Tuesday, 29 April, 2003
Mobile computer input devices
About four years ago I gave a presentation on Online Help for Mobile Devices at a Windows Help conference in Dallas. During that presentation I pointed out that mobile computing lacked only a new type of human interface in order to become mainstream. I also predicted that this new interface would be invented and become popular in the next three to five years. With only one year to go, I'm losing hope. The only significant change in human interface over the past four years has been handwriting recognition software, most recently in the Tablet PC. Voice recognition isn't appreciably better than it was then, and alternate input devices are frowned upon because they require user retraining. There has been even less change in output devices, although that's understandable given that humans have only two ways to gather large amounts of information: sight and sound. And we are much less able to understand and retain spoken information than textual and graphic information. Whether the inability to learn well through hearing (as opposed to reading) is genetic or just a lack of training is still unclear, although it's quite clear that seeing a diagram is much more effective than reading or hearing a description of the diagram.
At the time I gave the talk, I was betting that some kind of hand-motion input device would replace the cumbersome keyboard and the ineffective stylus. For output, I was sure that affordable and comfortable ear buds and monocular display devices (that fit like a pair of glasses) would be developed and become popular. The ear bud is a no-brainer, but the display devices just aren't there.
What's it all mean? Two things. First, it's dangerous to make predictions in this business. Second, that mobile computing will continue to be a niche market. The rapid development of WiFi technology has made mobile devices (laptops, notebooks, and Tablets more so than Palm-like devices) more useful to those who already possess such devices, but it's not enough. Mobile computing requires a mobile method of I/O—something that we can access without having to sit down and spread out. I think it's going to be a while yet before we see that.