A career between genres (and why that’s always been a challenge)

I’ve spent most of my career making music that sits somewhere between established lanes.

I’ve never made a conscious decision to be somehow ‘difficult’ or evasive, but that’s simply where it tends to land. It leans, it overlaps, it pulls from different directions at once, and by the time it starts to resemble something recognisable, it’s already moving again.

If I look back over the years, there’s a fairly clear pattern of genre-dodging belligerence to it.

Flicker Noise started as a collision of whatever we were into at the time - guitar music, early electronics, bits of indie, bits of Balearic influence. Lunatic Calm has often been something you could say “that belongs here”, largely off the back of Leave You Far Behind, but then you listen to the rest of our catalogue and most of it bears no resemblance to that one track. Elite Force, which on paper became associated with breaks, always stretched well beyond it with 4/4 tracks, slower pieces, vocal-led material, even tracks that sat closer to Massive Attack than anything you’d expect from a dancefloor set.

Under my Futurecore guise I drifted into abstract, leftfield downtempo, while PhrackR sat in an awkward space between breakbeat and drum and bass without fully committing to either. Zodiac Cartel, which for a moment looked like it might lock into something more defined as the blog house movement enjoyed its 15 minutes, but then that started to loosen and wander as well. Then came Simon Shackleton and the material moved again, initially into something more melodic and progressive, before shifting into much more sparse, song-driven territory with The Shadowmaker, which in hindsight was probably the least “placeable” body of work I’d made up to that point (especially given what had preceded it).

None of this was strategic. It wasn’t a conscious attempt to avoid definition - it’s just how I’ve always related to music.

As an avid consumer of music, I’ve never really understood the idea of being loyal to a single genre. It’s always felt far more natural to move across a myriad of styles, depending on my mood - from Beck to Underworld, Debussy to Mogwai, Steve Reich to Bolt Thrower, Arvo Pärt to Bicep, Godspeed! to Max Cooper, Motorhead to Dustin O’Halloran. There’s no neat, convenient, thread running through any of that beyond curiosity and a perpetually open mind. I just constantly gravitate towards anything that feels like it has something resonant in it (and that resonance can take any number of forms), so when it comes to actually making music, it ends up being a broad reflection of that - a kind of accumulation of a wild number of distinct, moving parts.

The problem, or at least the tension, is that the systems and categories we’ve built into the way we classify music have never really been built to deal with non-conformity … and now we’re in the vicelike grip of algorithmic analysis as our new ‘arbiter of taste’, the picture’s more confused than it’s ever been. Why? Because the Machine simply doesn’t understand finished pieces of music now, given it doesn’t cope with complexity and nuance particularly well.

In the 90s we had record stores, where everything had to sit in a physical genre bin. It was the same story with the music press trying to decide which page something belonged on, or radio needing to know which show to slot it into. There’s always been a reliance on clear categorisation - on tramlines, and guardrails. When I ran my Fused & Bruised label back in the late 90s, we were constantly running into these issues… not because the music lacked direction, but because there were too many directions in play at any one given time. We ended up creating our own frameworks just to make sense of it. Projects like This Is Latin Amyl or Day Trip to Brisco weren’t built around genre at all, they were built around ideas - loose concepts that gave people room to explore something broader without having to pin it down to a single sound. It made the label feel more like a collection of outliers than a scene, which was exactly what we wanted, but it also made it harder to plug into a system that relied on things being easily labelled.

“It’s not that it doesn’t fit. It’s that it fits in too many places at once”

That made us the perfect fit for the ‘Leftfield’ classification. Sometimes that felt like an ongoing ‘apology bin’… the place where anything that didn’t behave itself got filed away. The misfits, the renegades, those who just had to fuck with the program and play on the fringes. That was who we were and who I still am today. That same dynamic hasn’t really gone away - it’s just evolved.

In today’s world it shows up in algorithms instead of record store shelves or the hallowed pages of actual magazines, but the underlying need is the same. Streaming platforms, playlists, recommendation systems - they all depend on being able to define what something is in order to know where to put it.

That’s where things start to get interesting.

Recently, I ran a number of tracks from the forthcoming Elite Force album, ‘Reverie’, through a set of AI analysis tools, partly out of curiosity, partly to understand how the Machine sees what I’m doing - because if the Machine sees it as X but you promote it as Y, chaos reigns, right?

The results were interesting. It came back with tags like dubstep, UK garage, electronica. None of them were technically wrong and you could hear where these interpretations were coming from. The bass architecture pointed towards the deeper end of dubstep, the drum programming and percussive sound choices leaned into garage, and the broader musicality sat somewhere in that catch-all space people tend to call electronica (the modern-day ‘Leftfield’ perhaps?).

What the algorithm had done is successfully decipher the sonic components without coming close to describing the music. It wasn’t hearing the record as a whole, it was identifying familiar audio signatures within it. The curvature of a bassline, the shape of a transient, the way a rhythm swung. All of it perfectly logical, and all of it largely missing the point - a point that only human people can, at this point, interpret.

It’s a bit like describing a car purely in terms of its components. Four wheels, an engine, a chassis, glass. You’re not wrong, but you’re also nowhere near understanding whether you’re looking at a Ferrari or a fifteen-year-old Skoda. The information is accurate, but it isn’t meaningful - and that’s where the gap starts to open up to expose a chasm between analysis and understanding, because when a person listens to music, they’re not parsing it like that. They’re not isolating the bass architecture or categorising the drum transients in real time. They’re responding to the sum of it. The mood, the intent, the tension between elements, the way it makes them feel - all the things that sit above and beyond the individual parts.

