February 19, 2010

Future of the Internet IV

Part 4: A review of responses to a tension pair about the evolution of the architecture and structure of the Internet: Will the Internet still be dominated by the end-to-end principle?

The structure of the Internet

q4: Respondents hope information will flow relatively freely online, though there will be flashpoints over control of the internet

Respondents were asked to explain their choice and “note organizations you expect to be most likely to influence the future of the Internet and share your view of the effects of this between now and 2020.”

A number of respondents pointed out that some of the key issues related to the functioning of the internet and its architecture were laid out by Jonathan Zittrain in his “The Future of the Internet—And How to Stop it.”2 Several said their views matched Zittrain’s: that the internet’s basic open character and its ability to foster innovation are threatened by “tethered” appliances and applications that are controlled by vendors.

What follows is a selection of the hundreds of written elaborations and some of the recurring themes in those answers:

There is too much good history and good experience with the end-to-end internet to see it largely overturned. Openness has its own virtues and those who resist it will fall behind those who enable it. Users will rise up if there are too many restrictions that get in the way of the information they want and the content they want to create.

“The Net users will band together to keep the Net open. They will continue to choose open over closed and gated.” – Jerry Berman, chair, Center for Democracy and Technology

“There is too much at stake to allow intermediaries to control the pipe.” – Peng Hwa Ang, director, Singapore Internet Research Center

“Incumbent network operators in some nations will succeed in asserting increased control over applications, but, in the long run, they will be at a disadvantage. This will vary from nation to nation, and those which view the Internet as basic infrastructure and act to balance public interest and quality of life along with the return on investment of network operators will be at an advantage in the future. This is related to the issue of ownership. The question is not whether we will have ubiquitous high-speed networks in the future, the question is who will own and control them — private corporations, government, users? Ownership/control will be disbursed among a mix of organizations each at different network levels.” – Larry Press, California State University, Dominguz Hills

“There are large numbers of people working to protect the end-to-end principle within the confines of organized structures (ICANN) and outside of formal venues (alternate roots).” – Elaine Pruis, VP for client services of Minds + Machines and ICANN participant

“It seems to me inevitable that nation states will attempt to exert more control over the Internet. However, I think that these will be relatively small changes, so that the internet will remain relatively free.” – Hal Varian, Google, chief economist

“This will be an ongoing debate, particularly when traditional organizations see the Internet encroaching on their legitimacy and relevance in the Internet Age. These groups will flail around to protect their business models and perceived relevance, but there will be equally powerful capabilities emerging from the Internet community that will break through/counter those new controls/restrictions on the flows of information.” — Richard Forno, Software Engineering Institute, Carnegie Mellon University

Those who took the opposite view were not necessarily happy about it, but they argued there are too many powerful forces pushing towards more control of the internet for the end-to-end principle to survive. Governments and businesses have all kinds of reasons to control what happens online.

“Given events in China and Iran, I am going to take a rare (for me) pessimistic position. The forces of central control, politically and economically, are moving to recentralize the power they lost when the Internet grew explosively. The net neutrality debate in the USA seems to temporarily have restrained the cable and telcos from exerting centralized control over the architecture, but who knows what will happen politically with future administrations? Unless a sufficient number of people resist, I see more and more control and intermediation being forced upon us.” – Howard Rheingold, author of several prominent books on technology, teacher at Stanford University and University of California-Berkeley

“I hope for the egalitarian selection I’ve made. I fear that the greed of the fewer and fewer, more and more powerful conglomerates, cartels, multi-nationals and monopolies, will allow them to use their power over government to choke the power of end-users to access and share the powerful benefits of different-time, different-place, near-’instantaneous’, egalitarian/equalitarian communications.” – Jim Warren, longtime tech entrepreneur and activist

Much as I want the end-to-end principle to remain I see it dying. Most users today, and most vendors of network services, perceive the net as a system of applications not as a system that transports packets. I believe that the internet is headed towards being a ‘lumpy’ network more like the several mobile phone networks in the U.S. than the uniform internet of today. There are a lot of pressures to drive this lumpiness – We’ve got the desire of vendors to lock in customers, we have national competitions and firewalls, we have resource scarcity (such as IPv4 addresses that are driving the net to partition via network address translation, NAT, devices) [PS, I believe that IPv6 may prove a dud], reaction to excessively heavy regulatory systems such as ICANN, etc.” — Karl Auerbach, Chief Technology Officer, InterWorking Labs, Inc

