J2 RF 09 Amber Baldet Quorum & What Business Actually Wants in a Blockchain

chains - enterprises and business and central banks and and corporates but I don't spend a lot of time going the other direction so I wanted to try to feel free to ask questions as we go or let me know if you know all of this already but one way to phrase this is that blockchain quote-unquote has brought a lot of quote unquote new ideas to the c-suite so here are some of those ideas right now some of these some of these ideas or technologies are they're not really new right and depending on where you've worked if you've worked in distributed systems or if you've worked in information security you are probably familiar with them already but a lot of the times these kind of these things or these these words are relegated to people that work down in the basement somewhere of what we would call the ciphers right so what blockchain quote-unquote has allowed is a renewed interest or an initial interest in people at the top of the house at very very large firms around the world to figure out what this stuff means there's also new things here you know when you're talking about crypto economics or some things that are more specific to blockchain systems that really are new and so they're learning these things for the first time so there's a lot of I think I don't say mistrust but there's a lot of misunderstanding about what it is that businesses are thinking about when they're thinking about blockchain and when I go the other direction and I talk to people that have traditionally been in information security and they say well why don't you just use a database you know it's about getting those people to understand that if we do adopt this kind of technology as the core core layer of enterprise infrastructure in the future that we're going to need to support that in the same way that we support an emerging cloud and and other kind of technologies so in this case I can tell you that what these people are talking about is it's it's not that they don't understand what a blockchain is and it's not that they hate cryptocurrency and it's not that they really just want to distribute a database and a lot of times it's that they have a business problem and they just don't have the land which to truly express what it is that they're trying to solve and they've read a white paper they've read some press and they want to understand it better so we can be kind right so here's what's actually happening businesses are experimenting along this kind of spectrum right and a lot of enterprise distributed ledger projects probably fall closer to a distributed database public blockchain of course we're all familiar with but there's really this kind of fuzzy gray area in the middle that's what I guess I'd like to call mutual eyes to infrastructure and I if you're familiar with creating a proof of authority Network for example on a close parity network that would be something that you might be able to use for a neutralised infrastructure sort of implementation you can do the same thing by setting up a vanilla gift network and simple with your friends now of course any of the ether that you would create and that sort of instance has no value on a public market but you can treat it just you can do all of the same kind of things on your own test network or your private network that you would do on a public one but so you get businesses that think this is great we want to make sure that we are only transacting with people that we know and if we're creating a permission network we can achieve that right so anonymous participation with public transparent Ledger's is really really a problem for a lot of business processes especially if you're talking about financial services where you want to handle anti money laundering and know your customer laws and at the same time you have a fiduciary obligation to keep data private and also if you're trying to build markets and exchanges just as we're learning out in the public space if it's really hard to do that in a private decentralized or in a decentralized way without some concept of privacy with the way that auctions work right so every use case kind of falls somewhere along the spectrum and there's not really a huge push as far as I can tell to move core infrastructure of enterprises on to the public blockchain right now and I think that's great because crypto kiddies didn't turn out so well for anybody for a couple days and if you look at the amount of volume of traffic that happens that comes out of the business world I don't think we really want anything overwhelm overwhelming the public network so that's okay but in designing future connectivity what we want to be thinking about is what a network of network looks like and then it doesn't necessarily have to be all enterprise disaster recovery kind of systems over here and all anti fragile public blockchain anonymous transaction things over here but that pragmatically it's probably more likely that we're going to end up with some sort of hybrid network of networks that can hopefully talk to each other and so developing that kind of underlying standards and the way that that that can feasibly happen in a technical capacity is something that for example the enterprise aetherium Alliance has been working on but also you see technical implementations of whether you're talking about cosmos or polka dot or the inter ledger protocol or these other kind of cross chain projects so as businesses are experimenting there are some things that they really like and specifically are I'm talking about a permission blockchain which supports a rigorous or a robust privacy model but you can share primary records so a golden copy of something without necessarily surrendering information ownership to third parties that's real business challenge when people say reconciliation that's what they really mean is and it really comes down to trust so this idea that for some reason because you know other entities in a permissioned distributed ledger or blockchain that because you know who they are you somehow trust them I think is just fundamentally flawed right if we trusted everybody