 So hello, I'm Amy Caster. I'm an independent reporter and we are here in Lisbon where IOHK is having its annual company summit And I am joined here today by Robert Vaglione co-founder of Zencash and Charles Hoskinson who is the CEO of IOHK and the reason we have them both here together is because Zencash and IOHK have entered into a research partnership and We're going to learn a little bit more about what that is it involves a Treasury system and Implementing what's called a directed acyclic graph protocol so Charles you start by telling us little boys at a partnership. Is it a research partnership? What exactly is the relationship that IOHK has was then cash, right? So it's important to understand the history and how we got to where we're at. So it started Last year when I had a chance to meet Rob and his wife and Rolf and many others on the Zencash team And we really really enjoyed talking to them working with them Basically just answering questions such as you know, how do you grow a cryptocurrency? How do you grow community or should the science go? Where should the development go and so forth? So they invited me to be an advisor And I of course accepted because I thought it would be awesome to collaborate with them It's a pretty good decision. Yeah, it was it was a lot of fun And actually we've already had a heck of a lot of fun and Rob has gone all around the world chasing us about it I think Ukraine and not here in Portugal. I'll take what I can get There we go. So that was a that was fun So then then we started opening some discussions about going beyond an advisory role and Helping Zencash figure out where it wants to be as a as a cryptocurrency in particular on the research side One of our core competencies as an organization is our ability to to speck out from business requirements or real-life problems Where does the science need to be to actually verify that these things get solved? So there are things like how do you make Zencash a currency that has a very high throughput and it could actually be used as a Cash, you know can actually be used for 10,000 20,000 transactions per second over an arc of time and be useful in In the browser setting and other things like that There are things like how do we enhance and grow the privacy guarantees upon which Zencash is constructed And also how do we make Zencash more sustainable as a protocol? So these are things like treasury components and so forth and these are areas that I which K as a company is is quite interested in And we have a research agenda for But we as a company don't pursue intellectual property We don't pursue patents and most of our work is done within the academic circles So these are papers so there's no reason why this can't be shared or in some cases custom Localized for particular projects like Zencash So we were really keen to collaborate. The other thing is that Zencash already has a treasury system It's a system that needs to be upgraded and evolved over time But we we wanted to do a deal where we instead of working with a company or a person We're working with a cryptocurrency with a blockchain Right, so our first client liked that and we we hope to have many many more So that was kind of the the spirit of the initial agreement. So we we negotiated something out and we're just getting started So so add to that or so so so part of the partnership is implementing a treasury system, right? Right, and and I understand that they'll be using the the same version of the protocol that's been developed by OHK's researcher Ben Cheng, right. Well that yeah, so so he's building out the theory So he's saying what does voting need to look like and what are some of the security properties you need to hold and what are some of the Game theoretic things that you have to worry about, right? So that's kind of like the model then once you have that model There's a question of what does that model mean for Cardano? What does that model mean for aetherium classic? What does that model mean for Zencash? And that's part of what we have to figure out once the model is constructed because it's going to be parameterized So let me ask Rob. I mean, what does that entail? You're going to be implementing a prototype of The this treasury implementation in the first quarter second quarter of 2018. Yes So the prototype we think will be will be ready by probably the end of second quarter If everything goes according to plan in terms of implementing that that will be the next step So we'll receive prototype and we'll make sure that expect, you know Well for our system and then we'll move it to Implementation and then production after that. Okay, so I mean like When do you think we'd have like a fully implementing implemented treasury system on Zencash? And yes, that's a really good question. I mean, yeah, this is purely, you know pre You know, I just kind of early thought that I would say probably by the end of the year So so will this be like the first? I mean there are other treasury systems. We know with like Dash is implementing a treasury system to credit some kind of treasury system where there's actual voting and and taxing the Should we've explained what a treasury system is briefly, you know, we can explain that sure Well, just like briefly that it is both both it involves Voting at first there's the taxation of the system where they take about a 10% of the mining reward or yeah We can get into that. So you have three sources. So first off What is the point of a treasury system is the long-term goal is permanent sustainability? So imagine a cryptocurrency like a government and a government is a collection of departments and each department does something So when Bitcoin was established, you had one compensated department and everything else was volunteer So for consensus, that's the mining side totally paid for a great But for moving network messages for doing development for doing marketing for setting up companies for evangelizing all these things Completely volunteer and not compensated. So people like Dash bit shares Decred pivots and cash others have adopted the philosophy that maybe Some of these other buckets need some fair share funding as well from wherever this funding comes from So the next question is where does that come from and then who gets to decide? Those are the two other