My impression of Github copilot

I have just got an invitation to try out Github copilot. I didn't expect much from it, but I thought why not try it out and see what it can provide.

After installing the extension, I started writing a test function, and boom! it auto-completed the function body! I mean this is just mind-blowing, how much code this AI is trained on to be that precise?

(the published_at should have been set to null but that may be because it tried to mimic the previous test setup)

This exactly matches what they say about it on their official site:

Tests without the toil. Tests are the backbone of any robust software engineering project. Import a unit test package, and let GitHub Copilot suggest tests that match your implementation code.

I also tried writing a function name and seeing if it can generate something close to what I wanted. I only typed function is_video, pressed TAB and this was the result:

function is_video($path) : bool
{
    return str_contains($path, 'mp4') || str_contains($path, 'webm');
}

Sure you may want to tweak this function a bit, but it did an excellent job generating such code!

This is my first time using Github copilot and I can't wait to keep playing with it.

Some people have concerns about such tools replacing developers, but I don't think this is why these tools exist, they are there to help you save time and write better code.

I think the best way to describe Github copilot is what is written on the official site:

Trained on billions of lines of public code, GitHub Copilot puts the knowledge you need at your fingertips, saving you time and helping you stay focused.