Scope
The focus is the relationship between a person, their agent and the organisations that agent encounters. Who sets the objective? Which actions require confirmation? What evidence accompanies a result? The essays examine user-controlled interfaces, service semantics and the organisational decisions beneath them. Personal agents replacing branded interfaces is treated as a scenario to investigate, not an accomplished industry-wide transition.
Illustrative example
A traveller asks an agent to rearrange a disrupted journey. The agent can compare alternatives within the traveller’s stated constraints, but a fare difference or non-refundable booking requires a fresh decision. A useful interface shows the proposed change, its source, the permission being used and which parts of the itinerary remain unconfirmed.
Terms and contribution
Computational agents are an established subject in AI; Poole and Mackworth’s introduction provides that foundation. “Agentic Interfaces” is this journal’s organising lens. The layered account of foundational, vertical and branded agents is developed as an argument in the essays below, rather than presented as a universal architecture.
Read in this order
- Inverting the interface — Begin with the speculative shift in who controls the interface and whose interests an agent represents.
- From Interface Craft to Governable Delegation — Then examine the permissions, evidence and recovery arrangements that delegation needs.
- Designing for agency — Finally, extend the question from an individual interaction to organisational roles, supervision and accountability.