We all hate filling out Captcha forms every time we log into a website, but there’s a certain genius to Captcha that many people don’t know is there.
CAPTCHA stands for “Completely Automated Public Turing Test to Tell Computers and Humans Apart”. It’s… quite a handful. It came about as a creative response to an ever increasing problem: automated robots were causing havoc on the internet. Its job is a simple one: Figure out if you’re a human or a robot by using random words that were difficult for a computer to decipher. In fact, a lot of them were too difficult for humans to decipher either. The popularity of CAPTCHA grew. People were filling out millions of CAPTCHAs ever day.

CAPTCHA 1.0 was a creative solution to an important problem, but that’s not what makes it genius. Whether you’ve known it or not, each time you fill out a CAPTCHA form, you’re actually training AI.

In 2009, Google bought CAPTCHA. They replaced the random letters and numbers with pictures of words. To the normal user, this seemed like a normal evolution because nothing much changed. But under the hood, a lot changed. You were no longer filling out random words… you were actually transcribing literature. CAPTCHA gave users two words. One word they already knew was correct. This is how they verified that you were a human. The other word they didn’t know. It was simple a photo taken from an old book that they wanted transcribed. When you typed in the second word, you unknowingly transcribed a tiny piece of a book for Google. Google went from a smeared picture of an old, outdated book to an electronic word that thousands of people had confirmed is correct. This is why sometimes you could type in the wrong word and get away with it. That “I guess I was close enough” feeling you might have gotten from time to time wasn’t just you. Google didn’t know the correct answer either. By giving the same word to many users, Google could figure out what the correct word was. With millions of submissions every day, it didn’t take long to transcribe entire books.

Fast forward to 2020, the same basic concept is at play. Why is CAPTCHA so interested in having you identify pictures of taxis and bridges? Because you’re teaching their AI how to correctly identify them.

The original CAPTCHA figured out how to solve an important problem, but it was Google that saw the true potential, something the original company never noticed. Had they realized the true potential of their own creation, they would have sold for a far higher price.

As technology evolved, a second method was used for catching spammers. It was called the Honeypot Technique. It solves the same problem that CAPTCHA does, but in an entirely different way. Instead of requiring humans to enter numbers to prove their humanity, it decided to trick robots into making mistakes. What was amazing about the creative solution is that it was so user-friendly that most people never realized it was happening. The Honeypot Technique uses a simple premise: Both robots and humans will fill out all the fields they see… but what if some fields were invisible? They were there, but they were imperceptible to the human eye? So they added invisible input fields that only robots would see. Robots continually fell for the trick. They were’t able to realize that they were filling out a part of a form that humans weren’t able to see. The robots were tricked into revealing themselves, and it all happened in the background so that humans were never even bothered with it.

These are two amazing examples of creative ideas that solve an important problem in very different ways. Of course, the cat-and-mouse game between CAPTCHA and spammers continues. A developer recently figured out how to use a different AI recognition program to automatically decode Google’s algorithm. So the cat-and-mouse game will continue long into the future.

As you’re creating today, remember that problems can be solved in more than one way. The biggest weakness of an idea might just be one of its biggest strengths. The best creative solutions solve multiple problems. Your goal isn’t to design a Frankenstein solution by mashing up tons of ideas that cover the core problem. You want an elegant solution. Take a step back and view your problem from a position of clarity. Ask yourself “If this weakness were actually a strength, what would that mean?” You’ll likely see that there’s a trade-off. What you original thought was an objective weakness isn’t as black-and-white as you thought.

This is something I’m currently struggling with for this podcast. On one hand, I’d like to have a script to be quick and to the point. That’s a strength. But if I did the opposite, if I worked from a few bullet points and allowed myself to just start talking… that’d also be a strength. But it’d be a strength for a different reason. There’d be more authenticity to the show. You would have a better connection to me because I’d be off-script.

There’s rarely an easy answer to questions like these because they’re a trade-off. When it comes down to it, it’s a value judgement. When solving problems that require a trade-off, it’s important to first realize that there’s a trade-off. You have to let go of the idea that one idea or one path is objectively wrong. It’s only wrong based on your current assumptions. You have to admit that you don’t know the answer as well as you thought. Then, when it’s no longer a black-and-white issue, you’ll be confronted with the gray area of a trade-off. This is where you make a value judgement on who you want to be as a creative person. Don’t get bogged down in the binary choice of A or B, because both answers are right. Your job is to understand the trade-offs between your different choices. Only then will you see creative solutions other people missed.

Enjoyed the article? Listen to the Kaizen Creativity Podcast for creative inspiration throughout your week.