The spread of AI coding agents is changing not only how code is written but also how junior developers are trained. The traditional “mentor and mentee” structure, in which senior developers passed on experience by assigning some tasks to juniors, is weakening. The skills demanded of developers are also expanding from coding to design, communication and product understanding.

Yoon Seok-chan (윤석찬), a principal tech evangelist at Amazon Web Services (AWS), said at a media briefing on Thursday at the AWS Korea office in Seoul’s Gangnam district that in the past developers trained juniors under them by giving them part of the work. He said now senior developers work with AI agents and junior developers also work with AI agents.

Yoon said traditional developer mentoring methods are facing difficulties. He said the change is affecting hiring and work inside companies.

As AI rapidly takes over code-generation tasks, the role of developers is also changing, he said. Yoon said many developers generate most code through AI agents. He said a main task for developers is becoming giving agents guidance on how to write code and what to do.

Similar changes are being detected in the workplace. Park Sang-woon (박상운), an AWS Serverless Hero who works as an AI architect at MegaZoneCloud, said developers used to be people who developed programs but have now changed into people who develop products. He said companies need people who can take a leading role across the entire product process.

He said the era has arrived in which being good only at development has, frankly, become much less of an advantage because of AI. He said he sees far greater demand for people who communicate well, share knowledge effectively and understand a company’s vision so they can participate across the overall development process.

He said the change is also appearing in hiring. Park said he was asked in a recent job-change interview what the company should do going forward to grow. He added that developers were not asked such questions in the past, but it is now an era in which developers also need to think that way.

University student developers also said the focus for developers after AI has shifted from coding to the ability to view an entire system. Lee Eun-ji (이은지), a leader of the AWS Student Builder Group at Seoul National University, said that before AI, developers were thought of as people who code well. She said that after AI, she thinks the scope has broadened to people who can draw the overall picture and judge for what purpose to build a system.

AWS also sees the role of external developer communities as becoming more important as the learning and growth environment for junior developers changes.

Yoon said junior developers are facing multiple challenges, including reduced hiring, rapid technological change during the AI transition, and fewer offline gatherings since the pandemic. He stressed the importance of developer communities.

AWS is expanding support so that students can directly practice AI and cloud from the university level and participate in developer communities. AWS on Thursday unveiled “Student Rewards” for university students. It said it would invest more than $500 million to strengthen cloud and AI capabilities among university students worldwide.

South Korean university students are also eligible for support. Each student can receive AWS credits worth up to $579, along with educational content and exam vouchers for certifications.