 And we're live. Hi, I'm Dazza Greenwood, a scientist at the MIT Media Lab, coming to you live from Cambridge Innovation Center. Very posh, very innovative. And this is a prep video for the Open Music Legal Hackathon at the MIT Media Lab this weekend. And it's a collaboration very much between the Media Lab and the Open Music Initiative with our friends and colleagues and fellow entrepreneurs at the Berkeley College of Music. And I am very, very happy to have George Howard of the Open Music Initiative with us here. And he's been collaborating with me the last few months, prepping the issues and figuring out how to do an Open Music Hackathon and then make a legal dimension on top of it. So we've been keeping a few plates spinning and I think by Joe we've got it. And so George has joined us today to introduce himself, give us a little bit of an overview of what is Open Music Initiative and the goals and to talk about a learning session that he's going to be doing live at the Media Lab during the Hackathon during our afternoon session. So we can get a preview of that and you can see if you'd like to join him for that. And maybe if we're lucky we have time we can brainstorm a little or hear from George about what could be some, you know, the good types of things to do for hacking projects or even small discussion groups. So with that, thank you. Thank you for being a collaborator on this event George and let's make the most of it. Thank you so much, that was overly kind. You have done the heavy lift here and we are grateful to you, grateful to MIT and Media Lab and, you know, grateful for this opportunity. It's a really important time. As a bit of background, yes, as one of the original members of Open Music Initiative I'm kind of standing in here for my colleague who is the founder of Open Music Initiative, Panay, and he's a hero of mine on a lot of levels but he's been able to push this initiative forward at a time when it's just so crucial. And he, you know, Open Music Initiative is a non-profit it's a joint effort originally founded by Berkeley and MIT and the goal of it is really to try to create these open source protocols for identification of music and right towards the creators. And that's a big lot of language but if I can try to make it make a little bit more sense in plain English. The music industry is a complicated one and it's commonly complicated from a legal perspective because songs unlike other IP axiomatically have, two copyrights have inception. You have a copyright for sound recording and a copyright for composition and that the failure to match those two. So I'm Dolly Parton and I write a song called I Will Always Love You but then X number of years later Whitney Houston does a version of it. Two different rights holders there probably more than that if Dolly has a publisher Whitney Houston was signed to Arista Records and contractually, Arista Records is not legally allowed to issue that work without an agreement in place with Dolly. And so in the right soup of 19 whatever when the Bodyguard soundtrack came out makes today's right soup look like I mean it's just it's been accelerated by many orders of magnitude when you weave in things like samples typical Beyonce song or Drake song might have 11 or 12 different writers on it. Once things moved from analog to digital tracking them in my mind should have become easier but it's become more challenging in the net of it for me and some of this is just George speaking but it's consistent with OMI is that because we don't have these interoperable databases we are at minimum not opening up all the possible ways in which musicians and other rights holders might be able to better get their music out there to have impact and profit, et cetera. And in my personal opinion, if we don't address this we're just gonna see further consolidation around the music industry. The popular artists will just get richer and more powerful. The Spotify and the major labels will continue to dominate. And at the same time the Amazons, Apple, Google all of whom are using music as loss leaders to sell high margin products will drive the perceived value of music to zero and it just leaves out a huge swath of creators. And importantly it also does not facilitate innovation if you look at something like SoundCloud which have a billion dollars of investment capital into it and that was sort of sold for $150 million because they couldn't get their rights issues in place that sends a chilling effect into the marketplace for investment and incumbents just kind of keep doing what incumbents do. So it's a critical time and Open Music's whole sort of mission is really to at least explore these things look at better ways for interoperability we don't own databases, we won't own anything we are a coordinating agent we try to put and again I give all credit to Panos put the right people in the room and the right people include the big stakeholders now whether that's Netflix or Warner or YouTube or all the publishers but it's also the artists and it's also the students and it's also the entrepreneurs who are an adjacent or orthogonal spaces and try to make everybody come to a place where we don't always see things as sort of adverse and we don't see these databases as being proprietary in the sense of our only value additive process. One of the problems of the music industry has been all these siloed databases where everybody has a database they don't talk to each other and the perceived or real value of the databases being siloed has led to a real, real lack of sort of innovation and a little bit of markets. And so we're trying to work through these processes and we're doing so at a time where as many people know the Music Modernization Act was just signed into law now this is something that many people feel is a real step forward and it may very well be but as part of this law you have a mandate for new databases and it's important to remember. Actually, I just wanna use my three favorite words from the statute and it is a legal hackathon so I think I'm entitled to read do some blue book citation later maybe. Oh, public accessible database. Yeah, perfect, perfect. Public