Bitcoin Won't Save Music January 7th, 2016.
When it comes to the future, I consider myself one of the optimists. But sometimes our collective optimism can run too far away from reality when the technology involved is complex. And the largest disconnect I see right now is this: Though the music industry has very many problems that technology can address and potentially solve, Bitcoin will not be involved in any of them.
Why am I the first person to tell you this? Music industry folks who have a great grasp of the problems are typically non-technical, so can't fully evaluate potential technical solutions. Journalists aren't often technical either, but even if they are, they also have deadlines and limited reader attention spans to reckon with. Entrepreneurs and investors in the field are selling a dream that can make them rich and/or famous even if that dream never materializes. And the engineers who really grok the technology are too busy trying to apply it to actually explain all its limitations to anyone.
So, as entrepreneurial software engineer, investor, and long-time student of the music biz, I took a close look at all of the various proposals that have been made. What follows are my detailed rebuttals (on both technical and economic grounds), but the TLDR is, Bitcoin won't save music.
Ascribe is a platform for attributing art to creators. It claims to "create an unbreakable bond between a creator and their work." What they actually do is create and track ownership so long as whoever later wants to take ownership does so using Ascribe's proprietary tools. If someone gets a hold of your digital work, it will not carry within it any attribution at all. And it can be consumed, shared, remixed and in all other ways pirated if whoever is doing the pirating doesn't care to register for an Ascribe account.
Don't believe me? Here's what their FAQ says:
Q: Can ascribe help me to enforce my rights or deal with infringements?
A: Not directly. We try to give all creators the transparency on where their work is spreading on the internet. With this information, creators have more power to decide what to do. Our goal is to give everyone the chance to claim authorship, license your work how you choose and be able to trace how it spreads on the internet. By doing this, every work is attributed to its artist or creator.
When they say in the FAQ that their "goal is to ... be able to trace how [your work] spreads on the internet" they mean that they haven't yet built the ability to actually do it. They claim elsewhere on the site to be working on the functionality, but in truth what they are proposing is intractable. With more than a decade invested into this sort of thing at Google, copyright material still hangs around in search results and YouTube videos. Don't hold your breath for Ascribe to figure it out.
Even assuming there were some way to account for all of the activity around a digital work, the Bitcoin blockchain has the theoretical capacity for just seven transactions per second at maximum. In practice, the sustained rate is more like three and change. There are proposals to raise this rate, but months of controversy have not led to any of them being adopted, and the maximum rate increase proposed right now would be to 224 transactions per second. In order to handle all the Spotify streams that happened in the past minute you've been reading, the rate would need to increase by a factor of 200,000% !
There are other blockchains that settle transactions faster than Bitcoin, but there is no consensus-based log technology that can handle activity at the granularity that some articles might suggest. It's, plainly, impossible.
Besides, the Bitcoin community really dislikes small transactions intended just for storing information in the blockchain. They call it dust or spam depending on how friendly they're feeling. While the limit before your transaction is totally ignored by the miners  is fairly low, the standard transaction fee makes it prohibitively expensive to track much information. Oh, you didn't know Bitcoin transactions cost money? Right now the standard fee is on the order of 10 cents, and that will likely need to increase once the block reward falls later this year .
Ujo Music's ambitious aim is to build a system on top of the blockchain that integrates microtransactions, smart contracts, and transparent payments. Combining all this fancy technology attempts to address a concern that perhaps fans aren't ponying up for artist's work because they don't trust the current digital music gatekeepers.
In reality, the reasons fans don't pay for music are much more mundane. What they like, they can get for free or cheap, and because there's an abundance of other options, anything inconvenient will be skipped regardless of quality. So unless you're making better music, cheaper music, or easier-to-obtain music, you aren't helping. Ujo (and the blockchain solution they've built) addresses zero of these concerns directly.
If you want proof, Imogen Heap, Grammy Award winner (and Taylor Swift co-writer, to boot), using Ujo's technology crafted to fit her vision, was only able to sell a little over $100 of her brand new music, Tiny Human. How do I know? I just clicked the handy link that transparently reveals her earnings .
Why so low? Because even if you wanted to buy it, you can't spend cold, hard cash on it—you need Ether. What's Ether? It's the basic unit of a significantly more complex Bitcoin-esque cryptocurrency with significantly fewer users called Ethereum . But even if we were using Bitcoin, every currency exchange hoop a customer has to jump through to consume your digital works will reduce the customer base exponentially. Attempts to apply crypto-based tipping transactions have failed elsewhere, too, most notably with ChangeTip's declining userbase.
I wrote earlier about ways that microtransactions could actually help, but the punchline there is the same as it is here: short of top-down innovation in the standard credit/debit rails there are no awesome technology hacks to help us at the moment. Just plain old spending money on things and indirect monetization via advertisement.
Benji Rogers outlined a plan that echoed earlier thoughts from D.A. Wallach around tracking credits for musical work on the blockchain. As discussed previously, the blockchain is not a great place to put information of any sort. But unlike before, we can probably fit all of this data on the blockchain, so that's good.
What's bad is that the blockchain is fundamentally a list of transactions (or a ledger), but what we actually want here is a database. We want to be able to quickly write data, modify it later, and efficiently read it back out so that we know what to do with it. That's not the way a list of transactions works. Writing data to the blockchain takes on average 10 minutes, modifying data is impossible by design, and reading it back is possible but requires a 50+ gigabyte download and a server that is always connected to the network. Any proposal that would solve these individual issues with the blockchain ought to just go ahead and solve the whole problem without the blockchain being involved at all, since PostgreSQL would do just fine.
But the real reason Bitcoin won't solve the issue of bad metadata in the music industry is, that bad metadata comes not from outdated infrastructure or fragmented PROs, but from people who are fallible, occasionally lazy, and sometimes also just plain misinformed about the collaborator and rights details of a given song. If you solve the people part of this issue, you'll find the technical part was solved 40 years ago with the invention of the relational database.
Almost nothing, in and of itself. Those contemplating innovative ideas to remake the industry aren't wrong to do so. And in the face of complex problems, simple answers like "blockchain" can be understandably, seductive. But the oasis in this desert is just an illusion.
Bitcoin won't save music. And the sooner we all realize that, the sooner we can start working on things that can.
Disagree? I love a good debate. __aston__ on Twitter.
- Approximately. Hand-wavy math: 10B hours of listening per year, based on a reasonable increase from the 2014 number with the average song being 3 minutes. ↩
- The phrase "the miners" sounds ominous because it is. The majority of the computational—and thus decision-making—power in Bitcoin is owned by relatively unaccountable Chinese operations, only some of whom actively engage with the community. What miners do, roughly, is enter a lottery where the number of tickets they get is proportional to the amount of computation they do. Whoever wins the lottery gets the chance to commit the next block of items to the ledger. They also get a reward of 25 Bitcoins out of thin air. This is the only way bitcoins are created. ↩
- The plan is for there to be only a finite number of bitcoins ever created. So the block reward mentioned in the previous footnote has to decrease over time. For reasons that probably have to do only with convenience in code, the reward cuts in half when it changes. It changes based on how many bitcoins have been created, so it's hard to know exactly when "the halving" will happen, but mid-July 2016 is the best guess right now. ↩
- At the time of writing this post, the "Distribution" link was not working on Ujo, but that screenshot is from when it was. ↩
- Ethereum is, however, Turing Complete, which is cool. ↩