the essay, it's probably the best tool ever to do that. If you're curious and you want to learn more, it's also the best tool for that. And so it's like, it's to some degree, like if you spike on agency, AI will multiply that. And I. To your point, Shane, I think it is curiosity 100%. And I'm really hopeful that those curious, kind of unreasonable people now have like an Iron man suit for their unreasonable curiosity. And I think that's going to be just fucking awesome to see some people invent things, you know, with again, fewer gatekeepers, fewer people that tell them no, fewer people to ask permission to do things that could, you know, that, that because of this technology they can sort of prove it through action rather than prove it through persuasion.

Yeah, I think I, I completely agree and I'm so happy you added agency to it. And Shane, I know you agree with that too, but I, I think that, look, you are limited now by your willingness to go learn something. You are limited now by that is the only thing.

Right.

It's not that the tool's not there. And I think the amplification point I really like. And I think even though social media, a lot of times people have negative views on what it does, I think that a lot of people there are sort of like social media reveals, social media amplifies, it doesn't actually change. Right. You know, the core nature, I think AI in a really good way amplifies, you know, and I think that it is whatever you make It. And I also like, Brett, that there's all these debates and like, when is AGI coming? When is AGI coming? For all intents and purposes, if we had said five years ago that a tool or a computer could do today what it can do, and we had said that five or six years ago, we'd have been like, that's AGI. Like, that's pretty good, right?

If you asked anyone three and a half years ago, what is AGI? Yes, everything we have now would exceed that. We just move the goalposts, you know, which is fine. That's awesome.

And so, but, but I think the takeaway for that, of course, is like, okay, cool, don't wait for some new definition. It's already here. Go use it. Go and like, learn from it. Go and use that agency. And I think one thing you talked about, humanity. I think that people being able to do interdisciplinary stuff, like Range is one of my favorite books and I love the idea that Epstein brings up in it of sort of like, look, there are like finite games and infinite games. There's defined games and indefinite games and like life and our work is always going to be in the indefinite version. And that means that it's going to favor the generalist who's curious and high agency, you know, and so, and I actually prefer that future because otherwise, like with kids, it's like what you're going to tell them at 11 years old that they need to specialize in, you know, a narrower or a narrower thing. Like that's no fun.

And I was never talented enough in sports to do this, but there's sort of an analog to, you know, the kid who's playing one sport all three seasons, all four seasons. And you know, now conventional wisdom is actually, actually having a diversity of sports makes you better in the sport that you excel at and things like that. And I, you know, I, I just think it's especially for kids growing up, it's just a much more fulfilling life to dabble in a lot of things. It doesn't mean you shouldn't invest the 10,000 hours in the thing that makes you great. But, you know, I like, you know, this Ravi and board meetings and stuff. You know, I like pulling out obscure historical references and I like film and I like sports. And you know, the people who are the most interesting to talk to are the ones who aren't single issue voters, you know, that like actually have some depth to them. So I'm hopeful, I'm hopeful this technology will really amplify that.

One of my favorite Memories of all time was hitting a home run in high school. Okay, so I play baseball. So here I am, you know, top 10 player in my class. And I didn't play my first two years in high school because I had some injuries. But my junior year, I said, you know what? I want to play baseball because I love baseball and obviously had not worked on my skills in a long time. So I used to be a pretty good power hitter, but I became a 6 foot 8 singles hitter my junior year. But I'll never forget, you know, I've done some pretty cool things in sports, but I hit a home run to right center against Warren Mott. It was solo shot. And I love you.

Remember the pitcher's name.

That's so I, you know, I, I will, I will never forget that field, rounding the bases because it reminded me of little League. I was a master in little league. But I'm like, when you're running or.

You'Re like, slow trotter, fast trot, chain, what do we have?

It's fast. Always fast, you know, Come on.

Never going to show them up.

Barry Sanders, Walter Payton. Act like you've done it before, you know, but inside I was like, this is pretty sweet.

Totally nothing to me. Yeah, I got this.

Yeah, but there, you know, so I, I strongly, you know, agree with you, Brett, about having diversity of experiences. And that year of baseball for me, I didn't start. I came off the bench and here I was, a hotshot basketball player. But it was so fun to, to play a sport where there wasn't the pressure of going out. And, okay, I got to lead this team and I got to win a state championship. And I was so much better for it. And it was still a great teammate, a team. Those guys I play with, I still talk to now. And it's, it's, it's too bad that kids, young kids feel the pressure to not have that experience. And it's a shame.

