Whenever AI is discussed among developers, there’s often an ‘all or nothing’ mindset. Some think AI will write all the software, while others avoid it because they believe it’s not producing good code. It always seems to revolve around the code. But that shouldn’t be the case.
Developers can still use AI in other ways to help them progress and save valuable time from their daily tasks. Here is a list of activities I do with AI daily that aren’t related to writing code:
Gathering context (~10-15 min of manual work)
As projects get larger, it becomes harder to remember the purpose of each file in the codebase. Sometimes, you need to locate where something is defined, understand how a feature works, and identify which files to check to make your changes.
Doing this manually can take a tremendous amount of time, but with AI, you can find the context you are looking for in a matter of seconds.
Organizing commits and writing commit messages (~5-10 min of manual work)
When working on a feature, you often modify multiple files and domains to reach your goal. However, when committing, it’s better to organize your changes into atomic, specific, and descriptive commits instead of a single commit that covers everything.
AI can help you break down changes into small, manageable steps and then commit each one to your branch. Depending on the size of the feature, this method can save you a lot of time.
Writing a change description (~5-15 min of manual work)
Building software is collaborative, and code reviews are an essential part of teamwork. Once you complete building a feature or change, you’ll probably need a description of the modification to help others review it more easily.
AI can recognize your modifications and generate a detailed description based on a template or a previous example.
Even if AI isn’t writing any code for you, it can still save 20-40 minutes of manual work per feature/change. And even if you ship one feature/change a day, that’s worth it.
Leave a Reply