 beautiful okay all right okay so hack pencils down pencils down hack completes for now does not be over forever keep hacking after the event but right now see how far we got and so we're gonna have the next is we're gonna do quick report outs from the learning session discussion groups and hacking teams and I'm gonna ask everybody that was every hack into and have at least one person come on up to the podium here and we've got your slides already and so we'll just go through the slides and when your team comes up in the next slide come on up and speak to what happened who here needs a to plug in their own computer for something Kevin and cave on as and cave on okay that's not confusing okay so bring your computers and if you both independently verified that your HDMI friendly with this and you have okay terrific that is a political actually status HDMI friendly and it's gonna be ballot measure 12 so so basically and then what's going to happen next is we're going to sort of smooth on in from the report outs of the hacking teams into George Howard's basically four points of failure for blockchain in the context of music industry transition to a digital age in the open music initiative kind of motif and then we're going to go from that four points of failure and then further smooth on out to a facilitated dialogue and George and I will kind of go from the themes of your projects the themes of his talk and then we want to hear from really all of you at least once unless you know that if you move to say something please do and we want to kind of work with you and to co-think and address some of the challenges and opportunities and big questions that that George will help us to crisp up through the course of his talk sound good yes so you can't see it in camera lamp but there is an aggressive nodding of the head and we have an on-site chiropractor so we should be fine okay so let's see what's going on here all right so the first first project is Endor for the music industry this is so worse I think this is used up could you do a quick report out of what happened in the learning session for those of us who are unfortunate enough not to participate in it like if we were running an event that we didn't hear that look come on up here and we'll get each report out from the podium and then return to the circle here you go stop and door thanks first of all thank you for participating this very live discussion that we had a lot of raised and I think that we answered some of them you're welcome to address me after the session we could continue discussing those new questions that you might have but the idea was to talk about the way we can take data and you can create insights based on analytics and mainly thinking of maybe instructive ways such as and or technology to make it faster and make it available for everyone so kind of taking this point until the end and one of the things that we've talked about is how this problem is very very relevant to mainly all the organizations that have data have the sparsity of data scientists that do not work on all the business aspects of the company that want to grow their business from those insights and how we can do that better and faster with utilizing tools such as and or engine so once again thank you for for the time for the discussion I think that thanks for the whole indoor team for coming out sorry machine learning other days it's media lab metal layer some nice use cases here just quick follow-up question stop 50 cent did you email me your slides yet thank you so we'll rock on the spots slides back into the deck where can I find the deck DASA legal hackathon dot org invented handsomely right at the top near the top of the page moving up we've got vibe chain so by changing but come on up time to present so what we had to do was do search for our chain so for our song sorry so there's the catalog of songs on our song we want to be able to make a query from a mobile device and get the return assets from from our song so that's basically what we did the basic idea again you do a post to our API I think it's from a mobile device that goes to our song which returns an array of song that you can then display and let the user select and play music we try to get now we've got the first I guess layer down by you can put in one query item and really turn one song so that's I kind of the as far as we got came out to show you the demo thank you I'm a geek so everything is gonna be pretty deep in all black and white and we just haven't had time to develop the UI behind this but basically today you could go with a young Android or iOS go to your app store and download our song for your mobile phone for your device and we're one of the first companies that's actually storing the musical assets on on chain once you download the app then you could start playing with our catalog currently we only have four items there which is a pretty big release there isn't a lot of new material but there are musical assets that are coming on board so this project is about being able to search the catalog and again the catalog is on chain it's in our version of blockchain which in our case the implementations are done so why don't I start by showing you what the search criteria looks like since I didn't want to do a lot of typing I just put everything in a in a curl script and all of this by the way is in GitHub so if you could see this we're basically doing a post and in this case we're going for an album with the name tiny human and the song title tiny human and we expect a metadata that's associated with that to come back so all I'm doing here is basically executing that curl command and passing it through through a jq to get a to get a nice json representation so that's a payload that will be coming in for the purpose of this exercise we've got a song back the idea is tiny human and that's what we're after it has the song has two different two different it comes in two different effects there is a stereo version which is a typical presentation of music that you use with your with your applications and then we also have the 3d effect and it's one of those things that you actually have to listen to yourself it requires it requires hard headphones that are actually wired in it doesn't work with wireless or it doesn't work well with the wireless I should say and with the 3d effect you could actually feel the music traveling from from one ear through through your head to the other side it gives it gives that effect which is really cool once you start listening to it in any event going back to this so this is all the matter that is associated with that and you also see the uri for different assets that are that are part of part of this this this music product in this case there is the there is an immersive version which is the 3d version there is a stereo version and finally the jpeg version we decided to do it to do the metadata in this way because that would