"People don't realize how close the tech is to being ready" - Vitalik & Balaji Srinivasan at ETHSF

So, we did just say this was all about hacking hackathons, but we do want to keep a conversation going with people who are who are kind of talking about big ideas, important ideas. So, we'd like to invite a few speakers onto the stage. So, our moderator will be Kartik Talwar from ETHGlobal and SVA Angel and deserves a big round of applause for all the work that he's done. If you see him Way to go, man. [applause] If you see him, say thank you.

And everybody else at ETHGlobal as well. We have Balaji Srinivasan, the CTO of Coinbase, uh ex-21.co and earn.com as well, among many others. And Vitalik Buterin, the founder of Bitcoin Magazine.

Welcome. Thank you. [applause] Test, test. Okay. Um before we start, I'd like to just clarify that I'm not the only one that made this event happen.

There's about 100 people from volunteers to organizers to everybody who supported us. So, I think they should get a bigger round of applause uh than just ETHGlobal. So, we're going to cover a lot of stuff here. Everything from kind of high-level thoughts into this ecosystem to going granular into the tech. Um but, let's just start off with something simple.

Um I know you guys have a really good history of uh meeting a few years ago. Uh Balaji, why don't you tell us about how you met Vitalik initially? Yeah, sure. So, um the first time I met Vitalik was actually in late 2013 when I was at Andreessen Horowitz. And at that time, uh Blockstream or the ancestors of Blockstream and the ancestors of Ethereum were actually all co-founders of the same gigantic startup.

Uh Most people don't know that blockchain and Ethereum were once on the same side. And there's just huge conference room with a bunch people there and I said, "Who are the founders?" And everybody raised their hand at the same time. And I was like, "Oh, this is going to be interesting." So that's how I met Vitalik.

Awesome. So, we're also going to do a Q&A at the end. And if you have any questions for for Vitalik or Balaji, please tweet us at at ETHSanFrancisco or use the hashtag ETHSanFrancisco and we'll kind of moderate and and get the questions up at the end. So, Vitalik, your your title is Chief Scientist of the Ethereum Foundation or the Ethereum project. And Balaji, you are the CT of Coinbase.

What is it that you actually do at at these organizations? What is kind of your day-to-day look like? What are some of your responsibilities or projects you're involved with? Kind of give us an overview of what you do. Number one, fundamental research on things like proof of stake, sharding, plasma, protocol economics.

Number two, helping to kind of go a little further down the pipeline and actually write down the spec. Very often write proof of concept code for kind of testing, improving, and simulating different kinds of ideas or at least writing what would eventually end up kind of becoming the first version of certain pieces of code. Coordinating with and talking to other researchers. Do I kind of participating in the discussion around the Ethereum Foundation and Foundation's overall strategy. Fighting with Bitcoin maximalist trolls on Twitter.

Yeah, they're probably the big ones. Yeah, so I joined in May and typical, you know, typically what I do is one part technology, one part management. And I'm trying to we're driving Coinbase towards a you know, multi-asset or many asset and uh you know, many new crypto technology future. Uh and so I'm doing a lot of that. You'll see a bunch of that manifesting on Coinbase, you know, this quarter and in the quarters to come.

Um and a typical day is basically like, you know, can be writing API specs, it can be um it can be, you know, managing folks, it can be working through all the various legal details to get something new launched. Great. Um kind of taking a step back, how did you initially get involved into uh the space and sort of what motivates you and and why? So, I first uh joined the space back in 2011 and I heard about it from my dad and I actually ignored it because I thought, wait, how can kind of pure bits on a computer that are not backed by anything have value? Then I heard about it again from the internet and I thought, okay, this might be something that's kind of worth looking into.

And I popped into the Bitcoin forums, then you know, comes the story that I already told in many cases, many times about starting to write for Bitcoin Bitcoin Weekly, then moving over to Bitcoin Magazine. What interested me about Bitcoin was that it was really a kind of combination of different things that all interested me individually put together. So, I was already doing a lot of math and computer science. I was already interested in any economic and political topics. Um I was already kind of interested in open source software and Bitcoin and at the time it really was just Bitcoin kind of was all of these things put together and it seemed like an idea that was kind of at least interesting enough to kind of be an idea and a community that was interesting enough to at least be worth trying to sink into.

Yeah, so uh you know, I heard I first heard about Bitcoin, I think in either 2010 or early 2011 and was following it um through the you know the initial bubble and whatnot. And what gave me a lot of faith in it was uh you know the technology was awesome, but it was the recovery after that first bubble crash and then uh you know comeback over 2012 that made me think, "Okay, this might have legs not just as a technology but as a as a market phenomenon." Um I got very heavily into it late 2012 early 2013. Taught one of the first classes at Stanford on it in early 2013. Actually, a lot of a lot of blockchain entrepreneurs came out of that Um and uh and then you know the whole thing really took off in 2013.

It's great. I kind of just following from that um in 2014 when you thought about working on Ethereum, what was sort of the original vision and sort of how has that changed so far from your perspective as the years have passed? And And Balaji, if you're to comment on that later, how has your perspective still still still kind of being involved from the early on stages has changed um about the ecosystem? So, my uh vision for Ethereum actually started off being incredibly unambitious, right? Cuz I was working on this uh project called Mastercoin.

