China's four major artificial intelligence (AI) players attended the AGI-Next Frontier Summit, hosted by Tsinghua University's Beijing Key Laboratory of Foundation Models on Saturday.
Attendees included Yao Shunyu, chief AI scientist of the chief executive/president's office at Tencent (0700), Lin Junyang, head of technology for Qwen under Alibaba (9988), Tang Jie, founder of Zhipu AI (2513), and Yang Zhilin, founder and CEO of Kimi.
China's artificial intelligence industry is shifting to intelligent agents capable of completing complex and multi-week tasks, from models focused mainly on chat functions and benchmark scores, industry leaders said on Saturday.
Yao said in the consumer market, improvements in user experience do not necessarily translate into higher user retention. In contrast, in the enterprise market, companies fear not slowness but being "wrong and uncontrollable".
Lin said AI needs to be combined with embodied intelligence in the future, directing robots to conduct real-world experiments and accelerating efficiency in areas such as pharmaceuticals and scientific research.
On whether China can overtake the US, the speakers put the probability in the next paradigm at no more than 20 percent, citing differences in computing power investment, with the US focused on high-risk research and China on delivery and productization.
Cynthia ZHONG
