Latest news with #GithubCopilot
Tom's Guide
22-05-2025
- Tom's Guide
Microsoft is all in on AI at Build 2025: 3 things you need to know
Microsoft's annual Build developer conference is happening this week (May 19-22) in Seattle, and I've been here in person to see it live. I trucked up to Seattle to attend Build 2025 because Microsoft occasionally debuts new hardware at its annual conference, like it did last year by unveiling the Surface Pro 11 and Surface Laptop 7 with Snapdragon chips inside. I was hoping to get a chance to go hands-on with new hardware, but I was out of luck—Microsoft unveiled a new smaller Surface Laptop and Surface Pro weeks before the event, and they were nowhere to be seen at Build 2025. Instead, all I heard about this week was AI. From the first minutes of CEO Satya Nadella's Build 2025 keynote it became clear that Microsoft is going hard on AI this year. Since Build is a developer conference much of the talk is about how to build and troubleshoot these systems, so it's been hard to parse how these new AI tools will impact the products and services we use every day. That said, here are the three most interesting things I've seen at Build 2025! Microsoft is rolling out a new coding agent for Github Copilot that is designed to work like an automated programmer, so ideally a coder could do things like assign annoying or repetitive tasks to Github Copilot instead of wasting human time doing the work. I got to see this AI agent in action several times during Build 2025, and it's a great example of the sort of "agentic web" that Microsoft employees have been talking up during the event. Put simply, Microsoft envisions a future where we have many AI agents running around doing work for us in the background—and the Github Copilot coding agent can do things like plan, write and test code on its own, so ideally programmers can assign it tasks they don't want to do, then check back later to review and verify the agent's work. Get instant access to breaking news, the hottest reviews, great deals and helpful tips. The demos I've seen of this tool in action here at Build 2025 are impressive, and it raises interesting questions about how programming work will be done in the future. Microsoft Discovery is a new platform built on Azure that aims to help scientists conduct research more efficiently using AI agents. This is another example of Microsoft's "agentic" approach to AI, as Discovery is intended to let researchers create virtual teams of AI agents that can be instructed with specific data and guidelines to function as, for example, a "molecular properties simulation specialist" that can then assist with R&D and learn over time. Microsoft claims this technology is already being used at places like the Department of Energy's Pacific Northwest National Laboratory, where researchers have used it to discover a new, more efficient solid-state electrolyte candidate. I couldn't tell you what that means for science, but I can tell you this pushes Microsoft's AI further into the R&D business — and that means we'll likely see more scientific discoveries with AI help. This one is a bit technical, but I think it's worth knowing about because it helps reveal where Microsoft is trying to take the future of business. The new Microsoft 365 Copilot Tuning features let you more easily create and fine-tune AI agents in Microsoft 365 Copilot Studio without knowing how to code them yourself. It's the kind of thing you'll only care about if you already use Copilot Studio, as the new Tuning features let users more easily create multiple agents and use company data to train them. So while you might not be using Copilot Tuning to fine-tune an AI agent to order your morning coffee, chances are that in the next few years your coffee shop and businesses like it will be using these kinds of agents in all sorts of ways. Microsoft gives examples of law firms creating agents that can quickly generate legal paperwork using data and processes of the law firm itself, for example, and I expect companies will find inventive ways to employ these agents when this functionality starts rolling out in June to members of the Microsoft 365 Copilot Tuning Early Adopter Program. If you made it this far, you know by now that Microsoft really was all about AI at Build 2025. I couldn't escape it no matter where I went, and the live demos I did see were pretty convincing. I still don't trust AI to reliably get things right, but I'm starting to see how it's being honed and adapted to try and fit into a broad variety of different businesses. As a writer I have my own concerns about AI agents getting involved in my industry, but I can't help feeling a little optimistic about projects like Github Copilot and Microsoft Discovery. More help for researchers is always a good thing (once they get these AI assistants to the point where they aren't making critical mistakes), and as someone who can't code his way out of a paper bag there's real appeal to knowing one day I'll probably be able to ask AI to do all my coding for me. Microsoft certainly thinks so, because the company is betting big on AI in 2025 and beyond.
