The article suggested that the current structure of the telco industry would have to change "the economic basis of your mobile telephony service in 2019 is goin

The article suggested that the current structure of the telco industry would have to change "the economic basis of your mobile telephony service in 2019 is going to be unrecognizably different from that of 2009". I think there is something to this argument. I worked for a telco OEM in 2005. Then it was possible to buy the hardware to build an "NFL" city backbone in the US for about $50 million. Adding fiber, end user access, operations and installation would clearly add much to this but it is likely that within the next decade that a wireless network with a significant nationwide footprint could be built by a company capitalized with a few billion dollars while not carrying the legacy operations costs of the incumbents. This company would likely have no problem in driving network costs to the ground. This almost happened during the first internet bubble and could this time around follow from the current cloud computing trend. Adding compute services on top of network services could help to regain some of the margins lost by commoditization of the network.