Rise of the Machines: Most Artificially Intelligent Computers Can Pass CAPTCHA
Google is using a deceptively simple new test instead to weed out the bots.
-- Those distorted strings of letters and numbers you see on websites to prevent spam have instead proven to be more painful to humans than the robots the test, called CAPTCHA, is designed to block.
The moment Alan Turing, regarded as the father of modern computing, predicted is already here: The machines have risen -- and they keep getting smarter.
Google said its research shows that today's artificial intelligence technology can get past CAPTCHAs with 99.8 percent accuracy, rendering them no longer dependable for anything other than driving humans crazy.
The "Completely Automated Public Turing Test to Tell Computers and Humans Apart," or CAPTCHA for short, was named for Turing, who devised a test in 1950 asking if machines can think.
Anyone who has ever spent several attempts trying to type those barely legible strings of letters and numbers into a CAPTCHA to prove their status as a bona fide member of homo sapiens knows how frustrating they can be.
Google announced on Wednesday it is rolling out a simple one-click solution for websites that use its "No CAPTCHA" service to weed out spam from bots.
With the new experience, users will be able to simply check a box that reads: "I'm not a robot." Surprisingly, this could be the ticket (for now) for fending off the bots.
If the risk analysis back end still needs more convincing, the user may be redirected to a traditional CAPTCHA.
Alan Turing would likely be saying "I told you so," if he had lived to see 2014.
In June, a supercomputer posing as a 13-year-old boy named Eugene Goostman, became what is believed to be the first machine to pass the 65-year-old Turing Test.
Goostman pulled off the feat by successfully duping enough judges into believing they were conversing with a real human, according to the University of Reading, which organizes the annual test.