Thu

May 15
2008

Andy Oram

Andy Oram

Yochai Benkler, others at Harvard map current and future Internet

Harvard's world-renowned Berkman Center for Internet & Society is celebrating its tenth anniversary with a conference called Berkman@10. I'll report here on today's sessions, which were organized as a fairly conventional symposium (although as loosely as one could run it with 450 attendees). Tomorrow will be set up as an unconference, where the audience defines most of the topics and self-organizes into small-group discussions.

Whither peer production--wither peer production?

The Internet is not monolithic--as speaker after speaker today recounted--so it's not fair to expect an organization studying it to be monolithic either. The Berkman Center is diverse to the point of being hard to characterize, as I'll detail later, but one theme that echoed through the day was the collaborative production of value, or "peer production" as the economically-minded like to call it.

Yochai Benkler, author of key Internet analyses such as The Wealth of Networks (available in both printed form and as a PDF) and Coase's Penguin, or Linux and the Nature of the Firm, set out his stake near the beginning of the day when he called on the Center to move from creating tools to examining social change, such as what it's like for dissidents around the world to be able to work together.

Wikipedia is a central piece of evidence in Benkler's case--along with Linux, a connection I'll explore later--and also forms Exhibit A in the recent book The Future of the Internet (And How to Stop It) by Jonathan Zittrain, cofounder of the Berkman Center and keynoter at today's conference. (I won't cover the keynote because I have already reviewed the book at length.)

So it's quite in keeping that Jimmy Wales joined Benkler in an afternoon session at the conference. Benkler laid out the traits distinguishing both the process and product of peer production from what we're used to getting in the market: peer production is unpredictable, unstable, loose, and people-driven. Wikipedia matches those criteria so well that Wales admitted its fate is still up in the air.

David P. Reed, a key Internet researcher, described how science--another, very different culture of peer production--has been corroded during the past few decades by commercialization and over-regulation (specifically, the various Offices of Research Integrity that lobbyists have forced scientists to answer to). He then asked whether Wikipedia's turn for regulation by government will come. (Details were added to this paragraph on May 16 after a conversation with Reed.)

Wales didn't address the precise question, but admitted that "humans will eventually screw up Wikipedia just as we manage to screw up everything else," yet promised to keep it true to its mission as long he can.

The challenge for peer production, according to Zittrain, Benkler, and Wales, is to avoid seizing up and imposing new controls when things go wrong. Instead, one must learn to deal with the damage through side channels.

Money is a secondary question. Benkler pointed out tanies now pay employees to contribute to peer-produced products (Linux, where most development is now done by the employees of various companies, provides an obvious example) and said the ultimate impact of this trend is unknown. Possibly, the current in-rush of volunteer labor is a temporary phase in the evolution of peer production. In any case, we should carefully examine the process so that the projects can remain fair and keep people motivated when some are paid and some are not.

Besides production for money and production for fun, I noticed another motivation for contributions when Wales mentioned that India has become an increasing source of Wikipedia pages as computers and Internet access spread there. I sense that regional and cultural pride can drive many efforts--the feeling that "if that city over there can do it, why can't we?" This suggestion shows the need for models, which I'll describe in another section.

The conscious commons

Although Wikipedia was described by conference participants from many angles--as an amazing example of volunteerism, a triumph of people's ability to resolve conflicts, and so on--I think one key trait has gone unremarked: Wikipedia has reached such a high level of value that participants are willing to put its success above any other considerations. No matter how much someone desires to express opinions, they know that fighting hard enough to damage the entire venture would be counter-productive. So people usually settle among themselves. In other words, they are conscious about protecting their commons.

Think, as a metaphor, of a town commons where people not only graze their cattle, but water the grass and spread around the manure so that it's properly fertilized.

As civilization develops, we tend to get lazy about maintaining our commons. We discover the benefits of turning functions over to large companies (they gain efficiencies from scale and from the use of professionals) or to governments (who provide transparency and equitable distribution of resources, when done right).

This trend is not limited to advanced economies. Esther Dyson reported that, after she encouraged Internet users in one developing African nation to share wireless networks with their neighbors, someone complained to her that it was up to the government to reduce costs and provide wider Internet access.

