Today, GitHub is launching a new and improved version of its Copilot coding assistant powered by GPT-4.
In 2022, Copilot was introduced to the public and was based on the GPT-3. It works within a developer's editor window to speed up code writing by either auto-completing code segments or generating them from plain language requests.
See also: How reliable is ChatGPT?
Thomas Dohmke, CEO of GitHub, outlined his plans to revolutionize the developer experience with Copilot: “This program will not only provide autocomplete capabilities, but will significantly improve every stage of a software engineer’s journey.”
In a conversation with The Verge, Dohmke talked about the new Copilot X and the planned integration of AI into every step of the developer experience. “We are preparing for a future where AI will be an integral part of every stage of technological development,” he declared.
Advanced Copilot brings to life a ChatGPT -like experience . GitHub has integrated an AI chatbot into the editing environment of its development tools, including VS Code and Visual Studio . Using either text or voice commands, developer users can now engage with the interactive Copilot interface in real time!
Dohmke emphasizes that this chatbot does more than just engage in conversations. “It recognizes the full context of your editing environment,” he notes, “so it understands what you’ve entered into the coding window. It also understands error messages and debug output in your terminal window — allowing it to provide you with personalized guidance.”
According to Dohmke, Copilot can contribute substantially at every stage of development – from project conception to pushing code for deployment.
During capture and tracking, this tool will help the developer communicate their plan for construction in clear terminology that can be understood by the other members of their team.
As it enters the coding phase, Copilot will continue to benefit from Codex (a super-fast version of GPT-3), which auto-completes lines or sections of code instantly.
See also: Opera browser adds ChatGPT and AI summarization features
According to Dohmke, they now use different models depending on the purpose – a fast one for auto-completion and a slower but more accurate model for generating test cases, answering questions, and creating documentation.
Copilot can be extremely useful during the code review process, as it builds test cases to guarantee that the code works as intended and reveals potential coding errors to ensure they are addressed quickly.
As you deploy your code, Copilot can check to see if everything has passed rigorous testing before going live. If any part of the code hasn’t been tested, a notification will appear when you hit deploy – so says Dohmke! This ensures that only fully tested software is put into production and gives developers a sense of certainty about their work.
GitHub has reached an amazing milestone: over 1 million developers and 5,000 businesses are now using Copilot!
Copilot is now open for waitlist registration for developers who want to start using this innovative product. Registration will be granted in waves based on demand levels and as recommended by Dohmke , meaning more developers will be able to access it soon!
GitHub Copilot has already had a huge impact on developer productivity, helping developers achieve a 55% faster process. Dohmke believes that with the recently added chat features, this goal is achievable, and AI assistants like Copilot will be fundamental to how coding skills are acquired in the future.
Individuals can enjoy the benefits of Copilot for just $10 per month, while businesses and developers can harness the power of this tool for just $19 per person per month.
Source: fastcompany.com
