Make the heritage project before the proposal

AI lets us turn cultural ideas into working experiences while we are still figuring them out.

A heritage project often starts with material that deserves more attention. Handwritten records. A research database. Poems that rarely travel beyond their original language.

Then comes the work of explaining what it could become. Proposals, briefs, funding applications. The public experience arrives much later.

AI changes that sequence. We can make a small version first. Let people explore it. Find out what matters. Build the project around what we learn.

Our working demos show what this can look like.

Make the archive searchable

Peel Park Diaries uses AI to transcribe selected pages from Salford’s Parks Report Books, identify people and places, and trace themes across the records.

Suddenly, someone can search for a gardener or follow a thread about wartime life. The collection offers new ways in.

The transcriptions are clearly labelled as unchecked AI drafts. That matters. A working prototype can demonstrate an opportunity while showing exactly where more work is needed.

It also gives archivists and potential partners something concrete to assess before planning a larger project.

Turn a collection into a question

The Manchester History demo, Who Knew Whom, turns a natural-history research collection into a network of people, organisations, places and events.

The network invites questions. How did specimens reach museums? Who connected working-class botanists with scientific societies? Where did knowledge and money cross paths?

Its connections distinguish documented relationships from unconfirmed links and interpretations offered for testing.

That is an important design choice. A line on a screen can look convincing. Showing how much confidence it deserves makes the research open to challenge.

Making the experience helps us see what to investigate next.

Give culture new routes to an audience

Love Manifesto invites readers to explore poems, songs and stories through a world map. Selected texts appear in their original language beside English, with new translations identified as AI-assisted.

The working collection makes curatorial questions tangible. Why this poem? What does the translation lose? What other voices belong here?

Readers can respond to something they have experienced. A collaborator can suggest another work. A speaker of the language can improve a translation.

The prototype gives participation a starting point.

Build enough to bring people in

Across these demos, AI helps turn source material into something people can search, navigate and discuss.

That changes who can shape a project, and when. Archivists can challenge the transcription. Historians can question the connections. Communities can point out what is missing.

But they need to arrive while the project can still change.

A polished interface can make uncertain work look finished. The strongest prototypes keep sources, assumptions and unanswered questions visible.

For heritage and cultural organisations, this opens up a practical approach: choose a small body of material, make a working experience, and invite people to test it. Use what happens to shape the research, partnerships and funding proposal.

At Visioning Lab, this is thinking through making.

AI lets us put an idea in people’s hands earlier. What they do with it can tell us which project is worth building.