TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Ethics & Provenance · AGE 19—22 · POSITIONAL · HELD

Name, brief, audit, refuse

The refusal is the authorship, and without it the other three moves are styling.

The dilemma

The guesthouse in Mysuru. You briefed for bamboo, lime-wash and afternoon light, and the model returned a beautiful room that could be in any hotel in the world. The client loves it. The deadline is Friday. Do you ship it?

The choices

Ship it and call it warm and modern. Re-prompt once, take whatever comes back, and ship that. Or run the four moves: name the flattening to the team and the client; brief against it with named typology, materials and climatic moves; audit the re-briefed output for what the default left behind; and refuse to ship anything that reads as anywhere-in-the-world when the project is specifically here.

The consequence

Shipping the default builds the bias into a real room, with your name on it. A single re-prompt usually keeps most of the default: the lime-wash still reads as gypsum, the jali is decoration, the eaves do nothing. Nobody audited it. The four moves cost a day and produce a room that belongs to its place. The fourth move is the one that carries the weight: without the refusal, the other three are styling.

The case

The Lab's own case, from the essay this card comes from. Brief: warm, contextually grounded reception space, 4×4 m, bamboo and lime-wash, soft afternoon light. Output: pale wood, white walls, raked light, low ceilings. You could find it in any global chain and nowhere in Karnataka. The deep eaves, the jali for privacy and air, the high ceiling, the cross-ventilation — the reasons the type exists — had all been lost. The image was not wrong. The neutrality was.

Our position

The default is not neutral; it is a placeless default — the model does not compute a cultural average, it leans where its data and its tuning lean. Name it, brief against it, audit what comes back, and refuse to ship the placeless image. The refusal is the authorship.

Why we hold it

For twenty years "globally minimal" was one designer's private taste. A model flattens by default, in seconds, every time. That makes it a built-in property of the tool: visible, and so something you can refuse. More data will not fix it from inside the model. The only thing that stops it is a designer who will not ship it.

The strongest objection

"Brief against placelessness" can manufacture a new stereotype. If the remedy for the generic room is to demand that an Indian result look Indian, the model starts sprinkling jalis, courtyards, terracotta, carved stone, brass and arches onto contexts that never asked for them. A global language can be locally chosen; a conspicuously local one can be culturally fake — a jali added for a jury is no less decorative than pale wood added by a model. Specificity is not iconography, and the four moves do not say how to tell them apart. Add to that the practical objections: "refuse to ship" assumes a power most students and juniors do not have — in a firm the senior ships and the junior renders — and some clients genuinely want the global look, so refusing it is paternalism dressed up as ethics.

What would make us revise it

Models trained on well-documented regional corpora that stop producing the flat default by default. When the audit routinely finds nothing to refuse, the four moves become a checklist rather than a position, and the card is rewritten as procedure. And if the audit move cannot be made to distinguish specificity from costume, the position is rewritten around that distinction first. Reviewed every edition against what the tools actually return.

Prove it

Take a flat default output and run all four moves, ending in a re-briefed, audited result. Then explain why refusing to ship the flat image is the move that carries the weight — not the styling.

Take it to crit

Faced with a beautiful but placeless default, does the student run the four moves in order — or ship the pretty image and call it warm and modern? Ask to see the audit, not the final render.

How it works

The four moves map onto the rest of this strand. Name is the Tilted mirror card, said out loud. Brief is the six-layer brief from the Brief strand, aimed against the corpus. Audit is provenance turned on the output: what did the default leave in — and what did the re-brief add that this place never had? Refuse is the judgment this whole map says is still yours. The Documenting is representing card is what you do so that the next designer has less to refuse.

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