AI music is the new four-track recorder

I've been around the Swedish music business for a long time, which is a long way of saying I've had this exact argument before. Many times. Usually in a control room, with someone holding a coffee that's gone cold.

Early on, the engineers around me were certain samplers were the end of musicianship. A few years later the same people were certain Pro Tools was the end of recording. Then Napster, then MP3s, then Auto-Tune, then streaming, then bedroom producers with a cracked copy of Fruity Loops. Each time the funeral was scheduled, the music showed up anyway, and it sounded like the tool that was supposed to have killed it.

So when the team started talking to me about ai music and a game, I told them what I have told people at labels before about samplers: the kids who pick this thing up first are going to make the next decade of music, and the rest of us should pay attention instead of writing op-eds.

The tool changes. The argument doesn't.

The four-track didn't replace the orchestra. The drum machine didn't replace drummers (it gave us hip-hop, which turned out fine). Auto-Tune didn't replace singers, it became a sound, then a genre, then wallpaper, and now it's a tool again. Every one arrived to a chorus of obituaries. Every one ended up being how music gets made. ai music is that chorus, this round. I've sung this part already.

What it actually is

Strip the discourse off and you're looking at a synth that responds to feelings instead of chord charts. You pick a mood (euphoric, dreamy, aggressive, melancholy), and the machine hands you back a song. Vocals, beat, lyrics. Don't like it, make another. Like the chorus but not the verse, keep the chorus and try again. The same loop a kid with a four-track in a bedroom was running, compressed from a weekend to about a minute. It's play. The same play that got me into this business.

Why a game

Here's the part I care about. A studio costs money. A producer costs more. Knowing a producer is the most expensive thing of all, and that one isn't for sale. I spent years watching talented kids never get heard because they didn't know the right person at the right label, and the system hasn't got much fairer since.

A fourteen-year-old in a small town with a melody stuck in her head has none of those things. She has a phone. So we built a game where the phone is enough. Pick a feeling, hit release, hear yourself back. The career layer (venues, manager, fans) is the wrap. Underneath it's just: you made a song, that's a real moment, go make another one.

A point I want on the record

The songs you make in here are yours. They live in your account. They don't feed somebody else's training pipeline, and we won't quietly change that next year when a clever term sheet lands on the table. I've watched the rug get pulled on artists often enough that I asked about this on day one. It's in the pledge for a reason. I wouldn't have signed on otherwise.

The kids are going to be fine

Bedroom pop happened because cheap DAWs landed in bedrooms. Hyperpop happened because Logic and a Discord server landed in the same room. The next thing will come out of these tools, and someone in their teens is going to make it on a phone, on a bus, on a Tuesday. I want the door open when they get there.

A friend of mine, an engineer I've known for a long time, sent me a voice note last week. His daughter, fourteen, had made a song on her phone the night before. He played me thirty seconds. It was genuinely good. He didn't know what to do with the feeling, so he sent it to me, and I sat in my kitchen and listened to it twice.

Peter