Article written for
Société de Stratége
AGIR Revue Generale de Stratégié
Issue 20-21: La société de l'information
(janvier 2005)
There was a Belgium
Research Project on this topic.
I gave this talk at the end of the project:
Applying ECs to Disaster Relief.
Read
this in
Croatian, courtesy of Milica Novak;
Polish, courtesy of Grzegorz Adamski;
Romanian, courtesy of Alexandra Seremina;
Portuguese, courtesy of
Artur
Weber;
Lithuanian, courtesy of Modestas
Mice
and
Chinese, courtesy of Elisa Liang.
See also the 2011 Peering article on this
topic
and the follow-on 2016 article.
Much of the conflict between grassroots exploiters of emerging Internet and telecommunications technologies, and the vested interests that are threatened by them is largely due to the inability of the latter to predict disruptive effects and rationally respond with new business and governance models.
The Recording Industry Association of America (RIAA) is the poster boy for this phenomenon. Their failure to see the application of peer-to-peer file sharing to consumer music, and their stubborn reluctance to give up their monopoly over music distribution (and control over artists), has led to suing teenagers and reluctantly offering expensive on-line music services that are less convenient and comprehensive than the free ones. The question is why the RIAA would not anticipate the change and take advantage of it? The short history of the commercial Internet is rife with examples of such short-sightedness.
When ARPA released the Internet to commercial interests, the entertainment industry dismissed it as the CB radio fad of the 90s (in the USA), soon to join discarded 8-track stereos. The real future was in Interactive Television. Of course, in France, who would have predicted that that the beautifully designed, widely-distributed, and much-used MiniTel would be supplanted by the Internet? Hollywood, France Telecom, and Bill Gates were only a few to be broadsided by the power of the Internet in the mid-nineties.
Everyone understands, now and in principle, that the exploitation of Internet protocols creates disruptive technologies that change the viability of old business models, usually by eliminating the "middleman". The old middlemen, such as CD distributors, can fight back, but it is clear that they inevitably lose their distribution monopoly as users migrate to more sophisticated systems, such as FreeNet[1]. But proper exploitation can lead to business models, usually based upon aggregating information and services in a new way. Amazon and Google are the latest and most spectacular success examples of this possibility. What is surprising is that many smart people seem to have failed to have learned from this history: readers will already be aware of many current examples of industries and governments misunderstanding Internet technologies.
Contrary to the impression one gets from "Dilbert" cartoons, most of the people making decisions in governments and corporations did not get to their positions by being stupid. What is it that they "don't get" after seeing repeated examples of success and failure over the last 10 years? The explanation that such leaders are insufficiently technical to see the potential of novel technologies is weak. Take the case of US telephone engineers in the mid-nineties who thought that TCP/IP, the backbone protocol of the Internet, would have to be rewritten in ATM[0], the reliable protocol with which they were familiar. Of course today, the reliable qualities of ATM are often subverted in service of the flaky TCP/IP carrying the latter's traffic in a kind of emulation mode. Why did smart, knowledgeable engineers, familiar with protocols, not see this coming?
Further, in 1996, Robert Metcalf, one of the several "Fathers" of the Internet, said that the Internet would crash that year, or he would eat his words. In 1997, at the ACM conference in San Jose, California, he blended a copy of his column and drank it on stage. Even for experts, the resilience and power of the Internet has been difficult to predict because its control mechanism is based on emergent behavior. Unlike ATM, the basic Internet protocol depends upon routers making local decisions, none of which is guaranteed to deliver a packet of information to its destination, but the global behavior is that communications happens, fairly reliably.
A common example of this kind of emergent behavior is an ant colony. Godel, Escher, and Bach[Hofstadter], discusses "Aunt Hillary", an ant colony, in which it is pointed out that her behavior, and intelligence, is the result of the rather mindless interactions of individual ants following simple protocols of interaction that result in qualitatively different global behavior. This effect is well-analyzed in computer science as "cellular automata theory"[Mitchell-93].
