GitHub, the collaborative software development platform, has partnered with OpenAI to create an artificial intelligence tool that helps to write code more efficiently.
This solution, called GitHub Copilot, is capable of suggest code and sometimes full functions. All as the developer works within the code editor.
The partnership has enabled OpenAI’s artificial intelligence system to be trained with millions of lines of code stored on GitHub. In this way, you have learned to offer accurate help.
“We are designing GitHub Copilot to be smarter at producing safe and effective code as developers use it.”
GitHub Copilot is not looking to replace developers, but focus as a tool to improve productivity and facilitate the learning of various programming languages.
The core of GitHub Copilot, called Codex, is a descendant of GPT-3, OpenIA’s powerful AI model.
How does GitHub Copilot work?
GitHub Copilot makes writing code very easy. The system parses editor context and transforms comments into code. This can be very useful for quickly producing repeating code patterns.
In addition, GitHub Copilot suggests tests that match the code you are writing and makes it easy to develop software in any of the supported programming languages.
The changes are not applied immediately, the user can choose which suggestion to accept or reject. You can also manually edit the suggested code.
Copilot is able to learn from the code that it analyzes every day to offer more specific and useful suggestions. This thanks to the artificial intelligence model that gives it life.
The tool integrates directly into Visual Studio Code. It can be installed as an extension or used directly in the cloud through GitHub Codespaces.
Of course, it is currently available as a Technical Preview. A commercial version of GitHub Copilot is expected to be released later, although there is currently no set date.