Process

A transbliteration is neither a translation nor an obliteration, but both. These are Shakespeare’s sonnets, rewritten whole, then grown into sculpture. Each sonnet is first put through machine translation into Spanish, where it comes back as near-nonsense. The nonsense is then read for sound rather than sense, and whatever it triggers in my brain is written down. The Shakespeare acts as waypoints, but the journey between can be very different.

The method turned out to describe how the pictures are made, too. A diffusion model starts from pure noise and, held to a text, makes it slightly less noisy, then does that again, and again, until an image is standing there. It is the same loop: a noisy text, a neural network, and enough passes.

Every line of the finished poem becomes a prompt — material, period, place, light — and generates an image of a sculpture in a garden. Each set of images is grown from a single fixed seed, so a style can be judged by how far it bends to the words rather than by luck; it is also why one line often echoes the pose or the spacing of the line before it. Nearly every image is then grown into a 3D model.

The poems are read aloud as well, and the indentation is the score. A tab in this book is an aside, so how deep a line sits sets its pace, its level and the silence after it: an aside at the fourth remove is read about a quarter slower per word than the line it hangs off. The reading is synthesized a line at a time and assembled rather than spoken as a block, which is why the recording knows to the sample when each line falls. That timing is what the garden walks to.

The garden they stand in is a maquette. The sculptures are meant to be built, and the votes are the beginning of the decision about which ones.

Technical specifications

Medium

Poetry, AI-generated imagery, 3D digital sculpture, and synthesized voice

Dimensions

Images 1344×768 px; sculptures as glTF/GLB 3D models; readings as mono AAC audio with a line-by-line timing map; texts on the web

Tools

Transbliteration

Original poetic rewriting of all 154 Shakespeare sonnets by the artist

FLUX

Line-by-line image generation from structured prompts (material, artists, setting, era), run in Google Colab

TRELLIS

Image-to-3D generation of sculpture models (GLB)

Tortoise and Chatterbox

Voice synthesis. The reader is a blend of averaged conditioning latents, not a clone of one speaker; the poem is rendered a line at a time so the recording knows when each line falls

Latent interpolation

Slerp walks between line images, rendered as video

DynamicDiff

Custom Python/Manim tool animating Sonnet I's drafts, from Shakespeare into its transbliteration

Site

Hand-built static site (Python/Jinja2, three.js, model-viewer)