Researcher(s)
- Axel Rodriguez-Leon, Computer Science, University of Delaware
Faculty Mentor(s)
- Matthew Mauriello, Computer & Information Sciences, University of Delaware
Abstract
Evaluating User Experiences of Co-Creative Artificial Intelligence Integration in Video Game Character Creation
*Axel Adonai Rodriguez-Leon, Owen He, Arnav Taduvayi, Matthew Louis Mauriello
Department of Computer & Information Sciences
University of Delaware
Recently, several creative industries, including the video game industry, have experienced a shift in their workflows. This shift involves the increasing use of Generative Artificial Intelligence (GAI) by creatives in these fields. This trend has sparked debate among creatives, with some questioning the effort and creativity behind works that incorporate GAI, especially when used to speed up the workflow. With this in mind, this research looks at how such systems have impacted the creative workflow through the implementation and use of Co-Creative AI. We conducted a third iteration of a between-subject experiment where participants (n=80) were asked to create a video game character using several different interfaces (i.e., buttons, text, voice, and combined) and then rate their experience using standardized scales (e.g., NASA-TLX). The character creation task varied between two difficulty levels: low and medium. We tracked participants’ actions, session durations, compared their use of traditional UI elements with AI prompt generation, and received brief feedback from their experience. It was found that when given a combined interface that allows for the co-creation of a character with an AI, users will opt to use the AI alongside the traditional customization tools. However, the completion time was found to be longer, and user feedback was mixed. This suggests that while some users are open to collaborating with AI for creative tasks, opinions about this approach vary and may not be optimal.