Coinbase CEO Brian Armstrong on Cryptocurrency and the Future of Decentralization

In May, I sat down with the co-founder and CEO of Coinbase, Brian Armstrong. Brian started Coinbase in 2012 when Bitcoin was $2 and almost no one knew what it was, let alone believed in it. Today, it is the largest cryptocurrency exchange in the United States, and earlier this year, it went public at a nearly $100 billion market capital.

Garry: Well, Brian, thanks for hanging out, man.

Brian: Thanks for having me.

Garry: Where do we even start? Obviously, we’re on the other side of an incredible IPO. Congratulations, man.

Brian: Thank you.

Garry: I don’t know how to thank you enough. Just even being able to meet you and be a part of what you created, it’s unbelievable.

Brian: Well, it’s mutual. You helped me out a ton in the early days. I don’t know if people know that, but you were my first CEO coach. You made a big investment in Coinbase at the time, before it was really obvious, and that was just me showing you a bunch of numbers on a screen and asking, “What do you think about this and this?” You were giving me advice every week, and I’d come back. I remember at a certain point, you saw some numbers from a growth experiment and said, “You should show that to investors. It makes me want to invest, and it might make them want to invest.” Then, a light bulb went off my head, and I thought, “Would Garry invest?”

Garry: Yeah, totally.

Brian: You took a big bet on me as well, and you were always so optimistic about everybody’s startups when they were super young and early. That’s my biggest memory of you actually, from Y Combinator. You were championing everybody when they had these beta, barely functional products. You were their first customer, and you had their T-shirt and a bit of motivation. An early customer who actually believes in the product can go a long way in the early days. So, thank you as well.

Garry: Dude, I appreciate that. That’s what gets me up in the morning — believing before people believe and then helping people believe themselves. Then, what you do after that is go and make everyone else believe. How crazy is that?

Brian: It can be a self-fulfilling prophecy, actually. Just having somebody else believe in you can make you believe in yourself, and there’s a lot to that. A lot of good ideas probably just die on the vine because people belittle them or they make fun of them or they assume it’ll never work. Having someone actually believe in your ideas is nontrivial.

Garry: The interesting and hard thing, though, is to believe something that can be true. How did you know that early?

How Brian discovered Bitcoin

Brian: About crypto and Bitcoin?

Garry: Yeah, about crypto. It sounds like there were things from places you worked before that started to inform that.

Brian: That’s true. You never really know for sure, right? I remember having moments of self-doubt where I thought maybe I was crazy because nobody else seemed to really get it if I explained it to them. I hear some people make business ideas today where they’re doing a consultant type of analysis or a financial analysis with the TAM and the competitor set. For me, it wasn’t like that at all. I basically just read a white paper and thought, “That’s cool.” I couldn’t get it out of my head. It wasn’t like I had a really sophisticated analysis of the market potential because there was no market for it. It was basically just me thinking this seems like a good idea and interesting, so maybe it’ll be interesting to other people.

I was trying to think about all the ways that it could die. For example, would it be regulated out of existence, or would there be some kind of cybersecurity flaw? The more I spoke to people about it, the more I couldn’t find any obvious reason that would happen. I didn’t know for sure that it would work, but I felt like there was a high enough chance that it was worth making a go of it.

And you’re right. My prior work experience definitely did give me a little bit of insight into that. Working on fraud prevention at Airbnb, I’d seen how difficult it was to move money around all of these countries, both collecting money from customers and paying it out to the people who were listing their homes on the site. It seemed like every single country in the world had their own proprietary system with a couple of entrenched oligopoly companies that were charging opaque fees. It felt like this was never going to get rid of all the bureaucracy and be innovated unless there was a new system. I always think about it in terms of the internet. It wouldn’t make sense for each country to have their own proprietary internet and you pay a fee when you want to load a webpage from another country or something. Let’s all just run the world on one standard. It felt so beautiful and elegant to me. So those were some things that gave me a little bit of a hint that maybe it would work. I knew enough computer science and enough economics — not that I was an expert in any of it — but after I read the paper a few times, I thought, “This might work.”

