So back in 2006 or something, I was heavily into live coding. It’s a way of live composing, improvising and performing electronic music by writing code in real time. I was doing this at OKNO, organizing SuperCollider workshops, making short experimental performances, etc.

It met with a bit of a backlash because it was perceived as too geeky. Which, I mean, it kind of is. Watching somebody type code on a projected screen is hardly a punk rock concert. Or is it?

Apparently live coding is back with a vengeance. Algoraves are hype, people are dancing to code again, and somehow the whole thing feels more relevant now than it did twenty years ago. Especially because live coding suddenly feels a bit like the anti-AI. Live coding also generates things from text, but the relationship between the text and the result is completely different. You see what happens, and you see it happen exactly as it flows from the fingers of the artist. There can be randomness, recursion, feedback and generative systems, but there is no black box vomiting out something vaguely related to what you asked for. The instructions are explicit, the system is authored, and you can inspect it.

The AI revolution does funny things. Live coders now somehow look like the real performers, while electronic music has been plagued forever by people basically pressing play. Manual labor turns out to be one of the hardest things to automate, while scientific research, writing and huge code bases are suddenly under threat.

There’s another reason why returning to live coding makes sense for me. One of the things that made me gradually shift from being an electronic music composer to working mostly as a visual artist was realizing that, from the point of view of code, sound and visuals aren’t actually that different.

An oscillator, a moving object, a rhythm, an animation, a filter, a camera movement: they’re all values changing over time, relationships, functions, signals, events. Obviously the way we experience sound and image is completely different, but underneath, when you’re building them in code, the concepts start to overlap in really interesting ways.

It also connects to another problem I had been thinking about for a while. A few years ago I made Blue Brute, my interpretation of the tale of Blue Brute. It was a kind of stripped-down live cinema performance using mostly camera movement, 3D environments, props and electronic sound. I really liked the piece, but I was never completely satisfied with how to perform it.

The live cinema and electronic music parts require very different kinds of attention. One moment you are thinking rhythmically and listening to sound, the next you are thinking spatially and cinematographically. It feels like switching between two instruments all the time. Which is kind of funny, because in code they don’t have to be two instruments at all. Code doesn’t really care whether a number is moving a camera, changing the pitch of a sound, placing a building or switching a light. The same logic can sit underneath all of it.

So maybe live coding is a much better way of bringing those things back together. And that brought back a couple of older obsessions at once.

When I was working with Ofer Smilansky on ATK!, we worked a lot around architecture and I had an absolute obsession with Brutalist, Metabolist and utilitarian buildings. I tried making those kinds of spaces in 3D already back then, but honestly I just didn’t have the skills yet to pull off an entire project.

Void

Meanwhile, through my work at CREW, I’ve also gone pretty deep into simulated 3D acoustics, procedural environments and systems that respond to virtual geometry. So somehow all those old and new threads fused again. Live coding, architecture, live cinema, procedural generation and simulated acoustics.

Which is basically where MEGASTRUCTURE comes from.