Ever since Google last week announced a $4 billion secondary stock offering, I've been enjoying speculating about what the company is going to do with the cash. Google already has $3 billion cash on hand -- and is generating about $300 million additional cash each quarter -- so its total on hand after the secondary offering will be more than $7 billion. None of Google's announced software projects even come close to requiring such a fantastic sum of money, and it doesn't seem like Google would raise another $4 billion in order to manipulate its stock price in the short term. For fun, I have started to imagine that Google will offer a free derivative of the original Teledesic communications satellite constellation. (Hey, at least I'm not imagining that Google is investing in a space elevator, like this journalist!)
The history of Teledesic began in the mid-90s. With the backing of Bill Gates, Craig McCaw, and some Saudi investors, Teledesic proposed to launch an $8 billion constellation of about 850 satellites that would provide internet service globally. At first, the constellation was to be launched in Low Earth Orbit so that communication latency was minimized. Another feature was that the satellite network would be self-contained, in that the satellites distributed data among the satellite constellation, rather than from each satellite to a network of ground stations. While no production hardware was manufactured, the project obtained regulatory approval worldwide, an impressive accomplishment.
As the 90s progressed, the cost of fiber optic communications decreased as its capabilities increased, making ground-based solutions more attractive than Teledesic's self-contained satellite solution. Meanwhile, Teledesic changed its configuration to fly higher up in space with fewer satellites. In the early '00s, the Teledesic project was shelved, even though, as Wolfgang Demisch testified this year, the project still makes some sense as a business proposition.
Over the past year, there have been reports of Google buying lots of "dark" (i.e., unused) fiber optic cable, presumably to be used to keep up with the growing bandwidth demands of its array of services. After the telecommunications bust of the late 90s, the dark fiber market is depressed, and Google can pick up dark fiber cable cheaply. Indeed, I wonder if Google can pick up the bandwidth so cheaply that it could let everyone in on its economies of scale, such as it is doing by providing gmail for free. Google would not need to provide the extremely expensive "last mile" of cable to the home if it had a satellite constellation such as Son of Teledesic. Google would only need to have a network of fiber optic ground station hubs that would communicate with each satellite in the constellation. From there, GoogleNet is born as a free wireless broadband internet connection worldwide to anybody with a computer and a small satellite antenna.
In a recent previous post, I speculated that Larry Page and Sergey Brin will be among the next round of investors for SpaceX, since they have been interested -- at least in passing -- in putting Google into the rocket launch business. SpaceX's new rockets would make such a constellation a lot cheaper to put on orbit than was contemplated for Teledesic, and a Son of Teledesic constellation would provide ample steady business to SpaceX, as it wrings the costs out of space launch.
But how would Google make money out of such an arrangement? Using GPS as an example, it has become clear that big space infrastructure that provides a free service can make money for someone. Also, it is becoming clear that as the internet grows, Google profits rise in like measure. However, the next big game in search is localization. With a satellite service, Google could pinpoint to the foot worldwide where the user is located and provide relevant search results based on that knowledge. There are billions of dollars in profit locked up in knowing location.
In thinking about this possibility, I have come up with some questions to which I don't know the answers, but would be interested in discussing.
(1) Would GoogleNet be able to use free spectrum, like Wi-Fi/Wi-Max are doing?
(2) How much total bandwidth would be available? Would the available bandwidth need to be rationed by charging people for use of the service or by use of individual quotas?
(3) Would Internet Service Providers (ISPs) be killed by GoogleNet, or would GoogleNet be more of a supplemental service?
(4) What would be the services that GoogleNet would facilitate? Worldwide Voice-Over-IP wireless phone service, for instance? A worldwide satellite Television-Over-IP service?
(5) How much of an open platform would GoogleNet provide? Would GoogleNet allow P2P file sharing, for instance?