There Is Too Much Music Out There (Part 1)

Hello, hello, hello. Announcements, announcements, announcements. First off, all support towards the Writers Guild of America that went on strike this Tuesday. Certainly an important fight no matter where you sit in the entertainment or tech industry. If you’re in New York City, this Saturday is the Water and Music Summit in Bushwick, where if you use the code (‘FriendsofWM’) you can still get 10% off. Next Saturday, my friend Moistbreezy will be hosting an album release party at Trans Pecos, for her upcoming album Pure Imagination. Last but not least, if you’re going to be at the Music Biz Conference in Nashville, let me know, I would love to meet. Promo outta the way, let’s get into this week’s newsletter.

The chatter around “artificial intelligence” is at a deafening pitch. This intensified over the past couple of weeks with a certain “A.I. Drake” song; the mad scramble to have it removed from streaming platforms and further speculation on music’s future. Unsurprisingly, or at least I hope, I’m not keen on commercialized, generative AI. Especially in this form of subpar novelty works and tedious fear-mongering around automation. However, before I tackle AI, I wanted to chronicle the evolution of commercialized digital music. Already emergent contradictions within modern music may provide more headwinds for music AI than is fully appreciated by boosters and doomsayers.

Welcome to the Celestial Jukebox

Earlier this year, I read an excellent book Music & Copyright in America: Towards the Celestial Jukebox by Kevin Parks, a newsletter reader who was kind enough to send me a copy. The book details the century-long history of music copyright cases from sheet music to 21st-century fights over piracy. Towards the end, Parks highlights a little-known paper by the US House of Representatives in 1995. The report looked at the Digital Performance Right in Sound Recordings Act of 1995, which would eventually establish a digital performance right, and foresaw that easily duplicated high-quality online audio might be wonderful for consumers but represented a potential threat to those “who depend upon revenues derived from traditional record sales”. Doesn't that sound familiar?

A few years ago, Cherie Hu over at Water and Music penned an essay critiquing the “celestial jukebox” era (her emphasis): “What I want to talk about today is why this collective emphasis on volume is actually irrelevant to how streaming consumption and curation actually work today — and why we need a new metaphor for understanding music streaming in 2020, instead of aggressively clinging to the “celestial jukebox” with which we once fell in love.”

She pulls on the history of the phrase “celestial jukebox” which goes back to a 1994 book (Copyright’s Highway) by Stanford Law School Professor Paul Goldstein, where he forecasts what would become early file-sharing sites, Napster, and ultimately in many ways, Spotify. Hu pushed back against the emphasis streaming platforms placed on hosting an excessive number of songs when in reality most fans aren't crate diggers. The real value of a streaming platform is the known, not unknown, content. She goes on to point out that such volume further helps justify the value of playlists on platforms where there’s just so much content to stream but users aren’t sure what to consume next. Yet, a few years later, we see stories about the depreciated value of playlists, endless paranoia around TikTok, and now AI. The value of quantity, instead of quality, on a music platform feels even more bizarre.

Now, why this little history lesson? The multi-decade story of digital music is often told through the flash point of Napster, in a little guy (Sean Parker) vs. the big bad record labels way. This particular claim ignores both venture capital funding behind Napster and its peers and also that if the US government was writing reports on the impact of digital music in the mid-90s, no one can say it wasn’t on people’s radars. The utopian framing of frictionless access to music captured a generation who went to stores to buy music or maybe heard a song on the radio only to never come across it again. The promise of an endless catalog for cheap, especially during the financial crisis when streaming began to become mainstream, is incredibly appealing. Yet, that dream is hitting a changing reality.

Many of the user-generated music platforms that emerged throughout the 00s sit in the long shadow of the “celestial jukebox” idea. These took differing shapes of Myspace, SoundCloud, various music blogs, and other now long-forgotten sites unfolded the ability for musicians to easily share their music without needing to physically produce copies for people to hear. Over the decade, this opened up the floodgates in terms of the quantity of music that was out there for people to potentially hear and engage with. Yet, in those days with the music industry contracting and lawsuits flying, all within a still fragmented digital landscape, it’s easier to understand how this shift was thought to be revolutionary.  

Real change occurred in digital distribution options, yet with some hindsight now, it’s clear that the value of record labels, radio, marketing, and other more traditional ways of getting music in front of people didn’t vanish overnight. Instead, what increased was the number of hobbyists who thought they were closer to a dream of being a working musician since the means were being placed right in front of them. A similar trend extended across the 00s and 10s with YouTube, Vine, and most recently TikTok. Suddenly, these unprofitable businesses provided people with the ability to perform and distribute their works, be it music, short films, writing, etc., and expect – no, demand – an audience. These tech companies would frame this as “democratization”, yet if the final rewards of“success” degrade evermore with more people running towards that goal, then it'd be hard to argue artists are seeing real benefits of this technological change. Parallel to trend another one was emerging within the major record and publishing industries.

