At the Build 2016 conference, which kicked off on March 30, developers got a sneak peak at Microsoft's (NASDAQ:MSFT) vision of all pervasive artificial intelligence (AI). CEO Satya Nadella calls this "conversations as a platform," in which human language is the preferred user interface. If Microsoft achieves its vision, it will be a breakthrough that will change the way we interact with our computing devices.
One of the staples of science fiction has been artificial intelligence smart enough to have a natural conversation with. Such an advanced capability has been elusive, but in recent years, cloud based speech recognition services from Microsoft, Google (NASDAQ:GOOG) (NASDAQ:GOOGL), and Apple (NASDAQ:AAPL) have gotten closer to the fictional dream.
Critical to the workings of these systems is that they connect to large cloud based AI systems that can to a greater or lesser extent discern context information in order to provide an intelligent and useful response. Such systems are far from perfect, and quite often produce artificial stupidity.
Thus, Microsoft's goal of making speech recognition (combined with various AI based services) the underpinning of a new computing user interface might seem overly ambitious. Thus far, Google, Apple, and Microsoft have restricted their AI digital assistants to relatively modest tasks: basic speech recognition, a front end for search, and some general purpose commands.
Despite their limitations, the personal digital assistants such as Google's mobile app, and the Cortana and Siri digital assistants represent an exciting frontier in application development. These are apps that provide connectivity to specialized cloud based AIs. While the apps perform critical user interface functionality, the real computing horsepower resides in a server farm that could be hundreds of miles away from the user.
These digital assistant apps have been "first party" apps developed by the companies that develop the operating systems the apps run on. For the most part, the cloud services that back the speech recognition capability of the apps haven't been available for third party developers. This is the critical thing that Microsoft wants to change.
Positioned for Change
It may be that Microsoft is better positioned to make this change than either Google or Apple. Microsoft already has a very well developed cloud services infrastructure. What this means fundamentally is that developers can create programs that run on Microsoft's cloud servers. These programs can be either Web apps accessible via a standard browser, or they can be accessed by mobile apps.
While all three companies utilize cloud based apps, Microsoft probably has the best set of tools for developers to create their own cloud apps. In contrast, Apple offers developers almost no capability in this regard, even though Apple offers consumers an extensive suite of cloud based services. Google Cloud Platform offers services comparable to Microsoft, though perhaps not as mature.
What Microsoft is doing to make language recognition into a pervasive computing interface is offering developers the features and capabilities of Cortana to use in their own applications. Microsoft calls this the Cortana Intelligence Suite. In effect, developers will be able to build into their apps mini Cortanas, generically referred to as bots. These bots are hosted in the cloud and can be endowed with a tailored set of capabilities to suit the needs of the app.
Bot capabilities are not restricted to speech recognition. Microsoft plans to make available a broad array of what it calls "cognitive APIs" that provide various artificial intelligence functions beyond speech recognition. These include image analysis, knowledge and learning, and search.
In contrast with what Microsoft is offering, the capabilities currently resident in Android and iOS are much more limited. Android has some speech recognition and search APIs that developers can use, but nothing close to Microsoft's cognitive APIs. Apple offers even less for iOS: no speech recognition APIs and only a very limited "Spotlight Search" capability.
What we were shown at Build was all very carefully labeled as preliminary. Still, there were no embarrassing moments where things didn't work at all. The scope of the demos was limited however.
I don't think that Microsoft means to replace the Win10 UI any time soon, but it does want to give developers the chance to explore and work with the new APIs. As developers get experience and the tools themselves mature, I expect that we'll see more widespread use of bots and language recognition.
And widespread means not confined to Windows devices. Since the APIs are fundamentally cloud computing based, Microsoft is making them available for use with either Android or iOS.
As I pointed out in my recent article on Amazon's (NASDAQ:AMZN) Echo speech recognition device for the home, we're seeing a very powerful confluence of a mobile (or at least compact) device that provides the hardware/software front end and cloud computing based AI on the back end.
These cloud based AIs are more than just speech recognition engines. They are capable of learning from the behavior of the user in order to anticipate the user's needs and wishes. They have broad contextual awareness that makes their speech recognition more effective.
The companies that are out in front in this area are the companies with the best developed set of tools for cloud based apps, Amazon and Microsoft. Google as a cloud services company is not far behind, although their tools for developers may not be as mature. Apple, when it comes to developer tools for the cloud is just woefully inadequate.
For some time, we've been hearing Nadella talk about the "mobile first, cloud first world." It always seemed a little vague and tenuous to me. How this was supposed to translate into a business advantage for Microsoft wasn't clear. With the Cortana Intelligence Suite, the linkage between mobile and cloud is now well defined.
There may be stumbles along the way. Unleashing developers to create their own bots is fraught with a certain peril. Nevertheless, it's a vision of the future, which is very compelling, and almost certainly the right direction.
At times, especially over the course of Win10 development, Microsoft has seemed to meander, directionless. That's no longer the case. If Microsoft can execute on this strategy, it could translate into a profound impact on personal computing, as well as a profound competitive advantage.
Are Google and Apple as far behind as they appear? Perhaps not. Both have developer conferences of their own coming up in the next few months. We may see movement in a similar direction by both companies.
Disclosure: I am/we are long AAPL.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.