Computing power, distribution, and instructionable generative video AI. These are the key challenges facing the Chinese film industry when it comes to AI, according to speakers at the Shanghai International Film Festival’s SIFFORUM panel “Smart Tech, Immersive Worlds, and the Next Cinema Revolution.”
According to Yan Yijun, vice president of AI basic model building company MiniMax, computing is the most important element, calling it the “absolute core.”
Variety details
“For generated video models to achieve higher fidelity, what they really need is more computing power to iterate on refinement and experimentation,” Yang says. “The more you experiment, the more effectively you can train certain aspects. Large-scale computing infrastructure is important for each iteration of the model.”
For other panelists, the challenge lies in finding a distribution model that works in a seemingly limitless world of production. When full-length epic fantasy films are shot and completed in days, how will the industry deal with the flood of such content?
“This is a real contradiction. From a commercial point of view, it is a contradiction of redefining the value of a product,” said Bauhinia Film’s Li Tingwei. “For a commercial studio, one aspect is creation. [But] Then there is distribution, but how do you sell it? It also brings challenges. ”
The fickle nature of generative AI also continues to pose problems for filmmakers.
“One point is that what AI creators actually need is not just a more powerful AI, but a more obedient AI. Directors have a very complete cinematic vision in mind, but the AI often presents something unexpected,” said Nina Zheng, deputy general manager of ASUS China. “Filmmakers want very specific emotions, very specific changes in lighting, very subtle and very personal in terms of aesthetic decisions. But when these instructions are carried out by AI, they often can’t get there in one step and require frequent adjustments.”
However, when it comes to job losses due to AI, the panelists seemed to be optimistic. Significant changes in film production workflows are creating new jobs and positions across the industry, including AI Director’s Assistants specializing in AI-prompted engineering.
“This tells us that the more sophisticated the tools, the more important the people operating the tools become,” Zheng said.
This feeling was echoed by Huang Jianxin, the only film director on the panel and dean of the film department at Xiamen University. He argued that unlike other industries, film production consists of many special cases, making it difficult to automate every aspect.
“China’s two largest tea factories replaced 2,000 employees with 120 people and AI. As a result, tea quality, safety, consistency, and packaging accuracy have all improved significantly. When standard processes are the norm, AI becomes the rule,” Fan said.
“Creating works of art” [on the other hand]is individualistic and necessarily excludes certainty. Therefore, it is most natural for artists and AI to collide. ”
According to Huang, the conflict between artists and AI is ultimately not insurmountable.
“AI is closely related to cinema, because cinema was born from technology. Theater, poetry, and dance are the origins of art. Cinema is a technological medium, but it is a synthesis of all these arts.”
“99% of people couldn’t go to the movies because they needed money. Everyone was disenfranchised from participating in this art. That’s why so many young people love AI now. AI is the equalizer.”
best of variety
Sign up for Variety’s newsletter. For the latest news, follow us on Facebook, Twitter, and Instagram.