Garry: The crazy thing that I remember — I went into my email and dug up the first mention of your name in my inbox, and it was actually from Jason Tan at Sift Science — is that you were at Airbnb debugging with Jason at 2:00 a.m. I think that was the moment where I thought, “YC needs to fund Brian.” Anyone who loves the craft so much that you’re on at 2:00 a.m and your brain is engaged in something that is useful towards a goal really resonated with me as a founder because those were the times when I did my absolute best work. You’d wake up and think, “Whoa, I wrote all that code” or “I figured out this crazy thing.”

Brian: Yeah, I’ve noticed that actually a lot of founders have weird sleep schedules. There are some people who are working during the day, but there are so many emails and meetings that others can get distracted. They need a period of time, whether it’s super early or super late, where there’s nothing else scheduled, so they can just dig in.

I remember at that time, Sift Science was this great YC startup that was doing fraud prevention with machine learning. At Airbnb, I was trying to prevent some of the fraud that was happening, so I had looked at a couple of different solutions and what would it take to build our own. I thought, “These guys seem like they have a good product,” and we hooked it up and started to debug it. I was coding the whole thing and talking with them and going through their API docs, and it was a convenient time to get work done late at night because there were so many things happening during the day. There definitely were times in my life where I got so deep in the code — I wanted to get it working, and it was fun. There were other times where you got deep in the code, and nothing was working, and it was so incredibly frustrating. Two hours later, you see that one character was different. It was a labor of love, and I love building stuff like that. I definitely got super into it.

Garry: I love that story because a lot of people who are watching this video will see what you’ve created, and all desire is so mimetic, especially in terms of creating companies and becoming a founder. I feel like that’s the common thread: being a builder and being very focused on the idea. It’s not the technology on its own or the problem on its own. It’s probably not the sleep schedule but something about entering a flow state and then having this outcome.

Brian: Other people have written more eloquently about this than me, but there is something about this desire or this ability to look at the world and say, “This is how things are, but they could be different” — not just dismiss the idea out of hand because it’s so ridiculous. It’s being able to suspend disbelief for a minute and think, “Maybe this could work.” That’s one characteristic, and I think that builder characteristic is super important too.

Paul Graham wrote a lot about this. Remarkably determined and resilient people that just have setback after setback are just like, “I’m going to keep going because startups are certainly like that.” I think the builder piece is interesting too. The ability to code helps, but I don’t think it’s necessarily 100% required. The reason it helps is that when you’re on a shoestring budget and you’ve only raised a tiny amount of money and you only have one or two or three people, you could hire a firm to build your app. However, you almost never get the first version of the product right, so you’re overpaying a high-priced firm to build the app, and then you get the first version wrong, which everybody does. Then, you get the second version wrong, which almost everybody does. Then, you’re on the third version, and you’re out of money, and you’re dead.

So I guess it was almost a way to save money if I just coded it myself. I remember, for instance, I wasn’t very good at design, but I scraped together things as best I could, either using Twitter Bootstrap or even connecting with some friends at YC who created my logo for me as a favor. It was beg, borrow and steal, and try to get something out. If you iterate, iterate, iterate, you’ve got product market fit. Then, you can start to bring in enough people to get better and better stuff happening. In the early days, there were so few people who would even come join me that I had to make something happen with whatever I had.

Garry: Something I remember about your version of Coinbase in 2012 was that it was the perfect idiomatic expression of Bootstrap. It looked like Twitter, actually.

Brian: It wasn’t by any means the most beautiful site, but it was functional and it wasn’t sketchy. It wasn’t going to win any awards, but it wasn’t a huge detractor of the design either.

Garry: That seemed like a very important thing from the beginning — making a clean, well-lit place to actually buy this stuff and interact with it. It was interesting to see Coinbase evolve. A lot of people had the idea of doing something with Bitcoin, but you picked something so specific as a problem set, actually built it and started sending Bitcoin around with it.

How Coinbase grew from idea to reality

Brian: Like many people, I downloaded the Bitcoin Client, which was desktop software — it had a command line interface — and I started syncing my note on my Macbook. That was the first version of it that I tried. I knew that wasn’t going to scale to a large number of people who would want to do that. For one thing, it took a while to sync to the blockchain. It reminded me of email servers or Git servers because you could run your own email server, but nobody wants to; they want to run Gmail or something in the cloud that works on your phone. That’s the same thing with Git. Git is harder to use, but GitHub made it easier.

