I am not a robot.

It’s a common box that you have to tick on many websites as you log in - which apparently proves that you are a human, and not a robot.

Many people however, seem to be more robot than human as they struggle to get the correct answers - which in some cases could mean the difference between getting tickets to a sell-out show or not.

Captchas (an acronym for 'Completely Automated Public Turing Test to Tell Computers and Humans Apart') are puzzles designed to safeguard websites from spamming bots attacking networks and websites.

The captchas can be as simple as checking an 'I am not a robot' box or more difficult such as deciphering artistically drawn letters or numbers, or sometimes seemingly impossible like identifying images in a grid that contain a stoplight.

The bad news for those of us who already struggle to solve the current captchas, is that they are about to get a lot harder.

Since captchas were launched in the year 2000, bots have become much smarter, meaning that now it's the humans that need to keep up.

Researchers at the University of California, Irvine recently found that bots could consistently answer the captchas that showed distorted text with almost 100 percent accuracy.

With the advancement of artificial intelligence powered image recognition - bots can now also identify images including stoplights, bridges and the other common images that captchas ask humans to find.

Ironically, while humans still struggle with these image tests trying to decide if the tiny corner of a stoplight, or bridge within a square is enough to be selected or not, the bots have learned exactly how much is needed to pass through the system.

So what do we have to look forward to?

To try and outsmart the robots again, captchas are moving to puzzles that ask humans to solve puzzles that will seem to make no sense. As robots and AI struggle with nonsensical problems it is hoped that presenting puzzles that don’t look like real life could give humans the advantage again.

New puzzles might show you a picture of a wild animal that is dressed in human-like clothing including a vest and jacket. Rather than be in the forest, the animal might be floating around with some watermelons. The captcha request could be to click on the animal’s bowtie. Hopefully easy for a human, but perhaps more difficult for a robot with so many parts of the image being out of context.

Other puzzles might ask you to identify two objects in a photograph that are the same shape, or to click on the non-aquatic animal in a photo containing lots of different sea creatures.

So if you are just learning how to be more human than a robot in a test, get ready to redefine your humanness once again.

LISTEN ABOVE

I am not a robot.
It’s a common box that you have to tick on many websites as you log in - which apparently proves that you are a human, and not a robot.
Many people however, seem to be more robot than human as they struggle to get the correct answers - which in some cases could mean the difference between getting tickets to a sell-out show or not.
Captchas (an acronym for 'Completely Automated Public Turing Test to Tell Computers and Humans Apart') are puzzles designed to safeguard websites from spamming bots attacking networks and websites.
The captchas can be as simple as checking an 'I am not a robot' box or more difficult such as deciphering artistically drawn letters or numbers, or sometimes seemingly impossible like identifying images in a grid that contain a stoplight.
The bad news for those of us who already struggle to solve the current captchas, is that they are about to get a lot harder.
Since captchas were launched in the year 2000, bots have become much smarter, meaning that now it's the humans that need to keep up.
Researchers at the University of California, Irvine recently found that bots could consistently answer the captchas that showed distorted text with almost 100 percent accuracy.
With the advancement of artificial intelligence powered image recognition - bots can now also identify images including stoplights, bridges and the other common images that captchas ask humans to find.
Ironically, while humans still struggle with these image tests trying to decide if the tiny corner of a stoplight, or bridge within a square is enough to be selected or not, the bots have learned exactly how much is needed to pass through the system.
So what do we have to look forward to?
To try and outsmart the robots again, captchas are moving to puzzles that ask humans to solve puzzles that will seem to make no sense. As robots and AI struggle with nonsensical problems it is hoped that presenting puzzles that don’t look like real life could give humans the advantage again.
New puzzles might show you a picture of a wild animal that is dressed in human-like clothing including a vest and jacket. Rather than be in the forest, the animal might be floating around with some watermelons. The captcha request could be to click on the animal’s bowtie. Hopefully easy for a human, but perhaps more difficult for a robot with so many parts of the image being out of context.
Other puzzles might ask you to identify two objects in a photograph that are the same shape, or to click on the non-aquatic animal in a photo containing lots of different sea creatures.
So if you are just learning how to be more human than a robot in a test, get ready to redefine your humanness once again.
LISTEN ABOVE

News, opinion, analysis, lifestyle and entertainment – we’ve got your Sunday morning listening covered with The Sunday Session with Francesca Rudkin on Newstalk ZB.

Advertisement