Generative Narrative Amidst Large Language Model (LLM) AI – A Transnational Experiment
Submission Type
Panel
Start Date/Time (EDT)
21-7-2024 9:15 AM
End Date/Time (EDT)
21-7-2024 10:15 AM
Location
Algorithms & Imaginaries
Abstract
In this roundtable presentation, faculty presenters share their methods for co-delivering a transnational, interdisciplinary workshop that invites graduate and advanced undergraduate students from their respective American and French universities to work together to create generative multimedia narratives using the Omeka S database, ComfyUI, Javascript, Unity, and recent Large Language Model (LLM) AI, among other tools. Over the past five years, this Transnational Digital Creation Workshop has worked online and on-site in France to actualize creative, interdisciplinary making as an epistemological driver of new knowledges and of more reflexive and cross-cultural understandings of digital practices and lifeworlds, premises central to the field of electronic literature. Situating itself at the interface between digital technologies of communication, the humanities, and emerging creative practices, the workshop enables each interdisciplinary team of student researchers to collaboratively co-create a generative narrative using a variety of media and to tell a story that resonates with them, and with the global context in which these practices unfold today. Meeting via videoconference and online communication over the course of a semester and then travelling to work together on site in France at the end of the semester, student researchers gain knowledge of creative computing beyond their chosen area of study, learn to work on an interdisciplinary team, broaden their awareness of how digital communication, arts, and electronic literature vary across cultures and geographical sites, in spite of their apparent interconnectedness, and build connections through their shared generative narrative and the workshop process.
Reflecting on evolving models of, and methods for creating generative narratives since the workshop was first offered in 2019, we will illustrate how LLM AI might be experimentally re-oriented to the ends of computational creativity and generative narrative, in particular. Our two transnational student teams will present the generative narrative projects they co-created this spring using these methods and will also share their perspectives on the workshop process, the potential and constraints they encountered using LLM AI in this context, and the importance of a transnational perspective on generative narrative and generative AI today.
Recommended Citation
Shackelford, Laura; Szoniecky, Samuel; Giacovelli, Matthew; Grey, Deen; Coles, Erica; Shah, Harsh; and Belkessa, Ferroudja, "Generative Narrative Amidst Large Language Model (LLM) AI – A Transnational Experiment" (2024). ELO (un)linked 2024. 28.
https://stars.library.ucf.edu/elo2024/algorithmsandimaginaries/schedule/28
Generative Narrative Amidst Large Language Model (LLM) AI – A Transnational Experiment
Algorithms & Imaginaries
In this roundtable presentation, faculty presenters share their methods for co-delivering a transnational, interdisciplinary workshop that invites graduate and advanced undergraduate students from their respective American and French universities to work together to create generative multimedia narratives using the Omeka S database, ComfyUI, Javascript, Unity, and recent Large Language Model (LLM) AI, among other tools. Over the past five years, this Transnational Digital Creation Workshop has worked online and on-site in France to actualize creative, interdisciplinary making as an epistemological driver of new knowledges and of more reflexive and cross-cultural understandings of digital practices and lifeworlds, premises central to the field of electronic literature. Situating itself at the interface between digital technologies of communication, the humanities, and emerging creative practices, the workshop enables each interdisciplinary team of student researchers to collaboratively co-create a generative narrative using a variety of media and to tell a story that resonates with them, and with the global context in which these practices unfold today. Meeting via videoconference and online communication over the course of a semester and then travelling to work together on site in France at the end of the semester, student researchers gain knowledge of creative computing beyond their chosen area of study, learn to work on an interdisciplinary team, broaden their awareness of how digital communication, arts, and electronic literature vary across cultures and geographical sites, in spite of their apparent interconnectedness, and build connections through their shared generative narrative and the workshop process.