But nowadays, sophisticated manufacturing methods reduce gains from scaling, and the benefits of training ordinary people to perform useful roles outranks the value of employing a small professional elite over and over. In fact, Benkler mentioned that learning is a key part of peer production and a driver of its success.

Meanwhile, transparency has become more available to all actors, including governments, through communications networks.

So we've started to turn back to ourselves in order to support what we hold in common. The modern equivalents of barn raisings are the groups in the 1990s who came together to wire their local schools (before WiFi made that less necessary), or people who made the news recently by installing solar panels on each other's homes.

Most notable, for the sheer size of the effort, is the self-mobilization of communities, both locally and nationally, in the wake of the Hurricane Katrina floods and the failure of government response. My own synagogue has sent two building teams to the New Orleans area over the past two months; hundreds of others have made similar donations.

The remaining problem to solve is the equitable distribution of resources mentioned earlier. For this, peer production and the conscious commons have to go global. We need to feel an immediate connection to all creatures around the globe--and that leads to the most audacious proposal that came up at the Berkman Center today.

What will a Harvard for six billion people look like?

This bold discussion began with a taunt lobbed into the arena from a surprising corner, former FCC chair Reed Hundt. In my opinion, Hundt's tenure in the mid-1990s stood out for the FCC's recognition that its landscape would be overwhelmingly changed by the evolution of networks and the media transferred across them, but in the end proved too timid and compromised to pursue the implications of these insights.

There was no timidity, though, to Hundt's proposal that well-endowed universities such as Harvard help the six billion people who are now deprived of the education they need to make a decent living.

Charles Nesson, a famous attorney and cofounder of the Berkman Center, picked up the tune without missing a beat. He talked of the enormous amount of high-quality online material that Harvard is making available. The Faculty of Arts and Sciences recently voted to open access to all scholarly articles, and the law school soon followed suit.

(Ironically, Harvard is one of the few Boston-area colleges that doesn't allow the public into its brick-and-mortar libraries. This policy is understandable though, because to make them open would overwhelm them with the throngs of odd creatures that circulate among the literate classes who inhabit Cambridge.)

Dyson then interrupted with an astute distinction between content and helping people to teach themselves, a goal that is people-intensive and requires a lot of side activities such as making sure children have enough to eat.

Nesson countered by saying that the goal was not to provide sterile content, but to provide content that would stimulate children's interest and sense of play, which in turn would lead to a peer production of education.

As an aside, Dyson mentioned an invention she saw at Microsoft's Bangalore facility. Using software that allows a USB port to be multiplexed, one computer through a single USB port can support up to eight mice. Thus, eight children can play a game or manipulate items on a screen. This doesn't turn a Windows system into an XO (One Laptop Per Child) network, but it's an advance for needy communities.

Designing for cooperation

There was much worth retelling in Benkler and Wales's session, although a good deal of it can be found in other works of theirs. Benkler pointed out that our economic system is designed around the notion of human beings as "selfish rationalists," but that no society ever studied has many people who actually behave that way. He said at most 30% are primarily motivated by material rewards.

Now, as we know, the people so motivated can be extraordinarily productive, creating some of the most important technological changes in history. However, we also know that a large part of that 30% lie, cheat, and steal. Anyway, Benkler seems ready to try something different.

His talk involved questions instead of answers--a research agenda rather than a curriculum. He suggested we draw on the disciplines of organizational sociology and experimental economics. He laid out the intrinsic motivations we want to encourage for peer production--solidarity, empathy, trust, fairness--and started an exploration of extrinsic motivations.

The extrinsic motivations include rewards and punishments, along with transparency. The latter leads in turn to reputation systems, which embody twin goals: control (so others know whom to trust) and motivation (because contributors expect future rewards).

Wales, in a private conversation, demonstrated the cooperative spirit in his comparison of Wikia Search (which I described in an article yesterday) with traditional search engines. He said we tend to place too much faith in algorithms. Good algorithms are certainly valuable--particularly in searching for the long tail, as when someone knows only a few phrases in a book or song--but don't have to be sophisticated enough to prevent all gaming of the search engine.

"If the community decides something is spam, they can simply block it outright," he said. You don't have to insist on creating a search algorithm so smart that it pushes spam down in the results list.

