Barnacules, a development engineer who worked for Microsoft for fifteen years, released a video analysis of why Windows 10 is currently very poorly stable. Obviously Microsoft is unacceptable for this view, so Microsoft also emphasizes the role of artificial intelligence and machine learning technology in the upgrade process through an official blog.
Simply put, Microsoft will collect the feedback data of the device through the telemetry data. If the upgrade of a certain model is no problem, it will be pushed to other devices of the same model. Microsoft said that after the deployment of the machine learning algorithm, the device update was significantly reduced due to the kernel crash, and the crash caused by the driver problem was also reduced by five times. However, the most stable and quality improvement should be the user. It is estimated that most Windows 10 users will not agree with Microsoft.
Microsoft’s response in this official blog is actually very interesting because both artificial intelligence and machine learning are Microsoft’s previously deployed update strategies. This update strategy has been used for at least a year, but the stability of Windows 10 has not improved significantly, and it seems that various problems have become more. However, Microsoft emphasizes the stability of the major version update of Windows 10, and the former engineer actually refers to the current monthly cumulative update.
Microsoft did not mention the reason why there are so many problems in the routine update every month. At the same time, the data released by Microsoft only shows the reduction of the system crash frequency. It is estimated that this problem cannot be completely solved by machine learning.