Context is a design requirement, not a personalization layer

Before choosing the model

A system’s usefulness depends on more than the quality of an isolated response. A research design might need to begin with the conditions under which a person can use, question, and correct it.

The contextual AI project starts from that possibility. It remains an early direction rather than an account of a completed intervention.

What to learn first

Which languages are used for the task? What happens when connectivity fails? Who bears the cost of correcting an error? Which local practices does the proposed interface make difficult to express?

These are questions to investigate with people in a specific setting. They should not become a generic checklist that substitutes for local inquiry.

A connected question

The local knowledge note asks who owns the interpretation of context. PICHNET approaches that question through contextual advice and epistemic agency.