Obnoxious Character Recognition
At the heart of the game of cat-and-mouse played by bloggers and spammers is Captcha, purveyor of those staticky demands to enter the code exactly as shown above. Captcha is premised on the idea that brains are still better than machines at reading text, and that by forcing visitors to decipher a distorted piece of typography, the system can successfully distinguish between humans and robots. Of course, ongoing advancements in OCR technology have sparked a proportionate response in the impenetrability of Captcha, provoking an arms race whose chief casualty is the quality of life online. Next time you’re submitting to some real-world indignity — say, stripping down to your underwear at an airport security screening — try to look forward to the geniality of the virtual world, in which your own computer, from the comfort of your own home, will upbraid you for mistyping B89gqlIIl. And this after it went to all the trouble of obscuring the type using a three-dimensional distortion matrix, edge softening, gaussian interference, random occlusion, and your least favorite font. Puny human.
But happily — brilliantly! — Captcha’s inventor, Luis von Ahn, has inverted his own technology in the service of something grand. Von Ahn’s latest project, reCaptcha, replaces Captcha’s random gobbledygook with actual snippets of digitized books that computers have so far been unable to decipher. ReCaptcha uses each individual human intervention to improve the quality of digital literacy, a welcome relief for readers of this 1861 text that mentions modems (“modem art” is a common flub.) National Public Radio has the full story in this four-minute interview with the inventor himself. —JH