Recently, you may have noticed fewer websites prompting you to type those words that appear in a red rectangle. Instead, you may be seeing an “Invisible reCAPTCHA” button which prompts you to click a button to confirm that you are not a robot. You can thank Google for this. This is a system that analyzes your browsing behavior to determine whether it is being clicked manually or is the work of a bot. That means that there already exists a better system for determining whether someone is a human than the one we’re used to using. At this point, we don’t need traditional CAPTCHAs anymore. This makes browsing the web easier for you, while ensuring that no bots interfere with your experience. Unfortunately, it is still far from perfect and quite susceptible to being cracked.
The thing to note with regards to the new CAPTCHA-solving model developed by Vicarious is that it’s not as bad as it may immediately sound. I can assure you with utmost certainty that the Four Horsemen have not descended onto the land, the souls of the dead have yet to rise once more, seven angels are not, at this moment, standing before God to sound their trumpets and exactly zero scrolls of the Apocalypse have been unsealed. In short, the world has not ended. CAPTCHAs were, for the record, on their way out anyway. This just hurried the process along. If nothing else, I hope I can convince you that there isn’t anything to fear from a program that can recognize words written in squiggly letters.
Indeed, the CAPTCHA is both archaic and ideologically problematic. It relies on the ability of the human vision system being superior to that which can be simulated by a computer. However, the very goal of computer vision is to automate tasks that can be accomplished with human vision. To do this effectively, it is necessary to develop systems with symbol recognition capabilities on par with those of a human. On the other hand, it is only the inability of machines to pass the Turing test that keeps our favorite online forums safe from automated posts with links to malicious websites. Frankly, Google’s solution is not a permanent fix, as it can still be broken by similar methods. The issue, interestingly, is that the goals of security and computer vision are, in a way, at odds. This process in which new security features are developed and subsequently broken is integral to the field as a whole, to the point that there exist groups of “white-hat hackers” who test the security of information systems for the purpose of improving them, rather than to exploit weaknesses in them. Breaking systems like this is the first step to augmenting their security features to protect against the kinds of hackers who do harbor malicious intent.
All these people have done is develop a better method by which a computer can tell symbols apart. The goal was not just to solve CAPTCHAs, but to build intelligence that works like the human brain. This application is just a method of testing the system. In short, CAPTCHAs are only the beginning. Soon, we may be seeing robots doing many of the things we thought were reserved for human intelligence. Vicarious, the company responsible for this, has accumulated an impressive list of Silicon Valley’s biggest names, including Bezos and Zuckerberg, as investors. Also among them is Elon Musk, who famously claimed that artificial intelligence was “summoning the Demon,” firmly believing that AI would be responsible for World War III. If these developments are sufficient to convince even the most fervent doubters, then perhaps the CAPTCHA is a small price to pay for such powerful technology.
Eli Udler is a contributor for The Daily Campus. He can be reached via email at firstname.lastname@example.org.