AI Levels the Student App Contest: More Women, Non-Coders, but Shallow Solutions
The contest is a microcosm of how AI is reshaping software education: it broadens who can build and sharpens output polish, but it also rewards feature bloat and cookie-cutter problem solving, leaving real user research and engineering depth as the true differentiators for anyone entering the field.
The 2026 China Collegiate Computing Contest's mobile app innovation track drew over 1,000 universities and tens of thousands of students. Serving as a judge, the weekly's editor observed that AI has dramatically raised the visual and functional polish of submissions, and reshaped the participant pool: at least a third to a half of contestants were women, and many teams came entirely from design programs with no prior coding experience, eroding the traditional dominance of computer science majors.
The same AI tools that lower the barrier to entry also led some teams to over-stuff features and skip genuine user research, producing near-identical "stereotypical needs" across projects, especially those targeting vulnerable groups. Lacking engineering fundamentals, several promising ideas stalled at demo stage. The new elementary division offered a counterpoint: children hand-writing Swift in Playgrounds, despite messy code, gained real learning through trial and error, in contrast to AI-wrapped polished outputs.
The author reflects that the student phase is a rare golden window for non-utilitarian creation, and concludes that an era where anyone can build software is arriving.
AI collapses the coding skill barrier but not the need for domain judgment; teams that skipped real interviews converged on homogeneous solutions, so the tool democratizes who builds without democratizing what is worth building.
The rise of female and design-major participants suggests the 'Coder' label is dissolving faster than the developer ecosystem's institutions and career pipelines have adapted.
The contrast between children's hand-written Swift and AI-polished entries exposes a mismatch: contests reward output polish, but the messy process is where durable learning happens.
Children's interest drives them; creativity-driven is the greatest source of power in the AI era. People like me, a wiring worker numbed by the assembly line, become the easiest tools for AI to replace.
The time AI helps save is mainly for people to relax and find new interests and goals. If we only blindly improve efficiency, we fall into an even more numbing cycle.
Currently capital is just blindly improving work efficiency: AI drives humans.