Friendly Surveillance by Cell Phone

Intel shows off a technology concept that would show what friends are doing, and where, in real time.

Tom Simonitearchive page

June 30, 2010

We already help others spy us by manually updating various social networks with details of our activities. But even Twitter isn’t real time–you can’t know what a friend is doing between tweets. The images in this blog are a preview of a future where we can instead glance, in real real time, into our friends’ lives.

These “shadow avatars,” created by Wendy March and her colleague Kieran Del Pasqua, both at Intel’s labs in Hillsdale, Oregon, are ways to represent live sensor data from a cell phone. The image at the top of this post, for example, shows that Kieran is listening to music, while his avatar strides in space to show that he is also walking. A grid view allows you to instantly surveille all of your friends. And the image of Wendy’s avatar, at the bottom of the page, indicates that she is in a crowded place, and is browsing the web on her phone.

The Intel group has been doing ethnographic research to inform their design, following people around and noting down their actions, to help users think about what kind of information they want to share.

“Knowing that people may be glancing at your avatar is very different to your pushing out updates as you might with Facebook or Twitter,” says March. It’s the difference between broadcasting what you want people to know you are thinking or doing and allowing them to know at any moment what you’re actually up to.

The Intel project may only be a design concept for now, but all the kinds of data shown in the demo are accessible, For example, by tapping into a handset’s microphone it is possible to recognize the hubbub of a crowd, or music. “Combining sound input and GPS could show that you are in a car, or a train, for example,” says March, “and we can of course know what you are doing on the phone at any time.”

As for privacy, March points out that basic activity recognition could help ensure only true friends know what you’re up to. “As well as the algorithmic recognition of what you are doing, there’s also a level of human intuition,” she explains. Even minimal information about a close friend is enough for us to infer with reasonable accuracy exactly what she is doing, because we know so much about her routines and habits. “In some ways that extra layer is also a privacy shield,” says March.

Keep Reading

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.