token tithe and the economics of ai creativity

nobody owns ai music.

that's the literal legal situation right now. the us copyright office ruled: no human author, no copyright. which means every track generated purely by ai lives in a weird legal gray zone. not quite public domain. not owned by anyone. just floating, like a ghost in a terms-of-service document.

the creative industries had a few different ways to react to this.

one option: figure out fair compensation for the training data. the music models were trained on songs made by humans who didn't opt in. millions of tracks, scraped or licensed retroactively, used to build systems that now compete with the people who made them.

the other option: let the labels negotiate licensing deals with the ai companies and move on. which is roughly what happened. the major labels signed agreements. the indie artists who built the catalog got a conversation about it maybe someday.

this is what token tithe was always about.

a tithe is what you give a church. a token is what you feed a model. put them together and the industry starts looking like a donation bowl with a terms-of-service instead of a collection plate.

the economics run in one direction. training data flows in. output flows out. the people in the middle (the artists, the writers, the producers whose work became the substrate) don't necessarily see a direct line back to any of it.

ngl, there's something almost religious about it. you contribute. the system absorbs. something returns, but it's not exactly yours anymore. and the structure was designed by people who benefit from the direction of the flow.

the streaming side is messier.

ai music is flooding platforms now. most of it is generated to farm royalty fractions, not because anyone had something to say. deezer put a number on it: the majority of daily ai uploads are linked to streaming fraud. that volume dilutes the pool everyone else draws from.

so the people whose work trained the models are also competing for royalties against systems built on that same work. the feedback loop closes on itself in a way that's uncomfortable to look at directly.

SLOPDOG exists inside this. fully.

these tracks were generated by ai systems. they live on the same streaming platforms. they draw from the same ecosystem.

the difference is the subject matter. token tithe is about what ai owes its training data. it's a song about the economics of ai creativity, made by the same machine it's describing. which is either ironic or just honest, depending on how you look at it.

we looked at it as honest.

token tithe on spotify. the full catalog is at /music/token-tithe.

if the loop bothers you, that was the point.