This leads to a more subtle point that has to do with state of mind. Engineers are used to being in control. Prime ministers, CEOs, and engineers depend upon the sense of being in control. They are congenitally incapable of realizing when control passes to a protocol.
The important principle is that emergent-behavior protocols are the government. This is easy to miss, especially if you think you are the government. Or the phone company. So-called "anarchists" are not out of control - they are simply recognizing who, or what, is really in charge - the protocol.
This strong argument about the power of emergent-behavior protocols is increasingly widely-understood, and insufficient. TCP/IP did not cause an immediate disruption when it was released for commercial use. It could have. But it didn't. The disruption was caused by a new Internet layer: the WWW.
The WWW is an emergent control-based system because there is no central control: merely nodes interacting locally with one another via hypertext links. But it is the social behavior of the users that makes the difference. Who would have thought that thousands, if not millions, of people would start posting web sites of information, about everything, with no obvious commercial gain? Few did. Certainly not Bill Gates.
The WWW allowed individuals to participate, contribute, create, and publish to a wide audience. Interactive TV, the MiniTel, and even the Internet, did not. Interactive TV and MiniTel were "closed" systems. The content was not under control of the users. The WWW was created because the older methods of sharing scientific papers and data, FTP and Gopher, weren't quite good enough for physicists. The WWW enabled mere theoretical physicists, instead of only computer scientists, to easily share documents over the Internet. As it turned out, what was easy for physicists could be made easy for lots of people.
The Anarchist in the Library[Vaidhyanathan-04] addresses the disruptive usages of new technologies, but emphasizes anarchy, both capitalized and lowercase, as a motivation. We rather postulate a drive toward collective work and play. Malicious hackers and spammers are largely a by-product, as is the threat to existing distribution channels. People who share music aren't out to ruin CD distributors: that's a side-effect.
Music-sharing networks expose such misunderstandings about control and intent. Music executives failed to anticipate that people would collectively post their content for free. And then they mistakenly looked for someone in control to sue, finally falling back on suing teenagers.
The WWW and peer-to-peer file sharing networks became popular because they were superior methods of allowing people to publish their own content for collective benefit. So emergent control networks are important because not only can they scale, but also that they do. Without central control, it's cheap to add nodes. And if tools make it easy to add nodes, then the system is likely to grow. And such systems should be error-tolerant if not homeostatic. But these are only necessary conditions.
The final condition that is sufficient for network growth is when people want to add nodes, useful to others, to such a system. In the case of the WWW and music sharing, it is easy for people to see that they benefit individually from contributing to the collective: the network not only scales but increases in value. We hypothesize additionally that there is a general motivation to contribute to a collective effort. Let us call the combination of emergent control networks, tools, and incentives that motivate people to act collectively "Emergent Collectives".
Emergent Collectives mimic the emergent properties of TCP/IP. The effects of the overall system emerge from the interactions of the components, the protocols, data standards, tools, and people. They are hard to predict. They run counter to many established models. Trying to protect older systems in which a single controlling entity provides content (either for pay or power) is futile and even counter-productive. Back to the phone companies for another example. When it was clear that the Internet was a major new market, the only way they could see to take advantage was cell phones. 3G ("third generation"[2]) mobile phone users would access the Internet at 2mbps sitting still in a coffeeshop. In late 2001, when Siemens brought out its 2.5G (GPRS) mobile phone, it paid young women to parade around the Frankfurt airport with laptops strapped to their chest, like old cigarette vendors, connected to mobile phones, to demonstrate the future of the wireless Internet.
But the IEEE had long ago standardized 802.11b. This protocol, when coupled to a simple and cheap card for a laptop, enabled the user to connect to the Internet wirelessly at speeds more than five times what 3G has only ever promised: 11mbps were already available in public hotspots such as airports by 2001[Petrie-01].
