✘ Music industry 3.0 - what we actually want to build
And: The economic and social impact of independent record labels; billion-dollar battle over online piracy; Let it burn; Architecture for sound; The choreography of posting online
We’re in the “horseless carriage” moment of music tech. Remember those early automobiles that looked exactly like carriages, just missing the horses? They were dangerous but also a revolutionary technology. Eventually, they dramatically transformed how we got from A to B. This begs the question: what could music technology do if we stopped thinking in terms of what came before and started designing for what could come next? We see this time and time again. Just look at streaming - it wasn’t just a digital version of old systems. It reimagined everything, from production to distribution and consumption.
We’re at a turning point
The economist Carlota Perez created a framework for technological revolutions that tells us that every major tech shift follows a predictable pattern: irruption, frenzy, turning point, then synergy. Depending on how you interpret the current technological revolution, one could argue that we’re at that messy turning point now. The old system is obviously broken, but the new one hasn’t crystallised yet. The frenzy phase in music as a creator has given us 100,000+ track uploads daily on streaming services. In January, Deezer found that about 10% of this content was AI-generated, and by April, this figure rose to 18%, by September to 28%, and by November, it had risen to 34%.
The numbers are stark. CISAC predicts that by 2028, 20% of streaming revenue will come from AI-generated music. That’s a fifth of musicians’ income potentially evaporating in the next three years. It’s clear that AI is transforming music. The big question is whether musicians themselves will have any say in that transformation.
Most musicians aren’t resistant to technology. They do, however, demand to be included in building it. They should be at the heart of the process, not just as content suppliers for platforms. The entire industry exists because of the central position musicians have within it. That’s the essence of Music Industry 3.0.
At Amsterdam Dance Event this year, we put together a workshop with AFEM (Association for Electronic Music) and MTNL (Music Tech Netherlands). The brief was simple: design your ideal music industry and be as pragmatic or utopian as you like. Here’s what the group sketched out.
Creation: Building DNA into music from the start
It all starts at the moment of creation. Right now, producers work in DAWs disconnected from rights databases, collaborate through voice notes, and hope everyone remembers who contributed what. But what if we could embed “music DNA” right from the beginning?
Picture this: you open your DAW and it already knows who you are through a universal artist identifier - one ID that works across every platform, every territory, every use case. Like having an artist passport that’s recognised everywhere. As you drag in that vocal sample, the system logs the contribution. When you bounce the final mix, it’s more than a WAV file. It’s a complete package with all the necessary metadata embedded.
This ID would capture everything: who made what, who owns what percentage, what samples were used, and what the splits should be. And with a good use of AI, we can now separate stems and create unique identifiers for each component of a track. So when your beat gets sampled, remixed, then sampled again, the trail always remains alive. Technology alone won’t fix this. We need regulatory frameworks, whether through governments or CMOs, to ensure these standards are actually adopted.
The attribution revolution that’s already possible
Here is a fact: the technology to fix attribution already exists. Meta knows when you’re uploading a DJ set and detects any copyright infringements for take-downs. They can detect vaccine misinformation in milliseconds, flag nudity before you blink, and trigger mental health resources if you type certain words. They analyse everything that gets posted.
Yet 97% of the average DJ set is other people’s music, and 90% of social media posts containing music provide zero attribution. The solution isn’t complex. When someone uploads a video with music, trigger a pop-up asking for credits. Don’t provide them? Your post reach gets throttled. This seems like a quick fix, why? Let’s repeat: the technology already exists!
This extends beyond social media. The DATA framework - Data, Aggregation, Transparency, Access, thought up by one of the groups at the workshop - imagines breaking down the silos between platforms. Right now, streaming data lives with Spotify, viral moments with TikTok, radio plays with PROs, but artists can’t see the full picture in one place. A universal artist ID would aggregate everything. Not just for tracking payments, but for understanding audiences. Where are they? What else do they listen to? When and where should you tour?