accessible database and so, look, we're on the JD too. So when we talk about words have meaning and we all know as lawyers or law students or JDs or whatever that you try for precision but there's subjectivity. And so public, right? So that would mean, does that mean available, accessible to the public? Does it mean owned by the public? And so one of the big threads of open music that's galvanized has been around decentralized databases and the buzzword of the day and it comes up maybe more than I'd like even though I've been a big proponent of this technology from the very early days is this sort of idea of blockchain-based technology and one of the key tenants of blockchain is a decentralization element even though increasingly there are many proprietary blockchain databases. So, but to your point, public and open and accessible, those are all abstractions and there needs to be governing protocols and governance and who gets to set those standards. And again, we have since our inception attempted to put forth best practices around standards. And so I hope that this hackathon moment can be from a legal perspective look at issues around, okay, these two rights from the sound recording the composition, how can we facilitate transactions between those? Because again, another big piece of blockchain tech is this concept of a smart contract. Now we as JDs know the contracts have certain elements, you know, there has to be sort of this barge and force change and meet in the mind. There has to be consideration, there has to be capacity, right? And if you have those elements and certain contracts require writing some certain don't but a smart contract in the blockchain sense is not necessarily a legal contract. It's more sort of it's this than that rules. And maybe the best sort of heuristic around that would be something like the creative commons, which I think would be a worthwhile thing for people to familiarize themselves with and advance if they're not. Creative commons is a way to unbundle the six exclusive rights that one gets as a copyright creator in exchange typically for promotion because that then solves the consideration element at a time when you just can't, you can't exchange funds in a scalable way. Let me try to say that better. So creative commons is a really, really beautiful concept. It's been used by a billion people or so. And what it does is I'm an artist and I can say, okay, look, you can now use my work. You can sample, et cetera. In exchange for that, you agree to give me credit and there's your consideration. What one way to look at it, and I've written about this in the article is blockchain that could facilitate a creative commons-like effort, but in a machine readable and thus scalable way. So I could say I'm a songwriter and I have this work and I'm gonna ascribe these rules in this work and make them machine readable. And someone on the other side could search for that. And then if the rules are satisfied, it would self-execute via this sort of smart contract. And then importantly, another blockchain component that's very important is immutability. A record put on a blockchain is there. You can't take it down. You can't change it. It doesn't necessarily mean that all of the elements would have to be transparent. However, it does provide an audit trail. And auditability in the music industry has been a bugaboo forever. I mean, I was just, I did a talk at Suffolk. And I was there with someone from the coffee right on. In University Law School? That's a model and it's a great place. I'm gonna strong with technology nowadays. What'd you talk about? We were talking about, it's a similar kind of concept. So in Europe, they have something now, a mandate called the digital single market. This was something that was kicked around or originally put forth in 2015. And it's just so amazing the way times have changed. If you think about 2015, we were still pre-Brexit. We were still pre, let's build walls around things. It was much more the Euro had not begun to sort of have the stress that it does now. And the gesture there was, look, we can create efficiencies if we remove intermediaries. And so they wanted the gesture is, let's have a single market in which people can have portable technology. So I can buy or license or gain access to a piece of technology in Germany whether it's a Netflix movie or whatever. And then I can also have access to that if I go to Sweden. Because you know you can't now, right? And my Netflix account in Germany won't work the same way. Similarly, they put in place a set of standards, section 13 of the digital single market essentially mirrors our digital millonium copyright act, right? And the safe harbor provision thereof, which stipulates sort of guidelines around user generated content. If I'm YouTube in order for me to stay in compliance with the DMCA and people are uploading content, I have to have a technological solution to identify infringing works. I then have to be able to give notice, et cetera. And if I don't, I'm now outside of protection and the lawsuit will come fast and furious. The problem for the issues around this is those types of systems are expensive to build and to maintain. And so therefore they could be seen as favoring the incumbents. And if I'm a startup and I have to build these same systems, that's a budget item I don't have. So there's a lot of back and forth. Jim Berners-Lee is not in favor of it. Jimmy Wales is not in favor of it. So these are the types of tensions that we should really be talking about from the broad base. But many of these things go away. And I'd like to put forth this as maybe a sort of, at least a thought experiment as we think about what type of projects to hack during our sort of learning session is that one of the big problems and challenges of the music industry has been it's a system built on blanket licenses, compulsory licenses. So for instance, if you wanted to record a song that I had written, make a cover version of it, I can't stop you, right? In the United States, I cannot stop you from doing it. And I also have a ceiling, I cannot charge you more than X for that license, right? And that's a very, to