Home >> Industry

Seminar examines AI-Driven science documentaries

Updated: 2026-08-19 14:14 ( chinadaily.com.cn )
Share - WeChat
The post-broadcast seminar for the AI-produced popular science micro-documentary Accidentally Scientific is held in Beijing on Aug 6. [Photo provided to chinadaily.com.cn]

The post-broadcast seminar for the fully AI-produced popular science micro-documentary Accidentally Scientific was held in Beijing on Aug 6. Experts, scholars, industry representatives, and the crew members gathered to engage in in-depth discussions on topics including the standardization of full-AI production workflows, industry regulations for AI-generated audiovisual content, and pathways to market adoption.

The documentary premiered on streaming platform iQIYI on June 5. Within two months of its release, it garnered over 2 million views and topped iQIYI's documentary trending chart on its debut day. This 50-episode series, each episode lasting 5 minutes, delivered a remarkable market performance, setting a meaningful industry benchmark for "AI plus popular science" content creation.

The series explores "serendipitous" discoveries in the history of science, telling the accidental stories behind inventions such as the microwave oven and Coca-Cola, transforming complex scientific principles into lighthearted, entertaining, and youth-oriented narratives.

As a documentary project aimed at achieving a fully AI-driven production pipeline in short-form popular science content, the series has also drawn industry attention to the question of how artificial intelligence can be integrated into documentary production workflows.

"AI opens up greater imaginative possibilities for science documentaries, but the creative core always remains human," says Li Jinwei, the chief director of the series, sharing the team's creative process at the seminar.

"We didn't follow the traditional documentary production approach entirely; instead, we aimed to turn each piece into a little scientific story," Li explains. The scripts and creative concepts were still developed by the human team, while AI was primarily involved in generating visuals, motion effects, and dialogue. "Technology provides the means of execution, but the content creativity still comes from the creators."

Yang Juan, associate professor at the College of Chemistry and Molecular Engineering at Peking University, has long been engaged in science popularization. She notes that the strength of the work lies in its departure from the traditional lecture-style approach to popular science. Instead, it uses humorous expressions and internet vernacular to make science more accessible to younger audiences. However, she also points out that the more popular science content targets the general public, the more critical it is to ensure rigorous professional review.

"Science communication must both guarantee factual accuracy and make the content engaging and understandable for the public," Yang said.

During the seminar, Yang also cited the specific case of microwave oven principles to illustrate that certain scientific concepts are prone to oversimplification or even misinterpretation in mass communication, emphasizing the need for greater expert involvement in content verification.

Most Popular