It’s been a decade since the iPhone launched, beginning the race to win at the mobile web. Somewhere around 2012 the big internet and computing companies looked up and realized that the next wave of disruption was heading their way with the smart home. In response, we saw Google spend an insane amount of money buying Nest in 2013, Apple launch but fail to deliver HomeKit in mid-2014 and Amazon debut the Echo at the end of 2014.
Meanwhile, Microsoft was still transitioning to a cloud-based services model and failing to win over mobile users with Windows Phone. IBM was doubling down with its Watson platform and Facebook was proving that it could sell ads on mobile phones, buying WhatsApp and also investing in AI. It also started research into internet access that may prove prescient as net neutrality rules fade away.
Why the history lesson? Because after years of positioning, it has become clear how the next consumer computing paradigm will play out. The race this time isn’t around a computer or mobile OS, it’s around a digital assistant. Or more specifically, an AI that will travel on top of devices, specifically a phone, a home, a car and likely wearables.
Once you choose one, you’ll have Alexa, Siri, Google, Bixby or whatever platform in all corners of your life. It will assist you with mundane tasks, answering questions and maybe even driving your car. The device underneath won’t necessarily matter. Each assistant should eventually be something you can buy and enable in the form of software. And it’s possible that because the data gathered from your relationship with the digital assistant will be so valuable, you may not have to buy your smart helper. You may just download it.
Today this market is defined by the so-called smart speaker. Amazon is winning this race so far with estimated sales of roughly 20 million Echo devices through September of 2017 according to Consumer Intelligence Research Partners data. That same analyst firm estimates Google has sold 7 million Google Home speakers. Meanwhile, Apple delayed the launch of its HomePod smart speaker from December to early in 2018.
Microsoft has an assistant named Cortana which is embedded in Windows 10 and a third-party speaker made by Harman Kardon. Cortana is also available on iOS and Android through an app. Yet, this is basically a race between Google and Amazon. And while it won’t be a winner-take-all situation like Android and Apple have managed to divide the mobile world, only a few companies will win.
Any AI layer sitting on top of myriad devices that communicate using different radio standards and don’t have a universal device schema to define the device will require developers to build hooks into the AI. Even though Amazon and Google have provided skills for developers to easily hook into their platforms, developers aren’t going to want to support many platforms.
Which then leads us to think about what will make a great AI assistant. If I carry the mobile OS war analogy forward, what will be the means by which the AI platforms deliver the most innovation and value? With the mobile OS war, it was the App Store. Apple’s decision to let developers play on its phone and build startling new functionality helped drive adoption of the platform. Do you remember how awed you were when someone showed you the Star Walk app on the iPhone?
Today Amazon is winning the digital assistant market because it was first to hit the market with a form factor that’s accessible and useful, playing music and handling digital tasks like setting a timer or checking your calendar on command. If you think of a digital assistant whose expertise is in affecting change in the real world, either by taking care of household tasks like changing dials on thermostats or channels on the television and restocking the pantry, then Alexa is going to win.
No tech firm can compete with the logistics and distribution system Amazon has, and the smart home first-mover advantage helps tremendously.
But if we view a digital assistant less as something to fetch and carry for us and more as a butler who can anticipate our needs and smooth our lives ahead of us, then Google is going to win. Thanks to its massive knowledge graph, Google can already answer more questions, more fully than Alexa, and Google is adding smart home features that mimic Alexa at her best and even offer new functionality based on context.
For example, in Night mode, the Google Home can determine that people are likely to be asleep and will automatically lower the volume when it responds to a question. If your digital assistant needs to be a butler, context and deriving intent from that context will be king. There Google will win.
Both Google and Amazon have understood a huge lesson from the mobile OS war that will be essential in building digital assistants. They need to play with a large ecosystem of developers, device makers and service providers.
As for Apple, it’s pretty clear that it is following along where it sees the platform war going, but it so far has proven difficult to work with by companies trying to integrate with the HomeKit platform, thus alienating its ecosystem. Siri, its assistant, isn’t very good at understanding what people want to do or even the questions they ask, making it frustrating by comparison to use for basic questions and for completing home tasks.
With the launch of the HomePod, it feels like Apple missed the overarching strategic play in the future of computing. Instead of intelligence, it’s emphasizing the sound quality of the HomePod. Instead of making the device in a variety of form factors so people can strew them around the home, it’s putting out a single, expensive home for its speaker. In a battle for the soul of our future computer, Apple is emphasizing the form factor and audio output.
The real question is what we want from this new digital soul. Do we want it to act as our agent in the real world as Alexa does, or to anticipate our needs and then change the world around us to fit them as Google is trying to do?