Are you developing on GitHub? You can continue to use GitHub, but use this tool to quickly and easily automatically sync your GitHub versions to SourceForge, giving your projects one location and making your project available to nearly 30 million SourceForge users monthly. It takes less than a minute.
Microsoft’s GitHub unit created one of the first mass-deployed programs using OpenAI’s language generation tools: an application called Copilot that helps software developers write computer code. Today, GitHub is adding a chat and voice feature that programmers can use to ask how specific coding tasks should be performed. Excerpt from a report:
The new release announced Wednesday is called Copilot X. Thomas Dohmke, managing director of GitHub, said he showed one of his kids how to code a snake game in Python. The chat window can provide explanations about the role of code segments, create ways to test code, and offer fixes for bugs. Developers can also provide instructions or ask questions using their voice.
GitHub previewed Copilot in 2021 and gave it wide distribution last year. The original product included a completion tool that suggested code snippets as the developer typed. It attracted hundreds of thousands of developers in November, and its product name has become an acronym for Microsoft’s strategy of using such assistive technologies in a wide range of its products, from office software to security programs. With OpenAI’s ChatGPT chatbot making a name for itself in popular culture, companies are trying to follow Microsoft by incorporating the research lab’s tools into their products and business strategies. At the same time, competitors like Google (Alphabet) are launching competing chatbots.
GitHub Copilot is evolving to provide chat and voice interfaces, support pull requests, answer questions about documents, and adopt OpenAI’s GPT-4 for a more personalized experience for developers.
Thomas Dohmke says:
Our mission at GitHub has always been to innovate and give developers everything they need to be happier and more productive in a software-driven world. When we started experimenting with large language models a few years ago, it quickly became clear that generative AI is the future of software development. We’ve partnered with OpenAI to create GitHub Copilot, the world’s first generative AI development tool built using OpenAI’s Codex model, a descendant of GPT-3.
GitHub Copilot ushered in a new era in software development as an AI-powered programmer that keeps developers flowing with autocompletion of comments and code. Less than two years after its launch, GitHub Copilot is already writing 46% of the code, helping developers code up to 55% faster.
But AI-powered autocomplete is just the beginning. Our GitHub Next R&D team worked to go beyond the editor and evolve GitHub Copilot into an easily accessible AI assistant throughout the development cycle. This is GitHub Copilot X, our vision for the future of AI-powered software development. Not only are we adopting OpenAI’s new GPT-4 model, we’re introducing chat and voice to Copilot, and we’re making Copilot fetch requests, the command line, and documentation to answer questions about your projects.
With AI available at every stage, we can fundamentally redefine developer productivity. We reduce repetitive and manual tasks and facilitate complex work throughout the developer lifecycle. In this way, we enable every developer to focus all their creativity on what matters most: developing tomorrow’s innovation and accelerating human progress today.
A new development experience powered by AI
A ChatGPT-like experience in your editor with GitHub Copilot Chat: The editor has a chat interface focused on development scenarios and natively integrates with VS Code and Visual Studio. This interface does much more than just suggest code. GitHub Copilot Chat isn’t just a chat window. It recognizes developer-entered code, displays error messages, and is deeply integrated into the IDE. A developer can get detailed analysis and explanations of what blocks of code are supposed to do, generate unit tests, and even get suggested fixes.
GitHub Copilot Chat builds on OpenAI and Microsoft’s work with ChatGPT and the new Bing. It will also join the previously introduced extension of voice-to-code AI technology, now called GitHub Copilot Voice, which allows developers to give natural language instructions orally.
Co-pilot for pull requests : You can now subscribe to a technical preview of the first AI-generated descriptions for pull requests on GitHub. This new capability is based on OpenAI’s new GPT-4 model and adds support for AI-powered tags in pull request descriptions via a GitHub application that organization admins and individual vault owners can install. These tags are automatically populated by GitHub Copilot based on the changed code. Developers can then review or change the proposed description.
This is just the first step in rethinking how download requests work on GitHub. New functionalities are tested internally. GitHub Copilot automatically suggests sentences and paragraphs when developers create pull requests by dynamically incorporating information about code changes.
Additionally, a new feature is in the works: GitHub Copilot will automatically notify developers when they’re missing tests for a pull request and suggest potential tests that can be edited, accepted, or rejected based on the developer’s needs.
This complements efforts with GitHub Copilot Chat, where developers can ask GitHub Copilot to generate tests directly from their publisher—so if a developer doesn’t have sufficient test coverage, GitHub Copilot will notify them as soon as they submit a pull request. It also helps project managers define testing guidelines and supports developers in complying with those guidelines.
Get AI-generated answers on documentation : GitHub Copilot for Docs is an experimental tool that uses a chat interface to give users AI-generated answers to documentation questions, including questions developers have about the languages, frameworks, and technologies they use. For starters, the documentation for React is Azure Docs and MDN, making it easy to learn and grow quickly with developers and users of these projects.
This functionality also applies to repositories and internal documentation of each organization. Thus, any developer can ask questions about the documentation, idiomatic code or internal software of his organization via a ChatGPT-like interface and get answers immediately.
Recognizing that the benefits of a conversational interface are immense, GitHub is working on a semantic understanding of GitHub across public and private knowledge bases to better personalize GitHub Copilot answers for organizations, teams, companies, and individual developers based on their code and documentation.
In the future, new ways will be explored to index resources beyond documentation, such as: B. Issues, Enhancement Requests, Discussions, and Wikis to give developers everything they need to answer their technical questions.
Powered by the new OpenAI GPT-4 model
Work on redesigning pull requests and documentation is powered by OpenAI’s new GPT-4 AI model.
Although this model has only just been published, there are already significant advances in reasoning and code generation. With GPT-4, the state of AI is beginning to catch up with the ambition to create an AI-enabled programmer to assist with any development task at any stage of the developer’s experience.
Additionally, it helps GitHub Copilot better understand a developer’s code base to offer more personalized suggestions in PRs and better documentation summaries.
Co-driver for Command Line Interface (CLI)
After the editor and the pull request, the terminal is where developers spend most of their time. But even the most experienced developer has to dig through many pages to remember the exact syntax of many commands. That’s why we launch GitHub Copilot CLI. It can create commands and loops and set obscure search flags to fulfill your request.
Source: GitHub
And you ?
What do you think of GitHub Copilot X?
Do you think an AI-powered program will greatly improve the developer experience?
Are you planning to use this type of tool?
See also:
Microsoft 365 Copilot: The omniscient version of Clippy boosts AI, automates data analysis in Excel, writes documents in Word based on other files, and more
AI-supported programming assistants are driving a rethink in computer science education, researchers say
Copilot, GitHub’s commercial AI tool, faces criticism from the open source software community,
For his blind reproduction of blocks of code