Bill Opens the Gates for Speech

When Bill Gates gave a speech last Thursday at Carnegie Mellon University, he predicted that within five years people will discard their keyboards and interact with computers using touchscreens and voice controls—most notably in conducting Internet searches. While analysts and bloggers have pointed out that Gates has a shaky track record of predicting the way certain technologies will affect common culture, the Microsoft chairman’s anticipation of a greater uptake of speech interfaces seems more plausible, most notably because of the decreased size of handheld devices.

"The keypads are getting so small you have to put your finger in a pencil sharpener to hit the buttons," says Jim Larson, an independent consultant and VoiceXML trainer. "Before long, the buttons on cell phones will completely disappear. What option do you have besides a small screen you can touch or speak to?"

Microsoft’s SYNC, factory-installed in certain Ford cars and developed around core technologies from Nuance Communications and CSR, allows users to navigate their music selections or address books by voice—particularly convenient in an eyes-busy, hands-busy environment. New technologies, Larson notes, typically start in niche areas before becoming mainstream.  

A spokesperson for Microsoft's Tellme Networks, which provides voice-based mobile search, claims that "voice is on the cusp of evolving as the next natural interface" and that for it to be fully successful, a more robust network is needed.  

Perhaps because of that technological hurdle, consumers have been reluctant to fully embrace voice as a viable interface. "Everyone understands how buttons and touchscreens work," Larson says. "They’re less error-prone and private than speech recognition. And they’re more widely used now. Speech is in a coming-from-behind kind of race."

And speech, according to Datamonitor senior analyst Daniel Hong, might be too far behind. "Accuracy and alogrithms still have to improve to take into account ambient noise and quality of the microphone on the computer," he says. "Also, there has to be a change in user behavior, quite frankly I don't see a person who has a keyboard and who can type using speech recognition to supplement keyboard strokes for searching."

And while Larson predicts that keypads will eventually be obsolete—phased out in the same manner as old rotary dials—Datamonitor analyst Ri Pierce-Grove disagrees. "The keyboard and mouse have too many advantages to be entirely abandoned, and they will still be a part of the workstation," she says, adding that in five years, it’s likely to see consumers moving comfortably between a variety of different interfaces. She agrees with Gates that speech has the most potential appeal for consumers in mobile interfaces. "We should expect to see speech playing an increasingly important role in mobile search over the next five years," she says.     

Additionally, a Tellme spokesperson claims that the company’s voice service processes 10 billion search requests every year. This magnitude gives the voice platform enough information to learn and improve, thereby delivering more accurate answers.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues