The conversation at iLaw yesterday repeatedly came back to the danger of choke points on the Internet. The morning discussion led by Professor Yochai Benkler focused upon the broadband last mile. As Professor Benkler pointed out, given the high infrastructure costs involved in hardwired delivery of broadband to the home/business, the most price efficient way to solve the last mile problem would be to create a "perfectly regulated" monopoly. But since our government has proven a relative failure at creating "perfectly regulated" monopolies, we are left with the alternative method -- price competition. The government has gone to great lengths to promote an alternative to the incumbent RBOC monopoly (with, of course, the RBOCs kick and screaming) and, as a result, we now have a duopoly. The last mile broadband problem is now solved by Cable and DSL, to which Professor Benkler asks "do two pipelines a competitive market make?" The question implies the answer -- no.
Professor Benkler's proposed solution (and, of course, the proposed solution of innumerable startups in the space) is an Open Wireless Network. By taking advantage of unregulated spectrum and ad hoc networks, it is clearly possible to create an alternative to the last mile duopoly. The question at hand is what technologies will emerge to create the greatest possible network intelligence at the edges. And, perhaps more importantly, what will the existing duopolists do to maintain their control of the networks as these potential alternatives evolve.
Professor Lawrence Lessig took over where Professor Benkler left off. He was far more explicit in his concern about the choke points on the Internet. It is Lessig's strong belief that the best possible network is not optimized for any particular use but rather is optimized for all uses and is non-discriminatory with respect to content, source or client -- Lessig calls this the End to End (E2E) Network. Because the E2E Network is fully agnostic, Lessig argues that it has 3 important benefits: it will maximize competition, minimize predatory behavior by market leaders, and will not require massive capital expenditures to get to scale (consumers will fund the growth). Given that, any threat to the E2E Network should concern end users of the Internet.
If the E2E Network is heaven, Lessig argues that "we're on our way to hell." So what are the threats to a broad, fully-agnostic E2E Network? They come at every level. Policy based routing is a clear threat to the E2E Network. Large network operators want to decrease their costs and increase their profits by routing traffic preferentially based upon the content of the packets, the sender of the packets, the recipient of the packets, etc. The result is that the network is no longer agnostic -- one packet may have to take a back seat to another based upon what it is, where it's coming from or where it's going. Moreover, policy based routing has an impact at the content layer. As media companies grow larger and more consolidated, many of us express our concern about increasing corporate control over the messages being delivered. The answer to this concern in recent years has been, "don't worry about it -- we always have the Internet to distribute alternative messages." But as media companies become increasingly integrated to include elements of the network itself (e.g., AOL and Time Warner merging), there is a threat that all content will not be treated equally (e.g., Time Warner content will be delivered preferentially on the AOL network). Lessig sees the inequality of the network only increasing over time.
Lessig proposes that it is in everyone's interest to resist the attack on the neutrality of the Internet at all layers. At the physical layer, it is necessary to promote open access, as discussed by Professor Benkler. At the logical layer, it is necessary to create equal protection for packets -- networks should not be able to prioritize certain packets over others. And at the content layer, media consolidation should be resisted and, in particular, the integration of content and delivery should be forbidden. All three layers are clearly interrelated. If even one layer is no longer neutral, the entire E2E Network and all of its advantages is destroyed.
What is interesting to me about Lessig's discussion of the E2E Network and the value of its openness is how frequently, in my capacity as a VC, I see technologies designed expressly to alter the neutral nature of the Internet. For example, I see packet and content routing technologies of all sorts. And while I can understand how these preferential routing schemes might disadvantage some bits, the entrepreneurs developing these technologies are able to make strong arguments for their advantages in many other respects (cost, quality of service, network efficiency, etc.). So it seems to me that Lessig is right -- we are very much "on our way to hell." While the greater good may require a truly neutral network, the ability to maintain a pure End to End Network will ultimately be driven by economics. And I suspect that given the economics around network creation and maintenance and content creation and delivery, that there will be lots of compelling technologies that will be developed and adopted to increasingly alter the End to End nature of the Internet.