questions. So you have three ways you can fill that bucket one way is through inflation So you just print more money. That's what Bitcoin does to pay miners, you know block rewards So it started at 50 and then it went to 25 and now it's 12 and a half and it decreases by a factor of two every four years Another way is to take a proportion of transaction fees So instead of taking all the transaction fee and giving it to the miner You take a percentage of it you put it into that bucket or in other ways donations where people just put money in Problems donations as well. They're the purest form and they're the most consensual the issue is that they tend to be inconsistent Unreliable problem with transaction fees is that they tend to be not enough to sustain the system unless you have a huge Scale system with tons of stuff going on. So generally inflation is the predominant model So what does that look like in practice? Usually what you do is you say okay instead of giving 20 coins or 30 coins per block to the miner. We're gonna give 16 coins or 24 coins the miner We're gonna take that small sliver that we don't and put it into that bucket Okay, but this one I mean by parameterization. So we when we construct out this model We we say okay, here's how the voting works Here's how the treasury ballot system works But those parameters of what where the the filling of that bucket comes from is tunable And that's a decision Zen cash gets to make and Ethereum can make and Cardano can make and so that's interesting Why wouldn't you why wouldn't Zen cash go ahead and implement something a little bit simpler? Why are you taking this, you know broad-spec to do something? Yeah, no, that's exactly Why are you taking the step up then to liquid democracy? Yeah Oh, we think that that's just the future and that's the way these systems should go because the current models Which is very inefficient, you know, so we want to constantly improve and we just had this empirical mentality of you know Kind of put something on the mark out there, you know Test it see get feedback from the market from the system see how they're performing and then try to improve them And constantly improve them. So what we found here when we first read That IOHK white paper for the treasury system blown away because this is just a very elegant implementation of our core technology with Zero-knowledge cryptography and a game theoretic framework that improves and optimizes voting systems So this may not be the final system that you know We have as an industry, but this is a huge step up kind of like a step function jump up in technology You know and mingling technology with economics and I think in a way that will be very impressive for the market And for IOHK it gives you a chance to kind of test out this protocol in a real-world Situation the first time it'll be used in a cryptocurrency Yeah, well we plan on putting it in Cardano as well And our hope is to turn us into the IOHK reference treasury model So we're going to put it into many different cryptocurrencies ideally. I look at these things like the mechanisms of governance I don't think that should be a competitive differentiator for cryptocurrency I don't think we should look at sell well, you know Dash has this or Decret has this of course There's going to be differences But ultimately they should be using best available tools and then it should be other factors that Determined whether the cryptocurrency successful or not. So in my view, I'd like to build the best governance model I can think of with the best people and get it in as many crypto currencies as possible Because that overall increases your overall amount of funding available for research for development and for growth So does that does that mean this is going to be like an open-source protocol this Treasury system that any of the Cryptocurrencies can go and just kind of grab out of a bucket right put it into their Exactly, and that's what that's what we've done with prior projects For example, we worked on the Scorax project for I think now more than two years and we continue our commitment And Scorax is a modular open-source framework for building crypto currencies And it's already been used in several projects the most notable of which is waves where they forked it And they put their own magic into it and now it's something that's floating around the space So similarly, I'd love to do the same thing is to create a treasury framework And then other people can take that and build on top of it. That's a really interesting concept I don't know that I've heard anything quite like that before But but so so basically all the cryptocurrencies all these projects should be kind of keeping their eye on Zencast to seeing how this works In case they may want to implement a Treasury system right future and it sounds like it would make so much more sense to It's it's open source. Why not take this one and use that rather than starting from scratch and doing your own, right? It's battle-tested and it also it's Research this is the first battle Yeah, but it takes transparency and just kind of like good ethics to the next level in this industry And I think that's something we really have to start self-policing and when you have a system that kind of Dissociates, you know the funding allocations from the decisions of specific people So it makes these systems just much more fair on its transparent And then you have to wonder why would other projects not adopt them, right? Right because then this is almost kind of like a signaling equilibrium in the market where you can see projects are taking this you know Hyper transparency route with things being kind of more Systematized and fair and then you have other projects that remain opaque and kind of have discretionary authority over funding allocations You wonder why well Why wouldn't a project want to do something like a Treasury system because it's so difficult to get? You know VC funding yours worried about whether VC is gonna have some can say something to say over the project and Ico's well, we've seen what kind of problems those have and nobody knows what's gonna happen the regulations and if you get donations Well, you just can't really count on that. You know, right? It's like public radio Practical reasons why you wouldn't want to do that a a