The Idea Machine

I used to dream of a magical machine. It was about as big as a microwave, all silver metal (with rivets, of course), and it had little knobs, lamps, and indicators everywhere. On the left, there were two buttons: One was green and the other one red. On the right, the machine had a tiny slot. Whenever a person pressed one of the two buttons, the machine would start rattling and hissing, all of the little lamps would start blinking and all the indicators would start, well, indicating. After a few seconds, a small slip of paper, just like the one you find inside a fortune cookie, would emerge from the slot on the right of the machine. On it: An idea. But not any idea. A brilliant, life-changing idea. Or, if you had decided to press the red button, a horridly bad idea. You would only be able to use the machine once a day and, naturally, most people would use the green button almost all of the time.

It’s not hard to see why the image of such an idea-generating machine is so alluring: Generating good ideas is one of the hardest things to do. People outside of the creative industry often say: “Oh, it must be so hard to come up with new ideas every day!” But coming up with ideas, in general, is not the problem. The hard thing is to identify the ideas that work. Although we are all able to envision possible futures, nobody can be sure if an idea will actually prove to be valuable in the future. Only in hindsight do we know which ideas were great ideas in the first place.

Great ideas also don’t come to you in a flash of genius. You have to generate them. Research suggests that the first step to generate good ideas is to come up with a lot of ideas. If you generate about 10 to 15 ideas, you have a good chance of having generated a few good ones. What often happens, though, is that we tend to fall in love with the first idea that comes to mind. You also might have heard people say that the first idea is generally the best. But that is simply not true. First ideas are the most obvious ideas and often the most unoriginal and outdated. They might be based on a solution you already came up with long ago and that you now use instead of thinking about the real problem, for example. Or, it is an idea that many others would also come up with as their first idea. So when you start to evaluate which ideas are the most promising, try to be objective and generate as much insight into how and why an idea might work as possible. For example, by building quick prototypes. Researcher Justin M. Berg also found in an intriguing study that when we evaluate and rank our ideas by their potential creativity, it is not our favorite idea that will be the best one, but actually the second on the list. While the favorite idea might indeed work, the idea that participants thought was their second best, often proved to be more original and, ultimately, more successful. So if we were to build an idea machine, would it spit out the first or the second idea on the list?

The most interesting question, though: Would the machine generate all those ideas from combining seemingly unrelated knowledge from different domains? I would hope so. Because, as David Epstein brings out brilliantly in his latest book “Range”, the most groundbreaking and innovative ideas are often the ones that transcend the boundaries of one specific domain and ingeniously combine different concepts in new, previously unthinkable ways.


This is the 45th post of my 100 days of writing series. You can find a list of all posts here.


2 Webmentions