Microsoft Touts Virtues Of Windows 10's AI-Backed Updates Mechanism Even As Critical Bugs Persist

Windows 10 Update
In a new blog post, Microsoft extols the benefits of somewhat recently implementing machine learning into its Windows 10 update process. This is an evolving process, though from Microsoft's vantage point, it has been a success so far, allowing it to identify which Windows 10 PCs are the least likely to run into an issue when rolling out a new update.

The first broad scale use of this process was when Microsoft pushed out its April 2018 Update for Windows 10 (build 1803).

"We started with six core areas of PC health (e.g. overall PC reliability) to determine whether the feature update process went smoothly. With Windows 10, version 1903 (the May 2019 update), our third iteration of using ML in a feature update rollout, we can now evaluate 35 areas of PC health and the process will continue to evolve with additional health measures to improve your update experience," Microsoft explains.

A smooth update experience is more important than ever—Windows 10 is now installed on more than 900 million active devices.  At the same time, the update process is certainly not flawless. We have seen monthly updates cause issues, such as the most recent Patch Tuesday roll out killing Wi-Fi on some PCs and breaking audio on others. And of course the larger feature updates have had their issues as well. The October 2018 Update in particular was a train wreck of bugs.

Windows 10 Machine Learning Graph
Source: Microsoft

In theory, this should improve over time, assuming Microsoft's machine learning model is working as advertised. It already has to some extent, according to Microsoft's data—in the graph above, Microsoft shows how PCs chosen via machine learning to receive updates have fewer than half the number of system initiated uninstalls and kernel mode crashes, and five times fewer post-update driver bugs.

"Every release starts with offering the Windows 10 update to early adopters (such as Windows Insiders and those actively seeking out the update). Once the initial set of PCs has been offered the update, we monitor their update experience via diagnostic data (e.g. kernel mode crashes, system initiated uninstalls, abnormal shutdowns, and driver issues)," Microsoft says.

This is where machine learning comes into play. It analyzes the data and identifies potential issues, and makes subsequent predictions, nominating PCs that should have a smooth update experience.

There is quite a bit that goes into this, and Microsoft details the tech in its blog (hit the link in the Via field below). At the same time, Microsoft rightfully acknowledges "there is still much work to be done." Hopefully things will go smoother when the next feature update (19H2) arrives next month.