Stanford Seminar - Building the Cryptocurrency Ecosystem - Balaji Srinivasan of Coinbase

Thank you, Tom, for that kind introduction. So today I'm going to talk about the blockchain and why it's actually interesting and why it means value is now programmable. All right, so why do we think blockchain is the most important technology of this decade? Tom gave my bio. I won't belabor it, but basically, currently CTO of Coinbase, prior to that, sold the company to Coinbase Earn, Prior to that, general partner at Andreessen Horowitz. You may be familiar with the firm. I also sold the company for about $375 million. It's a genomics company. Literally came out of Biox, just down the street over there with my brother and a bunch of our friends and basically just a Stanford guy through and through. So I kind of know your pain. So that's me. But enough about me, let's talk about Bitcoin and Blockchain. So I like this particular image because I think it motivates. This is a screen cap of Wireshark. And you can see the Bitcoin protocol literally as a protocol on screen. And what's amazing about this is you can see that payments have been turned into packets, right? This is a transmission, this is a bitcoin block. You can see the bytes over here. And what's remarkable about this is what's absent, right? There's no hsbc, there's no Wells Fargo. Right. There's no countries, there's no licenses or anything like that. What we have done is literally digitize the transmission of value. And if you think about how big a deal it was to digitize books and music and movies and newspapers and so on and so forth, the underlying technology behind Blockchain allows us to digitize. Now, cash, stocks, bonds, mortgages, loans, derivatives and all kinds of new things, financial instruments that you haven't even maybe thought of yet. So this is really like the motivating thing. Once we've got something into the realm of computer science, well, all you folks can go to town. So this is going to be a lot of fun over the next 10 years. All right, so I'm going to give basically 10 things about blockchain, 10 things about Coinbase, and go through a fact and certainly answer all your questions. So just about Coinbase or about Blockchain, I want to review Bitcoin, Blockchain and Ethereum and then talk about Coinbase a little bit and do a Q and A. So first, just for those folks, just to level set, because I know there's varying backgrounds, I want to give the very Very, very basics. And in particular, this slide over here is if you only remember one slide and you don't want to remember anything else, this is the motivation behind why even invent bitcoin in the first place, right? So with physical cash, if I have a dollar bill and I hand Tom that dollar bill, there's an implicit property of the dollar bill, which is when A hands that dollar bill to B, A no longer has it and B has it, right? I had it then, now Tom has it. And anybody who's observing that can see that that physical bill was handed over. If I try to naively transplant that to the digital realm, okay, if I take the Federal Reserve serial numbers on that dollar bill and I just go and email them to Tom, okay, well, I still have a copy of those serial numbers and I can go and email them to somebody else, okay? And that is the fundamental issue with a naive representation of digital cash. It is the double spend problem. I could have spent those serial numbers with Tom and then with somebody else and somebody else. And so simply using serial numbers alone, that's not sufficient for basically scarcity in the digital realm. And so the way that we. Until the invention of bitcoin, the way that we represented digital cash was with a bank, right? Like a centralized actor that we trusted in the middle. And this actor over here is where the scarcity enters the system, right? When A sends money to B digitally, C is trusted to debit A and to credit B, right? So this is where the scarcity enters the equation. The thing about this is you're putting a lot of trust in C, right? They can debit and credit anybody. They can choose not to debit or credit. They can not allow a transaction to go through. In extremists, as in 2008, they could print billions of dollars for themselves. So this is something which is, you know, inelegant from a computer science perspective, to have a trusted central node, if you can. If you can avoid it. And so what Satoshi Nakamoto did is he came up with D decentralized digital cash. And so essentially, this central actor, this bank, was replaced with a network of miners. And any one of those miners could approve that transaction, that debit or credit between A and B. And so since any one of them could approve it, one of them disapproving it, all they would just do is be giving up some bitcoin that they would have mined otherwise. It's basically a way to combine transaction approval and. And senior ridge currency printing in the same unit. So the details of this aren't super important technically, at least for a talk like this, but the concept is important, which is we took physical cash, we tried to naively turn it digital cash, that didn't work. So we had these centralized actors, and Bitcoin dispenses with those centralized actors by having anybody in theory who can connect to the Internet with sufficient computational power can now approve transactions and push them through. Okay, so the cool thing about Bitcoin is once people understood this, and once people understood what a breakthrough it was, they said, oh well, you know, this data structure that these miners are updating, that blockchain, that fundamental innovation behind Bitcoin