Stephen Downes

This article starts from the perspective of a personal AI that learns about its user over time. We'll call that the 'personal context' for lack of a better term, along with the author, Duncan Davidson. How do we create and manage a personal context? Davidson lists some things that matter: provenance, for example, to distinguish between what the user wrote and what they read. Scopes are important, as they distinguish between work, home and family. Identities and types matter, to allow for a consistent ontology across sessions. Time matters: my views may change over time, and what I said recently on a topic should take precedence (all else being equal). The system should be available across devices and across locales, and interoperability should not require uniformity: "Different people and companies should be able to build different context systems for different ecosystems and trust boundaries. What they need is agreement on identity, permissions, provenance, and exchange, not one universal application."