Contextual Computing: When Your Devices Start Reading Your Mind

We are starting to cross that invisible line between anticipation and prediction.

You used to hope that you could rub an old lamp - like Aladdin or 'I Dream of Jeannie' - and a mysterious dude in pajamas with a strange accent (or the made-for-tv sitcom version) would then ask what you wanted.

Now? You just have to run your finger across the surface of your phone. Or the powerful computers in New Jersey that are programmed to trade your securities or your home thermostat or some other technological manifestation of your physical life will do it for you. It's pretty cool, until it gets creepy. But then our creep factor's boundaries are moving. It's just a question of at what point we'll decide enough's enough - though it is not at all certain we'll know - or that if we do, the devices will care. JL

Molly Wood comments in the New York Times:

Signals start to be predictive. When that happens, your phone can start anticipating your needs, interests and habits and give you relevant information, apps and even coupons and ads.

Like many people in this modern world, I struggle with the tension between the conveniences offered by the latest technology and the loss of privacy that comes with them.

Nowhere is this devil’s bargain more evident than in the blossoming field of so-called contextual computing.

When I picked up my phone earlier this week, it told me — without a single tap on my part — that my estimated commute time was 51 minutes and that I had a lunch scheduled with a friend. The friend’s Facebook photo showed up next to the appointment.

The phone also showed my other appointments that day and a customized feed of news and weather, and it gave me the flight status of an approaching trip.

Sadly, it did not bring me coffee.

My phone is trying to anticipate my needs based on what it knows about me — the context of my life. And what it knows seems like almost everything.

Your phone is constantly gathering what app developers call signals. These could be your commuting habits, which the phone can glean from its internal GPS, often within a few feet. Your phone could also gather your meetings, your future trips, your friends and family, your favorite sports team, the type of news you usually read and even things like your heart rate.

Things really get interesting when the apps that gather these signals start to be predictive. When that happens, your phone can start anticipating your needs, interests and habits and give you relevant information, apps and even coupons and ads.

These context-based devices and apps are already everywhere. The Nest thermostat, for example, is a context-aware device. It programs itself according to your schedule and can even tell whether someone is in the room. A coming speaker called the Aether Cone will analyze your listening habits to serve up the right music at the right time of day or week.

But right now, your phone is where the most interesting contextual computing is happening.

Google Now is the best-known app for predicting what you will want to do. The app generated my morning traffic report, which pops up as a notification around the time I usually leave for work, or when traffic is particularly bad (which it is most days). It also shows sports teams, travel information, tech news and previous searches, as well as nearby locations or events.

The new personal digital assistant built into Windows Phone, called Cortana, does similar tricks.

In theory, this can put the gadgets that know the most about us to work helping us live better lives. In practice, the context is there, but the predictions are not quite — yet. And the privacy trade-offs may be too much for some people to stomach.

Apps that know who and where you are can be undeniably useful, said Robert Scoble, a co-author of a book about contextual technology, “Age of Context.”

If you give an app access to your email so it has your travel itinerary, for example, it can alert you when there’s a problem with a flight, show you alternate flights and let you tap one or two times to rebook — especially if the credit card information is stored in the app. That situation played out for Mr. Scoble recently, he said, and “that night, I won at life.”

A generation of mostly brand-new apps is trying to add new conveniences by scanning your signals.

EverythingMe, available only for Android, does something Apple’s iOS will not allow: It takes over the home screen and then rearranges and recategorizes your apps. This type of app is called a launcher. A similar app is Aviate, currently available on an invitation-only basis.

Of the two, EverythingMe is more powerful and it required virtually no setup before it started learning from my behavior. It categorizes apps into obvious folders, like Social, Games, News, Weather and Music. Those folders also instantly generate related content from popular websites, as does any folder I create and give an obvious name, like Kids.

That’s slick, but not contextual or predictive — it’s just a nice way to organize your phone.

What EverythingMe does well is serve up a rotating collection of four home screen apps that change according to the time of day, my location and how often I use particular apps. In the morning, I get a custom icon, My Day, which shows calendar events and apps I use often in the morning, like Twitter. I might also see Traffic, Calendar and Hangouts, which I use for chatting with co-workers.

But once I get to work I might see just email, Hangouts, Calendar and something like Uber, which I use to get around San Francisco. It’s simple, purely anticipatory and generally accurate.

On iOS, developers can do a bit less: An app is restricted from taking over the home screen, or from seeing too much about what you do in other apps on the phone. So predictive apps for the iPhone tend to focus mainly on your calendar or contacts.

On iOS, Mynd does context-aware calendaring that shows photos of the people you are meeting with, which it gathers from Twitter and Facebook. It also grabs locations from calendar invitations and tells you how long it will take to get to meetings.

EasilyDo on iOS and Android handles calendar appointments, travel times and birthdays. It can grab travel plans and boarding passes from email, and will manage contacts. Of the two, EasilyDo has more potential, but it is cluttered and keeps pressuring me to upgrade to the paid version or download free songs.

Both require access to a lot of information about you, like all your calendar data, social profiles and email accounts, which feels invasive. It would feel less so, though, if either truly delivered on the promise of what contextually aware devices and apps could do.

But to be fair, no app has yet fully delivered on the promise. Listening to starry-eyed developers talk about the possibilities remains like a visit to the future. Actually using the apps brings you back to earth with a clunk.

“Contextual is a whole world,” said Ami Ben David, co-founder of the company EverythingMe. “It’s going to change computing as we know it, in my view. We’re going to start looking at computers as being smart, as having infinite computing power and infinite access to databases, and therefore able to talk to us and give us what we want.”

That sounds great! But that’s not quite what EverythingMe is doing yet.

Certainly in the future, phones will most likely be talking to wearables that are monitoring your movements, as well as things like your thermostats and even your cars. As the signals grow louder and more frequent, the context becomes more refined and accurate.

Location is the most powerful element of contextual tech, from what I can tell. But it’s also the element that makes me concerned about my privacy. When an app is already reading your calendar, reading your email, learning where you live and work, where you’re traveling, what you are interested in, is there anything it doesn’t know?

“I talk about this as an all-in or all-out world and a lot of people are in the middle of trying to figure out what they’re going to do,” Mr. Scoble said. “Yeah, it is freaky. Yeah, it requires you to give access to things that are somewhat scary to give access to, but the utility is going to be so deep that the people who don’t, really are going to miss out on life.”

Count me as one of the people in the middle. I’m still clinging to a semblance of privacy, even though it is probably futile. And every time I, like Mr. Scoble, win at life thanks to an app that can deliver exactly what I need, when I need it, I take another small step toward the dark side.