December 14, 2008

The Future of the Internet III

Scenario 7: The Evolution of the Architecture of the Internet

Prediction and Reactions

PREDICTION:  Next-generation research will be used to improve the current Internet; it won’t replace it. In 2020, the original Internet architecture is in the continuing process of refinement – it hasn’t been replaced by a completely new system. Research into network innovation, with help from the continued acceleration of technologies used to build, maintain, enhance, and enlarge the system, has yielded many improvements. Search, security, and reliability on the Internet are easier and more refined, but those who want to commit crimes and mischief are still able to cause trouble.

 Expert Respondents’ Reactions (N=578)
Mostly Agree 78%
Mostly Disagree  6%
Did Not Respond  16%

All Respondents’ Reactions (N=1,196)
Mostly Agree  80%
Mostly Disagree  6%
Did Not Respond  14%
Note:  Since results are based on a nonrandom sample, a margin of error cannot be computed. The “prediction” was composed to elicit responses and is not a formal forecast.

Respondents were presented with a brief set of information outlining the status quo of the issue 2007 that prefaced this scenario. It read:

Due to concerns over Internet security, reliability, and complexity, the National Science Foundation in the US is funding research into the building of a “next-generation” or “clean-slate” Internet. The NSF initiatives include the Global Environment for Networking Innovation (GENI – building a test network on which researchers will be able to try out their ideas) and Future Internet Network Design (FIND). The European Union is funding research through its Future Internet Research and Experimentation (FIRE) program. Creating an all-new Internet might solve problems like viruses, spam, phishing, and worms. But it would cost billions of dollars and there is a debate among experts about how long it might take. If a next-generation Internet is built, some people are concerned it will be characterized by intrinsic features that will allow governments and corporations to exercise more control over what happens online. So, the constant question remains: How do we raise barriers against spam, cybercrime, and terrorism and provide secure systems for digital transactions without infringing on civil liberties?

Overview of Respondents’ Reactions

Respondents clearly feel the current structure and basic architecture of the Internet will continue to underlie the technology. They believe there will be significant enhancements and updates, however, a “new” system will not “replace” the current architecture; transformations will occur gradually. They point to two major changes that are already running in parallel with legacy systems: institution of IPv6, the new protocol; and implementation of elements of the Semantic Web, which will make it easier to find and link related information. Some argue, though, that by 2020 there will be specific “walled gardens” (or restricted areas of interaction and information) that will be secure but also give control over the network to the garden creators. Others suspect there may be split networks or partitions in the Internet, especially as governments and corporations leverage security fears to retain power over who can do what on the network.  While protections are consistently added to the network, many experts think crime, piracy, terror, and other negatives will always be elements in an open system.

There was resounding support among these respondents for the idea that the current Internet architecture will be continually refined and not completely replaced by a next-generation Internet, with four of every five responses mostly in agreement with the scenario. Just 5% mostly disagreed. “Legacy computing platforms tend to last a long time, as will the Internet,” wrote David Moschella, global research director for the Computing Sciences Corporation’s Leading Edge Forum and a Computerworld columnist.

“The control-oriented telco (International Telecommunication Union) next-generation network will not fully evolve, the importance of openness and enabling innovation from the edges will prevail; i.e. Internet will essentially retain the key characteristics we enjoy today, mainly because there’s more money to be made,” responded Adam Peake, executive research fellow and telecommunications policy analyst at the Center for Global Communications (GLOCOM).

Scott Brenner, consultant, technologist, and Web developer for Fortune 100 companies, commented, “The current Internet won’t be replaced by a new system by 2020 any more than the highway system originally built in the 1950s has been replaced by a new system. Sure, the asphalt and concrete has long since been replaced, but no one’s suggested to let the forest reclaim the land while another system of roads is built (at least not on a large scale). The Internet of 2020 will be very different from today. But it will just be a many-orders-of-magnitude improvement over what we’ve got now.”

