a16z Podcast - a16z Podcast: Ethereum, App Coins, and Beyond

Episode Date: August 28, 2016

Bitcoin quickly made its way from a whitepaper to a production network, which is pretty amazing when you think about it. But its scripting/ programming language was initially, intentionally, limited for a few reasons, which meant that building new apps on bitcoin wasn't always easy. Enter ethereum in 2014 -- a public blockchain platform that moved away from the "Swiss-army knife" approach to a more general protocol approach. This would in turn allow endless (and entirely new) use cases to be built on top of the blockchain, whether smart contracts or "app coins" that allow decentralized crowdfunding and decentralized business models. The results, at first glance, may seem just like a new way of financing a company. But it actually goes much deeper than that: They're really software protocols that are almost replacing centralized companies or what those companies would do. The possibilities are endless... In this episode of the a16z Podcast, Ethereum inventor and co-creator Vitalik Buterin joins Fred Ehrsam, co-founder of Coinbase (an a16z portfolio company) in conversation with Chris Dixon. The conversation covers everything from the politics of open source (and value of network effects even when those networks split) to the challenges of mainstreaming and scaling tech. And what happens next?

Hi, this is Chris Dixon. This is the A16Z podcast. Today we're going to talk about Ethereum, and we have the co-founder of Ethereum, Vitalik Boutaren, and we also have Fred Ersum, who's the co-founder of Coinbase. Thanks guys for being here. I'm a pleasure. So, Vitalik, could you tell us what Ethereum is and why it's interesting? Sure. So I'll probably start off by going back a couple of years to when the project first started. So at the time, at the end of 2013, people were starting to get really excited about Bitcoin. And people were also, for the first time, started to get excited about the Bitcoin's underlying technology.

So about blockchains and decentralization. And people are starting to basically think about what other use cases can you come up with for these kinds of systems. So if you've been in the space for a while, you might have heard of Namecoin, which is a blockchain based and decentralized DNS that a few people came up with back in 2010. There were also some projects at the time that were called Covered Coins, which was trying to use the Bitcoin blockchain in order to store kind of more generic and different kinds of digital assets. There was also a project based out of Israel called MasterCoin. It was trying to be even more advanced. It was trying to support certain kinds of financial contracts, name registration, betting, and lots of other different use cases. And I looked at this project, and at first, I wasn't really too interested in any.

of them. Like I actually looked at the master coin crowd sale and it seemed like they were asking to basically throw your Bitcoin into some address and then get some other coin and then how are you supposed to even believe that this other coin is going to have any value and so forth. But I got into that particular community. I saw that there was like genuine development effort going into it. I saw that there were people that were genuinely excited and interested. And I even started working for that team for a few weeks. I then discovered that both that project and also pretty much every other projects that existed at the time that was trying to do what you might call blockchain 2.0 crypto 2.0, cryptocurrency 2.0, whatever you call it. And the weakness is

that they were trying to create either single purpose protocols or what I call Swiss Army Knife protocols. So by Swiss Army Knife Protocol, what I mean is basically, you know, five guys come together, sit in a room and they brainstorm for an hour and they come up with, let's say, 35 different use cases for blockchain technology. And then they say, okay, we now know the 35 use cases of blockchain technology. Let us make a protocol where we have 35 different transaction types for each type of transaction where maybe the first bite of the transaction represents the ID from like 0 to 34 represents sort of which application or which use case it's for. And then you have some special rules for handling each individual use case. So this was the route that most

of these projects were going towards. And I, after spending some time, thinking about this, started moving pretty heavily in the direction of trying to make the protocol more and more generic. So trying to see, at first, if we could take some of these 35 use cases and like expand them so that one particular type of use case could, you know, maybe with some kind of more generic mathematical formulaizing or whatever, encompass more and more different areas and potentially even applications that people didn't think of before. And eventually I realized that the thing that made the most sense is to basically just go on. all the way and create a blockchain where you have this principle that essentially there are no

features. So what there are, there are no features means in Ethereum land is essentially that there is no sort of specific transaction type for, let's say, registering an asset, even like issuing an asset, registering a domain, entering into a financial contract or whatever. All there is is this sort of very generic architecture that's just based on code. And whatever application you want to build, you can basically just write the code for it that describes what the state of the application is, what the state transition rules are, what kinds of transactions can go into it, what the rules are, what modifications, different transactions make to the state and so forth. And at the end of that, what you have is the system where no matter

what someone wants to do, they can just sort of write the logic up and do it. So that was like the basic principle. So that was the idea that Ethereum started off. back at the end of 2013. Projects evolved quite a bit. So the first version of Ethereum that I was thinking of building is just a very simple kind of meta-protocol. So something on top of the prime coin blockchain that would just make prime coin transaction data have sort of different meanings that would be understood by Ethereum-enabled clients.

