How YC Rewrote the Seed Playbook with Garry Tan

Transcript: (disclaimer: may contain unintentionally confusing, inaccurate and/or amusing transcription errors)

David: Hello, Acquired LPs. We are coming at you today with a very special episode with Garry Tan of Initialized Fame and before that—Y Combinator, Posterous, and has a long and very illustrious career here in Silicon Valley. We are going to talk about the evolution of early-stage investing.

Garry has seen it all. I think he started back in the early days when early-stage investing meant a $2 million Series A at a $5 million poster. Maybe that was even high. 

Garry: Yeah, it’s a while.

David: Goodness, things have changed. I'm so excited to have you here. Thanks for joining us and we can't wait to dive in.

Garry: Yeah, thank you for having me. Big fan of the show and it means a lot to me that you'd have me.

David: Likewise that you would come on. Let's just dive right in. We're going to weave in the story of Initialized along the way, but we thought maybe we'd start way back in those prehistoric days of what life is like?

Garry: So, Arthur Rock.

David: We've already done that on the Sequoia episodes. Not that far back but when was it that you started Posterous? Was it in 2005?

Garry: It was 2008, actually. 

David: It was 2008. It’s later than I thought.

Garry: 2005 I was still at Palantir, so I had just designed the logo for Palantir and built one of the major product teams. Before that, I was Stanford Computer Engineering, and the crazy story for me is that friends of mine were starting a company. And I was the lowest of the low PM at Microsoft.

Ben: What that’s like?

Garry: Great place to start, great reach. Friends of mine were starting a company with Peter Thiel. They flew me down to San Francisco to have dinner with Peter right when he wrote the $500,000 check to Facebook. He said, “Garry, what are you doing at Microsoft? You're wasting your time.” I said, “I wanted to work at a startup, but they weren't startups in 2003, 2004.”

He said, “I'm so sure this is the right thing. You need to quit your job.” He asked me how much a year I made. I told him it was $70,000. He said, “Well, how about this? I'll write you a personal check from my bank account to yours. This is your risk opportunity. Quit your job.” I said, “Thank you very much, Mr. Thiel, but I might make it to level 60 next year,” and I got on a plane and went back to Seattle. That company turned into Palantir. 

David: Oh my gosh.

Garry: I ended up joining a year later, they had hired away some of my closest friends who were way smarter than me. Bob McGrew, who now runs stuff over at OpenAI. Just so many smart people you get to meet in Silicon Valley over time. Once they hired away people smarter than, I was like, I need to quit my job at Microsoft. At the moment, when your friends are starting a business and you don't know anything about startups, tech, or how these things are funded. You say, well, I have a real job, and you say no.

David: We're talking about this with Kevin and Julia Hartz on the Eventbrite episode. I imagine you're right out of college. Your parents were probably (if you even told them about this) like, no way would they let you do this.

Garry: They’re like don’t do it. That seems unsafe. Yeah. The immigrant mentality for sure.

Ben: Garry, that's an amazing story, and it is absolutely one of survivorship bias because I want to share my story. I had a friend who was starting a YC Company. I had just moved to Seattle, and I was about to start my job at Microsoft. This friend lobbied and lobbied and lobbied. They’d come and get me to co-found the company with them.

For years, it looked like a huge mistake that I said, are you kidding me? I'd have to give back my signing bonus. I just moved here. This person told me something very similar that they would help definitely pay back the signing bonus. [...] your risk thing, they’d pay for my move. But as years have gone by, that company sold for exactly the preference. It would have been completely awash.

Garry: Yeah. That's tough.

Ben: For a while I was like, wow, that's one of several examples in my life where I really blew it on joining something early. Sometimes, not that my choice was the right choice, but unless it's Peter Thiel calling, the story doesn't always end the way you’re—

Garry: Yeah. These things are crazy risky. I often think about, wow, he was willing to pay that much upfront for an engineer. The rest of my career—even as an investor today—is now actually the inversion of that. Which is now I realize actually, it's the software engineers, designers, product people, and the builders who create the future. That's why we're able to do early-stage at all is that we look a lot more like them at that stage and so they'll pick us.

On the flip side, because I can still code a little bit and I still do design. I'm probably better at marketing now than I was 10 years ago.

David: You've diversified your skillset.

Garry: Yeah. That's right. We look more like them. Then that's the cool thing. I think that fits with the overall series. It's like we're talking about the traitorous eight—sending real typewritten letters across the country to financiers who were totally different from them. Now, what we're talking about—what you guys do and what we're doing—is we're no different. I think that time was for venture capitalists to say we're set apart, we're different.

David: We're in this ivory tower.

Garry: Yeah. The ivory tower is different. There's actually a ritualistic aspect that I was talking with Geoff Lewis at Bedrock. He has this theory that is really interesting. There is something to be said for I'm walking up the steps of Sequoia, this is what legends before me did, and I can be a part of that.

Now it's Zoom. If anything, now it's flipped. Now it's about the one on one conversation that you can have right here. That's why you have the rise of influencer investing. I know, we'll talk about it later, but that's part of the reason why I think YouTube is so important, for me anyway. I'm investing very deeply into it because I think it's a very interesting innovation in the course of how ventures are created.

David: VCs historically take a long time to catch on. YouTube was founded right around this time that we're talking about, and here we are in 2020 and people only just were starting to catch on.

Garry: 2005, yeah.

Ben: Garry, you talked about your time as a builder, and we're going to put a pin in this influencer investing and definitely come back to it because I think it'll be a nice way to round out the full story. Take us through founding Posterous, leaving Palantir, how you raised money for that, and how you went about getting enough proof that there's a there, there to invest more of your time.

Garry: There's nothing quite like seeing a super early-stage startup for your own eyes. Actually, I've always been really thankful for my time working with Stephen Cohen, Joe Lonsdale, and Alex Karp—just the founders of Palantir. Being able to build software from scratch. The more subtle interesting thing that I feel like I learned was how important it is to basically continue to hire people way smarter than you actually.

The cult making and the mythmaking of the startup very early are really underplayed. I don't feel like people talk about it enough, and Palantir, I think remains very good at it. The only cult that was stronger than our cult was the Facebook cult (I think). But it's interesting to see. Years later, that's an order of magnitude bigger as a company, which is fascinating to me. I think that actually is directly proportional. How strong your cult will result in how big your company ends up being.

Assuming you're in the right market and 10 other things that you need to survive. You need to be one of those survivors. A lot of things have to break your way.

Ben: What's an example of something that was done at Palantir to help build that cult brand?

Garry: Honestly, I think the simplest thing was even just trying to get the smartest, most capable, and hardworking people, which sounds really stupid simple. It seems like everyone should do that, but honestly, people just don't. When you think about hiring, the mistake that a lot of founders make—and honestly, I made this mistake at some level too when I worked on Posterous—was who can I get? And that's the wrong question.

