Recently I attended Planetwork, a three-day conference in San Francisco on new ways of using computers and online technology to help people work together for a better future(1). Some new communication and organizing tools, many available now, could make a big difference in AIDS and health. I have a background in computers and was the only AIDS writer there, and will report to the community on some of this new work, and why it is important for us.
Over 300 people, some from as far as London, heard 100 presenters, many of them well-known leaders in the computer industry, in three parallel meeting tracks. The conference was supported by registration fees and by industry partnerships, mostly with small technology companies.
This article will look at software for extending existing social networks, as a way to help people around the world work together on common goals and projects -- even when some of them have no computer, email, or Web access.
A key theme of the conference was establishing trusted communication online, to assist existing networks of friends and colleagues who already work together and trust each other. A semi-official "white paper" for the conference ("The Augmented Social Network: Building Identity and Trust into the Next Generation Internet") explored some of these ideas(2). "Trust" in this case refers to personal judgment, especially knowing someone and having confidence that you can recommend him or her to associates for a working relationship -- not to computer security against technical attacks. [And it does not at all refer to the so-called "trusted computing" concept, a major project of Microsoft and others to encrypt every email and document in your computer so that government and corporations can control what you do with it. For more information on that effort see John Markoff, "A Safer System for Home PC's Feels Like Jail to Some Critics," New York Times June 30, 2003; or see Richard Stallman, "Can You Trust Your Computer," http://www.gnu.org/philosophy/can-you-trust.html]
To show what augmented social networks could mean for AIDS, here is a fundraising scenario we expect to be happening within a few years. The software required either exists today or could easily be written. What will take time is for people to learn about this possibility and start using it together.
For this example we will take a hard problem -- raising money for local AIDS and health clinics and other grassroots projects in developing countries. But the same tools could work for many purposes.
It is well known that small amounts of money can save lives or otherwise make a big difference in poor countries (see the new documentary film, A CLOSER WALK, http://www.acloserwalk.org/). Today almost all money donated goes through governments, big nonprofits or churches, or other large organizations. For some projects this is the best or only way to proceed. But many people would be more willing to give directly through a personal connection. (We suspect that these two kinds of giving will be more synergistic than competitive -- that if donors had good ways to give directly they would become more personally involved, leading to more political will to support government, nonprofit, and church programs, rather than less.)
The problem today is that the people and personal networks who can donate are far away, both geographically and socially, from the people and personal networks where their money could do the most good. Most potential donors living in the U.S., for example, do not know anyone in any developing country -- especially in remote villages where many of the people live, and where a little money could often go farthest. And most donors today do not personally know who is really doing the most important work, but must rely on public relations and carefully crafted images, which they know very well are unreliable -- discouraging commitment and contributions. On the other side, those who have the first-hand knowledge usually have no good way to reach many donors and establish credibility.
How could computers help? Imagine that a few years in the future, you hear that $25 could save a life in poor areas through health care or famine relief (as is the case today). Perhaps you want to make a small contribution directly to an organization or person on the scene -- for example, a local group somewhere in Africa that is doing superb work but may not be part of any big charity or international organization. You want to contribute based on the personal recommendation of someone you trust. Usually there are people you trust -- but none of them are at the scene, so they cannot help you directly.
So instead of making dozens of phone calls to try to find a chain of recommendations that reaches from your personal network into villages in Africa, you go to a social-network Web site where basically anyone in the field (global AIDS or health in this case) could publish a profile for themselves -- including a list of people or groups they recommend.
For example, since you are reading this newsletter, perhaps you trust me. I do not know who in poor countries could best use your contribution -- but I do know well-regarded doctors, activists, and others who work or volunteer there, whose judgment and recommendations I trust. Perhaps none of them know what is really happening on the ground in a particular area, but they are closer than you or I, and will know people closer still.
Health activists and professionals who work regularly in developing countries could create a profile of themselves on a social-network Web site for international AIDS, or international health. Their profile would include a list individuals and organizations that they recommend as doing good work -- and who could also put their own profiles on the site, where they list others they recommend. Those listed need not have access to computers, nor speak English or whatever language is used on the site, because their colleagues who want to recommend them and help raise funds could work with them to prepare and submit their profiles. These profiles might include specific projects that need doing, with a budget for each.
At the simplest level, the way social-networking programs work is that you can click a link on anyone's profile to see their network of friends or trusted colleagues. Then you can click on any of those profiles to see that person's network, etc. Just with this much, you could start with one person or several people you trust, then check through the networks partly by trial and error to get closer to the kind of project you are looking for -- either geographically (Africa, for example), or by other categories. But software tools (discussed below) could greatly help.