But then as the project grew in developer interest grew, we quickly went on to creating our own blockchain and then came up with a fairly advanced concept of gas accounting that we have right now. And it kind of kept going from there. Can you give some examples of what kind of applications someone might, you know, that would motivate you a reason why you'd want to use something like Ethereum? What were the kind of things you want people to be able to build on top of it? Back when I started the project, I was thinking of fairly simple use cases.

And as an example, I'll actually give one that the foundation is using right now. So the idea is simple. Let's say you have a multi-sig wallet. And with Bitcoin, one thing that you could do is you could have a multi-sig wallet with the rules that say you need four out of seven keys in order to move any of the coins. With Ethereum, though, you could actually do something a bit more complex, and you could say, well, any one of the keys is allowed to move a maximum of 1,000 ether a day out of the wallet.

But if you want to go above 1,000 ether a day, then you have to go up to 4 out of 7. So if you want to have more complex rules like that, I think it's pretty obvious to why you might want to have those kinds of rules, because it's actually very similar to how your debit card works, where if you send small transactions, they generally go through with low security. If you try and set something really big, then the bank might complain about it ask for a phone call for some extra verification and so forth. But with Bitcoin itself, you can't do that because the protocol has a kind of inherent statelessness to it. Like, you can't remember the notion of, well, this guy already withdrew 500E3, six hours ago and another 500Ether or two hours

ago, so he can't withdraw anything else right now. Whereas with Ethereum, you can actually have programs that remember some notion of long-term state. So that's one example, just simple withdrawal limits. Another example would be any kind of financial contracts. I was also later on thinking more and more about either non-currency use cases or even use cases that I call sort of semi-financial. So non-currency use cases, I would take a domain registration as one example. So the sort of things that name coin is doing, except with Ethereum, the theory is that people could make 100 different name coins that each one of them could have different rules, and we could experiment a lot and see which one of the rules makes the most sense, and maybe different

rules even make sense in different contexts. You could have various different systems involving web IoT devices talking to each other. The semi-financial use cases all have to do with combining a payment element with some kind of like computing or information storage or manipulation element. So one of the first examples I came up with is that you could have a contract which automatically incentivizes people to store a file for some particular time. So let's say if I have an encrypted copy of my hard drive and I want to incentivize people to just store a backup of it just in case, then I could come up with this scheme where I sort

of hash the file into a Merkel tree, stick the Merkle word into a contract, and then the contract automatically enforces some rules, and those rules might basically say, you know, if you can prove to this contract every single day that you're still storing the file, and there's various different kinds of sort of cryptographic compact proofs that you can do, then the contract automatically pays you five cents a day or something similar. That's another example of something that you can do, and there's actually use cases for computation. So you can think of this in the context of different kinds of like verifiable cloud, including file storage, file retrieval.

And a lot of these use cases are actually even being kind of developed right now. So if you look at projects like Swarm, Golem, Ethereum computation market, a true bit, and a few others, they're all based off of this principle. There's also more kind of complex stuff that has to do with things like DAOs. So much more complex kinds of contracts where you incorporate kind of very complex. rules about that are in control of digital assets in various ways. And so that's an area where I still think that, you know, for obvious reasons, it's going to be a couple of years before we really learn how to do it well and safely, but it's still one of those that I'm

really interested and excited about. Yeah, Vitalik, I think the backstory you tell is really important about how this kind of organically grew out of Bitcoin. Like, in some ways, what you said could be simplified to. I tried building advanced applications on top of Bitcoin that didn't go so well. And then I decided I needed a more generalized system that would let me build generalized applications. So I created Ethereum. Is that like, is that kind of the gist of it? I'd say so. Bitcoin has, of course, a scripting language. It's just, it's deliberately very, very limited, right? That's right. Yeah. So when Bitcoin first came out, it was kind of a miracle that Bitcoin even got off the ground in the first place and still stands here today

with $10 billion of value tied up in it, right? It was really a miracle of a combination of cryptography and economics that the system actually worked almost right out of the gate. It was born into existence in Satoshi's white paper and actually worked right away as a production network. So I think it's important to understand that context. When Satoshi first outlined Bitcoin, he actually thought about including more in the scripting language, which was kind of a fancy way of saying the programming language in Bitcoin. But because he wanted to make sure the thing just worked at first, he intentionally made that programming language extremely restrictive.

I think the best analogy to use for this is if you try to program something into the Bitcoin blockchain, it's kind of like programming using a TI83. calculator or some very, very restrictive language that's very primitive. You know, Mike Kern built one of the more sophisticated applications on Bitcoin, I thought, Lighthouse, I think it was called. And it was sort of a crowdfunding app. And he told me it took him, you know, he's a super, super advanced, you know, Google engineer.