You should start with who is the smartest person I know, and it doesn't matter where they're at. Because if I get them, then our self-fulfilling prophecy becomes destiny. If I don't, then it doesn't. I'm not doing myself or them a favor by not going after them. We would just go. 

People would pass out yellow legal pads, and we would force everyone to step away from their computers, and it'd be like, write down the names of 20 people who were the smartest people you've ever met in any walk of life. It didn't have to be engineering, design, or whatever. It was just, who are the smartest people you know in your life. We take it into a backroom and cross-reference it and then those are like our hit list. It's like let's go get those people. We're going to take them to dinner, we're going to take them to lunch, we're going to meet them, we're going to chop down the tree, and we're going to go get them.

David: That mindset leads to doing things like what Peter offered to you. I can't believe I've never thought of that before. Yeah, you were ungettable. You were at Microsoft, but what if you just offered to personally cover the gap in your salary? It didn't work for you then, but a certain number of people, that's going to work with. And if you're not in that mindset of okay, I don't even go to try to get this person. Well then, you don't know, but once you're like, no, I'm going to try and get them. Well, what can I do?

Garry: Yeah. It just compounds from there because smart people want to work with smart people. I think that is testament and credit to what they've been able to build over the years. That becomes a self-fulfilling prophecy and then self-impelled because if you have a group of smart people who are very driven, it's a magnet that attracts a lot more people. Even if they know nothing about the mission.

Past 30 or 40 people, they started doing really awesome campus hires. That is actually very, very hard for companies to do. I actually haven't seen anyone do it that early. It tends to be in the hundreds of people by the time they feel like they have the cycles to do it. But if this is a 15-, 20-year business, it actually is really smart to start on that campus hiring sooner. And then it pays off in 2 or 3 years when you get the Stanford section leaders for 1-6, X, A, and b.

You have the best people from CMU. You have the best people from MIT. Generation to generation, the software engineers know. I need to go there because the smartest person I knew from that year went to that company.

Ben: You're surrounded by all these incredibly smart people (to keep using the word cult) at Palantir. Why leave and start a company?

Garry: I've always also had this thing where I can't really follow directions. What I realized is that's actually a founder mindset at some level. At Microsoft, it felt like I was caged. There were things that were obvious, but I was nine levels too deep in the organization to make the call. In retrospect, the most successful people I remember from the Microsoft days were amazing at hopping to level five or level six.

They knew their palm. They knew their VP. Those people would answer emails. I don't know how they did it, but that's actually the advice I give to incoming PM classes, Uber, Google, and elsewhere. It's like, hey, this is an org like any other. You just treat people like people. That I wish I spent more time doing earlier in my career. It was like, oh, all of these people are just human beings and they have wants, needs, and desires. Actually, you as the youngest person in the org will know more about reality than that person who's been in a bubble for years, and they want that.

That exchange, I think is fascinating. I think intrapreneurship is actually quite possible, but it's probably as hard or harder than entrepreneurship. Because entrepreneurship, you just say, hey, well, I'm just going to do it this way. That was for me the inevitable. Even today, I don't really care about power or control. What I realize is when things are broken, I do need to make the call.

That wasn’t me when I was 22, 23, and 25. Even when I was a dev lead and I worked directly for the founders. It was like, even that wasn't enough because I do have a very intransigent streak, and when I get something in my head, I must have it. The people who are founders like that just know that about themselves, I feel like.

David: This is actually the perfect tee-up back to early-stage investing and how that's evolved. This cult idea, I think YC really took that whole concept and started applying it to a community of founders rather than just a company. You were then leaving Palantir, going to start this company. How did you first hear about YC? You must have been applying in 2007 or 2008?

Garry: Yea, I remember the moment when I heard about the Y Combinator cult because it waltzed in the front door of our first office at Page Mill Road. It was like a CMU friend of one of the first employees. They came by for lunch. Obviously, the catered lunch things like come on by. I think we were trying to recruit them, honestly, and they had just gone through the YC batch.

It was the talk of the office afterward because they were like, ha, ha. This Y Combinator thing, it's like $17,000 or 7%. What a bunch of chumps. That was like what people said in the office. I was like, huh, 7%. I mean, that's not a lot of money, but they are their own founders, and they did raise money afterward. That was the kernel.

It was interesting to see the mockery because here's our cult in opposition to their cult. Anytime cults have bumped up, it's actually like mockery—it's war. I think that's one of the interesting things. It's like that was my first interaction with the Y Combinator cult was actually like, oh, that's garbage. I mean, that was what my team said. Even my other co-lead, he was like, I would never do that.

That was enough to get my attention because I was like, oh, that's interesting. Today, I do this all the time where anytime there's a very strong reaction to something. There's something there actually. I heard about it and later I found myself applying just a year later.

David: Do you remember who that team was that you have met that day?

Garry: I think it was ZumoDrive, and they ended up selling actually. I ended up getting to know David Zhao. He's a good angel investor now. It's just funny actually, how these things work out, but I applied. It was really Paul Graham's essays. I started reading Hacker News. I found Paul Graham's essays and then, of course, once you own the media, you own a little part of the 20 million developers and the world's brain.

I think Hacker News was actually a really powerful thing for Y Combinator and Paul Graham. He had his essays. He converted into media. I remember that the essay that did it was the one that said, hey, Garry, you weren't meant to have a job. I was like, oh, my God. You're like putting voice to this feeling that I've had for my entire life. I have just been in a cage this whole time, and I'm ready to be let free. 

It's funny because years later, I ended up working for Paul Graham. One of the big criteria that he and Jessica would often talk about after we met a team in those 10 minutes of an interview—often we would ask, were those founders animals, and what animals were they? Which is just this crazy heuristic, but very valuable. That was the feeling I got when I became a founder which is like, oh, it's finally my call. I have been uncaged from the systems that have prevented me from doing the things I want to do.

David: Obviously, you worked for Paul. You know him quite well. How purposeful was he about Hacker News and his essays being part of this? It's so hard to remember now. We're talking 2006, 2007, and 2008—so long ago in startup time, but he knew that the media gave you this direct pathway. You read that and you were like, Paul is talking to me. These are my people. How intentional was that?

Garry: I think it was highly intentional. If you read Hackers & Painters—his original book, which was his essays—he had always been that magnet for people like him. He could write about things that didn't even have anything to do with startups, and I thought that they really resonated. I don't intend to speak for that much of a startup done, but a lot of us were nerds and so was Paul.

He wrote about that very eloquently like secondary school and high school. That whole experience for a lot of people is like being babysat, actually. It's like horrible crushing amounts of boredom. Plus a social dynamic that designed to torture people like us, actually. It's like, are you serious? You zoom out a little bit and you're like, oh, actually.

