America Online is America's largest Internet service provider. Twenty-two million members get to the Internet through AOL. If it were a state, AOL would rank second in the nation in population, behind California. The company has a market capitalization of $125 billion--a bit less than the GDP of Denmark. And with its proposed purchase of one of the largest and most powerful media giants, Time Warner, many are beginning to ask, Should we worry about AOL the way the government worries about Microsoft?
Maybe. But to see why, we've got to look at something politicians don't talk about much--architecture.
At the core of the Internet is a principle of design described by network architects Jerome Saltzer, David P. Reed and David Clark as "end-to-end." The principle of e2e says, Keep the network simple, and build intelligence in the applications ("ends"). Simple networks, smart applications--this was the design choice of the Internet's founders.
The reason was innovation. Simple networks can't discriminate; they are inherently neutral among network uses. Innovators thus don't need to negotiate with every network owner before a new technology is available to all. The burden on innovation is kept small; innovation is, in turn, large.
AOL has benefited from this neutrality. Because regulators breaking up AT&T forced the telephone company to respect e2e neutrality, consumers of telephone service have always had the right to choose the Internet service provider they want, not the ISP the telephone company is pushing. This built an architecture of extraordinary competition among ISPs. AOL, by delivering what consumers want, has prevailed in this competition.
All this may change, however, as Internet access moves from narrowband (telephones) to broadband (predominantly cable). Cable companies are not required to respect e2e; they are allowed to discriminate. Unlike telephone companies, they get to choose which "new ideas" will run on cable's network. They get to block services they don't like. Already many limit the streaming of video to computers (while charging a premium for streaming video to televisions). And this is only the beginning. The list of blocked uses is large and growing.
This trend worries many. AOL fought restrictions when AT&T (after buying a gaggle of cable monopolies) proposed them. But now AOL, by buying Time Warner, is buying its own cable monopolies. And many are worried that AOL will forget its roots. Will the temptation to build its broadband network to protect itself against unallied content and new innovation be too great? Will AOL, like every other large-scale network that has controlled content and conduit, pick a closed rather than an open architecture? Will AOL become what it eats?
Compromising on the principle of e2e would weaken the Internet. It would increase the costs of innovation. If to deploy a new technology or the next killer application--like the World Wide Web was in the early 1990s or gadgets to link the home to the Net may someday become--you first have to negotiate with every cable interest or with every AOL, then fewer innovations will be made. The Internet will calcify to support present-day uses--which is great for the monopolies of today but terrible for the future that the Internet could be.