Every few months, a new article declares that blockchain will revolutionize the music industry (or TV, or any other media). Just about as often a new article declares the opposite. Benji Rogers seems like he could write either, or both: he refers to the excitement to build blockchain into everything as “kind of schizophrenic,” but at the same time he’s founded a company to bring it to music metadata.

Every time you stream, multiple parties get involved before payment even makes it to the performers and composers—streaming services, publishers, labels, publishing rights organizations [PROs], the list goes on. Each of those has their own database of songs, master records, who made them and who gets money. And they don’t always match. A musician himself, Rogers sees blockchain as our best chance to solve this bureaucratic tangle. 

In just a few months, Rogers’ company dotBlockchain Media will release the first public (and free) version of their technology, which will enable pretty much anybody to export their music to a .bc file and get involved. We spoke with him at length on just what that means, and what new avenues it can open for major labels and independent artists alike.


Why don’t we start right out with the blockchain? I’m sure that people hear “music” and “blockchain” and they probably just assume something like buying music with Bitcoin, which is not at all what you’re doing.

The most important thing is that I think that there is two types of blockchain as it relates to music and media industries. There’s blockchain for payments, which is moving digital currencies like Bitcoin and Ethereum around the internet, and there’s blockchain for data. And so, one of the ways that we looked at blockchain was, yes, there are ways to look at innovating how artists get paid by clients, but what problem does that actually solve? Is it going to pay artists more? I would argue, no, that’s not necessarily true that it will pay artists more. It’s just another way to pay.

Most of the companies do know how to pay artists, they just don’t have a way to, because they can’t get their ownership information. If you look at some of the lawsuits that have gone on in the music industry of late, it’s basically, legally speaking, Spotify or whoever they are have to pay the songwriters who wrote the songs. But there’s nowhere for the songwriters to write down that they wrote them.



When we went from vinyl to digital media, songs became files, and files go everywhere. Files are replicable. There’s a whole bunch of things you can do with files you couldn’t do with CDs or vinyl very easily. They can have their data stripped out of them. And so, why do we write about the songs everywhere but within the songs?

So you’re reducing the data loss by moving the recordkeeping into the song itself.

What we’ve done is basically built almost like a Slack or WhatsApp interface into every song, so that you can write down, “I am the songwriter,” write down, “I am the performer,” write down, “I am the producer,” write down, “I’m the engineer,” and wherever the song goes, the blockchain tracks any changes to that information. Today the average song is transferred to streaming services as a WAV file. WAV was introduced 26 years ago during the presidency of the first George Bush. We don’t use much technology from then—except for all of our media, right?

So it seems crazy to me that you would sit there and say, “We’re gonna build this brand new blockchain-backed interface for payments and it will smart-pay everyone who’s written into this database.” Okay, how do I get my information to the database? And then once I’ve done that, what happens if I copy the song onto my hard drive and push it somewhere else? All the data’s gone. You have to track it from where the song actually exists itself.

So let’s make the song, if you will, the surface upon which you digitally etch your information. So if you wonder who wrote the song, look inside the song. If you want to message the people that wrote the song to ask for permission to use it in a movie, message through the song.

There have been many attempts to do one database to kind of rule them all. But all those attempts fall over because A. who runs it, B. who pays for it, and C. what happens if not everyone wants to it? But blockchains have some very unique properties. The one I like the most for this use case is, if I tag you in a song, you have three choices. Approve the tag, which means it’s true. Ignore the tag, which means it’s true. Or deny the tag, which means you can figure out what is the truth. Which is really going back to sort of Tim Berners-Lee’s construction of a semantic web.



What we really thought was, if you get people to use the same system to their advantage, which means it serves everyone’s good to get this information in, then the songs become smart, and you can monetize them much, much faster than you can today. Because today, if a service wants to access a song directly, or in a different way it’s really slow to almost impossible to get permission do so. I’m stuck with all the rules that are set by other third parties, but if we create a smart song, literally Amazon’s Alexa or Google Home or Apple HomePod communicate with the song itself. You can effectively connect commerce at the song level, versus 17–20 parties having to get their information correct to pay the person.

It’s a simpler way of looking at the problem. I’m a big fan of cryptocurrencies. I do love the idea that you can transfer money really fast to the right people, but if you don’t have a method to identify who the right people are, aren’t you just speeding up the problem?

Essentially what you’re trying to solve is the transparency of authorship.

