Apple

Copying is Cheap, Variation and Selection are Expensive and Painful

The internet is a copy machine.

Kevin Kelly

Copying is cheap and enjoyable. Copying ideas consists of activities like TV watching, web browsing, news reading, attending a lecture or sermon, lunch conversations with friends and colleagues, listening to the radio, etc. Copying genes occurs through an even more ancient pastime. Copying is important in maintaining and disseminating[1] good[2] ideas and genes.

Innovation[3] occurs in the variation and selection phases. But variation is expensive, in nature -- random mutations of genes through irradiation and random mixing of genes through reproduction -- pretty unreliable and time consuming. And variation is expensive in design/R&D -- trying lots of ideas and actually implementing them costs time and money. Selection is the only way to cull genes (or ideas). It can be painful: death of individuals or extinction of species in nature, and trial and error in idea space constitutes letting go of pet projects and ideas in design, writing, programming.

Ideas are cheap (a dime a million) because they don't require any work, really. Ideas can be produced by just stringing together some disjoint thoughts, throwing out a dozen words, or even mistaking some thought for originality -- "the fine art of remembering what you hear but forgetting where you heard it" (Laurence J. Peter).

The work comes in concretely building something, which comes from a position of experience, skill, resources, know-how, or paying for some combination of these. World class designers, artists, musicians, programmers[4] and creative workers try out ten variations and eliminate the nine that don't work (some of this occurs invisibly in the creative individual's mind, but it is always happening). They then build off that concrete idea with ten new variations, implementing all ten painstakingly (possibly as outlines, concept art or paper prototypes, but always concretely and pedantically), and then mercilessly throwing away the nine worst ideas. Note that the designer never has to try a hundred or a million implementations at once to reach that "one in a million idea", just one in ten then another one in ten variations on that idea, then ten more, etc. -- that gets the designer somewhere near, without the million parallel attempts. That's one hundred fully realized ideas, and 99 of those are sent to the bin. (And this ignores the powerful effect of copying and varying based on other groups of designers and innovators out there, disseminating there ideas in parallel, competing and collaborating.) The idea of iteration gives intuition for why evolution and good design seem so compelling, inevitable and obvious yet uncanny and sometimes improbable.

Like compounding interest, the exponential power of building on previous gains helps explain why innovators continue to innovate, while those who simply copy the results of the innovation process continue to do only that -- copying results, never contributing new innovations. Cargo cult innovators fail to see what distinguishes the real innovators. Copying is cheap and may work for companies, at least in the short term, but in the long term, like a collapsing ecological niche can wipe out an entire species, outside forces in the market will select against corporations that can't play the variation and selection half of the game.


  1. From the Latin word meaning spreading seeds in all directions. ↩︎

  2. Where "good" is a circularly defined as "the ideas or genes that stick around". Seth Godin: "Ideas that Spread, Win" ↩︎

  3. Where innovation is defined as the to ability improve ideas/genes, to make them better, to make them more good, fit or "sticky", within a context or environment. See The Information Challenge on how information could be created by natural selection. The gist: the environment culls (some) noise out of the genome and what is left is (some) information on how to survive in that environment. Similarly, innovative companies cull noise out of the business/design space and what is left is information on how a product/service/company can survive in a given market. ↩︎

  4. For an example of Apple's iterative design process, see http://devimages.apple.com/products/video/wwdc09/wwdc09iphonesessionlist.pdf session 128, Prototyping iPhone User Interfaces: "Learn how to turn your personal vision into an elegant iPhone application design. Explore your ideas through rapid prototyping and experience the iterative design process that leads to a truly innovative user interface. With some cool tricks and a few lines of code, see how a working prototype provides insight far beyond a static mockup. Watch your design evolve from good to great." Note the word "evolve". Also I can't find the Stanford iPhone Application Development course lecture on iTunes U but an Apple engineer went over this iterative design process in detail. ↩︎