AI is everywhere, and it’s presence is rapidly expanding. The capabilities of machine learning systems and the power of the computing in these systems for consumer usage is unprecedented. Microsoft is one of the companies trying to press into the space in countless different ways, including Cortana. The coming Build 2017 conference on May 10th should give insight into where Microsoft is moving in this market space.
For starters, Microsoft has sought to insert AI into it’s standard issue flag ship programs. For example, Powerpoint now offers AI interaction with users, suggesting pictures and captions without being asked. The user can edit or reject the information, but the program will learn from the rejections and changes and make adjustments in order to improve future suggestions. Even Office has AI built in structurally. Word can make suggestions for spelling, grammar, and even word choice, all of which have improved greatly with the addition of machine learning into the mix.
Microsoft’s answer for Apple’s Siri, Cortana, is built on an AI platform, and is reasonably powerful and able to help users, but the vast majority of users turn off the speaking application on their computers. It seems that Cortana is not Siri, and many users have an iPhone and so don’t like the mixing platforms. However, Cortana does allow for searches by simply typing into the search bar by the start menu, and thus the system is still learning and predicting user expectations. At this point, though, Cortana just can’t compete against Apple’s market-dominating Siri. Without a more substantial phone presence, Cortana has an uphill climb toward garnering more users.
However, the place where Microsoft has concentrated has been on developer tools that serve the front of the marketplace rather than end users. The main focus has been on the Cloud platform Azure, where Microsoft is offering a number of extremely useful services. The addition of a machine learning service (Azure ML) that interacts with large data systems and provides predictive analysis has been helpful. They also offer IaaS (infrastructure as a service) virtual machines which allow users or developers to launch general Windows or Linux machines as well as pre-configured machines for popular software packages. They’ve also been publishing a large volume of programming interfaces on Azure, allowing developers to build their own language, image and speech intelligence into their apps.
Ultimately their focus is to start making AI and understanding a part of everything they’re sending out in order to encourage shareable systems. The final goal is that some of these integrations will take root in the most commonly used systems and will be taken for granted. As the direction of the company is shifting, the focus on AI is growing, and will likely continue to expand. As data sets rapidly increase in size and the demand for rapid learning and response systems increase, so will Microsoft keep trying to fill the holes the marketplace has provided for them.