But we can change that, you know, like, we can help with that. I think seeing people who are successful, who have done different things and, you know, I think, Brett, it would be easy to have an opinion of you based on your, you know, accomplishments, that you're like some sort of cyborg who's super specialized. And I think the idea of, like, no, like, he's just like a really curious, high capability, high agency person. I hope that kids see that. Speaking of kids, we talk a lot about parenting on our show. Right? You have three kids, and what do you tell them in terms of both on the. Hey, the world's going to change. And here's how you should think about, you know, what you do, but also just like, what do you and your wife value? What are the values in the Taylor household?

So, yeah, it's interesting when I go to a cocktail party, particularly because of my OpenAI roll, the number, like, the two questions I get are, what should my kid study in school? And are we in a bubble? And those are the two that I always get. I think we probably. I don't really know. I mean, I think it's very likely that the things we do in our jobs change pretty dramatically over the Next, you know, five to 10 years in most professions, most white collar professions, but even that with robotics, which is a slightly longer time horizon, really all professions in some form. And as a consequence, I think the idea that you're going to learn a vocation and that vocation will be stable through the course of your career is unlikely to be true. Almost zero percent chance. Obviously, it will vary by profession. Like if you're a masseuse, maybe that's not going to change. But if you're at work in the world of information, it certainly will change. And as a consequence of that, I think the sort of traditional liberal arts education view of you're going to learn how to learn feels particularly important right now. I think similarly, having sort of some plasticity of your mind, like this idea that your career is the start to the learning process, not the end, and that when these new technologies come available, one of the things you will need to do in your career is to be good at adopting those tools. I've used this analogy before, but I like it because we've lived it. Just imagine being an accountant before Microsoft Excel and after Microsoft Excel. Imagine Excel. Scary. Like ChatGPT. It went from 0 to 100 mph overnight. And you can imagine if you were in one of the big accounting firms, there would probably be the folks who are like, this is not the way we do things. And then there's going to be someone who's like, I learned pivot tables and I'm running laps around everyone else at the firm. And I say that just because I think analogies are somewhat problematic, because as Mark Twain said, history doesn't repeat itself, but it rhymes. There's going to be very big differences and it might undersell the change, but there's this mindset shift, which is like, your job is to optimize your tax rate or balance the books or whatever. Accountants do not type in numbers on a calculator. But I think day to day, you start to define your job by how you do it, not what you're actually accomplishing. And that's where that sort of ossification comes in, and that's where these new tools come out. You can get almost defensive psychologically around it. So with our kids, we're trying to make sure they can learn the basics. This is a tool to amplify agency, not to shortcut understanding a topic. But it's really wonderful. One of my kids was having a Spanish test and was using ChatGPT advanced voice modes to quiz them on the way into school. When one of my kids came home with Shakespeare and they had a question on it. We took a picture of the page and it was so much better than I was ever at Shakespeare, which I thought was really amazing. My son got a 3D printer and wanted to start an Etsy store. And he's using ChatGPT to learn how to do online marketing, to drive traffic to his Etsy store. And it's been such a great way. So I'm trying to, like, have them embrace. This is a tool to help you learn and to help you be successful. I think we're sort of at the point where using ChatGPT to write an essay is completely inappropriate. Not using ChatGPT to give you feedback on your essay might be equally inappropriate. You know, it's like, are you not. Are you, you know, are you using the tools available to you? So with that, I would just say it's like the tool has sort of exceeded the educational system right now. I mentioned my sister's a public school teacher and. And she does biology and chemistry, so it's been a little simpler. But imagine being an English teacher right now, and it's just really hard. Have it. You should, yeah. Have a ton of empathy for the situation. So I think we're still figuring out how to sort of have the infrastructure around education and, you know, children sort of catch up to the technology. But in the meantime, as parents, I think it's the wrong thing. I don't think this is like social media, where it's probably just unhealthy at a certain age. I think this is a tool and you should teach your kids how to use this tool. And in the. Really, the Steve Jobs quote, you know, computers being a bicycle for the mind. This is really. It's sort of a realization of that vision in a lot of ways. So in everything, we'll pull it out. And, you know, my daughter was learning how to code and she was asking me Questions. She was making a website for my wife, and she asked a CSS question, like, use ChatGPT. I didn't answer a question just because I was like this. I want to teach her how to answer the next 10 questions. So we really pull it into everything that we do and, you know, try to strike that line about what is lazy and what's, you know, actually, you know, helping their. Their educational process. And it's been amazing. My son loves, like, the idea of being a pilot and flying and, you know, we're. He can Learn more from ChatGPT than he ever learned from my wife or me. And so it's been really great. We've just been using it as an educational tool, and I honestly could not imagine raising kids without it at this point.