Networks of personal recommendation have always been important in almost every human activity. But usually it is difficult in personal networks to go through a chain of more than two or three referrals -- especially across international borders, time zones, and language barriers, or into regions that have no telephones, computers, or other modern communication. Even when communication is possible it is not feasible to make dozens of phone calls in different languages to explore social networks and establish a trusted chain of referrals, all for the sake of perhaps a $25 contribution.
But with a social-network Web site, potential donors or anyone else can explore these personal and professional networks whenever they want -- without necessarily setting up their own profile or making any other arrangements in advance. They can start by finding one or more people they trust who have profiles on the site, and then look for chains of trusted referrals to projects they want to support.
I would guess that in a specialized area like global AIDS, the average length of the shortest personal-referral chain from, for example, an interested U.S. citizen to a clinic or other project in a village in Africa, would be less than the proverbial "six degrees of separation," but more than the two or so degrees that can comfortably be managed informally. The first time a potential donor uses this system, he or she might spend an evening exploring the links through these social and professional circles, looking at perhaps several dozen of the thousands of profiles on the site, to find one or more chains of perhaps three, four, or five connections that work for them -- ending with an organization or individual they want to help. Since part of the work of setting up the Web site would be to provide a way that funds could be transferred legally, inexpensively, and conveniently, with a few keystrokes the donation could be done. Each recipient organization, individual, or agency would have a way to be notified that the contribution had arrived and credit was available; that would be part of their profile. They would also be able to contact and thank the donor and tell how their contribution was used, unless the donor asked not to be identified publicly.
* In AIDS we already have networks built on personal trust and long-time working relationships, and these function well. But we need better ways to connect the separated networks, so that they can coordinate better when appropriate. Social-network software -- especially databases of personal and professional recommendations, within a given field or for a particular purpose -- could support larger projects and activities, without the problems of a centralized, top-down command structure (which has not worked well in AIDS).
* Speakers at the Planetwork conference were quite familiar with academic studies of social networks, at Stanford University and elsewhere. One of the findings has been that important benefits like learning about a new job are more likely to come from loose social connections (for example, an acquaintance of a friend) than from tight connections (being part of the same corporation, school, or other social structure). Social-network software extends these loose connections.
* Software tools can help in searching the network for a credible chain of recommendations. For example, a potential donor could list any number of profiles of people or organizations they trust, as a starting point -- and also list any number of potential recipients, as a goal -- and ask the system to find any connections automatically. They could also specify potential recipients either geographically (all organizations in Africa, for example), or by kind of mission (such as caring for orphans), or both -- and then find the chains of trusted links and list them by strength of the connection. Searches could also specify keywords, names, or subject areas in the profiles. The software could also show who throughout the whole database recommends a particular project or organization. Other tools could allow users to click to see a visual picture of the whole social-network database, showing how much it is a unified network vs. a collection of separate groups -- and if the latter, then who are the crucial links between the groups. Potential donors could use their choice of such software tools, or just follow links one by one to find groups they want to support.
* Potential recipients (and others in the database) can speak any language, and do not need to have access to computers -- since those who are recommending them will be motivated to help them enter their project description and other profile information.
* The profiles could belong to individuals or to organizations interchangeably (in that the system handles them the same in either case, although potential donors can specify one or the other in their data searches or views). Profiles could also belong to other entities, including software robots (computer programs that do things you might expect a person to be doing) - - perhaps created by the database administrators to make the system work better. For example, a robot's profile could look like a person's profile, but change every day to reflect reputations or other information in the entire database -- perhaps in order to direct users to hot topics, special opportunities, and other time-sensitive information. Robots might also help donors combine to fund larger projects than any one of them could handle individually.
* A number of social-networking Web sites already exist, providing proof of concept. One often mentioned at the Planetwork conference was Friendster (http://www.friendster.com), which connects people for dating or making new friends through their existing social networks, unlike the usual matching services that connect strangers. Any such system needs a critical mass, and Friendster is now taking off on the East Coast of the U.S.
* Participants in social-networking software are not interrupted when their referral network is traversed; they do not have to answer a phone call or email. This makes possible much more exploration of existing referral network. Donors deciding how to use their money can look through hundreds of person-to-person links, instead of only a few.
* All the information on these sites is public and voluntarily given. Anyone who creates a profile can choose what they want to reveal about themselves. At the same time, social-network sites have strong pressures for honesty and accuracy, since peoples' friends and colleagues are on the same site, and can see how one presents oneself.
* Some churches raise money through public donations, in which members of the congregation parade near the collection basket in front of the group, and everybody can see what is dropped in. Other churches provide envelopes so that members can donate privately. The online referral database offers both possibilities, and can allow each donor to choose whether or not to make their donations public -- which I think will be the prevailing system.
* A fundraising site could be useful for connecting people even if they do not want to donate. Such visitors should be welcomed, as they increase traffic and general usefulness of the site -- and might change their mind and donate after all. In addition, it would be costly to police the site to keep out those who do not intend contribute, since the whole idea is to allow donors to explore possibilities, without any guarantee of finding a match that speaks to them.
