Meet Copilot Workspace: GitHub’s AI-enhanced Coding Assistant

Prior to its GitHub Universe conference scheduled for early autumn in San Francisco, GitHub unveiled Copilot Workspace, a development environment that leverages what it terms as “Copilot-powered agents.” These agents are designed to assist developers in various stages, from brainstorming and planning to building, testing, and running code using natural language.

What does Copilot Workspace offer?

Jonathan Carter, who leads GitHub Next, the company’s software research and development division, presented Workspace as an advancement of GitHub’s AI-based coding assistant, Copilot. This new iteration expands its utility with recent additions such as Copilot Chat, enabling developers to interact and inquire about coding issues in natural language.

“Through research, we found that, for many tasks, the biggest point of friction for developers was in getting started, and in particular knowing how to approach a [coding] problem, knowing which files to edit and knowing how to consider multiple solutions and their trade-offs. So we wanted to build an AI assistant that could meet developers at the inception of an idea or task, reduce the activation energy needed to begin and then collaborate with them on making the necessary edits across the entire corebase,” Carter stated.

As of the latest figures, Copilot boasts more than 1.8 million individual subscribers and 50,000 corporate clients. Yet, Carter believes the potential user base could expand significantly, spurred by the introduction of features like Workspace that have widespread appeal.

“Since developers spend a lot of their time working on [coding issues], we believe we can help empower developers every day through a ‘thought partnership’ with AI. You can think of Copilot Workspace as a companion experience and dev environment that complements existing tools and workflows and enables simplifying a class of developer tasks … We believe there’s a lot of value that can be delivered in an AI-native developer environment that isn’t constrained by existing workflows,” Carter stated.

Workspace, powered by OpenAI’s GPT-4 Turbo model, enables the formulation of strategies to address bugs or add new functionalities in a GitHub repository.

It utilizes insights from repository comments, issue responses, and the broader codebase. Developers receive code recommendations for fixing bugs or implementing features, complete with a checklist for validating and testing the code. This includes options to modify, save, refine, or revert changes.

The proposed code can be executed within Workspace and shared with colleagues through a link, allowing them to further adjust and perfect the code as needed.

Workspace processes tasks in a methodical sequence: it outlines a strategy, devises a plan, and executes it. Developers have the option to delve into these stages, closely examining the recommended code and modifications, and can rearrange, repeat, or remove steps as required.

“If you ask any developer where they tend to get stuck with a new project, you’ll often hear them say it’s knowing where to start. Copilot Workspace lifts that burden and gives developers a plan to start iterating from,” Carter stated.

GitHub has not yet decided on a specific strategy to commercialize Workspace, but it plans to use the initial preview period to gain insights into its utility and how developers interact with it.

By focusing on human oversight, Workspace might just be able to address some of the complications that arise from AI-generated code. The effectiveness of this approach will become clear as Workspace begins to be adopted by developers.

“Our primary goal with Copilot Workspace is to leverage AI to reduce complexity so developers can express their creativity and explore more freely. We truly believe the combination of human plus AI is always going to be superior to one or the other alone, and that’s what we’re betting on with Copilot Workspace,” Carter stated.

Featured image credit: Roman Synkevych/Unsplash