Imagine: How Creativity Works (16 page)

Don immediately went to work on his new cocktail project. He turned Monday and Tuesday at PDT into experimental nights during which he tested his strange new concoctions on customers. Most of the experiments were utter failures, like his attempt to carbonate a cherry. “Wouldn’t it be cool if you got a Manhattan but the maraschino cherry was fizzy inside?” Don asks. “The only problem was that the fruit kept on exploding.” Another experiment involved a gel cap that propelled itself around the drink — “Like a little submarine,” Don says — and kept the cocktail properly mixed. Unfortunately, the propulsion system altered the flavor of the drink; the martinis tasted like baking soda.

But Don refused to get discouraged. “I was having so much fun,” he says. “I was like a little kid in a candy store, except my candy was ninety proof.” And so he continued to experiment, searching for new techniques and ingredients to work into his avant-garde cocktails. For Don, the research was a chance to combine his long-standing interest in chemistry — he had been an engineering major at Columbia University — with his new bartending obsession. “I was really a novice behind the bar,” Don says.

“Unlike the bartenders, I didn’t have an encyclopedic knowledge of vodkas or single-malt whiskeys. I knew much less about liquor per se. But I knew a fair amount of chemistry, and that let me think about bartending in a slightly different way.”

Look, for instance, at one of Don’s first successful inventions: the bacon-infused old-fashioned. The drink relies on a process called fat-washing, in which a fatty food (like cooked bacon) is combined with an alcohol. The mixture is then chilled — the greasy globules solidify on the surface — and strained, so that no lard remains in the liquid. While fat-washing might seem like a weird concept, Don saw the experiment in terms of its elementary chemistry. “I was pretty confident it would work,” he says.

“Alcohol and fat have very particular atomic properties.” Don then launches into a lecture on chemical polarity, a phenomenon that’s caused by the separation of electrical charges within a molecule. “Polarity is why oil and water don’t mix,” Don says. “The fat is nonpolar and the water is polar, and so the molecules stay far apart from each other. I knew that the same principle applies to alcohol, which is also polar.” However, the flavorful compounds embedded in the fat — those stray molecules that give the bacon its savory taste — are polar too, which is why they easily dissolve in the bourbon.

It took Don a few weeks of careful tinkering — he had to adjust the ratio of lard to alcohol, and then figure out how long to refrigerate the mixture — but he eventually developed a booze that made him drool. “It was a delicious drink, a perfect combination of those oaky bourbon flavors with the saltiness of breakfast meat,” he says. “Unfortunately, it was a little much straight, even with ice.” And so Don began testing out cocktail recipes, whisking his flavored bourbon with a variety of mixers. After adding a dash of bitters for balance, he started searching for a sweetener, as the standard squirt of simple syrup just seemed too, well, simple. His breakthrough arrived at breakfast: “I was eating pancakes and I thought about how when you pour the syrup on the pancakes and you get some maple syrup on your bacon by accident . . . That’s pretty fucking tasty. So I decided to try maple syrup with the bacon-bourbon instead of straight sugar. And it worked. It worked really, really well.”

Don added the bacon-infused old-fashioned to the PDT cocktail menu, serving it to his bravest customers on the experimental nights. It was an instant hit. The fat-washed bourbon soon became the bar’s signature drink, garnering praise in Gourmet, Saveur, and New York. Don is now the chief mixologist at the Momofuku restaurants in New York, run by the chef David Chang. His latest cocktails make the bacon-infused old-fashioned seem conventional. There is the celery nori, in which Don steeps dried seaweed in apple brandy and then removes the seaweed and adds a dash of celery-flavored simple syrup. “I don’t know how this drink works, but it does,” he says. “Sometimes, you have to suspend your better judgment and just taste it.” Or consider his sesame-candy cocktail, which involves mixing cognac with toasted sesame seeds and a burnt-caramel syrup. Or his pickled-ramp martini, in which tangy onion juice takes the place of olive brine. And then there is Don’s clever riff on rum and Coke, which begins by fat-washing white rum with melted butter and then steeping freshly popped popcorn in it. The drink is finished with a dash of Coke. “I call it the ‘movie theater,’ ” he says. “On the one hand, it’s composed of really familiar flavors. On the other hand, it’s bizarre to taste them all together in a liquid. I like inventing stuff that’s just weird enough to make you think.”

