~/blog/why-im-shifting-from-visual-studio-to-vscode-for-ai-development

$ cat why-im-shifting-from-visual-studio-to-vscode-for-ai-development.log

Published

#VS Code #GitHub Copilot #AI #Productivity #Developer Tools

A Shift in My Daily Workflow

Something interesting has been happening in my development routine lately. Despite having full access to Visual Studio Enterprise—the IDE I've used for over a decade—I find myself increasingly reaching for VS Code instead. Not because Visual Studio is bad (it's still excellent for many things), but because the GitHub Copilot experience in VS Code has become so compelling that it's changing how I think about development.

This isn't a planned migration or a corporate mandate. It's organic. When I want to quickly prototype something, debug an issue, or write new code with AI assistance, VS Code is where I naturally end up. And I'm not alone—I'm seeing this pattern across my team at EUCARIS as well.

Why Copilot Feels Different in VS Code

GitHub Copilot is available in both Visual Studio and VS Code, so theoretically the experience should be similar. In practice, it's not. The Copilot integration in VS Code feels faster, more responsive, and more deeply integrated into the editor's DNA.

The inline suggestions appear almost instantly. The Copilot Chat panel is snappy and contextually aware. Switching between code and conversation feels natural, like you're collaborating with a colleague who's sitting right next to you.

VS Code showing GitHub Copilot inline code suggestions

Copilot inline suggestions appear as you type, offering context-aware completions.

Part of this is architectural. VS Code is built on Electron with a modern extension API that makes it easier to integrate AI features deeply. But part of it is also focus—GitHub's engineering effort on Copilot is clearly prioritizing the VS Code experience.

Background Agents and New Capabilities

The latest Copilot features have pushed this advantage even further. Background agents can now work on tasks while you continue coding. You can ask Copilot to refactor a module, generate tests, or explain a complex piece of code—and it handles these requests with impressive accuracy.

VS Code with Copilot Chat panel open showing conversation

The Copilot Chat panel provides conversational AI assistance right within VS Code.

What really changed my workflow is the ability to iterate on code through conversation. Instead of manually typing out changes, I can describe what I want in natural language and let Copilot propose the modifications. It's not perfect—you still need to review and understand what it's doing—but it dramatically speeds up the exploration phase of development.

The agent mode is particularly powerful. Rather than just autocompleting the current line, Copilot can now understand your broader intent and work across multiple files. It can create new files, update imports, and maintain consistency across a codebase—tasks that used to require tedious manual work.

Practical Tips for Making the Switch

If you're considering moving from Visual Studio to VS Code, especially for better Copilot integration, here are some practical tips from my experience:

  • Essential Extensions: Start with the GitHub Copilot and GitHub Copilot Chat extensions. For .NET development, add the C# Dev Kit.
  • Learn the Keyboard Shortcuts: Ctrl+I opens inline chat, Ctrl+Shift+I opens the Copilot Chat panel. Master these and you'll fly.
  • Use the @ References: In Copilot Chat, you can reference files with @workspace, @terminal, or specific file paths. This gives Copilot the context it needs.
  • Configure Settings Sync: Enable Settings Sync so your configuration, extensions, and keybindings follow you across machines.
  • Don't Abandon Visual Studio Completely: Some tasks (complex debugging, profiling, designer surfaces) are still better in Visual Studio. Use both tools for what they're best at.

The VS Code Extensions Marketplace has everything you need to customize your setup. Take time to explore extensions for your specific tech stack.

Team Adoption at EUCARIS

What's fascinating is watching this shift happen organically across my team. Nobody mandated a switch to VS Code. There's no official policy. Yet more and more developers are naturally gravitating toward it for day-to-day coding tasks.

The pattern is consistent: someone tries VS Code with Copilot, experiences the improved AI integration, and gradually spends more time there. They still fire up Visual Studio for specific tasks—debugging complex threading issues, using the memory profiler, or working with Windows-specific projects—but VS Code becomes their default.

We've started sharing tips and configurations. Someone discovers a useful extension or keyboard shortcut, and it spreads through the team. There's an energy around it that feels like the early days of adopting a new technology—excitement mixed with practical problem-solving.

The Future is AI-Assisted

I'm convinced that AI-assisted development is not a fad. It's a fundamental shift in how we write code. Just as IDEs with IntelliSense transformed development compared to plain text editors, AI code assistants are transforming what's possible in terms of speed and productivity.

VS Code, with its lightweight architecture and deep Copilot integration, is currently the best environment I've found for this new way of working. That might change—Microsoft is actively improving Copilot in Visual Studio too—but right now, the gap is noticeable.

If you haven't tried VS Code with Copilot recently, I'd encourage you to give it a week. Not for everything, just for your next new feature or side project. You might be surprised how quickly it becomes your go-to environment.

The tools we use shape how we think about problems. And these AI-powered tools are opening up new ways of exploring code, learning unfamiliar codebases, and turning ideas into working software. That's worth exploring, regardless of which IDE has been your home for years.

← Back to blog overview