else's records we wouldn't be doing this stuff in the first place so being able to transact or transmit information without surrendering information ownership is critical also I think we could do a lot to reverse the aggregation and centralization and this creation of data lakes which over time especially in a transparent public blockchain construct are going to at some point become surveillance lakes if we're not careful so figuring out the way to balance privacy with data transmission is critical not just for business but I think also for public endeavors also the way thinking through how we have historically treated information security and understanding how that could change if we were to use this sort of mutual eyes infrastructure so historically we've always had this if you're on the outside of the firewall you're you're bad if you're on the inside of the firewall you must be safe I think we all know that that's probably not right and instead moving to this kind of structure where you have various information security layers and defense-in-depth really mid throughout the entire stack it's a different way to think about cybersecurity and also just being able to drive adoption of significantly stronger encryption and data privacy standards you know if we can actually bring this technology into enterprises because they've been built from the ground up with with generally stronger more modern standards we can achieve a bunch of things that are on some compliance checklist somewhere about encryption at rest and other things just by using the tech so all that stuff is very attractive but privacy compliance governance always a challenge dispute resolution how do you actually integrate this with an existing legal framework performance and scalability settlement finality you cannot tell a corporate treasurer that their payment was 96% probabilistically settled and may or may not unwind at some point in the future if there's a hostile Fork of the network so figuring out how to to work around those things continues to be a challenge so business is worried about Justin over these right privacy of client data confidentiality of their business practices governance in their context governance to them means control of the network so making sure that they don't lose control of what's going on in what's okay which mostly is because they want to make sure that they can still meet the rules and be compliant with the rules that are imposed on upon them performance just meeting the needs of you know basic doing business the security we just talked about and efficiency so if this isn't better than what you've got today nobody's going to use it makes sense so it turns out the public networks are worried about the exact same things so from a privacy standpoint we're worried about anonymity what happens to your personal data what happens if this is stored on chain and then there is some sort of disclosure event right from a confidentiality standpoint what are the impacts to surveillance in the long term governance in that case we're talking more about the control of the network evolution and the EIP process if you're talking about a theorem specifically performance we all are familiar with the throughput and scalability challenges right now security systemic Lee and a fish see are you gonna launch your business here just because it's cool or is cuz it's better than what you could be doing with the rest of your time is sending a payment through a distributed ledger a better more cost-efficient way to do it or not so maybe these things aren't actually so far apart it's just that people are trying to solve the problems in different ways and maybe if we learn from people who have implemented global systems at scale we could be moving faster and if they were also listening to some of the challenges that we've had and you know things that are actually tested in a hostile production environment every day we would be building better more robust systems more quickly so enterprise versus enterprise so and all credit for this goes to the technical creator of quorum Patrick Nylund but Nielsen but he said Enterprise with a capital e versus Enterprise with the lowercase e it's a subtle distinction so enterprise with a capital e is software that is made for enterprises that means that it meets some sort of complicated fuzzy requirements these are all the stereotypes about enterprise software right it's industry specific conflicting implementations you end up with all these bespoke customizations different clients are running different versions you can't ever upgrade things you have to call someone there's always a vendor support contract its proprietary it's expensive it doesn't integrate with things right enterprise software it's great okay so that's capital e enterprise software and historically I think when we've talked about enterprise blockchain this is where people get stuck and we start creating all of these enterprise clients and then it looks a lot like existing enterprise software and not only is it very difficult to attract developer talent to want to work on something like this when you tell them that this is what you're trying to build when they could be building something that's a lot more like crypto kitty's but you know we have a next generation of developers and just people in general that are coming in to work at large companies and large businesses and they don't necessarily want to continue to use software that looks like it was developed in the 90s so maybe what we actually mean is that we want enterprise grade software so enterprise with the lowercase e that's software that is actually reliable software that has some sort of dedicated development and support team someone that you can call if you do need help it works with other software they are already using it supports common standards common languages they are already familiar with there is some sort of well-defined API s and clean interfaces the deployment model takes into account that not everyone is cloud native all the time you need to be able to deploy some of these things on-premises there need to be access controls and integration with