It took him like eight months. And I think, Vitalik, tell me if I'm wrong, it's one of your like 30. It's one of those examples in the back of a T-shirt, yeah. Yeah, exactly. It's like a 30-line script in Ethereum. It's just, Bitcoin just wasn't, it was actually sort of deliberately designed. not to have, right? I mean, it's deliberately kind of limited because for security reasons and all the reasons

you have, right? And I totally like don't blame Satoshi for doing that because Bitcoin was testing so many different completely new and concepts at the same time that nobody even knew if they were going to work. If there was the concept of proof of work, then there was the kind of social and economic concept of using economic incentives in order to secure a ledger, all of these ideas around decentralization, the way that the Bitcoin community would be able to kind of manage itself and all these different things that I think Satoshi probably definitely would have had

way too much on his plate if, you know, if we're getting hacks and collapsing in 2010 all at the same time as everything else was happening. It's already one of the great inventions of modern applied computer science. It was pretty impressive. So Fred, you were saying more, though, about the backstory. Yeah, exactly. So, I mean, if you're looking at this from like a historical perspective, I would I would phrase it as Ethereum as kind of standing on the shoulders of giants here, namely Satoshi's white paper. So if you've been sitting in the Bitcoin community for quite a while like Vitalik has or like I have, you know, maybe since 2010 or 2011, Bitcoin kind of became a big thing in 2012, let's say that's when Coinbase was founded. And it's been. getting increasing notoriety. But at the same time, if you're embedded in the community, you're kind of, or an outside observer, you've been sitting here for the last three, four years

and saying to yourself, okay, well, where are all the apps? And when you ask that question, then you try to sort of do a root cause analysis of it where you ask, okay, well, why aren't there that many apps getting built? And I think the answer you arrive at is the scripting language in Bitcoin makes it really, really challenging to build an app really, really easily. And in my mind, that's kind of where Ethereum comes out. Now, a counterargument, right, would be that this is the sort of layer two idea that in Bitcoin, you could build these applications at a higher level in the stack and these things like Lightning Network, or I guess, or whatever, right?

I mean, isn't there the counter argument architecturally is you could replicate a lot of what Ethereum's doing with another layer on top of Bitcoin? Yeah, that's the argument, although I would say, yeah, the answer there. is kind of sometimes yes, sometimes no. So one example of this is that there was a few papers came out of Cornell over the last few months. One of them was they published this at their financial cryptography 2016 workshop that they were trying to create a provably fair lottery on top of Bitcoin.

And it turns out that if you have N players in the lottery, then you basically have to have O of N cubed collateral in order for the thing to actually be safe. So you basically have to have every person submit enough collateral in order to cover everyone winning simultaneously. And then you actually have to have every person submit enough collateral to cover everyone simultaneously covering everyone. So the way the structure works just ended up being kind of incredibly complex and basically like worthy of a research paper. Whereas in Ethereum, the same thing could be done as they discovered was just O of it. there's a very simple approach for doing it with O of N squared collateral, and there's an even simpler, or a slightly more complex approach rather for doing with just O of two times N collateral.

And the reason why you can go from here to there is because there's this sort of fundamental difference between the kinds of state that Bitcoin supports and the kinds of state that Ethereum supports, and it's this kind of bridge that's sort of implicitly the our hearts to cross. the Cornell researchers actually did end up coming up with modifications to Bitcoin. One of them was this concept of covenants that would basically let you send to make addresses where if you send through those addresses or any transactions going from those addresses will themselves have to have particular restrictions on them. But the idea just ended up being kind of very complex to implement.

So I think in some cases there's this sort of fundamental notion of, like, like a different kind of statefulness that's really hard to jump over. But, you know, in other cases, you can kind of fudge your way over it. And I suppose the Lightning Network is definitely one of those examples. And one other thing I observe is just looking at, like, you know, the Reddit forums, our Ethereum and our Bitcoin, that the, our Ethereum community seems, I would describe it as much more of a computer science developer community, whereas the Bitcoin community, you know, for whatever reason, historical reasons, seems much more kind of interested in politics.

you know, I don't know, Snowden, WikiLeaks, Federal Reserve, is sort of like if you look at the kinds of things that are posted, whereas it seems like the talk of what you're building in the community is sort of more of a kind of a engineering, computer science kind of way of thinking of things. Is that fair to say?

I'd say so. In general, kind of Ethereum community politics, by which I mean both sort of politics regarding what to do with the blockchain itself and political issues in general as different and I guess you could call it sort of more mainstream in some ways. It's sort of emphasizes different things. It's definitely more focused on technology. Also, one of the comments that I've heard from an Ethereum developer is that he got the impression

that Ethereum people, in his own words, value optimality. This essentially means like a notion of sort of technical pragmatism and kind of building things in, the way that makes the most sense rather than the way that trying to, you know, want to do some kind of constitutional originalism based on white papers or written in 2010 or whatever.

