Before GitHub’s big annual conference in San Francisco, GitHub introduced Copilot Workspace, a dev environment that taps what GitHub describes as a “Copilot-powered agent”. This place will help developers come up with ideas, plan, test, run codes using simple language, and write different ideas. Thus, helping developers enhance their capabilities of performing different tasks.
Jonathan Carter, who leads GitHub’s research and development team, says that Copilot Workspace is a step up from GitHub’s existing AI tool, Copilot. This tool expands on features like Copilot Chat which help developers to discuss coding issues in simple 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 core base.
Jonathan Carter
Moreover, 1.8 million individuals and 50,000 enterprise customers use Copilot, and GitHub Workspace attracting more and more audience with its extended features. According to Carter,
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.
So, this new Workspace is providing a companion experience that complements existing tools and workflows, it’s aim is to simplify a variety of developers’ tasks, ultimately enhancing productivity and code quality.
Apart from such big success, GitHub faces financial pressures as Copilot reportedly incurs losses per user, thus highlighting the need for sustainable monetization strategies. This thing is already creating competition in the AI-powered coding tools market with some rivals like Amazon’s Code Whisperer and multiple other startups.
When a user gives Copilot Workspace a GitHub repository or a specific bug within one, it uses the OpenAI GPT-4 Turbo model to come up with a plan to fix the bug or even add a new feature. This feature looks into the repository’s comments, responses to uses, and the overall code to understand what needs to be done. After this, it suggested what code a user could fix or a new feature. Moreover, it also lists what needs to be checked and tested with this new code, and gives tools to change, save, improve, or reverse the changes.
You can run this suggested code right in Workspace and share it with your team through a link. Team members can then go into Workspace, where they can make more adjustments and improvements to the code as needed.
How to use GitHub’s Copilot Workspace?
If as a user you are looking for a way to start using Copilot Workspace, just click the “Open in Workspace” button next to issues and pull requests in the GitHub repositories. When you click this button, it lets you describe what you need to do in simple language, like “Add documentation for the changes in this pull request.” After a user submits this description, it gets added to a list of sessions in the Workspace.
This new Workspace handles these tasks step by step. It first creates a detailed plan of what needs to be done, then sets out to do it. Thus, developers can look closely at each step, see the code changes suggested, and make adjustments like recoding, changing the order of these steps, or deleting something as needed.
Carter stated, “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.”
GitHub has said that it’s still figuring out how to fully develop and sell Copilot Workspace as a product. However, they are planning to use the trial period to see how much it helps and how developers can use it in a better manner. Thus, they gather better information and make improvements based on real usage.
Despite the promises of AI-driven software development, some concerns remain regarding the quality and security of code generated by AI assistants. Some studies show the potential pitfalls, including increased code errors and security vulnerabilities associated with AI-generated code.
GitHub shows human review and employs AI-based vulnerability prevention systems to mitigate these risks. Yet, developers are actively using AI tools in their workflows, while a significant percentage already using or planning to adopt AI code assistants. So, while combining AI capabilities with human expertise, GitHub aims to create a collaborative development environment that maximizes creativity and productivity.
As GitHub launches a test version of Copilot Workspace, they are looking to get feedback from developers. This will help them improve the tool and understand its benefits better. Copilot Workspace is the highlighting step in using AI to help with coding, aiming to change the way software is developed.