The Microsoft-owned company has teamed up with OpenAI to launch a technical preview of an AI assistant for coders.
GitHub, the code hosting service for developers, has launched a new AI tool that is designed to act like autocomplete for software developers.
The company, which was acquired by Microsoft in 2018, collaborated with OpenAI to develop GitHub Copilot.
This will act as an AI programming assistant in GitHub’s visual studio code editor. The new tool will give suggestions for lines of code or entire functions inside the editor, acting like predictive text for coders.
In a blogpost, GitHub CEO Nat Friedman wrote that the tool draws context from the code programmers are working on.
“It helps you quickly discover alternative ways to solve problems, write tests and explore new APIs without having to tediously tailor a search for answers on the internet,” he said. “As you type, it adapts to the way you write code – to help you complete your work faster.”
According to GitHub, the tool works with a broad set of frameworks and languages and the technical preview works well with Python, JavaScript, TypeScript, Ruby and Go.
It will allow programmers to cycle through several suggestions, allowing them to choose their preference or manually edit suggested code. The tool is designed to adapt to these edits to learn and match each programmer’s style.
The tool is powered by OpenAI Codex, a new AI system that was trained on a large dataset of public source code.
According to OpenAI’s co-founder and CTO, Greg Brockman, the underlying technology of the new AI tool won’t be exclusively for Microsoft.
In an interview with CNBC, Brockman said OpenAI will release the Codex model for third-party developers to weave into their own applications later this year.
OpenAI was founded in 2015 with the aim of ensuring that artificial intelligence “benefits all of humanity”.
Microsoft invested $1bn in OpenAI in 2019, with plans to work on “secure, trustworthy and ethical” AI to serve the public, while focusing on constructing new Azure AI supercomputing technologies.