“The locked-down future is more realistic as things stand now. We’ve got a very cautious government, an international movement towards greater control, and a pliant public. I wish this wasn’t the case.” – Susan Crawford, former member of President Obama’s National Economic Council, now on the law faculty at the University of Michigan

As the internet becomes ubiquitous and increasingly important in commerce and politics, it will become increasingly important and profitable to control it. Control will be exerted through control of systems and architectures, networks, points of access, platforms for sharing, and content. ‘Dedemocratization’ processes are currently in the ascendancy in the US and other nations and these political trends are conducive to increased control by governments and corporations over all aspects of the internet. As in politics, the democratization (and dedemocratization) of the internet is not linear. But the trend over the next decade will be dedemocratization. We already see this in the attempts by China and Iran to control access and content and in the so-called Homeland Security legislation in the US to monitor internet activity.” – Benjamin Mordecai Ben-Baruch, market researcher and consultant to non-profits

There will be alternative networks for companies and individuals that prefer to have a more controlled environment for sharing and consuming content.

“A number of companies have built global IP networks that are not the Public Internet, but provide similar capabilities to companies that don’t want their traffic to mix with the Internet. These IP networks often offer QOS services, MPLS VPNs, MPLS pseudowire private circuits, VoIP services, telepresence services and so on. Soon they will offer secure cloud computing services as well, competing with Amazon and Google but not on the Public Internet. This activity will secure the Internet’s end-to-end principle on the Public Internet, because anyone who has a technical or business reason to want something different, can get their needs met on the private internets. This reduces the pressure on the Internet and the end-to-end principle remains secure. In fact, IPv6 means that you will see even more end-to-end capabilities on the Internet of 2020.” – Michael Dillon, network consultant for BT

Regional differences will become more pronounced.

I think on the whole the way the internet works will be mirrored by the country in which it is accessed. Although it may well be possible to access anything from anywhere online, in reality the physical requirements such as routers and access points will always mean that locality plays some role in what is and is not possible. We will start to see more regionality of control, and greater development of specific language based ‘sub-webs’ as the predominantly western centric web becomes less important to the Asian and African continents, and they develop their own dominant sites along with their own culturally driven methods of control and administration. However the vast nature of the web will still ensure that the majority of web content remains freely available, following the end-to-end principle.” — Rich Osborne, Web Innovation Officer, University of Exeter

The future will produce a hybrid environment with a bit more control exercised in the core of the internet for some purposes, but for other purposes will enable end-to-end practices. Some things will have to be managed, especially if the capacity of the current internet becomes strained.

“While obviously these are contentious issues, in some ways this is a false dichotomy. Intermediary institutions may well gain more control. However, there will be minimal restrictions on information availability, because that’s what consumers will demand.” – Tom Lenard, President, Technology Policy Institute

“I think we will have an outcome that is a hybrid of your two options. For many users, the end-to-end principle in its literal form is a pain–it means they have to install software and manage upgrades on a PC that is complex and insecure. Much better to take advantages of services that are professionally run. But I think the end-users will be able to maintain the ability to reach the content of their choice and use the applications of their choice. I think the crucial question is not where a function is located (at the end-point or from a service provider somewhere on the network), but the extent to which the end-user will preserve the right to choose providers that they decide to trust. The real question is about trust, not location.” – David Clark, senior research scientist for the Next-Generation Internet, MIT professor

I believe that we go through cycles of expansion and contraction in our freedoms in different environments, or spheres of activity. The new and often unmanageable is tamed, and brought under control. Perhaps that is a good thing, as that state helps to stimulate the next new thing. So, if I wasn’t clear, the internet will become far more controlled, managed, segmented than it is today.” – Oscar Gandy, emeritus professor, University of Pennsylvania

This is a very hard one. It’s complex and multidimensional. Big things like ‘net neutrality’ as well as smaller things like customized brokerage and management. So it’s not a zero-sum game. Both will increase. Fights to maintain net neutrality will be fierce, but specialized and proprietary systems will also flourish.” – Ron Rice, University of California, Santa Barbara