Reflecting on evolving models of, and methods for creating generative narratives since the workshop was first offered in 2019, we will illustrate how LLM AI might be experimentally re-oriented to the ends of computational creativity and generative narrative, in particular. Our two transnational student teams will present the generative narrative projects they co-created this spring using these methods and will also share their perspectives on the workshop process, the potential and constraints they encountered using LLM AI in this context, and the importance of a transnational perspective on generative narrative and generative AI today.
Bio
Dr. Laura Shackelford is Professor in the English Department and the founding director of the Center for Engaged Storycraft at the Rochester Institute of Technology. She received a B.S. in English from the University of Minnesota, Twin Cities; an M.A. and a PhD in English with a specialization in Literature and Science from Indiana University, Bloomington; and taught as a post-doctoral researcher in the English Department at Penn State, State College before joining RIT’s College of Liberal Arts. Her research examines contemporary fiction, narrative, and emerging digital literary practices with a particular interest in how literary fiction creatively draws from, and reflects on the computational, bioinformatic, and networked knowledges and digital media practices post-World War II information, systems, and bioinformatic sciences introduce. Speculative fiction that experiments with emerging digital media and spatial forms like the network, she finds, reminds us to attend to these socio-cultural and technological transformations with an eye to the possibilities these changing knowledges, scientific practices, and social systems might introduce to the benefit of women, other minoritized groups, as well as to the entangled, nonhuman material life forms with which we share our ecosystems. Her research and writing unfolds in conversation with women’s and gender studies and, especially, feminist science studies’ explorations into the material technologies, social contexts, and systems of power that shape knowledges and might re-shape knowledges—and the broader social systems they help to co-realize—in important, necessary ways, if thoughtfully diversified and approached more equitably. She is the author of Tactics of the Human: Experimental Technics in American Fiction (2004) and co-editor of Surreal Entanglements: Essays on Jeff VanderMeer’s Fiction (2021), and the author of numerous book chapters and journal articles on digital literary and narrative practices that contribute to understanding the potential and limitations of digital cultures, at present. At RIT, she teaches contemporary fiction, speculative fiction, narrative theory, storytelling across media (print fiction, graphic narratives, interactive narrative, digital games, animation, film, photography, digital literary arts), and women’s and gender studies courses.
Samuel Szoniecky - After studies in art history (DEA) and professional experience as a computer engineer, project manager and business company founder, I turned to research and teaching. Since 2006, I have been exploring new knowledge engineering methods for the development of collective intelligence, focused on stimulating, expressing and sharing individual intelligence. In parallel with theoretical research into the epistemological limits of symbolic languages and their possible overcoming through the use of analogical languages, I approach the question of socio-semantic cartographies with reference to philosophical, anthropological and information and communication science work. The aim is to lay the foundations for an interdisciplinary knowledge ecosystem dedicated to the development of digital humanities through the generic modeling of informational existences.
These research perspectives are put into practice through a number of national and international partnerships. In collaboration with Pierre Lévy, holder of the Chair of Collective Intelligence at the University of Ottawa, we have developed the first semantic web application using the IEML language: Evalactisem, a concept addressing tool for evaluating an individual's semantic activity. As part of an international research program for the Evalactisem, a concept addressing tool for evaluating an individual's semantic activity. As part of an international research program for the archiving of digital poetry, we have worked with Philippe Bootz on a web application to model "points of view" on a work of art, thus preserving its "power to act". In the same field, we are working with Jean Pierre Balpe on a new version of his automatic text generator and its development in artistic (M. Chevalier, G. Chatonsky, E. Vernhes, M. Jaffrenou), research (GAPAII) or educational (CréaTIC - proverbs) projects. In the context of city 2.0, we have developed an application with Nasreddine Bouhai for the global evaluation of urban territories. In the field of digital humanities, we are participating in the ANR Biolographes project, where we are designing a tool for modeling networks of influence between scientists and literary figures in the 19th century. Similarly, in the ANR Aliento project, we are working on digital devices for interpreting short sapiential forms.