Yeah, yeah. It comes down to the what questions can you ask? And that's the message. I tell my kids, like, look, yeah, it's an amazing tool, but you have to be really good at asking the right questions and questions. Beget questions, beget questions. And I tried to give them examples of things I've tried to teach myself. I do a lot of corporate speaking, and I put my, you know, put my speech in. I'm like, this doesn't make sense. Like, help me figure out a better way to communicate with the people I'm trying to reach.

Are you more honest with ChatGPT than, like, the people? Like, I always literally, like, when the doctor's like, how many glasses of wine do you have a day? You like, you alls around down, right? Like, you know, with, like, chatgpt hash, you're like, okay, I'll tell, you know.

A hundred percent, 100%. I mean, my. My wife is starting to question my relationship with ChatGPT.

The problem with that, guys, is when you're cheap like we are, and we share a subscription, you know, on Chat GPT, my. On my honest stuff shows up. And when the kids are using dad's CHAT GPT subscript. And so, you know, this is a good argument for why we all need our own individual subscribers and why ChatGPT strongly agree.

You should get accounts for everyone.

Nonprofit. You might be an unpaid share, though. You know, the, you know, you know, the thing that you said, though, on the writing piece on, like, it kind of moving past the education system. That's a place, you know, our kids homeschool. One of the places that's been kind of helpful has been like, even teaching me why I care that they become good writers, right? And reminding myself of why I care about this because now to your point, I'm sort of like, guys, you can use this to do this in an inappropriate way, but the thing that truly is you won't be learning how to express yourself. And like the thing about writing that is amazing. Like, I love that the board memos you and Claywright are long form. I like that they are written thoughts, they are not slides and they are.

Much harder bullet points.

They're actually, they're actual writing. And in the. I would imagine that the clarity it forces you all to have is the reason you write it. Much more so than anything else. But they're wonderful to read and I think that the people that think the clearest are the best writers. It's not about the vocabulary, it's about the clarity and the concise way to communicate. And so it's actually helped me even with the kids, which is listening, you can, you can, we can probably tell if you use chatgpt on your writing. But even if we couldn't, the real problem is you can't express yourself in a way that you're going to want to, you know, and there is no skill more valuable than that at some point of what am I trying to communicate and how. And so I even like that, you know, you guys practice that in like, you know, high stakes discussions.

Well, it's one thing that is obvious to anyone who's founded a company, but communication, clarity of thought are prerequisites to almost everything. Because everything. You could call it selling, but I think that word has a connotation. I don't mean. But you have to recruit employees, you have to bring investors on board. When you have a product launch goes really well or not well, or a competitor do something or there's an up or a down or whatever it is, everyone's looking to you to help say, what's going on, how should we think about this? How do we map a path forward? And it's. I think the people just need, the more you are effective at communication, written communication and verbal communication, the more successful you can be. Because great things come from movements with more than one person in them. You know, like most things aren't, you know, the can't be done by one person. Most great things just can't. And I strongly agree with you. And actually we will always think about like Clay and I will always talk about, are we writing enough? You know, and you know, as you scale a company and we're in a period of hyper growth I personally never quite experienced before, even having been at all the companies you mentioned and we realize if I say something out loud, it's actually mathematically less valuable than writing it because writing is a durable asset. Yeah.

Yeah, that's cool.

The future people can read. And I realized if I stay in a room, it's not only not available to people in the room, but you have the standard game of telephone that happens. And also anyone who arrives at the company that wasn't there in that moment doesn't get to benefit from it either. So we. It's funny to think about it that way, but we were like, you know, it's. We. We think about communication as a. I don't know, it's just. It's one of the most vital inputs to a culture. And I think sometimes people think about it enough and repetition doesn't spoil the prayer. You know, like, you need to, basically, with these things, especially hard things, you just need to drive it home over and over and over and over again. And I think it's really hard because if you're the communicator, you feel like you've said it totally and you're not thinking about the, like, all the stakeholders who need to hear it. And I think it's one of the most challenging parts of really growing up as a company.

I remember this was relayed secondhand to me, but I believe it to be true that part of the reason Patrick and John write so much at Stripe is because it's like letters to the unborn Stripe employee or the Stripe employee doesn't know they're a Stripe employee yet. So I thought that was awesome. I think some of the.

By the way, both of them are exceptional writers, too.

