After Tiktok Chinese businesses such as Kling Ramp Up AI for video

The Kuaishou’s Kling AI platform generates video from text and stationary images.

Nurphoto | Nurphoto | Gets the image

Beijing Chinese Fun in China has given data to companies-and now they are stepping up tools for artificial intelligence for advertising and movies.

Bytanse Tiktak is ranked first and third in the research firm Artificial analysis“Higher AI Text-Video Generative Models, which have been launched over the last two months. Google ranked second and fourth, while a short video based on Beijing Coffer Cling ai ranks fifth.

Despite some consolidation in other parts of the AI industry, “Competition in the AI video generation models are at an earlier stage, and some Chinese companies have appeared in the early leaders in this space,” Vay Zion said, an analyst at UBS Securities.

“We believe that AI video generation has a potential for reshuffling the content industry,” she said, “raising the efficiency of production, reducing barriers to create and unlock new monetization models.”

Using such tools AI users can upload one image or several, and direct AI to make video based on them. Other tools allow users to enter the text from which AI will generate the video.

More than 20,000 enterprises: from advertisers to animators are already using Kling AI to create video and a company based in Beijing, said AI in Shanghai this week. The latest version of Kling 2.1 can automatically add appropriate sound effects to match the AI video.

This is not just for users in China.

“Let it be the scale of users or commercial income, abroad for the majority,” said Zeng Yushen, Head of the Kling AI operations, Mandarin, translated by CNBC, said. She said the company plans to increase the support of the tool in places such as Japan, South Korea and Europe.

“This is what we watched, big AI models are becoming more globalized,” she said. “People don’t seem to care what the country’s goods are.”

CEO Figma Dylan Field on IPO Debut: Design Today publicly goes to public

Kuysha claimed to be a clearing ah Made more than 150 million yuan (20.83 million dollars) in revenue During the first three months of the year, and daily advertising spent on AI generative instruments during this time amounted to 30 million yuan. The company has not yet reported when releasing the results of the second quarter. Zen refused to share the cost of Kling AI training.

While the reduced cost of production provides a “significant” market, UBS ‘Xiong said: “Modern models remain limited by the clip length, movement sequence and management.”

Chinese companies AI companies also face competition from the US, except for the restrictions of Trump administration in China’s access to the advanced semiconductors needed to train AI models.

Amazon and Google has Running tools to create video from images or text. Issues come as Microsoft-When Openai launched its video generation model Sora to chatgpt subscribers in December – Almost a year after he revealed his capabilities in February 2024.

However, Kling AI has already launched for the public in June 2024. Users sign up and buy loans to create video.

Vidu, a rival tool from Startup Shengshu based on Beijing, launched by a worldwide user approximately 12 months ago, and approximately in March this year stated that an annual $ 20 million annual income is expected.

“Chinese companies tend to first try to determine the commercial” pain point “… areas where companies will pay for services, which has become a problem for AI applications,” said Paul Triol, a partner and senior vice president of China at the Dga-Albridge Group Advisory firm.

He pointed to how the Chinese launch 3DStyle uses the generative AI to develop new clothing styles and integrate them with internet to automated production.

US companies also apply for II in certain industries, Triolo said, but Chinese businesses can often integrate II as they face very competitive conditions and can recruit with a “very skilled” software engineer.

“AA AA MOVIEMAKER”

Chinese e -commerce giant Alibaba This week also remained on top of the trend, releasing the latest version of AI Hiving Generation AI called Wan2.2. The company claimed that with open source, users can control lighting, time of day, colored tone, camera corner, frame size, composition and focus length.

Open Source allows users to download the model for free and customize, if not commercialize, products with it. Alibaba claimed that since the open search of the Wan series “Wan” in February, the model has been downloaded more than 5.4 million times from the Hugging Face platform and similar in China called Modelscope.

“The II era ended in the film. We entered the II era as a director,” said Winston Ma, Professor of the New York New York Law School. He noted that in China, 1.4 billion population has given local companies “huge” data viewing video.

“Just like Tiktok took world markets storms with short video in the mobile Internet era, Chinese companies AI could well lead a generative revolution in visual digital entertainment,” said MA, the author of “The Digital War: How technological power of China forms the future of AI, Blockchain and Cyber Prapl.”

Avatars and games

Chinese companies are also building AI tools for more than just video generations.

Last week, Bake announced that its latest digital human technology is which AI- Running for sale in $ 7.65 million During the interactive session more than six hours in June – will be released for wider use of the industry in October.

In 3D -Visualization, Favorable Released its Hunyuan World model to create digital panoramic images generated from text and visual clues. Visual materials use the “Network” file format, which gamers can use gamers to edit certain parts of the image.

“In addition to support (Tencent’s) developers’ internal teams, the platform demonstrates Tence’s ambitions to standardize high accuracy asset generation and expand its impact on game development in China,” said Daniel Ahmad, director of research and ideas in Niko Partners.

Weekly analysis and understanding of Asia’s largest economy in your mailbox
Sign up now

Niko has found that more than half of the game development studios are already using II to generate the contents and reduction of time and costs.

But game development reflects broader problems in using AI on a scale to create video and graphics.

“While interest in II is high,” Ahmad said, “we have already seen a reaction to games that have poorly implemented this technology.

Source link