Day 2: Arthur Gervais (Liquidity Network)

Transcript

[Music] all right moving on now for cryptocurrencies to have any chance of knocking cash off its perch they need to at least be able to carry out lots of transactions simultaneously and have rock-bottom processing fees if they can do the former they can scale if they can do the latter then micro payments can flourish which hopefully enough is the title of our next block it's on scalability and and micro payments the first of our four speakers if my voice will hold up is there dr. Arthur Gervase he is the CEO CEO of liquidity Arthur thank you much thank you so much for the

kind introduction hello everyone what happened 2040 before crest why guess it was the first instance of centralized banking so what you have seen here is a balance sheet it's it was in Mesopotamia basically it tells you how much funds the particular person owns and well in this centralized because it's in one single place controlled by one entity and I agree much time has passed since then and this custodian based financing Cemetery business really flourished so we have seen banks and services being built around this all over the world and even more with the invention of the Turing machine by Alan Turing in 1836 the

internet and the computer revolution actually also revolutionized the financial intermediaries that our custodian based so that you have point-of-sales at every shop etc interestingly we have seen a dimension of privacy preserving banking which still is centralized with for example inventions such as ecash by David shown and we can see that these these these blockchain these these payment systems have been hardly they have been really scalable so they have been used by millions and millions of users and have helped a lot of businesses too a lot of user adoption almost ten years ago now so in October

on the thirty 31st of October it will be ten years was Bitcoin invented by Satoshi Nakamoto and this is the first instance of non custodian based financial financial intermediaries so non custodian means that the user so the one that is actually doing a transaction on behalf of his own funds is remains the owner of his own funds so it's in its so the user any innocence remains custodian and the issue with these open ngspice functions that we have seen is that they do not really scape so the reason is basically that the number of processing capacity of current block chains are just it's just very low it's

because there's no single entity that allows you to control the network or the ledger so the number of unconfirmed transactions or the number of transactions that are floating around in the network is just can be immense so the settlement capacity is very low we just reflected in the unpredictable transaction prices that you can find on open and decentralized Ledger's such as Bitcoin or ether iam so they have hit on average rates beyond 50 US dollars luckily it's going down since then but if they go down this doesn't necessarily mean it's a good thing it could just mean that well there might be less users actually using the system and this is not what we want right we want mainstream adoption so my question it would be is there

maybe there the possibility of a new market segment for banks and exchanges so if you look at the current existing services of finance intermediaries well these are custodian and the perform transfers and exchanges as in the capacity in the capacity of a custodian now they have regulatory certainty because they're acting in a framework that has been well established over many years but maybe this these finance intermediaries can now add a new type of service to their existing portfolio of systems to address a new market segment which would be still transfers and exchanges of currency or payments however in the capacity of a noncustodial financing to meet you so

what's interesting I believe is that on the regulation side there might be simplifications there might be it's not yet true everybody knows we are in a rather rather uncertain regulatory environment in the blockchain space but I know at least from a few countries that if you are not a custodian then this simplifies your requirements regarding banking license regarding kyc regarding AML so how would this look what is the requirement to have a blockchain based non-custodial fines finance intermediary so I believe one of the first priorities is user onboarding so a user should not be encumbered with for example purchasing crypto or having to deal with with complicated private

keys so it needs to be easy and it should be like if you install your app you basically already in the network right so this should be as easy as possible it needs to scare to millions of users millions of transactions and more importantly a merchant would need to be able to sign you would need to get a service level agreement so for example if I'm a merchant I want to be able to get a guarantee that I can get at least a hundred transactions per second from a customers right Amazon would never use the blockchain if they can't get a thousands of transactions per second as SS merchant so service level agreements or to professionalize the blockchain space I believe is very very important

in order for the Box chain space to to be used by more and more businesses worldwide so how to achieve these properties well obviously we need to so we need to solve scaling right and there are different avenues and there are many excellent teams working on different directions notably on the consensus layer basically changing for example the longest block chain rule to a graph or on the sharding side so partitioning the network into smaller shots and having like reaching consensus within a shot which would be faster or doing off chain transactions so which one to choose well maybe there there you can actually work on all these lab together and they actually complementary that's that's the good thing so we are

