Skip to Content

With Personal Data, Predictive Apps Stay a Step Ahead

Apps that proactively help people with their lives represent a significant departure from earlier approaches to software.

A new type of mobile app is departing from a long-standing practice in computing. Typically, computers have just dumbly waited for their human operators to ask for help. But now applications based on machine learning software can speak up with timely information even without being directly asked for it. They might automatically pull up a boarding pass for your flight just as you arrive at the airport, or tell you that current traffic conditions require you to leave for your next meeting within 10 minutes.

Promptomatons: Phone software mines personal data to cue up reminders and weather reports.

The highest-profile of these apps is Google Now, which is a feature of the latest version of the Android mobile operating system and was recently added to the Google search app for the iPhone. Google Now is trained to predict when a person is about to take certain actions and offer help accordingly. It can also learn about an individual to fine-tune the assistance it offers.

Google Now’s algorithms use the data in a person’s Google e-mail and calendar accounts and Web searches. The app learns where you live and work and when you commute so that it can offer a virtual index card showing traffic or transit information. Other cards offer boarding passes and other handy information at appropriate times (see “Google’s Answer to Siri Thinks Ahead”).

Bill Ferrell, founder and CEO of Osito, a company with an iPhone app that offers similar functions, calls this idea “predictive intelligence.” Osito’s system foretells a person’s actions and needs from location, e-mail, and calendar data and uses those predictions to go beyond offering just information or advice. It also presents ways for a person to take action. A flight reminder will include a button to summon a cab, for example.

Now that the first generation of this type of app has been well received, engineers at Google, Osito, and elsewhere seek to wring more insights from the data they collect about their users. Osito’s engineers are working to learn more from a person’s past location traces to refine predictions of future activity, says Ferrell. Google Now recently began showing the weather in places it believes you’re headed to soon. It can also notify you of nearby properties for sale if you have recently done a Web search suggesting you’re looking for a new home.

Machine learning experts at Grokr, a predictive app for the iPhone, have found they can divine the ethnicity, gender, and age of their users to a high degree of accuracy, says CEO Srivats Sampath. “That can help us predict places you might like to go better,” he says. The information will be used to fine-tune the recommendations Grokr offers for restaurants and music events.

These apps benefit from improved data mining techniques, but they’re also succeeding partly because of how they are presented to users. They are not cast as artificial butlers, a staple of science fiction that Apple tried to mimic with the voice-operated app Siri in 2010. Instead, Apps like Google Now are intentionally made without personality and don’t pretend to be people.

That plays to the strengths of today’s artificial-intelligence technology, says Mike Volpi, a partner with venture capitalists Index Ventures, which invested in a predictive iPhone app called Donna. “An assistant probably is one of the most tough use cases, because you set up the expectation it will be human-level,” says Volpi. Apple’s assistant has not become a core part of many iPhone users’ lives, he says, because software cannot recognize speech accurately enough. Apple may have exacerbated this problem by giving its app the capacity for witty repartee (see “Social Intelligence”) and running TV ads in which Siri appears to act with almost humanlike intelligence.

Hilary Mason, chief data scientist at the Web company, has mixed reviews for Google Now. She finds that it’s often serving up unnecessary information: for example, she says, she doesn’t need to be told that she is near a Staples office supply store, which is true in many parts of Manhattan, or be given a bus schedule every time she passes a bus stop. “It’s not quite tuned to what matters to me,” she says.

But still, it represents a milestone in computing, she adds: “Google Now is kind of a sucky product, but I use it anyway. It’s important because it’s the first time Google has taken all they know about us to make a product that makes our lives better.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.