Over the last couple of years, a slightly bizarre discussion emerged amongst cultural observers. The starting point was a bit of data from MRC Data, now Luminte, the data analytics firm that collects data for Billboard. Their analysis showed that "old" music consumption eclipsed "new" music (to be specific, "old" referred to music older than eighteen months). This set of data became the basis for Ted Gioia to write in The Atlantic about his own perceptions of younger people listening to older music, thus shunning younger talent, all while labels and publishers are reorienting themselves towards promoting and investing in older known catalogs. For the concerns of this newsletter, it’s important not to conflate these hypotheses.

Contemporary music does face a much higher degree of cultural competition than the music of previous eras. The cultural impact and reach of music in the 1980s is just going to be higher than in the 2020s. Music isn’t currently at the forefront of emergent mass media forms (radio, cable television) and there is now just an increased flood of work out there. All those YouTube videos and TikToks are taking away attention from something, that cannot be denied. Now, Gioia’s other observations on industry investment shifts are worth more reflection.

Last month, BMG announced that it was combining its frontline and catalog businesses, explicitly citing the fact that three-fourths of its business is found in older works. The rise of song catalogs is something I’ve covered for years, all while major labels appear to have all but given up investing in up-and-coming acts. The emergence of labels but also their peers in the distribution business willing to sign acts for only a single, rather than a traditional multi-album deal, shows a clear short-term investing strategy, rather than the longer-term bets made in previous decades. This isn’t to venerate the exploitative nature of multi-album deals but rather to point out that an artist signed for a single isn’t being given the same level of industry support to build a career.

The economic, rather than cultural, drive towards older music cannot be overlooked since if young people aren't being marketed artists with the same level of investment, then it shouldn’t be shocking that these artists don’t last. This is all happening while there is a cheapening of tools to create music, which mirrors the increased ease of being able to distribute these works. This produces an industry that is increasingly built to be highly derivative, and in fact, at the legal level, this is a further self-reinforcing condition.

Last year when I wrote about digital music “piracy”, I mentioned the book Democracy of Sound which gives a detailed history of how music copyright holders fought against unapproved copies of their work. An underappreciated thread in the book is that often the market for these works was what today may be referred to as "stans", or extremely devoted fans. That celestial jukebox wasn’t preordained as music’s natural state. The contemporary model of music streaming is in many ways an outgrowth of an expansive copyright policy that was being litigated first internationally and then domestically with the arrival of the internet. Despite Spotify’s marketing department, people are not inherently ‘better’ because of a diverse musical palette; but that extreme hobbyist character trait is pressed into the core logic of music streaming, which is the industry’s primary revenue source.

Ironically, it’s that misguided endless collector mindset, mixed with this ease of creation and distribution, that’s beginning to undermine the very platforms that host these works. Music Business Worldwide in recent years keenly observed that hosting all of this music isn’t cheap. Though I may quibble with the exact methodology, it is astute to highlight the real-world costs of maintaining not only the servers to host the tens of thousands of songs uploaded every week but also paying the teams to properly ingest, tag, and maintain these works. The ongoing, even if unsaid, drive to maintain the celestial jukebox is running into problems.

The other issue beyond the technical and labor costs that increase with swelling catalog sizes is the increased political tensions over content moderation. A great op-ed earlier this year by Tristra Newyear Yeager, the Chief Strategy Officer of the public relations firm Rock Paper Scissors, pointed out multiple cases at the state and federal level that could end up placing more onus on platforms around content moderation. The direction of the cases she cited is up in the air but does point to increased cost and legal uncertainty with maintaining such vast catalogs on top of straining their technical and legal sustainability.

This scene setting is to help better isolate how various parts of the music industry may approach “artificial intelligence”. Lazy commentators will evoke Napster, piracy, and some refusal to get along with the times but ultimately it’ll be artificial intelligence companies bending their business models into the record industry’s vision.

Unheard Labor

Contractors at YouTube Music won their union vote to join the Communication Workers of America. Bargaining, and ultimately winning, a contract is still needed for these workers, but alongside Bandcamp and Secretly Group, there are ever more nodes of unionization within the American music industry. In local news, the musicians at the Distinguished Concerts International New York (DCINY) are on strike to win a contract, as the DCINY continues to undermine the union, a typical recent move within NYC’s more established (i.e. flushed with cash) cultural institutions.