I mean, that goes another big difference which are just sort of, and then this goes to like the hard fork and all the kind of discussions lately is the attitude towards change. Like, you seem to take the attitude that I would say is very similar to, let's say, how the Linux community approaches things, which is you need

to build new features, you know, need to rapidly fix things. You need to make changes. Like, you need to stay up with the future. It seems like your approach is what is sort of more similar to, let's say, Linux or other kind of classic open source projects, whereas the Bitcoin community has the attitude that there's sort of this, you know, fixed code that should only be changed if it's absolutely required. Is that fair to say? And then maybe we could talk a little bit about the hard fork and all that. Yeah. Like, I think in general, the Ethereum attitude is probably a combination of a few things. To some extent, I would even say that this also ties into politics because for Bitcoin, like, practically speaking, there is this political element to the community. And if you're one of the people who's in it for the political elements, then, you know, even if the lighting network happens, never happens, even if part forks never happen, even if three years from now transaction fees for Bitcoin are like one. $1. Even still, it's good enough for donating to WikiLeaks. It's good enough for dark web

stuff. It's good enough for any kind of like internet privacy preservation use case or anything that's kind of really important. I feel like it's like there is this sort of ethos that it's trying to create something that's high value for a smaller community. Whereas I feel like the Ethereum community is trying to is more interested in building things that are kind of medium value for a larger community and that carries with it a different set of constraints because like one of the consequences is that you know if Ethereum ends up having like a 20 minute block time in transactions start costing you know two and a half dollars and like user interfaces don't work well and we don't have scalability and so forth then that practically speaking that means that like nobody's going to

use in the system. So in general, what this means is that people view it as a project, which is it's not something that's standing still. Conservatism as a philosophy, you know, in general, is kind of about preserving things, whereas the Ethereum is the stomach still a project, which is aimed at sort of moving towards something. And the thing that it's moving toward, like a large part of it does inherently require technical improvements. It requires things so like scalability, it requires things like coming up with some solutions to a transaction speed, requires coming up with solutions to privacy. It requires all of these kind of changes. And like people are more willing to accept, you know, that there's like uncertainty about what the ether supply is going to be,

that there is some uncertainty about when it's proof of stake coming out of like what number of threads is they're going to be in Ethereum 2.0 for scalability and all those different statistics. There are some people who are concerned. Like there are people who sort of come in. from a Bitcoin mindset and they say, well, this thing over here committed to 21 million forever, whereas this thing over here doesn't seem to have any commitment at all. And there's people for whom, like, that philosophy doesn't really sort of mesh too well. I would argue that, you know, in general, the Ethereum view is that it's definitely that the currency serves the technology and not so, and not the technology serving the currency. So, yeah, it's probably a combination

of all those different things. So let's talk a little about the hard fork and the, and the stuff that's been in the news, Ethereum Classic. Sure. Yeah, I mean, I think Vitolik just summed it up pretty well. It's basically, you know, we're seeing software development and money come together in a way that we've never seen before. Thus, we're sort of facing these questions that we've never faced before. On one end of the spectrum, you have this idea, right, that the network should be immutable. It's very principles-driven.

and that is the way that we can create the optimal sort of like digital money, if you will. On the other end of the spectrum, you have more of like the startup mentality, basically, which is we're really, really early in this space. We should be trying whatever we can try to see if we can get product market fit, basically. And in the case of Ethereum, I think that means people building apps that other people in the world find useful. So that's another sort of way of looking at maybe the Bitcoin community versus, the Ethereum community at the moment, or even in the case of the hard fork, the Ethereum classic community, potentially versus the Ethereum community.

Can you guys just explain that a little, the Hard Fork, a classic thing just for people that don't know? Yeah. So the basic gist of it is Ethereum made a decision as a community to do what's called a hard fork. A hard fork basically means you make a change to the software protocol. that is not backwards compatible. And the reason it's called a fork is because if you think about the blockchain as this string of transactions over time, it quite literally forms a fork where when this

change is made, there's a new history getting created on one new version of the fork and another history that's on the other version of the software that's being run. specifically what happened here is there was a decision to fork after the Dow, which was a large decentralized application with a couple hundred million dollars in it, was found to have a software bug and basically leaked, let's call it, $160-ish million worth of Ethereum causing a bit of havoc on the network. So there was this whole debate. Miners and users voted in a series of informal ways.

And ultimately, I think to Vitolik's credit, there was a suggestion made by the foundation or by Vitolik, which was to do a fork. Users and miners were given the option of which software to run. It seemed the majority of people decided that they did want to do this fork. And so Ethereum continues. Now, as a necessary outcome of this as well, there is another fork of the software, which is now referred to as Ethereum Classic, which is kind of the unmodified version of the blockchain. And again, I think you can sort of reduce this to people who support Ethereum Classic tend to have a hard ideological view.

that code is a law and that it's incorrect to modify transactions that have occurred in the past. The Ethereum view would be more one of we haven't even reached product market fit yet. We should be trying to make the system as good as possible and get ideological later on when it feels like the trade-off between moving fast and having extremely predictable outcomes for everything is more relevant. But Vitalik, maybe curious to hear your thoughts as well. I'll probably have to say that this whole fork and the circumstances around it are definitely something that was kind of unprecedented in the history of kind of cryptocurrencies generally.