me, a very unusual thing because in what other industry is there a ceiling where I can't say, you know, so, but they did that. They did that in Soviet Russia, which one of the features of their kind of command and controlling heights of the economy, but it is an unusual work of free market economy. Free market economy, you typically wouldn't do that. So you make it my point, right? I mean, is that the society we want to live in? In any case, so, but of course, if we were able to contract directly rather than having made a statutory license, then we could set whatever price we wanted. You know, one of the things that I say frequently is we don't know what the fair market value of music is because there's too many price-setting elements done by governmental institutions, the quasi-governmental institutions, but we do know that the market price for music currently is unfair, right? And so, so between that and then for instance, also like if my music is played on the radio or my music is streamed online, I'm not, there's no bargain for exchange there. These are either governmental or quasi-governmental like performance agencies, ASCAP, VMI or whatever that are sort of setting the prices on my behalf. And that made sense in an era in which you had to sort of model rather than measure in the same way that people would come to your house to read your gas meter once every, you know, six months and then they would extrapolate that data and price you accordingly. Now we know exactly how much energy you use and so we went from modeling to measuring. We can do the same thing with music. We could measure our usage in a very exact, but it may not benefit the incumbents to do that. A blockchain-based system in which we could say, hey, look, I'm gonna set a price for my work. I'm gonna make it searchable through metadata, what have you, and then buyers can find it. And if they buy it enough, you know, supply and demand, I'll raise my price. If they don't, I'll lower the price. And we create markets that way in direct licensing and that would really obviate all of this. And so I would encourage people to think about the projects that they're doing in that way. Think about, for instance, rights holders and look, as my bias is the incumbents, the big artists, they're fine. They don't need your help. You know, anecdotally, Universal Records is making $10,000 a minute just from Spotify. They don't need your help, right? And Drake doesn't need your help. The people that need help right now are the artists that, yeah, sure, they have access to the market, but now there's a new set of gatekeepers. You know, getting on a Spotify playlist is the analog to getting played on a radio station five or 10 years, 20 years ago, whatever. So we're kind of, you know, it's like, look at the new boss, same as the old boss. So I'd encourage you to think about creators from the standpoint of, I'm a writer, I'm making works, but my access to the market and my ability to sort of monetize those works and or control them is very, very abstract and very, very intermediated. And on the other side of the market, say I'm somebody that's interested in creating an application that uses music for help or something. There's no real easy way, and I know it's from personal experience, to get that type of license via a Harry Fox or ASCAP or any of those places. They're not set up that way. And therefore I can't get the music that I need with that risk of lawsuit or incredible expenses. So you have a marketplace that's very inefficient. You have a willing seller, a musician that's like, sure, I'd love to have my music used and to help that. And you have a willing buyer. I'm an entrepreneur, I want to use music to treat Alzheimer's, I need music. We should be able to search for this, right? And find a love match and set rules and it should be able to be done at scale. And we have the technology to do this. And through blockchain based technology, I'd encourage you to look at current blockchains, our API, other APIs and build around this and the legal perspective that you want to look at is, okay, you have these rights and they're not going to change. I would strongly recommend that you familiarize yourself with section 17 of the United States Catholic Code and the certain elements that delineate the enumerated rights. If I own a composition, right? I wrote a song, I'm Dolly Parton. I have an exclusive right to distribute, reproduce, create derivatives, display and publicly perform. In advance of these sessions, you need to know what each of those mean. I've written about it, I'll send links up. On the sound recording side, I'm Aris to records. I have the same exact right, but there's no exclusive right for public performance. So if a song gets played on the radio, Dolly Parton gets paid, Whitney Houston doesn't get paid. Everywhere else in the world both get paid because of something called neighboring rights. This is crazy. So look at those rights, figure out how buyers and sellers could more efficiently create use cases for these works outside of traditional things. Outside of, again, we don't need help getting music on Spotify. We don't need help. What we need help is if I wanna create my own Spotify. I mean, we shouldn't have three dominant streaming services. Everyone should have one, right? You, I bet I would love to know what music you're listening to. Why is it that you can't make your works, what you're listening to publicly available? And why can't I subscribe to that? And why can't we create incentives around that? Tokenization is something to look at, right? Tokenization is a concept that's getting traction dominantly because of crypto, but it's not a new concept. It's just gamification using some sort of alternative form. A rewards card is tokenization, right? Why can't we coordinate better so that we're rewarding fans for curating? And then when those curated playlists that a fan makes gets licensed into some restaurant, that that fan is rewarded, right? So there's all sorts of interesting things that can be talked about with regards to rights issues, with regards to contract issues, with regards to fiduciary duty, with regards, you know, and