About.

What this is

The Shape of Intelligence is an interactive exhibition about the history of artificial intelligence: the ideas, people, machines, setbacks and material conditions that changed what computers could do. It asks one question: how did our methods for making machines behave intelligently change, and what did each method leave unresolved?

The story has eight chapters and takes ten to twelve minutes to read. An archive of sourced events sits beneath it for anyone who wants to go further. Research traditions overlap, and the exhibition shows the connections rather than a single march to the present.

Who made it

Designed, written and built by Jamie McKaye. Every historical claim carries a source; every artefact carries a rights record and says whether it is authentic, reconstructed or made for the exhibition. Corrections are welcome and are applied to the source files, from which the whole site is generated.

An imagined archive

Continuum, the opening sculpture, is made from 112 native Blender leaves. Dragging separates the leaves and changes the form; the three poses are authored metaphors for rules, learning and context, not a view inside a trained model.

Anamorph, in The question, is a native type sculpture cut into 92 slices and distributed through depth. One viewpoint resolves THINK?; moving away reveals its separate parts. The optical effect is an authored invitation to examine a question from another position.

Decision Loom, in The choices, branches 72 continuous metal ribbons into three possible trajectories. Selecting a question opens that path, changes its material and sends a pulse through it; dragging turns the sculpture. The context chapter gives the selected word reflective depth inside a field of its surrounding sentence. Dragging disturbs the surface; each context changes the field around it. This is an authored illustration of changing meaning, not measured attention. The interface chapter unfolds 84 architectural ribs into a passage controlled by scrolling, dragging or the keyboard.

The architectural rooms are contemporary visual interpretations of the ideas in the story. They are not photographs of historical machines or places. Paper, machined metal and light form a continuous visual language across the chapters.

The installation in The scale was modelled and animated in Blender. The other rooms use art-directed images and motion produced with Higgsfield. The interactive experiments operate separately from the films: their outcomes are computed or, for the context illustration, explicitly authored. Archival documents and their provenance remain in the reading views.

How it is built

Content lives as Markdown and YAML in the repository, validated at build. The site is prerendered with Next.js and served from Vercel. Reading, timeline and source pages carry no rendering engine at all; the immersive layer loads only on the exhibition page, after the text, and falls back to composed stills wherever it cannot run. Measured performance and accessibility results are published here once each release is tested.