because First maybe you don't believe the technology is there or be you can't figure out how to balance economic incentives The biggest thing in all of these types of governance models is that voting is difficult democracy is difficult It's you have to have an incentive to participate, right? There's this idea called rational ignorance It's well understood in the economic circles where the cost of becoming informed on something is More than the value you gain from being informed. So the rational action is to stay ignorant on that issue So if you look at the United States, for example, we have these really complicated things like Medicare and social security and the federal budget and informed policy And and you can literally go get a PhD and spend your entire career studying just one of those things more accurately one dimension of one of those things because they're multi-dimensional and The vote you get is identical to the crazy hobo on the side of the street that things aliens are coming tomorrow Right. So when you have a system that incentivizes at that level outside of altruism There is no reason to actually invest a reasonable amount of time become informed these types of things So what ends up happening is people tend to vote for irrational reasons They tend to not vote at all and so participation tends to go down and quality of participation tends to go down So so basically and I've been reading about this a bit too And I think I think this is part of what you're saying that the two biggest challenges in implementing a voting system or one getting Incentive people to vote right they have to have some kind of desire or or get paid or something back for that Oh, there's all kinds of things. It's not just about who gets to vote It's also about what do you get to vote on an example? I love to give a fact that we were a nexus comp together an aspen and I did a presentation there And I said well how many people here voted for either Hillary or Trump and some people raised their hands And I said how many people here wish they had more choices than just Hillary Trump and everybody raised Regardless of the political affiliation. So it's less about Who gets to vote and also you have to consider What do you get to vote on right like how do you actually get on the ballot? Who gets to decide what gets on the ballot and what doesn't get on the ballot? Because that that in itself has a profound impact on the outcome of the election So you need to think about it sentence very carefully the voting class very carefully the balloting process very carefully and Time of deliberation very carefully as well the problem with a lot of voting systems is that we just say go make a decision Here you go, but in some cases it needs to be a very slow deliberate system Well, let me ask let me ask let me ask you all this what kind of things are you going to be voting on? Yeah, so I mean We do have a path in mind for ultimately like even goes back to the question of why are we doing this? All right for because there's always this trade-off between centralization and decentralization, right? if you if you had God making all decisions and This was an omniscient, you know Omnipotent entity that could just do everything perfectly you would want that right, but we don't have that we have human beings Who are imperfect so we decentralized and for us it's extremely important to decentralize because we want to be censorship resistant, right? So that's a whole part of the theme of Zencash is being censorship resistance and and we didn't talk about earlier, but you have this entire this messaging system where You know people can send secure private messages back and forth So we don't have situations like what we just saw in Iran where they they they blocked telegram, right? Because of peace of peaceful protests. That's why we're willing to take the resource You know it this is resource intensive to do this kind of thing And we have trade-offs and we know that we're getting some inefficiencies But we're making what we think is a more open fair system this censorship resistant And we we know that there's some wisdom in crowds and there's also some irrationality in crowds, right? Like Charles just mentioned so we're trying to create a balanced system and we're doing this incrementally So initially what we're doing like I view governance in two almost two stages You've got resource allocation, and then you've got parameter adjustments, right? Who determines the next hard fork and what goes into it? We're tackling the easier of those two problems right now, which is the funding allocation So that's that's what we're doing where we have a treasury funds for governance And we're just determining a better way to implement like so So let's touch on the second aspect of this research partnership between Zencash and IWHK And that is is which is really interesting to me is because I haven't heard a lot about this But you're implementing the the specter protocol. Yeah, well, that's a protocol. We might implement It depends upon at the end of the day the research development the Zencash team's opinion the community's opinion But that's the probably the best of the group that we've seen to start with and a really rigorous inquiry So why do we want to implement something this or what's the need? so one of the easiest ways to get higher throughput and performance with the proof of work system is to take your block interval and Reduce the block interval that means you make more blocks in a shorter period of time Which means you have more throughput you have more transactions per second. So why didn't Satoshi do this? I mean created Bitcoin and he set the original block interval of 10 minutes. So why didn't he choose four minutes? There are one minute or 30 seconds Well, the problem is that the smaller and smaller and smaller and smaller that you make that block interval the higher the probability You have what's called an orphan block. So what happens is Bob who's a minor will discover a block But right around the same time Alice who's a minor will discover her own block And so this group of the network because they saw that block first think Alice's block is right and this group of the Network because they saw the block first thing Bob's block is right So now you have fork you have two versions of the chain