Chinese fintech firm claims priority access to Sora API through Microsoft amid new text-to-video AI frenzy

Chinese fintech firm claims priority access to Sora API through Microsoft amid new text-to-video AI frenzy

Chinese fintech firm claims priority access to Sora API through Microsoft amid new text-to-video AI frenzy

A Chinese fintech firm is already attempting to take advantage of OpenAI’s Sora by claiming it will have priority access to the application programming interface (API) through Microsoft’s Azure system, in the latest example of how mainland businesses are rushing to associate themselves with cutting-edge generative artificial intelligence (AI) tools that are often not directly made available in the country.

Beijing-based Sinodata said it will be one of the first companies to apply for a Sora API subscription once the text-to-video tool becomes available on Azure, the cloud computing platform of Microsoft, which is OpenAI’s biggest backer. In a statement on its website on Tuesday, the company said it held a business seminar with Microsoft China on the potential application scenarios of Sora.

Chinese entrepreneurs express awe and fear of OpenAI’s Sora video tool

Sinodata will gain priority access by joining Microsoft’s AI Cloud Partner Programme, according to the statement, and become one of the US tech giant’s solution partners for data and AI in a deal the company signed last month.

Founded in 2003, Sinodata listed on the Shenzhen Stock Exchange in 2012. The company focuses on technology related to generative AI, Web3 and the digital yuan, according to its website. Its enterprise services include banking and data centre solutions.

Shares of Sinodata jumped 10 per cent on the news, reaching 14.40 yuan (US$2) before trading was suspended for reaching the daily limit.

Since OpenAI released demo videos generated by Sora last Friday, Chinese businesses and other parts of the tech community have shown strong interest in using the AI model. As with ChatGPT, though, some have expressed concerns about China’s ability to compete in this area.
OpenAI does not make its services directly available in China, along with a handful of countries that are sanctioned by the US. It also blocks Hong Kong internet protocol (IP) addresses, but Microsoft makes its Copilot products, which rely on OpenAI’s GPT models, available in the city.

05:03

How does China’s AI stack up against ChatGPT?

How does China’s AI stack up against ChatGPT?

Chinese internet giants have expedited research and development into creating their own text-to-video models.

In January this year, ByteDance released the MagicVideo-V2 text-to-video model. According to the project’s GitHub page, it combines a “text-to-image model, video motion generator, reference image embedding module and frame interpolation module into an end-to-end video generation pipeline”.
ModelScope, from the Damo Vision Intelligence Lab under Alibaba Group Holding, owner of the South China Morning Post, also launched a text-to-video tool. It currently only supports English input and video output is limited to two seconds.

Other public Chinese companies have sought to assure investors that they are watching developments in this area closely.

In response to questions from investors on Hudongyi, an interactive communications platform operated by the Shenzhen exchange, Shanghai-based game operator Giant Network Group said it expects the “breakthrough progress” of Sora to “improve the efficiency of content creation in the game industry”.

The Shenzhen Stock Exchange seen on October 23, 2019. Photo: Shutterstock

Intelligent interactive tablet manufacturer HiteVision also said on the platform that it expects education to be an important application of generative AI products. Education is an important market segment for the company, it added.

Zhejiang Enjoyor Electronics, a construction service provider for smart cities, said on the Shanghai Stock Exchange’s platform that text-to-video generation tools may be applied in multiple digital city use cases in the future, such as assessing and determining fault in traffic accidents.

Source link