Editors' Rating
Published: 16 Nov 2007
Peter Cochrane, the former head of research for British Telecom, had a nice line about the future: 'I do worry about dying before my laptop is proud of me'.
He'd better have a long life. Imagine how frustrating we are for our machines. We do destructive things like plug them into unreliable power grids. We make stupid mistakes like delete files we want to keep and misaddress email. We design houses that are anything from unsafe to unnavigable for robot vacuum cleaners and then complain that the rugs aren't completely clean. We drive each other to homicide; imagine what we could to a really intelligent machine.
In The Design of Future Things, usability guru Donald A. Norman considers the problem of how human-machine interaction design should change and develop as machines get smarter. There is, he says, a fundamental gap between the way humans and machines think and behave. Humans are fuzzy, emotional and vague; computers are precise, logical and have no sense of humour. The gap will only continue to widen as computers get smarter and more pervasive, he argues.
Worse than that, the consequences of design errors are going to get worse. It's bad enough if your car releases its airbags without warning. But in one of the book's examples a car with automatic lane-keeping controls keeps its driver locked in the inside lane of a roundabout in a 14-hour barrage of heavy traffic in the other lanes.
A key issue in designing future systems is improving feedback. Norman suggests using more natural sounds and responses. So, for example, instead of building vibrating strips into the road surface to prevent cars from straying out of their lanes, build the vibration and road feel into the car. Or add noise to what are now silent cars, and make it change in pitch and volume as the speed goes up. Today's silent systems are more dangerous precisely because users can't be sure what's actually happening.
In what may seem a shock reversal to those who have followed his lengthy career in usability issues, Norman suggests that humans will have to adapt to machines. He is not, he hastens to add, telling engineers it's OK to go back to bad designs that provoke human error. But he is acknowledging that humans are adaptable whereas machines are not, and that many things in the way we live have been changed to accommodate the needs of machines. We paved the roads for cars, designed spaces in our homes to fit appliances like washing machines and dishwashers, and we learn to type so our computers can accept input. Norman gives an idea of what this reversal might mean in practice in the final chapter, in which he imagines the machine's eye view of the history of machines, revealed through a conversation with a person.
You don't read Norman's books to discover how to design a web page today. You read them when you want to understand how to think about design in a broad sense.
Digg
Slashdot
Del.ici.ous
Stumble
Reddit