working liquidities working on the off chain layer the main reason being that you can actually compare off chain to Cash if you're having a computer architecture are you writing every time that you are you writing to disk every time that you modify that you that you do something on your PC you know you have like random access memory you have caches on your PC and these are much faster than the underlying hard disk that you go back to for persistency right so you can you can you can see there's an analogy between off chain and the blockchain while often is you can think of it as kind of a cache so it's much faster it allows you really basically the number of transactions that you can get this is simply limited

by your internet connectivity between the peers piled on the consensus layer and charting layer there are a lot of efforts going on but we are focusing on off chain at the current moment and if you're looking at off chain I have to specify we liquid looking at off chain payments there are off chain state channels as well so state channel is something that stores a state of a computer program while payments are really focused and optimized for the purpose of performing payments or exchanges and I believe you can get for for purpose specific solutions are always more efficient than

general extra Lucian's right for example the ether and blockchain grows much faster than the Bitcoin blockchain why is that well smart contracts are much bigger than regular transactions on the Bitcoin blockchain right I believe the current issue in blockchain is about six is so beyond 60 no it's beyond it's it's about six times bigger than the bit conversion I'm not yeah roughly and Bitcoin has been around for more than five years than its real so if you're looking at difference of change scaling solutions you can see to party payment channels which is something that has been pushed many

by lightning and this is a great great way on how to potentially scale Bitcoin and the different solutions you can have uni-directional payment channels so you pay in one direction you can have bi-directional payment channels you're paying two directions linked payments so if you want to pay over a hop to someone oh you could think of building payment channel hubs so basically one node or several nodes that have a star topology so these nodes are very well connected within the system a new class of off chain payment scaling solutions would be n party payment hubs and this is what liquidity is fundamentally building this is the technology that that that we're

providing so let me go bit into details about what is the payment channel so payment channel is an agreement between two parties so by it can be a bi-directional channel if both parties deposit for example one ether and they can do their payments in both directions up to one either this channel establishment here that you see is on the chain so this is not enough chain transaction you can think of it like you charge a prepaid card right and then you can use your amount up to the maximum amount that you have on your prepaid card so let's say you send 0 0 1 if in one direction what you need to do at this very moment is to update the off chain balance which you can see here if you do another payment in the other

direction you will need to invalidate the former state and basically create a new off chain state so at the very moment that you want to close this particular channel you would need to settle the latest state right you need to invalidate all former states and it's actually important too that you remain online to verify that your counterparty is not cheating that it's not like submitting in all state and in that case you can close the channel and then you're done right so there are a few challenges that you can see that payment channels have so one particular is it's nice to have a payment between two

parties and and four payments between two parties it's I think chants are a great solution if you however want to reach more than two people like more than one person more than one counterparty you would need to find kind of a route to particular peer so let's say we have this topology right there are three channels so far the middle channel is almost depleted so there are very little funds in there so what can you do well you need to find another route right so let's say there's another peer and these are connected as such so then you can route a payment here across this path and reach the final note the more intermediate hops you have the higher the likelihood that

your connection or your your payment's gonna fail because these are not professional entities it's a peer-to-peer network there's a lot of churn peers are leaving pieces are joining they don't sign an SLA with your service labor agreement right the collateral that they have it might be empty it might not be sufficient so there's a lot of dynamic adjustment that needs to be done and it's expensive and moreover those peers there may be nice to you but they won't do the service for free so they want to cut so there will be a transaction fee for every hope that you're hoping about another issue is that if you look at the collateral required to build for example a hub system with two party payment journal

technology you would look at how much collateral does this central node needs to set up so let's say we have four peers and we we estimate an average transaction volume of like one ether to each peer well in this case might be four ether which which could be bearable but by this particular instance but what about if we have a million peers so if you're a hop and you actually want to scale mainstream well then we need at least a million transaction a million payment channels here for a million payment channels we need a million on chain transactions that's in itself is already quite a huge requirement but not only that four million peers and let's say an

average transaction amount of 10,000 US dollar to each peer we would need 10 billion US dollars of lockup for this particular requirement and that's a lot of money that you need to lock up in a hot wallet on the rock chain so it's it and each of these are tied to a particular pier each of these channel so you can't kind of use the collateral at another at another pier so in order to alleviate some of those problems liquidity proposed revive revive is is a technology in order to make to party payment channels more scalable and it relies on the idea that if you have a cycle in your graph of your payment channels and let's say we have the