Over in France, there is a proposed adjustment to an already-established ticketing tax, which would include a low 1.75% on music streaming platform revenue to help further fund the National Center for Music. (I helped co-author a report on song catalog purchases for the CNM, so I’ll admit to being a bit biased.) Still, the streaming market has matured enough that taxes and heavier government regulations can enter the frame. Unsurprisingly, the CEO of Deezer and the trade organization of major labels weren’t excited by the proposal. O, also Italy told Facebook they must go back and negotiate with the Italian Society of Authors and Publishers; the saga continues.

Mandolin, a livestreaming start-up that raised over $12 million, including from The Orchard, iHeartMedia, Concord Music, and other venture capitalists, is shutting down. I articulated my critiques around livestreaming a couple of years back and now hold even stronger convictions that many of these companies were built on top of a bubble that was already bursting without anyone realizing it at the time. Either way, I hope folks can find new jobs, and maybe capital allocators will find better places to invest, but I digress.

A Note of Financialization

Two fund raises really caught my eye last month. Primary Wave invested $100 million into Times Music, one of India’s biggest labels, to begin purchasing catalogs. So, expect to see an even bigger push for Indian music in non-domestic markets. The other deal was Armada Music, the label founded by Armin Van Buuren, which is launching a $100 million fund with an eye towards electronic music catalogs and started off buying masters of KMS Records, founded by Kevin Saunderson; and masters and publishing of the producer Arty. The capital came from Pinnacle Financial Partners, who helped Cutting Edge Group with a recent fundraising round. Hipgnosis bought the “writer’s share of performance income” by the composer David Foster. Reach Music got the masters and publishing for Judas Priest’s first two albums and explicitly said they were going to be doing 50th-anniversary promotions of them. Nice when the marketing plan accompanies the initial purchase.

6 Links 2 Read

The Music Streaming Crisis - Conversation with David Turner - 8000Km

Got to join a fun podcast in providing a 360-degree view of the record industry. I got to meet Jorge, one of the hosts, earlier this year and I'm so glad he invited me to his show. It was a lovely time.

As Music Downloads Tank, DJs’ Favorite Platform Beatport Is Selling More Than Ever - Billboard (Subscription)  / 'It's Becoming Harder and Harder': Is the Dance Music 12-Inch on The Way Out? - Resident Advisor

The ripple effects of the pandemic continue to be felt in digital music finding a long shelf life for up-and-coming DJs, while the genre’s original format is becoming harder to maintain. Even if the economic value of the 12-inch declines it’s amazing that such a format lasted nearly 50 years, and would say that many folks interested in “alternatives” to music’s mainstream should perhaps look more at what still exists and how it managed such a survival.

Everything Is Interpolated: Inside Music’s Nostalgia-Industrial Complex - Pitchfork

Another look at the world of song catalog and how it’s starting to seep further into top 40 pop music. Even though sample-heavy music is certainly nothing new since the 80s, I cannot help but think the music industry’s quiet problem of nurturing new talent isn’t disconnected from the almost necessity of novelty driving younger acts. This isn’t to fetishize originality but such a recursive music industry ultimately will start to take a different mold than what came before it the longer this trend goes. Or maybe, I’m certainly hoping, this is another zero-interest rate phenomenon that’s being worked through the system.

web3 music in 2023: experiments, communities… and weirdness - Music Ally

Even in the middle of a bear market my skepticism towards “web3” is unwavering. This web3 panel summary hits on why, as I couldn’t get over the vagaries of the discussions and that the perceived difference between “web2” and “web3” makes less sense when no one is using this third iteration of the internet. Still, I recognize these are smart and passionate music technology people but I do wonder when the self-indulgent nomenclature fades back to simply music technology.  

How a Google Antitrust Case Could Determine the Future of AI - BIG

Matt Stoller neatly connects a few thoughts that were swirling in my head around the recent AI hype wave. First, all of this AI chatter is to the benefit of Microsoft in helping justify the billions they pumped into OpenAI, especially while the business model for their products isn’t well articulated. But, the more subtle knock-on effect, and more within Stoller’s wheelhouse, is that Google is now acting like it's no longer a monopoly of the search market, because of *scary voice* AI. Real uses of commercialized AI products do exist…but the hype cycle does feel powerfully skewed toward certain corporate narratives.

We Built This City on Authenticity - This Machine Kills

A great podcast on how cities sell themselves, and I bet readers of this newsletter will be able to see how music, in particular “independent” venues and communities, fit into the shifting character of cities.