GitHub Copilot now assists in project initiation with AI, not just project completion.

GitHub Copilot is expanding the reach of its Copilot code completion platform, powered by GPT-4, to cover all stages of development, from project inception to completion. The introduction of GitHub Copilot Workspace marks a significant milestone, aiming to streamline the initial stages of coding projects. Initially available in a technical preview for a select group of developers, Workspace will later be fully integrated into the GitHub Copilot platform.

According to GitHub’s blog post, Copilot Workspace will seamlessly integrate into GitHub repositories or libraries. Developers can communicate their project goals to Copilot Workspace through prompts, receiving tailored suggestions and a step-by-step guide to kickstart their projects. Users retain full control, being able to modify suggestions before executing the code. Additionally, Copilot can assist in completing the code as needed.

Since its launch, GitHub Copilot has emerged as a widely adopted tool and a standard measure for evaluating new AI models’ capabilities. While many AI models focus on specific tasks like summarization or writing, GitHub’s parent company, Microsoft, has introduced Phi-3, a compact model designed specifically for code-writing tasks. Other tech giants, such as Meta, have also unveiled their versions of large language models tailored for coding projects.

More info