What about voice?

I am not following voice-recognition and its potential applications but today I've been confronted to three papers about it in my daily scans. Even though it's still R&D oriented, each papers delivered some promising messages about a technology that I am skeptical about (based on previous research project and research readings). First there is this ACM Queue discussion by John Canny (University of California, Berkeley), which is actually a great piece about the future of HCI. Canny quote Jordan Cohen's work (formerly of VoiceSignal, now of SRI International)

"The killer application is probably going to end up being some kind of interface with search, which seems to be the very hot topic in the world today; for mobile search especially, speech is a pretty reasonable interface, at least for the input side of it,"

This "search" concept is what I ran across this morning in a Business Week article by Steve Hamm, there is a presentation fo a curious application called TellMe about voice-driven Web information:

The idea is to create mobile search services that can make it easy for those on the go to find people, businesses, and information. That goes for any phone, but especially those equipped with browsers. A tourist might bark "restaurants," "sushi," and "downtown" into his cell phone and then see listings, read online reviews, make reservations, and retrieve a map with directions. "It has taken us six years to get to this point, but now we can really start to deliver on our original mission," says McCue, TellMe's CEO. (...) Skeptics point out that despite technology advances, voice recognition still turns off many consumers, who remember past glitches. But experts say that will change when systems combine voice, text messaging, and graphic info from Web pages. Each mode will be used for what it does best. "People will be using voice to launch into their search, and they'll want to see the information on a screen," says David Albright, executive director for marketing for Cingular Wireless, which is working with TellMe.

Yes, of course these last pointed I quoted are recurrent, but as presented in this Speech Technology Magazine Issues, there are others applications:

Use your telephone or cell phone to talk with Google—search the Web for answers to your questions, extract the information chunks you need, and listen to the results...Rather than struggling to find the answer to a specific question by chasing links across a Web site, you can simply click a button on the GUI screen and be connected to a human or artificial agent... instruct your oven through your cell phones...

Why do I blog this? don't know whether it's apophenia but I ran across those 3 articles today. So what? I am still dubious about speech technologies but there seems to be confidence in this avenue.