I think like that was a breath of fresh air because it's like, oh, giving voice to the type of person like me. It's like that's me. I saw him go on to find success in a world that—we want to believe the real world is like college. It's that we get to really expand our minds, follow what is true to us, and we're deeply going down the rabbit hole of the thing that interests us the most. But the reality is, you go into the world and it's actually like high school. It's unfair.

David: Maybe even middle school sometimes.

Garry: Yeah. It's worse.

Ben: It's like the American corporate industrial complex. More people work at big companies than ever have before, and there are fewer small companies. If you look back 100 years—I'm going to be ethnocentric here—say in America, people's professions were way more distributed. Way more people worked for themselves or the local thing and not 500,000 people for the national bank.

Garry: Yeah. You go off into this world, you take people who are founders—who are basically fish out of the water their entire lives, and then you give them a voice. Of course, you become the Pied Piper.

Ben: Oh, man. I remember sharing those religious texts. I remember startups equals growth and stuff. There are so many that I would like. I think I assembled my list at one point as these are my favorite of Paul's writings so that I could share them, sort of easily referenceable way with people.

Garry: I still do that.

David: Gosh, you had the Acquired book club back in the day.

Garry: That's right. All of the startups done period is also its own cult, and then we have our high priests.

Ben: Since we're going to talk about the evolution of early-stage investing here. I want to ground some truths and some numbers. In this timeframe, when you're starting Posterous, could you have raised from VCs before going and doing YC?

Garry: In retrospect, yes, and then I just didn't know, honestly. My co-founder Sachin Agarwal was the lead engineer for Final Cut. I think he actually built maybe the first version of the rebuild. He's a world-class engineer, still a world-class engineer. A great PM. For me, I had left Palantir. The funny thing about Palantir is leaving Palantir—nobody had heard of it yet.

2007, it was like, oh, the startup who knows something to do with Peter Thiel. Cool logo. Was it Lord of the Rings? What's going on over there? Actually, that's one of the very interesting patterns I find. When I left Palantir, nobody had heard of Palantir, nobody cared. It was another 10 years until it became a thing that people really paid attention to.

By then they were on a path. They had 1000 employees and made hundreds of millions of dollars. That's, in a nutshell, the magic that happens. Y Combinator was the same thing. 2005, it was a fringe thing, like if you read Paul Graham's essays. He spoke at MIT, Harvard, and a few schools. That was the very first YC batch.

David: Were you in the California batch or the Boston batch?

Garry: I was in the last Boston batch in summer ‘08. I actually lived in Mountain View.

David: No way. OG.

Garry: This was 2008. We felt like our startup idea actually had a very short fuse because the iPhone was brand new, our idea was post by email, and we both had iPhones. Flickr, WordPress, and all of these other social media services had post by email, but it was like secretpassword@ flickr.com (or whatever it was). We figured out—it was actually my co-founder who figured out—we can just do a post@ and then you don't have to remember and then it's just a brand. It's post@posterous.com. Easy. 

I think Facebook and Twitter probably each had less than 10 million users still. It was really the proto era of social networks. It turned out that we were the way people wanted to post to their networks because it was easy to post photos that way. It's funny because I can point to the month Posterous stopped growing, and it was actually the day Instagram was released.

David: Oh wow.

Garry: Instagram, on the share page today, we see it every day and use it all the time. Billions of people use it every day, but on the day it launched, we were one of the checkmarks. It was Facebook, Twitter, Foursquare, and Posterous.

David: As the share sheet official that Apple launched?

Garry: Yeah.

Ben: No.

David: Oh, the one on Instagram.

Garry: On Instagram, yeah. Today, it's only Facebook, obviously, but back then, we supported 20 other social networks. I think we were in the top 200 sites by then.

Ben: You stopped growing when Instagram launched?

Garry: It was that plus we didn't share the database properly. I was a bad manager. You know how startups fail. It's like a litany of different reasons that in the confluence, recently, I did a video about why planes crash, and it's a similar reason to why startups fail. Which is like it's not one thing. It's actually five different things that all come together, and they become intractable and the plane crashes and kills everyone.

This was like one of the mistakes or one of the things that we weren't aware of. This is the fog of war part of it and Paul Graham would always talk about if you're an engineer, it's like being a physicist. You know what the next element on the periodic table will be, but at the moment, we didn't understand. I remember pitching Benchmark this.

I sat across Peter Fenton and Bill Gurley raising our Series A for Posterous, and we had 25% long term cohort retention, which is a little low. My friends at Mixpanel say, 20 or above, you could be a top site. We're in there. We got the meeting, but we didn't get the money. It was just funny.

Ben: Was this your first round after YC or was there a seed before?

Garry: There was a seed. That's the funny thing is being two Stanford grads who were both engineers, I didn't realize it. We probably could have raised. What's funny is, it would have been just money. What we got out of YC was actually the network. The people who taught us how to raise money, it was James Lindenbaum who created Heroku.

Josh Wilson at Freedom Robotics who actually—it's funny, I funded him now. It's funny how things come full circle. The crazy thing is we got a lot of help from other people who are just like us at that stage. We got very tactical advice that was extremely secret. This is now widely known, but Airbnb for instance. I think that they were already thinking about this, but we stole users at Posterous.

Airbnb was like the batch right after us, and I remember grabbing burgers with Joe. I was like, hey, we basically just steal users from elsewhere. We look at the APIs. We're like, hey. How do we get our name in there and become a utility for other people by piggybacking on other platforms?

David: Of course, you're referring to the Craigslist offloading that they would then...

Garry: I actually blogged about it once, and then he was like, hey, dude. Could you just delete that?

David: Oh, that was you? Oh, my gosh. I remember reading that blog post. I didn't realize that you were doing that.

Garry: No. This was like 2008, 2009, and then I think it ended up coming out much later. But the reality is you have to go where your users are, and I'm glad they did it, honestly.

David: Craigslist wasn't doing it.

Garry: They weren't. It was a bad UI. I like unbundling. It's interesting to just see how these networks evolve, and it's like secret knowledge. That is this essential—aside from believing X and nobody else believes X, it's also like secret know-how. That's where probably what you guys are doing with some of the incubations that you are doing with networks.

Honestly, what AngelList is doing and OnDeck. There are just so many spearheads, angel track by the first round. There are so many different forms of this that are very powerful today that didn't exist back then, which is awesome. I think it's awesome. A thousand flowers are blooming around both the cult themselves and how to make the cults. I think that's really setting the world on fire. All of that, I really applaud.

David: Okay. You're armed with secret knowledge. You then raised a seed round at the end of YC?

Garry: Yeah. It was actually Satish Dharmaraj who's an MD now at Redpoint, and he was an angel. I think he had just sold his company. He had created JavaServer pages before, and then it was Bill Lee, who actually runs Craft Ventures now, Eric Han. There are really strong angels who probably would have backed us without YC.

Ben: Listeners, you'll know Bill Lee's name actually from the SpaceX episode. 

