Rethink Trust 2018: Christopher Georgen, Beyond dApps, building an application protocol
Recording: Rethink Trust 2018: Christopher Georgen, Beyond dApps, building an application protocol
so thank you for that very warm and wonderful introduction Alexei again my name is Chris georgian and i am the founder and CEO of Topol so my talk today is really going to be in two parts first I'm going to talk about how we are using our own blockchain to accomplish a specific business process and then what I'm going to talk about really through that is how we went on an our engineering journey and where we ended up from a technical standpoint and that's going to be a weed into a workshop that our CTO and we blockchain developer are giving later this afternoon so the whole thing is kind of two parts first this talk and then this workshop later on so to get started we you know thirty thousand meters up what is topple topple is a blockchain that we built specifically to support investments moving into developing countries so it's an investment protocol an investment platform with various features baked into it but before I get any further I feel I probably need to convince you a little bit that we are qualified to be doing this so I'm Chris I'm a mathematician by training I've done some research in various areas of physics I'm joined today by my three co-founders Kim who's a statistician and economist Jim who's our CTO and just a phenomenal atomic physicist there's a long history of physicists becoming excellent developers as Alexey can tell you himself as well and then Nick are we blockchain programmer and there were guided by various advisors academic statisticians one of whom is sitting in the audience today from the Coffey's Institute at Rice individuals from the World Bank individuals from private equity all guiding us and telling us exactly how we're trying to or the best way to solve this problem that we're attempting to solve so that business process I was talking about again it's how do we move investment more transparently more efficiently into developing markets and we're doing this really for two reasons the first reason being social and the second reason being economic the social reason goes to what are known as the SDGs or sustainable development goals a set of 17 initiatives outlined by the UN back in 2015 saying how we all want the world to be a better place we want less poverty we want gender equality we want more clean energy and the list goes on as you could probably imagine these things don't come cheap they're expensive you know just change the world that's the social side and then the economic side is this economic opportunity that we have so if you look at emerging markets and say we want to increase the average income to maybe ten thousand euros a person still about a quarter of way it is here in the Netherlands that would be 32 trillion dollars in new economic activity each year so again there's a social problem and there's an economic opportunity and so this investment business process is what we're setting out to solve but it's a fair question to ask you know this problem exists this problem so big but why isn't it been solved yet now you might think that we're sitting in a blockchain conference I would say the reason it hasn't been solved yet is because the blockchain just came out it's not exactly that simple but that is part of it the other reason is it really goes to an issue of financing the reason there's still so much room for growth is because there's so much demand for money and there's not enough supply if you go to a bank to get a loan for you know a textile or manufacturing company and say Africa the bank will say fine I'd love to give you a loan but it's gonna cost you about 40% interest now I know there's a lot of engineers in the audience but sure some of you have ventured into starting your own businesses as well and I'm guessing you wouldn't have wanted to do that if the best money you could have gotten was alone for 40% interest so that's why this problem persists that's why there's still so much opportunity there and so this is exactly what I was talking about if bank financing won't work if there's not enough money in the banks to cover all these opportunities we need something else we need direct investment from anywhere in the world from any fund from any individual anywhere in the world to be able to get access to these opportunities and now you're probably starting to hear some things that would remind you why blockchain would come into play this idea of openness this idea of accessibility and that's exactly where the blockchain came in for us as I was saying our architecture focuses on this accessibility this transparency this how do we connect an opportunity halfway around the world with any investor who wants to get at it here in Europe or in North America and getting into our features a little bit that will become very very important when I give you a technical overview at the end of this talk what does this process actually look like so the first thing that you'll see up there are there's a lot of different actor types a lot of different roles that need to be fulfilled the first thing that's happening is different projects are coming to our blockchain and saying we want to whisk our opportunity so we call those producers and then those producers go to another actor type called a hub and all a hub is a local representative kind of a local intermediary on the ground that connects the producer to the blockchain and then once the hub has brought the producer onto the blockchain now there's risk analysis that goes on and that risk analysis is handled by another actor type that other actor type is known as an arbiter and what arbiters do