I’m grateful to Frank Coluccio for pointing me to George Gilder’s Fibersphere piece from 1992. I’m spending time these days trying to figure out what an open access fiber optic network would look like.
It’s astonishing what abundance could be unleashed by combining a few components: dark, unlit fiber; a coordinating entity that could ensure that different providers were using different wavelengths to communicate across that fiber; a small box with power and air-conditioning for that coordinating entity to operate in; and modulation schemes taking advantage of different frequencies. That’s it – and then you’d have potentially hundreds of thousands of “channels,” each possibly provided by a different vendor, each carrying the communications of thousands of knowledge-workers. It really would be the end of scarcity. Transmission would be the cheap element – device-manufacturers and coordinating entities would have to leap into innovation mode.
The way Verizon has built up its fiber network doesn’t allow easily for this kind of unbundling, for many reasons. The top reason is that the potential interconnection points (“splitters”) are out in the field, without power or air-conditioning, so no one else can interconnect there. Also the hardware, software, and protocol standards used by this network are hard-wired to Verizon. You could interconnect right near Verizon’s central office, but you’d need a lot of cooperation from Verizon.
I’m just beginning to understand that the architecture chosen by Verizon makes it difficult (if not impossible) to retrofit abundance and open access into their network. The company gets a lot of credit for bringing more fiber to more people. But what tradeoffs are implicitly being made?