And Mastercoin was this uh initially this well well and always this uh layer two system on top of Bitcoin that was trying to expand the utility of the blockchain going beyond just payments into other use cases including decentralized exchange, um betting, domain name systems, crowdfunding, like contract smart different kinds of contracts. And Mastercoin took what I described as the Swiss Army knife approach where they said, "Here's 23 use cases. For each use case, let's make a feature. And let's make a protocol with 23 features." And I um was uh kind of hired into this team very briefly.

And my immediate approach is, "Can we try to take some of these features and immediately start generalizing them, right? Can we start instead of creating 30 features for 30 use cases, come up with single features that can be designed in ways that are kind of more general purpose and abstract and allow a bunch of other sets of use cases. So, I started off just kind of generalizing two-party financial contracts and I basically said instead of just having one parameter, which is the leverage, you turn it into the whole thing into a mathematical formula. Then from there I even created this white paper that I called ultimate scripting and the idea was basically to replace Mastercoin's entire feature set with this a programming language and I sent this white paper off to the Mastercoin team and I thought like this is obviously such a powerful and amazing and obviously the right way to go and they said, "Hmm, this looks interesting, but this seems ambitious and this seems like this will be a very far along step in our road map and maybe we'll get to it in a year." And I'm like, "Guys, this is like so obviously the the thing that you're supposed to do and you're going to wait an entire year?

Well, screw that. I'm going to do it myself." And started doing that, wrote the white paper myself, sent it to 15 people and out of Ethereum came from came out of that and over [snorts] time the vision kind of expanded almost even more with the feedback from other people than out of my own initiative, right? Other people started coming up with ideas, building applications, and I just realized, "Hey, wait. You can use general purpose blockchains for this other stuff.

Hey, wait. You can use general purpose blockchains for this other stuff." And it kind of slowly dawned on, you know, probably both me and the Ethereum community in general that it really general purpose blockchains really are general purpose and really are usable for a much wider variety of things than we thought blockchains should be used for back in 2013. Got anything to uh add from your perspective of the project from the initial phases or [sighs] Do you have any comments to add from kind of your perspective from meeting and learning about the project early on to how you've seen it evolve uh now. Yeah, well, you know, in the early days actually um I remember uh some of the folks who were working with Vitalik talked about like colored coins.

Uh and that was a big thing and you could if you Google colored coins, you can see a video from that time period. And now of course we have called those tokens. Um and a lot of the or a lot of the concepts from tokens are are are similar. Um so it was interesting to see certain things which couldn't be done on the Bitcoin blockchain. Um you know, first there was exit in the form of Ethereum and now there's exit in the form of many other chains.

And so all that pent-up innovation has kind of come out over the last like year and a half or so. Got it. What does mainstream adoption mean to you? Um where do you think we are in sort of uh just the adoption curve for for cryptocurrencies in general um with Ethereum nearing 4 years, Bitcoin nearing 10 years? Kind of how do you look at that as a broader space?

So uh you know, where we are now I think is, you know, in tens of millions of users land, right? And there's a difference between that versus hundreds of millions and obviously billions, right? And truly mainstream probably in the US would be when we're at hundreds of millions of globally it's probably mainstream in the US. When we're at billions globally it's mainstream worldwide. Um I think it gets to mainstream in the US probably over the next decade or so and might take more like 10 to 20 years for it to get mainstream worldwide.

And that's still a very fast growth rate that kind of mirrors kind of the rise of the internet. Um and that's like one quantitative difference definition of what mainstream is. Um so back in uh some I think 2011 or 2012 I uh think I either wrote this article or kind of expressed this idea somewhere that there that I saw three possible paths for what at the time was just Bitcoin where one the analogies I made are kind of Esperanto, Linux, and the internet. And Esperanto was this kind of language that a that um a guy named I think Ludwig Zamenhof came up with in in the 1900s and he thought that if we create this artificial language which is kind of completely regular, maximally easy to learn, and it's politically neutral, then we can get the whole world to speak it. And but it ended up just never taking off and you know, it has a small enthusiastic community but it just never grows at all beyond that.

And Linux, you know, it like there's this you know, there's the joke in the Linux community that the the year of Linux on the desktop is always like the current year plus five or whatever it is. And you know, there's the dream and the dream never ends up being fully realized but even still Linux totally survives as a legitimate alternative to kind of more mainstream proprietary operating systems and it has achieved like extremely real successes in places that it didn't even think of before like, you know, in mobile through Android and and in servers. And that the internet is, you know, Bitcoin is the next great invention since the beginning of the internet itself and that whole kind of expansionist vision. Um I would say at this point we've definitely gotten enough adoption that we're way past the phase where we have to worry about it becoming Esperanto. Whether it looks more like Linux or more like the internet, I think we're definitely still unsure and it definitely still is one of those kind of time will tell sort of issues.

Like the big question is basically are there going to be applications that actually do manage to kind of use blockchains and use blockchains kind of front and center in ways that you know, could reach very large numbers of people. And realistically speaking, cryptocurrency is probably still the primary candidate for that but there are other candidates that are slowly starting to come up and if anything in that in that area takes off, then great. But even if it doesn't, you know, the Linux outcome is still I think of it, know, much stronger than anything that any of us were hoping for 4 or 5 years ago. And that would basically be a a combination of a kind of fairly niche user base for cryptocurrencies plus blockchains being used to notarize and secure a whole bunch of stuff, you know, you being used to improve the currency of a bunch of a bunch of mechanisms doing a bunch of things in the back end and kind of impacting the lives of millions of people indirectly. So, we'll see.