“I really want to check and believe the first option, but it seems that the rapid increase in use of the internet is not being matched with development and expansion of the infrastructure- e.g. the predictions that studies suggest we may run out of internet capacity in a few years. So, supply may not keep up with demand and that suggests a bleaker future for an open end to end internet.” – Alan Levine, Chief Technology Officer, New Media Consortium

I’m split on this question as well. I think we’ll see something in the middle. Understanding that information is static and that communication is information in action; communication is the key and access is the lock. Banking/financial institutions and merchant organizations will develop authenticating/verifying roles of business that will enhance commerce. Governments will develop roles of engaging in the communities to govern institutions and protect the constituents. Access will be open and unrestricted but transactions will be governed by rules of commerce and community.” – Jack Holt, senior strategist for emerging media, U.S. Department of Defense

The dictates of business will shape large parts of the online experience and more pay-to-play business models will affect information flows online.

“’If the entertainment industry gains control of the routers, it will stop being the internet, so your dystopian scenario could happen, but not as written.” — Clay Shirky, professor, Interactive Telecommunications Program, New York University

“I think a larger portion of the web will become fee based by 2020 and the information will be managed. Even web start-ups will be looking for viable business models, the monetary issues of Facebook & Twitter will not be repeated.” – Brad Adgate, Senior Vice President, Research, Horizon Research

“My view is that rights owners are well organised and have the sympathy of governments. (This view is reinforced by events such as the French government’s recent decision that Google Books contravenes French copyright law, and that French law overrides foreign law on the Internet in respect of services available in France). The answer probably lies between the two : the internet will ideally by based on the end-to-end principle, but there will be a growing realisation – based in a combination of education and legislation – that there are rights in information that necessarily place some restrictions on its use and re-use. The eventual balance will be dictated by a combination of circumstances, and possibly by alliances of odd bedfellows such as interventionist governments and the music industries.” – Peter Griffiths, former Head of Information at the Home Office within the Office of the Chief Information Officer, United Kingdom

“There will continue to be centralizing of the content and the resources. It is the way of business to grow and capture your competition. This the area where I have the greatest concern, in that it will become too centralized. Go into many small towns today and you will see a Walmart and many empty locally owned shops, while not inherently bad for the small town, it is a sign of the centralization and more power in the marketplace. In some ways the internet will continue to grow and be able to accommodate both the large and the small players, if a few specific areas are guaranteed open access. But there is a point at which the audience will stop growing and the centralization will take root. We are just at the cusp another growth spurt, so that will take longer than 10 years.” – Michael Nelson, Georgetown University, Communication, Culture and Technology program

The needs of users themselves will sometimes drive changes that bring more control of online material and less end-to-end activity.

“I think there will be a slight difference in the world as I see but I agree generally that the openness of the Internet will remain in place. What will change is that people will want to access customized solutions, use special purpose devices like the Kindle, and will have a need for better connections for various things like telemedicine. All of this will not be in sync with the end-to-end principle but it will use the Internet architecture and provide value.” – Link Hoewing, Vice President, Information Technology, Verizon Communications

There will be “content service providers” who are gatekeepers of many users’ online experiences.

The notion of an ‘Internet Service Provider’ will still exist by 2020, but the main distinction will be the service packaging of base and premium services, both core and related. Perhaps a more descriptive term might be ‘Communications Service Provider’ (CSP). For example, your CSP could provide you with broadband access, and in addition offer VOIP services. In addition, the CSP could offer IPTV or so-called ‘over-the-top’ services such as Video-On-Demand, for additional fees, on certain devices. Another service the CSP could offer might be mobile broadband. In the process of delivering these various communication services, the CSP must invest huge amounts of capital in telecommunication infrastructure before any revenue can be realized. These CSPs then act in their own corporate interests in order to maximize the profitability of the services offered. However, these CSPs effectively become gatekeepers for the ‘last mile’ of internet connectivity needed to deliver consumer services. This means that pricing and service offerings, availability, etc. are all under the control of the CSP. A recent example is Comcast purchasing NBC Universal. Now it would be at least theoretically possible for NBC content to receive special treatment on Comcast networks.” — William Luciw, Managing Director, Viewpoint West Partners LLC