Yeah. I’ll give you an example. You’re recording this conversation we’re having, right?

Right.

And the end result will be a WAV file or an MP3 or an MP4. Now, if you were to edit that file to make me sound foolish, and write something that I didn’t say, I would say, “Well, that’s not what I said.” We have an argument and you could say, “Well, no, here’s the recording of it.” Now I don’t know if it’s the one you edited or the actual one. There’s no provenance to it. So if we don’t solve the versioning of media, effectively, if you were to act badly (which I don’t think you will), you could alter anything I say and I wouldn’t know.

I could argue with you and say, “Hey, I didn’t say that,” but you could come up with a perfectly edited version: “Look, here it is,” and I wouldn’t know where the edits were. We would argue over it, but in the meantime, you’ve published it and the damage is done.

Think of one news story being recorded and disseminated out to 10,000 different outlets. That’s power, right? And if you can’t track the provenance of ownership, how do you ever pin it down? So a database, a blockchain database that just holds caches of information is all well and good, but if the media itself doesn’t speak to that database, what’s the point? We have a media and access problem. We don’t have a data problem. If you have the correct data, every single digital service provider knows exactly who to pay.

Spotify knows how to pay. YouTube knows how to pay. All these services know how to pay, but they often don’t know who, because there’s nowhere for that person to write that down. And we believe that we should write it down within the song itself. And that’s what gives the song power. You’re not just gonna be reliant upon these larger incumbent companies controlling all that metadata—arguably not very well.

I think we have to just rethink the versioning of our media. And it’s not the most exciting application of blockchains. But it’s what blockchains are actually good at. This person made a claim. The blockchain confirms it. What it actually has is a version history of all the created media. We’re starting with music because that’s what we know, but this is for all media.

It means that effectively, you can also know the provenance of a news story. You can always know the provenance of a movie. You can say, “Hey, how did you get this version? This is out of sync with what the rest of us have. Something’s wrong with it. Maybe it’s a virus. Maybe it’s something untoward.” And I think that that’s a real opportunity for us as we go ahead into this very digital world.

One of the most important things for making a blockchain work is that there are computers to actually compute the blockchain. It must be a tremendous chain to have metadata for millions of songs. Who do you imagine will run the nodes?

We’ve actually thought a lot about this. Today, arguably, the fastest scaled up blockchain is Ethereum, the second fastest is Bitcoin. But they’re having trouble confirming more than a small number of transactions per second. So can you imagine if you were to suddenly take the 10–20 trillion potential musical data transactions, that you have to put in there? It’s too much to think about. Whereas, if you’re just talking about the metadata changes, then you’re talking about a different type of blockchain.

What we’re using for this initial wave is Intel’s blockchain, called Hyperledger Sawtooth, which is their contribution to the Linux Foundation’s project. All this would be doing is tracking the amendments. No media on chain. It’s just doing music metadata, which means it can be much, much faster.



And it needs the participants in the music industry: labels, publishers, PROs, ourselves, independent artists can run a node if they want to. That’s what’s really powerful. It doesn’t require huge amounts of power to run it. It’s a bit like turning up—or more correctly put—adding more cloud capacity, if you will, with a lot more bells and whistles attached.

It’s a little clearer now. A .bc file that you would use to play music would contain its own metadata, and a reference to the blockchain itself to verify that?

Yeah. Think of .bc as like a .zip file. So if I take four files that are too big to email to you, and I zip them into one small one, I can send it, and it gives you permission to open that file at rules that I’ve set or you’ve set. So the same thing applies. The difference is, when I zip those files then, any changes made post that fact are written to the blockchain. So you would never just go command/Apple-I and alter metadata. You would go command/Apple-I and it would say, “Do you wish to propose these changes to this metadata and publicly write them to the blockchain?” So that the blockchain’s just gonna be tracking the changes and the variances between what’s happening.

Our plan is to roll this out in stages. Next year, the first public-facing app that any artist can use for free will be available. We’re a public benefit corporation; it’s designed so that anyone can use this thing. You will need stronger applications if you’re a bigger company, in the same way that I can use Gmail personally on the SMTP protocol but if I try to maybe email five million people, I would need to go to a more specialized platform to help me do that. Similar thinking with this. Anyone can use the Gmail version of .bc, to just message people and do whatever they wanna do.

