Microsoft is joining Google and Amazon in the race for your home. This week, at an event in China, the venerable tech giant trumpeted the arrival of Project Evo, a sweeping plan to build hardware devices that work a lot like Google Home or the Amazon Echo.
But this race is much bigger than some gadgets that sit on your coffee table. It’s a race not only for the hearts and minds of consumers, but for a world of business customers, too. The prize is more than just the best home digital assistant. The biggest spoils go to the company that rides its assistant to artificial brains that are far smarter—and creates a market for using these brains to do just about anything.
Giant Corporations Are Hoarding the World’s AI Talent
Artificial Intelligence Just Broke Steve Jobs’ Wall of Secrecy
Google, Amazon, and Microsoft are all racing to build systems that recognize and truly understand natural language—how you and I talk. If they can put this technology not only in your living room, but in your pocket and elsewhere, they can become the hub of everything you do online. Google wants to retain its central role in your life. Amazon hopes to move well beyond online shopping. And Microsoft doesn’t want to be left out.
But as these companies develop services for speech recognition and natural language understanding, they’re also using many of the same underlying technologies—loosely called deep learning—to build all sorts of artificial intelligence. They will slip this AI into their own apps, and offer it to a world of other companies via cloud computing services so that these companies can build AI into their own apps. In the years to come, AI cloud services may wind up as the biggest business for these three tech giants.
Multiple Futures
Project Evo is a way of bootstrapping all these opportunities. It’s a way of showcasing Microsoft’s AI talents and pointing coders and companies toward the AI cloud services of Microsoft Azure. More importantly, it’s way of gathering data that can advance the development of future AI technologies. How you talk to Evo will inform how other AI services will operate. Deep neural networks learn by analyzing vast amount of data, and to reach true natural language understanding, they need far more data than a company like Microsoft now has.
“This is just the beginning,” Chris Stone, the engineering director at Acquia, which helps businesses build online services. “These companies can learn what it takes to build a conversational UI and then implement that across everything else they do. That is what they are doing.”
This is certanly what Amazon is doing. The Echo’s digital assistant is called Alexa. And last week, Amazon unveiled the cloud computing service Lex, which lets anyone build conversational bots using the technology underpinning Alexa. The more services coders build for Alexa, the more consumers will find devices like the Echo useful. At the same time, Amazon wants to nudge coders and companies toward Amazon Web Services, its sweeping collection of cloud services.
Microsoft, meanwhile, missed out on the smartphone market and social networking. Its Bing search engine trails Google by a wide margin. But the company is strongly positioned to benefit from the virtuous circle formed by digital assistants, AI, and the cloud. It was the first big player to move speech recognition toward deep neural networks, a form of machine learning that has rapidly advanced the field in recent years. This fall, Microsoft released a paper claiming that its speech recognition has reached parity with humans. Though the research comes with caveats, it represents the current state of the art, at least among publicly available work.
These technologies already have played into its voice-driven Cortana smartphone assistant and its forays into text-powered chatbots. Granted, Cortana isn’t nearly as widely used as Google’s voice-driven assistant or Apple’s Siri. And its chatbot is most famous for producing a Twitter bot that ended up spewing racism. But as Google has shown, the technologies underpinning these services play right into conversational devices for the coffee table.
Yes, Project Evo trails Echo and Home, placing Microsoft in a familiar spot. Amazon was the first to market, and developers are building services for the Echo. Just this week, Google announced it will tie Home into third party services. But Microsoft is better positioned to compete here than it has been in other markets. It is competitive in AI as well as in the cloud. And it has a voice-powered platform that can feed those services both customers and data.
Amazon and Google also have the elements needed to benefit from this virtuous circle. The living room, AI, and the cloud will define the field of competition in the coming years. Success in any one of the them will drive success in the other two. It’s a battle for multiple futures.