and we're not quite sure which version of history is right So how Satoshi resolves this is he said ah, okay? We're gonna have a competition and whichever one of these forks gets a block built on top of it That becomes the legitimate chain and the other one is discard. It's called an orphan block So that block goes away. Okay. Well, here's the problem that Reduces the efficiency of your system because you now have all these wasted resources on this thread Everybody who worked there. So you've gotten less security amongst others theoretical and practical problems So what if you could build a protocol where it's okay to have Lots of these things and they get included in the blockchain And they have some sort of relationship to what ends up being the longest chain But it doesn't throw things down. It doesn't damage throughput or security So this was a line of research that started in Israel under two people obvious Ohara and Yonatan some of And the first paper was called ghost and it came out 2013 and actually we included ghost version of it in the Ethereum protocol So later on what they did is they built up the theory They built up the rigor and they created a new version of the protocol called specter which came out in 2016 And they've been since iterating and refining and thinking carefully about the model And there's likely going to be several commercial products that end up putting Are there any cryptocurrencies right now that are implementing the the now this is a Dicrylic I never directed a cyclic graph. We'll just call it dag Yeah, but but and there are different there are different versions of that, right? I mean, I think there's hashgraph and I don't know and biota bite ball, right? But they're not implementing the specter. No, that's a very different protocol, right? So the one that you're working on like what exactly is IOHK doing in terms of the specter protocol like are you just seeing the feasibility of implementing it? So there's a collection of things that have to be done So the first thing that we have to do is read the paper very carefully. Yeah, it's not easy paper There's 40 pages of math products, right? It's not a that's not an easy paper. So then so then we have to extract from that paper some form of an engineering Specification and then what we have to do is a trade-off analysis and say if the Zencash engineers were to implement this This is what you're giving up and this is what you're gaining These are the security guarantees that you can reasonably have these are the performance enhancements You're going to get etc etc. So that trade-off profile has to be generated and specification has to be generated It's understandable by I mean, this is like a big deal a lot of people understand like somebody writes a paper and they've got an idea And they've got these mathematical Specifications for but turning that into an actual piece of software is you can implement is is a gigantic step Especially if it's a very complex paper that hasn't been simplified a lot if you look at the Bitcoin white paper It's very small. Yeah, very concise and it's a very simple idea. Yeah specter is like this Super complex and really involve really deep and detailed paper that was written from the perspective of a mathematician Not from an engineer. So it's mathematically elegant It has a high degree of confidence that there's some correctness there But it's not something you can hand to a programmer and say I mean, this is what I HKS doing you're taking this 40-page paper and you're turning into something That is like specified. Yeah, that's why you're actually gonna specify it and formally verify it not a formal verification That's a bit beyond the initial scope The first step is just let's write a semi formal specification that an engineer can read that doesn't require gawking 40 pages of math proofs And and then what Rob gets to do is go to his people and say does this make sense for us Or are there other DAG technologies that are maybe lighter weight or have better trade-off profiles than specter? And then we can do that same task for them and then eventually they can pick one and we can help So that's the part I'm not sure about so you're you're developing the protocol that that Cryptocurrency your project can implement but but what so what role does Zen cash play they write the code Well, you're actually writing the right and we need to see if this protocol makes sense for us I mean, it's extremely fascinating. I would say this is the most fascinating protocol out there So so I HKS specifying it, right? And then they pass it on to you and you are using cash is actually writing the code, right? See I didn't understand that before right. That's pretty cool Yeah, we've done our job really really well That should be pretty straightforward for that to know how to do that kind of implementation Also, there's gonna be something you have to give up when you go from theory to practical And there's a well big back-and-forth discussion about what's reasonable to give up just like the parameterization of that Treasury model We can give them a framework We can tell them all the theory but ultimately the Zen cash team has to make a decision along with their community of What are those parameters going to be and what that profile is going to be? It's it's not our decision to make Okay, got it so so but I mean, what do you get out of? Putting all that work into you know creating a dag I mean is that something that you're really serious about augmenting tens of thousands of transactions per second low latency Confirmation times It's basically like an order of magnitude performance. Maybe several yeah, sir. I'm saying a performance improvement Yeah, I mean this takes us we also we call ourselves and cash and across the industry We talk about being used as an alternative currencies, but the reality is most of these systems don't scale very well They don't have high transaction throughput there are issues, right? Right, even even now we have a fairly active chain We have a good amount of transactions already on there for a project that's seven months old We're already surpassing a lot of more mature projects We're already starting to kind of bump up against some every now and then we hit blocks They're like wow this is kind of slow kind of packed, you know, right? We