GitHub Copilot is now available to everyone

Copilot is definitely one of the most interesting GitHub initiatives in recent times, to the point that during the nearly one year that it has been in the testing phase, it has aroused enormous interest in the developer community, with many of them wanting to join said beta to be among the first to test this system. GitHub encrypts those registered in the Copilot testing phase at no less than 1.2 million users.

GitHub, owned by Microsoft, which acquired the popular repository now just over four years ago, considers that it has already carried out all the necessary tests and that, consequently, Copilot is now ready to be used by all those users who want it. The bad news, yes, is that now that it is a service out of the testing phase, those users who wish to use it will have to pay a subscription, with a monthly price of ten dollars.

If you still don’t know what Copilot is, it is an artificial intelligence-based service that allows programmers get code suggestions based on what they’ve already written, the purpose that they have declared with natural language and, of course, the knowledge acquired by the tool. According to data from GitHub, the latest data from Copilot indicates that approximately 40% of the code uploaded to the platform has been generated with this system.

Copilot is optimized for programming in Python, JavaScript, TypeScript, Ruby, Go, C# and C++, but it is also capable of proposing suggestions for other languages ​​and also for a multitude of frameworks, something in which, furthermore, we can imagine that it will improve over time, thanks to learning once it is already available to all users . Copilot extensions are available for Noevim and Jet Brains in addition to, of course, Visual Studio Code and in the cloud with GitHub Code spaces.

Due to the complicated nature of AI models, Copilot is still an imperfect system. GitHub said it has implemented filters to block emails when displayed in standard formats, and offensive words, and is in the process of building a filter to help detect and remove code that is repeated from public repositories. But the company acknowledges that Copilot can produce insecure coding patterns, bugs, and references to outdated APIs, or idioms that reflect less-than-perfect code from your training data.

However, we must not forget that with Copilot we are talking about a tool to help the programmer, but not a fully automatic programming solution. We can expect, yes, that interesting improvements continue to be produced in it.

More information: GitHub