So those were my analogies, and I had this thought that someone is going to make the hosted wallet in the cloud for crypto. I was trying to talk myself out of that because I thought it was a really big responsibility. That’s going to be a big company that does that. You’re going to be storing people’s money. You’re going to have to have regulation. Early on, I made a Bitcoin Wallet that ran on an Android phone with a friend, and we tried to make it easier to use. We got a version of it out, but the day we launched it, I realized it was built wrong because it was trying to run the full Bitcoin node on the phone. Even if you were on Wi-Fi, it still took hours to sync in. I thought someone’s going to make the cloud version of this, but I couldn’t help myself — I just started tinkering and put it out there.

Garry: That’s an interesting thought pattern that seems to come up a lot: someone’s going to do X, but I’d rather do it.

Brian: Yeah, I think there’s a lot of truth to that. You notice there’s some big issue in the world or some schlep you had to go through  — I think people call it. Sometimes you think, “Man, that was such a pain. Someone’s going to have to do this.” Then, weeks go by, and you’re like, “Maybe I should do it. I just solved it for myself. Maybe others need the same thing solved.”

What the early days of Bitcoin were like

Garry: With crypto and Bitcoin in particular, you had to be a part of a fringe to accept it that early, but even within that fringe, there were ideas that were not useful or turned out to be wrong. One of which, I think, was being totally anonymous or pseudonymous at all times. You made a different decision for yourself very early. I’m curious because that was creating a fringe within a fringe that later became Coinbase.

Brian: Yeah, there were a lot of cypherpunks at those early Bitcoin meetups that I went to, and of course, Satoshi was anonymous. I think some people were of the opinion that we’re all going to be 100% anonymous here. I don’t have any issue with that, and I think the future is probably pseudo-anonymous like Reddit, where you have a username, can declare bankruptcy and create a new one. Your reputation goes away too. I think that’s probably going to be the future, but for starting a company, it’s different from what an individual might want. I knew that if this company was going to ask people to put in their credit card or their bank account or store their crypto, then it wouldn’t be okay for me to be anonymous.

I also went through the thought exercise which was, “If this thing is small, it can probably fly under the radar, and no one’s going to care, but if this gets to millions of people, there’s no way I can be anonymous.” I knew enough from reading about the history of PayPal and working at Airbnb doing this money movement stuff that there’s regulation involved. I was not willing to be on the run and never come back to the United States or something like that. I have friends here. I have family here, so I need to do this above board from Day One if I’m going to do it. So I felt like I wanted to create all of these credibility indicators about the company even when it was very tiny. Design is a little thing, same with how support works. I even remember having on the about page “press@” and “sales@” and random email addresses, and I was the only person in the company. I also put news articles and my actual name to show who created it. I also put it on my LinkedIn, so they can go read about my background and prove that I’m not trying to hide here. Getting into Y Combinator or getting certain investors were all little credibility indicators that started to accumulate to help us break through some of the noise.

Garry: You used a way of thinking that was if we get there, then we’re going to need to have this. You had to presuppose you could get to millions of users and then work backwards from there.

Brian: Totally.

Garry: That, I think, is actually relatively rare for a lot of people. They don’t focus on the outcome and don’t believe that they can get to the outcome. They focus on looking like a startup, and that’s just the beginning part of it. That’s putting the cart before the horse in a lot of ways.

Brian: Yeah. Well, there’s probably a way to take that too far, which is if you’re so worried about having this beautiful product, and everything’s perfect, and you never launch in the first place. Startups do have to “fake it till they make it.” You can’t be too precious about it, but if there’s an opportunity to create some kind of legitimacy along the way, certainly take it.

How Brian got into tech

Garry: Even going earlier, technology, software and computers have treated both of us incredibly well. What was your first computer and first interaction with tech?

Brian: Well, I feel very lucky on that front because I grew up in a home where my mother actually worked at IBM. She would take me to work sometimes, and I got to see computers. We got some of the early IBM 386 and 486 PCs in the home — this was growing up in the Bay Area in California. That was pre-internet, but I learned how to use DOS and play games on the computer and things like that at a pretty young age.

