Associate Professor
Dr. Tim Grünewald
University of Hong Kong
Realism and AI-Generated Narrative VR: The Redemption of Virtual Reality?
Futurists, science fiction authors, and technologists have long dreamed of photorealistic, open-ended, digital virtual worlds that are indistinguishable from physical reality. Generative AI is a necessary tool required for realizing this vision. Several AI generated solutions have recently been released (e.g. NVIDIA Omniverse and GET3D; Tri3D for Meta Quest; Meshi AI), and AI generated virtual worlds are already emerging. This paper considers how AI-generated, photorealistic VR will impact notions of realism by revisiting classical realist film theory. The arrival of the new narrative medium film sparked a debate among early film theorists about whether film is best understood as a realistic window upon reality (e.g. Kracauer, Bazin, Lukács) or a framed canvas on which the filmmaker creates a subjective interpretation of the world (e.g. Arnheim, Balázs, Münsterberg). I argue that early narrative VR presents a similar contrast between realist and formalist approaches to VR storytelling. While I reject the prescriptive tone common in early film theory debates, I claim that AI generated virtual models of physical reality are paradoxically required to fully realize realistic 6DoF experiences. While VR texts are always interactive, 6DoF experiences significantly increase the extent of co-creation of a narrative VR experience between the creator and the immersant. 6DoF photorealistic representations of reality more fully enable some of the affordances that realist film theorists celebrated in early cinema. For example, such VR narratives enable the immersant to experience the significance of fortuitous occurrences, indeterminate events, or mundane objects (Kracauer), they allow for a reinterpretation of the long take and deep focus advocated by Bazin, and present novel means of representing social realities (Grierson, Lukász). After introducing the current state of AI-generated VR and the foundational claims of realist film theory, I will analyze early examples of photorealist narrative 6DoF experiences as well as the latest, hyperrealist interactive VR experiences created for the Apple Vision Pro such as Encounter Dinosaurs.
BIO
Tim Grünewald is the founding director of eXtended Humanities, a lab for XR and humanities research, and Associate Professor at the University of Hong Kong, where he directs the Global Creative Industries Programme. He is the author of Curating America’s Painful Past: Memory, Museums, and the National Imagination (2021, Kansas University Press), the editor of Rethinking America’s Past: Voices from the Kinsey African American Art and History Collection (2019, University of Cincinnati Press), and the co-editor of Imperial Benevolence: U.S. Foreign Policy and American Popular Culture (2018, University of California Press). His work on collective memory, public history and narrative VR is published in journals such as Convergence, Communication, Culture, and Critique, Cultural Studies, Journal of American Culture, and elsewhere. His research projects on the National Mall and on VR film were funded by the HK Research Grants Council and won several research awards, including from the National Council on Public History (Washington, DC) and the Popular Culture Association (Atlanta, GA).
<< previous | next >>
Dr. Tim Grünewald
University of Hong Kong
Realism and AI-Generated Narrative VR: The Redemption of Virtual Reality?
Futurists, science fiction authors, and technologists have long dreamed of photorealistic, open-ended, digital virtual worlds that are indistinguishable from physical reality. Generative AI is a necessary tool required for realizing this vision. Several AI generated solutions have recently been released (e.g. NVIDIA Omniverse and GET3D; Tri3D for Meta Quest; Meshi AI), and AI generated virtual worlds are already emerging. This paper considers how AI-generated, photorealistic VR will impact notions of realism by revisiting classical realist film theory. The arrival of the new narrative medium film sparked a debate among early film theorists about whether film is best understood as a realistic window upon reality (e.g. Kracauer, Bazin, Lukács) or a framed canvas on which the filmmaker creates a subjective interpretation of the world (e.g. Arnheim, Balázs, Münsterberg). I argue that early narrative VR presents a similar contrast between realist and formalist approaches to VR storytelling. While I reject the prescriptive tone common in early film theory debates, I claim that AI generated virtual models of physical reality are paradoxically required to fully realize realistic 6DoF experiences. While VR texts are always interactive, 6DoF experiences significantly increase the extent of co-creation of a narrative VR experience between the creator and the immersant. 6DoF photorealistic representations of reality more fully enable some of the affordances that realist film theorists celebrated in early cinema. For example, such VR narratives enable the immersant to experience the significance of fortuitous occurrences, indeterminate events, or mundane objects (Kracauer), they allow for a reinterpretation of the long take and deep focus advocated by Bazin, and present novel means of representing social realities (Grierson, Lukász). After introducing the current state of AI-generated VR and the foundational claims of realist film theory, I will analyze early examples of photorealist narrative 6DoF experiences as well as the latest, hyperrealist interactive VR experiences created for the Apple Vision Pro such as Encounter Dinosaurs.
BIO
Tim Grünewald is the founding director of eXtended Humanities, a lab for XR and humanities research, and Associate Professor at the University of Hong Kong, where he directs the Global Creative Industries Programme. He is the author of Curating America’s Painful Past: Memory, Museums, and the National Imagination (2021, Kansas University Press), the editor of Rethinking America’s Past: Voices from the Kinsey African American Art and History Collection (2019, University of Cincinnati Press), and the co-editor of Imperial Benevolence: U.S. Foreign Policy and American Popular Culture (2018, University of California Press). His work on collective memory, public history and narrative VR is published in journals such as Convergence, Communication, Culture, and Critique, Cultural Studies, Journal of American Culture, and elsewhere. His research projects on the National Mall and on VR film were funded by the HK Research Grants Council and won several research awards, including from the National Council on Public History (Washington, DC) and the Popular Culture Association (Atlanta, GA).
<< previous | next >>