Building Technical Innovation for the Borderless Economy by Paul Riegle of Algorand

[Music] you guys okay so my name is Paul I am here from an other in to talk to you guys about a little bit about of our project so momentum we've have it I've been having in ways to develop and kind of get the ball right so I'm Paulie so my experience is both so I came from a country called car black point sacra security company I started more about sixty people of the only product manager at a time I left we had scaled all the way to 1300 through an IPO and if you saw it in the news maybe two weeks ago they were required for 2 point 1 billion dollars in an all-cash deal with VMware which doesn't suck I also have experience at large enterprises and IBM which is one of the largest ones you get up there 350,000 employees which I think we really need kind of perspective on building products having seen both a place with very little process and then having to kind of change how do you do things along the way up to scale up and then getting to the very largest and seeing all of the process all the time just real briefly if you're wondering and what is out there and so we're at the foundational public information list group of state protocol we're super technologically focused innovation where you see projects with these they don't see a lot of innovation three years mostly unrecognizable from what the underlying blockchain technology looks like now and we want to be the former that stands for front of that and one of the ways we do that is through our research team our founder for example Sylvia Connelly he's a cryptography pioneer is it Turing Award winner he's got mighty professor and each kind of underpins a lot of the underpins our research team and pushes on that technical innovation but just a kind of give you an idea of what our project is like one of the more substantial projects in terms of the folks that we have behind it that you'll see out there so we talked about Sylvia our CEO Steve Cochinos use Austin CEO founder of a company called fuse here in Boston familiar with them and also a founder that a company called Bladelogic if you've read the book the hard thing about our things play logic is it covered is often mentioned as the sort of nemesis of the main characters Sean Ford on the business side as well he was a CMO who logged me over the last four years founder he promised in an early employee at monitor group the research thing this is only some of our research team you can see a lot of mighty stuff on there we've got a team of 14 researchers who are focused entirely on innovation around economics cryptography game theory distributed systems and they are that's what their focus is full-time we also have a corresponding engineering team of about 14 folks that are really focused on a translated every research into something we can implement and then be really executing around what we can't build on top of those innovation I want to make sure we don't get to academic focus one make sure that not only we have that academic so I'll talk about like a couple inches deep here on our technology and then come back up and talk about the things that developers or the interaction that makes to so as far as our core technology there's a couple different approaches to consensus or agreement code that are out there right now you're familiar the etherium which i guess everyone is group work you're probably also familiar with delegating proof mistake where it's state based but I actually give it to someone else to go for me or bonded prove that state where I take some amount of my state I put that happen bond and that can get /etc there's definitely pros and cons because one of these and I'm here to tell you they're all terrible the definitely pros and cons it seems a lot of work protocols you have the lower transaction speed lower finality delegated through the state for the state you tend to have a very small team who's in charge of a very large group we have what we call a pure proof of stake there's no penalties there's no slashing you're not delegating my to someone else to devote to someone else and we do that in a couple different ways so first of all to take a step back we have when we when we agree on blocks we have an ever-changing sequence of committees so what does that mean there's effectively three steps in our block confirmation process we've got a puzzle a soft low surcoat each one of those is a separate committee and it changes for every block and a chicken for every round within every everyone each committee is roughly a thousand users that are randomly independently selected and they only have to send a very short single easy message in terms of their vote so what does that allow us to do I guess being a little bit deeper on the video time so who selects deputy right that's a a question for our consensus each committee member actually selects themself so they run a private lottery if you're familiar with verifiable random functions that's the cryptographic primitive that we use which Silvio McCollum is actually a co-founder of a number of years ago the RS have two really interesting capabilities or properties one of which is you can't cheat in Europe like you can't fake that you would want it and to you can always actually prove they did one if you combine those two things it's a really good way to be able to select committee members in this kind of situation if you win that lottery so the VRS is just buy one then you can participate in that committee and continue on the probability of you winning the lottery or being selected by the PRF is proportional to your total amount of money in the system sort of proof of stick so what does that provide well first of all it provides a really good decentralization around the agreement itself because literally everybody in the system is eligible to participate everybody in the system could be a bot composer everybody the system can be a lot confer or verify which is very different than a lot of other systems where you end up with a couple of classes of users these guys are the proposers these guys to the verifier these guys are the users that is not the case now granted super decentralized from that standpoint second date super scalable so the size of the committee we have two mathematical proof some back-end aside the committee does not need the size the scale of the user gets so even though you can go up two million times the users the committee sizes don't have to change and that means that we can still reach consensus at the same speeds even with a lot of users that you could with very few users and then finally on the secure side so what's interesting about this a selection process in vrf is that you don't know who the voters are so as an attacker I could say all of this is easy if I know who the committee member is I'll go corrupt that committee member and then I can break your consensus at the thing with the PRF is when you're a selected one you have won the lottery you can send one message that message essentially says two things one I guess dream one I have won that lottery - here's the proof the proof and three years of my building so that's one message which means you don't know who the committee member is until after they have voted oh and by the way if they've already voted they're no longer part of it because now there's a whole thing so you can't get to the individual voters because you don't know who they are and they're constantly changing by the time you know they are part of it so if you're familiar with the blockchain trilemma you know big to do centralization