that's essentially a tamper resistant database for storing arbitrary things of value. You can use it for storing cash and you can use it for storing stocks and bonds and so on and so forth. But Bitcoin itself wasn't that easy to work with. It wasn't built for programming, or it was actually. But Satoshi turned off a lot of the more sophisticated opcodes early on just to keep it simple and attack proof. And so what happened was more recently, a new blockchain was launched in 2014 called Ethereum, about five years after the launch of Bitcoin. And to understand the improvement that Ethereum gives, here's an example of a script in Bitcoin which is kind of almost like assembly language, like a stack based language. And here's what Ethereum solidity looks like, which a little more like English or JavaScript or what have you. And this programmability led to a huge new surge in crypto activity. And in 2017 it led to this massive ICO boom. And this went up from there were tens of billions of dollars to many hundreds of billions of dollars, and then down to $100 billion. But the thing about this is, even though the ICOs came up and they went down, we think of it today, it's the winter for price, but the summer for innovation. A lot of the stuff that got started back then is starting to work. For example, you've got decentralized prediction markets and decentralized loans and decentralized derivatives. That stuff is all starting to work. The winter for price, the summer for innovation. One thing that ICOs themselves, by the way, is they represent, in my view, the simultaneous disruption of venture capital, swift crowdfunding and cap tables. And I'll come back to that point, but essentially you can invest in something internationally with a group at the same time and have all of the property rights recorded in the blockchain. So anybody can see that's like such a big deal and so many innovations at once that people can sometimes lose a farce for the trees because they say, oh, you know, what are you funding? You're funding all this crazy stuff. But if you just look at it as an innovation in crowdfunding or in VC or Swift or cap table, it's actually a pretty big deal. Okay, so this brings us to the present day. So you know, it was bitcoin, then the blockchain, then Ethereum, ico, boom. And now the present day, the crypto winter. Winter for price, summer for innovation. Okay, so the way to think about the blockchain as a technology is it improves fundamental financial primitives. And one way of thinking about this is a blockchain. As I mentioned, it's a database for storing things of value. And it started with cryptocurrency, which is like digital gold. And more recently we've had these so called ERC 20s, these tokens which represent like equities. But you can also use it for other stuff, right? You can use it for non fungible assets like so called. It's very technical sounding but like in game items like swords and potions, baseball cards, collectibles or even eventually piece of real estate, right? You can use it for representing identity and frankly you can use it for representing almost anything scarce. Anything scarce you can think of you can represent in a blockchain. And you might say, well, why use a blockchain for things that are scarce? Well, once you've got a database that represents all the money in the world, the first thing someone's going to try to do is write themselves a million dollars, right? And so you have to actually say, well first I want that to be tamper resistant. Second, I don't want one party to be able to write themselves a million dollars. I want it to be decentralized, right? So that you know, the power is split up. No one person can write themselves a million dollars. They might need more than a million dollars to write themselves a million dollars. And once you start thinking of those kinds of constra database that represents things of value, you start to be led to the kinds of properties that blockchains have. And once of course, you can solve the problem of preventing someone from buying themselves a million dollars, you can stop them from buying themselves a million potions or a million acres or a million of anything else. So as I mentioned before, bitcoin turned all this stuff into packets. They're literally protocols, Bitcoin, ethereum, these things in that you can open up wireshark and see the raw packets that update the underlying blockchains. Just, it's actually completely amazing when you think about that. Anybody who's on the Internet can send and receive cryptocurrency back and forth. You know, just send packets back and forth. One consequence of this is, you know, there's, there's another, you know, people have talked about the unbanked and it's, you know, it's an important problem that I think crypto can solve in the long term. There's another group of unbanked which are, which are machines. And you know, we gave machines telephone numbers in the form of IP addresses. It used to be that you had to have a human who is associated with the telephone number. When you give a machine an IP address address, a machine can then autonomously communicate back and forth. Right? And you do that all the time. Whenever you've written a bot that goes and interacts with Twitter or Scraper, like that spider's web, and connects to other computers autonomously, autonomous machine communication was a really big deal. And we believe that autonomous machine transaction where machines can hold and send money on your behalf will likewise be a very, very big deal. Here is just in a simple GitHub gist, that string over there represents money. So if you can now transmit