Computers are finally becoming good listeners. But are they understanding us well enough? Windows 7 tries very hard.
“Start Firefox. Start Firefox. Start Firefox! START FIREFOX!! Why the !@#$ doesn’t this work??”
That is a quote from a popular YouTube video in which an enterprising young man, with a thick Eastern European accent, tries to use Windows 7’s speech recognition system. While the software application more or less does what he wants it to, it does occasionally act up. And stubbornly refuse to do things. Like summon the Firefox browser.
Hands off: Instead of typing, dictate a letter to the computer.
The young man’s predicament would be really, really funny. If only it wasn’t so sad.
These days you have computers that can process the most demanding multimedia applications. You have gaming consoles that can generate lifelike recreations of World War II battle scenes. You have iPods that can sense motion and geographic location. Hell, there are even computers that help chess grand masters train for their matches.
Why then, in 2010, can’t you still just dictate a letter, or even a newspaper column for that matter, right into a computer microphone? And then magically see it transcribed, if not punctuated, into a word processor document?
Actually you can almost do this today. And we’ll get to that.
But there was a time when universal voice activation and precise speech recognition seemed just around the corner.
If you ever spent time tinkering with older versions of the Windows operating system, circa Windows 98, you may have come across pieces of software by a company called Lernout and Hauspie Speech Products (L&H). L&H, at one time, was at the cutting edge of speech recognition technology. They made products that read out text, transcribed voice and even corrected grammar. So hyped was the company that the area surrounding the company’s office at Ieper in Flanders, Belgium, was known as the Flanders Language Valley.
And then in 2001 L&H exploded in a mess of bankruptcy and financial scandal.
Personally I’d given up all hope of being able to just walk around my bedroom, muttering away into a Bluetooth headset, seeing the words magically appear on a computer screen in one corner. Imagine how beautiful offices and newsrooms would be if you had hordes of people lounging in front of their PCs doing this.
Rejoice then. Because if you have a Windows 7 PC at home or at work, you can actually try this for yourself right away. Despite reading through reams of reviews and analysis of the operating system, I’d never paid attention to the fact that Windows 7 came with a full-fledged speech recognition system. With it, and with a lot of patience, you can actually achieve the incredible: dictate text directly into a document.
First of all you need a decent microphone (if you are reading this column and just can’t wait to try the software here’s a tip: You can plug your iPod headphones into the microphone socket and speak into the earbuds. Works often. Slightly gross).
Then search for Speech Recognition in Windows 7’s useful search bar. Double-click and you will be taken into a great tutorial that shows you the remarkable ways in which you can control your computer comprehensively using only voice.
The tutorial will, frankly, blow your mind. The demos work brilliantly. There are clever ways of entering text and summoning programs that you will learn in minutes. The problem, however, is with accuracy. The software gets better with practice, as it gets used to your voice, and precise transcription will take time. But till then you can have fun calling up a browser and surfing the Web using only your voice.
Windows 7 speech recognition is available out of the box, at no extra cost, for all versions of the operating system. And the only technological qualifications you need are the ability to speak and to plug a microphone into a socket.
If you have the operating system at home or at work go try it right away. But do it discreetly in a private place. Unlike the poor YouTube boy.
Play Things is the official tech and time-pass blog of Mint. Drop in for a dose of cool tech gossip and online merriment at http://blog.livemint.com/play-things