GitHub is already one of the most popular platforms for creating and hosting programming projects. It started last year an integrated integrated development environment (IDE) so that you can code right in the browser. Now it’s rolling out an AI-powered autocompletion tool called GitHub Copilot.
The goal of this function is to learn how to write code and automatically suggest lines and functions. In addition, the Copilot tool helps you find alternative ways of solving a problem and makes it easy to use new APIs.
GitHub developed this feature in collaboration with OpenAI, but it is not based on GPT-3. This is what the company says on its blog about its technical performance:
OpenAI Codex has extensive knowledge of how people use code and is significantly more powerful than GPT-3 at code generation, in part because it was trained on a data set that contains a much greater concentration of public source code.
The function is also able to automatically fill in code that repeats itself, Conversion of comments or descriptions into working code and assistance in creating tests for your project., and help you create tests for your project.
AI-powered autocomplete has been around for a while. When we spoke to Codata last year, the founders said the tool helped them write 25% less code – and did more than a million developers have already used it. At that time, another such tool was called Kite – primarily geared towards Python-based programming – had more than 250,000 users.
Entering this market from GitHub gives developers more opportunities to leverage AI to complete their projects faster.