State of Code Developer Survey report: The current reality of AI coding
Blog post from Sonar
The "State of Code Developer Survey" reveals that AI-assisted coding has transitioned from an experimental phase to a daily habit for developers, fundamentally altering software development workflows. The survey, conducted with over 1,100 professional developers, indicates that 72% of those who have tried AI coding tools now use them daily, with 42% of committed code being AI-generated or assisted. AI is extensively utilized across various project types, from prototypes to mission-critical services, yet its effectiveness varies significantly by task. While AI excels at generating boilerplate materials and assisting with documentation, its performance is less effective for tasks like refactoring or maintaining complex systems, highlighting a gap between widespread usage and perceived effectiveness. This discrepancy partly stems from confidence issues, as developers remain cautious about relying on AI-generated code for high-stakes tasks, resulting in a new bottleneck at the verification stage.