Yeah, so I actually I like that trichotomization of like niche, nerd, next big thing, right? But I would also add maybe to that Linux, even though it was nerd, it actually did win on the desktop just the next desktop, right? The mobile desktop, Android and and whatnot. And that's kind of an interesting way of thinking about the emergence of crypto is like Linux, it's initially a producer technology, right? It's something that makes developers' lives, you know, better.

It makes investors' lives better like power users. Um and it's actually more useful if you're building a new, you know, financial service or you're trying to do new things with investing than it is for the average person. And Linux went through that phase for a long time. It wasn't good for the average person, but it was great for the person who's building Google and, you know, Amazon and and Facebook and and so on. And then eventually it became useful for the average person through Android.

And I think that might be an emergence path for how crypto works where it's useful for all you guys. You can build, you know, new products and services with it. Um and then eventually, you know, just it becomes useful the average person um as it diffuses out from the developers and the producers. Those are Those are excellent points. I actually emerges with the The next question I was going to ask, so I'll just skip that, but kind of just deep diving a bit deeper into this, kind of with that narrative and and sort of these now possibilities, how do you 10x or 100x this ecosystem?

Anywhere from technological side to distribution to use cases, how do you think about that? Sure. So, one one angle that we're thinking about, there's there's many, but um one angle is there's at least three ways to to get crypto. You can mine it, you can buy it, you can earn it. Um and you know, we did the hardest way first, which is setting up, you know, mining farms, and then buying came on a little bit later.

And I think earning is still something that's open and interesting. Uh and with the acquisition of Earn by Coinbase, we're working on ways to make earning crypto as easy as just, you know, like buying it. Just click a button and earn. And if that works, then it would work anywhere there's an internet connection, because you just click a task and, you know, make money. So, that's one possible approach.

Um there's other approaches as well, which uh are based around taking all these new coins that are coming out and really stressing the utility, right? You know, investment is generic, buy, sell, send, receive, but utility is specific. You know, for example, send a shielded transaction with Zcash, or use BAT in your browser, or use REP for something, or use, you know, like ZRX for something. They've got unique things you can do with them. And each of those individual markets and niches, maybe those can be pulled [clears throat] into cryptocurrency as well, because those cryptos end up doing something unique.

Uh so, that's another approach potentially. The other suggestion on the market side that I've raised a couple of times is just making it easier for people to buy small amounts of cryptocurrency in corner stores. So, just increasing the accessibility of, you know, I want 40 bucks of crypto just enough to make to participate in some applications a little bit and pay for a bunch of transaction fees. Um because like if you don't have even the small amounts of cryptocurrency, then it's hard to participate in public blockchain applications that are even uh non-financial in nature. Um on the protocol side, I think uh solving the scalability challenges is obviously paramount, because otherwise, if the user base goes up by another 10x, then transaction fees are going to be back up going up to like five five or $50.

Um and another important thing is usability. And that includes kind of the traditional notion of usability of making things kind of look and feel user-friendly and intuitive, but also what I call like I don't tend to talk about usability versus security. I talk about usability of security. So, how do we make it kind of maximally user-friendly to use coins and to use them, but not kind of accidentally do things or have things happen to you that you really really don't want, like $100,000 of your money disappearing into a black hole. And like if those usability issues can be solved, I think it also cuts down the friction by quite a bit.

And I know there are quite a lot of great projects that are starting to work on that. So, I guess I have to follow on. Balaji, you you had a company earn.com that tried being able to kind of monetize users and then emails and being able to give people cryptocurrencies for the first time as like the first way to acquire them. What are kind of just expanding on Vitalik's answers on how do you get people to initially own small amounts of cryptocurrencies?

What were some of the learnings from from doing earn.com? Yeah, so we actually, you know, we got the company to like a 10 million run rate before we before we sold it. Um and the you know, what were the learnings on that? Well, in theory you can do lots and lots of different kinds of tasks, you know, paid tasks.

In practice, we found that those tasks that were crypto-native in some way were the most popular initially cuz that was the community that was familiar with it and that liked accepting crypto. Um I do think that over time, uh you know, as as we get better at smaller payments as Vitalik was talking about, as we have more people who are familiar with crypto, it'll become very common to just like, you know, or tweet and you're sending information, you know, or you you fav somebody's tweet or whatever, you're sending like a like a bit of information effectively, but plus a lot of metadata. It'll be very very common to do these tiny little payments as part of doing other actions within an app, right? So, the blockchain is part of it. It's an important part of the app, but it's an enabling part of the app and there's other stuff around it.

Um and like Earn is like one manifestation of that, but there's others. And so, that's one of the things I've I've thought a lot about. Do you think a bigger audience for for blockchains in general is machine to machine payments or is it broader or bigger on the user side if that makes sense? The biggest consumers, will they be machines paying machines or is it humans interacting with machines? Well, I I think it's a completely new thing that you can have machines holding money that like didn't exist before that you could literally put a private key on a machine and and actually have a true intelligent agent that could earn on, you know, on its own and trade on its own.

