Photo: Courtesy of Lü Yang

In a vivid display of technology and creativity, visitors witness a simple image of a cat being transformed into a lively and dynamic short video. The scene unfolds as a staffer uploads the image onto a screen, which then magically brings the adorable feline to life. The atmosphere is filled with excitement and wonder.

This is an advanced image-to-video technology a Chinese company showcased at the ongoing World Artificial Intelligence Conference (WAIC) 2024 in Shanghai. Many Chinese tech enterprises are displaying their generative AI large models, based on which they have developed various generative AI tools that allow users to create images, videos or music works within seconds.

Text-to-image, text-to-video, image-to-music... the mushrooming AI-generated content (AIGC) industry has become extremely popular, particularly after OpenAI's cutting-edge text-to-video model Sora went viral earlier this year.

How are Chinese companies developing in the AIGC field? How, and in what ways, will generative AI benefit the lives of Chinese people? The Global Times talked to several leading Chinese firms and influential organizations in the sector, who shared their experiences, observations and thoughts on this trending topic.

Photo: Chen Xia/GT

In the first echelon

On Friday, visitors gathered at the exhibition area of Chinese tech firm Tencent, to see new products including Tencent Yuanbao, an AI assistant application designed to help with various tasks and queries, as well as Tencent Yuanqi, an open platform for AI agents.

They interacted with Tencent Yuanbao and used it to generate their own virtual images of various styles, and later tried to select, use or create their own AI agents using Tencent Yuanqi.

Both Tencent Yuanbao and Tencent Yuanqi were developed based on Tencent's proprietary foundation model Tencent Hunyuan. Tencent Hunyuan supports a variety of multi-mode generation capabilities, such as generating videos according to texts, photos and video clips, the Tencent Hunyuan team told the Global Times on Friday.

Apart from some basic core functions, Tencent Yuanbao offers personalized services including creative drawing and oral practicing, the team said.

Photo: Courtesy of Tencent

Chinese tech company Kuaishou, renowned for its short video platform, brought its text-to-video generative AI model Kling to WAIC 2024. Kling is the world's first text-to-video model with Sora-level abilities that has been made available to the public, the Kling team told the Global Times at the conference.

According to the R&D team, the videos generated by Kling have good effects including large and reasonable movements, and high resolution and frame rates. Moreover, it can generate new videos based on ones that have already been generated, extending the generated videos by up to about three minutes, the team added.

Kling has drawn a lot of attention both at home and abroad since it started a free trial in June. "There are a few things worth applauding here. None of these videos deviates from the prompt much, and the physics seem right," commented MIT Technology Review in an article on June 19, which also listed some "obvious shortcomings" of Kling's trial version

"But unlike Sora, which still remains inaccessible to the public four months after OpenAI trialed it, Kling soon started letting people try the model themselves," it added.

In terms of industry progress and public popularity, China's AIGC development is in the first echelon globally, said Ni Kaomeng, initiator of the China AIGC Industrial Alliance, a nationwide civil group of China's AIGC industry insiders.

"In other words, it is the US and China who are at the top in this field," Ni told the Global Times during WAIC 2024. "The US does better in cutting-edge and underlying technology. But as for industrial application and service design, we Chinese have certain advantages."

Photo: Courtesy of Kuaishou Kling

Easy to use

The rapid development and popularization of AIGC in China has aroused a lot of curiosity and expectations, but also uncertainty among the Chinese public.

Many people reached by the Global Times said they want to try the AIGC tools, using them to either assist with work or leisure, such as music creation. But they are afraid of the technical complexity.

On Friday afternoon, a salon for general visitors was held at the WAIC venue, at which many Chinese AI content creators and industry insiders talked about how AI is affecting ordinary people and bringing opportunities to all walks of life.

Li Yichen, the salon's co-organizer, said they held the salon specifically for the general public who don't know much about AIGC. "Through the event, we hope our audience can feel that AI is not unreachable; AI is closely related to our lives," Li told the Global Times.

"AIGC is a big trend of future development. We look forward to more people knowing about the industry, getting interested, and probably joining in someday," he said.

Li is one of the most influential AIGC creators in China. In February, Li, along with dozens of peer creators, made the first-ever AI Chinese New Year gala online. The 150-minute gala, consisting of more than 50 AI-generated video or music programs, attracted millions of views.

The threshold of AIGC is actually very low for the public, as anyone can create an image, video or piece of music in just one minute with the help of AI tools, Li said. "Nonetheless, generating high-quality content requires an accumulation of creators' knowledge and aesthetics, as well as their ability in terms of continuous learning," he added.

Li was an early member of the China AIGC Industrial Alliance. The alliance is mainly engaged in popularization of AIGC knowledge and skills for both start-ups related to the field and the general public, said Ni.

Ni said that the alliance regularly holds non-profit open classes and workshops, inviting experts in generative AI research and application to share front-line experiences in both AI content creation and entrepreneurship. "We have also made Generative AI tool packages for various groups, such as students, public servants, and designers, who can download the tool packages for free," he told the Global Times.

Speaking of the promising application future of AIGC, Li said it can do a lot in professional areas including finance and healthcare. "For ordinary people, generative AI can be a new form of entertainment, as users can create their own images, videos and music works for fun with friendly AI tools," he added.