Hearing Voices ... Everywhere

By Samuel Greengard  |  Posted Thursday, October 11, 2012 21:54 PM

By Samuel Greengard

Speech recognition has come a long way over the last few years. The recent update of Apple's Siri provides a glimpse into the future of technology interfaces. It transforms a clunky five-step process into a single streamlined step.

Ask what the weather is in Missoula or Madrid, and you get an instant answer. Check on what movies are playing at local cinemas, and you see show times immediately. The virtual assistant also provides restaurant reviews and reservations through OpenTable, stock information and driving directions. Plus, it lets you place calls and dictate a text message or email.

Yet, despite some flashy capabilities, Siri still functions at a relatively primitive level. It's able to handle and automate an array of common tasks, but it's no HAL from 2001: A Space Odyssey.

For example, don't bother to ask which baseball team won the 1957 World Series or utter “Tom Cruise” and expect to get his bio. When I tried it, Siri suggested several people named "Tom" from my contacts.

There are also issues revolving around data protection and privacy. Earlier this year, Technology Review reported that IBM has asked employees to avoid using Siri for fear that sensitive contact information could leak out. That's because Siri voice data is stored on Apple servers.

Nevertheless, the use of speech systems will undoubtedly grow over the next few years. But why stop at smartphones, tablets and dedicated computing devices?

Car systems must improve as well. Part of the problem is the ambient noise level in the cabin, but many of these systems also rely on arcane commands and lack common-sense functions such as “open the rear trunk” or “turn off the headlights after 30 seconds.”

And what about household appliances? Wouldn’t it be great to tell the thermostat to switch to cooling mode and hold the temperature at 76 degrees? Or order the oven to bake for 30 minutes at 350 degrees and then broil for three minutes at 450 degrees? Or tell the television to display only channels currently offering news?

You get the idea. The possibilities are almost limitless. Speech systems could get a lot smarter.

Of course, we will eventually have speech recognition embedded in a dizzying array of devices, and it's likely that these systems will use biometric technology to customize responses.

Frankly, for now, I'm happy to see Siri simplify basic tasks on an iPhone and iPad. I'll take speaking a few words over a half dozen pinches, swipes and taps any day.