Why Does Technology Improve Exponentially?

How to think about the forces underpinning the great transformations of our time


If like me, you spend a lot of time wading through the internet looking for interesting ideas on technology and innovation it doesn't take long before you come across the concept of exponential change. It's ubiquitous, popping up in everything from marketing blogs to engineering manuals. For a certain brand of futurist, the idea is such a powerful one that it's been elevated to an article of faith; the first principle from which all other conclusions are derived. This kind of thinking is pervasive.

Naturally, when it comes to sacred cows my first instinct is to critically interrogate where they come from (and yes, I'm aware I have a problem with authority. I'm working on it). So I've been doing a lot of thinking about the idea of exponentiality and more specifically about why it seems to be so common in the tech world. Very often you'll see someone present a slide like the one above in a talk or presentation, and give an explanation of why exponential growth is such a powerful force. We've certainly been guilty of this in our own work with Future Crunch. What's always bothered me though is that what's missing is an explanation of WHY technology changes the way it does - why is it that exponential technology laws exist in the first place? 

One way of thinking about this is that the word technology is really just a fancy way of saying tool or technique. The purpose of which is to help humans to do something quicker, more efficiently or with greater safety. So when you have a technological breakthrough it means that we've invented a new tool or technique for achieving an intended purpose. Here's the thing though. New technologies don't necessarily have to replace old ones. That idea has been popularised by a recent spate of literature on what's popularly known as disruptive innovation. I've got a number of reservations about this kind of thing, most of which are covered in this excellent New Yorker article by Jill Lepore. More than anything though, I just think it's a glorified rip-off of Joseph Schumpeter's ideas from the 1970s. Schumpeter argued that capitalism exists in a state of ferment he dubbed creative destruction, with successive waves of innovation destroying established enterprises and yielding new ones. The more I learn though, the more I think that the wave metaphor is problematic, because it suggests that new technologies come along and 'wash away' everything that existed before. 

That's not how technological change works though. Yes new technologies come along, but they don't necessarily replace everything that's come before them. 3-D printers might offer a new way of manufacturing chairs but a number of Chinese mega-factories and individual artisans will still continue to make chairs the way they have before; albeit for a smaller market. As a result the market becomes more segmented (and hopefully more efficient). In other words technological innovation adds a new layer of complexity to a particular field or industry, rather than replacing it altogether. Less of a wave, and more of an archeological dig or a cross-section through sedimented rock. Even more importantly each addition to this civilisational pile of tools and techniques interacts with everything that's come before it. And that increases the number of ways you can achieve an outcome exponentially. Let's go back to the example of a chair to show you what I mean by this: 

Start with a wooden chair. It's the only way chairs have ever been produced, and everyone uses the same sorts of hand tools to get it done (although there may of course be varying levels of quality depending on the skill of the craftsman). 
1 technology (wood) = 1 possible product
Someone then comes along and invents a new technology to create steel chairs. In this new world, craftsmen now have three options for producing a chair. They can continue to use wood, since there will always be a market for people who appreciate the way it looks and feels. They can switch to the new, stronger material which allows them to make chairs more cheaply or with greater functionality. Or they can produce a hybrid, combining both technologies such as a wooden seat with steel legs. 
2 technologies (wood / steel) = 3 possible products (wood / steel / wood-steel)
The next breakthrough is plastic. As before, craftsmen can now employ the the new technique to make better or cheaper chairs. Or they could stick to one of the older techniques, involving steel or wood. Or they can produce hybrids using any combination of all three technologies at once. 
3 technologies (wood / steel / plastic) = 7 possible products (wood / steel / plastic / wood-steel / wood-plastic / plastic-steel / wood-plastic-steel)

What's happening here is that a linear increase in technological breakthroughs (1>2>3) results in exponential changes (1>3>7) in the number of different products or outcomes that can be achieved. The key thing is this idea of layering - everything that came before still exists in the new paradigm, side by side or jammed in between or stuck on at some kind of crazy side angle. And as more outcomes become available they open the way to ever faster rates of technological breakthroughs; the entire process feeds back on itself. It may have taken 2000 years to get from wood to steel and then another 400 to get to plastic; but it's only another 40 before you get to 3-D printed chairs made from polycarbonate or graphene. 

This doesn't mean of course that exponential growth is inevitable. The way these innovations play out is messy and complicated. In some cases previous technologies may happily co-exist side by side with new ones, in others they may disappear altogether or be relegated to such a niche that they effectively stop influencing the next round of innovation. That process is messy and unpredictable and it means that our future world is heavily, fundamentally layered; that problems have no simple solution that a single technology plonked on top will fix. Technological progress may tend towards exponential growth but it's not an article of faith. Rather it's the interplay of the complex social systems of economics, human geography, history and politics that come together to determine how the next layer fits onto (or sits partially side by side) with everything that's come before it. That complexity is both deeply exciting, and hopefully, humbling.