It wasn’t until the internet happened — I remember we first got dial-up internet and then DSL and everything — that I really started to get excited about it. I started to try to learn how to make HTML and web pages. I remember reading this book on how to learn Java in 21 days, and I was probably in eighth grade or something. I completely failed to understand it, but then I got HTML and thought, “HTML is simpler. I can learn that first.”

Garry: That happened exactly. That book and then HTML in that order actually.

Brian: Really?

Garry: Yeah, totally.

Brian: That’s so funny.

Garry: At Barnes & Noble. Maybe we ran into each other. Where was this in California?

Brian: In San Jose, California.

Garry: See, I was right down the street in Fremont.

Brian: There was something very special happening right there in the world at that time. I remember the very first job I had in high school. At the library, somebody put a piece of paper on the wall saying they wanted an intern to help develop their website. It was at this guy’s garage in San Jose — a cliché startup thing. That was my first job in high school, and I was making a little bit of money trying to build his website.

It was a remarkable moment in time historically. Looking back, we were in that primordial soup. I meet people now who had never even seen a computer when they were growing up. Maybe they saw a photo of one in a magazine or something, and then they knew they wanted to be in computers. They applied to college. They got in, and they got a computer science degree having never even seen a computer in real life, which is amazing. I feel very lucky to have had early access to it.

Garry: It’s the bicycle for the mind as they say.

Brian: Yeah, and there was something about it. Frankly, I was a very shy, awkward kid. I was not good with people, so I liked being on the computer by myself and reading books. It felt exciting to use it.

In high school, I didn’t really know what I wanted to do with my life. I was learning all kinds of things. I was taking some music lessons, and I was studying and doing all this stuff. I remember making this website with a friend of mine. He and I were trying to resell computer hardware. We had a PayPal account, and we made a simple website. I remember I went to sleep one night and woke up in the morning, and the hit counter said that 500 people had visited the website while we were sleeping. I thought it was the coolest thing ever because it was almost like a superpower. I was asleep, but something I created was helping serve the world for 500 people, which I could never do one-on-one. That gave me this rush, and I remember going into college and thinking, “I think I want to do something in business and computers or something.”

Democratizing information and money

Garry: The crazy thing now is — even thinking about this video, it’s going to be interesting because there might be hundreds of thousands of people who watch it — can you visualize right now being in a stadium with 200,000 people or something like that? It’s kind of insane.

Brian: It’s like the printing press of our day to give distribution. The internet changed all of our lives in terms that — just connecting people and having such an impact on anything you do and create.

Garry: Now, the internet is actually eating money, and you have created one of the keystones of that. Where do we go from there? How crazy is that, that we get to live in this time when money went from this roadblocked thing in a database someplace to something that is actually distributed and everywhere and not owned by a single entity?

Brian: I think people under-appreciate that — just how powerful it could be to take what the internet did for information, which is sort of democratizing it. Anybody could publish and do that for money because before the internet, there were all these gatekeepers. Publishing information was expensive. You needed to convince some studio executive to be on your TV show or get a gig at a radio station that broadcast to 1,000 miles. Newspapers were also these incredibly powerful institutions. Suddenly, anybody could do that.

People think money pretty much works today, and in some ways, it does. I can swipe my credit card for whatever and buy a coffee, but in other ways, we’re stuck in the era of radio, television and newspaper. If we make it so seamless where anybody around the world can participate in this global economy — and payments are instant, global, cheap and fast — I think so much is going to be unlocked. You can see little glimmers of it. In China, for instance, they have WeChat. WeChat payments are really seamless. You can do micropayments; they’re instant and pretty much free. We see all these little business models just flourishing on WeChat, and that’s just one country. It’s a large country, but it’s one country. So what would happen if we can connect everybody on the globe onto a payment system like that? We’re talking about orders of magnitude of potential unlocked.