Yeah. And they're exceptional writers because they're exceptional thinkers and exceptional readers. Right. And then they practice. But I. I love that concept. I think the Sierra writing. I think the Sierra communication ends up being like letters to the, you know, Sierra employee that doesn't know they're a S employee yet. And then the other thing is, I don't know. Like, I found that sometimes I think I have something in my head and I think it's really clear. And then I'm really humbled when I try to write it down because it's like, oh, wow, this is actually not nearly as clear as I thought, and I don't understand it as well as I thought. And then the discipline of writing, it makes me understand it much better. And, Shane, I know on the stuff that you write, I find the same thing. I think the clearest things are the things that you could write. You know, you can write down and tell the story of and make it concise and pithy and all that stuff.

I think I'm, I'm the opposite actually. I don't think I'm a very talented writer at all. I think I'm much better orating my thoughts. Maybe that's what I do for a living. But I found that using chat GPT, I become a better orator because I have to be very clear in what I'm trying to distill. And I don't, you know, I use voice dictation for most, most of my, most of my queries and prompts and if I'm not clear, I'm wasting my time. And so I, I'm very thoughtful as to what my inputs are verbally and it really gets the gears spinning, you know, in a way that I've never used my brain before.

Gone from I'm talking to I'm prompting. Yeah, when you, when will you start thinking about that? When you're talking to people, like, what prompt do I want to give this?

100%? But that's so true. Yeah, but I would want to argue that at least to me, the way you're describing it, I mean, we're just using writing to describe thinking, you know, and you can think with, you can talk or you can write. And my point is if you write down the themes of your corporate talks or you write down the whole thing, I actually don't know what you do. It to me is a very similar process of sort of like making sure that it is clear and, you know, communicative and actually saying what you want to say. And to me, the like, I think it's the same process. Like Mike Moritz, who is a beautiful writer, unbelievable writer, you know, I asked him one time, I said, what are the tips you have? And he's like, after you write it, just read it out loud. The most simple thing. He's like, just literally sit there and read it out loud and you'll be able to tell like, does it work? Where is it not? And it was so simple. The two piece of advice he gave were, you should read a lot and then you should read the thing you write out loud.

And I was like, that sounds terrifying, Robbie. When I play basketball, this is true story. I never watch film of myself. I hated the way it's like, you know, singers don't like to hear their songs. I never watch myself play basketball. I still don't like to watch any replay of any of my games because I'm like, gosh, I look terrible. I'M so slow and uncoordinated, and I didn't want that mental vision in my, in my head. And so I never watched tape ever, because I hated the way that I look looked on tape.

But what if you had, like, film in, in practice? Like when Coach K was like running film sessions? You obviously have to watch then, right?

Yeah, you know, just cringe, cringing.

Do you watch videos of yourself? You know, afterwards just being like, man, I crushed that. Look at that talk I just gave.

You know, it was interesting. I. I didn't used to. But at Salesforce, we did a lot of events. The running joke within Salesforce, where we're an events company that also makes software on the side, you know, and so speaking in front of, in person, there was tens of thousands of people and often online, millions, which was. And I did that on such a regular basis, I would start to watch and then try to, like, edit it. I had the exact same thing as Shane, actually. For me, just hearing my voice felt weird to me. It doesn't sound the same to me as it does on, like, Jesus, oh, listen to that. And then you're like, but I. A little bit like an athlete. I was just, you know, this is a product for game Salesforce community. I want it to be great. And I noticed little idiosyncrasies that people are too polite to tell me and the way, you know, just the mannerisms you have when you're stumbling over a thought and things like that. And I would notice and just kind of like try to get better at it. So I don't always do it, but I, I forced myself to get over the cringe so that I can make that product better. Now that's a much less vital part of my life.

So one thing that I will tell you guys. So, you know, in the venture business, obviously you're making investment decisions, right? Not the way our process works is on Mondays, you know, we're discussing the companies that are, you know, that we're considering trying to pursue. Right. And there's a memo that comes out on Friday that you read over the weekend. And then we discuss it on Monday and, you know, we'll make our decisions. We all vote at the end, 0 to 10. You can't vote a 5, you know, on whether or not you should do it. And the two things that are helpful and very humbling are like, looking back, tracking all your votes on everything that you've ever done. Right. And looking back a few years back of like, okay, well, what did you think? And then I Also write down for myself, what do I think of this company? What do I think are the key things that have to work and not. And because the partner meeting is like a living thing, right. You don't always say everything that you're gonna say. It's just naturally there are conversations that go one way or another. But looking back yourself, what did you vote on this company? What did you think when you read the memo? It is exceptionally helpful and exceptionally humbling. Right. Because you have nowhere to hide. You can't say, oh, we just did that because somebody else or we didn't do that because of this. You know what you voted and you know what you said about it. And so that's something that find to be helpful in mine.

