While Microsoft can hardly be called a pioneer in the personal digital assistant space—Siri predates Cortana by three years—the software giant has moved much more quickly than Apple to integrate this technology across a wide range of product types. And at Build 2016 last week, we got a fascinating peek at where Cortana is going next: Which is to say, everywhere.
The looming battle of personal digital assistants, which pits Cortana against Apple’s Siri, Google Now, and Amazon Alexa, is an important one, and is widely regarded as the next great platform fight. By combining natural language interaction on the client with a host of cloud-based technologies, including machine learning, these products represent a next-generation user experience that goes well beyond pointing and clicking/tapping and typing while looking at a screen. What personal digital assistants are really about is pervasive computing, the notion that you can find what you need at any time, and from any place.
Cortana first debuted on Windows phones in 2014, where it was barely noticed by the outside world. But the expansion of Cortana to Windows 10 for PCs and tablets in 2015—and later on Android, iOS, Surface Hub, HoloLens, IoT devices, and Xbox One—shows how important this work is to the software giant. Indeed, unlike any of its competitors, Microsoft is uniquely positioned to succeed in this market because it supports all major clients to some degree, and also has the cloud-based muscle and machine learning expertise to make Cortana truly useful and pervasive.
Sitting here today, Cortana’s features are, well, basic but improving steadily, and there is only minimal cross-platform functionality, also improving steadily. So you can set a reminder on your PC, let’s say, and have it fire on your smart phone, though this will be more reliable on platforms—Windows 10 Mobile, Android—where Microsoft is able to more deeply integrate the technology. In some cases—reminders again being a great example—you can actually access the feature completely by voice. And of course, Cortana can (sort of) “learn” about you by storing your interests, location, and other information and then triggering actions. And then there’s the silly stuff: Cortana can sing you a little song, tell jokes, and otherwise waste time on one-off demos.
But Microsoft has much bigger plans for Cortana, of course. Which is important because the competition isn’t standing still. Siri, still reeling from the “Is that rain?” stupidity of that Zooey Deschanel commercial, has improved immeasurably. But Siri, and, sadly, Cortana, are both outdone today by Google Now—which has almost preternatural skills—and Alexa, which is to personal digital assistants as TiVo was to home DVR: Everyone who owns an Alexa device, like Echo, loves it and recommends it to others.
At Build 2016, we learned that Cortana will proliferate across the Microsoft ecosystem—for example, in Skype, as a bot—and, as important, through third party apps and services thanks to new extensibility features. We saw the first bits of that extensibility functionality over the past several months, with Uber, LinkedIn and Microsoft Health integration, but the floodgates will open in 2016, so that any developer to can build on top of Cortana. And we saw that Cortana integration with Microsoft’s own platforms will only deepen, with lock screen-based access from the Windows 10 locks screen as an obvious example. The point, literally, is to make Cortana available from everywhere.
“This notion of Cortana being everywhere is really fundamental,” Cortana product lead Paul Bloom said during his Build 2016 presentation, Cortana: Learn How Cortana’s New Capabilities can Proactively Drive User Engagement with Your Apps. “In fact, Cortana has been architected to be in the cloud, a service-driven product. Wherever users want her, Cortana can show up.”
This is Cortana as a user experience. With your phone, PC, tablet, Xbox, or whatever actively listening—or because you touched off an explicit request—Cortana can use natural language capabilities to understand a query—about the weather or time, your schedule, a reminder, whatever—and then process it using a host of back-end web services—Bing, Azure machine learning, third party services like Uber or whatever—and then respond back to you in real time, again using natural language. And these interactions can be literal conversations, where Cortana prompts you for more information, or to clarify a point, and can occur over repeated steps.
But Cortana isn’t just reactive. It can also proactively prompt you because it knows about you, your schedule, and the things you want to be reminded about. (These features are improving in the Windows 10 Anniversary update.) It knows where you are, using location services, and can prompt you based on that information. Cortana’s ability to close the loop on task completion will only improve over time, and as this assistant is made available in far more places throughout your life.
And with that, Cortana will go from a Windows phone-based curiosity to a cross-platform wonder, and one that will hopefully fulfill Microsoft’s promise of making Cortana truly pervasive. There’s only one end-game for this technology: Cortana must be everywhere.