Alejandro Pisanty, director of computer services at the Universidad Nacional Autónoma de México, an active leader in ICANN, the Internet Society, and the Internet Governance Forum, noted, “Most of the clean-slate proposals that are being thought of in public would seem to underestimate the value of the yet-existing system, and the fact that the Internet’s strong decentralization makes it incumbent on the users at the edge to apply changes they often don’t master technically and for which it is difficult for them to pay. The abuse by intermediaries (from large telcos to small, local providers including small ISPs and Internet cafes) disincentivates change even further.”

Some survey participants responded that there are multiple conduits in the network now and that approach will become more formalized. “In fact,” wrote Anthony Townsend, research director for the Technology Horizons program of The Institute for the Future, “some parts of the Internet may fragment, as nations pursue their own technology trajectories.”

Townsend also agreed that change will continue to be an incremental evolution, writing, “The Internet is so vastly complex, incremental upgrades seem to be the only way to get anything done. Look at how little IPv6 there is. Places like China may make big leaps and bounds because there is less legacy.”

Joe McCarthy, principal instigator at MyStrands and formerly principal scientist at Nokia Research Center in Palo Alto, left the door open when he commented, “Too much is already at stake on the existing Internet to build a new one. However, the recent FCC rulings that will force everyone to switch from analog to digital television shows that the [US] federal government is not averse to forcing large-scale changes on its population in the conduits through which they must seek electronic information and entertainment.”

Hal Varian, chief economist for Google, wrote, “The research on next-generation Internet will pay off by allowing some retrofit of the current network.”

Just one respondent who provided an elaboration jumped in with a clear statement of expectation that significant system upgrading is needed. Ian Peter of Ian Peter and Associates and the Internet Mark 2 Project, a pioneer who helped develop the Internet in Australia and the Asia-Pacific region in the 1980s, responded, “It is unlikely that TCP/IP (be it v4 or v6) will survive much beyond 2020. Current Internet standards bodies and core Internet protocols are ossifying to such an extent that security and performance requirements for next-generation applications will require a totally new base platform. If current Internet base protocols survive, it will be as a substrata paved over by new-generation smarter ways of connecting.”

Many the many fears expressed over the politics that could be built into its architecture and the Internet’s deepening complexities respondents also indicated there is reason for optimism. “The Web must still be a messy, fabulous, exciting, dangerous, poetic, depressing, elating place…akin to life; which is not a bad thing,” noted Luis Santos, Universidade do Minho-Braga, Portugal.

Incremental Change Will Continue; There Will Not Be a Clean-Slate Internet

A number of survey participants noted that change is rarely delivered in a wholesale way in complex systems. “Successful solutions are almost always built on existing infrastructure, rather than starting from a clean sheet—simple economics,” noted Jason Stoddard, managing partner for strategy at Centric/Agency of Change, an interactive strategies company.  Walt Dickie, executive vice president and CTO for C&R Research was aggressive in his support for the scenario. “I don’t ‘mostly’ agree, I agree completely,” he wrote. “The utopian vision of a next-generation Internet birthed by wise and benevolent leaders will be preceded by flocks of flying pigs, peace in our time, and the Easter Bunny.”

Jeff Jarvis, blogger at Buzzmachine and a professor at City University of New York, commented, “Any media company that has tried to build the ultimate content-management system has learned this lesson: It’s never done, far from perfect, too expensive, and always behind. We will build on what we have.”

“The Internet is too distributed to undergo a clean-slate facelift,” wrote Susan Thomas of S2 Enterprises LLC. “Incremental innovation will reign, based on short-term pressure to monetize,” noted Peter Kim, a senior analyst specializing in e-strategy and management for Forrester Research.

Steve Goldstein, an ICANN board member whose job with the US National Science Foundation in the 1990s was to help diffuse the Internet globally, commented, “Depending on where in time one reckons the start of the Internet (~1970 or ~1980), it took about 25 or 15 years for a truly commercial Internet to develop (~1994), an another 10 years at least for it to become as feature-rich as we now experience it to be (recall Mosaic, first browser in 1993; fully functional browsers on phones in early 2000s). So, even if NSF’s and the EU’s experimental network technologies were to be successful in developing a revolutionary next-gen Internet, I would not expect it to displace the legacy Internet until after 2020. And, I am not a real fan of either the NSF’s or the EU’s ability to re-create another disruptive technology to displace the Internet as we know it. There is likely to be too little funding and too much cronyism for that to happen. On the other hand, I would expect to see some developments feed into incremental improvements in today’s Internet.”