But also a little while ago that got some pickles during the pandemic. I can't remember when, but. So what was your tech prediction that was the most wrong. And I brought up two. I thought the iPad would be significantly more disruptive to the PC market when it came out. There's a bunch of things I would almost put in the category of execution issues there, in my opinion, why it didn't. But then the other one, which is relevant just given the kids. Thing is, I bet my wife when my oldest was born that she would never get a driver's license because of the confluence of self driving cars and on demand and she just got her learner's permit, so I paid up. But it's interesting, you look out the window here and you've got Waymo's and all that. So like I wasn't that far off.

That's a good bet. That's a good bet.

I thought it was a good bet. I thought it was going to happen five years ago. And so it's interesting though, in the spirit of, I always attribute this to Bill Gates, but I'm not sure who said it, which is we overestimate what we can achieve in a year and underestimate what we can achieve in a decade. I think it's really important as a technologist to kind of remember your Mrs. Because as you're doing, going through what Bezos would call the one way door, those are the moments where you want to remember your past misses and understand like if I walk through this door, you know, and I'm wrong, you know, what, what's going to happen? And it's, it's, those are the moments, as I said, mainly for the one way doors. You don't need to bat a thousand at most things, but those one way doors are Significant.

Well, you know this Brett, like a lot of people now are sort of saying at least around here, like okay, an X number of years, 10 years, 15 years, you know, very. A very different percentage of kids will go to college, right? And I think that, I believe that that will. I think there's like elements of truth to that of like if there is one part of today's college bundle that is less valuable in a world where AI advances as it does. But I think it ignores the whole value of the college bundle, right? Of you know, there's a lot more than just like sitting in class and getting lectured to as to like why you go, you learn independence, you build incredible relationships. A lot of people meet their spouses, you know, like there is this, you learn other things, you learn resilience. There's all these other things that happen. But I actually think that part of the thing that happens on the predictions is we sort of assume excellent execution or we assume like a progress rate like on self driving. I actually think that that was an amazing prediction in many ways of like what would happen. But it's hard when you think about the human side of this. Of like we treat machine errors very different than we treat human errors. Right? Self driving is objectively way safer, you know, like you would imagine. There's some community that I would imagine is going to happen in the next year, one to three years where they basically are like it is immoral for us to allow teenagers to drive rather than for them to be in a self driving car, right? At least 10 times safer for everyone to get cruising around in self driving cars. But you know, when it's a machine that makes the mistake, you know, rather than the human, you know, we think differently. But I can imagine it's also been.

Really interesting just to see, I don't know, the cost and technology shifts there too. I mean not to go like too deep into it, but there's actually a really interesting PBS documentary like the Nova series about the early DARPA Grand Challenge where it's really. I don't know if you can find online. I was trying to find it unsuccessfully the other week, so maybe I'm just bad at Internet searching. But effectively it was the story of CMU and Stanford primarily and their rivalry to see who could complete the first DARPA Grand Challenge which was driving through this track, through the desert. The first year everyone failed. The second year, Stanford and CMU both completed the race. Stanford won, by the way. But it was really interesting because the dynamic in there was Sebastian Thrun had gone From Carnegie Mellon to Stanford and taken kind of a software first kind of approach like AI, ML, Computer vision and cmu, sort of famously one of the best robotics labs in the world took a very hardware sensor oriented approach. So whether it was sort of hyperbole and I love the software versus hardware which will hook any, any geek any day of the week. And it was a great documentary. But you know, it's interesting if you look at the trends since then of like sensors like LIDAR and all that and just a huge C code bases running these cars and the bandwidth going through it. And then you look at what at least my understanding of what Elon is doing with Tesla, which is using much lower cost sensors and replacing most of that code with just a model. A lot of that wasn't really technically possible before. So it'd be really interesting too because the last mile on these things is do you need to manufacture a whole new fleet of vehicles for the world to use to benefit from this? How much can you. And so it'll be really interesting to see this play out. I'm fascinated by it and there's some really smart people who think it's going to be impossible without these really high quality sensors. There's other smart people who think with appropriate models you don't need it. But I love this stuff. I mean this is where the most interesting problems in engineering lie. And so I, I love just how much, despite the fact we haven't really gotten to the point where self driving is completely pervasive, I think we're at this point where there's just so much diversity of approaches and engineering and there's some really interesting companies in this space and they're all, they're not all taking the same approach, which I think is kind of awesome actually. And so I'm very optimistic for the future and I, I'm glad that it's, there's not so much homogeneity and like the way people are approaching it just because in the world of engineering you're wrong a lot. So I'm glad that there's lots of different approaches being taken right now.

