AI coding assistants like Gemini 3 Pro and Claude Sonnet enable developers to create personalized software forks tailored to individual needs, bypassing traditional open-source contribution workflows. The author demonstrates this by building a custom Hyprland screen-share picker in an hour using AI agents, solving a long-standing personal pain point. This trend toward "software for one" raises concerns about security, code quality, and fragmentation of open-source projects, while simultaneously empowering developers to solve problems without waiting for maintainer bandwidth. Projects can mitigate risks by providing clear testing frameworks and linting tools that guide AI-generated code toward maintainable standards.
Sort: