To Solve Big Problems, Assume Everything Is Wrong and Ask Dumb Questions
What does it take for a computer hacker to solve a CAPTCHA (those garbly letters on website forms that make you prove you’re a human)? Apparently, just a little bit of porn.
When CAPTCHAs became popular in the early 2000s, they created a new front in the never-ending arms race between security experts and scammers. There was enough money to be made and lost on the wild web that good guys were constantly building walls to hold bad guys at bay. But rather than attack security walls head-on, the bad guys often simply found ways around them.
CAPTCHA proved a difficult foe. At the time, no computer algorithm could be built to solve the distorted word puzzles. Only a human could do it, thus preventing spammers from spraying their sketchy advertisements at high volume. But it didn’t take long for them to find a workaround.
Instead of trying to build a better algorithm, the spammers simply looked for some humans to solve CAPTCHAs for them. The biggest group of highly-motivated humans on the web at the time: a group I refer to as “porn lookers.”
The immense volume of traffic to pornography sites meant that a spammer stuck behind a CAPTCHA could simply copy the CAPTCHA from a legitimate site and stick it on a porn site, saying, “Fill this out for free porn,” and users would solve the fake CAPTCHA in near real time. The spammers would then simply take that real, human response to the fake CAPTCHA and use it to solve the real CAPTCHA.
All of this could be done with a simple bit of code. So much for security.
The spammers’ solution is an example of what psychologists call “lateral thinking,” or the ability to solve problems faster by reframing them. A business’s natural inclination when confronted with a similar type of arms race is typically to attack the competition head-on -- to build a better mousetrap, to answer the question as presented. A lateral thinker, on the other hand, questions the question itself.
Fortunately, this kind of hacker thinking isn’t just the domain of scam artists. In fact, it’s the kind of creativity that has led to breakthroughs throughout history, in almost every industry. It’s, as Twitter founder Biz Stone recently said, “the ability to connect dots you otherwise couldn’t connect.” It’s how we developed game-changing ideas from iPods to stealth jets to democracy.
Working sideways. As entrepreneurs, we can take a page from the hacker playbook in the way we approach our own business challenges. I call this lateral-thinking-for-good mentality, “smartcuts.” The basic idea: rather than play the game you’re expected to play, play a new one. Rather than attack the castle head-on, find a way around the walls (or get the enemy to come outside).
For example, when Salesforce.com launched its “Customer Relationship Management” product in 1999, the convention among software companies was to charge a large fee for a box of software that you installed on your computer -- think Microsoft Windows or Apple OS -- and to upgrade every couple of years (i.e. buy and install a new box of software).
Rather than compete head to head with the huge, expensive CRM software companies of its day, Salesforce.com decided to play a different game.
Instead of installing software, you just logged into a website where all the features worked online (making it faster for companies to get started). Instead of a big fee, you paid a subscription of a few dollars a month (the fee added up, so Salesforce.com still made money, but it made it extremely easy to sign up risk-averse new customers). Instead of having to buy and install new versions, your tools were upgraded automatically without you having to do anything (and you were less likely to switch to another company between upgrades).
Thus, Salesforce.com created a new industry, a new business model and a massive business. It was David throwing a rock at Goliath from a long distance instead of fighting him with a sword.
In Smartcuts (the book), I explore how businesses throughout history have applied lateral thinking in a variety of forms in order to accelerate their growth. In digging through academic research and conducting hundreds of interviews with superlative successes, I found that often -- in fact, almost always -- the most rapidly successful companies were the ones that regularly found counter-intuitive solutions to problems.
The problem for us, of course, is that learning lateral thinking by definition presents a paradox: How do you think of things that you wouldn’t think of?
The simple answer -- the lateral thinker’s approach to developing lateral thinking, if you will -- is to ask dumb questions.
“Creativity involves breaking out of established patterns,” said Edward de Bono, who coined the term lateral thinking in 1967. “Most of the mistakes in thinking are inadequacies of perception rather than mistakes of logic.”
To combat our tendency to assume too much, a lateral thinker will assume that everything they know and hear is wrong. Elon Musk, whom I profile at the end of the book, calls this, “getting to first principles.”
An easy way to actually do this is to physically write down all of the assumptions inherent to a question (“Do we have to install software on someone’s computer?”), to question the question itself (“What if we didn’t have to compete against software companies?”), to work backwards from the desired solution (“What are the different ways we could backtrack from a billion-dollar company to where we are now?”), and to systematically change perspectives when brainstorming and planning (“How would a five year old, a computer hacker, a magician or an ice skater approach this question?”).
The good kind of hacking. In 2006, a computer scientist, Luis von Ahn, figured out a way to hack CAPTCHAs to solve one of publishing’s biggest problems. No porn necessary.
At this point, the New York Times had been printing a daily newspaper for 155 years. This had added up to millions of stories, and hundreds of millions of words. When the web came around, few of those words were digital. This posed a problem, of course, because paper disintegrates, and computers are bad at reading photos of words.
At that time, the Times had made optical scans of its archives, but the back issues weren’t searchable. It would take decades -- and millions of dollars -- to hand type all of those stories, and computer algorithms weren’t accurate enough to do it automatically.
Von Ahn proposed a lateral solution. By this time, thousands of websites were using CAPTCHAs to cut down on spam (despite the workarounds that the porn spammers had found). Rather than putting random letters and words into CAPTCHA, von Ahn’s team wrote code that would grab words from photographed books and newspapers, smoosh them around so a human could still recognize them but most robots couldn’t, and then let everyday people transcribe them.
People were already filling out CAPTCHAs to access web forms or do other tasks, why not let them kill two birds with one stone?
Cleverly, von Ahn put two words in each CAPTCHA -- one that was already known, and one from an old book or New York Times story. If the person got the first word right, the code would assume that the transcription of the unknown word was also right. This proved 99 percent accurate.
The beauty was that even if spammers used their real-time porn hack to solve the CAPTCHAs, real humans would still be transcribing real words and building the New York Times archive.
Google soon bought von Ahn’s company, reCaptcha, for millions, with the goal of digitizing the rest of the world’s old books.
As we learn from good and bad hackers alike, thinking laterally is not just about finding roads less traveled. It’s about questioning where you’re going, whether you need a road at all and whether you can build something better than a road.
In other words, when someone says the shortest distance between two points is a straight line, a lateral thinker will say, “Sure, but it’s not the fastest.”