So, simply the fact that that's possible is is a new thing. Um I I think it'll take a while for the fullness of that to be appreciated, uh but uh you know, I I know that it's already a subculture, uh you know, the trading bot subculture. There's all these people who communicate about this stuff. They they do these algorithms online and there's people who do this from college students all the way up to like large hedge funds do this. So, so there's definitely starting to be that movement.

It's hard to say whether what the percentages will be in the future, though. Vitalik, if if you were to comment on some of the most underrated technological improvements in the Ethereum ecosystem, what would you say some of those things are? Underrated technological improvements, um I would probably say ZK at this point as ZK-SNARKs and specifically the extent to which you can use ZK-SNARKs already into improve the privacy and scale and scalability of applications. Like there are teams already. Like there's the the I mean, team being led by a very whitehead is probably the biggest one that are trying really hard on making the making things like privacy preserving tools for Ethereum using snarks to create more scalable ways of transaction batching techniques to create more secure and more scalable versions of plasma chains and things like that.

So, I think there's definitely there's definitely a lot of people that don't appreciate just kind of how close the technology is to reality even already. Vitalik, what would you say are the most overrated parts of the space? Well, you know, I always respect folks who are working hard and you know, respect entrepreneurs and and so on. I do think that the whole, you know, banking movement of, you know, blockchain not Bitcoin and and private blockchain and so on arguably overrated and you know, the reason is I think a blockchain without a native digital asset is something which you know, you lose a lot of benefits of a of a blockchain, I believe. Um, and I think, you know, private blockchains may have a V2 where people realize, "Hey, I'll do a like a private ICO.

It'll be set up as a like a joint venture where you have 10 companies in a consortium, but they have actually a percentage of the token." And then it actually might work again. But I I think the whole concept of blockchain without Bitcoin was it was a dead end. A question for both of you. What is your favorite data data structure?

Um, sparse Merkle trees. Me? Uh, I mean, standard binary tree is good. You know, hash map. I use them all.

Great. Um, Vitalik, what is Ethereum 2.0? Um, just can you kind of go into some details on what it is, why it matters, and uh So, Ethereum 2.0 is this kind of catchall term for a combination of different paths toward kind of seriously and fundamentally revamping the Ethereum blockchain that we've been working on over the last few years, where one of those is proof of stake Casper, uh one of those is sharding, one of those is the switch from the uh current EVM Ethereum virtual machine to Ewasm, the web assembly based one, and one of them is um a set of kind of miscellaneous improvements including things like abstraction, um improved protocol economics that I wrote a paper on recently, um among um other kinds of things like that.

And uh Ethereum 2.0 is basically the kind of new version of Ethereum that will take all of these different improvements and put them together and essentially create what we believe is the best possible design for a general purpose blockchain given what we know in 2018. And so the goal is to kind of have a roadmap to migrate Ethereum over into a kind of being that over the next few years. So so could you uh elaborate a bit more on when uh what does that timeline look like and sort of what goes into determining those things? Uh sure.

So one of the kind of big uh things that's made me a lot more optimistic about the whole thing over the last maybe four months specifically is that the research uh side has a really kind of congealed and stabilized a huge amount, right? It feels like the research side at this point is uh like there's no kind of significant unsolved theoretical problems left. It's basically all details and kind of safety checking and verification. Um and right now it's basically the kind of a software development slog, right? And we're familiar with software development slog.

So like if you recall Ethereum 1.0, going from the original first step of the Ethereum white paper to Ethereum 1.0 being released altogether took 19 months. And right now we're actually very significantly ahead of the initial white paper release. So like I would say right now with 2.

0 and with where the protocol is and where the software development is is probably where Ethereum 1.0 was possibly in maybe the fall of 2014. And that means and there are already you know like independent companies like for example prismatic lighthouse status with Nimbus and a couple of consensus parity starting to work on kind of independent implementations of the Casper and charting spec. And I expect test nets to be coming out fairly soon and then you know we'll have in parallel work happening on test nets and along with formal verification of the protocol auditing of the protocol auditing of the code and implementations and that will at some point kind of culminate in you know like some some form of main net release. So just kind of quickly digging into this What's sort of the rationale behind having other teams work in parallel and kind of how do you see that versus Ethereum foundation Sure the core protocol team also work on the protocol.

The Ethereum community kind of really strongly believed in this idea of multiple implementations basically since day one. So for those who remember the 2014 15 developments log right you had python get go geth and C++ basically developing the protocol in parallel all the way through with the intention that all three of these implementations would be viable implementations of the of the protocol that people could run and use. And the goal behind this was a large part of it was political like basically we the kind of what we perceived as kind of failures of development centralization in the Bitcoin ecosystem and the kind of arose from the existence of a single implementation implementation there is something that we thought that we wanted to avoid basically by saying there is no such thing as a reference implementation and the reference is the protocol and the protocol is, you know, the yellow paper and the protocol is distinct from implementations of the protocol." This time around, we're actually going a bit further because there is one version of the of this of the protocol which is made by kind of people who are formally hired by us by the Ethereum Foundation as full-time developers and that's the Python implementation. But, Python is a slow language and everyone knows that there's plenty of contexts in which Python is just set up that it is just not going to win, right?

