TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Foundations · AGE 15—17 · FACTUAL · DURABLE

Naming versus making

A sorter is checked against a label, and a maker against criteria you set.

The idea

Some models sort things that already exist: this is a chair, this façade is Art Deco. Others make something that was not there before: a new image, a new paragraph. Sorting can be checked against a label: the façade is Art Deco or it is not, even where the edges of "Art Deco" are arguable. Making has no single answer sheet. You check a made thing against criteria you set: is the dimension right, is the fact right, does the code run. That difference changes how each kind can fail, and how each must be checked.

Why it matters

You check a sorter against truth. You check a maker against your own criteria. Mixing up the two checking styles is how bad work slips through.

See it in the studio

A sorter tags your photo archive by building type. Spot-check fifty tags and you know its error rate. A maker proposes a façade. There is no ready-made error rate; you set the criteria, then you measure.

Watch for this

Expecting a maker to be "accurate" without saying against what. A sorter is accurate or not against a label. A made image, paragraph or piece of geometry can be checked for accuracy too — a dimension, a fact, a result — but only against a criterion you bring.

Try it

Use one tool to name the style of five buildings, and one to make a building "in a style." Write one sentence on how you checked each. The sentences will be completely different.

Prove it

Give one example of a naming task and one of a making task from design work, and state how each gets checked.

How it works

The textbook words are classification and generation. One famous model family, CLIP, learned to connect images and words so well that it can name what it sees. Generative models run that connection in the other direction. Later on this map, computer vision covers the naming side in full.

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