JAAK: A New Engine to Power the Web of Content

alright that works hello everyone it's me here so where I'm for Mackenzie I'm from Jack if anyone knows us we are in the music and media space the name of the presentation today is a new engine to power the web of content and I just really wanted to start first by by kind of acknowledging that you know in terms of music there are a lot of music blockchain startups in the space so it be a good idea really to just kind of recap why we're in the space and why we really think that music and blockchain really do make sense and yep take a word so first of all what is the problem and I think that the way that most people as consumers experience music is that we just think of a song as an mp3 so when we play a song or we hear a song playing we just think that's a song play the song that's what I need to worry about the problem with music is very hard to figure out who to pay and that's because a song is actually kind of complex so let's say that a few people get in the studio and they start working on some music you actually are creating a very complex ownership structure the reason why is a song which actually is not represented in the music industry don't really think of songs they have completely other terms to describe things songs are just for consumers but you have this thing called a recording there's a copy right now and that copyright is normally created when an artist makes a recording so if I go in a street and I record something I'm now I create this thing called a recording and that is not only owned by a label labels aren't recordings and in compositions you can think about recordings being the implementation of a composition so a composition or as according the us of work is really just the the nuts and bolts the mechanics of how this song was made the lyrics the melodies and so on and that's made by writers so producers you make the instrumentals or obviously the people who just write the lyrics and that's normally owned or represented by publishers and every single time when you publish our writer comes into a tracks you can have five producers ten producers five samples all of those things have their own ownership structure these are like micro businesses everything one of these are there they're argued in in in by lawyers and figured out my managers and accountants to try and create this structure and nobody has a real picture how this works now this is one song on Spotify there are over 30 over 30 million songs so you can imagine how complex that is to try and keep track of and the way we keep track of that is with proprietary databases now this tries to explain how when we think about how music goes from people in a shoe you're crazy all the way through to a consumer listening to on the other side it's actually quite complex because the song itself is a collaboration it's then pushed in the metadata and the music becomes separate as they always are and then they move down the supply chain in different ways now on the work side there is a standard identifier called a nice that we see it on the recording side a standard identify clitoris is our see people like to apply these things willy-nilly they just applied all over the place and it really mapped down to the actual unit we're trying to describe and as we want to keep sync between these databases we create these exports and we just send them around so this is flat files flying all around the industry and this is what we're kind of using to sync right we're just using these flat files now as we were saying before the problem here is that we end up with complete information asymmetry one person saying one thing another person saying another thing anyway to keep those two things together is to send this flat file and that means this fragmented content rights across all these databases and that means that even inside one organization which could be global and owned and controlled different things and assets in different territories it's quite hard to keep an internal consensus in that company and really the only way the music industry has known how to solve this is just to create a new database so we end up with more and more and more fragmentation and really we ended up with more and more databases and then we end up with parallel data reconciliation so again most things are traveling around via XML and spreadsheets and getting delivered to FTPS there's not very many API is either so actually syncing is very much manual and that's problematic because we are using proprietary databases and access is ownership if I want to tell you what I own I have to give you the data so it's quite hard to build out kind of one single database however one day the music initially did try to do that there was one big global database and it didn't really work for reasons we won't go into but centralized governance is problematic if I have ownership I have control and really that's why blockchains are great solution to this problem they help us avoid it we have a network of unbounded parties we have the need to collaboratively maintain a single dataset and we want to do it without putting in a central authority this is why block chains are great fit so what Jack what we're building is an open data network and we've kind of start to coalesce around inside there of a single source of truth and think about block chains as single source of true systems and actually thanks to incentive structures block chains a highly effective single single source of true systems and a great our highly effective knowledge production systems so when we think about single source of truth we try to think about it in the way that it's always been defined in enterprise and that is say if any one of us are working with an enterprise or a consumer we're known in the enterprise to one department as say a customer and a different person in a CRM and a different person to cells and so on and enterprise service bus is helping us to try to create a single source of truth about who that customer is that's what we believe block chains can help us do around rights so using the theorem ecosystem and web the infrastructure we can power a new paradigm for content on the web a single source of truth for content rights built on top of an open inclusive baiser network which you call meta now when we were coming up with this we ran into a few problems the first is a single source of truth we have to define truth and truth is pretty subjective in the music industry because data changes all the time and no one actually really knows if anyone is right he was on that song I wasn't in the studio how do I know that creates a problem and what we've tried to do is kind of a kind of a proxy for truth and what we've kind of moved to his authority so Authority requires really robust identity structure which kind of led us to building out this thing we call meta ID so you'll see here on the