TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · The Brief · AGE 16—18 · POSITIONAL · HELD
Name the job, not the deliverable
The purpose comes first in every brief, because a deliverable alone has no way to fail.
Our position
This card is about one distinction: purpose against deliverable. "Draw a diagram" tells the receiver what to produce. "Compare the two access strategies so we can choose one" tells the receiver why — and the diagram follows from it. Both may end in the same drawing, but only the second tells the model, and you, what a good answer looks like. We hold that the purpose comes first, in the first line, every time, and the deliverable after it. THE BRIEF, the next card, builds the full structure on this one move.
Why we hold it
A model fills whatever you leave unsaid with its own assumption. If you name only the deliverable, it guesses the purpose, and its guess shapes the work instead of yours. Naming the purpose also gives you a test: the output either serves it or it does not. "Draw a diagram" has no way to fail, so it also gives you no way to judge the result.
The strongest objection
In exploratory design the purpose may not yet be knowable. Drawing, modelling and making images are sometimes how the question gets discovered; the purpose arrives after the artefact, not before it. Requiring a stated purpose can close an open study too early and turn an exploration into a task. A student taught to always state the job may invent one to satisfy the rule, and then design against the invention.
What would make us revise it
Evidence from studio that open, deliverable-first runs in early exploration lead to stronger questions than purpose-first briefs do, judged blind on the work that follows — or models that reliably infer the purpose from the artefact request alone. We already teach that the dreaming register stays open on purpose; if that openness serves early design better than a stated job, this card moves to "name the job once you can." We test it against student work every edition.
Try it
Take three requests you actually typed this week that begin with a verb and an artefact: "render", "generate", "draw". Rewrite each as the purpose the artefact is supposed to serve. Run the old and the new. Which outputs could you judge, and which could you only like?
Take it to crit
When the student briefed the model, did they state the purpose the work should serve, or only the deliverable they wanted back — and can they show where that difference changed which output counted as good?
What this idea builds on
What this idea opens up
Sources
- briefing-ai-field-guide
- prompt-theory-primer
- Your Prompt Is Your Brief
- Prompt Literacy for Architects: Briefing GPT‑5.2 Like a Studio Team
Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.