Most people think about e-waste when the impact computers on the environment is discussed. The powerful images of Asian and African villages surrounded by piles of tangled wires and broken circuit boards have lead to significant revisions of waste export policies. Lead has been removed from solder, and other heavy metals have been banned.
Society has an insatiable appetite for computers. We simply can’t get enough. The computer power that put men on the moon can now be found in a Furby doll. Your iphone is four times more powerful than a Cray supercomputer of the 1970s. Just what are we doing with all this power?
In many cases not very much! The Furby tells us that much. Of course there are plenty of counter-examples. There are shockingly few areas of society in which computing hasn’t lead to some kind of increase in efficiency. The internet is driving the informational revolution.
Yet all these circuits need power. Between 2000 and 2008, computer power usage in Germany (the only country where stats are easily accessible) doubled. Across the world the factor is undoubtedly much higher and projections suggest this increase will just carry on. Hundreds of millions of Chinese and Indians want their own PC.
As a fraction of total energy usage computers and all the other infrastructure like networks probably use around 10% of the total world energy supply. Precise numbers are actually extremely difficult to estimate and the subject of debate. Nonetheless, increasing power demand of any type puts a real strain on non-renewable resources. It’s no surprise that strong proponents of building larger computer farms (like those that power Google) want them to use renewable energy.
We might hope that making computers more efficient could get us out of this mess. But look at what’s happened. The computers of today are vastly more efficient than those of the 1970s. So it becomes easier to put them wherever you want. New possibilities open up that didn’t before.
This is an example of the Jevons Paradox. Making a technology more efficient often leads to more consumption of the “fuel” it requires by opening up new opportunities.
Economists understand this in terms of whether the demand for the service provided by the technology is elastic (so it can grow) or inelastic (fixed). If demand is inelastic improvements in efficiency lead to a reduction in fuel use.
But remember what we said about society’s demand for computing? It seems to be completely unbounded…! So it’s hard to avoid the conclusion that improvements in efficiency won’t save us, and running as much of this infrastructure on renewable sources as possible, to essentially take it out of the climate equation, is the only solution. But clearly that doesn’t work for your home PC…
E-waste remains a huge issue though. Easy solutions to the technology hardware consumption issue are not obvious.
But we don’t want to be all doom and gloom. One really interesting new idea, driven by the internet, is using data mining to conduct environmental monitoring. There is an enormous amount of data that can be mined for its environmental ramifications. Government planning information, monitoring, health, economic data and so on, all have ties, whether indirect or direct, to the environment.
The key to making this effective is to be able to use this data to determine when problems are coming. That might sound obvious, but finding problems and then doing something about them is not a logical process. Just look at climate change. We’ve known the basic science behind it for decades, it’s not debated by the scientists who work on it. What to do about it, how to act, well that is a whole other story.
But the idea of setting up a number of web crawlers to trawl through the internet for this information is incredibly powerful. It won’t be easy to do – experts will be needed to help understand the linkages (perhaps even neural networks could play a role).
You can be sure Google is thinking about this.