Identity and Access Management programs there's also probably ongoing QA security auditing it works with your existing tools it's not just adding massive technical risk if you deploy this thing and then people can learn about it there's training it's not super niche it's not just like twelve people that know about this but you can go out in the world and you have some sense that if you build a business critical application on this software that it's going to be continued to be supported in the future and of course designed to scale so we have a ride to quorum 2.0 which was released in December that is not to say that I think that quorum is exactly nailing all of these things right that's that's that's not exactly the segue I was going for but rather that there's there's a difference between building enterprise distributed ledger systems that are meant to always run on a 24/7 uptime well supported cloud infrastructure supported by an incumbent technology company and software that can run somewhere that has power for hours a day and where the regional bank doesn't trust a central bank and where local people are getting smartphones but don't have any bank accounts so if we're trying to create an internet of value it's not just about connecting business to business at some point it also has to be about connecting business to consumers and it's not about just consumer to consumer or as I like to call them humans but we actually want to connect all of these people together and be able to provide not just financial services but we're talking about you know new cottage industries of all the kind of internet of value things that we're moving towards so if you want to create those kinds of marketplaces that are not simply isolated amongst the people that already have all of the assets in the world then we're going to need to do a better job of making software work not exclusively one for one / the other which will continue to bifurcate those power structures but to actually integrate these things together into one thing so quorum is a fork of the Geth client um you can run it in it's meant to run in a permission setting and that's because every node has an identity their assigned public private keys and you can basically I'll get into how this works in a second but you can address a smart contract to any other node in the network that allows you to have private smart contracts so what we're trying to solve for is governance right so you basically solved for your governance by having your own permission network where you set the rules you're solving for confidentiality so there's various privacy mechanisms within the network and we're trying to solve for some some security things so this gets a little nebulous but how you're treating consensus on the network so if you're running within a permission network you no longer necessarily have to be using proof-of-work however depending on how you expect these things to scale when you look at supply chain for example as a use case a lot of supply chain use cases started with you know five five entities and the network topology looks a lot like ARPANET in like 1970 right like you're directly connected to these people and you trust them and everything's fine but an actual supply chain might contain millions of participants if you're talking about somebody like you know a Walmart or a Home Depot that sits in the middle or an Amazon and then you have all of the kind of physical the physical supply chain crossing a financial supply chain that creates this actual way that we move goods and money around the world once you have millions of participants in your permission blockchain it starts to inherit a lot of the security properties of a public blockchain so we can't actually be designing something that will only work like I mentioned previously in just that clean trusted case which is what a lot of the pilots look like right now so how do we move towards something that has other properties so within quorum we have raft which you can use on your standard kind of closed enterprise system Istanbul BFT is that so this is a Byzantine fault tolerant algorithm but it also does not require the use of cryptocurrency it was created not by us it was created by a startup called AMI but this is to solve Network coordination amongst a larger group of participants who do not necessarily trust each other as much so with quorum you end up with its a theorem but plus permissioning privacy performance configurable consensus and depending on which algorithm you're using settlement finality so that's the goals of the project okay just wanted to make sure that I wasn't rambling for two hours okay so here's the generalized architecture this is not to scale right like the the Geth portion is actually a ninety ninety-five percent of all the code it's actually not changed that much where it's really been and that's intentional right there's very minimal modification what has been changed or what has been added in is constellation which is the privacy the ability to address these smart contracts around the network which can keep your business logic private and then above that all of your smart contracts would function exactly like they do above that you have your kind of standard a player you can also work with projects like web 3j which Connor was in here earlier talking about which mean that you can write your business app applications in Java and things that bankers know and love and you're in its we don't necessarily have to treat this like it's a completely new software okay so just in the interest of time I'm gonna jump forward so JP Morgan announced a partnership with Z cash last year at consensus and that was not the the goal of that partnership was to pull the zero knowledge settlement layer or zero knowledge security layer as they had coined it into quorum as it became a zero knowledge settlement layer so what that is is we end up here's kind of the the architectural model or I guess this is more of a this is how you can think about it so you end up with this the in the middle layer there that's that's aetherium so that's your shared blockchain everybody is you know your node