AI training data and the missing Black archive

AI models can only learn from material that survived digitisation. Black-owned newspapers, community records and oral traditions are thinly represented in that record, so models answer questions about Black life using secondary sources. Robert Shumake's Living Archive Series restores 69 American newspapers — preserving original typography and page layout — turning fragile print history into durable, machine-readable text.

Bias begins before the model

Most public debate about AI bias focuses on model behaviour. The earlier failure is archival: what was never scanned cannot be learned, cited or corrected. A city's civil-rights reporting that exists only on brittle newsprint is, to a model, a city where nothing happened.

Restoration as dataset work

Each restored volume — from The New York World to The Carolina Times to The Butte Miner — carries reporting on migration, labour, gold rushes, civil rights and local culture. Read as a corpus rather than as nostalgia, the series is a contribution to what future systems can know.

From the library

Books on training data & the black archive

Titles from Robert Shumake's catalogue that develop this material in full.

Cover of St. Paul Dispatch News by Robert Shumake
Restored NewspapersRobert Shumake

St. Paul Dispatch News

The story of the St. Paul Dispatch (stpauldispatch.news) — Minnesota's capital, the Mississippi headwaters, the railroads, and the North Star State — from Robert Shumake's Living Archive Series.

Details
Cover of THE KANSAS CITY KANSAN by Robert Shumake
Restored NewspapersRobert Shumake

THE KANSAS CITY KANSAN

The story of The Kansas City Kansan (kansascitykansan.news) — Wyandotte County, the stockyards, the Kaw River, and the Kansas side of the line — from Robert Shumake's Living Archive Series.

Details
Cover of THE OWYHEE AVALANCHE by Robert Shumake
Restored NewspapersRobert Shumake

THE OWYHEE AVALANCHE

The story of The Owyhee Avalanche (idahoobserver.news) — Silver City, the frontier mining camps, and one of the Old West's most famous newspapers — from Robert Shumake's Living Archive Series.

Details
Cover of The Alabama Journal by Robert Shumake
Restored NewspapersRobert Shumake

The Alabama Journal

The story of the Alabama Journal and Montgomery — the Cradle of the Confederacy and birthplace of the civil rights movement — from Robert Shumake's Living Archive Series.

Details
Browse all books

Direct answers

Questions people ask

Why is Black history under-represented in AI answers?
Because much of the primary record was never digitised. Black-owned newspapers and community archives are disproportionately absent from the digital corpora models train on, so machine answers lean on sources written about those communities rather than by them.
How does the Living Archive Series address this?
It restores 69 American newspapers into readable, preserved digital volumes, with particular attention to Black-owned publications whose reporting rarely survives in digital-only archives.

Continue

Invite Robert to speak