Time of India
20-05-2025
- Business
- Time of India
Google I/O 2025: Google's answer to Microsoft and OpenAI's AI coding agents, Jules is now available for everyone to try
Google has made its AI-powered coding agent Jules available to the public in beta, allowing developers worldwide to experiment with the asynchronous coding assistant. Announced days after Microsoft and OpenAI announced their AI coding agents Github Copilot and Codex, at Google I/O 2025 , the Jules lets users continue working on other tasks while Jules handles coding problems in the background. Jules integrates directly with existing projects and repositories, cloning codebases to a secure Google Cloud virtual machine where it can work alongside developers. The AI agent can build features, provide audio changelogs, update dependency versions, write tests, and fix bugs independently. "Jules handles bug fixes and other time-consuming tasks while you focus on what you actually want to build," Google stated when first introducing the tool in December 2024. The agent has already generated substantial attention in the developer community, with many comparing it favorably to OpenAI's Codex and Microsoft's GitHub Copilot. The AI coding assistant 's key advantage is its asynchronous workflow, which differentiates it from competitors that primarily function as intelligent autocomplete tools. While using Jules, developers can assign coding tasks and return later to review completed work rather than watching the AI generate code in real-time. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like This Device Made My Power Bill Drop Overnight elecTrick - Save upto 80% on Power Bill Pre-Order Google emphasizes that Jules maintains privacy when working with private repositories and doesn't train on private code. The tool operates directly within the Gemini app , eliminating the need for additional software. During the public beta period, Jules will be accessible to anyone with Gemini access wherever the service is available. While Google hasn't announced pricing details for the eventual full release, the company indicated that cost information will be provided as the platform matures. The release comes amid increasing competition in the AI coding assistant space, with Microsoft recently enhancing GitHub Copilot and OpenAI launching new coding capabilities. For developers interested in trying Jules, the beta is available now, though usage limits apply. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Axios
20-05-2025
- Business
- Axios
AI agents will do the grunt work of coding
AI makers are flooding the market with a new wave of coding agents promising to relieve human programmers of busy work. The big picture: Automating the routine aspects of technical labor will almost certainly transform and downsize the tech industry workforce — but there's no guarantee it will alleviate software development's biggest headaches. Driving the news: Microsoft Monday announced a new AI coding agent for Github Copilot that's good for "time-consuming but boring tasks." "The agent excels at low-to-medium complexity tasks in well-tested codebases, from adding features and fixing bugs to extending tests, refactoring code, and improving documentation," Microsoft's post says. Github's move follows Friday's announcement by OpenAI of Codex, a "research preview" of a new coding agent that can "work on many tasks in parallel." Notably, the Github Copilot agent is powered not by Codex or any other tool from Microsoft partner OpenAI, but instead by Anthropic Claude 3.7 Sonnet, per Microsoft. The intrigue: Tech leaders have sent mixed messages on just how much work they see ahead for programmers. Amazon Web Services' then-boss Matt Garman caused a stir last year when he suggested the need for human coding could disappear within two years, However, he later told Axios that his comments were taken out of context. "I think it's incredibly exciting time for developers," he told us last year. "There's a whole bunch of work that developers do today that's not fun." "If you think about documenting your code, if you think about upgrading Java versions, if you think about looking for bugs, that's that's not what developers love doing. They love thinking about, 'How do I go solve problems?' " Why it matters: Business transformations that start in Silicon Valley usually make their way into the wider economy. Silicon Valley's "dogfooding" tradition ensures that it will avidly apply new technologies to its own business first. Both Microsoft and Google are now claiming that roughly 30% of the code they produce is AI-written. Coding agents, like other generative AI tools, continue to "hallucinate," or make stuff up. But programs, unlike other kinds of language products, have a built-in pass-fail test: Either they run or they don't. That gives programmers one early checkpoint to guard against bad code. Yes, but: AI-generated code likely also contains tons of other errors that don't show up today. That will cause nightmares in the future as programs age, get used more widely, or face unexpected tests from unpredictable users. Zoom out: The software industry's assumption that what works inside tech will work everywhere else could be sorely tested when these techniques get pushed out beyond Silicon Valley. AI's usefulness in writing code may not easily transfer to other kinds of work that are less abstract and more rooted in physical reality — witness the many setbacks and challenges the autonomous vehicle industry has faced. Between the lines: Nobody doubts that AI means tech firms will write more code using fewer employees. But no one yet knows exactly where these companies will continue to find competitive advantage. AI models are much more likely to be interchangeable than human organizations and cultures. What's next: As coding agents shoulder routine labor, product designers and creative engineers will use "vibe coding" — improvisational rough drafting via "throw it at the wall and see what works" AI prompting — to do fast prototyping of new ideas. The bottom line: The biggest challenges in creating software tend to arise from poorly conceived specifications and misinterpretations of data, both of which are often rooted in confusion over human needs.
