TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Judgment · AGE 17—19 · FACTUAL · EVOLVING

Offloading and atrophy

A step always handed to a tool stops the practice, and the skill can thin.

The idea

Skills are kept by practice. When a step is always handed to a tool, the practice stops and the skill can thin, because the tool covers for it. Researchers call the handing-over cognitive offloading. The evidence is early and uneven. In one clinical setting, clinicians' unaided detection rate fell after months with an AI assistant — one observational study. In higher education, over-reliance is the most reported risk in a 2026 review, which also finds structured use can strengthen reasoning. This card argues with the capstone on purpose: if judgment is the job, what happens to judgment never exercised?

Why it matters

The capstone says judgment stays yours. This card asks whether you will still have it. Both can be true, but only if you choose what not to offload.

See it in the studio

A third-year student who has generated every section since first year is asked, on a site visit, to sketch a section of the existing building. They cannot. Not from lack of intelligence, but from three years of never needing to.

Watch for this

"I'll learn that properly later." Later is when the tool is not there, and the tool is not there exactly when it matters: on the site, at the jury, at the client's table.

Try it

For two weeks, do one task unaided each day that you would normally hand over: a daylight sketch, a room count, an area calculation, a paragraph of the design statement. Then do the same task with the tool. Record where you have slipped, and where you have not.

Prove it

State what in your own practice you refuse to offload, and why, with one piece of evidence from your own work that shows the skill is still there.

How it works

Two kinds of evidence, and they do not point one way. The clinical study followed nineteen experienced endoscopists at four Polish centres and found that their adenoma detection rate in colonoscopies done without AI fell from 28.4 per cent before AI was introduced to 22.4 per cent after some months of using it. That is a retrospective, observational result, in one field, and the authors are careful about causation. The education review found generative AI acting as a cognitive amplifier under structured conditions and as a substitute under unguided use, with over-reliance, reduced analytical autonomy and cognitive offloading the three most reported risks. Its authors list their own limits: a single reviewer, English-language sources only, a two-year window. That is why this card is marked EVOLVING. The direction is clear enough to take seriously. The size of the effect, how permanent it is, and whether it transfers from detecting polyps to designing buildings are not settled, and we will revise as the evidence matures. The honest counterweight is this. The same review found that structured use amplifies rather than substitutes, and a small 2025 experiment found that a tool designed to provoke, by critiquing its own suggestions, restored critical thinking that unprovoked use had lowered. Atrophy is a tendency of unstructured use, not a law of the tool.

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