CASE STUDIES
What Structured Memory Makes Visible
The Žiburio Archive is built on the CultureNet framework—a system for organizing cultural materials so they remain connected to the contexts that give them meaning. Rather than treating books, photographs, or programs as isolated items, CultureNet structures them as a network: people, places, institutions, events, and the relationships between them.
This makes it possible to move through the archive not just by item, but by connection—tracing how a festival links cities, how a name reappears across decades, how a single publication sits within a broader historical moment.
General-purpose AI can generate language about culture. The Žiburio Archive preserves what culture actually produced—along with the provenance and structure that allow it to be understood. These case studies show the difference.
The Same Question, Two Systems
Ask a general-purpose AI and the Žiburio Archive the same question about DP camp poetry. One produces language about the books. The other surfaces the books themselves — with provenance, connections, and the structure underneath.
Explore →
What AI Gets Wrong About Lithuanian Culture
We asked five leading AI models to describe a specific book from our archive. Every model produced confident, fluent, and wrong answers. The archive holds the actual object — and the distance between generation and truth is measurable.
Coming Soon
The Publishing Network AI Can’t See
One publisher — Lietuviškos Knygos Klubas — connects to 11 books and 9 authors across three decades in our archive. Tracing those edges reveals a publishing infrastructure that no language model knows exists.
Coming SoonWhat is CultureNet?
CultureNet is a framework for treating cultural materials as structured memory infrastructure — preserving provenance, lineage, and relationships so that cultural knowledge can be accessed without losing the structure that makes it meaningful. The Žiburio Archive is the first implementation.
Read the framework →