[an error occurred while processing this directive]

CORANTE



INDUSTRY NEWS today's headlines

Biotechnology
Communications
E-Business
Internet
Law & Policy
Personal Technology
Venture Capital

WEBLOG COLUMNS today's highlights

Amateur Hour
the "me" in media

The Bottom Line
the economics of IT

Brain Waves
neurons, bits & genes

Connected
nodes & networks

Copyfight
the politics of IP

Corante on Blogging
in media res

Got Game? NEW!
the future of play

IdeaFlow
creativity & innovation

In the Pipeline NEW!
drug discovery

Living Code
biology & information

Many-to-Many
social software

Moore's Lore
new technology

Open Mind
open source


About this site

Social software blends tools and modes for richer online social environments and experiences. Some examples of social software are weblogs, wikis, forums, chat environments, or instant messaging, and related tools and data structures for identity, integration, interchange and analysis. For more, see Liz's primer on what we're up to.

This group weblog is authored by Elizabeth Lane Lawley, Ross Mayfield, Sébastien Paquet, Jessica Hammer and Clay Shirky.



Corante.com Navigation

[an error occurred while processing this directive]
















powered by WebCrimson






MANY-TO-MANY: social software

By Elizabeth Lane Lawley, Clay Shirky, Ross Mayfield, Sébastien Paquet & Jessica Hammer


Sunday, August 31, 2003

Anti-social software: Interesting attack on /.

By Clay Shirky

There's a seemingly automated attack on the slashdot moderation sytem. The attack technique is quite simple: flood the comments with pointless crap. Each of the attack posts is titled Mod Parent (Up|Down|Sideways), a reference to slashdots moderation system, and each post is accompanied by the single word Crapflood, linked to the non-existant URL slashdotbot.com. (You can see the signature of the attack in today's Dot Com Era Fads story.)

The attack is interesting because although the form of the attack is a Denial of Service (DOS), it is aimed at slashdot's social infrastructure, not their technological one. slashdot is an example of political systems arising out of social dynamic. The "How did the moderation system develop?" section of the Slashdot FAQ reads like Federalist Papers #10 with the role of james Madison played by a Jolt-addicted perl programmer.

The effect of the attack is relatively simple: if moderators mod the crapflood down, they will use up their mod points (each moderator can affect no more than five comments.) If the moderator points are used up in moderating the crap down, they can never flag the good stuff. And if they can't flag the good stuff, the slashdot comments system dies, because it is an unreadable mess without the moderation system to help.

The answer, of course, is not to moderate the Crapflood down, but that requires moderators to override their natural instincts, and to leave obvious grafitti intact, whcih will create other signal to noise problems down the line.

Will be interesting to see how persistent the attackers are, and how the infected body politic responds.


7:53 am |

Thursday, August 28, 2003

Blogs for What Business?

By Ross Mayfield

Jimmy Guterman's new piece on business blogging (sub. required) is sure to cause a stir. He charges the blogging community as being "self-absorbed and elitist" and says its not essential for business. He cites a Forrester study to back up his claims:

You don't have to believe me on this. Finally, some data asserts that blogs are hardly a popular pursuit. If anything, blogging is more marginal than its critics contend. Forrester Research (FORR) conducted an online survey of 3,673 people and found that 79 percent of its respondents had never heard of blogs, 98 percent had never read one, and 98 percent said they'd never pay to read or write one. Blogs can be wonderful things, but if a mere 2 percent of Internet users read blogs, the pastime is far from mainstream. The Forrester survey notes that the typical blog reader has been using the Web for an average of six years. For the most part, blogs feature the Net elite writing to the Net elite. This continues to be the case only as long as the elite are underemployed.

I believe what Jimmy is saying is that there isn't a consumer market for blogging and that it isn't essential for businesses to address it. The problem is we are at the very beginning of a technology adoption lifecycle. Some serious companies have forecasted this market to grow and made their bets accordingly. Every time a journalist tries to wrap themselves around the existing market, what's visible are early adopters. What stands out are the leaders in using blogs for publishing, who benefit from preferential attachment as the earliest entrants. And if you take the innovator dialogue to seriously it looks like a one ring circus.

