Auren Voss~4 min read

Every ring shared a style. None of it added up to a signature.

A side project built alone, with no ring made and no one in the jewelry trade consulted.

  • Role: Art direction, design system, prompt architecture
  • Timeline: Side project
  • Team: Solo
  • Impact: Not measured yet
Skip the static, live it

Overview

Generating an attractive ring is easy, and I wanted to know whether AI could help build a recognizable design language.

I built a generative jewelry exploration system. Silhouette, structural behavior, negative space, material, surface, asymmetry, and emotional character were separate variables I could move one at a time. The first round worked, then stopped working. Every output converged on the same tall, skeletal, pointed ring.

Nothing here has been made, quoted, or worn. The house is Auren Voss; at least two boards still print its earlier name, Varex.

Worked with: no one; no jeweler has seen the work.

Scope

I owned

  • The design grammar: references turned into rules
  • The variable system, composable rather than one long prompt
  • Diagnosing the convergence, and abandoning a direction I liked rather than trying to improve it
  • The banned-vocabulary list, and the engineering vocabulary that replaced it
  • The form exploration engine
  • Three house-signature territories, and the standardised R&D board that made them comparable
  • Every judgement about which outputs were worth keeping

Research

I translated contemporary jewelry references into rules, because an image can be matched and a rule can be applied to a new case.

The grammar covered sculptural sterling silver, architectural negative space, asymmetrical balance, and structural ribs. I split it into four axes: shape, behavior, surface, and emotion. One long prompt gives one lever with everything attached to it. Four axes let me hold three still and move one.

Shape

  • skeletal
  • fractured
  • folded
  • layered
  • cantilevered
  • interlocking
  • monolithic

Behaviour

  • compression
  • torsion
  • suspension
  • interlock
  • shear
  • expansion

Surface

  • brushed
  • satin
  • forged
  • carved
  • oxidised
  • mirror-polished

Emotion

  • serene
  • threatening
  • mysterious
  • protective
  • predatory
  • ceremonial
The Glacial Era ring concept board: a tall, sharply pointed silver ring shot against dark ice, with a keyword list, a design-intent paragraph, a material and finish list, and top, interior, profile, and worn views beneath it, plus a millimetre scale reference and three silhouette studies.
One output in the board format: keywords, intent, material, six views, and a scale reference. The format is the part that made anything comparable later.

Decision: ask what the ring does

I stopped asking what the ring should look like and asked what it should do.

Two boards both claim “three eras, one bloodline,” and seven eras between them resolve to the same radial spike. The convergence was in the question, not the outputs. A look is a destination, and a behavior such as compression, torsion, or shear can be met by forms nobody has drawn.

I banned words with a picture already attached, including spike, claw, fang, and blade. Engineering language replaced them. The cost is that half the outputs got worse, and the other half were ones nothing else would have produced.

A three-panel board titled The Eclipsed Crown Collection, showing the Lunar Era eclipse ring, the Volcanic Era vessel ring, and the Ritual Era altar ring, each with keywords, materials, four views, and a design-intent paragraph. All three are tall, spiked, skeletal silver forms.
Three eras. One bloodline.
A second three-panel board with the same title and layout, showing the Desert Era solar spine ring, the Cathedral Era spire sentinel ring, and the Underwater Era tideborn ring, each set with a citrine or onyx stone. The silhouettes closely resemble those on the first board.
Also three eras, also one bloodline, and a shared-DNA row at the foot listing the six traits they have in common. Seven eras across the two, and one shape.

Banned outright

  • spike
  • claw
  • fang
  • talon
  • bone
  • gothic
  • blade
  • horn
  • armor

Used instead

  • compression rib
  • load path
  • tension arc
  • pressure seam
  • acceleration line
  • structural fin
  • directional mass
  • cantilever
  • shear surface
The Strike Collection board: three lattice rings named Intercept, Pursuit, and Terminal, each with front, side, and top views, a worn view, and a structural-intent list describing dual load paths, cantilevered mass, internal lattice that dissipates impact forces, and edges held in compression or tension.
The same three rings described only by what they do. Every annotation names a force (converging load paths, cantilevered mass, upper edge in compression and lower edge in tension) and not one of them names a shape.

Redesign

One standard board format made rejecting a direction as cheap as generating one.

Every board fixes scale, material, views, and annotations, so the only variable between two boards is the design language. “Do I like this ring” takes a second and never told me whether a language was forming. “Does this establish something recognizable” is slower and sometimes says no.

I compared three territories through it. One made objects feel dangerous because they were certain rather than sharp. One asked how little material could imply a much larger object. The third, Slice Architecture, translates a Calatrava structure and The Broad’s porous veil into a ring.

A six-column comparison board titled Weaponized Elegance, showing rings named Predator Claw, Hunter's Edge, Kill Switch, Raptor Crest, Voidbite, and Silent Strike. Each column carries the same rows: orthographic views, a worn view, a silhouette profile, a structural-language note, material and finish, and a scorecard.
Six directions in one row, with identical rows underneath so the silhouettes can be read against each other rather than admired one at a time.
A single-ring board for Apex Hunter 03: a swept lattice form with design intent, philosophy, and signature-element notes on the left, six orthographic views and three worn views on the right, and structural language, silhouette profile, lattice detail, and material notes along the foot.
The one that came out of it. The signature element is named as a structural path, a spine that begins in the band, accelerates across the finger, and leaves the form, rather than as a motif.
The Impossible Bloom R&D board: three open, fractured silver rings named Origin Burst, Inverted Emergence, and Fractal Fracture, with five orthographic views each, worn views, a 22mm scale reference, a variable-thickness diagram running from 2.0mm to 0.3mm, a hollowed-for-lightness section, and a philosophy panel reading “not a flower, not a symbol”.
The constraints, reported back on the board itself: a 22mm height, thickness stepping from 2.0mm down to 0.3mm at the edge, and a stated weight of 12 to 15 grams. The negative-space figure it prints is 60 to 75%, against a brief that asked for 75 to 85%.
The Slice Architecture 04 board, credited to Calatrava and The Broad, with two architectural reference photographs at the left, four orthographic views of a tall pierced lattice ring, an exploded axonometric naming ten components, an interlocking-mechanism diagram, detail studies, and a scale comparison against a coin.
The third territory, and the only one that pins its references to the board: a Calatrava structure and The Broad’s porous veil, named where anyone can check the translation.
An Apex Collection R&D presentation board, printed with the earlier name VAREX, showing three rings named Ascent, Coil, and Razor, each with orthographic views, on-hand photographs, and a construction and structure list describing load paths, torsional rigidity, and thickness transitions.
Three behaviours in one territory (a lunge, a strike, a pierce), kept on the one board so the comparison is between structures rather than between moods.

Scope

The platform is live, and you can use it.

I built it after the boards: a form exploration engine. The designer sets the forces an object is organised around, and it builds geometry, measures what it built, and scores it on seven axes. Its interface is not the direction I would take now.

Results

There are no results, and the scores on the boards are not mine.

Several boards carry an originality scorecard rating themselves between 8.1 and 9.6, and one scores its own social media impact at 9.5. The system wrote those, and no human graded them. The annotations are generated text, and typos such as “AESTETIC” show it. One brief asked for 75 to 85% negative space and the board reports 60 to 75%. Neither was measured.

The only person who has judged whether these add up to a signature is me

The grammar did widen the space, and the later boards reach structures the first round never did. A house language is defined by being recognized by someone other than its author. That is the whole of what is unproven. The model has no opinion about whether its favorite shape is a signature or a rut.

Sam Cusano