Garry: Oh, yeah. He was the first guy off of the first Tesla I think in one of the documentaries. He was number one off the assembly line. It was interesting because these angel groups were—Eric Han had funded all of them. The joke was Eric and Satish were SMTP, and then Bill actually started as a founder doing NNTP. He was at Berkeley, and they basically could have chosen HTTP.

Ben: Self-aware gains? Is that right?

Garry: He was doing news servers. The hosted news service for all the ISPs. I think that was the first startup that made him hundreds of millions the first time around. It's just funny. All of this is very nerdy and around literally the protocols. He's joking. Would always be, oh, man. I really wish I chose HTTP or SMTP instead.

Ben: Much more widely used protocols. In this 2008-ish timeframe, raising a seed round was not going to Initialized and going to (frankly) in the northwest, it'd be PSL Ventures. Which obviously I've got a big horse in that race.

David: There are no seed funds.

Ben: It's not like there are 100 seed funds out there you go pitch. It's like these angel groups or super angels. Is that the right time frame?

Garry: Yeah. I guess the advice back then, which I feel is still true is if you're going to have a Series A investor, have a couple, or just have zero. We ended up having zero. That was the right thing for us.

David: It's so funny, you were thinking of it as Series A, even though today for sure, that would have been a seed.

Garry: Oh, sorry. Actually later, Satish became a VC. We were one of his first Redpoint Series A’s, which I think still happens now, honestly. We did it a few times for Initialized. I think that that's going to continue to be something that happens. It's like you're going to have seed investors who are successful. They're going to build up their LP bases, and they will do a couple of rounds in a row.

David: Sequoia does it now, why wouldn't anyone else?

Garry: Yeah.

David: Did you know at the time that Sequoia was funding a lot of YC? Was that talked about? Was that known?

Garry: I guess, basically, at some point, Paul and Jessica had put all of their personal money into YC, and they knew it was working. What's interesting is even by 2008 when we were pitching—people even for the Series A, but certainly at the seed, we would go to the VC's offices and we'd meet with partners. They would actually just be like, hey, you guys are like Stanford plus worked at Apple and this other startup. Why did you take YC? Which is very interesting because, for us, we were like, oh, well we want to be like Dropbox and Reddit. 

It was just funny to see. It was three years in, and it was clearly multiple multi-billion dollar companies had already been funded by Y Combinator. But the VCs that you would meet, they didn't read the headline yet.

That's the other trend that seems to happen over and over again. The interesting thing that is happening is not being written about yet. You'll find out about it in two, three, or five years if you're lucky. That's probably good for us because that's how we make our money actually. How do we find this stuff before other people are onto it? That was always true. It just takes forever. 

These days, I just see the 10-year overnight success all over the place, and it's like, you'll see. And then it comes out of the fundamentals. For Y Combinator, I was the customer. I was an engineer who could build anything. That was the community I wanted to be a part of. It was probably six years. 2011 was the year that people realized that Y Combinator was legit, but that was the year—out of dumb luck—I ended up working there. 

Ben: Take us through that decision. You sell Posterous to Yahoo. Obviously not the outcome you had dreamed of based on early traction, but that chapter of your life is closed. How do you make that call?

Garry: Instagram came out in 2010. We flat-lined. Me and my co-founder are having a falling out, which I've actually talked about. Wrote about it in TechCrunch. I've done a couple of videos about it. I should probably do one with The Chin at some point. We weren't really on speaking terms for a good number of years. We're friends now. 

That's one of the problems with startups is that they become so encompassing. They embody your ego, and then the problem is when things don't go right, you start looking for people to blame. This is very common, and it's almost uncontrollable, which is really hard. But it also helps me now to be able to say like, hey, this might happen. Be aware of it. Maybe that'll help. It was basically, either I had to go or he had to go because we just disagree on how to pivot. It was clearly not going to work.

I ended up going to work at Y Combinator instead. I initially was just designing homepages. I just needed a place. They’re paying me, I think $30,000 a year. I was just literally part-time contract work to come in and help the companies. But I just got energized working with people and talking with them. It's crazy because these companies turned into great companies. I remember working with Joseph Walla at HelloSign. We were designing their homepage and first-time experience. Same thing with GOAT today, which is an amazing, amazing marketplace. 

Eddy and Daishin were so stingy. They were from LA. They didn't even get an apartment in San Francisco. They actually crashed with me at my apartment for a month or something. It was cool to just be around smart people who were making things that were awesome. When SV Angel and Yuri Milner came along and said, “We're going to fund every single company at YC.” I think Yuri Milner came in on an Anybot, one of these telepresence robots. He was in Moscow and it was nuts.

David: I think I was at DSP at the time, and I'll never forget. I was in, I think it was called Sports Cafe. It was right across the street from the new GSP. I used to eat lunch there for a lot of days. I'd see David Shaw—the football coach, and a bunch of people would be eating out there. I remember one day, they clearly had a football recruit coming through. This was like 2012, 2013. He was there with his family. Usually, you would see coach Shaw with these guys. But he must've wanted to go into tech because instead of him being there in person, he was there on an Anybot in a Sports Cafe. That was the pitch to come to Stanford.

Garry: That’s awesome. How funny?

David: It was amazing.

Ben: Was that Trevor’s company before YC?

Garry: YC’s west coast office was split with Anybot’s for a good number of years. There were awesome robots in the back, and dinner and slop in the front. It was bad chili. I shouldn't bag on the food that much. I always thought it was fine, but that's like the inside joke. It was Paul making the stuff himself for the first few years. 

David: When you joined in 2011, that was the beginning of the inflection point. 

Garry: That’s when VCs figured out that this was someplace they needed to pay attention to because one, you're lucky; two, you're good. I mean, Dropbox people we're like, oh yeah. They got lucky. They’re a bunch of people who get lucky. When Airbnb became a giant marketplace and a billion-dollar company, then everyone was like, this is the thing.

It was very interesting to see the number of applicants went up by a lot. Honestly, it was a different thing. $17,000 for two people, not a lot of money. My YC batch in the summer of ‘08 was the least successful batch on record. Even more unsuccessful than anything in ‘05 and ‘06 because Lehman died. We closed our angel around the day Lehman died. Going from $20,000 to $170,000 was a really big game-changer. It really helped the companies really fundamentally.

Ben: For folks who don't know, the $17,000 came in on the same terms for that 7%, and then the rest came in the form of the convertible note. 

Garry: It was uncapped though, so it went into the next round. That was how powerful—I think it’s since been blended. Now it's 7% for the $120,000 or whatever it is right now.

Ben: Okay, and just with the amount of early-stage capital that's flooded in, they’ve basically needed to do that to not look like a gigantic rip-off. How does that move?

Garry: Basically, Sam Altman and I had to take that program in-house. We raised a vehicle from institutional LPs in 2014 to take it over, and before it was always a handshake. It was like, okay, Yuri gets 75, Ron gets 75, and then afterward, we added Andreessen.