on the network you're all participating in validation underneath that you have the constellation network and that's what allows you to address smart contracts around the network it's a point-to-point network it is not a peer-to-peer network so one thing that businesses are very concerned about is meeting regular regulation but any sort of rules about in-country storage of data right there's actually rules that say that you cannot transmit encrypted payloads around the world it's kind of strange if you think about how the internet actually works but like don't bog yourself down with these things so what we end up with though is a system where you you really don't want to transmit encrypted payloads you definitely don't want to share store them on a shared blockchain but you also don't want to transmit them peer-to-peer so by making that point - point you've satisfied that requirement and you're also able then you can have your own kind of disaster recovery model around those the private payloads but we what we critically are doing there is taking a hash of this encrypt the encrypted payload and committing that back to the shared lock gen so everyone does actually validate that portion right so it is a it's a quote unquote known issue right now there is not consensus over the private state within constellation but that is not meant to be that way forever it's simply a priority roadmap item of resourcing to fix it there's some designs around it if you want to talk about that feel free to corner me later so what you have then is you have this shared blockchain in the middle which is basically your event coordination layer if you were to deploy a basic vanilla solidity smart contract that did not utilize constellation then it would just get deployed in that layer just like now if you utilize the quote unquote private four that activates constellation then those payloads are stored down here up on top that is the zero knowledge settlement layer and that is a UT EXO model of encrypted while cryptographically private er C 20 tokens so in that case you can basically as you're creating a work flow you have a business process for example let's say of a loan let's say and the loan is for like a hundred million dollars so you define that in your private smart contract that private smart contract issues all your er C 20 tokens just like you would do in a standard process but it creates them in such a way that then they sit on the shared ledger and they're shielded and the only information that is queryable about those tokens would be the amount that they're for and the current ownership every other business process that has to do with that loan about there's like a whole whole workflow process that goes around originating and then who owns it and how you're handling asset servicing on these things all that stuff remains private that means that very importantly let me just jump ahead to this one to show that so you are not running business logic through anything that is zero knowledge that would be ridiculously it would be very costly from a gas perspective but also just take literally forever instead the only time that you need to call out to the zero knowledge settlement layer is when you are actually effecting a an actual asset transfer from person a to person B or while the data wallet B so some of the benefits that this gives us though is that you can create actually private secondary markets so there are requirements that when you sell a portion of a loan forward the new owner cannot know who previously held it right so the the provenance that's considered such a benefit in a transparent public blockchain system actually just really doesn't work when you're talking about how we treat existing financial assets and there's any number of reasons around that so so what you get is two things that kind of pair off against each other to create this hybrid privacy model so so constellation allows you to have private smart contract execution but because they're private unless you architect your contract in a specific way you basically would lose the concept of provable mass conservation now you you could make it work but we didn't want to rely on developers to always make it work since developers occasionally make mistakes so instead you can call the very simple CSL api's which will create those tokens and then re-implement all of that for you so that's those things kind of pair off together on the other hands so we have prevention of double spend require specific application architecture mass conservation comes from CSL private smart contracts and constellation but we're not running anything like the EVM through through zero knowledge people tend to be confused about those things but hopefully that was clear okay so public transactions are broadcast to everyone within the network just like on regular theorem their process there right now we have the concept of gas within quorum gas is set to free there has recently been a pull request to re-implement the cost of gas and so I think we're already seeing this evolution towards as these networks scale that people actually do want to re-implement some concept of either I don't want to call it incentives but at least reputation in some way you need some way to to at least rate limit the network right so we're kind of toying around with different ways to do that without making businesses by crypto assets which they absolutely still refuse to do private transactions then our point-to-point and a hash of those are included on the shared logjam okay so much as dsl is to z cash blockchain is to Bitcoin that's the disclaimer on that okay so if you're interested in trying out Korman's ESL's ESL is very much a proof of concept there's a whole other phase plan where we'll address some of the proof malleability as well as some of the other they all of the performance increases of the Z cash shift to the new Jubjub curves should also kind of flow forward into the zsl implementation so we'll see right now generating that proof does take about 30 seconds again it only happens at the end when you actually move an asset but that should come down to like five to seven seconds