At first glance it seemed like it would be difficult for an independent artist to get involved, but it sounds like there’s a minimal cost for anyone to just jump in.

If it can’t be used by a singer-songwriter or a performer, it doesn’t work. It’s obviously a very expensive system to build and there’s a huge amount that goes into it. And yet at the same time, I think it’s just the smart way to go because it’s scalable. It’s within reach. We can do it.

And I think that we’re not going to be subject to the problems of scaling that Ethereum or Bitcoin are having. We’re just thinking of it differently. Blockchain for money is always going to be a different thing, in my mind. That’s not a Bitcoin or blockchain “purist” attitude, but it’s a practical attitude towards getting us to the deployment and adoption phase.

And cryptocurrencies require a huge proof of work, which is why they’re using more electricity than entire countries.

If you want to send anonymous digital cash between people, that’s brilliant. But what we want to do is get the correct musician’s and songwriter’s and other participants to the songs’ information into their songs so they can be paid, whether in fiat currency or in anonymous cash. We don’t really care what the currency is, as long as we can surface to those who want to pay, who to fire the payment at. And it’s just a very different attitude, because everyone’s been like, “Hey, we’re gonna build Spotify, but on the blockchain. Or YouTube on the blockchain.” And I keep saying, “Do we need that? What’s the advantage of putting it on what is today a markedly slower system? Is there a true, valid and important reason to decentralize this process?” It’s kind of schizophrenic, to just put everything on the blockchain. In the same way that when the internet was first coming out, everyone was like, “Hey, let’s get this thing on the internet.” What does it actually need to go on the internet for?

What’s key is that this is completely platform independent. So people could build new apps without the huge hurdle of making individual licensing agreements.

Yeah, it was very specifically designed with that in mind. Emails scaled because SMTP was given away, for free. You could use it. The internet scaled because of the TCP/IP and the protocols that run the internet. And then we made all the money on the application layer on the top. What .bc really is, is the protocol for media, and if we get it right, everyone can use it. It’s free to use. You can make money on the application side by building new streaming services, facilitating changes to existing streaming services.

So for example, if Spotify gets delivered 65 million .bc’s, which have every co-writer, every songwriter, every engineer, every publisher, every bass player written into them, then effectively, what Spotify could do is create playlists based on those variables. They have all the album codes, all of this rich data flowing. You could search your system and say, “Wow, interestingly enough, this person keeps listening to the same bass player without knowing it.” Surfacing that type of rich metadata actually makes the services better. And that’s what’s exciting.

That’s a good point, there’s a lot of potential for recommendation.

I think there’s also potential for things like sync licensing. Think of how many student films want to use music legitimately, and they want to pay if it makes money, but they can’t afford to do it up front. Well, if they can message the song and say, “Hey, this is the trailer. Can we use it for this? And if it goes well, could I do a license from you?” “Sure, why not?” Because today, it doesn’t make anybody any money to go and try and do a $5,000 sync license with a major label. They won’t listen. It doesn’t make enough money for them. Whereas if there was an automated way to do that, you would do it really fast. But that’s dependent upon the people who own the song and who are part of the song being able to give permission to do so.

It makes it all very conversational, and the cool thing about that to me is that musicmaking is conversational. Musicmaking is often done between people talking about the same thing. The problem is that we distribute it as if it’s not.

It also eliminates a lot of fear for independent artists. Right now it takes an anxious good faith in private services, like “I’ll put this on Bandcamp, and nothing will go wrong, I hope. It’s just audio that’s on the internet now and I don’t know what’s happening to it.” But this versioning gives you a way of knowing essentially that your work stays yours without limiting where it can be found.

If a machine is tracking this stuff, then a machine can read it. We’re going into a world in which people will either tell machines what to do, or be told what to do by machines, and we need to think of our songs as little machines, that go out into the world and commit commerce with platforms and bring that money back to us. And we have to get rid of this outdated notion that we require an interface in which music needs to be played in this specific way.

At the first company I built, PledgeMusic, we would sell, effectively, virtual songs. The songs didn’t exist yet. We pre-sold them. Just as valid a form of commerce as selling things that actually exist. And what was interesting was that we would just get fed these WAV files later on. I kept thinking to myself, “How does this work long term? Like licensing needs to be done so much quicker than this, but if there’s nowhere to put that information, how do we even begin to modernize it?” And that’s really what we’re trying to get to, which is the modernization of the media format and a new protocol for how media gets around the internet.