Error 404: Tune Found — The Spotify Glitches That Accidentally Invented a New British Sound

At some point in 2021, a twenty-two-year-old student from Sheffield called Imogen started receiving deeply strange recommendations from Spotify. Not just mildly unexpected — genuinely baffling. A playlist she'd built around post-rock instrumentals began populating with field recordings of Estonian railway stations. A Discover Weekly dropped her into the middle of what appeared to be a digitally corrupted Welsh male voice choir. An autoplay chain started normally enough and ended, somehow, in forty minutes of unidentified ambient noise that may or may not have been a malfunctioning HVAC system.

She didn't skip any of it. She listened to all of it. And then she went looking for more.

"It felt like the algorithm was having a breakdown," she tells us, "and the breakdown was more interesting than anything it was doing when it was working properly."

When the Machine Stops Making Sense

Streaming recommendation systems are, by design, supposed to be invisible. They're meant to feel intuitive — to surface the next logical thing, to keep you in a comfortable sonic neighbourhood, to reduce friction between you and more of what you already like. When they work, you barely notice them. When they fail, you notice immediately.

But a growing subset of listeners — disproportionately based in the UK, according to several Discord servers and Reddit threads dedicated to the phenomenon — have started treating algorithmic failure not as an irritant but as a feature. They're deliberately cultivating conditions that produce broken recommendations. Seeding playlists with contradictory genres. Rapidly alternating between wildly disparate artists to confuse the taste profile. Streaming obscure recordings with corrupted or missing metadata, which the algorithm then attempts to categorise based on... whatever it can find.

The results are frequently chaotic. They are sometimes transcendent.

The Metadata Miners

The corrupted metadata angle is particularly interesting, and it's where several British musicians have started operating deliberately. When a track is uploaded to streaming platforms with incorrect, incomplete, or intentionally scrambled genre and artist information, the algorithm has to make guesses about where it belongs. Those guesses are sometimes spectacularly wrong in ways that create genuinely novel listening contexts.

Tom Bassett, a producer from Bristol who releases music under the name Clerical Error, has spent the past two years uploading tracks with carefully broken metadata. Genre tags that combine incompatible descriptors. Release years set decades into the future or the distant past. Artist names that are slight misspellings of genuinely famous acts, which causes the algorithm to occasionally slot his music into playlists alongside the real thing.

"I once ended up in a Coldplay radio station," he says, with visible delight. "Fourteen seconds of my stuff between two massive Coldplay tracks. The comments on that playlist were extraordinary. People were furious. A few people said it was the best thing on there."

He's not wrong that it's a strategy. But he's also genuinely interested in what the broken context does to the listening experience. A track that sounds one way in isolation can sound completely different when the algorithm has inexplicably placed it between a Mongolian throat singing recording and a 1970s British library music cue. The error creates meaning that wasn't there before.

The Glitch Community

The Discord server Imogen now helps moderate — called, with some affection, Playlist Purgatory — has around two thousand members, the majority of them in the UK. They share screenshots of particularly deranged recommendation chains. They post recordings of streams that went somewhere unexpected. They swap techniques for producing broken listening experiences, and they discuss, with genuine critical seriousness, what those experiences mean.

"There's a whole conversation about whether this is actually a new genre or just a listening practice," Imogen explains. "Like, is the music itself different, or is it just that the context the algorithm creates is different? And I think the answer is both. The context changes what the music is."

This is, if you squint at it, a genuinely interesting aesthetic argument. Context has always shaped musical meaning — that's why the same song sounds different at a funeral versus a wedding versus three in the morning on your own. What the glitch community is arguing, essentially, is that algorithmic error is a new kind of context-making, one that can produce effects no human curator would think to attempt.

Some of the sounds circulating in these communities are remarkable. Heavily compressed electronic tracks that clip and distort in ways that sound deliberate but apparently aren't. Folk recordings with digital artifacts running through them like interference patterns. Ambient pieces that loop incorrectly, creating rhythmic structures the original artist never intended.

Accidental Subgenre or Legitimate Movement?

The question of whether any of this constitutes a coherent artistic movement is one that the community itself debates constantly. Several musicians have started making music specifically designed to function within broken algorithmic contexts — tracks that include deliberate digital artifacts, that are engineered to confuse metadata systems, that are intended to be discovered by misfiring recommendation engines rather than human curation.

Delia Marsh, a Glasgow-based electronic artist who performs as Null Pointer, has released three EPs in the past year that she describes as "algorithm-native" — designed from the ground up to exist in broken digital spaces. Her music incorporates the kind of compression artifacts and sample-rate errors that streaming platforms sometimes introduce during encoding, essentially pre-corrupting her work so that it sounds at home in glitched contexts.

"I'm not interested in the pristine version," she says. "The pristine version sounds like it's asking for permission. The broken version sounds like it's already somewhere it shouldn't be, which is where I want it."

Whether Spotify's engineers view any of this as a problem to be solved or a feature to be preserved is unclear — the company declined to comment for this piece. But the community's suspicion is that the more the algorithm is "improved," the less interesting it becomes.

The Error Is the Point

There's a lineage here that the glitch community is often quite aware of. Glitch music as a formal genre has existed since the nineties, with artists like Oval and Fennesz deliberately incorporating digital errors into their work. The British experimental tradition has always had a fondness for the broken and the malfunctioning — from the musique concrète experiments at the BBC Radiophonic Workshop to the industrial noise of early Throbbing Gristle.

What's new is the scale and the accessibility. These broken listening experiences are happening to people who weren't looking for them, on the most mainstream music platform on earth. The avant-garde has accidentally gate-crashed the algorithm.

Imogen's playlist, the one that started all this, is still active. She's curated it now — deliberately — with the same broken logic that the algorithm originally applied by accident. It has around three hundred followers.

"Most of them found it because Spotify recommended it to them," she says. "Which I think is funny. The algorithm keeps accidentally doing the right thing."