a string, you can transmit an arbitrary amount of value, which is pretty awesome. So a lot of the stuff in the late 90s that people talked about, when they talk about intelligent agents or bidding entities and so on, that's all actually starting to happen. And the primary context in which it's happening are things like exchanges, especially decentralized exchanges. You can go and write a program on your laptop that will literally make you money as you sleep. By arbitraging these decentralized exchanges, I think over time, in like the 10, 20, 30 year time frame, every price on the Internet, every Amazon.com product will have a cryptocurrency interface. So if you have some arbitrage strategy for like wool versus fleece sweaters or whatever, you can actually arbitrage anything because everything will be priced in crypto. That's, I think, where arbitrage stuff goes in the medium to long term. Okay, fourth point. You know, people talk a lot about the potential of blockchain, and I do think it has a lot of potential. But, you know, it's actually something where, you know, in some ways it's overhyped and in some ways it's actually underhyped. Right. That is to say there are fundamental financial primitives that the blockchain's already delivered, like a 10x improvement on, and I mean a quantifiable 10x. So, you know, the easy one, the obvious one is gold, right? So bitcoin is infinitely lighter than gold, okay? Like, it's faster to transport internationally. It's cheaper to custody. And to give a concrete example of this, you know, when the German Bundesbank tried to repatriate a bunch of gold from New York, it took them years, it cost them millions of dollars to bring back billions of dollars in gold. And while that might be an infrequent operation, it's an extremely important, valuable operation that frankly, underpins, like, a big chunk of the world economy. And if you've improved on that, that's a big deal, right? Gold is $2 trillion industry. If that was the only thing that blockchain ended up doing, that'd be a pretty big deal. But it's more than that. Tom's going to buy some bitcoin now. So it's more than that. With Ethereum and other kinds of cryptocurrencies, that represents a 10x improvement over Swift. Right? Swift. If anybody's sent a wire internationally, that takes days to clear, typically. Whereas with Ethereum, I can be on the phone with somebody and just like, I can wait for that email to hit, I can wait to just see it in the blockchain, and I can say, oh, that transaction is done. Boom. I can now get the deal closed. And so the speed difference here is the difference between a physical mail, like sending a letter, which takes two to three days, versus an email, which takes minutes. And think about how big a difference that made when you went from physical mail to email. Just the metabolism of business changed. When you go from two to three days for international commerce to minutes, the metabolism of international business changes. Third point, crowdfunding. So in 2015, if you go to Wikipedia and you look at the list of the largest crowdfunders, you will see that it's like maybe 10 million, $15 million. By 2017, 2018, a large crowdfunder was $4 billion. So that's like a 400x improvement in about two to three years. And that's more than 10x. That's like 400x. Now you might argue, well, what are those things funding? They're all funding ICOs and crypto stuff. It's all this ouroboros, but it's actually something where you could fund a building, right? You could fund pretty much anything now that folks are customized to using crypto to do this kind of Thing, It's a powerful new innovation in crowdfunding. And I could go on and on and I probably write an article on this, but just a few more examples. So, like Delaware incorporation in many studies, the World bank has a study called Doing Business. And in this study they actually say, hey, you know, how long does it take you to incorporate in your country? Does it take you weeks or months? Or does it take you days? And it takes you days. Well, you're a great country for business, great country for startups. Well, Ethereum's shattered that, destroyed that. Because it's not days anymore, it's minutes. You can literally get a smart contract into the blockchain in minutes, right? And that's amazing. That works internationally. You're not incorporating in Delaware, you're incorporating on the blockchain. And it might seem like an apposite comparison today, but lots and lots of stuff, all these VC terms, all of these bits of Delaware law are currently, as we speak, being encoded into smart contracts so that incorporation on the blockchain will become probably the default in like, I don't know, 10 or 15 years. And, you know, there's more like the time to set up bank accounts, the time to audit. You know, in 2013, people talked about, you know, bitcoin as a scam. But today, the big four, when they audit a blockchain firm, they use the blockchain as the source of ground truth that they check all of your transactions against. Okay? So from a scam to the gold standard in like five years. And so the point here is we're 10Xing all of these financial primitives, like basic, basic fundamental things and how we move money, how we think about money. And that's going to have consequences, good consequences, hopefully. Relatedly, not only have blockchains 10x lots of things, they've already created many billion dollar entities, right? So, you know, three that have risen in particular, the folks who founded new coins, the miners who run these gigantic server farms and the exchanges. And each of these three groups has created and