An introduction to GitHub Copilot

[blackgirlbytes]

When GitHub initially introduced Copilot, I was suspicious. In addition to not understanding the technology, I was uncertain how it would affect my job and the industry. However, as I started to use Copilot, I gained an appreciation, understanding, and even eagerness about the future of AI pair programming tools ✈️. I’m passionate about Copilot because it helps increase my confidence and velocity as a programmer. I’ll share how you can use Copilot to enhance your developer experience too!

What is GitHub Copilot?

GitHub Copilot is an AI pair programmer that helps you code faster with less work. The user experience is similar to the Smart Compose feature within Google Docs and Gmail, which autocompletes sentences while the user is typing. Similarly, GitHub Copilot instantly draws context from comments and code and instantly suggests individual lines and whole functions instantly ⚡.

What is GitHub Copilot Labs?

GitHub Copilot Labs is a separate extension that comes complimentary with GitHub Copilot. Please keep in mind that Copilot Labs is an experimental extension developed by the GitHub Next team. Currently, Copilot Labs consists of a VS Code sidebar that houses distinct features starting with “explain this code” and “translate this code.”

Is GitHub Copilot magic 🪄?

Although Copilot feels like magic, it’s not. To suggest lines of code, Copilot needs a programmer's brain. Under the hood, GitHub Copilot is powered by OpenAI Codex. OpenAI is an AI research lab that developed Codex, a model that translates natural language into code. Codex is a version of GPT-3, another model developed by OpenAI. GPT-3 (Generative Pre-trained Transformer 3) uses deep learning to produce human-like text. Many of your favorite applications use GPT-3. For example, Duolingo specifically uses GPT-3 for French grammar correction. Codex also features natural language understanding but is fine-tuned for programming. By leveraging Codex, Copilot can help computers interpret people’s intent and empower developers to code with confidence.

Is GitHub Copilot perfect?

GitHub Copilot does not always spit out perfectly well-written, up-to-date code. Occasionally, some of its suggestions are downright hilarious. However, it’s more commonly accurate, and it continues to improve 🚀. Here are some tips to ensure technologists receive their desired suggestions:

• Tip Welcome to GitHub product feedback #1: Try writing clear, understandable comments to get the best results from Copilot.
  
• Tip Feature Request: Select a column when adding an issue to a project #2: Prompt GitHub Copilot to give you other suggestions by using a few keyboard shortcuts. You can learn more about viewing multiple suggestions here.

How does GitHub Copilot benefit developers?

Coding faster

Whether working as a software engineer or a developer advocate, I've found myself in moments where I'm writing code at the last minute 🕰️, or experiencing a coder's block. Reading suggestions from Copilot reminds me of how to solve common and not-so-common algorithms.

Writing tests

While I understand that writing tests for your code are essential, I don't enjoy it. I often take forever to start – trying to remember the difference between keywords like mock, nock, beforeEach, and afterAll. Then, when I get the pattern down, it feels like a tedious process for me to reach a particular test coverage percentage. This is where Copilot can come in handy. I can write a few comments describing what I want to test; then, Copilot may suggest the code needed for the test. I can also start writing tests, and Copilot may provide anticipated code for other needed tests. Learn more by reading Using GitHub Copilot to Automate Tests by Colby Fayock and Rewriting tests from Cypress to Playwright using GPT-3 by Gajus Kuizinas.

Writing better comments

Sometimes, writing great comments ends up on the back burner. There's a tendency to assume your code is clean, readable, and self-documenting until you return to your code the next day and you're unable to decipher the hieroglyphics. Since well-written comments tend to trigger better results from Copilot, I'm more motivated to write explicit, relevant comments.

For example, Copilot can understand Spanish comments!

Copilot can help you write Regex!

Pair programming with a coworker

I'm not fantastic at pair programming, especially when I have less experience than my colleague. I also experience anxiety about people watching my every move while I code. What stops me in my tracks is that I often try to remember the syntax. With Copilot, my pair programming buddy spends less time watching me Google the "Reduce JavaScript array method" for the umpteenth time. Instead, GitHub Copilot suggests the correct syntax for me.

Using GitHub Copilot is a game-changer, but don’t take my word for it. In a recent study, “60–75% of users reported they feel more fulfilled with their job, feel less frustrated when coding, and are able to focus on more satisfying work when using GitHub Copilot.”

How do I install it?

GitHub Copilot is available in Visual Studio Code, JetBrains, and my favorite, Neovim. Copilot is available for developers today and is free for qualified teachers, students, and maintainers of popular open source libraries. Sign up for GitHub Copilot 👩‍✈️today!

What’s next?

GitHub Copilot has inspired the evolution and development of other AI pair programming tools. It also enables developers and open source maintainers to think outside the box and create confidently. There is much more to share about GitHub Copilot and how it can improve our lives as developers, so stay tuned for my next post.

