Sunday, January 05, 2025
34.0°F

What is the word on Word?

| March 20, 2011 9:00 PM

Dear PropellerHeads: The last time I closed Word, it tried to send a list of words I had misspelled to Microsoft. Why would they want this information?

A: We applaud your skepticism regarding software that reports back to the mothership. You don't have to be a conspiracy theorist in a tinfoil hat to ponder Microsoft's motives here... but it sure helps! In this case, though, letting Word phone home is all for the greater good.

Microsoft calls this feature "Help Improve Proofing Tools." It's activated when you enroll in their "Customer Experience Improvement Program." You did sign up, right? In the months after Office is installed, your "custom dictionary entries" - the words you told Word really are words and not typos - are sent to Microsoft.

If enough people flag a certain word - say, Wikipedia - as valid, Microsoft might add it to Word's built-in dictionary so future versions recognize it. This is one example of what computer book publisher and technology trendspotter Tim O'Reilly calls harnessing collective intelligence.

The concept isn't new. Google soared to popularity with their PageRank technology, built on the idea that a person who links their web page to another person's web page is effectively vouching for that other page. They later introduced spelling corrections by analyzing how users corrected their own mistakes on follow-up searches.

Amazon got in the game early, too. "Customers who purchased The Greatest Frauds of All-Time also purchased Dr. Phil, Season One, on DVD." Sound familiar? Collective intelligence, indeed.

A newer example is the ReCAPTCHA project (recaptcha.net). CAPTCHAs are those pictures of wavy words you're shown whenever you buy tickets or make purchases on-line. Identifying the words proves you're not a computer program designed to unfairly hog all the tickets. Until recently, those images were generated randomly by a web server.

But now scientists at Carnegie Mellon have devised a way of using words scanned in from books. The offer? "Our software couldn't recognize this word. Tell us what it is and we'll complete the sale." The Internet Archive (archive.org) uses the results to help digitize books for free public consumption.

We're not sure we like this trend. Sure, it's an ingenious work-around for the limitations of artificial intelligence. But could it also be the first step in training people to clean up after technology's messes? Maybe.

We PropellerHeads suspect we're being suckered into doing all the hard work while somewhere in The Matrix a robot laughs at us maniacally. If we're handling the tough tasks, what are we paying the machines for? What? We're not paying them? Oh...

Incidentally, if you support this effort but are afraid of accidentally sending your Twitter login name (BigBoi75) back to MS HQ, you can selectively remove words from the list.

Now if you'll excuse us, we have to go clean up after technology again. The Roomba horked up all over the stairs and our stupid robot dog pinched a couple of double As in our shoes this morning.

When the PropellerHeads at Data Directions aren't busy with their IT projects, they love to answer questions on business or consumer technology. E-mail them to questions@askthepropellerheads.com or contact us at Data Directions Inc., 8510 Bell Creek Road, Mechanicsville, VA 23116. Visit our website at www.askthepropellerheads.com.