The other story folks pick up on is unclueful attempts by businesses and PR firms to market to bloggers as an emerging and influential segment. Any attempt to treat bloggers as a segment will fail. Today the influence of participants who act more as producers than consumers is the attraction. The number of participants is growing at 400% per year, and that's before AOL's entry.

But the real story in the consumer market is how increasing numbers of real people are using blogs , but as a way to communicate an form their own communities. Its that skinny tail of the power-law distribution that's going to wag the market. A way to share with friends, communicate post-by-post and remain open to new people joining your community. Conversational Networks provide the most value to your average Jane.

Rick Bruner does make the case that there are lots of businesses using blogs in the consumer market and points out this is like the web in 1995 and where the weblog as publishing market is headed. And many of them are making money. I agree that more evidence in this area would help, always does, but give it time for these new ventures to tell their story.

There is another story of weblogs and business that is less visible becasuse the real action is behind the firewall. At Socialtext we are adapting weblogs for use within enterprises. Weblogs are one Enterprise Social Software tool, because they are necessary but not sufficient for communication and collaboration.

The enterprise market is entirely different than the consumer market. What is in common is an efficient, and dare I say fun, way of having conversations that contribute to productivity. Maybe its time we start telling more of our customer stories, but the distinction between consumer and enterprise needs to be made.


6:13 pm |

It *is* all one big site

By Clay Shirky

Tom Coates of plasticbag points to an interview with Adrian Holovaty, in which Mr. Holovaty says:
...I believe in my heart that people should come up with their own publishing methods. Frankly, it’s boring to surf the blogosphere and see so many sites using the same, tired weblogging tools. The same basic templates, the same “post a comment” form, the same URL schemes… It’s almost as if they’re all small parts of one huge site.
Leave aside the first-order stupidity of that idea ("I believe people should come up with their own sort algorithms. It's boring to see everyone sorting data the same way") -- what Holovaty really missed in his worship of individuality over substance is that the weblog world is, to a first approximation, one huge site, and this is part of what makes it so vital.

This is the conclusion of "Bursty Evolution of Blogspace" (free reg required, and worth it for this paper), which I wrote about back in May. This paper concludes that there is, at the core of the weblog world, a Strongly Connected Cluster (like a giant clique, a set of weblogs far more interlinked to one another than one would expect from random linkages), that is better analyzed as a single vast site, than as a handful of individual sites. (Note that the Strongly Connected Cluster is much larger than the oft-referenced group of "alpha bloggers", however calculated.)

The Strongly Connected Cluster is new, having come into its current shape in mid 2002:
"But up to this point, blogspace is not a coherent entity -- the overall size has grown but the interconnectedness is not significant. ... In 2002, however, a threshold behavior arises, and the size of the component increases dramatically, to over 20% by the present day. This giant component still appears to be expanding rapidly, doubling in size approximately every three months. Clearly this growth cannot continue and must plateau within two years."
Like a colony creature made of independent organisms, whether a beehive, an ants' nest, or a colony of fireflies, the individual output of webloggers is far more synchronized than any random system would be -- if it weren't, sites like Technorati and blogdex would be no different than random surfing.

Holovaty came this close to articulating the most important difference between weblogs and the earlier "personal home page" era, where people linked to New Kids on the Block sites rather than to each other, but he missed it, because he mistakes the surface for the substance. Having lots of small parts of one huge site turns out to be a really good idea.


12:58 pm |

Tuesday, August 26, 2003

Wikis, Grafitti, and Process

By Clay Shirky

Every time I show a wiki to someone who has never seen one, I invariably see the same two reactions: "That's pretty cool", followed seconds later by "It'll never work." This second reaction is understandable, as wikis take a radically different attitude towards process than almost any other piece of group software.

Process is an embedded reaction to prior stupidity. When I was CTO of a web design firm, I noticed in staff meetings that we only ever talked about process when we were avoiding talking about people. "We need a process to ensure that the client does not get half-finished design sketches" is code for "Greg fucked up." The problem, of course, is that much of this process nevertheless gets put in place, meaning that an organization slowly forms around avoiding the dumbest behaviors of its mediocre employees, resulting in layers of gunk that keep its best employees from doing interesting work, because they too have to sign The Form Designed to Keep You From Doing The Stupid Thing That One Guy Did Three Years Ago.