* Donation guided by online exploration of networks of personal trust, relationship, and recommendation could become widely popular. There are many successful models and precedents for online activities that involve other people -- either in one's own social network, for games or discussion among strangers, and for activism in issues important to society in general. Here is a way to combine these kinds of interpersonal activity online -- and participate meaningfully in donation decisions, with the help of the best personalized expert guidance anywhere.
* This kind of social-network software could help in many areas outside of fundraising -- for example, in improving elections of political candidates. Already, probably most voters decide whom to vote for mainly through personal recommendations. But today the chain of personal referrals, if traced back, is likely to be found to originate in television ads or other crafted manipulation (you may have noticed that news reports are most informative about what is really going on just after the election, when they can tell the truth because it no longer matters). Online referral networks could improve the political process by providing more people with trusted connections to a variety of experts -- who, unlike political advertisements, do not always have pre-established positions they are paid to support regardless of the facts. Another political use would be to help people judge the credibility of action alerts, even from organizations they do not know -- providing a much larger potential base for public mobilization and response to the most important alerts.
* A down side of referral networks is that recommendations can be traded -- you scratch my back and I scratch yours -- or even sold. This kind of activity, which can span the range from normal social process to corruption, can occur wherever personal referrals are used. For example, several decades ago scientific papers started being rated by how many future papers referenced them; collusion quickly developed and was quickly recognized, but is still with us today.
In the referral database, one way to reduce this problem would be to have rating agencies, companies in the business of investigating claims and publishing relevant findings, good or bad, on the database. They would not need to audit everyone, but might check those who are most important on the database, for example those who can change large flows of money. Rating agencies could also help with the "transitivity" problem: if A recommends B and B recommends C and C recommends D -- with all the recommendations within the same professional context and purpose -- then to what extent can those who trust A's recommendations also trust D's? (If enough money depends on answer -- for example, if A is a celebrity and D is a widely respected expert at a critical scene -- then an independent agency might investigate and provide its findings on how well fans of A know what they are getting with D.) Donors could follow or ignore these agencies' reports as they wish. [Note: I added the idea of rating agencies, which I not hear at the conference. Many presenters mentioned online reputation, however.]
Interestingly, the four kinds of entities on the database that have been discussed so far -- individuals, organizations, robots, and rating agencies -- all use the same format, a profile of the individual, with an unlimited number of referrals to others. The software need not even distinguish between these different entities. (There is at least one other kind of entity - - imaginary characters that could represent factions, themes like reconciliation, or abstract themes that may be nameless -- or could serve to bring separated social networks together. These fictitious people are also handled by the database exactly like real people. Such characters have developed, totally unplanned, in existing social-network software, and can become among the most popular "persons" in the system -- showing the need for this kind of device.)
One of the most useful presentations at the Planetwork conference was "The Multiple Dimensions of Emergent Media," by Mark Graham, a computer conferencing pioneer who founded Peacenet and was president of Whole Earth Media, in addition to many other credentials. His talk referred listeners to over 30 Web sites of new-media tools and experiments(3). He also suggested some guidelines for successful projects in this field -- guidelines not just for technology, but for effectiveness and building constituency in whatever one does:
* Think big but keep it simple;
* Solve problems, don't invent solutions;
* Do your homework -- know what has been done before;
* Connect with and learn from others;
* Be true to your vision;
* Focus on deliverables and keep it real.
Ultimately the purpose of technology is people. The point of the fundraising scenario outlined above is to make possible a personal connection between people who have very different lives, and who otherwise would not have a trusted channel between them. With such channels available, donors can make better decisions, and are more likely to become personally involved.
Next: Blogs (Web logs) and their software tools, Moveon.org, online communities, RSS and news aggregation, instant uploading and publication of photos from camera-equipped cell phones, online video editing, advanced Google techniques, and more.
[Note: This article first appeared in AIDS Treatment News #392, June 27, 2003, under the title, "AIDS, Computers, and Organizing: Part I, Toward a Revolution in Fundraising? (A Report from the Planetwork Conference)"
(1) PLANETWORK: Networking a Sustainable Future, conference June 6-8, 2003, at the Presidio in San Francisco, http://www.planetwork.net
(2) THE AUGMENTED SOCIAL NETWORK: BUILDING IDENTITY AND TRUST INTO THE NEXT GENERATION INTERNET. For a 2-page abstract and a link to the full 77-page paper, click "WHITE PAPER" on http://www.planetwork.net
(3) Links mentioned in Mark Graham's talk are at: http://mark.path.net/planetwork
Copyright 2003 by John S. James. OK to copy and distribute for non-commercial purposes.