If you have any questions or want to share your love for GitHub Copilot, feel free to comment below!

[GuiBarradas]

Do you guys recommend the use of Github Copilot for begginers?
Could it interfere with the development of my logical thinking?

[blackgirlbytes]

I would say that GitHub Copilot helps you come up with ideas when you're stuck, even if you are beginner-level. I've used GitHub Copilot to write Python (a language that I don't often use because I'm a JavaScript developer). Through using GitHub Copilot, I've learned so much about Python's syntax. However, it does help if you understand the logic of why it's suggesting certain answers to you because the suggestions are not ALWAYS perfect. Accepting the suggestions blindly may break your code.

I still lean heavily on my logical thinking to get the best results from GitHub Copilot.

Here's what I normally do:

• Deactivate GitHub Copilot
• Think about what I want my application to do
• Activate GitHub Copilot
• Write a comment for the first step
• Look at Copilot suggestions and accept the best one
• Run the code to see if it works (if it doesn't; I may edit the suggestion..sometimes the syntax is slightly off or it's using an older version of an API)
• Once it's working, I'll press enter to see what else Copilot might suggest are next steps
• Turn it off again to take some time to think about what I want to do next

This is what works for me, but each person will find their own flow.

[janesroberta]

Vocês recomendam o uso do Github Copilot para iniciantes?

[blackgirlbytes]

I would say that GitHub Copilot helps you come up with ideas when you're stuck, even if you are beginner-level. I've used GitHub Copilot to write Python (a language that I don't often use because I'm a JavaScript developer). Through using GitHub Copilot, I've learned so much about Python's syntax. However, it does help if you understand the logic of why it's suggesting certain answers to you because the suggestions are not ALWAYS perfect. Accepting the suggestions blindly may break your code.

I still lean heavily on my logical thinking to get the best results from GitHub Copilot.

Here's what I normally do:

• Deactivate GitHub Copilot
• Think about what I want my application to do
• Activate GitHub Copilot
• Write a comment for the first step
• Look at Copilot suggestions and accept the best one
• Run the code to see if it works (if it doesn't; I may edit the suggestion..sometimes the syntax is slightly off or it's using an older version of an API)
• Once it's working, I'll press enter to see what else Copilot might suggest are next steps
• Turn it off again to take some time to think about what I want to do next

This is what works for me, but each person will find their own flow.

[habolanos]

Is there any performance indicator for copilot?

[blackgirlbytes]

Could you expand on this question more? Are you asking if Copilot shows how performant your code is? I'm not sure I understand.

[pleabargain]

Is there a user group for copilot for users to share their experiences? Are there copilot webinars?

[blackgirlbytes]

Folks often share their experiences on GitHub Copilot here: https://github.com/orgs/community/discussions/categories/copilot

And in terms of webinars, we don't have many yet! However, I have a talk on how to get the most out of GitHub Copilot here: https://youtu.be/inr1fFxvFAw

I would also suggest checking out the GitHub Youtube channel in general: https://www.youtube.com/@GitHub/featured

Hope this helps!

[danielapassos]

Can GitHub Copilot be used with other Git hosting services besides GitHub and how does GitHub Copilot handle code conflicts and merge conflicts?

[uyriq]

as far as i know (me a tru newbee) GitHub Copilot does not handle any repos problem nor merging conflicts, it is about code.

[blackgirlbytes]

That is correct. At the moment, GitHub Copilot doesn't handle merge conflicts. However, we do have a technical preview/beta of a feature within Copilot X that was released announced called Copilot for pull requests: https://githubnext.com/projects/copilot-for-pull-requests

[uyriq]

Great that you told me this news, thank you! By the way, I wanted to know if it will be possible to chat with Copilot just like with Bing because chat is better than comments.

[blackgirlbytes]

yes actually we also announced something like that. You can chat with Copilot for code-related issues. This is in a technical preview. Watch this video to see all that comes with Copilot X: https://youtu.be/4RfD5JiXt3A

[cognivore]

I want to buy CoPilot for my employees who are in my organisation, but I would like for them to be able to use it with their own private repositories.

Does anyone know if they will be possible?

[cognivore]

@blackgirlbytes I'm sorry to ping you, but you seem to be the most knowledgable about these things and the policies.

[blackgirlbytes]

Hi, thank you for pinging me again. I didn't notice that you commented. Yes, if you purchase Copilot for Business and do a team plan..your employees will be able to use it in their own private repositories as well!
Copilot is an extension into an IDE. Once installed and authenticated, it doesn’t care where the files are coming from...as long as the dev can open a file in their IDE - Copilot will start providing the suggestions.

[Vibhu315]

hey @blackgirlbytes ,
what's are the differences between github copilot and github copilot labs?