Wikis dispense with all that -- all of it. A wiki in the hands of a healthy community works. A wiki in the hands of an indifferent community fails. The software makes no attempt to add 'process' in order to keep people from doing stupid things. Instead, it provides more flexibility, a crazy amount of flexibility, and intoxicating amount of flexibility, allowing massive amounts of stupidity and intentional damage to be done, at will, by roving and anonymous posters. And it provides rollback.

Bad things happen on wikis. All the time. As historyflow shows (w00t!), pages frequently get deleted outright. But, as historyflow also shows, in a healthy community they also get restored, quickly.

Programmers have known for decades that a version control system covers a multitude of sins, and wikis embrace versioning as the cardinal virtue. With versioning, there's no need to try to prevent bad things from happening, so long as they can be quickly undone. "Detect badness? Get back to the last good version, then start out again from there."

I was recently reminded of this marvelous property when checking out wikitravel.org. I noticed a page for my city, and checking it out, I saw that it was an entirely fake review, making reference to places and events that never happened. It looked like an attempt at humor writing (though a fairly lame one), but of course the side effect (or perhaps intentional effect) was to undermine the goal of the site.

Seeing this, I simply deleted the current content and put an "Add content here" stub on the page, then went to Recent Changes to see if there had been other such grafitti entries. The same IP address came up in two other places, both also fake entries, and I deleted them as well.

Looking at the timestamps in recent changes, I saw that our budding satirist had spent an hour and three-quarters working on his trio of masterpieces. They were on the site less than two hours when I came along, and I undid everything he had done in two minutes.

And this, mirabile dictu, is why wikis can have so little protective armor and yet be so resistant to damage. It takes longer to set fire to the building than put it out, it takes longer to grafitti the wall than clean it, it takes longer to damage the page than restore it. If nearly two hours of work spent trying to subtly undermine a site can be erased in minutes, that's a lousy place to hang out, if your goal is to get people's goat. Better to go back to posting Microsoft trolls on slashdot.

The freedom from process is quite remarkable, and is also the hardest thing to explain about why wikis don't just fall apart with the first attack.


9:38 pm |

Instant Message Industry Insider

By Ross Mayfield

Many-to-Many gains another neighbor in the Corante hood who gets Social Software.

Stowe Boyd has launched a new Corante blog on the Instant Messaging industry.

He is launching with two scoops: on Oracle's late entry into IM and Tipic's early entry into IM Blogging.

You might recall that Stowe wrote a great piece on Social Software for Darwin Magazine. Welcome!


7:05 pm |

Monday, August 25, 2003

Wikitravel

By Ross Mayfield

Wikitravel.org is a new collaboratively developed world-wide-travel-guide.  A wonderful use of wiki technology, inspired by Wikipedia.

Check out Itineraries such as One Month in Southeast Asia which is a great example of the strength of a wiki structure.  Will be interesting to see if pages that start as copies of the CIA World Factbooks, such as the Gaza Strip, will evolve to include on-the-ground travel alerts.

Its content is available under a Creative Commons license, who blogged it.  Its just getting started, so give them (you) a hand.

[tip 'o the hat to Adina]


4:50 pm |

The Network is the Market

By Ross Mayfield

Tribe.net is a Craig's List meets Friendster style Social Networking Service that is just coming out of beta. What's different is the explicit transactional nature of the network, emphasis on tribal organization tools and how it relies on social capital to underpin transactions

In the interest of full disclosure, Tribe.net founder and CEO Mark Pincus is an investor in my company, Socialtext. That means I have had the benefit of talking with Mark about what he is building, and its an excuse for me to explore some Social Networking themes.

Within the Social Networking models paradigm, Tribe.net connections are declarative, making it an Explicit Social Network. Declarative like Ryze or Friendster where connections are browsable. Connections between people are confirmed ties, like Friendster, and you can only see the network that you are four-degrees of separation within. Ryze, it should be noted, took some steps recently to display both confirmed and unconfirmed ties.

But the purpose of Tribe.net is less connections, but the information flow they enable: tribal constructs and messages (classified ads). Information flow is to facilitate trade.

