Go to the Globe and Mail homepage

Jump to main navigationJump to main content

Nokia Oyj Lumia 900 smart phones sit on display during the 2012 International Consumer Electronics Show in Las Vegas, Nev., on Jan. 12, 2012. (Daniel Acker/Bloomberg)
Nokia Oyj Lumia 900 smart phones sit on display during the 2012 International Consumer Electronics Show in Las Vegas, Nev., on Jan. 12, 2012. (Daniel Acker/Bloomberg)

Earlier Q&A

Are our smart devices becoming too smart? Add to ...

What if your smartphone – or laptop, or remote control – could anticipate what you want and could intuitively respond to your needs?

It’s called context-aware computing. Small sensors gather information about a user’s location, preferences and environment, and your device learns about you.

Your phone could track your diet and exercise and help you improve your health. Your TV remote could offer program recommendations based on who is holding it. Search engines could display more accurate results based on a person’s browsing history.

Context-aware computing has been gaining steam as costs fall and technology improves. But it also raises serious questions about privacy that experts say must be addressed as these systems come to our homes and offices.

Ted Selker, one of the foremost experts on context-aware computing, joined us to talk about the possibilities – and the risks – of submitting personal information into such systems. Mr. Selker is the associate director of mobility research at Carnegie Mellon University, Silicon Valley campus. From 1999 to 2008 he was an associate professor and the director of the Context Aware Computing Lab at MIT.

This is an edited version of the live discussion. To see it as it appeared originally, just go to the frame at the bottom of the article.

Ross, from Waterloo: With the advent of context-aware computing coming into vehicles, how do you see a vehicle reconciling information and preferences of the driver vs. passengers while a vehicle is in motion?

Ted Selker: Indeed, the context does resolve many things; we don't want the music to drown out the firetruck siren, for example.

The goal of understanding the situation is so interesting in a car. I built a concept car in which we found that giving feedback - of any sort - too often was distracting.

By giving only a little positive feedback and much, much less negative, car coach can actually improve your attention to problems.

Dave M, Globe and Mail: How fast do you think the tech industry will move on context-aware computing? Do they see money here?

Ted Selker: The automatic grocery store door opener that was first built in 1962, was wonderful in this way. It expects that a person leaving a grocery store has a basket in front of them and needs help opening the door; people loved and love it.

There is tremendous money in solving peoples problems without being inappropriate. Context is one great way to avoid being inappropriate.

Context sensing is creeping into everything. At first the phones require us to turn them off when in certain situations, soon, as my new Bluetooth, they will know when to be turned on. My Bluetooth, connects to my phone ... only ... when on my ear, saving it from answering calls from my backpack. That is contextually aware computing.

Chris: Could a smart phone on the Costa ship that just sank have helped passengers make better decisions about what they should have immediately done with no information from the captain and crew?

Ted Selker: Until they got wet, the phones could have told people that are on deck where they are, inside the ship, the Wifi could have helped people know where they are. People have been saved in avalanches using cellphones as beacons...

Cellphones could have been useful in many other ways, starting with being used as flashlights, communicating information before it's too late, and even being part of some broadcast messaging if things were set up for that.

gmpartsman: Intelligence implies awareness or sentience. How could artificial intelligence even be measured? What appears to be intelligence is merely clever algorithms.

Ted Selker: Intelligence is measured behaviorally by the Turing test; can a person distinguish a person from a bot? The first time that a bot passed the Turing test was with a system called parry. Parry was an AI system that simulated a paraniod schizophrenic in 1973 at Stanford.

Jeff: I love the idea of this technology so much but it certainly straddles "the line" in terms of how much is too much? For example, I love using Gmail but the ads Google presents to me at the top of my inbox give me the creeps. How big a barrier is privacy for growth of this technology?

Report Typo/Error
Single page

Follow us on Twitter: @GlobeTechnology

Next story




Most popular videos »


More from The Globe and Mail

Most popular