Is GitHub Copilot Ruining Code?


In this episode we cover "New GitHub Copilot Research Finds 'Downward Pressure on Code Quality'" by Visual Studio Magazine. This is a summary of a research paper put out by GitClear that looked at 153 million lines of code over 4 years to assess the impact of AI usage on code quality. The paper concluded that Copilot usage is steadily reducing code quality overall, and it looks like it's only going to get worse. We discuss the strengths and weaknesses of the paper, our own personal feelings on this, and Evan gets a little too existential.


References from the episode:

- Why I'm No Longer Using Copilot by Dreams of Code on YouTube

- The Primeagen's reaction video