The Network is the Market

Social Networks (unlike Political or Creative Networks) are fundamentally transactional. We each have a group of people, no larger that 150, that we passively track and trade with. We have relationships of a kind with each of them and are aware of their relationships with each other. What we are monitoring is social capital. When someone wrongs another, if you have both people in your Social Network, you become aware of it and adjust your tacit social credit ratings.

Unless you are a little weird, you don't keep explicit ratings of your friends. You make a note to self, sometimes without full cognition that you are, move on and let it effect how you trust others in the future. Trust and reputation is fundamentally social credit.

Explicit ratings have a place, in larger networks/markets where the vast majority of transactions are with people you have at best a representational relationship with. Commerce with strangers is really scary, because they can dupe you without consequence. We all know how eBay scaled low-entry commerce through explicit ratings with social credit rather than financial credit. Allowing people to bank their actions, not just their assets. But there is value not just in scaled and scale-free networks.

This weekend we had an old-fashioned multi-family garage sale. First dibs were given to the other families to take what they wanted without having to pay. There were other garage sales on our block, kids ran from sale to sale having fun and a couple swaps occurred. Now I live in a strange neighborhood, mostly with overworked VC parents, but garage sales are the one time where people really hang out with their neighbors. The point is not only do we prefer to trade with people we know. Trade can be an excuse for conversation. The positive externality is each transaction can form the basis for more than transactions, but basic social capital that underpins relationships.

Richard Wilhelm recently made an argument for increased use of social capital instead of financial credit for Internet commerce:

...Mutual trust, when it exists, is a far better and more efficient alternative; it substantially lowers transaction costs, and it can offer a big competitive advantage. One World Bank study, using a regression analysis covering the 1980s, suggests that a 10 percent difference in the degree of generic trust among the citizens of a nation is reflected in a 0.8 percent variance in that country’s rate of economic growth. With average annual growth worldwide in the range of 1 to 3 percent during the same period, it is easy to see the payback in building trust...

We must bring to cyberspace “social capital,” the notion popularized by the political philosophers James Coleman and Francis Fukuyama. Social capital represents the matrix of behavioral norms and reciprocal expectations that allow any social network to function. These informal constraints provide the essential context within which societies can establish formal institutions, procedures, and rules of law. The core of social capital’s process is self-restraint, a willingness to forgo potential advantage...

[Source: Strategy+Business (reg. required)]

Value of the Small

Tribe.net is built upon the Value of the Small -- the smaller the network the stronger the ties and the more valuable the information flow. The weakness of the Value of the Small is if the network's design allows new participants and information flow.

Tribe.net allows you to post or search for classified ads within varying degrees of distance -- both relational (degrees of separation) and locational (geography). It therefore attempts to capture the Value of the Small by affording users the ability to constrain network size under search, but broaden it to weaker ties as needed.

Social credit is not made explicit through an eBay-style reputation system. Instead, it relies on conversations and connections for restraint. The challenge for all markets, however, is liquidity -- how the market scales. Unlike a Private Network like LinkedIn, where information flow is squarely constrained by risking social capital at every node in an information flow, Tribe.net relies upon explicitness. Relationships are declared, so if someone defaults on a transaction and there is enough communication to reveal the default, the defaulting party may risk their relationships.

Tribe.net encourages conversation by having almost no constraints in messaging modes. You can post listings (classified ads) for free, can send direct messages and post to message boards. If the cost to communicate is near zero, feedback loops proliferate to support social credit.

Benefits of Organization

If there is one thing people naturally reward others for in our complicated world, its organization. When someone brings people together its a valuable form of emergent leadership. When someone aggregates things together in a usable form, it becomes a resource. In both cases we reward leaders, designers and bricolageurs with social credit and sometimes even pay them.

Ryze's best feature, IMHO, was the creation of what they used to call tribes and now call networks. They allow people to construct networks for a multitude of purposes. I formed the Blog-Network to allow bloggers to find each other within Ryze, which has grown to 550 members.

Similarly, weblogs allow people to construct their own Conversational Networks. A blogroll constitutes explicit connections. Its easy to set up a blog with many authors and enable comments for open contribution. The emergent order between blogs form something comparable to a social networking service.