Only in recent years have most large telecoms dived into the public hotspot business. But this is too little too late. Free hotspots are now ubiquitous in many urban areas. You can stand on many corners in San Francisco and have a choice of free connections. The Hackerbot builders[3] report seamlessly completing a download of the Linux OS while driving across Seattle.
The essential mechanism of cell phones is a kind of cellular automata protocol: cells hand-off to each other and the global behavior is that phone sessions continue. Why has an Emergent Collective formed from 802.11 and not cell phones? Because it is not possible for users to contribute to a cell phone network, other than by paying to use it.
WiFi has the properties of an Emergent Collective. 802.11 is open in that the default settings on access points allow users to use one another's systems. The hardware is cheap and the software is free, mostly built-in now. The system scales because there is no central control and users are inclined to contribute to the collective effort of providing ubiquitous free broadband access by adding nodes to the ad hoc network.
The telecoms have been thinking that by being in control of the 3G radio frequencies, they would be in charge of the wireless Internet. They never thought people would offer broadband access to strangers for free using 802.11. (Nor did the cable companies.) They have not had the imagination to consider alternative business models[4] and must now consider 802.16 (WiMax) and 802.20 as possible 4G technologies that skip 3G.
The Emergent Collectives we describe here emphasize structured technical networks as well as a will to collective action. This does not cover all of the examples of Smart Mobs [Rheingold], such as flash mobs, but would cover the open source movement and Seti@home as people add nodes to a network. But all exemplify the common impulse to contribute to a collective. The reader can probably think of other important examples, but let's jump to a post-modern world.
There is the potential for new Emergent Collectives to change our society in more fundamental ways. In the early 80's, some of us predicted that companies would mostly downsize to a few hundred people and that almost everyone else would become a private contractor. Today, this vision is not quite so far-fetched, having been independently stated by so many people that "Anarchist" calls it the "Northern California Ideology" (NCI).
An important principle of this ideology is the creeping commoditization of everything. As we understand commonalities among things that previously seemed disparate, we communicate about them at a simpler level. One trend is the marketplace that started with buying and selling of electronic components on an closed marketplace. We now agree better on how to classify products and services. The more we agree, the more they become commodities. And if it is a commodity that an individual can supply, then that individual can be an independent one-person business.
Further, we can see emerging Internet-wide workflows, both in businesses and in engineering. Distributed supply chains today only deal with established vendors. But it is easy to imagine a near future when the emerging technologies of Web Services and the Semantic Web enables products and services to be found just when they are needed and under the best terms[Petrie-Bussler-03].
The logical progression of these developments leads to increasing digital interaction of individuals for both commerce and play. Play is already happening with Internet-based games. We have the glimmerings of individual commerce with E-bay. It is also possible that new search technologies will suffice for a majority of commodity goods and services.
The technology is not here yet, so you can't easily advertise, in a way that a software program can read, that you can supply the service of, say, designing database schemas. Could you, you would probably no longer work for one company, but rather sit at home and take jobs as they were offered electronically from customers who found you via automatic search. That's the vision of the NCI.
A contributing technology might be social networks. Suppose that you use these networks of trust to find partners in some new enterprise, whether to build a new Burning Man camp, or develop a new start-up? Or further out, suppose that none of you have a project in common, but as each person works on his/her individual project, there are pieces of information that would help one of your friends with their own project? Now you are engaged in collective work.
You may all be working on designing parts of the next space shuttle. Or you may each be writing a novel. Your design for thermal insulation, or the research you do on eighth century architecture may be of use to some one else. Technologies for facilitating such information sharing for mutual benefit is a natural future step in the increasing sophistication of Emergent Collectives.
But a mistake of the NCI is not to include the soft component of Emergent Collectives. For instance, until the USA has a universal, single-payer health system, most people won't want to be independent contractors handing over $700/month to a health insurer each month. Thus, the NCI might occur first in western Europe. Emergent Collectives are based on social factors as well as technologies. Is there any hope of analyzing such factors?