The three ages of crypto

Garry: It feels like there are three different revolutions happening. If we time travel back to 2012 when we first worked together and you told me, “Hey, Bitcoin is going to be worth $50,000 or $60,000” — who knows where it’ll be when people are watching this video? It might be millions by then, who knows? —  I guess I just wouldn’t have believed you. That seems like the first revolution. Then, the second one is, you could argue, DeFi. The third — you’ve talked about this — is probably just decentralized all organizations, social media and maybe government. There’s all three of these things happening simultaneously.

Brian: Yeah, I think people still haven’t fully appreciated the different phases of crypto. They’re still thinking about it. Most people, if they’ve heard of crypto, are still thinking about Bitcoin as kind of like gold or digital gold, as some kind of investment you can make in some scarce thing. That’s really phase one: crypto as this new investment category.

Now, we’re seeing the second phase of crypto with DeFi, meaning crypto as a new financial system and all these pieces getting reinvented there with lending and insurance and exchanges being recreated in a decentralized way. Then, the third phase is probably crypto as this new internet application platform, which could be creating new identity systems. Social media could get reinvented there, like you were saying. Governance, whether that’s startups being recreated with their cap tables or even governments in some voting, could be reinvented in all kinds of ways.

Garry: And research, like with ResearchHub.

Brian: That’s a good example of a startup that is creating a token to try to bring together a community. What if the early people listing their homes on Airbnb could have gotten a percentage of ownership in the company or the early drivers on Uber? The standard model most Silicon Valley companies follow is they issue an option pool of 10% or something. They give those to the employees, and then they reissue the option pool as more employees come in. There’s probably going to be companies in the future where they issue an option pool for employees and then an option pool for the customer base because both are helping build this new thing. The board will re-up both of them whenever they run out, but you’ll end up with these new companies that are owned by the customers, the employees and the investors. It’ll be enabling new things like that.

Garry: Yeah, ResearchHub is fascinating. I think you wrote one of the blog posts that posited what ResearchHub could be.

Brian: I’ve been helping that project get off the ground during my free time, which is fun. It’s helping me better understand what a crypto startup of the future is really going to be like.

Garry: I think about that, and I think about what research organizations do. These are some of the biggest, most important, most prestigious institutions out there. Then, here’s basically a decentralized autonomous organization that is a website and also has a coin, but what’s more important than just the coin is how people interact, how things get funded and what gets funded. How does that actually push human knowledge forward?

How blockchain will change science and research

Brian: The financial services we talked about were mired in this bureaucracy and legacy systems, and I think scientific research is very similar in that regard. It’s very archaic the way a paper gets submitted to a journal. A journal takes a year to review it, and you have random people you don’t even know peer review it. Maybe it will get accepted to this journal. By the way, you have to pay the journal to submit your paper, and they charge the people reading it, and it’s basically just a PDF. Why isn’t scientific research happening more like open source software at that speed?

We saw a glimpse of that happening during COVID where because of the urgency, people started putting out their research on social media, and you’d get peer reviews the next day on Twitter or Medium or whatever. I think that’s the future. The currency of academia is citations, and I think it should actually be driven by something like Research Coin. If you’re creating breakthrough innovation or you’re helping the curator answer questions or one of the many ways that you can contribute to research, you should be accumulating a piece of ownership in something for that. The people who are making these breakthrough innovations like CRISPR should be billionaires in the way that startup founders are.

There’s this big gap between academic research and then a breakthrough idea will leap over into being commercialized once in a while. A lot of wealth generation happens in startups when things are commercialized. Frankly, Coinbase is an example of that. I read a research paper written by Satoshi Nakamoto. I thought, “This is a great idea. I could probably help commercialize it.” A lot of the wealth got generated over here, but the people writing the research papers aren’t generating the wealth, so their fake currency is these citations. There’s no reason that needs to be the case. Why isn’t a lot of research just published with some kind of license? For example, if you want to commercialize this, give us 1% of the revenue for the first 10 years. There should be just a button that says “license this.” Whereas today, you have to go to these tech transfer offices at universities, and there’s tons of lawyers and bespoke processes. Let anybody start doing the research. Let anybody license it, whoever wants to go and commercialize it. There needs to be a well-oiled path between those two worlds.

