

Episode 51: The Architecture of Knowing – with Jessica Talisman
Sep 21, 2025
Jessica Talisman, an information architect with over 25 years of experience, shares her insights on the critical role of structured semantics in the age of AI. She discusses the ontology pipeline, emphasizing the significance of taxonomies and hierarchies in creating reliable systems. Jessica also highlights the need for clear governance and communication to maintain taxonomy integrity. The conversation dives into balancing user-friendly language with backend stability, and contrasts big-data methods with symbolic knowledge approaches to enhance AI quality and reliability.
AI Snips
Chapters
Books
Transcript
Episode notes
Information Architecture Is Infrastructure
- Jessica Talisman frames information architecture as infrastructure and semantic engineering, not just front-end work.
- She emphasizes librarian practices to improve findability and interoperability across enterprises.
Begin With A Controlled Vocabulary
- Start with a controlled vocabulary to reconcile synonyms and clean noisy data before building hierarchies.
- Encode terms into machine-readable structures to connect front-end navigation with backend semantics.
Stable IDs Enable Flexible Labels
- Use stable backend identifiers to allow multiple display labels while preserving a single meaning.
- Parent-child relationships enable inference and help AI traverse and contextualize concepts.