it's probably worth noting that JP Morgan also worked with the Monetary Authority of Singapore along with consensus to do an a second zero knowledge implementation for a project they were working on it does not use the SL what it does allow you to do is to use a simple a simpler zero knowledge primitive to shield an account balance within a private smart contract so it was a slightly different kind of a project but now we have at least two different ways to think about zero knowledge in this kind of a kind of a system and what you can do there the reason you would want to do that is if the specific problem or the use case they're working on is called gridlock resolution and right now Maz Monetary Authority of Singapore sits in the middle of all of their kind of their banking infrastructure and at various points through the day they issue these sweeping instructions that clear the gridlock so like if I owed you 20 bucks and you owed him 50 bucks but you can't pay the 50 until you get the 20 that's actually centrally coordinated right now so they wanted to figure out if there was a way to build the same thing but without that central coordinator and we were able to demonstrate and do that on quorum with the privacy features and because those balances sometimes drop below zero you would want to use something that looks a lot more like this account balance structure than something that probably looks like the UT EXO model of zsl so next steps out that I think that's the next steps of what we're doing with CSO and then so there's a lot of people that are working with quorum right now probably people that you've heard of like Accenture and Microsoft maybe other people that you may or may not have heard of like chronicled and amber data and block apps amber data is not my company it's a reference to it's a reference to nick szabo saying that blockchain was like locking data in amber so sadly i whenever they're in the room though i tell them they have to tell me it was me okay and then within each a there is a quorum test net there's a couple dozen institutions either on there or queued up to join so people are playing with this stuff there's also a lot of actual projects and pilots that are happening out in the world but we don't have time to talk about all of them so it's that as a test network corporations are actually playing with these things Thomson Reuters for example is actually like publishing a data or Oracle of market data that you can subscribe to and use within your smart contracts and I think this stuff is real and it is happening so that's very exciting and then just to kind of get the imagination thinking here's something that I've been interested in for the last year or so that I've been trying to get off the ground we'll see what happens with it long term but why couldn't we launch something that looks like a public bug bounty program on quorum right so if you were to create a network that becomes something that public what does public mean right maybe it's actually what is semi-public mean if you can join a slack channel for example but it requires an invite but all you really need to do is click on a button to get access and registered to do that why couldn't we set up a quorum network that functions that same way so in essence you would register and you would know who you are and if you do bad things that are outside the bounds of what's allowed in the network maybe you get kick band but you could have somebody that you know so censorship resistance not a thing here right but you could also have some sort of either reputation or gas allocation system so that you can let people test out you know test various kind of security things there so I think that would be really interesting way to kind of play with with changing kind of the boundaries around what does a public or a permission network mean and I don't really think that we're at risk of anybody blowing an actual like guess Oh day on a quorum Network so I probably don't have to pay anything out on that okay because it would be much more valuable elsewhere that's the joke okay so um to learn more get involved you can either kind of there's a variety of websites I would recommend probably just github is the easiest everything that we've built around guess is Apache for constellation MIT for the CSL layer the rest of quorum does inherit the GPL LGPL from Jeff but it's relatively like tightly controlled so we're trying to keep all of the interfaces around that probably something slightly more permissive like Apache and there's also a slack I think there's about 400 people on the slack at this point pretty active and finally never forget that all mentions of hash tag crypto actually referred to crypto Politis the pliosaur reptile from the middle jurassic period of england so this is how I've rectified people that are upset about crypto either meaning crypto currency or cryptography that's me I'm uniting the tribes so you're all wrong so at this point if there's any questions or comments feel free we have a couple minutes I think [Applause] yes oh do we need mics or anything for that I think for the livestream they want a mic thank you why did you guys choose to go with zero knowledge CK snarks as opposed to ring privacy sure I mean ring signatures being a method of obfuscation and not encryption I think we were trying to see if we could actually play with --zk snarks because they're considered kind of the best-in-class privacy solution there are other Enterprise aetherium focused platforms like clear Maddux platform photon that are looking at that and I don't think that in the long run they're not necessarily mutually exclusive this also there's a bit of a a just pushing the envelope kind of thing of sitting in a lot of rooms where people say that privacy is going to come from SGX or something else and saying no we're gonna do it with applied cryptography here and let's let's prove that we can make this work I don't know when these things are in production at scale what they will exactly look like and if this is the end-all be-all of privacy forever that's not the goal but to prove that we could get large businesses behind