In his presentation, Wales said that a majority vote is not enough to ensure quality content. If only 70% of editors like a Wikipedia page, something is still wrong with it. Therefore, discussion continues until everybody is happy except a few unreasonable people who are usually disruptive in other ways as well. He said that unanimity is not the goal, but consensus.

His contrast of unanimity and consensus struck home with me, because the exact same distinction (using the same words) is made by community organizers in the international network created by the historic Industrial Areas Foundation created by Saul Alinsky in 1940. As a volunteer for a local community organization, I know its power to build consensus as well as its success at building power.

And it's worth nothing that Barack Obama spent years as a community organizer with the IAF, while Hillary Clinton wrote a thesis on it as a young student (and turned down a job offer from Saul Alinsky). Someday, community organizing experience could well become a prerequisite for a management job in any business or government position.

Seeding and modeling

Another aspect of peer production escaped discussion today. Zittrain mentioned in passing that Wikipedia began as a set of comment forums on Nupedia, which had reached the limits of its growth with seven articles from paid experts. Small though this starting point was, I believe these seven seeds were critical to show what could be done and give volunteers models to emulate.

Consider also that a worldwide, Internet-based development effort on an operating system could not begin (although all the variants of BSD were produced by volunteers using more traditional team methods) until Linus Torvalds seeded the effort by publicizing his budding kernel.

So peer production requires models. Not coincidentally, Dyson said that social change also requires what she called models for courage. An Asian journalist from the audience claimed that most Chinese Internet users think government censorship is a good thing. People will push for change--but most of them to see someone else start.

The dilemma of openness

"Open" was probably the most frequently uttered word of the day. Nesson, in his opening remarks, chanted of "open talk, open access, open education..."

But the problem, as I explained eight years ago, is that openness, in a context of unequal power, just puts more power in the hands of those who already hold it--those with the guns, the funds, or other ways of controlling the public agenda.

In small ways, blogging and efforts such as the Sunlight Foundation (represented by its head Ellen Miller at the conference) take power out of the hands of its current possessors and distribute it more widely among the public. The Internet can also help democratize fund-raising, as both the Ron Paul and Barack Obama campaigns proved. But there is still a lot more that governments and large institutions can do with information than ordinary people.

The Berkman Center broadens

Although Berkman's ten-year anniversary formed the occasion for this conference, celebrating the anniversary was not its main goal. Thus, I saved a description of the Berkman Center for the end of this article.

As I mentioned earlier, the center far from monolithic. It is a conglomeration of many people, both lawyers and non-lawyers, who study the Internet and add their efforts to empower its users.

Legal studies of the Internet were by no means a new field when the Berkman Center was founded. In fact, current director Terry Fisher says such studies were already a "fad." The Berkman Center is distinguished in many ways, such as by its independence (although it has corporate sponsors in addition to Jack N. and Lillian R. Berkman's gift) and the caliber of its professors and fellows.

But in my opinion, the most salient contribution of the Berkman Center is its devotion to new research instead of pure theory. (Another such research center is Do Tank.)

At the conference, one example of this valuable approach was a fascinating visualization of blogs in Iran--the fourth largest blogging community in the world--and of which sites are blocked by the Iranian government. As one would expect, most blocked sites are written by secularists, reformists, and ex-patriots.

It's also impressive, however, that most sites by these groups are allowed through the government's filters. Too much blocking, as the Berkman researcher said, would lead to a loss of legitimacy for the government.

Among the major research and production activities at Berkman are:

  • The OpenNet Initiative--a tracking system that reports government censorship worldwide
  • Global Voices Online--a blog for people who previously had no way of reaching the public outside their nations
  • StopBadware--a service that recognizes infected web sites and (in cooperation with Google) interpolates warnings when users try to visit them

These projects go far beyond the field of law, and in fact, law school head Elena Kagan announced at the conference that the center was moving outside the law school to become a general Harvard institution.

The Berkman Center also exemplifies the openness they speak about. Long-term and temporary associates mingle with invited guests and passers-by. If you're in the Boston area and are interested in where digital networks and media are heading--technically as well as politically--Berkman events are among the best places to spend your time.

 
Previous  |  Next

0 TrackBacks

TrackBack URL for this entry: http://blogs.oreilly.com/cgi-bin/mt/mt-t.cgi/6499

Post A Comment:

 (please be patient, comments may take awhile to post)






Type the characters you see in the picture above.