Netflix has integrated artificial intelligence into the production processes of roughly 300 titles in various capacities, according to Lee Sung-gyun, the streamer’s senior director of Asia-Pacific production. He outlined applications spanning previsualisation, production design, set dressing concepts, editorial work, and visual effects. Lee stated that AI serves as a tool to improve production methods rather than dictate creative direction, comparing its role to previous technological innovations that enhanced how stories are made without determining why they should be told.
Director Kang Yoon-sung reported that AI workflows could reduce feature film production budgets by 30 to 40 percent. He credited this savings not only to replacing traditional computer-generated imagery and visual effects work but also to simplified shooting processes and shortened production timelines. Kang noted that these cost reductions have revived projects that previously stalled during budget planning stages, potentially opening doors for genre and science fiction films within the Korean industry. He demonstrated this with a test project for a military-operations film created entirely using Chinese AI software earlier this year, rendered at cinema-quality specifications.
However, industry figures highlighted significant limitations of AI-generated content. Kang acknowledged that while AI produces technically high-quality imagery, it often diminishes emotional resonance with audiences by circumventing the collaborative discussions, creative effort, and accumulated experience traditionally central to filmmaking. Lee concurred, asserting that human intervention remains necessary to convey emotion effectively. Both speakers opposed the notion of AI replacing actors, characterizing performers as emotional creators whose mechanical reproduction through AI cannot achieve authentic human expression. Instead, they identified potential applications in stunt work and stand-in roles.
Kimball Thurston, chief technology officer at visual effects studio Wētā FX, characterised machine learning applications as “assistive AI” designed to support rather than alter creative intent. He noted his company’s decade-long history with machine learning and flagged technical challenges, including colour accuracy issues when AI models trained on video images fail to reproduce precise colour specifications. Netflix indicated it operates under strict internal guidelines and tests new tools before implementation, prioritising careful evaluation over rapid adoption. Thurston noted that Wētā is developing open-source tracking tools through the Academy Software Foundation to document AI use and provenance across projects.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI