Redefining usability in the age of generative AI: Towards a new evaluation paradigm

Abstract

Generative AI has revolutionized how humans do a range of things, from content creation to decision support. These AI systems are no longer static tools but dynamic collaborators capable of evolving in real time. This shift necessitates a fundamental rethinking of usability models since they normally depend on systems behaving in a steady and predictable way. This white paper provides a new approach to measure how beneficial generative AI systems are by looking at how well they adapt, trust, co-create, and react to their environment. We look at innovative ways to assist designers, developers, and researchers keep up with the times, such as AI-driven user testing, experience-based metrics, and models for ongoing review.