I’m a big fan of prizes as a way to incentivize research and development.  Not only is money a big reward, but there’s also prestige, something not found with your basic paycheck.  From the X-Prize to Netflix’s recommendation software, prizes are proving to be a growing model.  I didn’t expect, however, spammers to use a prize model to improve CAPTCHA cracking software.

CAPTCHAs are those random letters you often have to type on online forms.  These prevent automatic bots from overloading systems.  But for every CAPTCHA created, someone eventually cracks its, leading to a constant arms race between spammers and security professionals.

Spammers are reportedly offering rewards up to $500,000 to anyone who can develop a way to crack reCAPTCHA, one of the leading CAPTCHA systems (used by Facebook and Craigslist).  Luis von Ahn, the co-creator of CAPTCHAs, said “If [the spammers] are really able to write a program to read distorted text, great – they have solved an AI problem.”  New Scientist notes this technology could be revolutionary optical character recognition software.

While I’m not trying to encourage more spammers (I already hate CAPTCHA which, if Google has its way, could get even more complicated), the prize model itself is still a great idea I hope to see more of in many different areas.