This article from GigaOm is interesting. Basically, the article explains that carriers are benefiting from consumer interest in social networking, which has migrated to wireless platforms. Interestingly, this was not an application carriers had in mind when they made investments in 3G infrastructure. How will carriers justify their upcoming investments in 4G?
The nature of telecommunications is that large investments must be made before revenues can be realized. The telecom network has to be largely built out before it becomes valuable to consumers. As a result, investments have traditionally been conservative and tied to applications. The telephone network was tied to voice communications and associated services, etc. A significant exception was the Internet, which was built with no particular application in mind, but then it was also built with government funding. It wasn't privatized until a commercially interesting application set emerged (email and the web).
The phenomenon that this article describes indicates that unexpected applications (social networking) are a significant new revenue source for carriers. Will this change their investment model? The other thing that is addressed, though not explicitly, is that these applications require collective action from independent entities whose interests are partly common and partly at odds. Applications need broadband and want it to be cheap. Social networking works better with smart phones, which are more costly than "regular" ones. So manufacturers need joint marketing agreements with carriers, yet they want the ability for users to control applications and configuration.
If high-revenue end user applications can no longer be predicted or managed by telecom carriers, how will their investments be justified? It seems that carriers need a strategy and a set of tools for managing investment risk. Physical commodities have futures markets for this purpose.
One of the reasons why I am interested in markets for capacity (most recently spectrum) is that derivatives (such as futures) are possible, which allow for industry restructuring. In the absence of explicit risk management, carriers will integrate applications with carriage, which leads to "network neutrality" concerns.