Can you ever turn off your engineer brain or is that just the lens? So I mean when you're watching a football game. No, I, I, I appreciate it, you know, I, I appreciate it. I, I, I, I in a lot of ways feel the same way about how my world view, but you know, my engineering friends, that seems to be a theme that they just cannot turn it off. And I just love to see the different applications of Watching a football game or watching theater and trying to figure out, okay, how does an engineer think about this and decipher this?

It's a fun. I think engineering is just the application of science to the real world. And what I love about engineering is its messiness. You know, it's. Mike Tyson said, everyone's got a plan until they get punched in the face. And to some degree, engineering is science getting punched in the face. And for a practitioner, someone who really identifies as an engineer, that's where all the interesting problems are. How do you scale up a supply chain? It's not just about solving the technology problem. If you think about sustainable energy and energy storage, like, all the interesting problems, I mean, there's a lot of science problems, but there's a lot of just engineering problems, like, just make it work. And then particularly with AI, you know, large language models are at once super intelligent and also kind of idiotic at the same time. And you're like, okay, how do you apply that? I was just on stage yesterday with the head of AI at Cigna Healthcare, which is a partner of Cira's, my current company, and, okay, if you have these large language models, which famously have so many problems, how do you apply it in healthcare, which is one of the most regulated industries in the world? That's an engineering problem. And I think that's just fascinating to me and the pragmatism it requires to say, hey, it doesn't matter what the theory is. What matters is the results it delivers for a customer in a regulated industry. I find it just. I could do it all day. I would do it for free. And thankfully, there's a. The free market means I don't need to do it for free.

Brett, there's a. There's a good quote on what you just said from Jamie Dimon, which is, in theory, theory and practice are the same. In practice, they're very different. Right. And I think that, to me, that's a great photo. Right? It's, like, perfect. It is, like, such a perfect quote for, like, the. Look, do you want to go and get punched in the face or not? Or do you want to talk about it and do you want to talk about what might happen? I think Your point on LLMs, that's so interesting, I completely agree. Which is like, yeah, they are 100x better than the median human in some respects. And I don't know if it's a hundred x worse, but, like, way, way, way worse than the median human and other, like, specific things. And so it's sort of like, how do you deal with something that is superhuman on like math intellect, but like, can't work for more than X amount of minutes without being prompted again and again and again or, you know, whatever. And you're like, I don't know, like, that's a new problem and how you deal with it. And I think your healthcare example is really good.

Yeah, it's fascinating because I've been a software engineer for most of my career and computers have gotten to the point where you don't really think about transaction costs that much. I mean, there's certain areas you do. If you're doing a high frequency trading platform, you probably think about it a lot. But most things, you know, transaction costs and latency have gotten so low you don't think about it a ton. We've sort of moved up the stack through Maslow's hierarchy of needs of application development, which is good. That's just progress. And then large language models come out and they're orders of magnitude slower than most pieces of technology that we've been building on for a long time. They can take seconds to do things rather than, you know, milliseconds or microseconds. And then similarly, they're non deterministic, which means the same inputs can produce different outputs. But as you said, Ravi, they're also mildly super intelligent. And so you have this thing that's expensive and slow and non deterministic. And we've been, you know, if you. Black Friday is coming up. We have a lot of retail customers here at Cira. And you know, when you think about Black Friday and being a retailer, it's like, how many nines of reliability? Which is for those listeners who know what that means. You measure reliability, like what percentage of page views are successful? And you want 99.99. And that's four nines, which is, you know, best in class. Five nines is virtually impossible to get. And then you have these models and you're like, well, they measure the reliability in eights and it's occasionally amazing and occasionally just make shit up. And you're like, okay, how do you bring that into like the day where you do all of your business for the entire year? Like one of our customers is minted the greeting cards kind of like, if you think you have a peak season, you ain't seen anything till you see mint at peak season and that. Talk about a cool engineering problem. And it's not. It's like apples and oranges, you know, do you care about quality, speed, latency? How can you Apply this technology to balance those things, and you don't want to, you know, actually hurt in any way. And I just find that, like, what a fun problem. I mean, for an engineer that's, like, you know, put me in the game, coach. Like, I got this. I want to dive into that complexity and come out with a solution that actually works. And we did a blog post recently about how to achieve higher reliability from these models, and it's fascinating to me. It's just a new. In some ways, it's not new because we've been dealing with distributed systems for a long time. But what's fun about it is just. It's this whole new class of problems that didn't exist before. And for the right people, it's just the most fun time I've ever had in my career being an engineer.