know we have to make this like two three order magnitude jump and an ability if we're gonna be taking seriously as a Transaction mechanism so just so people understand This is like this is like a big problem that all blockchains all cryptocurrencies face This is scalability issue and a lot of people are working on a lot of different ideas on how to fix that So this is this is kind of This is really interesting Oh the way that I see scalability right now in the industry is a lot of projects are looking at parameter changes Which are kind of like linear scaling solutions, right? This would be kind of like an exponential scaling linear or else they're doing things on a chain where they Take and then use the blockchain to settle on but now this is like a completely different concept right and the one thing that I still don't understand and this may be really too Technical to get into here, but there is a big difference between the specter protocol and The existing DAG protocol that's being implemented in the three other projects we mentioned Yeah, and if specter is kind of fine tuned for proof of work Iota and hashgraph is not So that's one of the inspector was made with an attempt to improve Bitcoin like systems in mind So when you reduce block interval, how do we get more blocks out of that? It's okay to have orphan blocks and we have paths of things Whereas Iota is like a completely different notion of how how this ought to work Which is less understood and not really rigorously reviewed as as much as what we've seen with proof of work So the hope is that we can reuse many of the security guarantees that proof of work has already provided But just get these orders of magnitude performance improvement. So just again like You you're still gonna have your your proof of work because we know that Zencash is based on Zcash Which is based on Bitcoin, so it uses that sort of basic Bitcoin Code and and and the the proof of work. Well equi hash Yeah, so this is a big thing for me is you know, like the big difference here like Charles said It's it's really tailored for proof of work system. So we still have blocks. It's just the blocks are arranged in a DAG You know versus transactions being arranged into a DAG, right? Yeah, and so and so that's nice because anytime you can reuse prior results or prior security Then you can usually drop in lots of cool theory there that has been vetted and tested when you're inventing new theory That's a much more ambitious much much Harder thing to do like for example, I would you keep with us or a Boris research agenda? We have invested now a year and a half into building this big lattice of theory At all these papers and we've been accepted at Euro Crypt and crypto and a CNS and financial crypto But we spent literally millions of dollars and we're just now starting to get to a point where we have practicality in our system Whereas when you look at things like proof of work, there's nine years of history here There's there's a lot of good ideas that have been vetted and tested and proven not only in the academic world But also in empirically in the in the cryptocurrency space as a whole so we say if we can just take something that's tested tried and true and Put something like specter in if we get that to work well Then it's the best of both worlds you get those performance improvements But then you can still rely upon prior results and other things exciting as proof of work actually has certain characteristics Which are very hard to replicate proof of stake. We think we have ideas to do that in the or Boris research agenda They'll take months to years to complete But right now we have papers for example our sidechains paper and on an rack of proof of work That's not only a good protocol for sidechains. That's a very good protocol for SPV clients So basically you can get like clients with equivalent levels of security or near equivalent levels of security as a full note Which is just currently not possible popular for impossible for any kind of system with this with Sticking to proof of work. It's pretty straightforward how to drop a system like that into a protocol Even if it had specter whereas if this was done with a proof of stake system It's like a original research thread. It'll take a year maybe two years for us to see if it's even possible to emulate So it's a lower risk more pragmatic way of approaching. This is some really cool stuff And I'll definitely be keeping an eye on Zen cash to see This is what I love work about working with IOH K is if you look at this industry right now Most projects are just kind of hobby shops to just kind of throw something out there and see they can see how it goes IOH K takes a scientific approach to actually doing rigorous peer-reviewed research and then implementing it Implementing incredible products. So this is something that the industry needs a lot more of so working with these guys There's no brainer for me. Yeah We need more of this, right? So this is really interesting. So you'll be working on DAGs specter protocol and Treasuries and what's the future any other collaborations between IHK and Zen cash? Yeah, so IOH K's Recently entering entered the snark space. And so we're very keen to do something interesting there We have a center. We're setting up at University of Edinburgh under one of our research fellows named Markov Cole wise wonderful guy very smart and The lab's goal is to look at snarks from the perspective of both verified computation as well as privacy So we're quite interested in doing things like making them more efficient Reducing the amount of size they have getting rid of trusted setups wherever that's possible these types of things So I see a very natural potential future collaboration there, but more abstractly You know, it's not just Rob and I or a roll for Rob and I or Zen cash and IOH cave And it's more of a community thing where the Zen cash community as a whole can say hey, you know The XYZ are really interesting to us. We'd love for you guys get involved in that We can potentially expand the relationship accordingly, which is what makes this so fun Is that that dynamicism and that back-and-forth that you get so it's not always top-down. It's more bottom-up right? Great. Well, we'll be looking forward to that CK snarks as well. Okay brilliant. Thank you