“The Machine can anlyse the parts, but it can’t understand the whole”

The Machine, by contrast, has to reduce it in order to understand it, which is fine, up to a point. The problem is that those reductions are then used to decide where the music goes, who it gets shown to, and how it’s framed. And if the music doesn’t sit cleanly inside one of those definitions, it ends up being scattered across several partial matches rather than landing properly anywhere.

You can already see how that plays out.

If something is tagged as dubstep-adjacent, it might get shown to listeners of that subset, but only a small percentage of them will connect with it. The same goes for garage, or electronica, or whatever else gets pulled into the mix. Each tag is a fragment of truth, but none of them fully describe the whole, which means the audience that might actually understand it is harder to reach, not easier … and that’s before you even start to get into how those same systems respond when you shift direction entirely from your earlier work.

When The Shadowmaker came out last year under my Simon Shackleton guise, it was a fairly sharp turn from what I’d been doing previously. More space, more focus on songs and lyrical expression, way slower tempos and less reliance on the kind of dancefloor-adjacent structure and functionality that I’d previously been known for. From a creative point of view, it made complete sense (to me, anyway), but from the Machine’s point of view, it was almost impossible to place. When it first came out it was being recommended to fans of Plump DJs and Stanton Warriors, which couldn’t have been further from the mark, and when the algorithm doesn’t know where to put your music, you’re effectively starting from scratch, but with the added challenge of existing inside a feedback loop of early responders saying ‘hell no, I want some functional breakbeat’.

What’s become clear is that there’s a disconnect between how music is made and experienced, and how it’s categorised and distributed. The more that music sits between those established categories, the more pronounced that disconnect seems to become.

This isn’t just a “me” problem.

The ones who choose not to evolve tend to find a lane and then fully occupy it, feeding the Machine with more and more identikit material to consolidate their position and identity. The Machine rewards them with exposure, receiving confirmation signals in return and consolidating their audience. The Machine is happy. The artists are happy. There’s no tension, and no resistance. But there’s also no artist development.

The artists who do choose to evolve in a more radical way often feel this friction, unless they’ve attained what I like to think of as ‘Standalone Brand Status’. Take Radiohead as an obvious example. If you remove everything we now project onto them - the scale, the mythology, the fact that they’ve become this almost untouchable reference point - and just listen to the music across their catalogue, it doesn’t sit still for very long at all. It shifts, it mutates, it contradicts itself. Try and pin it down to one thing and it immediately slips out of your hands. They would be an algorithmic nightmare, if they weren’t a global brand.

The Chemical Brothers are another good example. When they first appeared on the scene as Dust Brothers back in the mid-90s they were the archetypical ‘Big Beat’ duo, but as they’ve grown and spread their wings off the back of a hugely successful early single, they are no longer genre-constrained. Same with someone like Four Tet, who’s managed to exist across multiple spaces without ever fully belonging to any of them.

What’s interesting is that once artists reach a certain point, that problem just… disappears - not because the music suddenly becomes easier to categorise, but because the artist becomes the category. The name carries enough weight that it overrides the need for explanation. People aren’t asking “what genre is this?”, they’re just engaging with it as part of that artist’s world, but until you get to that point, you’re very much operating inside a system that still wants things to be clearly defined.

“The moment you start explaining your music, you’ve already reduced it”

The core tension here (and this is probably the part I find most frustrating) is that the moment you start trying to describe your music in genre-orientated terms, you can feel something slipping. It’s subtle, but it’s there. You go from talking about what it is to trying to explain where it fits, and they’re not the same thing at all. It ends up feeling like you’re back-pedalling and somehow apologising and reducing the finished song before anyone’s given it a listen. You’re breaking it down into fragments instead of presenting it as the whole thing it was meant to be.

No one listens to a record like that.

No one sits there thinking, “ah yes, I can hear the UK garage influence in the hi-hats, and a touch of dubstep in the bass design.” They either feel something or they don’t. It either connects or it doesn’t. It’s as simple, and as complicated, as that.

So maybe the question isn’t how to make music easier to place. Maybe it’s whether the systems we’re relying on are actually capable of handling it in the first place. The more you look at it, the more it feels like they’re built to reward clarity over curiosity… and if that’s true, then trying to force the music into those frameworks is always going to be at odds with the reason it exists.

At the same time, there are small shifts happening in the opposite direction. Spaces are opening up where the filtering is human again - where someone takes the time to sit with a piece of music, to think about it, to place it in context, to write about it without immediately reducing it to a handful of tags.

That’s part of what I’ve been finding on Substack.

Writers, musicians, people who’ve spent years immersed in this world, not trying to categorise everything perfectly, but trying to articulate what something actually is, or what it feels like, or why it matters. It’s slower, more subjective, sometimes messy and ragged but it carries passion, which is something that’s very easy to lose in more system-driven environments.

Whether that’s enough to counterbalance the direction everything else is moving in… I don’t know, but it does suggest that the answer might not be about fixing the system at all. It might just be about paying closer attention to where the music actually lands when it’s given the space to exist on its own terms… about reintroducing human judgement into a process that’s become increasingly automated.

And maybe that’s all you can really do.

Because for me, at least, the instinct has always been the same - to make music that reflects whatever I’m drawn to at the time, without worrying too much about where it’s supposed to sit… and to accept that, for better or worse, it’s probably always going to exist somewhere in between.