Are you finding those annoying little puzzles you need to complete before you can proceed on websites harder to solve? You aren't alone and there might be a good reason for it.
Experts have confirmed that Captchas, which were developed in the early 2000s, are indeed getting more difficult, and this tells us "something terrifying about the future", said Metro.
What's happened?
The Captcha, an approximate acronym that stands for Completely Automated Public Turing test to tell Computers and Humans Apart, were designed to identify a human from a bot. Evolving from transcribing random distorted text, to ticking a box to confirm you are not a robot, and clicking on all the motorcycles and traffic lights in a grid, they have "become ubiquitous" as a way to deter fraudsters, scammers and abusers, said The Times.
"At their heart is a simple equation", added the broadsheet. The Captcha "has to be easier for a human to do than a machine", but as machines and AI have become better at tasks and identifying images, those developing the Captchas have "had to up their game, quite literally".
A report from the University of California Irvine found that bots could consistently answer Captchas’ distorted text with almost 100% accuracy. The researchers put this down to "advances in computer vision and machine learning", as well "Captcha farms", which are "sweatshop-like operations where humans are paid to solve Captchas". Humans' accuracy rate is lower: it ranges from 50–86%.
In his most recent Netflix special, the stand-up comic Jack Whitehall asked the audience: "Is it just me, or have those 'I am not a robot' tests started getting harder?" Experts disagree on the answer. “Yes, I think so," Cyril Noel-Tagoe, principal security researcher at Netacea, a company that combats bots, told The Times, adding that "we've done a lot of research on this internally".
But a spokesperson from hCaptcha, which develops the systems, disagreed. "hCaptcha challenges are not actually getting harder for people to solve, as measured by time needed and percentage of people passing on the first try," they said, but "any change in a familiar system can create a temporary perception of difficulty".
They explained that "several years of relatively slow progress" in visual AI meant that "people got used to seeing a small number of questions about bicycles or crosswalks", but today there is "more diversity" in the process.
Why does it matter?
If the process continues to get harder, this could cause problems for people trying to book tickets for popular shows, said Metro. The fact bots can solve more Captchas and their "ever-increasing level of difficulty" could mean "the difference between getting tickets to a show or standing outside the concert hall, looking for a scalper".
We shouldn't expect a return to the easier days of Captchas, Kevin Gosschalk, the founder and CEO of Arkose Labs, which develops Captchas, told the Wall Street Journal. "Things are going to get even stranger, to be honest, because now you have to do something that’s nonsensical," he said. "Otherwise", bots "will be able to understand" the tests.
The fact that artificial intelligence has caught up with a security measure intended to protect us from it, has troubled many, and this is not the first time that online security measures have unsettled people. Last year, people were "not impressed" when they discovered "what really happens" when you confirm you are not a robot, said the Daily Mail. The action prompts the website to check your browsing history and collect your data, it reported.
The data collected includes "your time zone, IP address, screen size, browser and plugins", as well as "key presses, mouse clicks, browsing history and things we may not know about", said BBC Science Focus.
0 Commentaires