Finding new paths
If Xu's story shows how students are beginning to find their place in AI, Du Xingjian's path reveals how that shift can unfold after several years in the tech industry.
Du, who studied media in graduate school, moved into tech in 2020, drawn by the rapid growth of the internet industry. She began pursuing internships and roles in the field, eventually landing her first job as a data product manager at Chinese services platform Meituan. She has since pivoted again and is now building her own AI startup.
Before ChatGPT-style generative AI models took off, Du said, the tech industry was largely dominated by major companies, leaving limited room for individuals and small teams. AI, she believes, has created more opportunities for smaller organizations and solo entrepreneurs.
Today, Du shares practical, hands-on experience with AI on social media under the handle "Hahadu". She describes her approach as "pragmatic AI": using AI to solve real workplace problems, rather than chasing flashy effects or treating the technology as a status symbol.
For instance, she uses AI to create content for e-commerce operations, repurpose livestreams into short clips and analyze product performance data.
Du recalled a designer friend who had spent tens of thousands of yuan learning AI but still struggled to use the tools in his own work.
"I walked him through his workflow and pointed out that only the design stage was worth improving with AI," she said. "If he could master that step and scale it up, he would actually make a profit."
"People love learning new AI knowledge but struggle to connect it to their own work," she said."They are eager to use AI to streamline their work, yet often overlook the costs and benefits. Not every step will need AI optimization to bring significant improvement."
Du emphasized the importance of connecting AI knowledge to one's own work instead of blindly following trends.
"AI information is vast and fast-changing, and chasing the latest can be overwhelming," she said."Start with your daily work and needs and figure out where AI fits into your own tasks."