Ben: It was just Yuri and Ron when this deal happened.

Garry: Harj actually negotiated it. Harj is a partner, and he's back at YC now.

David: Wow. That must have been wild. You joined the cult, you go through it, and then you're back as a cult leader.

Garry: I was a very low priest.

David: A hype man. 

Garry: Yeah.

David: What were the conversations you guys were having? To my mind, at least this was the moment where all of this started changing. Seed fund started forming—all of this. I think the decision on the part of YC—maybe it wasn't even a decision, maybe it just naturally happened—to start bringing in so many more companies. You had so many more applicants. What were the conversations you guys were having about that?

Garry: Paul Graham built all of the first software, and then Brett and I basically rebuilt it right after him in 2013, 2014. But it was always run by software. Even the first batches were Arc. The interesting thing that I remember is every single batch, Paul Graham built it himself.

Ben: That’s Paul’s own flavor of lisp. That’s what Arc is, right?

Garry: Yeah. That’s his own flavor of Lisp. I think every single year, they basically tried to keep the acceptance rate constant about 2%. Sometimes it floated up above to 1.3%, 1.4%. But the rate of innovation is just going to go like that. If anything, as interest and it goes up, there will be more innovation.

David: Success rate is always going to be the same.

Garry: If you select fewer companies, you're actually reducing the rate of innovation, which is the opposite of what we're trying to do. There is a method to the madness, and actually, there were definitely moments where YC felt broken. The batch that Coinbase and Instacart, both of those were maybe a little bit too big.

David: Was that 2012, 4 I think?

Garry: Yeah. I think there was one that blew up to maybe 75 or 80. Then that one was so big that the transaction costs of reloading who the person was, and what their company was became a little bit too great. That was actually when Paul Buchheit—the creator of Gmail decided, hey, let's try doing a little bit of sharding, which is very straightforward. It's like, let's separate this out into groups, let's have group partners be the people who work on their cohorts. And then there's more continuity in those less transaction costs, which just made sense. 

David: Oh my God. You guys took the Constellation Software approach to scaling YC.

Garry: Yeah. A little bit. I think they're still doing it. I think it was Geoff Ralston and Michael Seibel who decided to break it up and in not just one day, but two days. Now there's Tuesday and Thursday. The idea for scaling continues. I remember very specifically when The Social Network movie came out.

Ben: That changed everything for all of us. 

Garry: Isn't that funny?

Ben: You grew up as this freaking nerd. You just like computers. This was junior, senior year—I was in college. This thing that I've been doing on the side that was extremely uncool, which is making an iPhone app instead of being social. Everyone that looks at you and they're like, so, are you going to do that? That’s what’s going to happen to you? I’m like, of course not, but it made the whole world wake up to this cult that was going on.

Garry: Totally. That’s our eternal September. That movie. It's over now. We should have packed it up back then.

David: I was just talking with friends the other night who rewatched it. I need to come back and rewatch it. They said it was just as amazing in 2020. It was like Liar's Poker. When Michael Lewis wrote Liar's Poker, he intended that to be lampooning the industry and like, don't go do this. And it had the exact opposite effect.

Garry: Yeah. Pop culture is powerful. I guess that just goes back to it's all mimetic desire. We're a bunch of monkeys. We basically just ape whatever we see, and my desires come from other people.

David: Oh my gosh. Well, they go Facebook at its core.

Garry: That’s right. The most mimetic company ever.

Ben: We'd be remiss here if we didn't spend a good amount of time telling the genesis of Initialized.

Garry: Oh yeah. Very accidental.

Ben: We're caught up-ish here. I think 2012 was your first fund. Something very small like $7 million? Is that right? 

Garry: Yeah, Harj, Alexis, and I were the younger partners at YC. It was actually Alex Bangash at Trusted Insight who gave us our first money. I think Jessica actually introduced us. Jessica Livingston actually introduced us. All of the partners were investing their own money. We were just younger partners. Twitter hadn’t IPO’d yet. We thought, well, maybe we would just try it out, and we'd write $50,000 or $100,000 checks.

Alex taught me a lot about LPs and this whole other world. He was a little bit Y Combinator for my fund life because I think I really had no conception of the macro picture of how the world's capital was allocated. Just the degree and how much there was—it's just staggering. All of tech is maybe $100 billion—may be early-stage anyway. But the global economy is $10 trillion. Trillions and trillions of dollars. We are truly just this mote of dust on just something far larger. 

But the kernel of it, I read about working for on-hedge fund software for Peter Thiel, which was like Joe Lonsdale made me read The Dollar Crisis. He made me read about what happened in 1971. Today, of course, you go on Twitter and you see these graphs of the working-class wage is flat, and then education, housing, and asset price bubble. Something really did happen at that moment. I guess that's why Bitcoin was so obvious to me when I saw it because something weird is happening. It's water. We're just swimming in it.

You should have Alex on the show because he's a fascinating character. He worked at Bell Labs. He's an engineer, too. He went to Wharton for his MBA. Then he discovered, oh, there's this whole world of professional institutional LPs who allocate most of the world's capital. He basically wanted to help them get into funds like us. He took a really big chance for us. He was like, “You're at the right place at the right time.” You can deploy capital into companies’ pre-demo days and have worked with them for 10 weeks. 

Working backward, that seemed obvious. On the other hand, in that moment, he was the one who gave us one of our first institutional LPs and put together the first $2 million, $3 million of our $7 million fund.

Ben: The important insight that you're talking about here is, sure, the $100 billion that exists in our early-stage universe. Even now when there's tons of money that are flooded into venture capital represents this line of dust around the edge. On the other hand, the NASDAQ is a $10 trillion market cap exchange, and $3 trillion of it was funded by Sequoia. These companies are just this little dust, and that’s why this ecosystem wasn't so magical.

Garry: That's why this isn't my thesis, Y Combinator’s thesis, or your thesis, which is the world is being remade by technology. What's funny is I'm still operating by the principle that it's all nerds, but actually, there are some jocks in there too now post The Social Network. That's how it is, and that’s okay.

David: Now it would look like, no-code and everything. It's just been so democratized.

Garry: That's a great thing. The reality is, I looked around and I didn't realize there was so much money, but I sure didn't know there were so many problems. Coming up, just your friends, the people you grew up with, they're all the people who can build. Don't tell me that there's too much money chasing too few people and too few good ideas. Go outside. There's a billion problems that could only be solved by people.

Don't tell me there aren't enough smart people. If anything, the people are just trapped in these systems and basically machines that don't allow them to fully realize their potential and solve the problems. How do we fuel this whole machine and make it better through money?

I realize that that's both highly capitalists and highly technological utopian. But that is what I deeply believe at my core is this is what the world needs the most of, and so properly applied capital allocation, plus finding the best builders in the world.