I see those are the types of things. So on that note, you know, listen to George everybody and like seriously think about that and I'll do my job and extract from his remarks on some of the best of including excerpts from section 17 and what those enumerated rights are. We put them on the hackathon page. And MCA is 512, the DMC is the 512, section 512 of the Digital Money Copyright Act really lays out the state department position. And go to legalhackathon.org and then if you click on resources, it's a little bit down our event page, you'll see what we've got tutorials that have made over the last couple of months with a number of blockchain and other technology providers that have some open source that you can use to get started. Enddoor is one of them that's for analytics and we've got some nice data on like how many plays music or art chain coming. I just wanted to run through the tools that are there. Our chain and our song, they've just flown some people out. They're here, we've got a demo of how to use that as a way to track rights and to make things available. Here's another tutorial and just last night, Monax presented in New York last weekend and we've got a tutorial for how to use their new Blackstone open source release and to do a co-writing agreement if you're working with other musicians. That's the use case that we showed but you can use it more flexibly. Open law also has a demo here. It's a consensus node for how to do music licensing and very creative ways. And then finally the Accord project has smart legal contracts where you can, both Open law and Accord project let you kind of show the legal terms on one side and some of the kind of machine processable automated terms of the smart contract on the other side and they're template driven. So there's a few tools that you can start with as you're starting to kind of think through how you would actually achieve some of the types of business models and the legal framework technology that George just implored you to try. I'm sorry to have interrupted you George, what were you saying? Oh I'm so sorry to interrupt you. Those are great tools. So I know the Archangel is very, very well and who's coming, is Greg Mayer coming or who's coming? Kayvon is here, someone named Aura is here and I think there's one or two other people that are coming for Sunday. Yeah that's cool. And they, you know, they, yeah that's great. So some really, really wonderful tools and you know, the Archangel is essentially, you know, their USP is it's just much faster transactions which is a big problem with blockchain, right? Scalability issues. So Ethereum is struggling the way it is right now just because, you know, is a scalable, I think, you know, SEC elements too and I don't know how much you're gonna get in the ICOs and those sorts of things, but not too much I hope. But if so, for those of you that want to explore Archangel I strongly recommend that you will have people on site hack that can hack with you. The tutorials are perfectly good if you're another site and there's a number of them around the world happening this weekend. And in the second, in the second video in the Archangel section, I took a while, well, we had the guys here yesterday to actually highlight some of what George was just mentioning. I'm speaking, you know, whatever, through the fourth wall now to you, the audience to George but you should know to George, we went through the composability of it and the extensibility and how they basically can scale up very quickly using Kubernetes. And so these properties could be really essential when you look at the deployment of a production scale, you know, music machine. Yeah, no, I'm so glad that you mentioned it and it's because like, you know, like with any new technology, right? You get with Clark's Law Rights, sufficient technologies are distinct from magic. And so we'll know blockchain has adopted when people stop saying blockchain, right? Nobody says TCPIP. One of the reasons that people say the internet, right? But the reason that blockchain has struggled and I just did post, oh, I should send you this too because I just posted just a recently Forbes article on the four points of failure around blockchain startup elements right now and it's a two-piece thing but I can send you the second piece. Maybe I can get it done before Sunday. Maybe it's done, try to get it up. Yeah, we'll wait for a learning session. Go on. Yeah, but one of the issues is that like, people are overestimating smart contract technological efficiency right now. It's just not terribly efficient, right? And so if you ask too much of it and this is really important in the hacks, right? Smart contracts right now are very good at sort of, if this than that single sort of deviation type thing. If you start saying, okay, let's do something to solve for samples and what you, it's not if this than that, it's if this and then this also and then this, this and then it just breaks and it's too, it doesn't work. So you have to be very careful about using sort of controlled compositions. Archane is a good example of kind of Moore's law advancement, right? Like right now things are slow but if you believe in Moore's law generally and have to be kind of not, in this case, like you know this isn't gonna be a problem forever any more than the speed of the internet wasn't a problem for very long. So that's a really, really good sort of. Speaking of problems or challenges, you've asserted that you're Dolly Parton at least a couple of times. Identity is one of the issues that you know. How do we even describe the identity of a musician in a way that's discoverable and unique and you can address for, well, tribute to songs and credits but also to get royalties to. I think that's another sort of eddy of this whole thing, right? Yeah, it is for sure. Yeah, yeah. And I mean, look, a couple of things. One, there is a sort of immutability quota. There is sort of proof of stake. There are ways in which that we can, I mean, identity is a big bug of the world right now. And so there are ways that we can combine sort of technology as well as sort of reputation scores, right? One thing to be very careful