AI video generation at CNCC 2026 weighs realism against emotional appeal
AIPanelists at a CNCC 2026 forum on October 23 said AI video generation is shifting from visual quality toward narrative coherence and emotional expression. Speakers said single foundation models reliably produce only 15 to 20 seconds of video before character drift and narrative breaks appear.
Source details
Leiphone · leiphone.comPublished · added here
Original title: 要更真实，还是更动人？AI视频生成走到了分岔口 | CNCC 2026