Tribe.net, as the name implies, has a dedicated focus on empowering people to create their own communities. Members are explicitly listed, conversation occurs through a message board, member listings are aggregated and events are scheduled. Tribe organizers are afforded convenient mechanisms for promoting their Tribe within the larger network. The cost of group forming is plummeting to zero.

Its too early to say if Tribe.net will succeed. My sense is it largely depends upon the cultures Tribe leaders foster, how the network scales and if norms of reciprocity beget social capital. Initially, social capital within the network is weak and the market for listings will be illiquid. Unlike LinkedIn, more valuable transactions will not occur from day one. Since it doesn't place constraints on connections or information flow it could grow rapidly.

It can be said that this Social Networking Service serves a special niche between eBay and Craig's List, between Newspaper Classifieds and Garage Sales, between Friendster and LinkedIn -- where the market is the conversation.


3:25 pm |

Social Networking Infographic

By Ross Mayfield

A terribly insightful Infographic on the trends in Social Networking services like Friendster.

11:11 am |

Thursday, August 21, 2003

Visualizing how wikis work

By Clay Shirky

Martin Wattenberg and Fernanda Viegas, in IBM's Collaborative User Experience lab have created a tool called historyflow that lets you see the history of a wiki page. The site has many of their observations on observed patterns, and more of mine are below.

The site has visualizations of several pages -- here is the wikipedia page on Abortion:

The X axis is time. The Y axis is amount and position of text. The colors are users, one per participant. The bright->dim progression represents text that survives multiple edits. The right-hand column is the text itself, whose color-coding matches that of the graphic.

And here's a less hot-button topic: Calculus

Several salient details jump out. First, many of the wiki patterns we've been describing over the years are now visible -- you can see the vandalism in the abortion diagram, for example, where the whole page gets deleted, and you can see how fast it gets healed. (There are views that represent elapsed time, and views that space edits out equally. The Abortion view above is elapsed time; Calculus is shown in "Even Edits" mode.)

You can also see that the healing is a full restore -- in times of vandalism, further editing of the page is put on hold as the goal is simply to roll back to the state just before the deletion.

- Next, edits come in waves, a pattern familiar to anyone on a semi-latent discussion list. Assertion begets reply, a brief flurry of activity occurs, then the action settles down (and presumably reappears on another page.) Like the shells of a nautilus, the space taken up by a living conversation on a wiki is a small subset of the whole.

- Next, most users are editors. Text, once added, tends to be shifted rather than deleted, and the shifting grows more granular with time, as people re-write, inserting their own edits into existing sentences and paragraphs. Deletions do occur, but they are relatively infrequent and fairly surgical (not counting vandalism, of course), and generally appear after the page has existed for a while.

- Text added at the bottom of an existing page seems to have a higher-than-average deletion rate. (I draw this conclusion from looking at a number of additional pages, not just the two shown here.) This seems to be a "BBS/graffiti" problem, where the added text is commentary on the page as a whole, rather than an embedded change.

- The respect for others' work means that the page tends to expand to express on multiple views rather than simply swinging from one pure point of view to another (exactly the result the wikipedia was designed to achieve.) It also means that pages tend to increase in length as additional views come into play, making it clear that at least some wiki gardening (deletion, compression) is required to keep every page from suffering Malthusian growth.

- It's also good to be early. This is a second-order effect of the users-as-editors pattern -- only the first user is writing in an uncomplicated environment. Whatever flaws or omissions there may be in the original text, at least some of that text generally survives through all subsequent edits. (The wikipedia entry on Capitalism is a rare exception, where almost all of the original text gets replaced during the many subsequent edits.) Like starting a mailing list thread, the benefit of starting a wiki entry that energizes other users is that there's a good chance that the topic will unfold in a context you help set.

Wattenberg and Viegas have also made a version of the software that can read history from MoinMoin entries. Here, for example, is the page from the Atom wiki called ProjectNameProposalII:

Note that here, unlike the wikipedia, there is much less editing and much more inserting, that the inserting tends to be blocky (signed paragraphs), that as a result the early text is hardly touched, and that the increase in page size is steadier than in wikipedia entries. This, in other words, is a picture of a wiki being used as a bad BBS rather than as a good wiki, exactly the pattern noted earlier in History, Personalities, Wikis Redux.