Did you see what Toby said recently? Toby, the founder and CEO of Shopify, he was recently on John Collison's podcast, and he had this incredible quote, which is like, the. The best gift you can get in your life is like, a beautiful problem that you will never fully solve. And it's just, like, exactly what you're saying. Right? And I know you are, you know, close friends.

That sounds horrible. That sounds horrible, Ravi. That sounds horrible.

No, the unwinnable game somehow related in our history because we have very similar interest.

You know what? That's such nonsense, your take on this. And you know why? You're like, my favorite moment of my life was guarding Kobe Bryant. That is an unwinnable game. It is the perfect problem. You will never fully solve it. You know exactly what to do. You can only play it well. I mean, you've said you've never felt more like, you know, professionally alive than that, but now you look skeptical, so I don't know. You tell me.

When he. When he shot 11 for 33 and we beat him to win our 23rd game in a row, 22nd game in a row in Houston, I was pretty good. I felt like I was. I felt like I won that day.

Okay. Okay.

The next time I played him, not so much, but that day, it was solved.

Yeah. It's not that you never solve it. It's never perfectly solved, you know, but I thought that. That. I just thought that was so. That was so indicative of your point of, like, the engineering mindset chain, you know, being kind of everywhere, and it's like a kindred spirit, you know, with you on this. Brett, I never asked you this before. You and Clay spent a ton of time together. Ton, ton, ton. Of time together. You decided to start this company. What goes into picking a co founder? You know what, when you guys choose each other to go do this together and build this company, first you have to make a decision. Do you want to have a co founder, both of you? And then, you know, you choose each other. And you. I think you've made an analogy to a marriage in the past. Like, I'm very curious. How does that work?

I've thought a lot about this. There's a. Who's the guy who wrote the Blind side? Michael.

Michael Lewis.

Michael Lewis.

Another blue guy. Guest, Brett.

He's amazing. I would read him. If he, Any book he writes, I'll read it. Whether or not I think I'm into the topic. His book will make it interesting.

On the back cover of one of his books. And he doesn't. I, he says this isn't true. It's definitely true, but it's like, you know, I would read a thousand page history of the stapler if it was written by Michael Lewis. And it's like, oh, 100% true. Right? Like, I want it.

I just, you said that.

I'm like, I would definitely, I'm like, I'm gonna ask ChatGPT to like, you know, write it in the style of Michael Lewis.

And by the way, Shane, your point of curiosity. He has to. What comes across, I believe, in his books is just this true, authentic curiosity on the topic. But anyway, going back to the co founder thing, he wrote this book, I think it's called the. The Undoing Project.

Yes, yes, yes, yes.