Garry: I’m struck by how different a decentralized autonomous organization run by software, designed by a deist creator — someone who thinks about the needs of all the stakeholders and people involved and then writes code — can put the right situation in place where magic can happen. In ResearchHub’s case, what we hope is that ResearchHub helps to create a whole wave of innovation and pushes forward human knowledge. I think it’ll do it. I think that we will have a generation of Einsteins, maybe in 10 or 20 years but probably sometime during our lifetimes. That can and will happen — maybe less and less in the auspices of a university and maybe more and more in something totally decentralized like this.

Brian: Totally, yeah. I think about YouTube. YouTube started off, and people would put up cat videos or copyrighted material or whatever, and it was not taken seriously. Now, YouTube is so big. It’s better to have 100,000 subscribers on YouTube than it is to have your own TV show on NBC or something. I don’t watch NBC, but I watch YouTube all the time.

I think, similarly, when you make these new systems, like a DAO or a ResearchHub or whatever, a lot of the early participants are a little quirky, a little weird, and they’re on the fringes. They’re probably not going to be, in ResearchHub’s case, some tenured professor with 30 years of experience. They might be a grad student who’s new in their career and decides they want to take a bet on it. By the way, a lot of the most active users on ResearchHub are grad students and PhDs that are earlier in their career, so we’re getting stuff that’s not being talked about sometimes in mainstream academia.

DAOs are an interesting thing. The governance component makes sense for a DAO. You can have money go in, and proposals go into the DAO. People can vote yes, and the money should go here and there. When I was helping to create ResearchHub, I was thinking, “How is the actual product going to get built?” The DAO is a more esoteric concept at the high level, but we have ResearchHub, a corporation which is actually called ResearchHub Technologies, and it is a technology vendor to the DAO. The DAO can have a vote and decide to allocate money to this vendor, which is a C corporation that builds the website and the app and pays the AWS bills,  but it is a technology vendor to the real company — the DAO. I don’t claim to have this all fully figured out, but I’m experimenting more in this area because I do think DAOs are going to be really important in the future. A lot of case law had to get figured out around a Delaware C Corp in that jurisdiction; the new jurisdiction is online. I think DAOs are like the Delaware C Corp of the cloud, which is where a lot of the future is being built. The DAO is almost like the real thing in my mind, and a Delaware C Corp just happens to be a technology vendor that’s providing services to it.

Garry: It reminds me of the time Yuri Milner came to speak at YC, and of course, people asked him why he invested in Facebook. He said that all of the world is reforming into a global brain.

Brian: Yeah. That’s kind of what the internet is.

Garry: Yeah. When you think about a DAO — I guess it’s like a synapse — it’s sort of a protocol for different cells in this global brain to work together, which is not centralized command and control.

Brian: In science fiction, there are these hive minds and stuff like that. I think the internet is turning the human race into more of a hive mind. I look at the number of WhatsApp or Signal groups that I’m in, and the number of links that get sent every day compared to three years ago (or even a year ago) is increasing and increasing. The information sharing amongst all these people in both little and large groups is just going up and up. It is almost like we’re creating a hybrid brain of some kind.

Reading about some biologists, they talk about how human beings are about 90% chimpanzee and 10% ant or honeybee, both of which are of a colony-type of mind. So we do have a hive mind aspect to us. For example, you see a bunch of people at a party dancing all in unison or people doing yoga where they’re all moving in unison —

Garry: It’s very natural. In our bodies, we can viscerally feel that experience.

Brian: Right. Even in companies, we do rituals like that sometimes where everybody will do a standup meeting, and we’re a little bit like an ant colony or honey bee nest or whatever, but we’re also different.

Decentralization in a centralized world

Garry: What do you think about this? When I think about decentralization, it does stand against tyranny in a lot of ways. When you think about tyranny, you think about command and control, and perhaps the worst things in human history come from hubris or too much control. One of the salient artifacts of a decentralized autonomous org is that you can fork it. If the creator of the DAO no longer serves the wishes of the stakeholders, it’s actually a lot easier. There are way fewer modes, but that might actually be really good for humanity.