So, it's um like if and so that kind of naturally sets it up for kind of competition between outside entities to create the client clients that people will end up using. And and there will be many people that that use Python as well, hopefully, but not in every context. And um you know, so the Ethereum Foundation basically gives grants in part and other companies are giving grants as well to and so with in the case of status, you know, Nimbus is itself funded by Status's very large ICO raise, right? So, the goal is to have this kind of robust and a community ecosystem and on top of that, it also mean it has this very pragmatic dimension where it insulates um the Ethereum community against the risk of any one of these teams failing and any one of these teams just ending up to not being able to figure out how to do software development correctly. Yeah, one of the things actually that Brian Armstrong and me and Fred Ehrsam were impressed by by Ethereum in the early days was that it did have a protocol spec.

It It did divide a protocol from a specific implementation and had multiple implementations. Um you know, one of the things we've actually been thinking about doing recently is um publishing like a Coinbase API standard um that would have maybe a similar effect for what we do with security because not all blockchains um provide the ability for you to separate cold storage, signing, and broadcasting. And that's like a critical kind of thing. Um and there's other kinds of things like how can you do staking, for example, in a way that it's not all with a hot wallet and what have you. So, there's aspects of potentially security where we could publish API specs and have people code to them.

So, we've already started to receive a lot of questions from the audience. So, if you have any other other questions, we'll keep uh ask them on Twitter with the hashtag ETHIS NYC, um and we'll kind of roll them up in a few minutes. But, I'll ask one that I just received, uh which is kind of for both of you. Uh what do you wish people in this space would stop doing that you constantly see happening a lot? Uh just stop fighting online.

Build stuff, you know. I mean, I would edit that slightly to uh fight online constructively, but [laughter] I mean, I do think disagreement is is important, and proposals need to be criticized, and like I need to be criticized, and Vlad Vlad needs to be criticized, and like and like every single research every single idea, regardless of uh the sort of of who it comes from, you know, it needs to kind of pass the trial by fire because ultimately it will pass the trial by fire by people who are much more malevolent than uh you know, actual trolls inside the Ethereum community. Yeah, that makes sense. Um kind of branching off from one of your previous answers, should everything be open source in this space? How do you kind of think about what does it mean to uh have like the IP kind of advantage to what should be private versus public, and kind of how do you think through that, or how should one think through that in this space?

Well, you know, a possibly perhaps trivial but useful, you know, limit case is your private keys are not open source, right? So, obviously not everything is going to be open source. You're going to keep something private. And then there's a continuum there between, okay, what is as valuable, what are the crown jewels like your private keys, and you know, what what can you put out in public? And one of the awesome things about blockchains is they offer a new way to monetize open source things of all kinds.

I think we're going to see blockchain versions of many other kinds of protocols. We've got already blockchain competitors for DNS, but we'll see blockchain competitors for SMTP and, you know, other other kinds of protocols coming out there. I'm already seeing them. And so, I think a larger percentage of open source projects can be monetized. So, we'll see a larger percentage of things that will be open source for their blockchain component.

Mhm. I mean, I would say the general principle is that it's more it's relatively speaking more important to open source things that are closer to the bottom layer and closer to the center, and less important to open source things that are kind of closer to the edges and periphery, and particularly kind of not protocols, but software that participates in protocols. So, one trivial example of something that I that I where I think the interesting bits will be proprietary pretty much forever is software involved in like market making, arbitrage, and generally participating in financial markets. But, you know, the financial market itself being a fully open source decentralized exchange is something that's absolutely that I think is great, right? So, like that that is just one example of, you know, the where the core periphery boundary is one is one good way of thinking about it.

One of the things that's interesting is in this space, you know, people will say, oh, you could just take a coin, you could take a chain, you could take a smart contract, and you can just fork it endlessly, and where does the scarcity come in? And, you know, related to the concept of, you know, do we open source everything? One thing I think is implicit is a lot of value derives from the backlink structure. Like, the reason that with the BCH BTC split, um that everything didn't just go to BCH in the same way as to BTC is a lot of things were pointing at, you know, the the longer chain, right? And so, the backlink structure I think is something that is a form of scarcity we don't normally think about that is distinct from whether the source code is out there in public is whether things are actually pointing to and trusting that particular database or that particular, you know, source code or uh instantiation.

Okay. I'm also going to start merging in questions that are coming in from the audience, so we'll kind of just have a mix of both versus doing a a separate segment on just audience questions. Another interesting question that came up is kind of what do you think are some of the cool things, anything from technological to community, uh that's happening in other blockchains that Ethereum should adopt? Um on the uh on the community uh on the community side, I I one one of the things I love about Zcash is this kind of separation between Zcash Foundation and Zcash and Zcash Co. And where, you know, like Zooko is part of one and not the other, and the end and they have a And I I think that they've actually done a kind of quite good job of just setting up a community in a way in a in a way where there is a kind of limited level of influence and kind of both one kind of person and one organization.