current setting here so you have like one either the we are let's say we have the left most node you have one if the deposit there and there's a zero ether deposit to the to the peer that you actually want to send something to so you could go over the expensive route expensive because the intermediate peer will take a transaction toll or you could run revive you could revive the channel that is currently depleted so you make a rebuffed instance you use some of your collection on the other channel and then you go to the cheaper rod okay so that's a very simple example it can get much more it can get much more complex within yeah more complex topologies so there's peer reviewed

academic work that was published last year and the sources open up on the sources open on github liquidity is an extension of real life so liquidity and liquidity we built efficient of chain payment humps and we believe these are capable of offering professional financial intermediary services while not being a custodian and so not being a custodian means that we would need to rely on open and decentralized blockchain because otherwise if the blockchain is owned by an entity in the end its custodian and the first feature that I would like to show you is the user onboarding that we can do in

liquidity right now already it's a decoder so called off chain registration so with just having a scientist EP message you can reach you can join the network so you basically have your private key of your of your standard for example if you in client you sign a message I would like to join liquidity and you send this to the payment hub and at this very moment which is this is free at this very moment you can receive an off chain transfer from another peer so note you cannot send funds without depositing them so if you register and you have zero funds you cannot send anything which is standard however you can receive already off chain funds from someone so let's say I'm stalling my

liquidity app and now my friend has liquidity and here's some off chain or so we call it liquid ether then I give him five bucks and he sends me five bucks or worth of ether tomorrow to my phone or it could be five five bucks of stable coin right we support we will support ear see 20 tokens so the user onboarding as such is really easy because the user is not encumbered with any unchain ether per CEO or on chain tokens right so you can really stay within this environment of off chain and only if you really want to you can withdraw from the off chain world and go back to the untrained world and then for

example go to an exchange and cash out so after this option registration you receive some of chain transfer and you forward it to appear so what are the implications of having an non-custodial finance intermediary that performs transaction execution and also allows exchange operations so if you look a bit at the history of exchanges you can see that in some ancient times or nowadays even if you go to some marketplace you're always custodian of your own funds right but the number of transactions that you're doing is not very high so the stock exchange basically wasn't isn't it it's an exchange where you're operating with the custodian so you're operating with a

bank and it allows you to have quite a high volume but even in the stock exchange the settlement is typically done always at the end of the day right so if you're looking at centralized exchanges such as kracken buy-ins you can see that these are custodians right you have a high of transactions you have frustrating but you don't own the funds as a user right you know you're working with a custodian you're not custodian of your own funds and here's the beauty about decentralized exchanges they brought back the control of the funds to the users and there are many excellent teams are working on really great decks as nowadays the only issue that we see with

current taxes is that they won't scale because they're just relying on the underlying blockchain for the transaction swap yeah an order book might be off the chain but they where the transaction swap is actually happening it's again on the chain which limits you to the current limitations of opened its interest options so in liquidity as it's a noncustodial finance intermediary that allows both the transmission of crypto assets as well as the swap of crypto assets we can perform a so-called decentralized exchange which is off the chain completely off the chain this means the order book can be off the chain as well as the transaction swap can be off the chain and I believe

such an exchange could reach the speeds of centralized exchanges of custodian exchanges so in addition to the off chain transfer we have the option swap in liquidity so liquidity is implemented and operational on the assumed test net since March 2018 so this year you can try it on your own browser you just need meta mask so basically you can deposit you can open a connection so you have off chain registration then you deposit some amount of funds on on one instance and then you can transfer it to another instance that also did an off chain registration and then just forward it to

more peers so it's fully working on the onlythe um testament so to finalize the the presentation I would just like to iterate that we have a simple free and user onboarding process so it's option registration we can scale to millions of users and we are currently working to get them not ready along with mobile apps so that people can can can join there and we're planning on on issuing service level agreements as soon as the software is more reliable there will be additional scientific papers that will be added to to explain how the liquidity Network works as well as the the accompanying

source code that will be released once it's ready so that other people can also run payment hubs and we really envision a network where the multiple entities that can operate payment hubs and connect together right so I see it as a blockchain on top of another blockchain thank you very much [Applause]