News

GitHub Copilot wizard already incorporates GPT-4

GitHub has announced Copilot Xa new version of your Copilot coding assistant with GPT-4, which results in an increase in the possibilities and functions of the tool to reach more phases of the development process and code writing. The original Copilot, released last year, incorporated the GPT-3 model, and worked within the developer’s editing window by automatically completing lines or sections of code. It was also capable of generating code from natural language requests.

On top of this, Copilot X offers developers an experience more like ChatGPT. To do this, GitHub has incorporated a chatbot into the editor environment. It appears in various development tools, such as Visual Studio and VS Code. Developers who use it will be able to interact with it through text requests, or through voice.

According to him GitHub CEO Thomas Dohmke, with this release they seek to take GitHub Copilot X to all phases of development, and to be part of all the steps of its entire life cycle. That is to say, that it is not only used for autocompletion. Dohmke has also pointed out that his chatbot does more than just have a conversation, since it will have the full context of the development environment of its users. In this way, he will be able to understand what a developer has written in the development window.

That way, you’ll understand error messages as well as debug output in a terminal window. With data like this, you’ll be able to offer personalized advice, and play a leading role in all phases of development. From project planning to final code deployment. For example, during planning and tracking, GitHub Copilot X can help a developer describe what they want to build, and do it in natural language and in a concise way. In this way, it will facilitate its understanding to the rest of the team members.

At the coding stage, GitHub Copilot X will continue to rely on an older model than GPT-4. It’s called Codex, and it’s a variant of the GPT-3 model, whose mission is to automatically complete lines of code. Therefore, on GitHub Copilot X they are going to use different models for different use cases: Codex for autocompletion, which is faster, and a slower and more accurate one for generating test cases, as well as for answering questions and questions. for documentation.

In the code review process, Copilot X can help create test cases for your code, in order to verify that it performs the actions for which it was developed. Also find and fix small flaws in the code, which can lead to a bug.

As for the code deployment phase, Copilot can check again that the code that you want to release to production has been fully tested. If any part of the code is not tested, the wizard will display an error message when the developer clicks the deploy button.

So far, more than a million developers and more than 5,000 companies have used Copilot, which is available in two plans: the first, for developers, costs $10 per month, while the second, for enterprises, is priced at $19 per developer per month. Those who want to use gitHub Copilot X can Sign up on the waiting list to access the service in his web page.

Deepak Gupta

Deepak Gupta is a technical writer with a 10-year track record in business, gaming, and technology journalism. He specializes in translating complex technical data into actionable insights for a global audience.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *