Why AI won't fix your broken IA
Information Architecture (IA) is the discipline of structuring content to make it usable. It’s tempting to think AI can build IA or that AI search renders IA redundant. But building an IA requires real, not artificial intelligence. If AI was a person, it would be optimistic, cocky, incredibly fast, hallucinatory, sloppy and weirdly selective in its application of common sense and ethics. I show how the development of the semantic models that underlie IA requires human curation, human logic, and a human grasp of ethics. What’s more, when AI is used to good effect, it relies on an explicit, well-conceived IA.
Who is this talk for?
Those interested in the limits of AI or in IA, content modelling, or search.
What will attendees take away?
AI is not a cure-all for large unstructured sets of data and documents.
Good IA design requires accountability, human logic and human ethics, outside the scope of AI.
When AI is most useful, it's leveraging good IA.
Maria McMillan, now operating as The Information Architect Limited, has 20 years’ experience working as an IA in Aotearoa across the public, private and non-governmental sectors, and at councils, universities and libraries.
These days, she spends most of her time wrangling information chaos, document silos and spiralling data bloat into user-friendly websites, intranets and document management systems. Maria loves a good metadata schema and will die on the hill of a single source of truth.