The problem with Captchas
Contributed by Joel Zimba, Special Projects Coordinator, MDTAP
According to a recent Ted Radio Hour episode, n.pr/1bOzlZ, we spent 500,000 hours solving Captchas world wide. Chances are, you run into them all the time.
A Captcha is the little box, usually on a website, which asks you to type the blurry words/letters that appear in a picture. The idea is to have you prove you are a human, and not some sort of automated tool—effective, but terribly annoying. They are far more than annoying to users of screen access software. Captchas are more like an immediate accessibility barrier. There is just no way for a screen reader to tell what is in the blurry picture. If there were, then an automated tool could do the same thing, resulting in the Captca being pointless.
Ironically, many Captchas are being used to do important work, like digitizing books or language translation. The very Captcha which prevents me from independently logging on to many popular websites or making purchases may also be producing a book in a more accessible format.
The only thing worse than a visual Captcha is an audio Captcha. It’s the same idea, but the execution proves to be far more difficult. An audio clip is played and the words or numbers must be typed by the user. It sounds simple, but the sound is often deliberately garbled to make the task more difficult for those pesky robots. Challenge your friends and family to the audio Captcha solving challenge. If you happen to know a World War II short wave radio operator, you too may be able to buy Warp Tour tickets this Summer.
If you use FireFox running on the Windows platform, you can try the WebVisum extension. It might solve some Captchas, though it takes far longer than the average of ten seconds sited in the NPR piece. Anyone else is out of luck. We welcome your innovative solutions to the Captcha problem.