David: That's one of the reasons we do this show. We tell these stories.

Garry: I actually now realize I'm a weirdo. You guys are weirdos for believing this. We are weirdos, but that's fine. I'm fine with that.

Ben: Non-consensus and right. I think I've probably said it a dozen times on this show, but you have to be non-consensus.

David: That's right. The younger guys within YC, you and Alex start thinking about—we work with these teams. We know them. We're getting this amazing team, this whole cult within Y Combinator. We should have a fund. Had Paul and Jessica thought about this before?

Garry: They quietly invested, Paul Bucheit, and Geoff Ralston. We just wanted to write checks of $50,000-$100,000 like they were kind of like theirs. It’s just other people's money. We were just seeing if it was good. I feel good about Instacart because I definitely flagged them and helped them get into YC about halfway through.

Ben: That was like a 14 times pivot, right?

Garry: Oh, Instacart. Instacart was kind of that the whole time.

Ben: I thought it was something where he had 14 different ideas and tried all of them.

Garry: Oh, yes. That was before I even met him. He sent me a 6-pack of beer using the Driver app. He had the Driver app and the Consumer app working. He had thousands of items with photos, inventory, and price data. I don't know how he did it. Did Stripe exist by then? I think it did, so you could put a credit card in. He sent me a 6-pack of beer using his service, and it was operating with two drivers in Mountain View with the local Safeway. 

I was one of the first 100 people to try it. It was a basic intelligence test. I mean, you're reading 1000 applications every six months at YC. It wasn't like we didn't see that idea. Uber was out. I think Zimride was about to do their Lyft version, that was actually just normal people in cars. It was a new idea, and the key insight was you could do it asset-light. You could deploy a large workforce.

David: DoorDash did the same thing. 

Garry: Yeah, exactly. That was the year after. There was this broad acceptance that maybe it was coming, but it was still like we don't know if it's really gonna work. And then it turned out yeah, it worked really well. If anything, a lot of the people who missed it said, oh, it's been tried. It's called Webvan. They didn't spot the “Why now?” That's why “Why now?” is such a crazy important question. We put it and highlight it in our memos. It's like, “Why now?”

Ben: What was your “Why now?” On Instacart, why that point?

Garry: It was just literally 60%-70% of people who had access to phones could have smartphones. That was the first moment where you could actually have a purely mobile distributed workforce that was onboarded literally off of Craigslist again. There's probably a Craigslist thesis in there, but it's been out there for a long time already.

Ben: We talked a little bit about the founding moment of Initialized. You were only part-time at this point, right? 

Garry: We’re full time at YC. That's why it was kind of tricky. We had inside information. We work with teams for 10 weeks. It was pretty straightforward to know these people are awesome, and this is a real market. I didn't have any conception of how hard demo day is until I left YC because you are looking at this list of hundreds of teams, you know nothing about them, and you have to meet them. 

Whereas when you’re on the inside, I met with every single team. Everyone wanted to help with their homepage. The question was very simple. It's like, shoot would I go work there? And if the answer was yes, meeting Brian Armstrong was crazy, stupid, simple to be like yeah, I want to work for that guy because he was a domain expert. He did anti-fraud at Airbnb, a great engineer. He built the first thing. He was obsessed with the Satoshi Nakamoto white paper. I had bought Bitcoin off of Mt. Gox, and it was a disaster.

I was like, this is an obvious need, so yes. It was totally undersubscribed. I went for a walk with Brian recently and he was like, one thing we didn't do as well back then that I think is still a really difficult thing is how do you tell a YC company what their valuation should be on demo day?

There's always these conspiracy theories that YC partners are just trying to get it up as high as possible. But the reality is these are rational, self-interested actors who are also maybe doing it for the first time. It's Lake Wobegon up here. Even Brian was like, "I didn't really know." 

Ben: This was before the safe, right? 

Garry: This was a note. I think the note was still what people did.

Ben: The founders would set the terms. The founders would say, "I'm raising a note with this cap."

Garry: Yeah, I think it was in the mid-teens, and it was really high. He didn't get his [...] on demo day. That's a credit to him because it didn't matter. I thought the lesson for some time was if it's hot, don't do it. And that’s actually also wrong. 

Ben: Why was that wrong? 

Garry: It's not that things that are hot can't be good. It's that if it's hot, it's not necessarily good.

David: It's actually not an indicator. 

Garry: Yeah, and that's the problem. If you have very little data and then we're social animals, then we think that we read signals into things that are random. That's the danger with things getting hot. It's weird now because when we were nobody investors—it was just very bewildering to me that people would be like, oh my God. They are investing. We have to invest. Now they're doing that to us and then like, this is just a guess, guys. I'm not complaining because it's good. It lowers the cost of capital for our companies. All of these things are great.

Ben: Let's just give you a moment to flex here. For people who aren't as familiar with your work, that first fund was $7 million. I think your most recent fund was $230 million, and you've funded companies such as?

Garry: Coinbase and Instacart in the first fund. Opendoor and Cruise Automation in the second fund. Flexport also, Patreon in that second fund.

Ben: Right. So not shocking to any listeners now that people would take signals from when you are participating.

Garry: Yeah, totally. I appreciate that, but the reality is even now, we do not have a crystal ball. This is the opposite of what I should be saying on this channel, but the reality is we're just trying to be rational human beings. I actually really like how Matt at Benchmark talked about it, which is, this isn't about like seeing things that aren't there. It's about seeing things that are in front of us actually.

Ben: Seeing the present more clearly.

David: I think it's that venture capital is not about seeing the future. It's about seeing the president very clearly.

Garry: Yes. Thank you. I am notoriously bad at verbatim quotes, but that was the general idea.

David: It's so true, though. The second fund was in 2013, right? 

Garry: Yeah. We're still working at YC and people would come out. Honestly, I also have really underestimated how easy it is to just write a $50,000-$200,000 check. Then the second fund was $40 million. We learned the hard way that percentage ownership matters a ton. 

I mean, that fund is going to be great because I think they're going to be six or seven unicorns in the end. It's still going to be a great fund, but just not as good because it's a way bigger fund. We didn't 5X the size of the check. It's just math at some level beyond that.

Ben: Yeah. Garry, why does percentage ownership matter? Because this audience is full of founders who have never been on the VC side of the table and who have heard from a VC, sorry, we're not going to be able to get there because we can't get a big enough chunk of your company. A lot of founders are going to look back across the table and say, the price is low now. There's going to be a huge multiple. Why do you need to own so much of it?

Garry: I guess at different stages, it's different reasons. At that stage what I realized was LPs are still looking for things that look like Sequoia or Benchmark from like the ‘90s. There is a little bit of still cargo cult thinking and LP-dom. 

