Wednesday, November 19, 2014

How Much Danger Does Google Pose to Other ISPs?

After Google Fiber, the notion that a major app provider might actually become an Internet service provider is no longer a possibility, but is a reality. The only issue now is how far that might extend, and which other firms might decide to do something similar.

Facebook appears likely to emerge as a satellite-based Internet service provider in Africa. And both Google and Facebook now own assets that produce unmanned aerial vehicles that could be used to supply Internet access.

Google is testing balloon-based Internet access for rural Australia, a test being conducted in conjunction with Telstra, existing Internet service providers have to be thinking about how far Google, Facebook and others might be looking at the ISP business.

Add to the that the fact that Amazon already is a specialized type of mobile virtual network operator, using AT&T’s mobile network to deliver content to Kindle devices.

There has been speculation that Apple might someday want to do something similar, perhaps becoming a provider of services that connect to any available mobile network network, something that has become a feature of the iPad.

Given the fact that, by perhaps 2020, 80 percent of all Internet access globally will use a smartphone, one has to wonder when that might become a focus for one or more application providers.

GoogleNet is Google’s vision to offer global, near-free Internet-access, mobile connectivity, and Internet-of-Things connectivity using a global, largely-wireless, Android-based, “GoogleNet,” according to Scott Cleland, of Precursor, a site with an admittedly “anti-Google” orientation.

Critics argue Google is following a business strategy of identifying markets with a valuable stream of consumer data, then creating an "open" or "free" product to induce adoption and “undermine the business model of existing market participants,” according to Fairsearch.org, an entity funded by Google competitors.

Once it gains dominance, Google then “closes” the market and excludes competitors, Fairsearch argues.

One does not have to accept the premise of the argument to agree that something important has happened. Google effectively disintermediates and commoditizes the direct relationships Internet or communications or entertainment suppliers have with their customers.

And telcos, cable companies and Internet service providers might have to worry more than they used to about Google. Initially, one might argue, Google was about businesses built on bits in virtual worlds. That is no small matter, as Google arguably has created rival communication products that displace products supplied by communications companies.

But Google now is moving into different realms, including “atoms in the physical world.” including Google Fiber, an Internet service provider operation that competes directly with cable TV company and telco high speed access and video entertainment products.

Beyond that, Google has invested in, and is testing, high-­altitude Wi-Fi balloons, and unmanned aerial aircraft that might also be used to support Internet access.

All of that, building on earlier Google investments, creates at least the potential for a “global Internet access” capability that would disintermediate other existing Internet service providers, as Google Fiber does in a growing number of U.S. cities.

Google bought Skybox Imaging (satellite technology) and plans to spend $1-3 billion on “180 small, high capacity satellites at lower altitudes than traditional satellites” to enable two-way Internet access.

Google also bought Titan Aerospace, a supplier of solar-powered, high-flying drones. Project Loon likewise is testing use of balloons for Internet access, most recently inking a deal to test them to provide Internet access in Australia, working with Telstra.

Google also operates its own global undersea network, including investments in four cable systems.

The issue is how widely Google’s ambitions might extend.

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...