captured billions of dollars in total revenue. It's comparable sort of to the 90s Internet when it was kind of the infrastructure stage and search engines and Akamai and AOL and so on were getting built out at that time. 6th. Now people talk about blockchains as being trustless. And bitcoin is about no trust. I think a much better way of thinking about it is that they give you a choice of who to trust. Okay, so before you just had to trust a bank, right? You had to store Your money in a bank. You can't just go and withdraw like a million dollars from the bank. Or you could, but you'd be on a lot of lists. And today what you can do is you can certainly still store it in a bank should you so desire. You can store it with a company like Coinbase or, and this is very new, you can store it on your computer, you can store it on a USB key, you can store it in the cloud, you can basically self custody an arbitrary amount of money. And that's a major innovation because it's like you can't carry a bunch of gold bricks around with you. You might be able to carry around suitcases full of cash, but it's a little bit conspicuous. And now you can carry an arbitrary amount of money. So that's a very new thing. You don't have to trust the bank if you don't want to. And places like Venezuela, this is actually truly a killer app number seven, I hinted at this before, but just to drill into this point, one can argue that the single most important data structure in Silicon Valley is the cap table, namely, the very simple table that shows which people or entities own what shares of the company, your common stock, your series, seed, A, B, et cetera. That data structure, that little table that Excel spreadsheet maintain, thing is what has aligned people from every country around the globe to build Silicon Valley. And the scale at which it's aligned them is tens, hundreds, sometimes thousands of people pre ipo. And then you build this huge company and then everybody hopefully does well together. But the state of the art for managing the most important data structure in Silicon Valley is an Excel spreadsheet or a Google sheet, which is kind of amazing. And now what we've got with the blockchain is you can 10 or 100 or 1,000, that number of people, something like BAT, for example, is 54,000 addresses that hold BAT. And you can see exactly who owns what, when. And this is fully programmatic. So one way I think about it is if aligning a few hundred people at a time built Silicon Valley, what does aligning a few hundred thousand people at a time, or a few million people at a time, what can that build? So we can now align human beings and incentivize them like never before. That's a really, really big deal. So number eight, one way I think about this is some of you here will be like, oh, I got to see it for. Too late. I miss mobile. It's over. No, it's not over. Just beginning. So one thing we think of as blockchain first, or crypto first is like the new mobile first. It provides a weapon with which to go and attack and open many industries or areas that were either close to competition or seemingly impregnable or not even thought of as the kind of thing you could innovate on. So I already mentioned international wires and precious metals and messaging. I'll talk about the social bid in a little bit more. But there's also other areas that you normally wouldn't think of as things you could even innovate on. For example, like DNS, right? With crypto you can innovate on DNS. You can innovate at the protocol level with things like Blockstack or ENS or Handshake where you can register domain names. These are also scarce entities. You can register them in a blockchain and you can route, you know, things through them. You can have incentivized versions of BitTorrent, incentivized versions of SMTP. All these protocols, you know, that you previously thought of as kind of, you know, non monetizable are now potentially improvable and monetizable. And there's also, you know, commons like Wikipedia, open source projects where you can now take a fresh gloss in them and think, okay, how could I maybe tokenize this? How could I make this even, even better? And so, you know, this type of stuff, you know, there's, there's so many different crypto first companies that are kind of swarming and attacking these different verticals that I couldn't even enumerate all of them. But this gives you some of the flavor of them. Coinbase Adventures alone has made like 40 investments over the course of last year. Ninth point. So blockchains break network effects, okay, and they give a completely new tool a can opener for going after scaled social networks and two sided marketplaces. One way of thinking about this is if you think about Metcalfe's law, you think about network utility. Typically, the larger the network, the higher the value to the user. So If Facebook gets 2 billion people, supposedly runaway advantage, you can't display something. But there's something really interesting about crypto, which is that if you have a token whose value is associated with the startup, the token's upside is maximal when the size of the network is smallest. So for example, an early investor in Snapchat had more upside than somebody who invested in the Series D round where it already had 10 million plus users. So upside is itself valuable. And now you've got a second term in the equation. Which can balance out that first term some kind of hypothetical future value that for the first time you can programmatically issue to your users for sign up and other kinds of actions that they take within your app. Right. I mean, imagine if Facebook like had