Google announced something for everyone yesterday at its 10th annual I/O developer conference. There were more details of a new version of Android; new messaging and video-calling apps; a built-in new VR platform for Android; and a good-looking Amazon Echo-like smart speaker called Google Home.
There was even a cool new research project called Instant Apps that will let users run portions of apps from the web without installing them first.
But the biggest theme stressed by Google CEO Sundar Pichai and his lieutenants, over and over again throughout the two-hour keynote, was that Google is doubling down on artificial intelligence as the next great phase of computing. And they believe Google can do it better than anyone else.
While mobile devices are still immensely important to Google, I believe this year’s I/O marked a turning point from a Google which sees search and mobile as its north stars to a new Google which aims to win with artificial intelligence over the next decade. Pichai said early in the event that "it is truly the moment of mobile" and then spent most of the rest of it talking about human-like software that — while able to run on smartphones — is designed to be device-agnostic.
This year’s I/O seemed less centered on Android and more centered on what Pichai called the "pivotal moment" he believes personal computing has reached. It’s the moment when machine learning, speech recognition, and natural language processing — coupled with the immense trove of personal and aggregate data that Google has collected for years — can make computing more human.
Pichai peppered his remarks with bold claims about AI and Google. At various points in the presentation he said, "We believe the real test is whether humans can achieve a lot more with machine learning assisting them," and "It's not just enough to get them links, we really need to help them get things done," and "We want users to have an ongoing, two-way dialogue with Google."
He is, by nature, a genial and humble man, so he also threw in some caution saying, "We believe this is a long journey" and "We believe we are just getting started."
Of course, this isn’t new, and Google isn’t the only one to have the thought. Apple’s Siri voice-controlled assistant came out in 2011. Google itself has had the proactive Google Now, with voice commands, for years. Facebook and Microsoft are doubling down on automated chatbots and, right now, the hottest product in this area is Amazon’s Echo home device with its Alexa voice assistant.
But Google spent much of the keynote touting something called the Google Assistant, which isn’t a product, but a technology that will power many products. The company says the Assistant "will extend across multiple devices and many contexts."
For instance, it’s Google Assistant that will allow the Google Home gadget to turn off your lights, play wake-up music, warn you of a delayed flight, and change dinner reservations due to that delay.
It’s Google Assistant that will allow the new Allo messaging app to suggest intelligent, in-context comments based on text and images in a chat — for instance, correctly identifying a photo of linguini with clams, suggesting an Italian dinner, and making the proper restaurant reservation. And it’s Google Assistant that will let you have an Allo chat with Google directly with no humans involved.
It’s Google Assistant that will save you typing on your smartwatch by suggesting intelligent replies to texts received there based on their content, not a canned list.
It’s by no means certain that Google Assistant will consistently do all the hard things the company showed off in promotional videos at I/O — like conducting a conversation in which it remembers the topic of the last question asked and understands that the next comment or question often refers to the same thing.
And it’s also uncertain that, if AI proceeds to become as human-like and integral to life as Pichai seems to hope, that it will be a good thing for society. Or that AI will be allowed to proceed unchecked by government regulation.
It’s just as uncertain that people will want to use the company’s new vehicles for Google Assistant. Lots of people resist adding yet another messaging app when they already have WhatsApp, or Facebook Messenger, or iMessage. The same goes for converting FaceTime and Skype users to Duo, or the growing legion of Echo acolytes to Google Home.
Plus, Facebook, Amazon, Apple, and Microsoft are traveling this same road, and will continue to slug it out with Google.
But Google knows a lot more about each of us than any company save perhaps Facebook, and it makes the underlying platform for 85 percent of the world’s smartphones. So it has a fair chance to do very well at providing intelligent assistance at scale.
Google’s biggest problem will be reining in its own propensity for putting data mining over privacy. For instance, the Allo messaging app is only encrypted end to end in a way that Google can’t decrypt if you select a special "Incognito" mode. Apple’s iMessage and WhatsApp default to that level of encryption.
Speaking of Apple, it has to be worried about today’s Google announcements. Unlike the very good iMessage and FaceTime, which succeeded where Google had failed in the past, Allo and Duo will be cross-platform, meaning they will run on both iOS and Android, potentially drawing away users.
More importantly, Apple’s own efforts at AI are hampered by its admirable but costly focus on privacy as a first principle. It has limited itself to working with local info on the phone itself, while Google can use all that data it has scooped up from search and other cloud services.
But the big takeaway from I/O for me was that Google has laid down a huge bet on computers getting more and more human. It’ll be fascinating to see if that bet pays off, how competitors respond, and what the consequences are for society.