I think it's changing. People like Alex Bangash and Trusted Insight. There are institutions that are coming around to the idea that seed is its own asset class, and it can be done in a different way. Because people are still thinking a seed is like being a Series A investor, you have X number of partners, and you can sit on X number of boards.

Actually, at the early stages, it's totally different. I think those rules shouldn't apply quite in the same way. But all that being said, until you have the DPI and this is top 1% or top 5% returns, which I'm blessed to have now. At the moment, raising a fund is hard. If you talk to any VC, it takes years. We're lucky. When I left YC, it took us nine months to get to a first close, and then hit the hard cap in a year and maybe two months or so. That was really fast, actually, and it's hard. You are sort of constrained.

David: By and large, the large pools of institutional capital are not the types of institutions that embrace doing things differently.

Garry: Yeah, and the DPI results are still from 20 years ago. I think this will change in the next 10 years, but for now, people are still going off of what they know is true. Again, it's a long time scale. It's a 10-year overnight success. This will change, and it's perhaps changing right now.

Honestly, AngelList is doing a big chunk of it. I respect the hell out of Naval. He is pushing forward the state of the art because he is recognizing the world is absolutely chock full of capital. We have a fully institutional space LP base, and they are awesome. It's awesome to have an institutional LP base of the top universities. But it's being able to unlock entire tidal waves of family offices that didn't have access. 

Honestly, 500 Startups' and Dave McClure and Naval with AngelList—all of these are unlocking that. Dave was the only person who would get on a plane and meet family offices in all parts of the world. I think that's under-recognized. It really is about access and the world is full of money. 

Now, the miracle is not only happening in Silicon Valley. It's happening all over the world. This is actually the core story of early-stage ventures in the past 10-15 years. It used to be all of us trying to copy what Sequoia did 20 years ago and then today, we get to bounce around the ball. We get to find our own markets.

The only rule is still—what do they say? You gotta put moolah in the coolah. You still have to get the DPI back, right?

David: That’s a Douvos quote.

Garry: I agree with that.

Ben: Can you define DPI for us?

David: It's distribution to paid-in capital.

Garry: Yes, that's right. That's the actual multiple that you get out.

David: There's TVPI, which is the total value of the fund. That includes markups. Distribution means you have realized this exit.

Garry: Yeah. You can't eat on TVPI is the thing. Your endowments cannot actually pay for salary, buildings, or anything that they actually use with TVPI. They just get marks, and marks are actually even more dangerous. 

I do empathize a lot with the LPs because it's a very bewildering time. Software is eating money right now. People are starting to say that incumbent businesses in the public markets that do not have tech at the core of them maybe won't have future cash flows in another 10 years. That's starting to happen, which is the opposite of what I thought. When Twitter IPO’d, I gave the money to my wealth manager. I was like, get me hedge. And it was like absolutely. It was 2013. It's the stupidest possible time. There is easy 10X in tech.

Ben: I want to offer a devil's advocate position to the notion that you're constrained by the number of boards. You were basically making an argument that an old school way of thinking is that you have a constrained number of bets you can make in a portfolio. But if you look at things like 500 Startups where you're making very small investments, having very small ownership stakes. A more negative way to phrase this would be spray and pray approach. That can produce great returns too.

Garry: It's not the approach we're choosing. We've been very explicit about not doing that.

Ben: Yeah, I wanted to ask you directly. How power-law distributed are your returns? If we look at Initialized II or Initialized III—there's data on Initialize I, so we won’t go with that. How much of the returns come from one, two, three companies?

Garry: For Fund I, it's 70% Coinbase and 30% Instacart. We got 2X DPI. I think 11X or 14X. I'd have to double-check with the TVPI on it. We're really happy about it, but it's also easy because it's a small fund.

Ben: It's reasonable to say that the rest of the companies basically contributed no cash return to the LPs.

Garry: Yeah, and it's tricky because there are some things in there that did hit billion-dollar status and then just got wiped out like benefits.

David: Right, this is the difference between TVPI and DPI.

Garry: That's the TVPI versus DPI. The TVPI can go away.

Ben: To bring it back to the beginning of the episode, this is what happened to the company that I decided not to join when I decided to keep my job at Microsoft for another couple of years. Paper markups are not liquid.

Garry: Yeah, it's tricky. I think going back to the spray and pray argument, it's tricky. SV Angel is one of the best investors I've ever seen, and I respect the hell out of it. I actually think their numbers are amazing, and they don't take an institutional LP-based at all. 

BoxGroup is another that I think is doing it really, really well, and they don't have percentage ownership. I talked with David Tisch recently. He's doing an amazing job and blazing a different trail. 

The other thing is you can get ownership as long as you're in it. That's the other way to make it work, and I think SV Angel is a good example of that. They've made an incredible amount of money on SPVs. I don't know as much about how Box is doing on that, but I suspect they are also doing amazing at having access and then just doubling down on the winners. 

It's not that different than what we were doing when we were first starting. You just have to have the relationship and be in the right pond. The rest takes care of itself.

Ben: Like you said, it's access.

David: Since you've been on Initialized, you raised $125 million fund III, right?

Garry: Yup, $125 million. That was the risky one for us. That one was that you're not going to be at YC anymore, and it's a much bigger fund—jumping from $40 million to $125 million.

David: Then to $225 million and then $230 million for the most recent.

Garry: That's right. 

David: Now this whole early-stage landscape looks so different. You were one of the first. I mean there was a [...] baseline probably around the same time. Maybe a little earlier than you started.

Garry: I love those guys. I really respect what they do.

David: They're so great, but they're like such a small group. Now, you throw a stone in San Francisco. It lands at the doorstep.

Garry: Yeah, I'm an LP. I know a bunch of them. They're my friends. They have their rolling funds, and I got to support them.

David: I guess at the same time the big funds, which for a while kind of abandoned seed, they're now back in force, at least that I see. They're leading seed deals all the time. How do you navigate this as a founder these days?

Garry: I'm excited. I think it's great. What's funny—and this may be a dangerous thing to say, but I'll say it anyway. It's like I made all the money that I really care about. I don't need a lot. The scary thing about ventures is if you spend enough time here, you'll start finding people who will never be happy with any amount of money and power. 

To me, this is the most growth mindset industry that exists on the planet. Let's make the whole world a lot bigger. The cool thing is the world doesn't need me to believe that, it's just happening on its own, honestly. All of the people we talked about—whether it's Paul Graham, Naval, or so many people—they're actually just coming at it. They’re bringing more capital. They're bringing more smart people. It's not just Y Combinator, whether it's Alchemist. There are so many really great places. 

What you guys are doing in Seattle. It's just amazing to see new places. EF, Entrepreneur First, they're minting companies that are great today. I think the whole market is continuing to grow, and the most dangerous thing in this market—and you meet investors all the time who act like this—is zero-sum. They're like, oh my God. I hate those people. They're crowding me out. They're increasing the valuation. The post-money is unbelievable. It’s absurd. This is going to end in tears.

