If people get over their self-consciousness about looking like cyborgs, then head-mounted displays only help relieve part of the problem of wireless connection to the Internet. Displays provide “output”-the information coming from the network to the user. The other half of the circle is “input”-sending responses back out into cyberspace. And it ain’t easy. “Output is hard,” notes Michael Karasick, chief technology officer of IBM’s pervasive computing division. “Input is harder.”
Just think of what it’s like using the keypad of your cell phone to send typed messages. You must press the same key three, even four, times to type a single letter. Some people may thrive on these difficulties: Adam Lavine, CEO of FunMail in Pleasanton, CA, reports from his Asian travels that the Japanese have a word, which translates roughly as “lightning thumb,” referring to people who are particularly adept at typing messages on mobile phones. But for those of us with ordinary thumbs, something more comfortable is in order. And it’s coming. Targus and other companies sell folding keyboards that work with PDAs. Last fall, Electro-Textiles demonstrated a prototype PDA fabric keyboard that could actually be rolled up for storage.
Some of us won’t ever have fingers dexterous enough to do our talking for us. But few of us have trouble talking, and that’s where voice recognition comes in. A number of cellular-phone manufacturers, including Motorola and Nokia, have been offering voice recognition for contact lookup for a few years. Samsung has incorporated voice navigation into a wristwatch-like phone. Conversay, which provided the voice technology for Samsung, has also ported its products to the Microsoft Windows CE Pocket PC platform.
However, voice recognition doesn’t necessarily make for a good user experience. “In a device, you need to make sure you’re not trying to force it to do things more easily done with a single button click,” says Williamson.
This could be tough in an area where restricted vocabularies-think “yes” and “no”-are more commonplace than sophisticated sentences. A general interface that would allow anyone to talk freely is overly optimistic, according to Dr. George White, senior vice president of technology at NetByTel, a Boca Raton, FL-based vendor of voice commerce systems. “True speech recognition requires enormous intelligence, which we can’t get into these little devices, and it will always be that way,” says White. But that doesn’t mean more sophisticated voice recognition will never be mobile. For complex voice recognition, White and others think the mobile device will do some preprocessing, then pass the sound over the Internet to a server that can handle more elaborate processing.
Each approach to more fluid input has its champions, but some experts think the ultimate solution will not come from a single technology but from a combination of approaches. Omar Javaid, chairman of New York mobile commerce consulting firm Mobilocity, shares White’s skepticism about voice recognition. “We call it the Star Trek interface,” he says. “In all likelihood, no single method of input will serve, and mobile electronics will probably best serve with a combination of all these methods and others as they become available.”
The ultimate interface, some argue, would be the complete integration of human and machine, with chips and sensors implanted under the skin to detect the user’s every intent. Issues of social acceptance aside, there may be practical reasons to keep devices at arm’s length. Doug Armstrong, CEO of AppForge, which manufactures mobile and wireless software development technology, remembers being part of a Navy research project that was monitoring the brain activity of pilots to try to let them “think” the weapons on their F-14s and F-22s into action. But the result was like telling someone not to think of the word “persimmon”-the pilots would think the trigger word even when they did not want to fire.
At this point, tech consultant Mike Foster, whose Palm may have cost him a payday, would consider almost any high-tech addition to improve his experience. “I can see myself stopping in a restaurant and putting all that stuff on,” he says, but only if it were sturdy and durable. In the meantime, he’d settle for a better screen. “I understand the physical limitations, but at some point having 640 by 480 on a Palm in color, that would be great,” Foster muses. “I’d pay three times the price for that.”
For now, Foster will continue trudging through the Internet, wishing under his breath for improved displays, better keyboards, friendlier interfaces. If vendors build it, he will buy it-as will others, no doubt. But so far, interfaces for handhelds are nothing more than tiny nozzles, and that makes the wireless Internet industry a little nervous, as they watch the flood of data gushing down the pipes.