I was scared like anyone else that “AI will replace me.” But with my experience using AI tools daily, I’ve embraced the cliché:AI doesn’t replace me — it speeds me up. AI doesn’t replace me — it speeds me up. But only because I’m learning how to use it properly. Let’s start with Github Co-pilot. only As a frontend developer, my job involves building features, debugging, and constantly improving things across large codebases. Copilot has become a tool I rely on almost every day — not to do the work for me, but to help me start faster, move smarter, and catch things earlier. Today, I’m breaking down how I actually use GitHub Copilot inside PhpStorm — where it works really well, and where you still need to lead. really well Context is the Superpower What makes Copilot inside PhpStorm so useful is this: It understands the file, the code, and the project context. It understands the file, the code, and the project context. It understands the file, the code, and the project context. It understands the file, the code, and the project context. This means I can write a comment like: // Service to validate student assignment submissions // Service to validate student assignment submissions …and Copilot will generate a solid starting point — not just random boilerplate, but something relevant to the structure of my project. This kind of head start makes a huge difference when starting something from scratch. It saves me from staring at a blank file, remembering syntaxes and starting empty functions. Instead, I’m editing and evolving, which is much faster than writing from zero. But here’s the thing: You need a strong mental model of what you’re building before AI can be helpful. How I Use GitHub Copilot Daily Let me walk you through some real-life use cases that have made the most impact: Kickstarting new features I often create starter code for new features by prompting Copilot based on what the file or module is meant to do. It gives me a first draft, and from there, I build step-by-step toward the final version. Kickstarting new features I often create starter code for new features by prompting Copilot based on what the file or module is meant to do. It gives me a first draft, and from there, I build step-by-step toward the final version. It’s not about skipping thinking — it’s about saving time on repetitive scaffolding. Reusing patterns from past work A while ago, I experimented with dnd-kit to build a custom drag-and-drop interface. When I later needed to build sortable question lists for a live feature, I pointed Copilot to the earlier implementation. Reusing patterns from past work A while ago, I experimented with dnd-kit to build a custom drag-and-drop interface. When I later needed to build sortable question lists for a live feature, I pointed Copilot to the earlier implementation. It recognized the logic and followed the same pattern — helping me repurpose tested ideas in seconds. That saved me hours of research and fiddling. Code reviews — before anyone else sees my PR One of my favorite ways to use Copilot is to run a pre-review on my code. I have a simple custom prompt that I use for this. Code reviews — before anyone else sees my PR One of my favorite ways to use Copilot is to run a pre-review on my code. I have a simple custom prompt that I use for this. It flags inconsistencies, suggests better naming, logic issues — even potential edge cases. It also suggests how I can improve accessibility and how the code can be more maintainable. This gives me confidence before sharing my work with the team. I treat it as my private reviewer. Debugging: Smarter logging suggestions Copilot can be a game-changer to help debug code. One great trick is asking: Debugging: Smarter logging suggestions Copilot can be a game-changer to help debug code. One great trick is asking: “Where should I put console logs to help you debug this error?” It often points me to helpful log spots. Then I run the app, collect logs, and ask Copilot what they mean. Sometimes this saves a lot of time. Sometimes… not so much. AI can go in circles — and you need to know when to stop and rely on your instincts. The Truth About AI-Generated Code Let me be clear: You can’t just accept what Copilot gives you. AI can hallucinate. It can suggest things that look right — but aren’t. If you don’t understand the micro-details of your system, it’ll slow you down. Using Copilot well requires skill. You have to guide it, check its work, and sometimes work around its wrong turns. But when you do — when you know how to prompt it smartly, use context, and course-correct — it makes you faster, not just lazier. Final Thoughts I didn’t learn this overnight. It came from trying, failing, and figuring out what works over the past year. But today, GitHub Copilot is one of the few tools I’d be reluctant to work without. Not because it’s perfect — but because it makes me better. And that’s what good tools are supposed to do.