Left we have this thing called meta ID and message basically an open beta network that can be used to build decentralized databases based on mass IDs where participants are incentivized to provide trustless services to the network such that they can be automatically penalized by providing proof to a small contract if the surface hasn't been provided if any of you are in the swamp talked earlier on or the one earlier in a week you've heard week two talking about sort of swindle and a service layer of a structure that's exactly what we're we're implementing here so with the mess hiding the provider can convert face into metal objects and push them into the networkers meta streams then anyone can provide generic indexing services that's what the validations are doing here they basically enter into a service contract with a smart contract and they provide service in the network and using Merkel proofs a node can just check that the actual information they provide as a meta stream is included in that index therefore without graphical API we can basically query that graph of objects and serve it up to adapt now the interactions between contributors to the media graph DB so moving ahead a bit quick in respect to music media rights information we can use the method network to implement a decentralized database which has a defined schema and it's maintained by validators the integrity of this specific database the interactions between contributors to the what we call media graph DB here which is basically participial to insert or update data into this database and validators is governed by a swaps western game rather they just put up a stake which can be claimed or burnt if they fail to perform their duties much like proof of stake now what that allows us to do is build out this object which we call smart content using message media graph we basically create a new index which is called smart content and the validators work to basically make sure that everyone conforms to that schema and it extends the media graph schema to include assets licenses metadata and anything else required to facilitate a commercial transaction so you can see here we have our graph and basically this is a swarm manifest and it has a kind of a Content address link that we can just find it and it returns the rice information the type of right the percentage the actual owner which is a Messer ID and you'll see that we have these checks here and checks basically just to show us that the person who actually owns this thing has signed it or some delegate thereof and that's really what the indexes or validators Herat do working to maintain to make sure that this thing is adhere to so this structure facilitates trustless kickthrough license licensing using small contracts and service guarantees using swearin Swindell contracts where we get usage analytics and automated real-time payments for free so what are we working on at the moment a few things so if you'd heard before the j-team has been contributing quite heavily to swarm and we started off by working on PSS slide earlier in the year PSS being a routing protocol which again is being spoken about just off this talk in the breakout hall and PSS is a routing protocol so meta nodes can communicate to each other and broadcast updates across the network we also worked on a network simulation simulation framework which allows you to view the nodes that are in a swarm and then we've also begun work on pot and pot is basically the infrastructure or technology that will be behind swarm databases or swarm DB and we'll be working on that alongside a few of the other projects who are engaged as a swarm team over the coming months we've also been working on Phillippi centralized applications what you're seeing here is an application that we were calling Jack for reciting naat it's really called the smart content app it's basically adapt that allows you to build a smart content and manage it so theory would be you enter into the network with your master ID and you're able to load up any of the assets you actually have already built in and then you can customize them add licenses at assets so on and push them into the jack network and it uses pretty much many of the components we've really discussed small content indexing media graph index Messer swarm ENS and aetherium just kind of trying to use the entire stack so Lisa's first thing is the Mentawai paper we've been working on it for a while now some people may have really seen her first draft already and that really defines the structure that we're talking about here in terms of the service level agreements between nodes which which is quite powerful because it would allow things like licensing where you just enter into a contact with the license and can provide all kinds of services including the actual storage serving up the streaming of the assets again I use a little live peer people talking about that as well and this kind of framework really I think will drive out all the next generation of that development the second part is the method up and the metal ID really allows you to create a public address and then connect all these other authorized providers to it so you can imagine signing up as an artist with a mess or ID and then being able to connect any of the other trusty providers like your label or your publisher or you know your Instagram anything where you like to dock your identity and then the meta test net which allow you to start building these decentralized databases we plan to push that in early 18 so you'll actually be able to try this stuff out a final hat tip Tom see on the CEO I'm building other stuff so I once wanted to let you know the people who you can speak to if you're interested in learning a bit more fred and victor obviously my co-founders for his lead application engineer on jack victor obviously the architect working on Jack and Messer who is Marshall who's heading up Messer he's the best person to speak to about all the stuff I just described and Oren as well he's also working on Messer let me have Lewis who sees the developer working on PSS and Luke who's who's heading up meta ID so again if you do have any questions and you do wanna reach out those guys please you just go and reach out to them and you can find us on we do have a github github.com / meta - network there we have most these things defined conceptually we have documents we also have the mess ID framework the first version of that which is open source we plan to open source more things over the coming month so please do check it out we obviously are on git er so again just get er with the same thing slash lobby and then if you do want to join us on slack and talk about things that may be less technical community don't metal - network dot IO and you can just jump into the sack there and it's because any other guys yeah that's the end of my presentation and thank you very much [Music]