Intelligent Machines
Watching Viewers
Making television smarter requires understanding why it is our favorite gadget.
Do you want a Web browser on your TV? If history is any indication, your answer is probably a resounding no. We don’t blame you.
In the past few decades, the technology industry has labored under the delusion that consumers would love their TV sets to behave like computers. Many tombstones now stand in place of devices built by very smart people, with incredibly smart technology inside, that made no impact. Our own company, Intel, had multiple failed attempts.
Even today, with more consumer electronics to choose from than ever before, the TV remains the most-used electronic device in the home. It is often at the center of our living rooms and bedrooms. It is where we go to relax and to gather with friends and family. For many, watching TV defines being at home.
In 2005, our group at Intel took a fresh approach. Instead of trying to build a television with PC-like features, we asked people how their TV experience could be improved. Instead of starting with assumptions about how TV had to change, we began by finding out what people loved about it.
Our ethnographers visited India, Japan, the U.K., and the United States, sometimes watching people watch TV, sometimes watching with them. We wanted to understand how people lived with their TVs and the other people around them. The results directly informed the design of the processors at the heart of new devices like those running Google’s TV software and D-link’s Boxee Box (see “Searching for the Future of Television”).
The first thing we learned was that people love TV just as it is. They love their shows and they love its simplicity. TV is always there and doesn’t ask too much of them. A story they care about is always just one button away. When we asked people what they would want from a TV with computing power, they didn’t talk about computing. They talked about TV. Their top three answers were that they want access to their regular broadcast TV, want access to broadcasts they have missed, and want to know what shows their friends recommend.
Delivering on all three requests does require computing. Giving viewers the shows they missed takes a combination of DVR and Web services. Telling them what friends enjoy is a mix of social networking and automatic recommendations. But it doesn’t require building a TV that behaves and feels like a computer. Recognizing that, using social science to inform computer science, has given us a new generation of smart TV devices like nothing that has come before. Let’s hope they fare better than their predecessors.
Genevieve Bell is director of interaction and experience research at Intel; Brian David Johnson is a futurist and director of future casting.