And another one of the interesting ideas that I think Zooko came up with is this idea that control, like practical control over protocol development and control over trademark ownership should be maximally separate. And that's I think and I an an interesting idea that probably in Ethereum and probably other public blockchains should consider adopting because, you know, unfortunately, there isn't really a good way to kind of have a trademark not be owned by anyone, but this does seem like a kind of pragmatic pragmatic approach that does kind of minimize the centralization harm that can come out of that issue particularly. Balaji, I guess to add on to that, what are some of the technological advancements that you're tracking personally outside of what's going on at Coinbase? Um well, certainly we're looking at all the privacy coin related stuff, you know, ZK-Snarks, Starks, bullet proofs. Um we are, you know, I I I can't talk about everything we're looking at, but, you know, we're obviously aware of um you know, stablecoins, we're aware of uh you know, what's going on with um like things like token curated registries.

You know, we we we pay attention to a lot of the stuff, but the thing is Coinbase's role is sort of to mainstream blockchain, which means that by necessity we're not on the cutting edge, right? Like if something is truly on the cutting edge, if it's on the bleeding edge, it's not really ready for tens of millions of users typically. So, we try to, you know, do what Apple does, which is take a technology that's in its spring, and it's proven itself a bit, and it's not going to have like a like a CVE on day one, and when it's ready for getting to tens of millions of users, then we'll mainstream it. So, those are the kinds of things that we're looking at. This is a really good segue into our next part, which is we're we're going to steal the underrated overrated segment from uh Teller account's podcast.

So, I'll uh I'll name a handful of topics uh and and the way you kind of play this is uh you have to say is it underrated or overrated and a quick answer on why. So, we'll start off with uh ZK-Snarks. I'd say about correctly rated at this point. Like they are power they are very powerful and uh like general I think general purpose private uh privacy uh general purpose zero knowledge proofs is like a really huge advancement in cryptography, right? Because if you look I was barely following cryptography maybe 10 to 15 years ago back when zero knowledge proofs were this kind of application specific protocol and you have to come up with specific protocols for specific use cases and it was like super lame and now it's I mean the fact that we that we're fairly close to the point where you can make privacy preserving protocols for just very large classes of things involving just through abstract programming languages is huge and I do think that, you know, the community recognizes that.

Underrated or overrated? ZK-Snarks or STARKs? Well, I mean I think ZK-Snarks are still underrated. People in this audience know what they are, but after Sapling you know, comes out in a in a few weeks, you know, I think it just gets mainstream to a much greater extent and we we haven't even yet explored where it's going to get. And then with STARKs and Bulletproofs I think, you know, it's going to go to the next level because you know, we eliminate the trusted setup and we have, you know, even faster you know, transactions and possibly smaller, you know, proof sizes.

So I think all those technologies are underrated relative to what the general population knows about them. Non-fungible tokens, underrated or overrated? I on the margin I'd say overrated. It's um I don't know. I'm like suspicious of things that get too buzzworthy.

So I throw token curated registries in the same bucket. I guess we'll skip the next question then. Yeah, I mean I I'm an investor in CryptoKitties. I I do think that in the long term NFTs will be important. I think like today if you go and look at DappRadar, there's not a lot of volume in you know, buying and selling you know, collectibles and NFTs.

Um but I do think that as they start getting hooked into video games and they're the back ends for things like their potions and swords within video games. I think that's going to be the thing that pulls them up in the medium term. So overrated maybe today but underrated in the long term and eventually we'll get you know what people have always talked about blockchain real estate and so on. It'll take a while. What about stable coins?

Globally speaking I'd say underrated. I still think people and a lot of people don't realize the extent to which stable coins solve one of one of people's biggest concerns with the with the cryptocurrency space which yes that in order to use these applications you have to expose themselves to volatility and with the stable coins a lot of that goes away and you can actually use applications that just look like regular applications involving dollars and I'm also you know very happy that very recently it seems like there was this big outburst of kind of new stable coin companies coming along like there's there's circles announcements there's I mean like true USD there's like and crypto DT over and there's I I forget there's like a there's a bunch of them now right and that mean like first of all and it's nice to see if there's more it's nice to see if these new ones are kind of more you know more reputable and audited and so forth than the ones that came before and it's and I'm also very happy to see that you know maker down and die is kind of doing well and continuing to prove itself and holding its value despite you know like overseeing each dropping in price by more than 80% I think it's still a very impressive accomplishment for a crypto back stable coin to stay to stay stable through that period. So and in the long run you know this these kinds of tokens are a way kind of way for traditional financial systems to kind of upload themselves to crypto land if they want to. Yeah. I I also I I would say stable coins are underrated today for a couple of reasons.

But with that said, I distinguish between two kinds of stable coins at least. There's multiple kinds, but I'd say there's one class of stable coins that I'm somewhat bearish on, and those I've considered the like Rube Goldberg machines, smart contract things that try to manufacture a stable result out of a fundamentally unstable underlying digital currency, I would argue. And then by contrast, there are the ones which I'd call, let's say, audit heavy, tech light stable coins, where really what they're selling is trust, the trust that you can go and redeem it for good old-fashioned US dollars or or whatever fiat currency underneath. Um and those I'm more bullish on because I think those are actually what people really want with a stable coin is they want the trust of redemption, um which means stable coins are more like an insurance business than like other kinds of crypto businesses cuz it's a question of what is the risk that I won't be able to redeem something. If you can solve those problems, and if you can have a stable coin that people really trust in that doesn't have redemption risk, then I think in the medium term we'll think of, you know, like FinTech before stable coin is like mobile before iPhone in the sense that every FinTech company, if you could just have a digital dollar, a blockchain-based dollar, if you didn't need to do a banking partnership or something like that, you just go and program right away with the digital dollar based on the blockchain.