is they use prediction markets so that's an aggregation of lots of different models lots of different projections to get one risk score for each opportunity so now we're at three actor types and then the investors come in and actually get you invests in that opportunity maybe it's a textile plant in South Africa maybe it's a farming cooperative in rural Colombia and so they can look at the opportunity that's been risk assessed and they can invest in it via a smart contract but then I said there was a social component to all this as well so now we have a few more actor types that we need to introduce first is the party on the ground that's certifying that some social impact was achieved that there was some socially positive outcome maybe poverty was reduced maybe this is a woman-owned business maybe megawatts of solar or wind energy are being produced we need to have some outcome that we need to certify and add to the blockchain and then what we can do is we can actually fully tokenize that outcome and then we can have a market for it so there's an element of exchange as well playing in so that's our business process so that was me as a CEO saying to our engineers this is what I need there's a problem and this is how I want you to solve it so then it was their task to get building and this is exactly what they've been doing over about the last year so this building was kind of a I would say a three step decision process the first one was do we need blockchain and why do we need a blockchain well if anyone's familiar with you know the first version of the blockchain Bitcoin this idea of transparently moving money between parties that don't trust each other that was Bitcoin you can probably see how this is applicable and what we're doing as well you know we have parties that don't know each other and they need to move value between each other so in addition to smart contracts and other things this idea of value transfers core to both the blockchain and what we're doing okay so Chuck we need a blockchain but here's where things get a bit worse for the blockchain as I hopefully made a parent we have a fairly complex feature set there's a lot of different actors and there are even things about identity management and supply chain integration they ain't get into but they're all necessary in this business process so anyone who's tried to write DAPs or decentralized applications in aetherium can probably tell you you know the second you go over a few hundred lines it becomes very very dangerous as our last speaker outlined for us it becomes very very expensive if you want to do it in a public setting and it becomes very fragile in a lot of ways because aetherium and iosing Cardno are still going through so many improvements themselves it's fairly likely that your smart contracts can be broken at the next major release so this complex feature set prevented us from simply saying we can adapt on top of aetherium so that told us we need to build our own blockchain so when you build your own blockchain usually your mind would go to a project like hyper let your fabric where I can customize things I want and I can really have good control over it but now this third thing is probably the most contentious point I can make at an enterprise walk chain conference so please understand we took the decision very very seriously when we said we wanted to go permissionless and now none of this is to say that it's a permissionless permissioned argument it's merely to say there are some business cases where you can build consortiums where actors trust each other where actors are all relatively similar in size somewhere in field and it makes sense for them to all come together sign contracts and trust that the other nodes are doing what they're supposed to be doing but our business process international investment isn't like that you have investors you have people on the ground you have intermediaries you have shipping companies that are all of different sizes of different reputation levels and those parties will simply never form a consortium together and so if you can't form a consortium you can't use a permission blockchain so we had to be permissionless and then the outcome of those three decisions is exactly the reason that I'm standing here now we bought the blockchain Aleksei found out about that blockchain we had some conversations and I thought this was an awesome conference he thought it was a cool idea so what you have behind me is that technical specification that came out of our business process and those decisions we built one of the few box chains that exists in addition to waves and a few others in score ax I'm sorry not in Sparks but on score X in the Scala programming language which is excellent for block chains because functional programming again as we learned very important for security but the nice thing about Scala maybe a little bit different than Haskell is it's cross-platform out of the box so I can write it on my Android phone I can run it on Mac icrons-- Linux run it on Windows all right away without having you really change anything and then getting into the permissionless side of things we had to come up with a not come up with but implement a really secure proof of stake system if we wanted to be permissionless but still have all the security guarantees of something like hyper OS your fabric and then finally we needed smart contracts obviously you know those investment contracts that I mentioned earlier they need to be smart contracts and they need to be supported but this talk isn't just meant to be about how Topol is using the blockchain it's meant to be part of a conference about how we're all learning to use the blockchain and what we can all do you know with this