Exponential Growth and the NBN
Exponential growth is something we poor humans have some trouble with. It’s not a hard concept to understand, I’ll explain it quickly in a moment, but in the kind of prediction-making we all do day-to-day, it’s rarely brought into the equation. This effects all kinds of things, from high profile business decisions to political discourse.
Before we get to exponential growth, let me first explore everyday, vanilla growth. Growth can be measured as a positive change in anything at all - the reduction in size of a computer chip, the rise of wages, the output of an economy, efficiency gains, whatever. All of us understand that as the world goes along, from day to day, we are progressing and improving in all manner of different ways, and growth is the tool to measure, record and understand this progression.
Where things become exponential, is that growth occurs on top of existing growth. The percentage used is actually a percentage of the previous year, not the beginning year - this is the acceleration of growth. The best illustration of this acceleration is the 7% rule. If something, anything, grows, improves or increases by 7% every year, the output would have doubled over 10 years. A 7% annual increase becomes a 100% increase from the beginning over the course of 10 years. And then the following 10 years sees another doubling, so at the end of 20 years whatever you’re measuring is 4 times as much as when you started. The next 10 years? 8 times as much. 16 times as much. 32 times as much. Every 10 years, at only 7% growth.
Of course, not everything grows by 7%, but everything that increases will double sooner or later. This doubling becomes shorthand in discussions of exponential growth, and I think it’s important for us to consider in all our decision making about the future. How frequently does the amount of data used by mobile phones and wireless internet double? Well, as the telcos are finding out, it doubles about every year!
The NBN is such a project where exponential growth needs to be factored in, because the technology sector doubles so frequently that costs and time frames change substantially over the course of a long term project. Over the 8-10 year construction time frame for the NBN, you should be thinking in terms of about 3-4 doublings. These are doublings in terms of the reduction in the cost of equipment, doublings in the capabilities of technology, doublings in the reduction in size of chips and microprocessors.
Of course, such doublings don’t impact every aspect of the NBN build. Things like the cost of labour and the speed of physical construction don’t change so substantially. But the technological aspect of the NBN does change substantially, and I see that impacting mostly on where is considered too remote to be in the fibre area.
As you know, the NBN fibre will not reach 100% of every single household in Australia. Cue images of remote farmsteads, the neighbours a three day drive away just to get a cup of sugar. In reality, such situations don’t occur very frequently at all. Most remote communities are just that, communities, consisting of dozens of households and a few shops in relative proximity to one another. The costs in getting backhaul to these communities, constructing a POI building to house all the service providers hardware, and linking the houses in the community to the POI with fibre will reduce substantially over the next 8-10 years. That means that as the NBN progresses, what is considered “un-fibreable” will become closer and closer to that ultra-remote image in our mind’s eye.
The practical implications of doubling touch almost every sector, so when you’re thinking about the future, don’t think about what has happened in the past time frame, but think in terms of how many doublings you expect to occur over the future time frame. Even if you’re completely uninformed about how many doublings will occur in each sector, you’re still making a better prediction than when you’re thinking linearly.