James Jay Horning, chief scientist for information systems security at SPARTA Inc. and a former fellow at Xerox’s Palo Alto Research Center, wrote, “Telephone managed to eventually supersede the legacy infrastructure of telegraph wires, but I don’t see any correspondingly disruptive technological advantage that will cause the clean-slate Internets to replace, rather than supplement, the current one. I see a rolling transition, rather than a clean break.”

Christine Boese, a researcher and analyst for Avenue A-Razorfish and Microsoft, responded, “The groups funding and building these so-called ‘new’ platforms are delusional. Not that there never will be new platforms, but they won’t come from any of those groups. If such a new platform should magically appear, it will arise from inside the current Internet, and it will be fully backward-compatible and inclusive. There will never be a ‘clean-slate Internet,’ unless our culture does an Atlantis and dumps our beautiful Alexandria on the Ethers into the sea.”

Hamish MacEwen, a consultant for Open ICT in New Zealand, wrote, “Looking at fundamentals such as the calendar, after lunar/solar, there was Julian, after Julian there was Gregorian. Will there be a replacement. No. Some basics reach a state of ‘good enough’ and we move on to other things. So it is with the Internet. So it was with Ethernet. So it was with SMTP. Could it be better, yes. Is it good enough, yes. IPv6, yes, but there’ll be a lot of IPv4 for a long time to come, probably still in 2020. ‘Those who want to commit crimes and mischief are still able to cause trouble.’ Now there’s an eternal verity.”

The Move to IPv6 and Semantic Web Will Create New Online Opportunities

Some survey participants noted that the Internet is a system of networks (including the research networks Internet2 and Lambda Rail) that is already undergoing the most major overhaul since its beginnings, as improvements in the technologies of the architecture are introduced and it transitions to Internet Protocol version 6 from IPv4 and as it also begins to weave in the added features of the Semantic Web, a longtime project of Web-innovator Tim Berners-Lee and the World Wide Web Consortium.

“Internet2 is providing today the promise for advanced networks of tomorrow; unexpected jumps in optical networks will permit new types of access to rich media data and HD-based imaging,” wrote Don Kasprzak, chief executive officer of Panaround.com and a former system engineer at Apple Computer. Paul Jones, director of ibiblio.org at the University of North Carolina-Chapel Hill, noted, “The work already under way on National LambdaRail and Internet2 is showing evolutionary improvements.”

Paul Greenberg, president of The 56 Group LLC, commented, “There is no reason to create a new ether out of whole cloth. With the implementation of the address protocol IPv6, which provides an infinitely large number of Internet addresses, we don’t have to worry about it running out of ‘space’ so to speak. The new forms of the Web, like Web 3.0—the Semantic Web—will begin to show us how to interact with the Web in context, ways we can hardly imagine now will provide us with new directions. The idea of specialized search will unlock much of the so-called ‘dark Web’—that portion of the Internet that isn’t really being searched with Google or any other engine for that matter. Yet, there is always the possibility with something that covers as much ground as the Internet for breaching it. If it is secure, given the old problems it has, there will be someone who will creatively find a way to commit criminal mischief. So problems will continue but there really is no reason to create a whole new Web.”

