Intelligent Narrative Technologies
The 16th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment
By Stephen Ware and Rushit Sanghrajka
There were three workshops held at AIIDE-20, held virtually October 19-23, 2020, including Experimental AI in Games, Intelligent Narrative Technologies, and Artificial Intelligence for Strategy Games. For more information the AIIDE conference, please see aiide.org.
Intelligent Narrative Technologies
INT returned for its 12th meeting in 2020 with two excellent keynote talks and a wide variety of topics on applying AI to games and other interactive stories.
The 12th workshop on Intelligent Narrative Technologies was held this year as part of the AAAI international conference on Artificial Intelligence and Interactive Digital Entertainment. INT brings together a multidisciplinary team of researchers interested in artificial intelligence, narrative theory, game development, psychology, social justice, and many other topics.
This year’s workshop featured two keynotes. The first, by Steven Poulakos, demonstrated several research prototypes developed by Disney Research | Studios. They included tools for assisting animators and for delivering interactive story content to users via screens and augmented reality. The second was given jointly by Kristin Siu and Eric Butler and described their experience integrating AI into Elsinore, a recently released time-looping interactive story based on Shakespeare’s Hamlet. They discussed the challenges of scheduling important story events and collaboratively authoring interactive story content.
Nine papers were presented, covering a wide variety of topics. Most were technical. Three papers focused on the familiar difficulties of reasoning about agent beliefs and theory of mind via classical regression planning, dynamic epistemic logic, and Monte Carlo Tree Search. One described a human-in-the-loop process for building patterns used to find interesting stories in large simulations. There were also two welcome data-driven approaches, one about collecting corpora of narrative video game text and one on measuring how and when player choices meaningfully affect the outcome of a game’s story. Two papers were more philosophical. One discussed the importance of reflecting on the cultural values communicated by interactive storytelling work via Critical Technical Practice, and one discussed a unified definition of the concept of a “quest,” which has been given many definitions.
The workshop also featured a tutorial and an accompanying paper on Camelot, a tool for quickly building 3D interactive narrative game worlds. After the tutorial, Alireza Shirvani, lead developer of Camelot, hosted Zoom Plays Camelot which mirrored Twitch Plays Pokémon by inviting participants to take control of one or several characters in a virtual world to tell a collaborative story. The goal of the event was to encourage researchers to imagine how they could plug their own techniques for interactive storytelling into a Camelot experience.
INT first met as AAAI symposia in 2007 and 2009, and since then has met continuously, usually as a workshop at various conferences, including Artificial Intelligence and Interactive Digital Entertainment, Foundations of Digital Games, and the Electronic Literature Organization. In 2016, INT was a special track at the International Conference on Interactive Digital Storytelling. After having a joint workshop with WICED in 2018, and its absence in 2019, we were excited to bring it back. Due to the COVID-19 pandemic, it was held via the Zoom tele-conferencing platform with low registration fees, making it the most accessible iteration of the workshop. At over 50 attendees, we believe it was also the largest. Moreover, organizing this workshop over Zoom allowed for talks to be recorded within the platform, with some participants consenting to have them published on YouTube. We hope that this practice allows the community to access work presented at INT in more ways.
This INT workshop was co-chaired by Prof. Brent Harrison of the University of Kentucky, Ph.D. candidate Rushit Sanghrajka of the University of Utah, and Prof. Stephen G. Ware of the University of Kentucky.
Stephen Ware is an Assistant Professor in the Department of Computer Science at the University of Kentucky.
Rushit Sanghrajka is a Ph.D. Candidate in the School of Computing at the University of Utah.