The success of Don Lee is a story of creativity coming from an outsider, a person on the fringes of a field. It’s a parable about the benefits of knowing less — Don was a passionate amateur — and the virtues of injecting new ideas into an old field. After all, when Don invented the bacon-infused old-fashioned, he wasn’t a cocktail expert. He hadn’t taken any fancy bartending classes or mastered the subtleties of Kentucky whiskey or studied the history of the old-fashioned. (In fact, he was still working as a computer programmer.) “Basically, I experimented with fat-washing because I was bored and nobody told me not to,” Don says. “I’m sure most bartenders would have told me it was a terrible idea, that it would never sell, that I was wasting perfectly good bourbon. But the laws of chemistry told me it should work, so why not try? I guess my only secret is that I didn't know any better."

1.

In the late 1990s, Alpheus Bingham was a vice president at Eli Lilly, one of the largest drug companies in the world. He was in charge of research strategy, helping to manage thousands of scientists working on hundreds of different technical problems. At the time, Eli Lilly’s business was booming — the company was flush with “Prozac profit” — but Bingham was starting to worry about the future. The company was throwing vast sums of money at its scientific problems, desperately trying to develop the next blockbuster drug. Unfortunately, this expensive investment was producing tepid results; Bingham was beginning to wonder if there wasn’t a more efficient approach to drug research. “After spending years on a problem, we’d often end up with a solution that was so imperfect it was virtually useless,” he says. “And those failures weren’t cheap.”

For Bingham, the most troubling aspect of the drug-development model was its complete unpredictability. He had no idea which problems were solvable and which ones weren’t; he couldn’t anticipate how long the questions would take to answer, or where these answers would come from. “That’s what really worried me — I had no idea how to manage the R and D process,” Bingham says. “I didn’t know who should be working on what. And that’s when I started to wonder if all these supposedly impossible technical issues were really impossible. Maybe we just had the wrong people working on them? Maybe someone else could solve them? I always assumed that you hire the best resumé and give the problem to the guy with the most technical experience. But maybe that was a big mistake?”

These troubling questions led Bingham to a radical conclusion: if Eli Lilly couldn’t predict which scientists would find the answer, then it needed to ask everyone the question. Instead of assigning its problems to particular experts inside the company, the corporation should make the problems public. “Needless to say, this strategy broke every rule of corporate R and D,” Bingham says. “Like every other company, Lilly was very secretive about its research projects, for competitive reasons,” he says. “You didn’t want anyone else to know what you were working on.” Bingham, though, was convinced that this secrecy came at a steep cost.

And so, in June of 2001, Bingham launched a website called InnoCentive. The structure of the site was simple: Eli Lilly posted its hardest scientific problems online and attached a monetary reward to each challenge. If the problem was successfully solved, then the solver got the reward. (The money was the incentive part of InnoCentive.) “Mostly we just put up these really hard organic chemistry problems,” Bingham says. “I assumed there was little competitive risk, since a lot of these technical problems had also bedeviled our competitors. Frankly, I didn’t expect many of these challenges to ever get solved.”

A few weeks passed. The InnoCentive site was mostly silent; Bingham thought his pilot project had failed. But then, after a month of nothing, a solution was submitted. And another. And another. “The answers just started pouring in,” Bingham says. “We got these great ideas from researchers we’d never heard of, pursuing angles that had never occurred to us. The creativity was simply astonishing.”

After less than a year of operation, the website had become an essential R & D tool for Eli Lilly, allowing company scientists to benefit from the input of outsiders. By 2003, the site was so successful that it was spun off from its parent company and began featuring challenges from other large corporations, such as Procter and Gamble and General Electric. “These companies did the same thing Lilly had been doing,” Bingham says. “They’d post the stuff they couldn’t solve, put up a little prize money. Like us, they didn’t expect any useful answers. But then they’d often get the solution from some researcher living halfway around the world. It was thrilling. We felt like we’d accessed this great pool of talent.”