The power shift would be seismic. Artists would own their data by default, granting access as needed. Want a booking agent? Show them your live performance metrics. Negotiating with a label? Share your true fan engagement stats. Planning a tour? You’d actually know where your listeners are, not where an algorithm guesses they might be. All this information already exists and so does the technology to bring it together.
Human verification in an automated world
Let’s confront an uncomfortable truth: 80% of tracks on streaming platforms have never been played. Not once. Much of this is AI-generated content uploaded purely to game the system, to pad out those “Deep Focus” playlists, to syphon micropennies from the pool.
Now, imagine relevance filters: human verification that a human was actually involved in making the music. Like Twitter’s blue tick, but for every upload. You’d verify once (potentially using blockchain to keep it decentralised), and that verification would work everywhere.
This should stop bots from uploading 10,000 tracks a day and drowning out the bedroom producers and everyone else. As one workshop participant noted, Gen Z is struggling to develop musical taste because there’s so much noise. The algorithm can’t distinguish between a track someone laboured over for months and something generated in three seconds. And more tragically, the new listeners cannot either.
The approach could be nuanced—concentric circles of classification. At the centre: fully human-created music with community verification. Next ring: AI-assisted but human-directed. Outer rings: functional music, background tracks, fully synthetic content. Everything has its place, but we’re transparent about what’s what. This fits quite neatly with recent publications around AI by BMI and STIM showing the industry might be ready to adopt this.
Exposing the hidden value in live music
As we were at ADE, we contextualised most of the discussion through the lens of electronic music, a culture that was built in clubs, precisely the place where the money trail goes bleak. DJs guard tracklistings like state secrets, venues don’t track what’s played, and producers whose tracks destroy dancefloors see close to no monetary benefit.
But what if we treated club plays like radio plays? Not through hardware in DJ booths, but through AI agents trained on specific genres. An agent that knows Tech House from Dub Techno, that can distinguish a bootleg from an original, a remix from a re-edit. These agents would monitor not just audio but social listening - hashtags, event listings, Shazam data - building a real-time picture of what’s actually moving dancefloors.
It might feel like data intelligence and surveillace, but think of it instead as infrastructure for the dancefloor. The data would do more than flow to CMOs for royalty distribution. Producers could prove their cultural impact when negotiating deals. Venues could demonstrate their role in breaking new music when applying for public funding. We could finally understand the real impact of electronic music.
Flipping the entire monetisation model
An excercise in radical rethinking. What if, instead of platforms licensing in bulk, listeners had to request access to play music? Right now, DSPs negotiate blanket licenses and artists get whatever trickles down. But imagine if every play, stream, sync, DJ play all required permission, with rights holders setting their own terms.
Maybe you let students stream for free but charge brands full price. Maybe your tracks are free for underground DJs but expensive for festival main stages. With universal IDs and smart contracts that actually work (because they have complete information), the technology that could handle the complexity is out there.
A DJ plays your track at Berghain? Payment hits before they’ve mixed out. Your song soundtracks a viral video? Compensation scales automatically with views. No more chasing payments through intermediaries who each take their cut. The money flows directly, transparently, and immediately.
Even the streaming model itself could become more nuanced. Why does background music at a coffee shop count the same as someone playing your album on repeat during a breakup? Active listening - full plays, no skips, repeats - could pay more than passive consumption. We could finally value music based on its actual impact on listeners, not just raw play counts. The ground has already been laid for this. We need to follow the money and let the business model lead the change.
The uncomfortable truth about power dynamics
Watching these ideas emerge in that room was pure joy! But we all knew the elephant in the room: none of these technical solutions matter if the industry’s power structures remain unchanged.
That’s why the most important flip chart might have been the simplest one: “Bring everyone around the table.” Not a table where everyone protects their interests, but one that invites genuine collaboration and where we admit that change is necessary and needs to happen NOW. This needs to include the platforms drowning in upload volume, the labels struggling to break new artists, the collection societies unable to distribute accurately, the managers figuring out how to build their artists a community, and, of course, artists and creative technologists.
