TERMINAL VELOCITIES: An Interview with LEGIO X
“I don’t overthink it—I just let the pieces come together naturally once the vibes feel right.” - LEGIO X
VELOCITIES TERMINAL, LEGIO X, 02:10 / 2816 x 2048 / 25 fps, AI video created using Neural Frames, audio created using Udio (Still image from the video).
Born in the ‘70s and based in South Florida, LEGIO X is an artist and self-taught musician. At the beginning of the Web3 wave, between 2019 and 2022, he observed and appreciated works by artists like AL Crego, ex_mortal, Elbi, loackme, Max Capacity, polygon1993, and others. However, it wasn't until late 2023 that he began minting his own creations—and LEGIO X hasn't stopped since.
“I dabbled with AI for years before fully diving in around 2022. From 2015 to 2021, the results I got weren’t nearly as interesting as the work being done by artists training their own GANs. That changed when I discovered DALL·E Mini in 2022,” he recalls, reflecting on the origins of his current AI practice.
With a new solo exhibition at objkt.one Gallery this Thursday, March 13th, LEGIO X presents TERMINAL VELOCITIES, a series where he transforms retro-futuristic computer office scenes into mystical cybernetic raves, fully embracing what he calls the "visual chaos" generated by AI.
The collection consists of ten pieces, each depicting figures in a hypnotic dance—at times resembling working-class people, at others, business executives—set against stock market graph lines and computer wireframes. All of this unfolds to the pulse of electronic synthesizers, in a hypnagogic pop style, with a precisely chaotic black-and-white aesthetic punctuated by flashes of RGB.
Below, I speak with LEGIO X about this drop and the artistic vision behind his work.
TTMYU, LEGIO. X, 01:43 / 2816 x 2048 / 25 fps AI video created using Neural Frames audio created using Udio. (Still from the video).
Could you share a bit about your professional background? I read that you are also a musician, a drummer. And I’m curious: what exactly is The LEGIO X BIG BAND?
"Professional background"? Not sure about that—I never set up a LinkedIn profile, so I can't really claim any official professionalism. I'm a self-taught musician and have played most instruments—sometimes even well enough to "fake it" alongside "real musicians." Guitar, bass, and keyboards are my strongest suits, but my drumming style completely disregards conventional rhythm, timekeeping, and structure. I've been told I play like a Jackson Pollock painting, which is probably the best compliment I've ever received.
The LEGIO X BIG BAND was a bizarre one-off "anti-music" project from a few years back. I programmed a long loop of synthetic drums, electronic textures, and basslines, all in different time signatures, stretched out to an hour in length. Then I recruited extra musicians to play along—bass, keyboards, guitar, sound effects—all using this chaotic backing track as a guide. It was essentially an MK-Ultra experiment on how much psychological mayhem a musician could endure while trying to keep up with a backing track that made very little musical sense. Later, I mixed everything together and added narration from a very drunken, very sweary Australian ranting about God-knows-what.
For anyone curious to hear some of my solo music, here are a couple of tracks:
"Medieval Kaos Ultra"—a bonkers sound collage, also from 2020
How and when did you get involved with Web3?
COVID was a mentally dark period for me, and I found a lot of inspiration in the work I was seeing on Twitter from various GIF and glitch artists. Their work felt like magical totems, hypnotically suggesting a prismatic escape from the daily black-pill matrix. From 2019 to 2022, I was deeply engaged in appreciating the art of AL Crego, Klaus, ex_mortal, Elbi, Max Capacity, loackme, polygon1993, and others—though only as a viewer through the Twitter lens.
In early 2021, a friend of mine who runs a record label in London did an NFT project with Kidmograph, tying it into one of his most popular albums. It was released on Nifty Gateway and sold extremely well. That was when I saw Web3’s potential as a game-changer, but I wasn’t in a rush to mint my own work. My first-ever mint didn’t happen until the very end of 2023, when I was selected for a show curated by Le Moon. My second mint was funded later that year through a "right-to-mint" initiative by Claire Silver/Accelerate Art.
Did you start using AI right away, or did you experiment with other techniques before incorporating it into your image production process?