Um and I think that's a really important medium-term thing that could happen if we get a working and functional stable coin. I think one important point in favor of crypto back to back to Rube Goldberg machine stable coins is that they're more likely to survive an unfriendly regulatory and banking environment. Yeah, so that's a So that's a good question, and I think if you're trying to fundamentally represent a US dollar, I think you can't, you know, avoid the question of interfacing with the US banking and regulatory system. It's kind of like this, if if you had an app that got popular enough on Facebook's platform, eventually Facebook will knock on your door, right? And if you have something which is representing hundreds of billions of dollars, eventually, you know, there will be some nexus to the US government and so on.

It's just like it's kind of inevitable. So, that that's kind of a a fundamental premise and and we can disagree on that. Kind of on that note, how do you how do you look at decentralized organizations? Overrated or underrated? Um I would probably say somewhat overrated in the short especially in the short term.

And part of the I will this is definitely a change in opinion. I was kind of more idealistic about this subset of the space maybe three or four years ago. I think the problem is that like first of all, if you try to create complete decentralized organization organizations, there's just game theoretic issues and misalignments in pretty much every kind of governance that you can come up with. But also I think a lot of the value it does that doesn't necessarily needs to come from kind of fully fledged, you know, like decentralized corporate corporations in the sky running on the blockchain. It can come from incentive alignments run that come out of kind of smaller and more purpose specific gadgets.

So, like the kind of DAIKO that that I came up with at the end of last year is probably one example of that. I mean things like assurance contracts, things like different kinds of like public good fund funding mechanisms, different kinds of you know, like quadratic voting gadgets or whatever. I think kind of more self-contained things like that are where I mean you can kind of see more kind of stable and realistic value in the short term. Yeah, so I am I'm bullish on DOs, very bearish on DAOs. Like like I don't believe in the autonomous aspect for a while.

Um and why bullish on DOs? Well, I think you know, we have more and more teams that are remote work-based in multiple jurisdictions and I think in the not too distant future quote incorporating or signing contracts on the Ethereum blockchain or other blockchains like that, smart contract-based blockchains, are going to is going to become pretty popular for remote teams and you can imagine like a blockchain-based Slack or a blockchain DocuSign a BlockSign that just allows these decentralized organizations to agree on something, vote, and then allocate funds, you know, with a proof of stake vote or some other kind of vote, just even a multi-stake kind of vote. And I think that's kind of cool where you don't necessarily need to take the entire DAO concept, you know, you can take pieces of it and just be like, oh, here's how, you know, we get multi-sig, you know, for three people who live in three different countries under three different legal systems. We don't have to have them incorporate in Delaware, we can have something on the blockchain that's pretty lightweight. I think I'm bullish on that kind of thing.

We'll do like one last one on overrated underrated. Off-chain scaling, what what are your thoughts on that? Ah, so I'm actually, you know, [clears throat] I really believe off-chain is very underrated in the sense of for most applications, unless you're unless you're Vitalik, unless you're Zuko, unless you're somebody who's a cryptographer or, you know, a very, you know, knows systems programming and so on. Most people aren't going to be able to do that and instead they should keep most of their state in a normal Postgres database and have blockchain functionality, right? So, you know, blockchain like the blockchain will provide balances, it will provide addresses, your users can send and receive funds, and you can keep your images and your normal user data or whatever in Postgres like a normal kind of thing.

And so then your app has two data stores. It's got Postgres and it's got, you know, the blockchain, you know, over here. And I think that kind of model is going to be like a pretty important one, a hybrid model. And moreover, anything that you can do off-chain, you should do off-chain because the you know, blockchains are relatively low capacity right now. You want to batch transactions, you don't want to pay fees, uh you want to try to do as much as you can off-chain and then just write the absolute minimal digest on-chain.

So, I think that's going to be a bigger and bigger thing. One thing I wrote about 2 years ago that I'd love to see, maybe somebody here can hack it tonight, is um something similar to PostGIS. Right? So, PostGIS is basically a set of add-ons for Postgres that adds, you know, range queries, XY, you know, latitude longitude support to Postgres, which is normally just a standard relational database. I'd love to see a post-chain that does the same kind of thing but adds blockchain functionality.

So, it makes it easy to ingest a blockchain, do relational queries, uh you know, to to search and and sort blockchain-based data structures, um and and an add-on like that. And that's the kind of thing I think would help a lot where you're doing a lot of stuff off-chain and then you have a digest that is written to the blockchain and it costs you something. I think that's going to be an important paradigm. Uh we'll move on to sort of the next couple of questions we've gotten in. Uh Vitalik, what role has uh threshold relay or and other improvements from the Dfinity ecosystem have played into the design of Ethereum 2.

0? So, we're not using threshold relays specifically and both and both because we prefer we prefer being a leaner on on cryptographic assumptions and also uh because it's well, because we Well, okay, just to kind of double back a bit, right? What threshold relay is is this the kind of in-block Well, it's two things. One is it's a consensus algorithm and two it's a a in-blockchain random number generator. And it relies pretty heavily on kind of BLS signatures and distributed key generation.

