GitHub, owned by Microsoft, today launches its Copilot AI tool, which helps propose lines of code to developers in their code editor. GitHub originally partnered with OpenAI to launch a preview of Copilot last year, and it’s now generally available to all developers.
Priced at $10 per month or $100 per year, GitHub Copilot is able to propose the next line of code as developers type in an integrated development environment (IDE) such as Visual Studio Code, Neovim, and JetBrains IDEs. Copilot can propose complete methods and complex algorithms in addition to standard code and assistance with unit testing.
More than 1.2 million developers have signed up to use the GitHub Copilot preview in the past 12 months, and it remains a free tool for verified students and administrators of popular open source projects. In files where it’s enabled, GitHub says nearly 40 percent of the code is now written by Copilot.
“Like the rise of compilers and open source, we believe that AI-assisted coding will fundamentally change the nature of software development, giving developers a new tool to write code easier and faster so that they can be happier in their lives,” says GitHub CEO Thomas Dohmke.
Microsoft’s $1 billion investment in OpenAI, the research firm now headed by former Y Combinator president Sam Altman, led to the creation of GitHub Copilot. It is built on OpenAI Codex, a descendant of OpenAI’s flagship GPT-3 language-generating algorithm. However, GitHub Copilot is controversial. Just days after the preview launched, there were questions about the legality of Copilot training on publicly available code posted on GitHub. Aside from copyright issues, a survey also found that about 40 percent of Copilot’s output contained security vulnerabilities.
Microsoft isn’t the only company working on automated AI tools to help with coding. Last year, Google-owned DeepMind unveiled an AI system called AlphaCode designed to write computer programs “at a competitive level.” AlphaCode was tested with Codeforces, a competing coding platform, and achieved an “estimated rank”, placing it in the top 54 percent of human coders. The challenges are different than what a regular coder would face, but they show how AI coding systems can help coders in the future.