We learned that Music Industry 3.0 isn’t about the technology per se. It’s about who gets to decide how that technology gets deployed. This workshop proved that when you get music people in a room and ask them to design a better system, they don’t create utopian fantasies. Try for yourself and see how they create practical solutions that respect both creativity and commerce.
Where do we go from here?
2028 isn’t that far into the future. If AI really takes a fifth of creator revenue in two years the path forward can’t be about waiting for a saviour platform or killer app. We need to work together to apply coordinated pressure at multiple points in the value chain.
Another takeaway from the workshop: the people who make this industry work still believe it can work better. We’re not techno-optimists or techno-pessimists. Most of us are techno-constructivists. We know the tools exist to build something better.
This workshop was part of ADE 2025, co-hosted by Music Tech Netherlands and AFEM (Association for Electronic Music). Thanks to everyone who participated, drew on flip charts, and dared to imagine something different. The conversation continues. Where do you see the opportunities for change?
LINKS (by Maarten)
🐺 The economic and social impact of independent record labels (Brodie Conley, Matías Muñoz Hernández)
“When viewed through the lens of profit, the redistributive nature of this model is even more striking: for every dollar invested, labels generated $0.77 in profit after covering operating costs, and 77% of that profit ($0.59 per $1 invested) was distributed to artists. This reflects a structural commitment to directing the majority of economic value back to artists, rather than concentrating it at the top.”
✘ This is one to digest slowly and really take in. There’s so much here to highlight, but the parts about the money really hit home. Let’s spread this research far and wide!
🏴☠️ Supreme Court hears billion-dollar battle over online piracy (Devin Dwyer)
“Copyright owners insist the risk of being sued creates an incentive for internet service providers to help root out online piracy and suspend the accounts of those suspected of dealing in protected material.”
✘ I genuinely thought this article was misplaced and came from 2001, not 2025.
🔥 Let it burn - toasting marshwellows over Adland (Zoe Scaman)
“The work is disseminating – into in-house teams hungry for real talent, into collectives forming around shared values instead of shared overhead, into direct relationships with brands who are tired of the holding company tax, into new shapes we haven’t even named yet. The demand for strategic thinking and creative excellence and brand building hasn’t evaporated. It’s just finding new vessels.”
✘ Understanding what’s happening in the world of advertising helps us understand every other industry. Music is as much about marketing as anything else and seeing how the infrastructure in Adland is changing helps us see how music will change, too.
📐 Architecture for sound: On Kevin Shields, SUNN O)))), and the demands of maximalism (Derek Blackburn)
“Both approaches require large systems, but for opposite reasons. Shields seeks instability and overlapping behaviors; SUNN O)))) seek immovability and pressure. The rigs look similar only because both philosophies demand a scale that resists miniaturization.”
✘ First, I love both My Bloody Valentine and SUNN O)))). If you haven’t experienced them live, please find a time to do before you can’t anymore. This piece does a great job of explaining how art thrives in the face of friction with technology.
💃 The choreography of posting online: Read an interview of Maya Man (Emma Grimes)
“My philosophy of authenticity is that it doesn’t exist in the way people wish it did. I don’t believe it’s possible to perform in a way that’s authentic. People will say, I just post for myself, which is a lie. They say that because they feel it’s morally better to be that way, and I really disagree with that. It’s okay to feel like you’re performing and even want to perform a bit. That’s not evil. It’s a condition of living. I’ve adopted a [Erving] Goffman-esque philosophy of performance online. Everything is a performance. Goffman was writing before the internet, so he is talking about socializing in general, which I also think is true. It’s been kind of freeing for me to subscribe to this notion that authenticity does not exist.”
✘ Everything is a performance. This newsletter included. That workshop at ADE included. You reading this included. Work with it, not against it.
MUSIC (by Rufy)
I have been walking through the streets of London listening to Dry Beetle by Rene Wise, try it sometime!! He is my favourite techno producer at the moment.
Enjoyed reading ideas and insights with a tech AND creator forward mindset. It feels very hopeful to know the possibilities of AI that help the artist community and not replace it. Great article!
A very interesting article. Thank you for your work!