Miguel Chevalier has been making art with computers since 1978 — long enough that “digital artist” undersells how early he started, and how consistently he’s stayed with the same basic question: what happens to an image, and to perception itself, once a machine is involved in making or reading it. His new show, “Beyond Retina,” opened November 5 at The Mayor Gallery in London and runs through December 18. It’s the most direct version of that question he’s posed yet, built explicitly around how vision is changing “in the age of artificial intelligence.”
A show that spans mediums, not just media
“Beyond Retina” isn’t a single-format exhibition. Per the gallery’s own listing, it brings together robotic drawings, dye-sublimation digigraphs on aluminium engraved with laser technology, 3D-printed sculptures, video works on screen, and an interactive installation — a genuinely mixed-medium survey of how the same underlying ideas look across very different physical and digital output forms. The show’s framing, per the gallery, traces sight “beyond the human eye into intelligent cameras, algorithms, neural networks and AI systems,” shifting the emphasis from the resulting image to the nature of perception that produces it.
Four decades of the same question, different tools
Chevalier’s practice has always moved with whatever the current frontier technology happened to be. He began tackling what one biography describes as “the question of the hybrid and generative image” in the 1980s, well before generative art had a name most people would recognize, and has returned to recurring themes — nature and artifice, networks and flows, virtual architecture — across four decades as the tools under his hands changed from early computer graphics to large-scale projection to, now, the AI systems “Beyond Retina” takes as its subject. He’s held artist residencies tied to that technological curiosity since at least 1994, when he worked at the Villa Kujoyama in Kyoto.
Why a robotic drawing belongs next to a neural network
What makes the show more than a survey of AI-art tropes is the throughline: a robotic arm executing a drawing and a neural network generating an image are doing structurally similar things — translating an instruction set into a mark — separated mainly by era and complexity. Putting a robotic drawing, a laser-engraved digigraph, and an interactive AI-adjacent piece in the same room makes that continuity legible in a way that writing about any one of them separately wouldn’t. It’s also a useful corrective to treating “AI art” as a sudden rupture rather than the latest step in a much longer history of artists handing some portion of image-making over to a machine.