allow the mobile app to make concurrent calls and try to satisfy retrieval of all these assets in a power over that would be my presentation I'll be happy to answer any questions okay so this is so what I think would make sense to do is now that you presented it's kind of you can go to the next to the other two presentations here from George and then let's get some feedback on your presentation but they're a little bit of feedback that's appropriate now is put your hands together for that that was all done on the fly pretty good so there's an implementation of our song the Arling and our chain infrastructure to this very interesting use case provide change okay so let's see who is next thanks for putting that in there so next we've got on go okay so Kevin Sandoval is going to join us James who's the computer science grad student at MIT that worked with Kevin had to run so he's here in spirit thank you James for coming and packing with us okay let's hear about I go and I go project Mike yes so I came here mainly because I was really interested about being able to sample anything and not get sued for it so James really helped me just think about how to how to build it but essentially just from a very unsophisticated you know a quick take this is kind of like the mock-up that we did of what it would look like a very not friendly way but it's kind of like you search up a song Miles Davis so what very popular jazz song you play your stream you buy it more importantly what I'm more interested is like a sampling license and you would click on that and then it would give you the options you know you can number one remissing sampling which is where I'm being a concern about but also you know the commercial aspect of TV shows movies advertisements and then other whatever else you want and then the licensing fee which would be you know X amount free $5 a thousand if you're Kanye West and believe it's worth that much and you can be done to fiat credit or crypto and then you have the options downloaded right there or send the file to Gmail your drive whatever it may be and then you would just pretty much kind of like a smart contract I guess have the licensing agreement of you know miles agrees license song to produce a dealer for X amount and X amount of revenue generated by the new song you know revenue splits and then that I pretty much pay I agree to pay that amount and just share the proceeds with with this and that's pretty much it yeah not yeah that's it's it was like half an hour that I had can't really do that much but it's just it would be in color this is like the very unattractive not aesthetically pleasing version of what it would look like you know but any questions comments about it like that's that's kind of came up with perfect and so so let's we're gonna get into deeper feedback and this touches on a lot of things I know George got a raise but hey the first and foremost is I think round of applause for bringing this today to something that is what we would call it at MIT that's engineerable so congratulations okay thanks so now the next next as pronounced I go this is when we pray on Sunday that the screen will work please work please work pray harder okay thank you for your prayers so now we have what is this one which one is this skip a day skip one oh there was a bit there so here's another team up with Endor and Berkeley College of Music in a library context of you introduce yourself and run us through the idea yes thank you hello my name is Emil I work at the Berkeley library and I just was exploring how I can use libraries as data with a technology such as Endor and the first few steps was me running with my generic needs as a manager at a library and learning what are some limitations of things that I'd like to figure out but if I don't have actual behavior data I cannot predict the future in that case but Daniel walk me through situations where I do have transactional data that could lead me to make the decision what I make like what books to buy when the space have available in my library it's decreasing decreasing every year but I did not get clearance before I came here to bring my own library data with student data so we use data from last time that I had shared with the world to try to estimate how the process will be to compare song playback data we're pretending that is album checkout data from my library system and the use case in bold the last one was the one that Daniel actually fleshed out by grabbing the data from last seven which is given a set of engaged users find more potential engaged users because for example at Berkeley we feel that the music therapy students are heavy users we want to find other on tap students from a particular major that we want to promote more of our search so there's a information how to access this last FM data it the whole data set is 20 gigabytes and the second link is an analysis of the quality of the data and the patterns within it so like I said earlier I mostly focus on album data but then we use the visual track playback to imitate that now Daniel would like to say something about the process so to take you a bit through the process event at first we kind of dug into the data itself we played around we found some trends just by looking at the raw data then we took all the users or the most engaged users from the platforms from the transactional data there from there we ran our engine black here in black and green and we found more potential users who are going to be as heavy or similar to the heavy users we've worried before that and at the end the right side now you can see the sorted list of user IDs of people who are going to be very heavy just like as the one just like the ones we are worried for now the exact the exciting part about this was that it was very easy and we ran on partially encrypted data in fact it was so encrypted that we couldn't actually get the actual user data at the end only the IDs but in a real-world scenario me as an operator I would reach out to these people and ask them for to be a bit more engaged or even reward them with a coupon or a specific campaign so I started my discussion with Daniel about physical media like books and scores but I would like to expand that to analyzing electronic data or online based data like e-books or streaming audio hits but sadly I need a lot more data from the actual vendors and some of my own subsystems something else I want to explore getting hits and I like this I've been to work his own student registration system so I can track that information like instruments as soon as play and majors and I know the keys next step would be to get clear to share sample data from working with companies like Ender and I specifically like to follow up with working with Ender some more research sharing data there's also everyone better than sharing my use cases just to give a different perspective so thank you so next is just gonna go right to the