Brian: Yeah, totally. Balaji Srinivasan has a great talk on this. It’s called Silicon Valley’s Ultimate Exit, and it’s about the concept of voice versus exit. If you can fork code or you can fork a DAO, that’s very powerful. I think about this for Coinbase too because in some ways, we’re building this very centralized command and control top-down thing in a decentralized world. I think there are some checks and balances against that, which is good. For instance, we’re using the open crypto protocols, so if people decide they’re unhappy with Coinbase someday, they can take all their crypto off Coinbase and put it somewhere else. There is a freedom for people to leave, which creates a great sense of accountability.

We’re also making it easy for people to store their own crypto. I actually think self custody of crypto is going to be a big deal in the future. It already is today, but it’s going to be even bigger in the future. We got started with people, custody and crypto on Coinbase. Still, the majority of people who come to us, including institutional customers, don’t want to store it on crypto; it’s too scary. However, with things like Coinbase Wallet, we’re making it easier and easier. There are over a million people that are using it to store their own crypto now. That’s even more decentralized, which I really like, so we’re going to keep investing in that. I’m thinking about that too. How do you create a company that stands the test of time? I’m not planning to go anywhere, but someday if I pass away or whatever, how do you make sure that the company has good values and makes good decisions? These are all really hard problems, which I don’t feel like I have all the answers to.

Garry: I think you have a new initiative. Is it called Coinbase 10 Percent?

Brian: Yeah.

Garry: So Coinbase 10 Percent is taking 10% of your time and resources and actually working on totally new things, perhaps like self-custody, that at the moment, seem maybe far off, but now are really pretty core to what the product will be and is.

Brian: I’m glad you brought that up. Coinbase Wallet, our self-custody wallet, and Coinbase Commerce and other things we’ve created have come out of that 10% bucket. When we were smaller as a company, a lot of those were really just coming from me when I got excited about something and wanted to make a team to go do it. What I realized is that, as we got bigger, it can actually be dangerous if I’m the gatekeeper or the bottleneck to creating these new kinds of applications. There’s actually even been instances where people inside Coinbase — like entrepreneurial people, we have a lot of entrepreneurial employees — have come and pitched me and been like, “Hey, I think we should build this. I think we should build that.” I’ve either said, “I don’t think it’s a good idea” or “We don’t have time right now.” They’ve left and those have become huge companies.

Garry: Totally.

Brian: We actually modeled it a little bit after Y Combinator and things like that. I don’t want to have a committee where 10 people have to say yes. That’s basically the opposite of innovation. A lot of innovative stuff is kind of contrarian.

Garry: You need 10 believers. There should be 10 people who would just go in and say, “This is going to be it.”

Brian: Yeah. Anyway, that’s what we’re doing with the 10%. We’re hosting our own little internal thing where if companies have one, two or three employees or whatever and you think this is a cool thing you want to spend your time on, you can pitch it to us. There are a couple of options, and we’re open to any option. If you want to stay at Coinbase and have your Coinbase salary and equity, then you can keep that, but we’ll own it;  it’s a Coinbase product. If you want to leave and work on it, Coinbase Ventures will invest, or other investors like Initialized can invest.

Garry: Totally. I’d love to.

Brian: There are even maybe hybrid versions of that too. Who knows? I’m open to various ideas. I’m trying to make sure we’re a company that has innovation happening all the time now, and it’s not stifled by us being too big or too bureaucratic or me having to think something is a good idea because at this point, I’m pretty sure I’ll miss out on the next big thing. There are so many cool things happening in crypto.

Garry: That’s awesome. I think that a lot of people who would be in your shoes and are in your shoes think they’ve made it. Instead, what I’m hearing is that this is actually a place to create the next thing.

I think that there’s a strong chance that the things that a billion people use in the future will all be crypto native. They might be DAOs. To me, it sounds like if you don’t have an idea already and are a great builder, Coinbase is actually the place to go. It’s going to point you in the direction of the people networks, the capital networks and give you the culture and know-how to actually build the future that both of us want.

Brian: Yeah, I hope that’s the case. I wanted to be an entrepreneur, and I had a tutoring company I had started — it didn’t really work, so I sold it — then I joined Airbnb. Airbnb was doing well. It was growing. I learned a lot. I met interesting people. That really helped me when I did have the right next idea, which was Coinbase. I’ve tried to create Coinbase in a similar way, which is that I think anybody who wants to create a company at some point or just wants to be a part of building something great inside Coinbase is welcome. I would check out coinbase.com/careers. It’s a great place to build not only a career in crypto but to learn the entrepreneurial skill set. You can get it funded internally or leave and do it, and we’re totally happy either way.

