Clickbait headlines: Human vs. machine - Can you tell the difference?

You won't believe how fast you can fall out of shape! You won't believe how little this NBA dancer says she was paid! What you don't know about eating peanut butter will shock you!

Here's what won't shock you: silly clickbait headlines may soon be generated by automated software.

Thanks to Click-o-Tron, created by Lars Eidnes, a programmer from Norway, software is now creating clickbait headlines as terrible as if they were written by a real-life human being.

Think you can tell the difference between Click-o-Tron and something a human wrote? One link below leads to Click-o-Tron and the other leads you to CNN. Pick the human.

• Director Robert White On How He Gets Away With Murder
• You won't believe the $%#! you've been eating

CBC Radio technology columnist Jesse Hirsh explains it all. Listen to his full interview with CBC Hamilton's Conrad Collaco by clicking the image above. You can read a summary of the interview below.

Tell us about Click-o-Tron

​It's one part joke and one part experiment and one part inadvertent utility. The joke part is that it was designed to make fun of the phenomenon of clickbait. To perhaps argue that the value we place on attention grabbing headlines is misplaced. If a piece of software could create clickbait better than a human, perhaps humans could spend their time in better ways than creating clickbait. The developer did this to mock clickbait but he hit a chord, and so Click-o-Tron.com has become quite a phenomenon. It seems people are entertained by the idea that this smart software could come up with entertaining headlines.

• Why Hamilton won't survive a Zombie Apocalypse

Are media companies using this?

Not yet. But the dangers of criticizing technology is that you might actually improve it and that might be the case with clickbait. He created the software to mock clickbait but he may have come up with a pretty good clickbait engine. If anyone wanted to use it, not to ridicule, but to actually create (automated) attention grabbing headlines it might be an effective utility. 

It is often the case that if you're too successful with criticism you reach innovation. The experiment began by taking 2 million articles published from the likes of Buzzfeed, the Huffington Post or Upworthy — basically the prime practitioners of clickbait — and feeding the words into the software, then mapped out the relationship between those words and the headlines. At first the headlines it created were nonsense but over time those words started working together, started to make sense. At Ckick-oTron.com people can vote, up or down, on these headlines to indicate if they find the funny, or insightful or do well to mock other clickbait headlines. 

• Want your skin to look 20 to 30 years younger? Try this

Don't we hate clickbait headlines and have built some defence to them?

We hate clickbait because it works. We feel manipulated because it speaks to our psychology but for every successful clickbait headline there are dozens that we ignore. However it's difficult to defend against curiosity. It's difficult to not click on something you actually think is interesting.

The other frustration is when those headlines don't deliver. The ones that promise substantive content but deliver superficial fluff. Perhaps the moral is that we could spend more time on substance. We could spend more time on investigative reports and more time on analyzing the important issues of our day without worrying how we are going to get an audience.

Google is already acting as a filter. On my smartphone every morning Google already lists articles that I might be interested in and it uses different criteria. Some are listed because they are trending and it's news. So, to go back to your point maybe it's a battle of algorithms. One to create headlines to grab our attention and another to filter out their manipulation of our attention. 

• Want to eat like a teen again? Science may have found a way