Micheál Ó Foghlú, the research director for the Telecommunications Software & Systems Group at the Waterford Institute of Technology, Ireland, and a member of W3C and an active participant in next-generation research, wrote, “In the short term, we need to put effort into migrating from IPv4 to IPv6 to respond to the looming crisis in IPv4 address space…My research group, the TSSG, plans to participate in the research efforts of GENI/FIND and FIRE and already have to some degree; these are medium- to long-term, and cannot come to fruition by 2020. Most of this work will involve various overlay networks (over the IPv4 or IPv6 Internet) but some will take a clean-slate approach, and any clean-slate approach is very unlikely to be widely deployed in the next 12 years. So I hope to see a healthy IPv6 Internet with a legacy IPv4 Internet both operating in 2012, and lots of interesting ideas from research being deployed as overlay networks over that basic infrastructure. I do not see the private telecommunications infrastructure adopting the open-Internet model, though it may use IP technologies, so there will still be a number of interesting networks in 2012, most using forms of IP. One other interesting trend in network infrastructure development is the use of carrier-grade Ethernet, pushing previous LAN technologies into use within a wider remit, such as metropolitan networks. The promise is that these are cheaper to deploy and manage even compared to IP networks. IP will still be needed to interconnect these networks, and IPv6 will be needed.”

Todd Spraggins, strategic architect for Nortel Carrier Networks and president and chairman of the board of directors of the Communications Platforms Trade Association, responded, “The Internet can never be ‘replaced,’ as the next best thing will not overlay it but be integrated, thus always having the appearance of being extended by the uninitiated outsider.”

Developments Will Be Driven by Security; There Will Be Increased Privacy Concerns

Many respondents say a further surrender of privacy in exchange for security will play out in a big way before 2020. “The arms race between the good guys and the bad guys doesn’t slow or stop—it goes on hyper-overdrive,” predicted Sean Steele, CEO and senior security consultant for infoLock Technologies. “Average business users and consumers will have more, not less, security in transactions and communications, but will be required to use more invasive technologies and techniques, such as biometric authentication (e.g., fingerprint recognition, voiceprinting, iris/retina scanning, etc.).”

Robert Eller of Concept Omega, a marketing and communication company, agreed that identification will be based on genetic information.  “We will eventually only be able to interact with the Web with a personal biometric/genetic code which will imprint on any interaction we provide,” he responded. “This should remove all forms of fraud or spam. To allow for privacy in 2020, laws are required for government access to this data when reason for fraud/misuse are evident.”

Bertil Hatt, an Internet researcher employed by France Telecom and Orange who is completing a Ph.D., predicted that in 2020 “most piracy has been solved through licensing, although corporate-secret appropriation (CSA) has taken the lead. Most malware used to come from rogue countries who have been so ostracized for harboring spam-, virus-, or worm editors that they finally took part in global agreement on extraterritoriality of digital crime and e-terrorism. Phishing is still rampant, perpetrated by very small actors, but widespread knowledge and Bayesian filtering considerably limits its impact.”

Thomas Lenzo, a business and technology consultant with Thomas Lenzo Consulting, wrote, “By 2020, beyond technology, there must be multi-national initiatives to coordinate efforts to fight cybercriminals; laws must change to combat evolving cybercrimes; nations must cooperate in their arrest and prosecution. There must be a unified global effort to deal with those countries that encourage or employ cybercriminals.”

Chris Miller, senior vice president for digital operations for Element 79, commented, “The how, when and what we use to access the Internet will change (smartphones vs. computer; anywhere vs. home/work becomes the norm). Hackers will continue to be a part of society but their mischief also drives innovation as it does today…Cybercrime or cyberterrorism takes on more priority. Look at the Middle East and Asia shutdown and delays due to the FALCON cable cut. At this time we don’t know what caused it.”

Leonard Witt, an associate professor of communications at Kennesaw State University and author of the Weblog PJNet.org, remarked, “When have we ever stopped crime? If it is a choice between having some criminals around and having a repressive government, I will take the former; they are much easier to deal with.”

Respondents held out hope that privacy protections can somehow be preserved. “If enough people demand privacy protection, that will improve, too,” noted Peter W. Van Ness, president of the Van Ness Group, a Web-development company. “If we do not demand it, privacy will be traded away for increased security and reliability; that is not a good trade.”

