Imagine a world where you don’t need to search for something on your phone, because your phone finds it and tells you first. Maybe even before you realize you want to search for anything.
It’s a world Google expects soon, one that poses an existential quandary for the search giant: What happens if the one doing the telling isn’t Google?
That’s one of the chief concerns of Aparna Chennapragada, the Google executive who has run Google Now, its AI-driven smart personal assistant, since the start of this year. Her aim is to have Google Now surface stuff that’s highly relevant to users — local weather, favorite sports scores, calendar appointments — but also in ways that can be predictive, prescient. She wants it to answer questions people may not even have thought to ask.
“It’s not just a shortcut,” she told Re/code. “There’s information that you may not know that you need.”
Search is evolving swiftly, moving to the realm of apps and intelligent voice-based exchanges — all of which currently have little room for Google’s core business of selling search ads. Google has labored to amend that. The company has released a series of features designed to magnify and entrench its presence within mobile search, which now dominates the desktop. Chief among them is a push to let Google Now burrow deeper into our digital lives.
Google Now launched in 2011, one of the two priorities of returning CEO Larry Page. The other was social, which has had a turbulent journey since. But Google Now, an intelligence layer on Android and the Google app, is beginning to spring to life.
In January, the service started integrating with popular apps, pushing notifications customized around personal data. This summer the product team is rolling out Now on Tap, a new Android feature that weaves Now into apps (and, likely soon, mobile websites). It’s search without the search bar — and, sometimes, without the query.
For Chennapragada, the product mimics Google’s approach to the Web. “The nature of information has changed, of course,” she said. “There are some building blocks there that are the same.” Among them is indexing context based on merit. Her aim isn’t to replace search, but to remove any impediments to speedy retrieval of information. “Can we get it to you in the most efficient way?”
“I want a smartphone to be smarter,” she added, repeating a favorite mantra, “and do a lot more heavy lifting.”
That heavy lifting demands significant machine-learning resources. The fact that Google has hung much of its mobile strategy on Now’s success signals the notable ascendance of its manager, particularly in a company chock-full of veteran executives.
Three “bets” for the personal assistant
Chennapragada cut her teeth at Akamai Technologies, a cloud services provider, during the dot-com boom. She came to Google in 2008 as a product manager with YouTube before joining the search team two years later.
Yet search, she added, was always her driving interest. “One way or another, I was in the forefront of information and information discovery,” she said.
She moved to Now in the summer of 2013, then replaced the product’s initial chief, Johanna Wright, who is currently an Android VP, at the start of 2015. Chennapragada has a reputation as a more commandeering product lead than Wright; Googlers want to work for her.
By her telling, Now is evolving from its initial version, which she described as “very intuition and anecdotal-driven.” If you woke up every morning and checked the traffic for your commute, Now would start feeding you notifications with the travel time once you hit your alarm.
The personal assistant is maturing. To do so, it’s marrying Google’s vast data architecture with the company’s two-year effort to index and archive apps as it did with the Web.
Chennapragada spelled out the three-pronged direction of the product — what she called the “bets” her team is taking. The first bet was embedding Now with Google’s full “Knowledge Graph” — the billions-thick Web of people, places and things and their many interconnections.
The second is context. Now groks both the user’s location and the myriad of signals from others in the same spot. If you enter a mall, Now will tailor cards to what people in that mall typically ask for. “Both your feet are at the mall. You shouldn’t have to spell it out,” Chennapragada said. “Why should I futz with the phone and wade through 15 screens?”
And this is where the third benchmark for Now comes in: Tying that context to the apps on your phone, or ones you have yet to download. In two years, Google has indexed some 50 billion links within apps. In April, it began listing install links to apps deemed relevant in search. Indexed apps will be included in Now on Tap when it arrives in the latest Android version this fall.
Now, in many ways, aspires to be a smarter version of the app store, linking users to apps whenever they signal the need for them. If it takes, the product would significantly restructure the entire mobile ecosystem, with Now as the gatekeeper — much the way Google was for the open Web. And that could spur a whole new fight for primacy on our phones.
“It’s one of the key pieces of brokering that Google can do,” Chennapragada added. “How do you match the right piece of content? That’s the third piece that we’re building out.”
Looming Cupertino threat
It’s a battle that’s just getting under way. Microsoft’s Cortana and leaner startups like SoundHound and Sherpa have similar voice-search offerings. Others, such as URX and Quixey, are building out app deep-linking capabilities.
But only one force can rival Google’s ability to link personalization with the device and the content therein: Apple.
Like Google, Apple is indexing apps in its world. At its developer conference last month, the Cupertino company unveiled a pair of features — a more advanced Spotlight Search and a “proactive assistant” technology akin to Google Now — that hinted at a broader ambition. Several sources say that Apple is laying the groundwork for a system similar to Now on Tap: Search for something in Spotlight or with Siri, which uses Bing results, and soon the results will simply take you directly into a relevant app.
It’s a process that would circumvent Google entirely.
It would hurt Google significantly, too. The company will bag $5.6 billion from its iOS Safari users this year, according to calculations from UBS. Its search deal with Apple is set to expire in 2015.
Apple, unsurprisingly, did not comment on its search plans.
Chennapragada declined to comment on rival services as well. Multiple AI observers, however, give her team an edge for one reason: Google is far more advanced in intelligence. Apple has to make significant progress to match the machine learning capabilities that make Google Now and algorithmic search possible.
For her part, Chennapragada has an almost giddy response to the reality that the intelligence technology, until recently a pipe dream, is in fact possible.
“As a computer scientist, I’ve studied this thing and said, ‘Oh, that thing, it’s not going to happen for a while,’” she said. “But it’s happening.”
Correction: An earlier version of this story incorrectly stated that Chennapragada has run Now since 2013. She joined the Now team that year, but began leading it in 2015. We regret the error.
Additional reporting by Dawn Chmielewski.
This article originally appeared on Recode.net.