Top Reasons NOT to use GitHub Copilot

[colindembovsky]

Generative AI, Copilot, blah blah blah - who needs it? You're the Ultimate Programmer, so why would you want some pretentious "large" language model helping you? You're a lone wolf that don't need nobody (or no thing) to "help" you - it will only get in the way of your staggering intellect.

Well this post is just for you - top reasons why do DON'T need GitHub Copilot.

1. You prefer writing code in Notepad

GitHub Copilot only supports four IDEs: Visual Studio Code, Visual Studio, IntelliJ and NeoVim. But you prefer to code in Notepad. Or vim. Or emacs. All those plugins and breakpoints and live debugging - it's overrated. You can debug in your head just by looking at the perfect code you wrote. And you can quit vim whenever you want to.

So what if Copilot integrates seamlessly and fades into the background as you code?

2. You like writing boilerplate code

Constructors. Getters. Setters. Who needs to think about business problems when you can write real code. I mean, you learned how to do it when you did your Intro to Programming course, so you want to make sure you get your money's worth.

So what if Copilot is really good at writing repetitive, boilerplate code, thereby keeping you focused on solving business problems?

3. You know every regex expression.

Only newbies need to test their regex expressions using regex101. You don't need Copilot's help to validate obscure string formats - you just do it in your head no worries about human error.

So what if Copilot can generate regex and easily dump out obscure formats and formulas so that you don't have to remember or search for them?

4. You know every API

Who needs to look up how to invoke common APIs? Once you've seen a Swagger doc you can call any and every method in that API forever.

So what if Copilot knows how to call APIs that millions of developers use daily?

5. You like copying and pasting from StackOverflow.

Speaking of searching for stuff - you love StackOverflow! What's better than googling a question and then inevitably landing on StackOverflow where there are a bunch of random answers that may or may not be correct that you can copy from? And who doesn't love renaming all the variables and fixing all the formatting errors (tabs vs spaces anyone)? Not that you need to search for stuff anyway - your infallible memory is a giant library of endless code examples to draw from.

So what if Copilot can get answers for you without you having to leave the IDE... er, file... and follows your naming conventions and styles?

6. You don't need unit tests.

Unit tests - that assumes your code could be wrong. And why spend time programming code that tries to break the code you just coded so perfectly? If you did write unit tests, they would be the ultimate tests.

So what if Copilot can quickly generate tests, mocks and find multiple test cases just by analyzing the code you're testing?

7. Comments? What for?

You don't have to document your code. Your code is so perfect that people can tell what it's doing just by seeing your code. Besides, no-one else will ever need to look at your code unless it's to learn how to program perfectly. Da Vinci didn't have to "comment" the Mona Lisa, did he?

So what if GitHub Copilot can generate code based on your comments, and that the comments stay to help document your code?

8. You'd rather leak IP by pasting code into ChatGPT.

You're not like those Samsung developers that leaked sensitive information while copying code into ChatGPT, right? I mean, you'd never be asking an AI for help anyway.

So what if Copilot encrypts data and uses an HTTPS and (at least for Copilot for Business) never keeps any of the data or solutions?

Conclusion

This GitHub Copilot thing is totally overrated. It's not going to change the way you work or the perfection of the code that you crank out as you consume coffee and cold pizza. No way.

Happy (not) Copiloting!

[rakotomandimby]

Having code written quicker reduces workstation uptime or allow more resting time for eyes

[spenserblack]

I'd be curious if there could be a study that compares carbon footprint of not using Copilot vs using it, taking into account time saved when developing with an AI assistant. E.g. The carbon footprint of developers of equal skill level creating a simple Twitter clone with and without Copilot.

[khoadaniel]

Excellent post, excellent click-bait and excellent twist.

[computationalknowledge]

I thought I was good at sarcasm.... KUDOS to you Sir~!

[johnwildes]

Great post @colindembovsky! When can we expect Terraform and or Bicep support...that would help me out a lot!

[Sbk34]

10 reasons why I should.

1. I'm not seeing to many other volunteers.

2. To learn this I am going to need someone or something to teach it .

3. It's going to suggest a lot of stuff I might have thought of.

4. The suggestions are going to be extension of what my brain may not have known was available.

5. I don't have many real friends. Lol

6. Ai is fun to use as a learning tool.

7. This learning tool is I believe progressive in the sense it's always growing the intelligence world.

8. When I wake up late night or early morning and need to get the ole brains wheels working and looking for a explanation why it is is .

9. The suggestions begin with the user .

10. Im pretty stoked to have found it .

[JudsonOrum]

Thank you Sir I decided to use GitHub Copilot

[spenserblack]

Thanks for finding these studies! In my experience Copilot definitely needs to be guided a bit. The first suggestion isn't always what you want, unless you've written a decent amount of code to heavily hint to it the style of solution you're going for.

Another study couldn't find that it statistically significantly saved time. That study found that developers had to spend time debugging and understanding the code generated too.

This is probably a good argument for manually writing tests instead of letting Copilot do it. "Copilot tested itself and concluded that it wrote good code." 😆
You don't want to have tests that you don't understand or need to debug. And, once you have confidence in your tests, it should be easier to find where Copilot generated unwanted behavior.

In my experience, I've appreciated Copilot the most with highly repetitive behavior. For example:

type Token int

const (
    FOO Token = iota
    BAR
    // a ton of other tokens
)

var tokenNames map[Token]string{
    FOO: "Foo",
    // Start accepting Copilot suggestions here
    BAR: "Bar",
    // a ton of other tokens
}

This could have possibly been done with an editor macro, but in this example Copilot does save time IMO. The simplicity leaves little room for Copilot to make a mistake.

So I guess I'm saying that Copilot works best when you understand its limitations and let it speed up simple tasks, saving the complexity for humans.