And I know there have been like forks in much small. smaller blockchain. So things that have like a $120 million market cap that have like a much more sort of centralized and smaller developments team where they were just able to pull it off. And like in some cases, it did well after that. In some cases, it didn't do well. But here, you know, we have something that's like by many metrics sort of second place among the cryptocurrency space in general with a billion dollars of value with lots of different companies developing on top of lots of different people watching it. And actually doing the thing that, you know, lots of people. said should never happen, which is what's called a controversial hard fork. So it's not something that's particularly novel technically, because Ethereum actually already went through two hard forks even before this. One was to implement the homestead changes, and one was five days after launch to implement the difficulty ice age, which is something I will go into right now, but generally essentially has the purpose of kind of putting a sort of finite lifetime on the Ethereum blockchain. So basically forcing us to kind of fork to add new features at some

points within the next year and a half. But those forks basically passed through completely smoothly with no issues whatsoever. But here the difference is that there actually is a minority of people that's a non-negligible minority of people that actually really feel strongly about kind of continuing the old version. And on the Bitcoin side, there is this political divide where a large portion of people believe that basically, you know, this sort of thing should never happen and it's something that's kind of very bad and disenfranchises people and so forth. But before this moment, it was something that seems to be kind of completely unknown. It seems like we really just have like no idea of exactly what would happen in such a scenario. And now, you know,

what we've had is a situation where it finally has happened and we actually finally have seen what the results are and what the results could be. like we've even basically seen the blockchain quite literally split into two and both forks continual and kind of end up having their own kind of prices on the market so this is all stuff that i actually theorized about quite a bit i had a meet up in silicon valley at the end of april where it talks about protocol governance and one of the scenarios that i had was that you know if there really is a kind of chaotic a sort of maximally chaotic hard fork really like nobody knows which side is going to kind of quote win right right up until the last second

and people sort of persist in running both chains, then the theory would be that eventually, like, there would be a few days of confusion, and then people would, at some point, realize that, you know, basically side A agrees that side B is not going away. Side B agrees that side A is not going away, and they'd, to some extent, make peace. They'd try to do things to prevent transaction replay attacks. They would end up being listed separately on exchanges and so forth, and, like, the situation would slowly normalize. like it feels kind of like almost like a sort of like a national secession scenario although it's obviously kind of very very different in lots of ways but what I

what I was theorizing at the time was just theory and now we've actually sort of seen a situation actually sort of happen there are a lot of kind of very interesting variables to it that I didn't anticipate so one variable that I didn't anticipate is that you know the Ethereum classic chain would be sort of almost dormant and nothing would happen for four days and then it would get

treated on Poloniacs and then all of a sudden it would get big. So the thing that I was kind of expecting was actually for, you know, either classic to not be a thing at all or for it's to kind of almost immediately hit the ground running, but instead you have this sort of weird leg time. And I was also, I think in general, kind of surprised by the level of adoption of classic, especially given the way that the situation looked like in the hours leading up to the fork where it seems like everyone was just kind of either agreeing to it or accepting it. Of course, you know, had we known that things would turn out the way they had turned out, there were things that we would have done differently.

So like one of them would have been including, you know, transaction replay protection, straight into protocol directly. And that's something that we may well end up doing like even for pretty much every normal future hard fork, possibly now that we know what the benefits are. financially, right? It actually worked out for Ethereum holders, because if you add up Ethereum and Ethereum Classic, the value. Yeah. Yeah, the ETH plus ETC is above sort of ETC before the fork, which is actually interesting because, and I feel like this is one of those sort of great kind of political debates that's, like, it has parallels even like way outside Cryptoland,

which is that people in generally, you know, like these sort of ideals of unity, harmony, network effects, people kind of working together, and doing the opposite of that is sometimes sort of perceived as either kind of weakness or confusion or chaos and so forth, whereas the thing that it might not realize is that there are times when a split actually can increase total value. Outside of Wachshan land, one of the examples of this is that there's been some fairly decent amount of research. Like if you read some of the behavioral economists, so economy and so forth, they'll talk about this a bit, which is that mergers very often end up reducing the value of the companies that go into them. So, look, there are times when, you know, coming together is a sort

of negative value event, and there's times when splitting up as a positive value event. In this case, it definitely does seem like it's actually been positive value on that. Now, the question, of course, is why. And I think the answer here is interesting because, like, the general sort of viewpoints, I think, in cryptocurrency land before this as a sort of notion of Metcalf's law, which is that the value of a network is kind of n-squared with the number of the users. And so you want to have all the users on a single network to maximize value. But here we've basically learned that I think that there's some cases where it actually goes in the other direction. And in this case, I think one of the reasons why is that if you look at

what the heaviest sort of ETC support is right now, even if you look at, you know, Barry Silbert has been sort of very positive on it. Charles Hotskinson has been very positive on it. Then Chandler-Guo has been switched to being positive on it lately. Then even the Satoshi no Komodo Institute, which has just wrote an article yesterday that was positive on it. So people who up until this day were pretty much like diehard Bitcoin maximalists. It's either Bitcoin or nothing else. And, you know, even Barry sobered even admitted to saying that, like, he before ETC, he never bought, like, another cryptocurrency.