site for this you're probably wondering how will I get to know what everybody seems to know at this hackathon go to legal hackathon.org and here's a little pro tip if you click on view me on GitHub you can unveil the underlying reality of the data one of the things we have here is a cool little wiki if you click on the wiki you can see you know more documentation all the tools and one of the things we have is you have learning session open music initiative to go to open music initiative you can get sort of the essence of George's remarks here's a little bit of background George and a preliminary video that we did set up the the date is this the article George you want the background when you're speaking sorry oh it is there so this one okay so now we're gonna go from those report outs to fill out a little more context with this article is some sort of a primer on two rules for success for blockchain music entrepreneurs and this remarks today are kind of a preview of the part two of this article but all told it's going to be four points of failure blockchain points of failure in the context of the music industry so with that George just like all of us here you're hacking to the last second which I love and admire but it is time thanks George thank you all for we don't need to look at Dolly so so yeah thanks for doing this right I mean it's not like there are other things to do one of something but this is important I think the timing is really crucial more crucial than than people would have you believe about the music industry as I said earlier it's really it's not a great time but one of the things that's been driving me of late is I'm obviously a true believer in the blockchain space of me I saw an image in up there in the tiny human thing and you go back on my articles back three four years now the very first ones the Genesis block was Imogen and I sort of writing out the spec in public for my silly and you know so she's a true believer we're all true believers in that time I think that it's pretty hard to dispute that blockchain as a tech has become what what you know not seem to tell you would describe as antifragile so even as you know Ethereum might get danged and hit and hurt and everything else I think we can we can plausibly say that blockchain is a tech isn't going anywhere and one in fact blockchain is a tech has been around a lot longer than people would care to talk about so I'm not concerned nor do I really care about blockchain as an underlying technology I do deeply care about music and art and I am driven by this idea that this technology actually might allow for some of these very interesting ideas to perpetuate but at the same time that blockchain as a technology is getting stronger and more in use blockchain as an answer or a way for music entrepreneurs or the music business or whatever to evolve I would say is that a deeply fragile state I say this for a couple reasons one that the the the startups and I won't mention any names so to speak that have kind of been kicking around for a while have have little to nothing to show for and if anybody wants to dispute that with me I'm all ears I'm one you know barely functional human being so I'm likely missing things but I can't find any examples the second thing what that's happened what happens with that then is it sends what's known as a chilling effect and we're all lawyers right a chilling effect to the investment community they say well look we've dumped money into this it's not working we'll turn our money elsewhere right and that's quite honestly a tactic that the the historic music industry has used forever right they've used forever barriers and moats and obfuscation to keep status quo status quo and and legal threats and everything else so that any entrepreneurial spirit that would that would in any way potentially disrupt the incumbents were very very quickly shut down so why what is going on what are the points that need to be to be addressed and looking at the the you know the the hacks whatever and they all are are within these four points and so this is all said with love and with you know again take it take it as you will but I keep seeing the same repeated problems the very first one is this the problem lies around the content itself so I love your what's the name of your company what is it with an H though right what does it mean oh I love that can you get that URL or do you have the URL for that so so like with yours samples is is a is a big bugaboo in the music business right and and I get completely or I think I get as much as one can from an eight-second pitch your gesture of hey let's open this up and you bring up one of my favorite songs of all time right Miles Davis and so what that's a competition he actually did right many of his companies that are famous he did not right exactly and and so he wrote so what and he in his publishing company maybe own that underlying right which is known in the music business and if I'm telling you stuff you already know I apologize but I bet everybody does it right so the that underlying rights that underlying copyright is known as the composition so he Miles Davis and maybe his publishing company on that on the other side Columbia Records put the record out that released that right and so Columbia Records owns the other copyright which is a sound recording for that's on kind of blue right so they own them and it's one of their most protected copy right they will sue the bejesus out of anybody that wants to do anything with so I would love to take 30 seconds of so what right and and put it into some work of my in order for me to legally do that or to do that without prima facie infringement I would need to get approval from Miles Davis's a state well no no Miles Davis is a state in or the publisher right and then I'd also have to go to Columbia Records and get approval from them neither of whom have any compulsory reason to do so in other words if I wanted to do a cover version of so what which would be a terrible motherfucking idea right but but not without my not without plausibility of my life right I could sit down with my guitar and say I want to cover so what in some drunken state and and legally they can't stop me no but hang with me they cannot stop me from doing a shitty shitty cover of so what neither the publisher nor the label because labeled having to do with it and the publisher can't stop me it's what sounds a compulsory license and I can pay the ridiculous fee right the fee set by who sets it huh yeah Congress right your United States representatives right our bicameral system are setting the maximum dollar amount that I can the Miles Davis can pay for so what if you can wrap your head around that explain that to me I'd love you for it I cannot think of any other industry in the world except for I was giving this exam maybe a dozen dogs like oh yeah well in Soviet Russia they do that I was like exactly right so I can't