something where the folks who signed up for it, the first hundred thousand or two hundred thousand or a million users or end users, received some slice of Facebook's equity for doing that. Well, that would be a very big deal. You'd actually start making money by signing up for all these dumb social apps. So that's actually something which I think is a very powerful tool for going after scaled social networks, two sided marketplaces and things like that split some of the upside with the user. Tenth and relatedly, blockchains, I believe are going to transform social networks. So for the last 10 years, people have been, you know, liking and poking and messaging each other, you know, wasting time. The next 10 years, you know, well, we've done making friends, it's time to start making money, right? So basically those edges we think are going to now have money on them, cryptocurrency, and you're going to have new kinds of social networks. What are we going to call them? Maybe digital economies or crypto economies. The fundamental difference being that your time on them is not wasted. Your time on them is actually like a form of work where you're actually remunerated for it. Right. And you know, some of these, you know, maybe existing social networks might be able to retrofit, but I think you're going to have to construct wholly new social networks that are based on these kinds of behaviors from the beginning, where earning is part of the equation. So, you know, finally, just on the topic of blockchain, blockchain is like a partial move away from the cloud and towards privacy. Okay. An amazing stat from earlier this year is that Ledger sold its millionth hardware wallet. Right now I think of a hardware wallet, actually this is earlier last year, early 2018. I think of a hardware wallet as being a pretty arcane kind of thing, right? But when a million of them have been sold, you're probably eventually going to have a billion sold in some form or the other. And you know, one possibility is hardware wallets like a phone charger are bundled with a phone. So you get like three devices, you know, the phone, the phone charger and the hardware wallet rather than two. There's lots of different distribution models you could imagine. But the fundamental concept is I think hardware wallets are going to become very popular. And that's a move away from the cloud and towards privacy. Because the whole concept of a hardware wallet is you're not keeping your keys in the cloud, you're keeping it locally. One possible future model is something where your data is on the cloud, but it's encrypted with the private keys that you're using in your hardware wallet, and so nobody can read it other than you, and it's only decrypted and re encrypted just in time. I think things like homomorphic encryption and so on will make that possible. So let me Briefly just give 10 things about Coinbase and I'll take some questions taking a long time, I think. So you can think of Coinbase itself as like the interface between fiat and crypto. And, you know, we have both the retail and like a professional interface. And on the fiat side, we're building like a Wall Street 2.0, right, like custody and indexes and commerce. And on the crypto side, we're doing like blockchain first apps like Earn and Coinbase Wallet. And we work closely with regulators. You know, our goal is to remain the safe, legal and trusted part of crypto. And you get all these licenses and so on, money transfer, license, all that stuff. And our mission is basically to create a more open financial system for the world. And what that means is kind of two different groups of users that would benefit from that. The first group are developers, investors, power users of money who are trying to move money around in different kinds of ways, do fancy things with money. And the second group are the marginalized, the folks who are unbanked, the folks who do not have bank accounts. Both of these folks, in different ways, are pushing on the financial system to make it more open. And those are our folks. What does that mean, by the way? What does a more open financial system mean? It does mean upside. It sometimes means downside. Upside is that you can raise $35 million in 30 seconds or disrupt remittances. The downside is there's new kinds of hacks and new kinds of scams and so on. But I do believe in the medium term we'll mitigate the downsides of things like multisig or what have you. You know, Coinbase. We're kind of thinking of ourselves as bridging from the investment phase to the utility phase, which is to say that crypto has sort of the opposite emergence model of social media. Social media piled up. The users first and then had to prove out monetization. And crypto proved out monetization first in almost the most literal sense and now needs to prove out utility. Right. But we're seeing that happen. You know, I can go to lots of different things with, with earn, with decentralized finance, and we think that's a big part of the next five years or so. Coinbase is actually named after where the blockchain starts. So the so called Coinbase transaction is where the new crypto is created in Bitcoin. And that's why we think it's also where the blockchain starts for users. Just like Windows, you started using it with the start button. Or Google Search is where you started browsing the Internet, log on with Facebook, how you started using social media. Coinbase is where you start using crypto. So we think of ourselves as kind of like this sort of on ramp, this start button for crypto. We're making headlines. You can read more about us online, and that's about it.