OpenAI’s latest LLM opens doors for China’s AI startups

At the APSARA conference in Hangzhou, hosted by Alibaba Cloud, startups in China confirmed from AI their efforts to develop large linguistic models.

Corporate efforts are followed by announcing the latest LLMS from Openai, including pre -training pre -training model Opened by Microsoft. This model aims to address difficult tasks and pave the way for progress in science, coding and mathematics.

During the conference, Konal Cellen, the founder of Moonshot Ai, stressed the importance of the O1 model, adding that he has the ability to reshape various industries and create new opportunities for emerging companies from artificial intelligence.

Zhilin stated that reinforcement and expansion learning may be a pivotal development of artificial intelligence. Talk about the scaling law, which states that the largest models that contain more training data work better.

“This approach pushes the ceiling of artificial intelligence capabilities,” Chilin said, adding that Openai O1 has the ability to disrupt the sectors and generate new opportunities for startups.

Openai also emphasized the model’s ability to solve complex problems, which it says works in a similar manner to human thinking. By improving its strategies and learning from errors, the model improves its abilities to solve problems.

Zhilin said that companies that have enough computing energy will be able to innovate not only in algorithms, but also in foundational artificial intelligence models. He sees this pivotal, as AI engineers are increasingly relied on learning to reinforcement to create new data after exhausting available organic data.

Stepfun Jiang Daxin agreed with Zhilin but stated that the mathematical force is still a major challenge for many startups, especially because of the restrictions of American trade that hinder access to advanced semiconductors.

“The mathematical requirements are still great,” Daxin said.

One of the informed in Baichuan Ai said only a small group of Chinese artificial intelligence companies-including Moonshot Ai, Baichuan Ai, Zhipu Ai, and Minimax-in a position that allows them to make extensive investments in reinforcement learning. These companies – referred to as “Tigers Amnesty International” – are largely involved in the development of LLM, pushing the next generation of artificial intelligence.

More APSARA conference

Also at the conference, Alibaba Cloud released several ads, including the QWEN 2.5 Model family version, which is characterized by developments in coding and mathematics. Models range from 0.5 billion to 72 billion teachers and support for about 29 languages, including Chinese, English, French and Spanish.

Specialized models such as QWEN2.5-ROMER and QWEN2.5-Math have already gained some traction, with more than 40 million downloads on platforms embracing the face and models.

Alibaba Cloud has been added to its product portfolio, as it offers a model from text to video in the photo generator, tongyi wanxiang. This model can create videos in realistic patterns and animated drawings, with possible advertising and filmmaking uses.

Alibaba Cloud revealed QWEN 2-VL, the latest version of the vision language model. He deals with videos for a period longer than 20 minutes, supports the answer to video list, and is improved for mobile devices and robots.

For more information about the conference, click here.

(Photography: Guy_ai_Wise via x)

Do you want to learn more about artificial intelligence and large data from industry leaders? Check AI and Big Data Expo, which is held in Amsterdam, California, and London. The comprehensive event was identified with other leading events including the smart automation conference, Blockx, the digital transformation week, and the Cyber ​​Security & Cloud.

Explore the upcoming web events and seminars with which Techforge works here.