Former Kuaishou AI Tech Researcher to Develop Large Model

Former Kuaishou AI Tech Researcher to Develop Large Model

Li Yan, former head of multimedia understanding at Chinese quick video platform Kuaishou, has gone on to set up Yuanshi Technology, an AI firm which is especially engaged within the analysis and improvement of multimodal massive fashions, within the second half of 2022, 36Kr reported.

Li, who graduated from the Institute of Computing Technology of Chinese Academy of Sciences, was the core determine of AI know-how analysis and improvement at Kuaishou. In November 2015, with the assist of Su Hua, then CEO of Kuaishou, Li arrange the primary inside deep studying group, with the objective of constructing an algorithm mannequin to establish unlawful video content material.

With the buildup of AI know-how, Kuaishou requires the power to display screen video content material. In 2016, Li modified the title of the group from deep studying to multimedia understanding. Besides fixing safety compliance issues, the group was additionally engaged within the analysis and improvement of algorithm fashions in varied varieties, comparable to voice, textual content and picture.

Several folks conversant in the matter mentioned that Li provided to go away in Kuaishou in 2021 earlier than establishing Yuanshi Technology within the second half of 2022. This time, Li discovered essentially the most appropriate technical path for content material understanding – the multimodal massive mannequin, which refers to a man-made intelligence algorithm primarily based on textual content, picture, video, audio and different modal knowledge for studying and coaching.

As early as 2018, Li publicly emphasised the significance of multimodal know-how. “Video is a comprehensive form of information, incorporating vision, hearing and text, with user behavior another type of modal data, so video itself is a multimodal issue. Therefore, multimodal research is a very important subject for Kuaishou.”

SEE ALSO: China Needs to Accelerate Developing General Cognitive Models

At current, pushed by the current launch of ChatGPT, AI competitors amongst Chinese enterprises has turn into much more fierce. Baidu, ByteDance and different Internet giants are at present competing with one another whereas some startups are benefiting from their very own knowledge benefits to edge out competitors. The newest motion is the launch of Dialogue Writing Cat, the Chinese model of ChatGPT, which is developed by Meta Sota, an AI service supplier within the pure language processing discipline. At current, customers can use the chatbot on an online web page and in small packages after being invited to register. The platform demonstrates a greater means within the Chinese language and coaching knowledge and is predicted to make a breakthrough within the authorized discipline.

Sign up as we speak for five free articles month-to-month!



…. to be continued
Read the Original Article
Copyright for syndicated content material belongs to the linked Source : Pandaily – https://pandaily.com/former-kuaishou-ai-tech-researcher-to-develop-large-model/

Exit mobile version