Web Doomsday: How Realistic Is It?

By Peter West • 13th November 2013 20:30

Originally posted for the
Web Science COMP6044 module, University of Southampton

Despite our modern-day dependency on services offered through the web, few of us have given thought into what the consequences to us would be if the web were to disappear. It might simply seem unlikely, and not worth planning for. However, there are many potential causes, intentional or not, for widespread loss of access to the web. I will outline some of these in this post and will argue that it is realistic and that we as a society should be prepared.

The size of the technical infrastructure required to deliver the web to our fingertips is huge. The larger a system, the more points of failure. Because the system is stretched over a large geographic area, where there are climate extremes, natural disaster is a large risk. Storms, earthquakes and erosion can easily break vital equipment. Such a large system also lead to scope for technical failures.  The Northeast United States blackout of 2003 left 55 million people without power, for many as long as two days. This was caused by a software bug.

Much of the internet infrastructure was designed before there was significant demand for the web. Technical limitations have already affected the performance of the web. We ran out of IPv4 addresses (which each person requires to connect to the internet) in 2011, and ISPs have been slow to adopt IPv6 to solve the problem. The technical infrastructure may also be prone to attack, whether it be through cyber warfare (e.g., military assault) or malicious intent (e.g., hacking commercial infrastructure).

There may be political and commercial motivations behind changing the way we can access the web. The Great Firewall of China prevents people in China from accessing a huge portion of the web. Commercial motivations include ISPs prioritising particular services (e.g., Comcast’s proposals to degrade high-bandwidth services such as NetFlix), or companies choosing to change of discontinue services (e.g., Google withdrawing Reader).

Whether it be storing our photos with a cloud service or becoming reliant on a social network for communicating with friends, depending solely on these presents a risk to ourselves, in that loss of the web or these services will potentially be detrimental to our lives. The next post will be a case study into the Northeast blackout of 2003, in which a huge power cut led to widespread panic, and will consider the societal and economic impact of the event.