And I personally dislike distributed key generation because it's like big extra load of craft that you have to add into a the add into the protocol spec somewhere. And we're taking a somewhat different approach where we where we use BLS signature aggregation, which also uses BLS signatures and also gets kind of very high efficiencies from a combining signatures together in the same way as threshold relay does. Um but we use this in combination with a different kind of a proof of stake algorithm that uses a ghost based fortress rule. And I mean this is nice and with the goal that we're targeting is actually we do not want the security of the blockchain to be fully dependent on the security of a random number generator algorithm, right? Like we we want to be able to prove that the blockchain is secure even if the attacker chooses what all the random numbers are.

And surprisingly enough, you know, with you in a blockchain algorithm where you have enough validators participating at the same at the same time, you can actually do that. Um then for random number generator, you know, we have Justin's entire line into VDF research and while that's technically not necessary for the blockchain to work, that does kind of increase the security margin of the system. So it decreases the the minimum safe committee size for example by maybe a factor of two or three in practice and it increases the safety tolerances by a few percent and at the same time a random numbers are something that's useful to a wide array of layer two applications. And so that it like we are you building in a random beacon in chain, but we are using verifiable delay functions instead of verifiable random functions. So it's somewhat different technology.

Balaji, a question for you from the audience. Um what geographies will we first see a mass adoption in terms of a use case for for cryptocurrencies and and sort of why is that the case Or will that be the case? That's a great question. Um, you know, there's there is a thesis that it's going to be like the Venezuela's of the world and so on where cryptocurrency has mass adoption. Um, I actually, you know, as sympathetic as I am to the, you know, to the Venezuelan people and as much as I do believe that that does illustrate an important use case for Bitcoin.

Um, I do think that cryptocurrency is probably going to have to prove itself in the developed world first and then it will filter out to everywhere else. And the reason for that is um, any technology in the early stages has bugs and horrible UX and various shortcomings that only relatively wealthy people who have the time and the energy and don't have a chaotic situation around them and they have good internet and they can download a multi-gigabyte blockchain and they don't have a light wallet, all that type of stuff. Only folks who have all of that, you know, time and internet, um, I think are going to get this thing off the ground and then eventually get it to hundreds of millions and then get to billions. And I think it'll have a similar emergence mode to smartphones where the developed world subsidized them and got them to scale and then they were cheap enough for the developing world to buy into them. So, I'm a huge long-term bull on like banking the unbanked with crypto and so on.

I do believe that eventually does happen. I just don't think that's where it starts. Yeah, and I'd probably generally agree. I I'd say while we're in the enthusiast phase, it'll continue to be the case that the map of where the blockchain nodes are will roughly track the map of where the the brightest night lights are if you look if you look at the the earth from space. But in in the long term, I definitely do and I also agree that you know, like banking the unbanked, kind of improving um, financial systems in cases where kind of inner cross-national kind of cooperation uh, um, is currently weakest and where existing governance structures are weakest and most hostile is one of the is one of the most interesting areas where they can be applied as well.

So the interest of time, we'll just ask two more questions. One just to kind of come back to something that was said earlier. Coinbase was involved with givecrypto.org and the Ethereum Foundation is involved with a grant program. How can others get involved with these initiatives?

Yeah, so we have a charitable arm called givecrypto.org. So you know, givecrypto is just spelled like that. That you know, Brian set up and you know, that does charitable work in the space. So for the Ethereum Foundation grant program, I mean there's still ways to participate, right?

One is to apply for a grant and the other is if you have a lot of money then you yourself can just become a grant giving body and you can either decide whom to give grants to yourself or you can kind of piggyback onto you know, who the Ethereum Foundation gives money yourself because those lists are completely public. So in either of either of those two is highly welcome and you know, we're on the project side we're actively looking for projects. So doing work in scalability, particularly usability including what I call usability of security for you know, and that includes both the wallet and the kind of smart contract programming side. Including you know, things like formal verification. It includes kind of it includes layer two systems of pretty much every kind.

So it's if you are one to build a piece of the Ethereum ecosystem, the opportunity and kind of do that as a full-time thing, the opportunity is there for you. And lastly, Balaji already answered what he would like to hack on potentially this weekend. What would you make if you were a hacker at this event? Well, Fomo 3D was pretty awesome. So you know, something like that I think really is you know, it's just is something you could only do on the blockchain, you know.

Uh so I'd love to see stuff like that um which I think push the limits of what you can do with smart contracts. Uh and uh you know, I'd love to see something like post-chain what I what I just talked about. I think that'd be very useful. Um and uh yeah, I mean those two in particular I think would be cool. Uh reputation systems are probably still kind of un an interesting area that we probably haven't seen.

There was a bit of a lot of enthusiasm back in 2014, but then it kind of petered off and I think it we might be at the point where the infrastructure is there to kind of make another go at that whole area. Um I would I also say kind of in privacy-preserving applications are in general are an interesting category. Um Anything kind of develop any kind of developer middleware um kind of in particular kind of more possibly decentralized or alternatives to existing kind of more centralized forms of middleware. Um solutions in wallets and kind of and securities um solutions in other important area. Like basically what like and also in general just look at kind of what what what you think are pain points that are seen by kind of developers in the ecosystem and try to have a go at solving any of them.

Perfect. Well, Balaji, Vitalik, thank you so much for taking out the time to share your views with us tonight and wish all of them the best of luck in that. [applause]