I dabbled with AI for years before fully diving in around 2022. From 2015 to 2021, the results I got weren’t nearly as interesting as the work being done by artists training their own GANs. That changed when I discovered DALL·E Mini in 2022—I started obsessively making images whenever work was slow. I still use many of those early, primitive images when building my current models.
Most people were using Midjourney, so I made a conscious effort to explore tools that were more "off the beaten path." Before AI, I was into stop-motion animations and experimental videos, often based on photo sequences shot on a handheld camera. Recently, I’ve been stockpiling new photo sequences, aiming to merge some of my older, more "analog" workflows with my current AI-based processes.
I always try to create as much raw material as possible—but in the AI process, sometimes the "raw material" also functions as a finished piece. My process used to involve animating sequences of images and finding the right velocities for a visually pleasing flow.
With most of my AI work, it's still about linking images together in a way that "works"—but AI allows you to do it much faster and steer it into places that weren’t possible before. That said, there’s a depth and richness to making videos by hand (using cameras, for example) that I don’t think AI can fully achieve yet.
ELECTRIC AVENUE, LEGIO X, 01:37 / 2816 x 2048 / 25 fps, AI video created using Neural Frames audio created using Udio. (Still from the video)
Do you also create AI-generated music for your pieces?
Yes, but that’s only something I’ve started doing recently. In the past, most of my music was created by human collaborators.
Could you tell us a bit about the concept and creative process behind Terminal Velocities? What inspired the audiovisuals?
This series has definite nods to AL Crego and Françoise Gamma—MacPaint textures, primitive 2D or wireframe renders of humans, with control bars and software menu windows visible. But whereas their work is very "pixel-perfect," I embrace the visual chaos generated by AI. The AI is essentially trying to create something outside its intended function.
Most of my prompts ask for clean vector wireframes, but they run on custom models that are deliberately imprecise and messy. I started with a few basic prompts and settings, then ran them through a couple dozen different models, generating 5 to 10 minutes of video per model. Some models "clicked" better than others.
Then I introduced unexpected elements—steering the imagery toward ritual and spiritual dancing (like whirling dervishes). Suddenly, what started as cold, corporate-office aesthetics turned into mystical cybernetic raves.
Your work has a futuristic-retro feel, possibly due to the use of synthesizers. It reminds me of late-'70s to early-'80s music. Even your static works have that neon, RGB-heavy aesthetic. Can you explain how you define the soundtracks for your audiovisual pieces? Do they come before or after the visuals?
For the most part, I aim to evoke specific eras when creating these pieces, but I allow for some flexibility. Sometimes, slight mismatches work in unexpected ways, and I keep them.
I made over 800 songs for this project, then narrowed them down to about 60 favorites. While reviewing visual footage, I would play those tracks and see what clicked. Sometimes, a perfect match happened immediately. Other times, I’d isolate 90 seconds of audio and search for a visual counterpart—or vice versa, looking for the right soundtrack to fit a certain visual sequence. The process involved a lot of back-and-forth and chance. I don’t overthink it—I just let the pieces come together naturally once the vibes feel right.
14) HEARTBURN BIBLE FIRE SYRINGE, LEGIO X + shaun keenan, HOT CROSS NUNS Series, animated GIFs 512 x 768 PIXELS , 25 FPS, VARIOUS LENGTHS (Still from the GIF)
Do you enjoy collaborating? Which artists in this space do you admire? Could you name a few?
Collaborations are awesome, for sure. I'm especially proud of the HOT CROSS NUNS collab I did with Shaun Keenan back in November—hoping to have more collaborations in the pipeline this year with Shaun and other artists.
I admire many artists in this space and try to collect their work whenever possible. A quick (but not-so-short) shortlist would include Frank Manzano, Santiago, Shaun Keenan, little cakes, Lily Illo, 0009, klaus, ex_mortal, RJ, Yuri_jjjj, skomra, solz, nuv1914, Pipi Universal, Sabato, Gozo, Xu0xo, michi_asu, Hasdrubal Waffle, Objektpermanenz, Kaparetti, Genresis, and many more.