Movie Creators Believe AI Will Transform the Craft, Potentially Beyond Recognition

New generative technologies are showcasing impressive previews, yet how will they truly reshape the film and television industry? While short-term transformations might be minimal, experts in filmmaking and VFX believe the future could hold changes beyond our current comprehension.

At a SIGGRAPH event in Denver, discussions by Nikola Todorovic from Wonder Dynamics, Freddy Chavez Olmos of Boxel Studio, and Michael Black from Meshcapade alongside the Max Planck Institute, explored how generative AI and similar technologies might alter media production. They agreed that immediate benefits may be limited, but the pace of innovation hints at significant, unpredictable impacts in the future.

The conversation began with the acknowledgment of the current limitations of video generation tools.

Todorovic highlighted a common misbelief that AI can serve as a magical one-click solution for final VFX shots, a goal that remains elusive. Without the ability to tweak outcomes, these “black box” tools fall short despite ongoing improvements in user experience, which is gradually incorporating 3D and filmmaking vernacular.

Black underscored the inherent challenge in using language to capture the essence of visual creativity.

“While certain movements and poses have well-known names enabling definition, countless actions fall beyond the scope of our language. This linguistic gap stems from our internal, generative behavioral model, not applicable for visual images. Explaining a vision is challenging; instead, demonstration via motion engages and transmits understanding through observation,” Black explained.

This philosophical angle reveals the intrinsic limitations of text-prompt imagery systems, hampered by the inadequacy even of the extensive technical and artistic lexicon used in production and post-production processes.

Chavez Olmos shared his personal journey, illustrating how geographical constraints limited access to the filmmaking industry, a challenge AI technology can overcome by distributing expertise more globally. “Leaving Mexico was my only option then, but today, AI can open these doors without the need to relocate,” he reflected.

However, Black voiced concerns over the potential downside of this accessibility.

“Handing someone a high-performance vehicle doesn’t make them an elite racer. Similarly, while the widespread ability to produce films sounds promising, the quality of such films will likely vary greatly. The true value lies in the potential emergence of unique creators who would otherwise remain unheard,” he opined.

“We’re on the brink of witnessing a novel form of entertainment, perhaps a hybrid of video gaming, film, and real life. Unlike the passive consumption model of current film media, this new genre could promote an active, participatory form of storytelling,” Black envisioned.

Before such advancements take hold, Chavez Olmos anticipates a gradual warming up to AI-generated content, much like the initial mixed reactions to early CGI films. “We may see our first fully AI-produced film by the year’s end, and I believe audiences will eventually embrace this new medium,” he predicted.

“The crucial lesson from Pixar is the paramount importance of story, connection, and heart. A film that resonates on these levels will captivate audiences, regardless of whether the characters are AI-generated. The allure of witnessing human talent and emotion at its finest will always hold a special place,” Black concluded.