The research explores the influence of Copilot, an AI coding assistant, on software development practices. It reveals that users frequently struggle with understanding Copilot's suggestions due to their length and complexity, impacting readability. The study identifies 'INCOMPREHENSIBLE SUGGESTION' as a prevalent issue and emphasizes the importance of code explanation features to mitigate this problem. By changing how developers allocate their time, Copilot may enhance coding efficiency but also impose a new review pressure, making understanding and verifying suggestions critical for effective software development.
Some users mentioned that the code suggestions generated by Copilot are excessively long, resulting in reduced readability. This indicates that when Copilot provides relatively complex suggestions, or when users lack coding experience in a particular domain, understanding the code logic and verifying its correctness can be time-consuming.
In our research, INCOMPREHENSIBLE SUGGESTION ranks as the fourth most common Suggestion Content Issue, and the code explanation feature is also frequently requested. This reflects the challenges users face when interpreting AI-generated code.
Collection
[
|
...
]