Corporate and Government Control Are Among the Primary Concerns About Network Change

Many respondents’ negative remarks about the diffusion of a “clean-slate” Internet were prompted by concerns that some see the “do-over” efforts as a threat to civil liberties. “The Internet is not magical; it will be utterly over-managed by commercial concerns, hobbled with ‘security’ micromanagement, and turned into money-shaped traffic for business, the rest 90% paid-for content download and the rest of the bandwidth used for market feedback,” wrote Tom Jennings, of the University of California-Irvine, creator of FidoNet, the first message and file-networking system online, and the builder of Wired magazine’s first online presence. “Notice that ARPANET was handed to commercial interests; it wasn’t turned into a national/international resource for citizens (and don’t tell me that mega-corporations are citizens).”

Nick Dearden, campaigns manager for Amnesty International, the human-rights organization, responded, “All I would like to do is point out the risks. The Internet has, in many ways, grown up from the grass roots, it wasn’t controlled by governments or corporations. That fact has led to it being a useful space, beyond normal social controls that we see, for instance, in the broadcast and print media. As governments and companies extend their control—sometimes to near-monopolies—over sections of the Internet, this space has closed down. On the surface, controlling spam seems like something few people would argue with. But in China, the war against spam has actually been used to crack down on all matter of political activity. The only way to protect free space is to ensure that any systems created to deal with real problems on the Net—e.g. child pornography—are grounded in human rights and protect fundamental freedoms like freedom of speech. To date these rights have taken a back seat in discussions of Internet governance, and I’m therefore fearful of how new-generation research will be utilized.”

Howard Rheingold, Internet sociologist and author, noted, “The Internet’s end-to-end architecture is being compromised when the Great Firewall of China filters packets and blocks data for political reasons, and the architecture of participation that made the Web possible is under attack when broadband providers break ‘network neutrality’ for commercial reasons. But the problems with replacing something as widespread and flexible as the present Internet—with all its problems, which may indeed necessitate radical redesign—are economic, political, and formidable. Who is going to design, govern, deploy, pay for the new system, and how are the world’s major political and economic players to agree? Starting the Internet was simple back when everybody trusted Jon Postel. The world lacks that technopolitical simplicity today.”

Theresa Maddix, a research analyst for ForeSee Results, responded, “NSF initiatives, GENI, FIND, and FIRE are all well-intentioned and led by very bright individuals. However, the information wants to be free. It was the release of the Internet from government hands and agencies that allowed it to explode. Google and others are always building better spam filters. Cybercrime still is much lower than non-cybercrime.”

Don Heath, a former leader of the Internet Society and member of the U.S. State Department Advisory Committee on International Communication and Information Policy, noted, “The Internet has achieved its remarkable success because it was not controlled by any one entity or government. As soon as governments attempt to exercise control or otherwise regulate the Internet, its usefulness will greatly diminish.”

The delicate balance of all interests was pointed out by Jerry Michalski, founder and president of Sociate, a technology-consulting firm and former managing editor of Release 1.0 and co-host of the PC Forum. “The Internet is what it is because commercial interests and government agencies didn’t know what it was (DARPA aside),” he responded. “There is no way to build anything like that anymore, so I have no hope that something better can be built, or that everyone can be migrated to it. That said, I’m worried about Net neutrality and I see many ways in which today’s Internet could be hobbled significantly or improved greatly over time, with no big disjunction.”

There Will Be Other Networks or Partitions; ‘Walled Gardens’ Will Be Leveraged For Control

Some respondents expect that various motivations will cause more separation of networks. “Those with resources and security concerns will have access to ‘better’ and more secure channels. Speed and security will increase for everyone but someone will figure out how to partition off areas of the network for elites,” wrote Ted Coopman, a communications technology lecturer at San Jose State University.

Michael Zimmer, resident fellow at the Information Society Project at Yale Law School, responded, “The most likely scenario:  A secure architecture to complement the existing Internet backbone for those who want to use it. One alternative view might be a new Internet-like infrastructure emerging tailored specifically for secure mobile-data transfer, capitalizing on the rise of mobile telephony.”

