Microsoft is shaping the future with AI that aims to do much of the work on your PC. The goal is for AI to handle tasks while developers focus on creativity. CEO Satya Nadella mentioned that AI produces up to 30% of the code for some projects. I’ve been digging into how Microsoft developers actually use this tech.
Some developers are not fully convinced AI can completely take over their work. They worry about the errors that automated systems might create, leaving humans to pick up the pieces. However, Microsoft is pushing forward, aiming to integrate AI everywhere to make things run more smoothly.
Amanda Silver from Microsoft’s CoreAI team explains, “We’re looking at where there’s developer toil and inefficiencies.” With over 100,000 code repositories at Microsoft, there’s plenty of room for AI to make a difference. The aim is to go beyond just suggesting code to actually automating parts of the development process.
In May, Microsoft added a coding agent to GitHub Copilot. This feature allows developers to delegate tasks. It creates its own work environment and even submits pull requests when done. According to Silver, this has saved developers significant time: around 30 minutes on simple tasks, up to two weeks on complex ones.
Microsoft is meticulous in measuring how AI impacts productivity. Some metrics include the time saved and how many tasks the AI completes. Despite positive reports, there’s a concern among some developers about how often AI tools are being used. Studies show that while most developers are aware of AI tools, usage varies widely.
For instance, although Microsoft claims that 91% of its engineering teams use GitHub Copilot, a survey found that only 51% of developers use AI tools daily. There seems to be a disconnect between executive optimism and actual daily practice among developers.
Silver mentions that certain teams, like the Xbox analytics team, have seen remarkable benefits from AI. They upgraded their services with minimal manual effort due to Copilot, which streamlined their development process.
Still, challenges remain. Some employees fear that autonomous AI could eliminate entry-level roles, leaving more experienced developers to oversee machines rather than code themselves. There’s also skepticism about how effective AI tools really are, with some developers joking about how little they actually use them.
As Microsoft continues its journey into AI, the hope is that these tools will allow developers to shed tedious tasks and return to innovative projects. Silver highlights that most developers want to create and problem-solve, not get bogged down in maintenance.
While the specifics of AI’s contribution to Microsoft’s code are hard to quantify, it’s clear that the landscape of software development is changing. AI isn’t perfect. Developers still need to review AI-generated code to ensure quality.
In summary, Microsoft’s venture into AI is both exciting and fraught with concerns about job security and productivity. If managed well, AI could indeed free developers to focus on what they love most—creating and innovating.
For more insights on this topic, you can read the 2025 AI Developer Study.
Source link
Microsoft,Notepad,Tech