and scalable secure we feel like our consensus algorithm is a great star for being able to solve that trial and actually achieve all three of those properties we're not done we talked about being innovation focused this is just a start we've got a lot of innovation a little bit of that so from a high level there's not just against consensus algorithm is a bunch of other properties - that are super interesting for technical benefits of our blockchain then while computation required we talked about anyone can participate also this energy supervision our founder wrote this slide he's Italian he put super after energy yeah it's very energy-efficient because there's no you're not solving at college you're not trying to do anything special you're literally sending one boat with a message secondly we protect both against further context and network attacks so what do I mean by that when you think about a partition if you're never mr. partition you usually have to make a trade-off and that trade-off is am I going to choose liveness continue to create blocks on the chain everybody might create am I going to choose safety which is to say there won't be a formula for most chains choose like let's so for instance as example when there's a partition each side of that partition will continue to create blocks when every digit comes back together one of those chains has got to decide yeah I'm actually the real change you drop all the transactions that is not the case an outer ring we've chosen sinking which means women in network partition we don't create a block until we have enough votes to bring them together and actually reach consensus what's interesting about that we've kept a lot of really good feedback from around from our financial customers and our financial developers but I really want to know if there's a transaction that I've created is it real is it final do I have to wait around and it might get dropped later that is never a word without right and it's it's the third thing is we are going to operate at Enterprise speed what does that mean because however we run a long-running task $10,000 will be distributed across the countries with evenly distributed stake which program is taken Congress is the hardest configuration and we could consistently run at thousand transactions per second with no degradation in our performance that's only one half the equation though block latency also had very important we create blocks and every five seconds since launch we launched about two and a half months ago we created about 1.8 million blocks all of those blocks have been under five seconds in fact our average time is about four point three seconds so not only are we can process a very high number of transactions well when you send a transaction it is fine I'm completely done within five seconds and then the final one is this immediate transaction finality because we've chosen safety because no Forks when a block appears in the blockchain it is fine there's no we got together with the question ok great five second blocks how many confirmations will have to wait for word you don't have to wait for any confirmations it's the first one we've talked about this it lets your application run as fast as it needs to run it takes the scalability question just the second one which is kind of interesting those actually brought to us by companies that have been building at it increases the users confidence in your application why why is that or why did we get that feedback if you have any experience sending crypto to friends or to an application and you send it and then you wait 15 minutes to see if it's real or not and then you use a system where you send the crypto and algo in our network it's there it's done five seconds later it's a fiat like experience it's like using PayPal it's like using demo and it takes your users from a mine space where it's like using this cool new not quite their technology to worse traditional mindset third reducing cost structures this is depends on the type of business but our transaction fees are very low there have a thousandth of an alga over to native currency and it won't increase until the blocks get over a thousand transactions per second so the cost records release is significantly depending on your business model and I'm going to combine the last two so we really strive to make the development process movement and using we've got SDKs across four languages go Java JavaScript and Python we've got a provider pure state which is a partner of ours who's put on the API service if you don't want to run a note it's very pure alike I expect you are within Europe and we also want to be a technology partner which i think is very different for the blockchain world in the traditional world again in the traditional world the way you choose a technology provider like hey I'm going to build something on top of the platform okay let me go do my research online and they call each one of them let me talk to them let me set meaning let me can see the roadmap hey what is your support process look like all of these things that businesses need not just a large traditional business with you businesses as well to make your experience better aren't available another public information with James and Alan Duran has the ability we are publicly Commission lists so you get all of those benefits but you also get the benefit who is driving the innovation and can meet you and talk to you and help you when you need help and we found a really good reaction to that capability as well which results in any like traditional logo slide it's a lot of stuff I'm not gonna go through but then I guess the one interesting thing here is that we've been launched for two and a half months and every one of these logos our fear is either building on a program has built a tool that supports Algar and accepts Aldo's which is our native currency or as a it's supporting the project in some way like providing no hosting services and this kind of traction is early we feel pretty good about to get one level down of some of the use cases that are buried in that slide that market gaming we've companies that are building telecommunications and payments applications this is a particularly good application on algorithm because of that transaction speed and because of that finality building insurance decentralize insurance models assets open ization is with a company called asset block which is launching a state government organization of real estate and allowing it to get into markets and users that haven't been traditionally haven't traditionally had access to that market median entertainment company called Oh toy does digital rendering and GPU rendering is building to centralize rendering service on top of a grand enterprise solutions data analytics this is an interesting one so a company called flip side crypto which is also based here in Boston has built a data analytics suite where any application who is building on top of that room gets a free service from them which gives you insight into your users where they're going spending what their activities are by public-key not actually user information but super super useful and something that's not necessarily available propose this early in their crime but we are just getting started we're super innovation focused there's a bunch of things coming down the pipe which I would just build out here some of this stuff is I'm going to a little bit more detail in the other so one of the things we're really focused on when we think about what we are think about ourselves as a platform for economic exchange so if you're