using this and they are that's just it was worth proving out thank you for this torque you just mentioned that quorum is a kind of enterprise grade theorem so question is is it a project that is promoted by a team enterprise aliens is it who's making the support of this product is it will it be licensed in the future and I wish I had answers to all these questions I mean running and I didn't get a lot of time to talk about this which I said I was going to on Twitter but running this project or an open source project within a highly regulated entity has been a huge learning experience I mean JP Morgan had never did not have a github presents before this had never really open sourced anything that was a real community project and there have certainly been challenges in doing that but it's also been great to have that kind of support so it's it's kind of goes together I think the enterprise etherium Alliance is a standards body they do not own code at this point and there are a variety of an enterprise etherium or Enterprise clients that aim to be enterprise grade within EE a so quorum is just one of them it's always been my personal opinion which others may or may not share but that any individual client isn't going to win the market by excluding others or being named the thing on day one it's a bit about people wanting to use it and people wanting to develop on it so I don't think there's a need to do that that said what eea is working on is a technical specification for what aetherium should look like if enterprises can use it and any client that exists within a EA should be able to be either certified EE a compliant or say that they meet 80% of these things or what not so there has been a lot of work in the last year to make that spec exist I'm hoping that it will be publicly available soon but my personal feeling on this as I hopefully came through throughout the presentation is that I really don't know if Enterprise anything needs to exist what we need is the Internet as the internet exists that can be used for personal use or business use but you use a variety of features if you are running a you know high security kind of a business I feel like that's the direction that we should be going with an enterprise aetherium client but we are not we're simply not there yet and it's it's a little bit like VHS Betamax I guess but I feel like in the next year or so there's simply no way that the market is going to support the number of clients that exist and are available so some consolidation is inevitable and who supports it right now there's a bunch of people in the quorum working group a te a there's the 400 people in the slack Channel there's the actual tech team at JPM which are basically like l3 support if something is on fire because it's a very small team but there's also if you have if you have a larger business like Accenture Wipro Infosys cynic Rahn like all these companies consensus they all are doing paid engagements supporting quorum wait for the she's gonna give you the mic and before you just that if etherium could learn from some of the privacy mechanisms that you've deployed without a some kind of political trade-off in terms of trusted setups oh well the okay so the powers of Taos thing is happening now and we're participating in that we being like me on my personal time and Patrick but so I think the some of the concerns about the trusted set up hopefully will go away in in that sense if you believe in the new mechanism of the new ceremonies I think you know what what aetherium integrated with adding like the zero knowledge primitives last year is a good step what zsl has been is an application that actually functions like it's an application of that technology and what you can learn from that I think is that the the governance of public etherium being necessarily slower and more considerate and involving thousands of participants means that nobody can just say hey we're gonna do this thing and we're doing it because I said so right and that's that's great for for that for public aetherium for quorum we were able to say why don't we see if this can work we're just gonna do it and we believe that the trusted setup is fine or I guess as we say initializing parameters or something that's not so polarizing as trusted setup so if you believe that that setup is flawed there's nothing I can do to like fix that for the public aetherium community that's that's like a religious debate there's an adequate answer good any other questions okay oh sorry one more yeah go ahead thank you will it be able to scale as fast as a theory on planetarium scales yes yes so one of the really good things about staying close together is that we inherit almost all the benefits of both security advancements and scaling advancements now one thing that's interesting is that some applications of quorum that I hear about now are people that want to run quorum almost as sidechains to public aetherium and they'd like to do something that might be expensive too expensive to run on the public chain or they maybe just want to be masters of their own universe and be the only ones that are allowed to deploy a smart contract because they're building a videogame or something and they want to let people that want to play their game onto that network but nobody else and then maybe you commit the results of the outcomes of your games back to the public chain like people are thinking a lot more flexibly about these things and so in the same way that we benefit from the scaling efforts happening right now and there are several of them I've probably talked to you more privately about I don't know if we're ready to broadcast them all over the world but in the same way that we benefit from that work the public chain also can benefit if we're able to communicate back and forth as some people are working on JP I'm not doing that development work but others are that quorum or doesn't have to be quorum literally any private networks could help also scale public aetherium okay those good if you have other questions feel free to grab me after the fact I think there's drinks tonight and whatnot so thank you very much [Applause]