As humans living in the modern world, we have to argue that we aren't robots often enough, and we don't even live in a sci-fi futuristic dystopia. It doesn't matter whether you check an "I'm not a robot" box, teach an AI what constitutes a "sign" (hint:the "correct" answer is the one that the majority of other users chose; try to think like the crowd), or solving math problems, the goal is always the same:to stop bots from messing up websites, and maybe using the humans solving captchas to scan books, train character recognition software, images or generate advertising revenue. But there are more captchas than meets the eye, and they're far from foolproof.
ContentsWhy do we need captchas?What exactly is a captcha?Captchas:making the world a better placeIn conclusion:am I a robot?CAPTCHA stands for “Fully Automated Public Turing Test to Tell Computers From Humans ”, which, besides being a really fancy acronym, tells you most of what you need to know. The idea is, as Google's reCAPTCHA motto says, to create an "easy for people, hard for bots" task.
If you've been browsing the internet for a while, you'll remember that for most of the 2000s, the most common type of captcha was a garbled strip of text containing a string of alphanumeric characters. It's not a very secure form of captcha anymore, but when Google acquired reCAPTCHA in 2009, it was still good enough to get most bots out of it. Since then, Google has moved to more secure "I'm not a robot" boxes (which actually monitor behavior like mouse movement and browser information to check if you're a bot) and user recognition challenges. 'picture. Audio captchas are still around, though, and they're surprisingly easy to crack with voice recognition software.
Image recognition captchas have had their own set of issues, as they can be a bit confusing to interviewees. As mentioned above, however, there is no right answer - since the computer doesn't know which images are shop windows and which are schnauzers - it simply accepts the majority human opinion as correct. If 75/100 humans decide to label a blurry image of a mop as a schnauzer, the computer will assume the mop is a schnauzer and mislabel you if you don't label it as such.
But there are plenty of other captcha options as well, and they can get quite creative. These are just a few of the ideas that have been popping up on various websites.
the slide lock captcha:
The math problem captcha:
The drag and drop captcha:
the image orientation captcha:
the logic/grammar captcha:
There are also captchas that you never see, like the honeypot captcha, which involves adding an invisible field to a web page, waiting for a bot to fill it in (humans won't do that, because they can't). see), and then launch the bot. Then there's the "invisible captcha", often associated with their box, which monitors how you navigate a web page (mouse movements, scrolling, clicking, general behavior) to see if that should give you a captcha. image recognition as a double check.
You may not know it, but the cumulative hours you've spent proving you're not a robot may have made a difference. reCAPTCHA, now Google's captcha service, was originally conceived by Luis von Ahn (now best known as the founder of Duolingo), as a way to use wasted brainpower to digitize books. By presenting users with a scanned word from a book or newspaper, this system could both confirm a user's identity and conduct a kind of opinion poll on what the word was. If enough people agreed on the word, the scanning system would accept the answer in the ebook version.
After implementing this system, it took only two years to digitize the entire Google Books library and the entire New York Times archive. In 2012, they switched to using humans to enter house numbers pulled from Google Street View.
In 2014, things took an ironic turn toward robotics:image recognition captchas. These work on the principle that machines aren't very good at understanding what's in an image, but, as described above, they've been pretty good at training AIs to do just that. Since this captcha will eventually fade from one job, it is being phased out in favor of the less visible behavior/tracking focused ones.
As artificial intelligence, deep learning, and a host of other advancements will be underway over the next few decades, captchas will also need to evolve. Most of the existing captchas have already been cracked, and it's getting easier and easier. Training a machine to read distorted text captchas now takes about 15 minutes. Maybe the only thing left in the future will be biometric captchas (hope you like facial recognition scans!), or maybe we'll wake up and find that the singularity has already been reached, and we were the bots all the time.
Image credit:Chippee via wrong house numbers google recaptcha
