People are just learning how the ‘I am not a robot’ CAPTCHA checkbox actually works… and their minds are blown
FOR years, the role of “I am not a robot” CAPTCHA tests have held up everyday activities on the internet relatively unscrutinised – until recently.
People are only now finding out the exercises are not really intended to block robots masquerading as humans from websites, but something else more sinister.
What was once an inane, slightly tedious activity that briefly halted your internet movements is now a hot topic and people are “mind blown”.
How CAPTCHA (Completely Automated Public Turing) actually works and its purpose goes far deeper than clicking a button or correctly identifying a traffic light.
According to the cybersecurity firm Cloudfare, CAPTCHA watches the every movement of your cursor.
“Even the most direct motion by a human has some amount of randomness on the microscopic level: tiny unconscious movements that bots can’t easily mimic.
“If the cursor’s movement contains some of this unpredictability, then the test decides that the user is probably legitimate,” they tell netizens on their website.
However, the powerful technology doesn’t stop there and might even end up checking out your search history.
Google-owned software reCAPTCHA “may assess the cookies stored by the browser on a user device and the device’s history in order to tell if the user is likely to be a bot,” says Cloudfare.
BBC’s panel show QI confirmed this further as host Sandi Toksvig sought to educate us all on what the real meaning of this seemingly easy exercise is.
“Let us say, for example, before you tick the box you watched a couple of cat videos, you liked a tweet about Greta Thunberg, you checked your Gmail account before you got down to work – all of that makes them think that you must be a human.”
Sandi continued: “Essentially when you are clicking ‘I am not a robot’ you are instructing the site to have a look at your data and decide for itself.”
If the software is still not satisfied you are a human, then it can throw yet another test at you. “It’s slightly spooky,” she said.
The QI video has been doing the rounds online and not everyone is happy with knowing the truth behind the CAPTCHA tasks.
“Mind blown,” one wrote.
Another said it “feels like an invasion of privacy”.
A third said: “Just realized google been using us for years to train ai on identifying pictures by using the captcha/proving we are not robots things
However, many online users seemed to just rally against the existence of it and the inconvenience it causes – regardless of what it’s for.
“Captcha test is so annoying. Wdym I’m a robot?” said one.
“I’ve failed so many captcha challenges lately but I promise I will be one of the good robots. You’ll see!” another joked.
A third mused on what happens if a robot got caught. “How many robots have gotten away with checking the “I’m not a robot” Captcha? What happens if they get caught? Is there an AI jail? How long is the sentence? Who is the judge? Will they stop clicking the box if released?”
A few years ago rumours began to surface that the software was not developed to banish bots but to instead help to train them in human ways.
Techradar reported back in 2018 that “you’ve been training AI for years without realising it”.
“All those visual puzzles add up to AI advances,” they said.
Techradar claims that millions of people have been unknowingly doing work for big companies, like Google.
Without realising, they reported, online users have been performing small tasks as part of CAPTCHA that needed doing anyway, such as transcribing words for the purpose digitising books.
More sinisterly, you also might have been training AI on how to correctly identify objects on maps, through the picture-based exercises.
In that sense, AI can been trained to recognise those objects in other images in applications like Google images or Google photos – building up a sturdy database for machine learning.
As QI’s Sandi Toksvig said: “It’s slight spooky”.