Cambria Ravenhill, manager of national channel planning at TELUS Communications, wrote, “The Internet will split into the ‘official’ Internet, where most civic life and corporate and government transactions occur, and an ‘underground’ Internet fueled by scarcity economics.”

Jay Neely, founder of News Armada, a Boston-based online news and community company, commented, “If government encouragement does not occur within the next 5 years, while there will still be refinements made to existing infrastructure, the process will be too slow for some organizations, and we will see development of separate networks, like Internet2 for universities. While unlikely, it is possible that a future technological mega-corporation could build an Internet-like infrastructure that competes with publicly available Internet; concerns about civil liberties and tracking are even more valid in this scenario, but may be overlooked by the general public due to the convenience of the advanced infrastructure.”

“There will be two Internets,” predicted Garland T. McCoy, founder of the Technology Policy Institute, a think tank focused on the economics of innovation, “one for ‘us’ and one for the financial institutions, security folks, spooks, government agencies, major corporations, etc. That almost exists today.”

Mark Youman, principal at ICF International, a Washington, D.C., consulting-services company, wrote, “The current Internet will be improved rather than replaced wholesale, but it will be one of MANY global networks. Institutions, industries, and other groups will construct independent networks when the Internet becomes too overrun or corrupt to serve their needs. Access to these networks will be part of what defines the ‘haves’ from the ‘have nots.’”

Could a division of networks lead to another type of visible divide and possibly even an “Internet class war”? Some respondents mentioned the possibility. John Jordan, an associate professor of communications at the University of Wisconsin-Milwaukee, wrote, “Like a highway, significant parts of the Internet likely would need to be shut down and closed off in order to receive a major upgrade. Unlike a highway, the public will not stand for this, necessitating that instead of a completely new Internet infrastructure, people get slightly better service over time as patches and upgrades are made, but this leaves open potential problems familiar today. At the same time, private business ventures and new housing developments in exclusive neighborhoods will experiment with and implement new Internet architecture, leading to a point in the more-distant future where there may be two Internets, creating a true Internet class war.”

Benjamin Ben-Baruch, senior market intelligence consultant and applied sociologist for Aquent, predicted that there will be two Internets, the original and the next-generation. “Those with the resources to move much of their communications and functions to this new architecture will do so—and early adopters will have to pay hefty costs to do so. But along with this high-cost barrier will come control of this new environment. Part of the digital gap in the future will be between those who operate on both the current and next-generation platforms and those who are limited to the current Internet…Security and privacy on the current Internet will be increasingly compromised. There will be two reasons for this slowly but steadily decreasing security and privacy: (1) Hackers and pirates will develop security-breaking technologies faster than security technologies can be developed and rolled out.  (2) As the secure, next-generation platform is developed, hackers and spammers and pirates and other Internet criminals will focus on the much easier but very lucrative prey on the current Internet.”

It’s Possible – Even Before 2020 – That Some Revolutionary Idea Could Shake Things Up

A few respondents noted that breakthroughs incorporating influences from biology, nanotech, and other sciences could push Internet evolution in new directions. “By 2020, two major advances will have significant impact. The first is bioengineering and nanotechnology, allowing the Net to be ‘embedded’ into individual humans (scary, eh?); the second is quantum computing that will significantly alter the current electrically loaded computing engines,” predicted David Hakken, a professor of anthropology at the Indiana University School of Informatics who studies social change and the use of automated information and communication technologies.

Roberto Gaetano, an ICANN board member who also works for the International Atomic Energy Agency, commented, “I concur in seeing the development of the ‘next-generation Internet’ as evolutionary rather than revolutionary. However, I am wondering whether by then we would not start seeing something that is started based on some new concept that we can’t even figure out today. And I wonder whether this is not likely to come from a socio-cultural environment that is completely different from ours.”

Cite this publication: Lee Rainie and Janna Anderson. “The Future of the Internet III.” Pew Research Center, Washington, D.C. (December 14, 2008) http://www.pewinternet.org/2008/12/14/the-future-of-the-internet-iii/, accessed on July 22, 2014.