is that blockchain doesn't obviate real law. In fact, we could say it's gonna put stress on it because it's gonna increase transactions but it does provide sort of an audit trail. And then there are technological solutions in which you can put sort of proof of stake there. And over time, you can develop reputation score in the same way you might as like an Amazon seller or something. But it is an issue for sure because people sort of equate Bitcoin at least with anonymization and that doesn't necessarily have to be true. I will say to be careful also about metadata, right? Which is a sort of maybe a second order issue to what you're talking about. But a lot of people are saying, well, look, the issue that we have to solve is are we matching Dolly Parton with this or the other thing, the writer and the performer and all of this. And again, if we think about it, like one of the first killer apps for the iPad iPhone was Shazam and that was 2007. And again, thinking about Moore's Law, we know that we have technology that identify sound recordings very, very well, right? Where things fall apart, but on a content ID session and any other type of identification is identifying a composition. So, but we know that with some sort of combination of AI and machine learning and human intervention, we should be able to catalog and rapidly identify down to the STEM level of everything, right, of every song. And at that point, metadata goes out the window. Now it's much more about recognition via, okay, no, I not only know that this was Dolly Parton who wrote this, it was whoever played base on it and that type of thing. And you can create incentives to do that, right? I mean, what is Wikipedia but a sort of validation mechanism for people to think they know more about things than others and then away an iterative improvement. And by the way, Wikipedia has an audit rate. You can see chain vlogs through that. So if we were to be able to incent people to say, look, I'm going to provide information about this song, this work. And then over time, give them benefit either through tokens or just validation. I mean, nobody's getting paid to be an editor for Wikipedia. I mean, it's a very small group of people and yet people do it all the time. And over time, you get some notion of truth around that. We could be doing the same thing for musical works and then through smart contract distributing better, right? So that's an important element to consider. The other thing that I had asked people to think about is right now, the underlying metadata is structured to work, work such as it is for text-based search, right? So I will type into a search box. I want to hear this song or this whatever. But we both know that text-based search is on the way out, right? I mean, now we're searching with our voices and we search very differently from our voice in the current with our voice that we do when we're typing something in and the metadata is not ready for that. And so we're add to my mind the precipice moment as people are demanding too much of the metadata because it's structured for search via text, but now we're searching with voice, that ain't no work, right? And then that gets you to sort of preference engines and why are preference engines so bad right now in algorithms generally, right? Where it's like, I mean, I did an example in class one time where I was talking about how Spotify is trying to decouple artists and songwriters or whatever from the playlist. They just want people to say, hey, give me music to cook by or whatever. So I said this, I said, hey, Siri, give me some music to cook by. And it said, okay, it's a little wheel spun and then it said, okay, we found something for you George. Here's the song from the band called Spoon. And Spoon's in Austin, Texas, like indie rock band has nothing to do with cooking, but the algorithm saw cooking in a band called Spoon and put two and two together because that's a failure of the algorithm, right? And the problem is like we all discover music exactly the same way. Our friends tell us about music, right? And algorithms aren't our friends. And also in this society right now, which I think it's a really good political time to sort of think about this in the Me Too era, who's dominantly coding these algorithms? Algorithms don't spring fully formed from the earth. Algorithms are coded by humans and humans have implicit bias, right? Well, who's dominantly the computer scientists and the engineers who are writing these algorithms with white males? So now we're tuning algorithms to be informed by our gender and racial biases, however consciously or otherwise, right? And facial recognition. So all of these elements sort of all come together around issues of identity, around issues of search. And those are the types of thought experiments and those are big, big, hairy audacious goals and broad things, but have at them, right? They do go to, that's the type of thing that we need to be thinking about in a hack. I'm talking too much. Here, here, well that was excellent. And if any of you are interested in following up on that final thread that George mentioned on, in a sense the ethics and the social implications are algorithmically based, data-driven, model-based processes in the music context. Tap me on the shoulder. There's some very interesting things happening at the lab. So if you're here in person or you can telegram in if you're at one of the sister locations. We've got the Algorithmic Justice League. We have the Assembly Project. We've got a few pretty pioneering, thoughtful approaches to try to address and balance out or counteract somehow these types of key biases and other unintended consequences that slip into our algorithmic based decision-making. So with that said, I wanna just thank you again, George. And thank you Open Music Initiative and thank you Berkeley College of Music for being our collaborators and looking forward to your learning session on Sunday afternoon. And everybody come on out. Come hack the law with us and let's see if we can take back music for the people. Oh man, from your lips to God's ears, right on. Thanks, I can't wait. Thank you so much, David.