This particularly long and thoughtful answer from Doc Searls, co-author of the “Cluetrain Manifesto,” is reported in full. It does not fall on one side of the tension pair or the other. It addresses different ways of thinking about the internet’s architecture, especially in the future:

“There will always be a struggle to reconcile the Net’s end-to-end principle with the need for companies and technologies operating between those ends to innovate and make money. This tension will produce more progress than either the principle by itself or the narrow interests of network operators and other entities working between the Net’s countless ends.

“Today these interests are seen as opposed — mostly because incumbent network operators want to protect businesses they see threatened by the Net’s end-to-end nature, which cares not a bit about who makes money or how. But in the future they will be seen as symbiotic, because both the principle and networks operating within it will be seen as essential infrastructure. So will what each of does to the help raise and renovate the Net’s vast barn.

“The term infrastructure has traditionally been applied mostly to the public variety: roads, bridges, electrical systems, water systems, waste treatment and so on. But this tradition only goes back to the Seventies. Look up infrastructure in a dictionary from the 1960s or earlier and you won’t find it (except in the OED). Today are still no institutes or academic departments devoted to infrastructure. It’s a subject in many fields, yet not a field in itself.

“But we do generally understand what infrastructure is. It’s something solid and common we can build on. It’s geology humans make for themselves.

“Digital technology, and the Internet in particular, provide an interesting challenge for understanding infrastructure, because we rely on it, yet it is not solid in any physical sense. It is like physical structures, but not itself physical. We go on the Net, as if it were a road or a plane. We build on it too. Yet it is not a thing.

“Inspired by Craig Burton’s description of the Net as a hollow sphere — a three-dimensional zero comprised entirely of ends — David Weinberger and I wrote World of Ends in 2003 (http://worldofends.com). The purpose was to make the Net more understandable, especially to companies (such as phone and cable carriers) that had been misunderstanding it. Lots of people agreed with us, but none of those people ran the kinds of companies we addressed.

“But, to be fair, most people still don’t understand the Net. Look up “The Internet is” on Google (with the quotes). After you get past the top entry (Wikipedia’s), here’s what they say:

  1. a Series of Tubes
  2. terrible
  3. really big
  4. for porn
  5. shit
  6. good
  7. wrong
  8. killing storytelling
  9. dead
  10. serious business
  11. for everyone
  12. underrated
  13. infected
  14. about to die
  15. broken
  16. Christmas all the time
  17. altering our brains
  18. changing health care
  19. laughing at NBC
  20. changing the way we watch TV
  21. changing the scientific method
  22. dead and boring
  23. not shit
  24. made of kittens
  25. alive and well
  26. blessed
  27. almost full
  28. distracting
  29. a brain
  30. cloudy

“Do the same on Twitter, and you’ll get results just as confusing. At this moment (your search will vary; this is the Live Web here), the top results are:

  1. a weird, WEIRD place
  2. full of feel good lectures
  3. the Best Place to get best notebook computer deals
  4. Made of Cats
  5. Down
  6. For porn
  7. one of the best and worst things at the same time
  8. so small
  9. going slow
  10. not my friend at the moment
  11. blocked
  12. letting me down
  13. going off at 12
  14. not working
  15. magic
  16. still debatable
  17. like a jungle
  18. eleven years old
  19. worsening by the day
  20. extremely variable
  21. full of odd but exciting people
  22. becoming the Googlenet
  23. fixed
  24. forever
  25. a battlefield
  26. a great network for helping others around the world
  27. more than a global pornography network
  28. slow
  29. making you go nuts
  30. so much faster bc im like the only 1 on it

“(I took out the duplicates. There were many involving cats and porn.)

“Part of the problem is that we understand the Net in very different and conflicting ways. For example, when we say the Net consists of “sites,” with “domains” and “locations” that we “architect,” “design,” “build” and “visit,” we are saying the Internet is a place. It’s real estate. But if we say the Net is a “medium” for the “distribution” of “content” to “consumers” who “download” it, we’re saying the Net is a shipping system.