Now they're starting to sort of be interested in getting into Ethereum because either they feel like there's just sort of a version of Ethereum that has their philosophy attached to it or something similar. Like, that is one, you know, I think one of the reasons why people sometimes believe that, you know, this notion of kind of emphasizing change. choice is a good thing because, you know, if you have choices, then you're able to kind of attract larger groups of people that have different values. It's actually similar to Volaji Thrinivasan's sort of voice and exit philosophy. It's interesting. It's sort of like, you know, eBay and PayPal when they split. It was like $75 billion before the split and then they make $85 after the split. There's actually a decent precedent for like Wall Street when companies split up, right? I mean, so the Bitcoin, the conservative people can have their own version of

Ethereum and there's a, you know, everyone gets what they want. Yeah. I mean, purely from a like things being tried in the world point of view, which again, I think is where we are just in the timeline of digital currency, it feels like a good thing, you know, personal views about the hard fork aside, or anybody's personal views about the hard fork aside, I think it's a good thing that we're just trying multiple things in the world. Like, I think it's very easy to think about these.

seemingly political issues and take a side and sort of vigorously defend a particular side. But at the end of the day, if you're just rooting for digital currency to do great things in the world and to figure out what's going to work best in the world as quickly as possible, then it's actually a really great thing that multiple things are being tried, even if you happen to agree with one approach and not agree with another. So when Bitcoin was coming up, people with Bitcoin projects would come to people like us, venture capitalists, to raise money. With Ethereum, people have been crowdfunding much more and not coming to people like us,

which I would think is a bad thing. I actually think it's a great thing because I think it just sort of increases the amount of innovation. But so, like, Vitalik, you, like Ethereum Foundation itself, I believe, was funded through Ethereum crowd sales, right? And then the Dow and Auger and, I don't know, there's a whole bunch, there's a whole long list, Fred and Vitale, you guys have better than I do that have all. Storageo, file, coin. Part of it's enabled by Ethereum itself, right?

because a lot of these things are built on top of Ethereum, and then part of it is just kind of the spirit of the community. Yeah. So, yeah, I mean, it's an interesting mechanism, right? Like, as you're saying, the sort of normal route for a company is, let's get an idea together, let's hopefully get some type of a prototype together. We're going to draw up some on-paper legal incorporation documents. Then we're going to go to some angel investors or a venture capitalists, raise some money.

if all that goes well, hopefully we IPO on the NASDAQ or the New York Stock Exchange at some point in the future, right? And this whole thing is kind of this closed process, right? Where venture capital itself, like a lot of the funding comes from a couple of guys sitting on Sandhill Road out here in the valley. And this, you know, just in the way that I think Bitcoin and Ethereum are sort of turning finance on its head into a software problem and a fundamentally open network instead of a close. network, we're seeing the same thing happen for projects to get funded and financed. So, yeah, I mean, we've seen, what, five or six different projects at least raise a couple million dollars in funding just through crowdfunding on the blockchain. And I think at first glance, this stuff looks like just a new way of financing a company. And it goes a lot deeper than that. It turns out that these things really aren't companies, projects might not even really be the right word for them. They're really software protocols that are almost replacing a centralized company or what a centralized company in the world we know today would do.

So one example of this, and who knows if this project will actually work or not, is Steam, which is a decentralized news and social network. network. And the model is instead of having a centralized company sort of create and extract rent from owning that network. So like Facebook created a Facebook network and makes a bunch of money off of it every day through ads or Reddit as the central hoster of that community, it's a decentralized protocol where people can share in this case news information with each other. And if you contribute to the network, then you get paid in the native token of that network. And that represents either a liquid payment you can cash out for dollars or something that feels kind of like equity in the network where if the network's successful and it presumably

goes up in value over time and you benefit as an early contributor. So this is sort of like a whole new business model that I think we're seeing sort of emerge in the world. It's, I think the best way maybe to describe it is if you saw internet first companies kind of in the 90s as a concept that creates entirely new or companies that look entirely new from a structure perspective, although it doesn't make sense for everyone. The same thing with mobile first companies in the 2000s. I think we now have these sort of like blockchain or decentralized first projects where, again, it will redefine how people look at the business model.