Garry: There’s this video that I’m very obsessed with and is a visual for what you just said. It starts with someone opening a bell pepper, taking one seed and putting it in the ground. Then, you see this time-lapse of a new plant just growing out of the ground and blossoming. Then, it bears more bell peppers. The cycle repeats, and this is sort of the thing that you’ve done and can keep doing actually.

Brian: Totally. When I look back on what I’m most grateful for in my life, it’s probably a couple of people — you being one of them — who took a bet on me, and that enabled me to do what I wanted. They saw the potential before I maybe even saw it in myself. I think that’s what we wanted to try to recreate. How can we do that for a whole bunch more people? A lot of employees at Coinbase have recently gotten some liquidity, and how do you put capital to good use?

My belief about the world is that prosperity comes from innovation. Innovation basically comes from there being more freedom in the world to go try new things with science and technology. If we have had some success here, how do we pay that forward? How do we show people that this is something anybody can do? If they’re determined enough and they want to go do some variation of it — it doesn’t have to be exactly how I did it; everyone’s going to have their own path — go try it. The first one probably won’t work. I tried creating probably 10 different startups in my life. Coinbase was by far the one that worked the best, but about eight of the nine other ones failed. One of them rounded to basically zero, but it didn’t completely fail. So try something. Try anything. It almost certainly won’t work, but you’ll learn something. Try something else. It almost certainly won’t work, but try it again. Go get a job somewhere and learn for a while, and do it on nights and weekends — whatever you need to do. If people keep doing that and actually believe it’s possible and have some role models, then I think a lot more startups will get created.

Garry: The crazy thing when I sit down and think about it is that when I wrote one of the first checks for your seed round, that was probably one of the first, either fifth or sixth, seed checks that I’d ever written.

Brian: That’s awesome. I actually didn’t know that. I thought you were some super experienced angel.

Garry: No, definitely not. I had paid my credit card off maybe the year before. I had to sell some secondary to Ashton Kutcher just so I had money to pay for my wedding. Then, I got hired at Y Combinator, and I had never done it before. Obviously, nine years later, I am only super emboldened by this experience because not only can any great builder build something that touches a billion people, but you can just keep doing it actually. We can actually keep making more and more of it happen; it’s the most growth mindset thing that exists on the planet.

The IPO is 1% of the Coinbase journey

Brian: Totally. That’s the really crazy thing. In some ways, we’ve just had this really great outcome with the direct listing and everything, but it’s not the end. We’re actually only 1% of the way into this thing because now — I think a lot of people misinterpret this actually. There’s a lot of negativity out there sometimes about billionaires and greed — once you have a place to live and food and stuff, there’s not that many things you can spend money on, so how are you going to allocate that capital? Hopefully, you’re going to allocate it towards helping build more things, which will create more jobs and more economic growth. Most of the people I know that have gotten some amount of liquidity, whether from Coinbase or other things in life, aren’t buying Lamborghinis or anything; they’re actually trying to think about how they can pay it forward, whether that’s philanthropy or starting new businesses. Both are really good ways to help the world.

Garry: Man, I don’t know where else we can end. I hope this is the first of many times, honestly. Every time we get to hang out, I learn so much, and it’s great just seeing how you’re continuing to grow Coinbase. If Coinbase is at 1% of the progress bar, what will happen in the next 10 years? It will be fantastic.

Brian: It’s just fun to build stuff. I’m sure not everything I try will go well. I hope if I have some certain amount of money put away that I just don’t need to worry about my kids going to college or whatever, then I can take big risks with the rest of it. We’ll see what happens. I think it’s just fun. We’ll try a lot of stuff, and hopefully, we can all be on “Team Human” to help things get better.

Garry: I like that. Brian, thanks for hanging out.

Brian: Thanks, Garry.

For more conversations like the above, along other leadership and startup topics, be sure to check out my YouTube channel and subscribe. To stay up-to-date with all the latest Coinbase news, follow them on Twitter.