TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Foundations · AGE 16—18 · POSITIONAL · HELD

Why this matters here, now

The image tools imagine India thinly, so this literacy must be built here, now.

Our position

AI is arriving in Indian architecture education faster than the answer to who will teach it. At the same time, in the Lab's own tests, for the typologies we have tried — ordinary housing, tier-two streets, regional vernacular, technical drawings — the global image tools imagine India thinly while sounding confident about it. It is not one gap: a Mughal monument, a Bengaluru apartment block, an Adivasi settlement and a Chettinad house are represented very differently. Building this literacy here, now, in India's own institutions, is how we defend ourselves and how we describe ourselves, at once.

Why we hold it

Two things add up: the institutional push is real (a national mission, school frameworks, university adoption), and the representation gap is real where we have looked. Teachers who wait get both problems at once, in the same semester.

The strongest objection

Where is the evidence for the premise? The Lab's tests are small and its own, and "India" is too heterogeneous for a national claim made from intuition. The gap may be real for some traditions, regions, languages and drawing archives and absent for others: metropolitan commercial interiors are probably well represented; a village house in Koppal probably is not. Until a benchmark exists — Indian typologies, regional materials, non-metropolitan streets, technical drawings, terms across languages — a sceptic can dismiss the whole position as a plausible story without numbers. And digitisation-first carries its own risk: rushing India's built culture into open datasets may mainly feed platforms elsewhere, on terms set by whoever digitises first.

What would make us revise it

A benchmark run — ours or anyone's — showing the tools represent the typologies Indian students actually design at parity with the West; or evidence that early AI literacy in Indian schools of design mainly increased dependence on external platforms rather than local capability. The Lab intends to run the benchmark itself after this edition and to publish it either way. Reviewed every edition.

Try it

Generate "an Indian street" ten times. Then ask for a street in Kannada — "ಕರ್ನಾಟಕದ ಒಂದು ಬೀದಿ" — and in Romanised Kannada — "Karnatakada ondu beedi" — ten times each. Count the clichés in each set, and note which prompt the tool understood. Then photograph your own street. The distance between the sets is this card's argument, in pictures.

Take it to crit

Argue for or against: "Indian design students need AI literacy before Indian datasets exist." Use one generated image and one real observation as your evidence.

What this idea builds on

What this idea opens up

Sources

Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.

Age grows from 11 at the centre to 22 at the edge, and six sectors show the learning strands. Tab into the map and the arrow keys step from idea to idea, following the links where there is one. Enter opens the idea under the cursor, and E reads out its links and the reason recorded on each. Press slash for Search, question mark for the full key list, and Escape to leave. Open Ideas for the complete readable list, including what each idea builds on and what it opens up.

LOGIKA · RBDS AI LAB INDIA
ON-RAMP · AGE 11 · FIRST ENCOUNTERS, NOT GATES — IDEAS · — DEPENDENCIES
DONE
OPENS NEXT
SOLID — STANDS ON