An architectural method for digital reading

The reader highlights.
The AI illuminates.

AI Augmented Reading is a method by which a reader highlights any descriptive passage in a digital text and receives a contextually appropriate visual β€” inline, instantly, without leaving the page. A working prototype with two generation paths: schema-bounded rendering for reliable, verifiable visuals, and a generative path for open-form descriptive content.

View the demonstration

The concept in numbers
2
Generation paths: schema-bounded & generative
1
Working prototype, available to view
5
Steps from highlight to inline visual

Chemistry is built out as a schema-bounded domain with deterministic rendering; a generative path produces open-form visuals for other descriptive content. The method is designed to extend to further domains.

The taxonomy the method is designed to span

Every type of descriptive passage.
One method.

The categories below are the design vision — the full range of descriptive content the method is built to address. Chemistry is implemented today as a schema-bounded domain; the others indicate where the method is designed to extend.

πŸ«€
Anatomical
Cardiac cycles, pharmacological cascades, cellular mechanisms
πŸ§ͺ
Chemical
Molecular structures, reactions, equilibrium
● Built β€” schema-bounded
β˜€οΈ
Astronomical
Orreries, planetary systems, orbital diagrams
βš”οΈ
Tactical
Battle formations, military dispositions, terrain maps
🌊
Meteorological
Storm systems, pressure fronts, weather patterns
β›΅
Geographical
Voyage routes, expedition tracks, cartographic journeys
πŸ›οΈ
Architectural
Floor plans, site diagrams, spatial reconstructions
🎡
Musical
Waveforms, instrument layers, acoustic dynamics
🍽️
Culinary
Recipe flows, ingredient diagrams, preparation stages
πŸ•ΈοΈ
Relational
Character maps, timelines, narrative structures