Participation Interfaces for Human-Centered AI

Accommodating these stakeholder groups during system design, development, and deployment requires tools for the elicitation of disparate system interests and collaboration interfaces supporting negotiation balancing those interests. This paper introduces interactive visual "participation interfaces" for Markov Decision Processes (MDPs) and collaborative ranking problems as examples restoring a human-centered locus of control. Human-centered design has long been a software design philosophy centered on users [14, 9]. In extending this practice to human-centered artificial intelligence (HCAI), the term "human" sometimes refers to users, but increasingly often the term refers to humanity writ large. As a matter of expediency, practitioners of HCAI distil the collective design target from "humanity" to that of the "stakeholders" affected by the intelligent system in question.