SpinRoom: a city for AI-music creators, revealed one door at a time
The situation
Discord communities built around AI music have developed a real culture of live listening sessions. A DJ hosts. Artists drop tracks. An audience tunes in together through a voice channel. The format works, and people show up for it.
But the culture ran on improvised infrastructure, and it had a bigger problem underneath. When a session ended, everything it produced disappeared into chat history — no record of what played, who made it, or how it landed. A whole scene was making real work and watching it vanish the moment the song finished playing. There was no home for any of it: no catalog, no proof of authorship, no way to turn a good night into anything that lasted.
What was actually wrong
The live session itself was held together with a notepad and a dozen browser tabs. The DJ built a queue by hand, watched a busy text channel for submissions, opened each link one at a time, and tried to stay present for the room all at once. A listener who did not already know a specific DJ or server had no way to find a session at all. Word of mouth was the whole funnel.
And past the session was the real gap. An artist who played a great track had nothing to show for it afterward. No artifact. No standing. No record that a room full of people heard it. And no answer to the question that follows anyone making music with AI tools: who actually made this? The Copyright Office's January 2025 guidance made the stakes concrete — what is protected is human authorship documented before the AI generates — but nothing in the scene kept that kind of record. The work was real and the evidence was nowhere.
What I built
SpinRoom, built as a city rather than an app. That framing is the whole design, not a metaphor bolted on afterward.
The free front door is the live listening session: a web DJ booth, a Discord bot, and a real-time public audience view working as one system. An artist drops a Suno link in Discord; the bot pulls the title, artist, and cover art, builds the queue, and opens a thread for an inspiration note and lyrics. The DJ gets one screen for queue, playback, and voting. Anyone can walk into the audience view with no account. And crucially, every play now persists as a permanent, scored record instead of scrolling away.
Behind that door, the city opens one step at a time. The Welcome Center is the moment the platform puts your name on the door — one account, one song record. The SoundCard is a receipt, issued automatically the first time a song plays for real people, with a "heard by" count instead of a play count. SoundStage turns those records into a public standing with quarterly seasons and five ways to win. Creator Station becomes a permanent address assembled entirely from the record — every session, every room, every score.
Then the deeper city: a Song Catalog that is the permanent home of a creator's body of work; a Trust Office that keeps the contemporaneous record of how each track was made and answers "who made this?" on its own merits; a Lyric Studio where AI-assisted writing logs every step as provenance rather than a prompt; and a creator Storefront and Marketplace where artists sell music, art, and lyrics at prices they set and own. Twenty-four features in three layers — recognition, catalog and tools, and commerce.
The reveal is the product
This is the part I am proudest of, and it is the reason the city framing is real. The recognition layer is free — no plan, no trial, no expiry — and the features are uncovered in a deliberate order, because each free step creates the pull toward the next. You hear it before you decide to go inside. You register to put a name to the sound. You earn the artifact by playing for a room. You compete for a standing. You end up with an address. By the time anything costs money, the creator already lives in the city.
The record is the spine that holds it together. The card, the board, and the station are all assembled from the same session record — the community generates it by showing up, and the platform's job is to keep it honest. Nobody is asked to remember to document anything. The proof is a byproduct of the night.
What changed
- The live session runs itself: the queue builds from Discord, the archive is automatic, the DJ can just host.
- The record survives. A play leaves a permanent, public, scored artifact instead of vanishing into chat.
- Authorship gets a contemporaneous record — the evidence is built at creation time, not reconstructed after a dispute.
- The recognition layer is free, with no trial and no expiry; catalog and commerce are opt-in subscriptions.
- Honest status: in alpha, built and operated by one person. The recognition core is live in alpha; the rest is partly built or finished-in-design and shipped in the open, labeled as what it is. No usage numbers, and I will not invent any.
What this says about how I work
I build around a culture instead of over it — submissions still happen in Discord, where the scene already lives. I treated the narrative arc as the core design decision, not the marketing, because how a platform reveals itself is part of what it is. And I report status plainly: this is in alpha, and it says so on every screen.