Empowering youth in AI with new start-up bases

Shanghai is strengthening its push in artificial intelligence by supporting young professionals through new initiatives at the ongoing World Artificial Intelligence Conference.

A new start-up base for overseas returnees for AI and a national youth challenge connecting university students with industry problems has been launched in Xuhui District over the weekend.

The Innovation and Entrepreneurship Base for Returned Talented Personnel was unveiled at the Shanghai Foundation Model Innovation Center, an incubator that hosts more than 100 AI start-ups. It was launched during the NextGen Returnee Voices Tech Exchange Forum of WAIC on Sunday.

The base is designed to support young Chinese professionals returning from overseas by offering workspace, funding access, and resources to turn research into applications.

Speakers at the event emphasized the need for international collaboration in AI and ethical development, especially as the field rapidly evolves.

At the Pioneering Conversation on AI with Chinese Youth of WAIC on Saturday, young innovators took part in the national AI "Challenge Match."

The competition pairs top university teams with technical challenges submitted by companies and research institutes. The goal is to create practical AI solutions with direct market or social applications.

More than 13,700 students from 506 universities participated. Projects covered robotics, cybersecurity, and AI-powered tools for public safety and urban planning.

Li Xuelong, CTO of China Telecom and the head of TeleAI, called the challenge match "a powerful way to match real problems with real talent," adding that it also helps companies discover new solutions and recruit skilled graduates.

Huawei Cloud has used the match to solve many real technical problems, drawing more than 6,000 entries from over 1,000 universities, the company said. One breakthrough involved deploying billion-parameter AI models on edge devices.

He Tianyao, a team leader and past winner, shared how his group used AI to improve kilometer-level positioning accuracy in PTZ (Pan, Tilt, Zoom) security cameras in a previous challenge match.

"That experience taught us how to translate research into real-world solutions," said He.