It doesn't mean it makes sense for everyone. Yeah. In general, I've always been kind of a proponent of crowd sales as this sort of very experimental but interesting, different way of getting funding for things that have historically been very hard to get funding for. And I do think that they have their problems. So like some examples would be just the fact that with a traditional crowd sale, the incentives to the developers are all up front and there is theoretically not as much incentive for the developers to kind of keep working on the project in the long term. There's the fact that, you know, quite honestly, a large number of them are for projects that aren't very good and some of them are even dishonest in various ways. But like even still, I think on the whole, it's like bottom level protocols or something that's been like so horribly underfunded. You know, like in the mainstream world, if you look at even something like the hard bleed bug, which happened in large part because open SSL security research was just like so horribly underfunded, like a world where basically your protocols.

protocols actually do sort of more naturally get developer attention seems like a very attractive thing. Almost all widespread protocols are 20, 30 years old and were government funded, right? I mean, because there's basically no funding since then for protocols. I'd probably make an exception for some things that just sort of ended up as being soft standards that what came out of Apple and Google. But Apple and Google these days are in some ways acting more and more like sort of governments of the internetian, probably in both positive and negative sense of the word. Yeah, Vitolik, how would you sort of, if you think about maybe the dynamics that would exist

in the world, if there are more protocols and less sort of centralized companies, what do you think the outcome of this idea of creating app coins or crowdfunding or more protocol development at whatever angle you might take? What does that look like in the world in 10 years? I think that in general it definitely has the potential to make a lot of things kind of both more competitive and more efficient and kind of allow a lot of the benefits of network effects without the harms of, you know, having a single monopoly company control, like an entire particular industry. One of the examples that people quite often bring up is even looking at something like Uber, where, you know, if you look at it naively, you know, if you look at it naively, you know, you might think that there's only kind of two rules in the Uber ecosystem, which is that you're either a driver or you're a passenger. But if you look at the ecosystem kind of more deeply,

then you realize that there's actually a lot more functions that are part of that sort of ecosystem. So one of them would be filtering drivers, even things like checking that the drivers are good, checking that they have insurance that they're credible in general. There's providing a search engine, providing an interface, providing kind of ratings of verification for passengers, several other smaller tasks. And one of the ideas in this protocol-centric notion or world has always been that you can take these kinds of industries and actually sort of decompose them

in this kind of very multilateral way where allow people to like very separately focus on each individual components where, you know, the theory is that you might, if there's some company that's really good at machine warning, that it might apply its machine warning to, you can even think of it as a sort of special kind of mining where the tasks might be things like matching up drivers and passengers. Tasks might be things like, you know, finding efficient routes, finding arbitrage opportunity is finding cheaper ways to move some product from one place to another. Yeah. It's a sort of unbundling almost, right? I mean, it's Uber without the central controller of the network, Uber, and what you're seeing in this new model is a lot of

the value that was once extracted by owning the network moves into sort of this general ownership model by all the people who contribute to the network in one way or another, whether they program the software or their, you know, they contribute their labor to the network or whatever it may be. And then the business model, as you're sort of saying, really becomes one of these value-added services around the edges of the network, whether it's verifying drivers or... Very similar to the web or something, right? Where you have the network is community-owned.

People go over there's on top of it. If I could just wrap it up by getting you guys to talk about your predictions for the next five, ten years about Ethereum and Bitcoin. You know, I guess one, what's the future of Ethereum? Will it coexist with Bitcoin? What are some of the applications you think might be built over the next five, ten years? You know, what do you think the future of cryptocurrency is? Is it just the very beginning?

Is this like 1985 in the Internet? Matt or 1995, or where are we? One of my favorite quotes with regarding the two or the general sort of, this is the X or X of the Y pattern, is that history doesn't repeat, but it does rhyme. You might say that this is, you know, this is equivalence to stage X of some other technology Y, but in reality, there's going to be a lot of similarities, and there's also a variable going to be a lot of just fundamental differences. So just as one simple example, you know, the internet, or Bitcoin started off.

is almost being, you know, run by crypto anarchists, but the Internet sort of being run by the government. And, you know, those are kind of polar opposites and you're going to expect to see sort of fundamental structural differences coming out as a result of those, of that kind of difference. With regards to where we're actually at, I think that in the last six to 12 months, we've actually been starting to see projects start forming around actual use cases,

and we've been starting to see all these different ideas get tried, Like, in some cases, it's some kind of like POC being run inside of a bank. In some cases, it's an actual startup making a product and actually pushing it out to the public. In some cases, it's still projects that are under development. And we've been seeing these increasingly many industries, including, you know, finance, IOT, whatever else. So I think that in the next few years, like, we're going to start seeing some applications. I don't think there's going to be one killer app, but I think there's going to be a few that we're going to see. start finally sort of breaking out into the mainstream. And there are, I think, a few kind of