going to be a great platform for economic exchange what do you have to be able to do well first of all you gotta be able to represent all of the assets right so we can represent our own our own native currency Yahoo but what if you want to however those assets we're super focused on that that is coming into this short term I'll tell you a little bit more about pixel is a is a technical optimization or aggregating signatures are using BLS standard we're helping the formula in fact we're driving that formalization a number of other chains including self validating transactions are super interesting allows those do not have to keep the state for accounts we have archived in our notes if you want to keep say for the account you can handle you can do a lot of very interesting things in terms of Lake nose like clients and more entertaining atomic swaps so when I didn't think without swaps I think cross chain I want to give you either you give me the coin we're focused on the interchange because of these assets they're coming and the proliferation on that asset so now that means we want to be able to do something that's a little bit more interesting than the traditional Afghan law contract doesn't have something downsides that you see there and then posted steel trades and full fledged smart contracts frankly full threats full-fledged smart contracts is something an absolutely in a roadmap and we will have but we think we can solve a lot of those use cases without full fledged smart contracts which actually which can increase the security profile of decrease the attack surface of your about the patients and make it a little bit more easy to get off the ground and get going and creating your own small pockets so I just wanted there's a lot of text in camera don't worry about the I want to talk about creating your own assets so this is super flexible internally in terms of the assets you can create it covers the full spectrum that's fun durable sunglasses hats like cryptocurrency stable coin system credits and in fact on this table points Braun tether I don't you saw this announcement about you month gonna have to go tether has announced that they are going to launch on Alderaan to could be using this functionality non-functional tokens like real estate tickets collectibles etc an example of this would be the asset block has a real estate and then restricted assets things like securities some types of real estate certifications I keep saying layer one over creating them in layer one and I'm saying like I'm focusing on that for a reason layer one assets are enjoy a security profile that layer to do more suits at spar contract that's heads do not enjoy the security profile so they get to take advantage of all the inherent protections that we put in for our native currency and and I'm sure you've had the experience where either you had you made a typo on your smart contract or you're copying pasting somewhere else and then there's a vulnerability and it sweeps through a whole bunch of you know here's to 22 contracts or you're just like why do I have to continue to copy paste the same thing that everyone else is using and all those sort of go away with a layer one implementation it also provides some interesting inherent compatibility what I mean is any application that's built for any asset on other end will automatically support all of the assets on top I think so from a community building in a network of that perspective you get a really interesting flywheel where there isn't hey I know my app I support these four but I want to support those other seven but it's a lot of work it's some tweaks and I just on time so if someone can convince me that it makes my under sense almost I support them in the later one model you already support them you don't have to do anything more which is it I think a really interesting thing for communities and then the last one it's super easy you send one trans actually network with like five parameters and you're outside of creating five sides right so there's a lot of benefits to that later on capability and then once you have this ability with the proliferation of assets you want to be able to exchange them I want to give you out those I want to take your boat I want to give you a ticket and take your elders or whatever it may be and so we're introducing this capability of in church anatomic swaps you can exchange any asset or any other elder an asset with guaranteed simultaneous execution of the exchange without any locks or time or without cash back on track so there's no multiple steps there's no waiting this is some inherent technical innovation that we're bringing to market about where it's really think it's an improvement on atomic swaps that are out there now there's a bunch of use case to do this is the obvious of a bilateral exchange you have something you want there's also it supports multilateral exchanges so I want to give you the town those who want to give him a boat but only if he gives him a car we can put that into one transaction and make sure that either all of those happen at the same time or they don't with the same structure that you would for just a simple bilateral exchange and there's cool use cases around around that is all right we're gonna match your donation of and it will only occur if the other denomination occurs as well also for where decentralized exchanges a couple days ago hi Dex which is one of the largest decentralized exchanges and that's that they're going to build on out of there and they're going to be taking advantage of interchange palette swaps as well as some other functionality we have to do what we think really takes a leap forward and decentralized exchanges in terms of the user experience in the speed with which you can trade our goal is to allow the decentralization that people want but still give them the traditional Fiat world feel in terms of the user experience and that's really what we're trying to mix together from a developer resources perspective if you guys are interested in checking it out the developer gallery and calm I've got work is our sort of jumping off spot valid documentation it's got our tutorials we've got sample applications we've got SP Kennedy that I mentioned earlier there's also an interest style of EDI gateway that pure snake offers we have online forums developer they don't act on that there's just a lot of ways to get involved here I definitely recommend that you reach out either do Twitter at Adyar and through our community forums community alder org work through that developer site to get any more information that you might need [Applause] yeah so I made this but still not creating my own layer one hazard no it's so it is not we don't have smart contracts this is a it's a I would consider saluting after I'd like to you read it off chain even though it's getting executed on changing the actual creation of its own so are the way that you created is you literally send a transaction it has like five parameters and you need transaction to make a transaction and it just gets created on the backend by the north side and then when you're sending a transaction instead of sending a standard transaction as you would for an algo which would just be simplified from two amount it's from two amount that's all of it so you're using all the same infrastructure it's going through all the same consensus code it's as if it was created by the protocol itself so maybe so it's a separate there is no EDM there is no etherion not bad there isn't it there and we are not appearing we are not around we are a wholly separate watching [Applause]