“If you put a gigabit in people’s homes they will be inspired to find new ways to use it," says Kevin Lo, Head of Google’s fiber access program. "We have no idea why you need a gigabit today, but when we all had dial up you could not possibly imagine watching video over them."
"It’s not about doing email faster, it’s about doing those new things that you don’t do today," he says. Video is the obvious practical application that could use bandwidth of that sort. Some of us might question whether new ways to watch TV is really such a huge innovation, though.
Unicast entertainment video, especially the same sorts of content you can watch on a subscription TV service, is a pretty dumb way to use bandwidth, some of us would argue. As any engineer will tell you, there always are multiple ways to solve any computing problem. You can process locally, or process remotely, substitute local processing for bandwidth, or bandwidth for local processing power.
If what a user wants to do is watch professional video on demand, then local storage such as provided by Tivo is a reasonable solution. Unicast is better suited to relatively low-volume types of content, or interactive content.
What also remains to be seen, as Google builds out its fiber to the home test markets, is whether Google really will try to build symmetrical 1-Gbps networks, as it has said, or will, for cost reasons, do something a bit less grandiose. The other angle is whether Google really will try to offer such bandwidth at prices roughly comparable to what telcos and cable companies might charge for 10 Mbps to 20 Mbps services.
It always has seemed unreasonable to think that Google has any magic answer to the costs of building such infrastructure. Sure, it will always make sense to choose test beds where aerial plant can be built, because that is cheaper than underground construction, unless there is universal duct space available, allowing any new provider to simply pull a new set of cables into the existing conduit.
So if construction cost is not amenable to significant cost reductions, one would have to look to revenue to make the business case. But here again, it has seemed unlikely that a sustainable business case can be built solely on "broadband access," in competitive markets, where the other contenders might have multiple services to sell.
Sure, Google is running science experiments, not trying to create a self-sustaining business. But you also have to wonder how much actual application innovation can happen if such isolated test cases.