AI Challenges Shape the Future of Chinese Filmmaking at Shanghai International Film Festival

During the Shanghai International Film Festival’s SIFFORUM, a panel titled “Smart Tech, Immersive Worlds, The Next Film Revolution” identified three obstacles to AI-driven filmmaking in China: insufficient compute for high-fidelity generative video, the need for new distribution models to handle a surge of AI-generated content, and difficulty making AI obey precise artistic direction.

Variety Magazine

The panel featured Yan Yijun, vice president of MiniMax, who called compute the “absolute core” of video fidelity; Li Tingwei of Bauhinia Films, who warned of a commercial clash over product value; Nina Zheng, deputy general manager of ASUS China, who urged a more obedient AI; and director-professor Huang Jianxin of Xiamen University, who contrasted AI’s success in standardized processes with cinema’s artistic unpredictability.

Variety Magazine

Huang cited the two largest tea factories, which cut staff from 2,000 to 120 after AI adoption, while reporting improvements in quality, safety, consistency and packaging accuracy.

Variety Magazine

Panelists agreed AI will reshape Chinese film workflows, spawning roles such as AI directing assistants that blend prompt engineering with creative oversight. Simultaneously, the ability to produce full-length fantasy epics in days threatens market saturation and forces a redefinition of cinematic value.

Variety Magazine

The panel stressed that expanding compute infrastructure is essential for higher-fidelity models, while distribution frameworks must evolve to accommodate output. They acknowledged AI’s unpredictability and emphasized iterative human oversight, noting that AI-focused roles highlight the growing importance of technical expertise in creative teams.

Variety Magazine

Critics warned generative systems often produce unexpected results, failing to capture subtle lighting cues or specific emotional tones in a single pass. This unpredictability fuels a “natural clash” between artists’ visions and AI’s standardized processes.

Variety Magazine

• “For a generative video model to achieve greater fidelity, what you really need is greater computing power to repeatedly refine and experiment,” — Yan Yijun, Vice President, MiniMax

Variety Magazine

• “This is a genuine conflict. From the commercial perspective, it is the contradiction of redefining product value,” — Li Tingwei, Bauhinia Films

Variety Magazine

• “One point is that the AI creators need is actually not just a more powerful AI, but a more obedient AI.” — Nina Zheng, Deputy General Manager, ASUS China

Variety Magazine

• “The two largest tea factories in China have replaced their 2,000 personnel with 120 people and AI, and the quality of the tea, safety, consistency, and packaging accuracy have all improved significantly.” — Huang Jianxin, Director, Xiamen University

Variety Magazine

The rise of AI directing assistants signals a stage of workforce transformation. Observers expect that solving compute bottlenecks and scalable distribution models will decide whether AI expands creative possibilities or drives market oversaturation in Chinese cinema.

Variety Magazine