think of any other but but so you get what I'm saying I know but so why are you trying to build a business around that because you're not going to get like that so so problem number one point of failure number one with blockchain based startups you understand this but many do not they don't understand that the complexities around copyright law I'm begging people who are trying to do startups in the music industry I'll this one to work with what's known as control compositions by that I mean works that are owned or controlled by the person who writes the song and performs a song because therefore they can license it they can say sure and I'll set a price for you to build that process where I'm gonna license a snippet of so what you have at minimum and I bet it's more than this at minimum two parties that would have to be coordinated so you've got an order of magnitude more complexities at inception it's not going to work right you could use an Excel motherfucking program it won't matter your issue is not the technology your issue are the rights problems yeah so that's problem number one problem number two point of failure relates to the state of the art with respect to smart contract technologies and so and likely I will again am I not allowed to do that the point is like this is important so if you're wondering about why we're doing this today sorry I want to emphasize one thing the technology is in a sense not the barrier here no could do this in a motherfucking spreadsheet so let's look at the actual barrier yes that's right the actual barrier is the failure to understand that music copyright law is really really hard and complex and is is and it's been statutorily designed to make it almost impossible for startups to access that's the point of failure right I mean what the Clarks whatever law sufficient technologies and distinct from magic right will no blockchain has adopted when people stop saying the word blockchain nobody says TCP IP I don't care what technology means I genuinely am different to that right so but speaking of the technology the second point of failure relates to the current state of the art the smart contracts right smart contracts work exceedingly well for things like supply chain management and that's why people are using them for that because it's a simple if this then that routine right it's simply okay if this rule is satisfying another one will self execute going back to point one with music typically having a number of coordinate a bowl or needing coordination rights holders it's one thing if there's a controlled composition that says if somebody wants to license this work at this price and it is met blockchains all day long if on the other hand it's well you can license this work but there has to be approval from this party who also has to get approval from this party and then this party and if they say no then this and it goes right smart contracts are very binary in nature they don't do nuance well and yet and we're asking too much of them so we're asking too much of the technology and I know the art chain technology maybe better than I should I think it's phenomenal and maybe this is the thing that will allow us to to accelerate it but we all know in this room that the cost and the time to scale transactions on a blockchain right now ain't great and then trying to ask smart contracts to do something they just can't do will again cause things to break really really fast the third one in this this words me about the lines company and I talked about this earlier and that is that is network effect there's been too much hype around just the general idea of blockchain and then if you do some sort of music startup that happens to be on a blockchain and you say the word blockchain people will give a shit they do not right customers do not care customers desperately need some sort of value proposition that improves their lives without forcing them to learn new skills period full stop and unless and until we can do that there will not be adoption and we will not have two-sided markets so we'll have a bunch of crap on a blockchain that maybe it's good maybe it's bad doesn't matter people won't buy the investment community will go well blockchain was the problem no you're a bad marketer you don't understand that promoter score you don't understand all the things that needs to have to happen to shift the burden of marketing from yourself or your company to the users because you don't have a USB and all of those things have nothing to do with the technology but you can't simply rely on the technology to be your marketing tool number four this is controversial this is George speaking for George and not George speaking for open music initiative it's a wonder they let me out of the house point number four you must assume evil with respect to the major labels and publishers maybe they will be nice and good in 30 years in the music industry I have never seen that happen they do not cut off the branch of the tree upon which they sit however they are not done they like their systems the way it is but they also get upward pressure from technology etc too often they will do things that's known in the business world is sandbag which means I'll say sure company XYZ we'd love to do a blockchain technology project with you company XYZ goes oh this is great they release press releases and all this and that the other thing company X blockchain partners with company Y incumbent company Y incumbent sets it up for failure company Y incumbent gets to them in good faith or in good perception of faith say we tried it didn't work think about over and over and over again so these are my little cautionary tales obviously they're not universal obviously there are there are elements in each one of them it can be massaged and maybe work to your your benefit but in looking at these things and I've done more of these things and I care to know I see in each of them these the same kind of potential points of failure not understanding the fundamental elements of copyright relying too much on unlicensable work so you can't get a data set over expectation of the technology generally smart contracts specifically this will change thinking that that you can somehow magically create a market just because you're using the word network blockchain in other words forgetting network effect and number four assuming anything other than evil with strategic partnerships at the incumbent company super fantastic so that that I'm very very glad that we got that um kind of on uh recorded so that we can go back and actually really think about this and make this part of our uh make this part of our portfolio part of our agenda but first things first um let's now have a want to get a kind of a better background screen going here we can figure out my password for where I just did some people some human faculty and we're like looking at like the slight glare behind me and uh that's the good hack