Two Israeli statisticians and psychologists. And I read that book and I found it fascinating because it was real story of partnership in my mind. And these two researchers who one plus one was definitely equal to more than two when they were working together, but also the complexity of partnership, the ego that gets involved. And even if, you know, once that partnership dissolves, they never quite did anything as great. But people are very complicated. And the reason I use the analogy for co founders of a marriage is if you visit a priest or a rabbi or a marriage counselor before you get married, you'll hear a lot of advice that's very similar, which is it's not where you are to now, it's how you grow together. And it's all about communication because if you have the marriage you hope you have, your by definition will be a very different person on the other side because no one's the same person 50 years in. And I think the same is true of co founders because, you know, you might have a technical co founder and a business co founder. Pick the most stereotypical architect types of founders and that's the value you have in that moment. But if you're successful, the real question is, you know, what do you, what are you going to be 10 years in? What are you going to be when the company's at a thousand people and how does that partnership continue to thrive? And I think it really does require a real deep sense of trust. And I think people, it doesn't. Co founders don't need to last forever. Lots of businesses have gone through co founders, but I really believe it can really be magical. And I've had really successful co founder relationships. I'm still really close friends with, for example, Kevin Gibbs, with whom I started quip. Once you're in the arena together, it's a bond that no one can ever take away from you. But I've also had complicated partnerships in business. So we spend a lot of time thinking about it, talking about it. The thing I would say is I could not imagine starting a company without a co founder. I just couldn't imagine it. It's really fucking hard. I remember the day before we launched FriendFeed and we had done this New York Times exclusive and just because of one of my co founders, Paul Buche had started Gmail and I had started Google Maps. So we had enough street cred. The New York Times was going to cover our launch and the product just wasn't ready. Like we, you can't, we didn't time it earlier, right? So I'm up to like 48 hours in a row, just writing code, trying to get up and running. I was in a fetal position in my living room thinking, what have I done? Like, not only will my business fail, but I'm going to be humiliated. You know, it's like just, it's just a really intense thing. And I think people, what's interesting about founding a company is that there's a certain vanity to it, whether you not you care to admit it, because you know, your, your brand is attached to what you're doing. And especially for someone like me who's accomplished a fair amount of my career. You know, there's this weird thing like people are sort of assholes online, which everyone knows, but somehow when you've been successful, people are really, it's like, they're like this guy, I'm just like, this guy's a doofus and da, da, da. They don't say the word doofus. As, by the way, they say other more colorful things. And so, you know, you have to sort of put on your big boy pants and, like, say, okay, I'm going to put myself out there. And it's so much easier to do when you're doing it arm in arm with someone who's on the same team as you. And there's sort of this, like, unconditional bond. Like, we're in this together. We're going to, like, hopefully succeed together, but if we fail, we're going to do that together, too. So I think it's one of the most important things. I think people sometimes overthink skill set and underthink the relationship just because a company is going to grow and you're. I think it's actually how you evolve. That relationship is more important than anything else. And what's super fun as you know, Ravi, like, I could spend every minute with Clay. I just think the guy's the most remarkable person I've ever worked with. And so what's amazing is, like, we're not sick of each other at all. Like, we call each other, like, I, you know, like, I could talk to him 25 hours a day if it existed. And so I. The coolest thing is, like, Sears going really well. I think we have an opportunity to create something that truly, hopefully becomes generational, that is outlives us. And Clay and I joke, like, our definition of success is we're sitting on a porch at his house in Napa, you know, complaining that the young people never listen to us anymore and, like, the companies moved on without us. And that's our definition of success. Success. But it ends with Clay and me on the porch. So it's like. Which is my definition of, like, lifetime partnership.

I do think just one thing. One. You know, we thought about having you and Clay together on to talk about, you know, Sierra, but we actually wanted to do it separately because I think each of you is so unique and interesting. And Clay, of course, is like, one of my very best friends in the world. I think there are moments where you guys have, like, 1% of the Napa porch thing today where somebody does something inside of Sierra that you guys describe. You're like, we didn't even know this was happening. And you have such an incredible team. And it's just awesome, right? Like, I think about some of the people that are. You guys have brought in, and so you get to see hints of that already. And I think that it's really. It's awesome. And I. I cannot think of a better moment to ask you, like, our signature closing question. I Don't know. We don't. We don't really have enough listeners to have a signature question. But, you know, the way we often close our conversations. Brett, who are your glue guys? Who are your glue people in your life? It sounds like Clay would probably be one of them. But these are the people that, you know, hold it together. And it doesn't have to be for business. It can be anywhere.

Definitely Clay, Definitely my wife, obviously, and my best friend Andy, who I've been best friends with since I was 16. Actually, earlier than that, freshman year of high school, we were taking the bus home together. His version of this is we were both wearing Nirvana shirts and kind of grunted at each other and, you know, decided we were going to start a band together. We've been friends ever since, so actually just went to the Niners game with him on Sunday, flew up from San Diego, so.

That is awesome. Shane, do you know what that reminds me of? When Brett said his best friend, that's the same answer Coach K gave when he talked about Mo. You remember, like, yeah. And when you say best friend, like, as a person in your mid-40s, you sound like you're like 12 years old, but in the most wonderful way, right? Of, like, same thing. Coach Kay and my best friend, Mo. And it's like, it's just so fantastic to hear someone talk about, like, a friendship like that that's lasted decades. And it's the same sentiment, like, you.

Know, die for you, you know, and, like, we've been through a lot. We were in shitty bands in high school together, been through a lot, seen our kids grow up. So, you know, it's. There's a different kind of friendship, you know, it's the best.

Thank you, Brett. This is. It's awesome. It's so. I very happy that I get to spend so much time with you, and you are an absolute glue guy, and we really appreciate it. In addition to all you've accomplished, I think that's another honor that we can bestow upon you.

Thank you so much, Brad.

Thanks for having me.