Many software development organizations introduced AI-powered tools like GitHub Copilot in 2023 to boost their delivery process. The aim was to deliver more and faster by speeding up the coding process.
I was skeptical about the claim that faster coding can make the entire software development process more efficient. Coding is just one aspect of the process, which also includes analyzing business requirements, designing proper architecture, collaborating with other team members, and many other steps. Even if coding is done 50% faster (as per GitHub analysis), it doesn’t mean the entire process will become 50% more efficient.
AI-powered tools can effectively and efficiently add new code to a project. However, they are not yet proficient in conducting thorough refactoring or reducing tech debt. As a reminder, one of the principles of the Agile Manifesto emphasizes the importance of simplicity. Maximizing the amount of work not done is essential, and this principle is also valid for the amount of code written.
A new report on the impact of AI-powered tools on code quality has been published recently. It’s an interesting read and worth considering. Also, the latest State of DevOps 2023 report states that “Some analysts and technologists hypothesize that AI will make software teams more performant without negatively affecting professional well-being. So far our survey evidence doesn’t support this.”.
To be fair, I’m a fan of many AI-powered tools, including GitHub Copilot. However, in my opinion, it’s too early to draw any concrete conclusions about their longer-term impact on developers’ productivity, code quality, and so on.
Please check “Coding on Copilot: 2023 Data Suggests Downward Pressure on Code Quality” report for more insights.