TL;DR:
- GitHub introduces Copilot Enterprise subscription for code customization.
- Copilot Chat, a chatbot powered by OpenAI’s GPT-4, will enter general availability in December.
- Copilot Enterprise offers personalized code support and model fine-tuning.
- GitHub expands Copilot Chat to GitHub.com for enterprise subscribers.
- GitHub partners with third-party companies for Copilot integrations.
- Copilot Workspace, a natural language bridge for developers, is expected in 2024.
- GitHub enhances code security with an AI-driven autofix feature.
Main AI News:
In a strategic move to further empower businesses with cutting-edge code support, GitHub has announced the launch of its Copilot Enterprise subscription tier. This offering is designed to enable companies to tailor their Copilot pair-programmer to align seamlessly with their internal codebase. The news was unveiled at the annual GitHub Universe developer conference, which also shed light on other significant updates.
Copilot Chat, initially introduced by GitHub in March, has been making waves in the developer community. Following a successful public beta launch for business users in July and availability for individual users in September, it is now poised to enter general availability in December next month. This groundbreaking chatbot resides within the developer’s integrated development environment (IDE), providing them with the ability to seek assistance regarding the code they are actively working on. This includes tasks such as bug identification and proposed fixes, as well as offering inline feedback on specific lines of code. Powered by the latest OpenAI large language model, GPT-4, Copilot Chat is available as part of the standard Copilot subscription, priced at $10 per month for individuals and $19 per month for businesses.
The Copilot Enterprise subscription, set to debut in February 2024, will take code support to the next level. It builds upon the existing business plan by offering additional capabilities. This includes the option for companies to personalize Copilot Chat according to their unique codebase and refine the underlying models. The concept is simple: businesses can seamlessly integrate Copilot with their codebase, allowing developers to receive tailored suggestions for their internal, private code. This development also coincides with the expansion of Copilot Chat to GitHub.com for Copilot Enterprise subscribers. This expansion enables developers to explore their code, documentation, and pull requests more comprehensively, with Copilot Chat providing summaries, suggestions, and answers across a wider range of inquiries.
GitHub CEO Thomas Dohmke emphasized the significance of this advancement, stating, “With Copilot Chat connected to your repositories on GitHub.com, Copilot Enterprise allows your teams of developers to quickly get up to speed on your codebase, search through and build documentation, get suggestions based on internal and private code, and quickly review pull requests. By placing the collective knowledge of your organization’s codebase at their fingertips, your developers will not only write code faster, but deploy your next application, feature, or update ahead of the curve of your competition.”
As it stands, Copilot Chat already functions within private workspaces in the IDE, though it requires users to clone their repository locally. Copilot Enterprise eliminates this constraint, enabling AI-driven discussions on code and associated documentation in the cloud. Additionally, companies can fine-tune the model to enhance Copilot’s ability to complete code and address specific codebase-related queries.
Mario Rodriguez, VP of product management at GitHub, shared the company’s vision, saying, “Our ultimate goal is to deliver a Copilot that is conversational, ubiquitous, personalized, and trustworthy, and that’s exactly what we’re doing with Copilot Enterprise.”
AMD, a prominent chip manufacturer, was among the early collaborators with GitHub to test this feature. They reported that fine-tuning the Copilot model enabled them to gain support for hardware design languages like Verilog, a feat not achievable with the standard Copilot version. Alexander Androncik, AMD’s senior director of software development, praised the customized Copilot model for delivering accurate, high-quality AI suggestions tailored to their specific product design style.
In related news, GitHub announced its plans to bring Copilot Chat to the GitHub mobile app in the coming months. Additionally, it will expand support for JetBrains suite of IDEs, extending beyond the previously supported VS Code and Visual Studio code editors driven by popular demand.
GitHub’s commitment to enhancing Copilot’s capabilities is further underscored by the launch of the Copilot Partner Program. This initiative will foster collaborations with third-party developer tooling companies, including Datastax, LaunchDarkly, Postman, Hashicorp, and Datadog, to create integrations for Copilot. The aim is to expand the ecosystem and unlock new possibilities and use cases for GitHub Copilot, ranging from improving database query performance to managing feature flags and A/B testing.
GitHub is excited to embark on this journey with an initial roster of 25 partner companies, with an early-access program available for those interested in joining.
Finally, GitHub offered a glimpse of Copilot Workspace, an innovative feature slated for release in 2024. Described as a “natural language bridge” for developers, Copilot Workspace streamlines the transition from concept to executable code. When a developer opens an issue in Copilot Workspace, they receive an autogenerated plan outlining the steps to implement the desired changes. While this plan is customizable, developers can also steer the AI to align with their preferences, making it a powerful tool for efficient code development.
In the realm of security, GitHub is building upon its 2020 features by introducing new AI enhancements. The secret-scanning and code-scanning capabilities, which facilitate automated vulnerability detection and secret detection in public code, are now augmented by an “autofix” feature for code-scanning. This feature expedites the process of introducing fixes to CodeQL, JavaScript, and TypeScript alerts in pull requests. These AI-generated fixes offer precise, actionable suggestions to address vulnerabilities, enabling developers to seamlessly incorporate them into their codebase.
GitHub’s commitment to facilitating frictionless remediation aligns with its mission to empower developers with tools that enhance code quality and security.
Conclusion:
GitHub’s latest announcements at the GitHub Universe developer conference signal the platform’s unwavering dedication to providing innovative solutions and tools for developers and businesses alike. The introduction of Copilot Enterprise, the expansion of Copilot Chat, and the collaboration with third-party partners through the Copilot Partner Program all underscore GitHub’s commitment to empowering the developer community and driving innovation in the software development landscape. With these advancements, GitHub continues to be at the forefront of code collaboration and security.