In the '70s, computer scientist Doug Lenat entered in a computer game of competing naval fleet designs. Doug developed a computer program that played human players and won in a national contest in 1981 and again 1982. His program won by using the rules of the game in a way that no one had anticipated: "exploiting counter-intuitive loopholes"[Lenat-82]. The secret to Doug's success was computer simulation prior to playing, using the game rules, in order to find these unexpected winning strategies.
If business and government leaders wish to not to be blindsided by new technologies, and their own prejudices, then perhaps they can develop simulation techniques to predict new Emergent Collectives. There exist today relevant technologies. A social simulation might be a new kind of "SimCity" [5] or "The Sims" [6] that could help determine, for example, the effect of a single payer system or the behavior of vehicle drivers or the spread of 802.11 access nodes. Cellular Automata [7][8] are already a kind of simulation. There are already simulations of business models [9] and organizations [10], and even construction projects [11]. Simulating with computational logic the effects of the interaction of various policies and laws is the research subject of at least three groups at Stanford University alone[12][13][14].
Our modest suggestion is that industry and government actively look for technologies that might have the characteristics of Emergent Collectives, perhaps using simulations in order to avoid being blinded by previous models. They can then attempt to build new business and governance models that take advantage of new technologies, instead of attempting to suppress them. This approach, while admittedly futuristic, has the advantage that it offers new possibilities to businesses and governments imaginative enough to travel along with the people and technologies forming new collectives.
Acknowledgment: Marty Tenenbaum provided early inspiration for and an insightful review of this article.
Notes :
[0] Asynchronous Transfer Mode (ATM) protocols make single-path connections between source and destination using circuit switches. TCP/IP data packets flow through routers and their path is not guaranteed.
[1] http://freenet.sourceforge.net/
[2] http://www.bizforum.org/whitepapers/wireless.htm
[4] http://www-cdr.stanford.edu/~petrie/802.11-Stockholm-2001/
[7] http://www.fourmilab.ch/cellab/
[8] http://www.cnd.mcgill.ca/bios/bub/CAs.html
[9] http://www.valuebasedmanagement.net/methods_business_simulation_modeling.html
[10] http://www.epm.cc/products.html
[11] http://www.stanford.edu/group/CIFE/CEE320C%20Spring%202002.htm
[12] http://logic.stanford.edu/
[13] http://eil.stanford.edu/regnet/
[14] http://complaw.stanford.edu/
[Lenat-82] D. Lenat, "Heuretics" Theoretical and Experimental Study of Heuristic Rules", Proc. of AAA-82, AAAI Press, pp. 159-163, 1982.
[Hofstadter] D. Hoffstadter, Godel, Escher, and Bach, Basic Books, 1979.
[Mitchell-93] M. Mitchell, P. Hraber, and J. Crutchfield "Revisiting the Edge of Chaos: Evolving Cellular Automata to Perform Computations," Complex Systems, 7, pp 89-130, 1993.[Petrie-Bussler-03] C. Petrie and C. Bussler, "Service Agents and Virtual Enterprises: A Survey," IEEE Internet Computing, 7(4), July/August, pp. 2-12, 2003.
[Rheingold] Smart Mobs: The Next Social Revolution, Howard Rheingold, Perseus,2002.
[Vaidhyanathan-04] S. Vaidhyanathan, The Anarchist in the Library, Basic Books, 2004.
Addendum 1
Wikis, are a
subset of HTML plus online editing, that have been around since the
mid '90's but which have become popular in the 21st Century as
people discovered Wikiscould be used
to collectively build repositories. These applications are examples
of Emergent Collectives that are specializations of the larger
WWW.
Addendum 2
The Cathedral and the Bazaar is an insightful description
of how hackers contribute to the open source movement. The emphasis in
this paper (Emergent Collectives) is on people's desire to contribute
to a collective efflort. This is clearly illustrated in The
Cathedral and the Bazaar but had this been the emphasis, a
different analogy might have been used. One might say that the
motivation in a bazar is commerce for profit and the motivation in
working on a cathedral is to contribute something, even if only a
brick, to some great collective effort.