What I realize at some point is you will always sound smart. You will be the sage when you are negative. People want to hear it. You will always have an audience when you are the permabear. But permabears don't make money, is the thing. I would rather be as big tent as possible. It's like every geography in the world is having startup activity today, and that's beautiful. I think people can escape. Basically working in a big company is a little bit like going to work in prison. I mean it's not that bad.

Ben: But you really embrace that.

Garry: It's true. That's my lived experience. It's probably different from other people's, but it's just true to me. I want that freedom for everyone. 

I believe in the individual who's a builder who can run their own destiny, but they have to do it all. It's a tricky thing. It's not enough to be a great engineer because that's like being a great cog in one wheel. You have to be the machine builder, and that's so much harder. You can't just be good at design. You can't just be good at engineering. You've got to be great at all of it, and it's so hard.

Ben: All right, I'm going to get super esoteric for a minute. Let's take social media companies as an example. There was an explosion of Facebook, Instagram, Posterous, and Tumblr—these companies that eventually consolidated to very few, if not one. Between 2011 and 2019, let's keep rolling funds out of this.

We had probably $200, $300, $450-ish million seed funds get started and raise a fund one. Are there dynamics at play? Are there winners that take all dynamics or network effects that are going to lead to a consolidation in seed investing the way that it did in your previous life with blogging? Or do you think that allocating capital is a different thing?

David: There used to be a lot more accelerators.

Garry: Yes and no. I can't really tell. It feels like there are as many accelerators today that are being started. It's just that they peter out. Will there be consolidation? The reality is I hope not, because, in the end, there's this idea of the Series A crunch. But the thing is, at that point, it's just one of the metrics. 

I think seed gives people a chance, and then that's all they're really entitled to at some level. Actually entitled is the wrong word because people aren't actually entitled to it at all. They shouldn't be anyway, and people do feel like maybe they're entitled to money now, which is brutal, and that's scary to me.

David: If you look back over this whole period of history, we've been talking about all these step function changes of YC gets started. There are all these companies, 12 a year (or whatever) that are getting started. Well, that's way too much. You can be bearish on that. 

Then why YC starts taking a lot more companies like, oh, man. Seventy companies a batch, no way that's going to work. You'd be bearish on that.

Garry: But there are multiple billion-dollar companies in that one, right? Which is nuts.

David: It's like all these seed funds, all this capital comes in the space. Man, way too much capital coming in the space. That's not going to work. But if you look back on all of these cycles, they all worked. Their power-law dynamics to each of them, but they all worked. Net a lot of value was created by all of these expansions.

Garry: I think there will be a new Series A firm. There will be new series B firms. I think the seed firms will grow up. We've grown up from writing $50,000 or $100,000 checks. Those companies turned into unicorns, and they allowed us to create a firm that I look at what First Round has done and I respect the hell out of Josh. 

One of the first things he did was he took me out to dinner, and he was like, "So you got this $125 million dollar fund. What do you need to know?" I'm like major, major respect for that because that's the most growth mindset thing, and he didn't have to do that, but even today he's like that. 

I'm like big tent as heck is the right thing. This is the most growth mindset thing in the world to be doing. Then it comes back to the founder. Can you make a flywheel that works, and does it solve a problem? It's not like there aren't enough problems. There are a ton of problems. 

I'm so long on software that I think we're just scratching the surface. Even today because we funded Instacart, now what I realized is we might be able to integrate the whole supply chain in a way that's never happened before. We funded Shelf Engine, which is a grocer to the distributor. We funded Silo—which is a distributor to the grower, and then Instacart. 

Even within the store, we funded Standard Cognition, which is cashless checkout, and all of these things work together. I didn't have a thesis about this. My only thesis was that really good engineers are going to go and remake all of the planets. I think we're scratching the surface on it. All of the other stuff about, oh, are there too many seed funds or too many startups? It's up to the founders, honestly. Hey, I was a founder. Luckily, we did end up making money for our seed investors anyway, but I screwed up so many things. 

Even when you screw up I got to buy a house in Noe Valley because of my screw up. That's the weird thing. It is a crazy, crazy privilege to work in tech. It's a crazy privilege that I never acknowledged before that we can build software and experiences. It is a privilege, and that's why people hate us. That's fine. People hate tech today because it's scary. It's pushing the world in ways that are unintended.

Ben: Being a builder is totally like having superpowers where you can get leverage on your abilities to achieve things that humans previously couldn't do, and you're right. Being able to wield that, particularly when people get sole control over their company, it's scary to the rest of the world.

Garry: Startups that become tech giants, they are basically autocratic dictatorships. The lucky thing is we get to fund the next thing that arms the rebels and fights the good fight. The hard part there is either you die a hero or you wake up to find yourself the villain. But the cycle continues.

David: I think it's time for a social network part two.

Ben: That is a good place to leave it. Garry, before you run, tell us about what you're doing on YouTube. And close that thread from earlier on the influencer investor concept.

Garry: Oh, God. I feel like standing on the shoulders of giants—almost certainly—whether it's Paul Graham's essays, Mark Suster’s or, Fred Wilson—this was all new, honestly. 

It was the ivory tower. Now, it hasn't been for a while through blogging, and I think that direct access to people is very, very powerful. I met Casey Neistat. He came by the Initialized office. I loved meeting him. He's a really, really engaging person, but to see how our interns interacted with him was fascinating because it was like they had been friends for years and years. He'd live stream his life for years and years, so they actually were. 

I find it very powerful to be able to just spend 10 minutes with people every single week for as long as I decide to do this. See where it goes. I get very depressed thinking about the state of tech, capitalism, and all this money. But I realize there needs to be the power for people to create jobs and to create new products and services that solve problems. I still believe in that.

As much as there's black mirror out there, I do think that there is a space for someone who can help people just get to where they should be. If Paul Graham were to do video, I'm never going to do what he does. You can't imitate it. 

I'm going to do the Garry Tan of YouTube, and then we'll see what happens. I hope that it leaves behind something that's more powerful, more valuable to founders. I'm trying to tell my personal story and honestly, I've been trying to be as vulnerable as possible. These are the reasons why Posterous didn't make it, and honestly, we could have been it.

We could have exited for a billion like Tumblr, or we could have run the table like Instagram. I could have been a contender. But let me tell you about the places where I didn't get it right and here's where founders don't get it right. 

My job is not just money, but also to help you get to that place where I want you to go where you want to go, which is the promised land to get to a billion users, to make a thing that really matters. That's a tech positive, capitalism positive message that I just believe. I want more of that. I think the world needs a lot more of that.

Ben: Well, I'm uplifted. I feel much better about my next meeting. Garry, thank you for coming on and joining and spending the time with us today.

Garry: Thank you guys for having me. We'll do it again.

Ben: LPs, we'll catch you next time.