I've been using GitHub Copilot since it came out.
There are two areas I'd recommend developers look into:
1. USING AI AS A COPILOT
GitHub Copilot fills this role. I drive, and GitHub Copilot autocompletes, suggests, and verifies. I stay 100% in control while pair-programming with it.
If you are a Visual Studio Code user, this one is a no-brainer.
Here is what I love about GitHub Copilot:
It handles everything I need: spec-driven development, documentation generation, testing, and I can use it with any model I want (including open-source models).
There's nothing to rethink or configure. I just turn it on and start coding.
2. ACTING AS A COPILOT TO AI
The GitHub Copilot CLI fills this role. You can plan and verify, while the agent refines, breaks down tasks, and executes.
That's all you need.
Two different modes. Two different purposes.