GitHub Copilot Chat: AI-assisted coding chat interface, supporting multiple IDEs

GitHub Copilot Chat is a chat interface that allows developers to ask coding-related questions directly in supported IDEs and get answers. It is currently in public testing and is available to all GitHub Copilot personal users in Visual Studio and VS Code.

Originally released as a beta version of an AI-driven software development extension for all GitHub Copilot commercial users, GitHub Copilot Chat is now available for free to all individual users, providing support in Visual Studio and Visual Studio Code editors. This extension leverages advanced natural language processing (NLP) technology to provide assistance and answers in natural language directly within supported editors, eliminating the need for developers to consult documentation or search online forums.

In a blog post, Shuyin Zhao, Vice President of GitHub Copilot, wrote:

"By reducing the need for context switching, it simplifies the development process and helps developers stay focused and motivated."

GitHub Copilot Chat also enables individual contributors to suggest security vulnerability patches, thereby enhancing the overall security of open-source projects. GitHub also provides detailed information about other features of Copilot Chat, including:

• Real-time, customizable best practices, tips, and solution suggestions
• Code snippet analysis that explains complex concepts
• Providing fix suggestions to reduce the number of vulnerabilities found in security scans
• Offering hints, explanations, and alternative solutions for identifying issues during debugging

As Copilot Chat becomes available to more users and offers more features, there is still a demand for it among users.

GitHub conducted research on the usability of Copilot Chat and concluded, "Even developers who have not previously used the Copilot Chat feature experience a significant improvement in the quality of code writing and reviewing after enabling Copilot Chat."

However, Rafa Darader questioned this research, stating on Twitter:

"Hey, I like Copilot, but this research hardly touches on quality. The metrics outlined are either about how developers feel when adding Copilot to their workflow."

Can quality only be measured by looking at the part involved in generating something?"

Finally, to use Copilot Chat, users must have a GitHub Copilot subscription and VS Code users must have the latest version and log in to VS Code with the same GitHub ID that has access to GitHub Copilot.