As AI coding assistants such as GitHub Copilot become more popular, their impact on productivity is being questioned. Despite initial appearances of enhanced efficiency, these tools may actually be contributing to a decline in overall code quality.
This observation comes from a recent report published by software engineering platform GitClear, which evaluated 211 million lines of code spanning from 2020 to 2024. The analysis by GitClear indicates a significant reduction in code reuse in the past year, raising concerns because code reuse is a widely adopted practice for developing redundant systems.
Various studies have recently been conducted on AI coding assistants, revealing mixed results. A survey by software vendor Harness found that most developers spend more time addressing bugs and security issues in AI-generated code compared to code written by humans. Conversely, a report from Google showed that while AI can expedite code reviews and enhance documentation, it does so at the expense of delivery stability.