“These metaphors are very different. They yield different approaches to business and lawmaking, to name just two areas of conflict.

“Bob Frankston, co-inventor (with Dan Bricklin) of spreadsheet software (Visicalc) and one of the fathers of home networking, says the end-state of the Net’s current development is ambient connectivity, which “gives us access to the oceans of copper, fiber and radios that surround us.” Within those are what Frankston calls a “sea of bits” to which all of us contribute. To help clarify the anti-scarce nature of bits, he explains, “Bits aren’t really like kernels of corn. They are more like words. You may run out of red paint but you don’t run out of the color red.

“Much has been written about the “economics of abundance,” but we have barely begun to understand what that means or what can be done with it. The threats are much easier to perceive than the opportunities. Google is one notable exception to that. Asked at a Harvard meeting to explain the company’s strategy of moving into businesses where it expects to make no money directly for the services it offers, a Google executive explained that the company looked for “second and third order effects.”

“JP Rangaswami, Chief Scientist for BT (disclosure: I consult BT) describes these as “because effects.” You make money because of something rather than with it. Google makes money because of search, and because of Gmail. Not with them. Not directly.

“Yet money can still be made with goods and services — even totally commodified ones. Amazon makes money with back-end Web services such as EC2 (computing) and S3 (data storage). Phone, cable and other carriers can make money with “dumb pipes” too. They are also in perfect positions to offer low-latency services directly to their many customers at homes and in businesses. All the carriers need to do is realize that there are benefits to incumbency other than charging monopoly rents.

“The biggest danger for the Net and its use comes not from carriers, but from copyright absolutists in what we have recently come to call the “content” industry. For example, in the U.S. the DMCA (Digital Millenium Copyright Act), passed in 1998, was built to protect the interests of copyright holders and served as a model for similar lawmaking in other countries. What it did was little to protect the industries that lobbied its passing, while at the same time hurting or preventing a variety of other industries. Most notable (at least for me) was the embryonic Internet radio industry, which was just starting to take off when the DMCA came along. The saga that followed is woefully complex, and the story is far from over, but the result in the meantime is a still-infant industry that suffers many more restrictions in respect to “content” than over-the-air radio stations. Usage fees for music are much higher than those faced by broadcasters — so high that making serious money by webcasting music is nearly impossible.

“There are also tight restrictions on what music can be played, when, and how often. Music on podcasts is also essentially prohibited, because podcasters need to “clear rights” for every piece of copyrighted music they play. That’s why, except for “podsafe” music, podcasting today is almost all talk.

“I’ll give the last words here to Cory Doctorow, who publishes them freely in his new book Content:

… there is an information economy. You don’t even need a computer to participate. My barber, an avowed technophobe who rebuilds antique motorcycles and doesn’t own a PC, benefited from the information economy when I found him by googling for barbershops in my neighborhood.

Teachers benefit from the information economy when they share lesson plans with their colleagues around the world by email. Doctors benefit from the information economy when they move their patient files to efficient digital formats. Insurance companies benefit from the information economy through better access to fresh data used in the preparation of actuarial tables. Marinas benefit from the information economy when office-slaves look up the weekend’s weather online and decide to skip out on Friday for a weekend’s sailing. Families of migrant workers benefit from the information economy when their sons and daughters wire cash home from a convenience store Western Union terminal.

This stuff generates wealth for those who practice it. It enriches the country and improves our lives.

And it can peacefully co-exist with movies, music and microcode, but not if Hollywood gets to call the shots. Where IT managers are expected to police their networks and systems for unauthorized copying — no matter what that does to productivity — they cannot co-exist. Where our operating systems are rendered inoperable by “copy protection,” they cannot co-exist. Where our educational institutions are turned into conscript enforcers for the record industry, they cannot co-exist.

“The information economy is all around us. The countries that embrace it will emerge as global economic superpowers. The countries that stubbornly hold to the simplistic idea that the information economy is about selling information will end up at the bottom of the pile.”

  1. Zittrain, Jonathan. The Future of the Internet-And How to Stop It. Caravan Books. 2008. Note: Zittrain is on the advisory board of the Pew Research Center’s Internet & American Life Project.