InnoCentive has continued to expand at a rapid clip. It now features problems from hundreds of corporations and nonprofits in eight different scientific categories, from agricultural science to mathematics. The challenges on the site are incredibly varied and include everything from a multinational food company looking for a “reduced-fat chocolate-flavored compound coating” to an electronics firm trying to design a lithium-ion battery for a solar-powered computer. There are calls for a spray to protect corn stalks from insect damage, and a request for a software program that can “analyze the emotional responses of consumers in a crowded retail space.” More than two hundred thousand solvers have registered on the site, people who come from every conceivable scientific discipline and more than a hundred and seventy countries.

The most impressive thing about InnoCentive, however, is its effectiveness. “When it comes down to it, the only reason companies use the site is because it works,” Bingham says. “It solves their hardest problems.” And this success isn’t merely anecdotal. In 2007, Karim Lakhani, a professor at the Harvard Business School, began analyzing hundreds of challenges posted on the site. According to Lakhani’s data, nearly 40 percent of the difficult problems posted on InnoCentive were solved within six months. Sometimes, the problems were solved within days of being posted online.

Think, for a moment, about how strange this is: a disparate network of strangers managed to solve challenges that Fortune 500 companies like Eli Lilly, Kraft Foods, SAP, Dow Chemical, and General Electric — companies with research budgets in the billions of dollars — had been unable to solve. By studying how these challenges got solved, Lakhani was able to better understand the surprising success rate of InnoCentive. He could see why the online amateurs were able to answer questions that had frustrated the experienced scientists.

The secret was outsider thinking: the problem solvers on InnoCentive were most effective when working at the margins of their fields. In other words, chemists didn’t solve chemistry problems, they solved molecular biology problems, just as molecular biologists solved chemistry problems. While these people were close enough to understand the challenges, they weren’t so close that their knowledge held them back and caused them to run into the same stumbling blocks as the corporate scientists. “Our results showed that when the solvers rated the problem as outside their own field, they were more likely to discover the answer,” Lakhani says. “Solvers were actually bridging knowledge fields — taking solutions and approaches from one area and applying them to other different areas. We have often heard that innovation occurs at the boundary of disciplines and now we have systematic evidence that this indeed is the case.”

Ed Melcarek, a seven-time solver on InnoCentive, perfectly exemplifies this finding. Although Melcarek has a master’s degree in particle physics, he has never solved a physics challenge on InnoCentive. Instead, he peruses the chemistry and engineering categories on the site, searching for problems that might benefit from his expertise. A few years ago, he helped Colgate-Palmolive come up with a new way of injecting fluoride powder into tubes of toothpaste. (The old method sent plumes of fluoride dust into the factory.) Melcarek’s elegant solution involved imparting an electrical charge to the fluoride while grounding the plastic tube — the particles directed themselves straight inside. “It was really a very simple solution,” Melcarek told Wired. And yet, the same fix had eluded Colgate engineers for decades.

There is something deeply counterintuitive about the success of InnoCentive. We assume that technical problems can be solved only by people with technical expertise; the researcher most likely to find the answer is the one most familiar with the terms of the question. But that assumption is wrong. The people deep inside a domain — the chemists trying to solve a chemistry problem — often suffer from a kind of intellectual handicap. As a result, the impossible problem stays impossible. It’s not until the challenge is shared with motivated outsiders that the solution can be found.

Bingham likes to tell a story that demonstrates the power of InnoCentive. It involves a company that was trying to invent a polymer with a very unique and perplexing set of chemical properties. “Nobody was optimistic that InnoCentive could help the client,” Bingham says. However, after a few months, solvers on the website came up with five different solutions to the problem. “The company paid for all of the solutions,” Bingham says. “They paid awards to a person who studies carbohydrates in Sweden, a small agribusiness company, a retired aerospace engineer, a veterinarian, and a transdermal-drug-delivery-systems specialist. I guarantee that they would have found none of those people within their own company. They would have found none of those people if they had done a literature search in the field of interest. They would have found none of them by soliciting input from their consultants. And they probably wouldn’t have hired any of these people anyway, because none of them were qualified.”