new technology to better our business practices which is why what we are announcing today what we are releasing today and we'll be covering in our workshop is what we call project Atta and what project Edda is is it's one of the very very few frameworks one of the very very few ways that you can roll out your own permissionless but still enterprise specific blockchain so whatever your particular enterprise cases you can customize this framework to your needs but it's still fully permissionless so if you have a use case that a consortium really won't come together for maybe your actors are just widely different maybe they don't trust each other maybe you should be looking at a permissionless set up and project edit would be exactly that and i won't get too much into the technical specifics of add up because that's what this workshop I think it's at 220 or something is all about so I don't know where my time is right now yes so it seems like there are time for questions so with this I will just thank you all for the time thank you to auch saying the conference organizers for having us and I hope to see you around so thank you yeah questions yes of course [Applause] so many questions are there any questions no right so the phase one right now is we've been running a test net for about six months and we will be we're running pilots right now and we'll be watching the main blockchain at the end of the year so that's the the end of the testing phase is where we are so our business model because it's a permissionless proof of stake blockchain we can rely on what are known as transaction fees so our we have a token model that exists where transaction fees from all these investments are collected and the foundation that's operating the blockchain collects a portion of those transaction fees by holding its own tokens big intern intern yes of course how big do we expect this network to become in terms of notes so there's very interesting debates going on right now about delegated versus undiluted proof of stake and for those who are unfamiliar delegating means I pledge my tokens to another node and they stake on my behalf Ouroboros Genesis and all the Ouroboros starting with Prowse support delegation so we would expect our delegation and our number of active nodes to be a few hundred is what we would expect large enough to still be considered fully decentralized but small enough that you could expect walk times around one to three seconds thank you very interesting how do you how do delegate its notes gets chosen so very different leaf so how did delegated notes get chosen really depends on the proof of stake system you're using in or Boros sorry I'm not sure why I'm crackling like this in or course you get to be a node that is staking just by having enough stake pledged to you so there's no voting process there's no selection it's simply a minimum amount of stake that you need to clear so it's much more fluid which is very very important for security stand from a security standpoint because if you have to vote on 21 nodes or something like that it's much easier for them to form a cartel and if anyone's really into consensus cartels are the one thing you always always have to avoid otherwise you have no security left as I see Topol is much more like least it's built on delegated proof of stake also and build upon JavaScript can you tell the difference between the list and top so the main difference between whisk and something like topple or some anything that comes out of the project at a framework is our watching is actually not a JavaScript blockchain it's a Scala blockchain what we do instead is we use JavaScript for these smart contracts using the NASS horn engine and something known as the NASS for an engine to take javascript smart contracts and migrate them into the java virtual machine that's actually our our blockchain so it's a scholar watching with JavaScript smart contracts oh good - Mike stop other questions hey I've enjoyed this talk and I had a question about your staking so when you stake your vote two nodes how are you gonna incentivize people to actually vote for one node or another so I wouldn't say it's about incentivization to choose one node over another just like if you look at Bitcoin or aetherium mining pools right now there's very very little difference between them it's really a commodity or utility so what a delegate would do is they would take some small amount of that transaction fee just for themselves and it will really just be a race to the bottom for that but again because it's a much more fluid set you don't have to worry about the same governance concerns as you know if you vote on 21 nodes and they're gonna stay in for a really long time because there's you know it's less important who the nodes are because they come and go right and is the profit sharing between the delegates and the nodes is that negotiated between the two parties it's so not individually but each node it can have you know can be charging a different rate yes okay so it's up to the node correct and the reason that I would stake my takings towards a node is I would get some part of that transaction fee right so just like yeah it works the same way as proof-of-work mining pools so the majority of the transaction fee still comes to the person that originally holds the stake but the reason you might not want to be doing it yourself is because maybe your internet connection isn't very good or maybe you occasionally want to turn off your laptop or something like that you don't have constant very very high quality access to the network is why you would delegate other questions all right we supposed to have a coffee break for now for 20 minutes or so so please enjoy coffee [Applause]