limiting factors that have been preventing that from happening up until now, whether they're sort of receding on a similar schedule. One of them is just the technology in general being kind of perceived as being trusted, so like how much people are willing to work with it, how much people are willing to trust it, how much people are comfortable with using it. There's obviously the sort of government's and regulatory issues. There's also the user interface. face issues and, like, just practical security concerns, then there's the constraints that have to do with the technology itself. So, like, there's a huge number of applications, and unfortunately, it's like most of the

applications that people tend to be excited about. That would be viable if wallchains could do 50,000 transactions a second, but, like, totally not viable if watchings can do 15. So, you know, Ethereum, Bitcoin, Bitcoin, Classic, and Bitcoin with Segwit, like, all four of those are just pretty much completely useless for a lot of what people want to do because the scalability just isn't high enough. And that's just one practical thing that we're going to just have to get over in some way. I think that technological barrier is just going to have to be solved over the next few years or so. And then there's general kind of ecosystem maturity

and the sort of higher level infrastructure around the technology is being sort of trusted enough in order for people to be willing to put like huge amounts of capital behind it. Yeah. I mean, it does it does feel like we're sort of hitting our stride in terms of people actually building interesting things in a way that we really haven't for the last three to four years in Bitcoin land. It feels like we're just at the very, very beginning of people figuring out what really useful services can I build using a blockchain, specifically Ethereum, I think, is where most of the interesting stuff is happening. And what is the right kind of model around that?

look like. I mean, it took quite a while for somebody to synthesize Bitcoin in the first place as a, you know, as an idea that combined a decentralized network with the right economic incentives to keep the thing running. And I think we're basically seeing that, but people are now applying it to all sorts of different ideas, businesses, applications. And it's going to take a while for people to hone in both the economic model for those and technologically just to build these things in the first place, right? Like we saw with the Dow, it's not so easy to build a decentralized application that has a bunch of money involved. So if you look at like a lot of the projects that are being built right now, let's take Auger or prediction market, I mean,

they're sort of thinking about these things on a daily basis and struggling through them as one of the first people that are really building a decentralized app. So it's, you know, how do I really easily do an initial crowd sale for my project? Like there's, not a standard library out there that lets you have a smart contract that offers up some percentage of your app coins to be sold and have people send money to an address. And then you distribute kind of the app coins based on how much each individual contributed on a pro rata basis. Like that should be a standard software component that just isn't quite built yet. Other people are working on even like board governance libraries really more or less.

So it's like it's early and people are sort of building the tools, I think. think in the way that we saw people building easy-to-use kind of software scaffolding or frameworks in the earlier days of web apps. So, like, you know, we don't have sort of templates for building apps in the way that things like iOS or Ruby on Rails or Bootstrap gave us easy ways to kind of construct web apps out of different components. That being said, like, I think as people hone that, hone these things, iteration will start to get quicker. And I think the first thing we'll really start to see is this idea of like X without the X getting built. And we've talked before about like Uber without the Uber.

I think that'll that'll be tried for a number of different ideas. You can you basically can apply it to any business right now that's based on network effects and take the central owner or operator of the network out of the equation. For me, a key moment will be the day I can't wait for is when there's an application built that's used. by people who aren't interested in cryptocurrency, right, and just like crosses over and you don't even realize it's built on top of Ethereum or Bitcoin or whatever it might be in the same way that, you know, people that use web browsers and whatever go to Facebook don't think about HTTP and HTML and things like that. It feels like we're at a very exciting time now

because I think a lot of the work that you guys both have done, where we're now starting to see kind of this new wave of applications. And at some point, I think some of them will start to work And I don't know, that's my hope. My hope is we're sort of exiting the infrastructure era and entering the application era, I guess. And it seems like that, but it's always hard to know. I think realistically, infrastructure and applications are both going to continue progressing

and parallel for the next few years or so. Yeah, yeah, I mean, it's done. But we're starting to see more and more applications and the infrastructure is good enough that you can start to build on top of it reliably. Yeah. It almost feels like Ben Horowitz told this story to Coinbase one time where he talked about how in the early days in Netscape,

they had somebody asked them to come build an online virtual shopping mall, and that kind of caused them to ask all these questions of what the Internet could do that they hadn't previously solved. I think out of that SSL, the protocol for securely transmitting data over the Internet was invented. So I feel like we're in a similar stage where, like, there's enough tooling and infrastructure out there to build an application. It's just challenging. And then as developers kind of work through building these early applications, more and more of that infrastructure will kind of get built in parallel, I think as Vatolic is implying. Yeah, I mean, Nescape, people, people, a lot of

people forget this. Now, Nescape built JavaScript. They built cookies. They built SSL. And they built them because they had to for their customers, because of the kind of the academic research internet before hadn't needed those things. So no, you're right, that these things will be built out. As the applications get more sophisticated, the infrastructure will continue to get built out. Okay, awesome. I think we're out of time. Thanks so much for Talek and Fred. All right. See later. See later.