AI-Driven User Interface Design for Solving a Rubik's Cube: A Scaffolding Design Perspective

Dezhi Wu, Hengtao Tang, Cassidy Bradley, Brittany Capps, Prathamjeet Singh, Katelyn Wyandt, Karen Wong, Matthew Irvin, Forest Agostinelli, Biplav Srivastava

Abstract: In this paper, we describe an AI-driven platform, ALLURE, with an embodied chatbot that teaches the user how to solve a Rubik’s Cube. Our AI algorithm consists of macro-actions, which refer to a set of moves that are not transparent or usable to users as a black box. Through the integration of explainable AI and conversational user interface designs, we created a novel AI-driven visual user interface inspired by scaffolding design paradigms to engage users. We conducted a set of initial usability testing with ten users, and usability findings imply some important AI-driven user interface designs to engage users for problem solving.