GitHub Copilot was announced with much fanfare in June, 2021. However, the technical preview courted controversy over copyright infringement and open-source code laundering. Here, we look at how the perception has changed around GitHub’s AI pair programmer six months into its existence.
What is GitHub Copilot?
Sign up for your weekly dose of what's up in emerging technology.
The OpenAI Codex model powers GitHub Copilot. Codex is trained on natural language and billions of lines of source code from the public domain, including GitHub repositories.
How does Copilot work?
GitHub Copilot generates multiple suggestions on the context of the file being edited based on the comments in the file, and the code you’ve written before.
Here’s how it works: You write the docstring or the comment, suggesting a code snippet. In some cases, just the function name or a part of the function is enough to generate the rest of the code. The code is uniquely generated for you, and you own it.
Copilot in action
Copilot makes ten suggestions at a time. The accuracy level of suggestions improves as you type more code. Check out a high-level overview of GitHub Copilot.
Copilot: Pros and cons
Though Github Copilot aids in perfecting the lines of code, the programmers risk atrophying his/her creative muscles, said a Bengaluru based developer. Relying heavily on such AI pair programmers could lead to codependency, especially for rookie developers in the process of picking up the tricks of the trade, he warned.
Copilot is trained with public GitHub repositories, and might suggest code snippets from old libraries or modules, so it’s important to review the large bunches of code it throws up, Daniel Diaz, a technical writer, said in his blog.
Lizzy Lawerance from Protocol, recounted tech consultant Ady Ngom’s experience with Copilot. A tool in Copilot had anticipated the exact function Ngom wanted to type.
Copilot, which Ngom had installed only moments before, was one step ahead of him. “Can it listen to what I’m saying? Maybe that’s how it got it,” Ngom joked on a November LinkedIn live broadcast. “I was like, holy moly, you are just reading my mind right now. It was a surreal moment,” Ngom said.
Buddika De Silva, a software engineer from Project X Space, wrote a detailed blog after using Copilot for five months.
“Behind the Copilot is some IT giants like Microsoft and OpenAI. Microsoft is already serving AI services with their Azure cloud. Microsoft can serve an improved version of Copilot without needing a programmer to code like real AI programmers. The most valuable resource for any software development industry are the software developers and the most costly resource too. But it looks like, Copilot can replace developers, one day,” he said.
Copilot was built with OpenAI’s Codex tool and Microsoft-owned GitHub’s monstrous code database.
Ulf A, a Chief Data Scientist at Rodinia Generation said, “By my estimate, I am now 2-3 times more efficient when programming because Copilot saves me about 90% of the Google searches and documentation posts I would normally make. I’m not uncritical about what this means for software development in the future, but it’s a total game-changer for now!
Saumya Majumder, Founder & Chief Web Engineer @AcnamHQ. Tweeted:
I am truly amazed by the potential and craziness of #GitHubCopilot. I mean, it seriously is crazy. Initially, I was very sceptical about how good it will be or will it help me or not. But now, it literally completes my thoughts as I am thinking them. Like it can read my mind.