In his new book, How Google Works, Eric Schmidt argues that “we are entering … a new period of combinatorial innovation.” This happens, he says, when “there is a great availability of different component parts that can be combined or recombined to create new inventions.”
For example, in the 1800s, the standardization of design of mechanical devices such as gears, pulleys, chains, and cams led to a manufacturing boom. In the 1900s, the gasoline engine led to innovations in automobiles, motorcycles, and airplanes. By the 1950s, it was the integrated circuit proliferating in numerous applications. In each of these cases, the development of complementary components led to a wave of inventions.
Today’s components are often about information, technology, and computing.
Would-be inventors have all the world’s information, global reach, and practically infinite computing power. They have open-source software and abundant APIs that allow them to build easily on each other's work. They can use standard protocols and languages. They can access information platforms with data about things ranging from traffic to weather to economic transactions to human genetics to who is socially connected with whom, either on an aggregate or (with permission) individual basis. So one way of developing technical insights is to use some of these accessible technologies and data and apply them in an industry to solve an existing problem in a new way.
Regardless of your business there is a core of knowledge and conventional wisdom that your industry is based upon. Maybe it’s logistics, maybe it’s biology, chemistry or storytelling. Whatever that core is, “that’s your technology. Find the geeks, find the stuff, and that’s where you’ll find the technical insights you need to drive success.”
That’s also the area to look for — where conventional wisdom might be wrong. What was once common sense becomes common practice. When everyone agrees on some fundamental assumption about how the industry works, the opposite point of view can lead toward disruption.
Another possible source of innovation is to start with a solution to one problem and then look at ways to use the same solution on other problems.
New technologies tend to come into the world in a very primitive condition, often designed for very specific problems. The steam engine was used as a nifty way to pump water out of mines long before it found its calling powering locomotives. Marconi sold radio as a means of ship-to-shore communications, not as a place to hear phrases like “Baba Booey!” and “all the children are above average.” Bell labs was so underwhelmed by the commercial potential of the laser when it was invented in the ‘60s that it initially put off patenting it. Even the Internet was initially conceived as a way for scientists and academics to share research. As smart as its creators were, they could never have imagined its future functionality as a place to share pictures and videos, stay in touch with friends, learn anything about anything, or do the other amazing things we use it for today.
Schmidt gives his favorite example of building upon a solution developed for a narrow problem.
When Google search started to ramp up, some of our most popular queries were related to adult-oriented topics. Porn filters at the time were notoriously ineffective, so we put a small team of engineers on the problem of algorithmically capturing Supreme Court Justice Potter Stewart’s definition of porn, “I know it when I see it.” They were successful by combining a couple of technical insights: They got very good at understanding the content of an image (aka skin), and could judge its context by seeing how users interacted with it. (When someone searches for a pornography-related term and the image is from a medical textbook, they are unlikely to click on it, and if they do they won’t stay on the site for long.) Soon we had a filter called SafeSearch that was far more effective in blocking inappropriate images than anything else on the web—a solution (SafeSearch) to a narrow problem (filtering adult content).
But why stop there? Over the next couple of years we took the technology that had been developed to address the porn problem and used it to serve broader purposes. We improved our ability to rate the relevance of images (any images, not just porn) to search queries by using the millions of content-based models (the models of how users react to different images) that we had developed for SafeSearch. Then we added features that let users search for images similar to the ones they find in their search results (“I like that shot of Yosemite-go find more that look just like that”). Finally, we developed the ability to start a search not with a written query (“half dome, yosemite”), but a photograph (that snapshot you took of Half Dome when you visited Yosemite). All of these features evolved from technology that had initially developed for the SafeSearch porn filter. So when you are looking at screen upon screen of Yosemite photos that are nearly identical to the ones you took, you can thank the adult entertainment industry for helping launch the technology that is bringing them to you.
How Google Works is full of interesting insights into the inner workings of a company we're all fascinated with.