Wattenberg and Viegas are both infoviz superstars (Martin was the author of MarketMap, and Fernanda worked on Chat Circles in Judith Donath's Sociable Media Group), and historyflow is still in its early days. Among other things, they are considering creating a generic history format to be consumed by the application, so that visualizing any flavor of wiki (MoinMoin, kwiki, Usemod, etc) would just be a matter of transforming the history database into the standard. If this happens, it could unleash a lot of comparative work on wiki development patterns -- large vs small, in-company vs intra-company vs world, etc.


4:08 pm |

Julian Dibbell on in-game theiving

By Clay Shirky

Weird post by Julian Dibbell (a tautology, I know...), in the middle of his quest to make his living buying and selling virtual goods. This one concerns fencing goods in the virtual world:
So there I was, stuck between a dirty deal and a quick 5 million gp profit. I'd been stolen from in the game before, and I knew how much it hurt. Players can use hiding and thieving skills to slip into your house right under your nose and walk away with everything they can carry. It's not just impoverishing, it's humiliating, and I wasn't eager to be part of any such business. That Jammaster had lined me up as a fence a full day before committing the crime didn't help...

3:04 pm |

Dave Rickey reviews Designing Virtual Worlds

By Clay Shirky

Dave Rickey, who I linked to earlier regarding an interesting experimental design in virtual economics, has a slashdot review of Dr. Richard Bartle's recent book, called Designing Virtual Worlds. (Dr. Bartle was a co-creator of the first MUD, and author of the famous paper Players Who Suit MUDs, an early work of online sociology.)

Rickey gives the book high marks, saying

The general focus on the "players eye" view is a very important attribute: too often, discussions of virtual worlds have the "God's Eye" designer's view from orbit, and forget that in the end it's the ground-level "fun or not-fun" experience of the players that makes or breaks a design.
He also reiterated his original objection to the book, namely that it is more concerned with convincing academics that games are a form of art (though he does not note Bartle's skepticism about games as a substrate for community, a key feature of his original review on Engines of Creation.)

2:51 pm |

Wednesday, August 20, 2003

Email is Dead.

By Ross Mayfield

Wired has a good article on how Aggregators Attack Info Overload, a perfect excuse for me to go on a rant on how information overload will kill email lists within the short term.

I have posted at length on Blogging to Prevent Email Overload, Occupational Spam and how email should be a one-to-one medium.

Joi says email is officially broken because 17% of messages are rejected as spam. Never mind the false positives, the point is that your average F500 employee spends 3-4 hours per day using email, almost 50% of email is spam and 30% is occupational spam. Email volume is growing at 30% per year, invading our time and effectiveness. Email is no longer a productivity tool.

Kevin says the SoBig virus may be the last straw and We have to confront the reality: either email is broken, Microsoft's email software is broken, or those two statements are the same..

Commercial spam will not be solved by regulation or filters, trusted email networks that use challenge-response to confirm ties could be an interim solution at best. The only solution is changing economic incentives. Occupational spam cannot be solved by opt-in or opt-out techniques.

Email Lists have a rich history and many of value still remain. The first application that dramatically lowered the cost of group forming. But the reason email lists will be replaced by RSS and Pie/Atom/Echo is simple: cost.

Administering -- Administering email lists is a costy form of hell. Add/drop/policy/bugs/viruses all suck the time of administrators unless they buy very expensive tools. Chris Pirillo decided to move some of the more popular commercial lists to RSS becasue of the personal costs on his time.

Moderation -- Garbage in, garbage out -- unless there is a human check. Editorial functions don't scale efficiently and have coordination risks. Great moderators do create great lists, but only if the value they create is worth the cost of repetitive enforcement.

Reading -- Costs are twofold: time to read irrelevant messages and time dealing with subscripton management that has high search costs.

As email trends of volume and spam continue unabaited, these costs are compounded. Until recently there has not been an alternative. As consumers and enterprises adopt weblogs and aggregators a critical mass of users is already available in early adopter segments.

By contrast to Lists, Feeds are not opt-in or opt-out -- they are optional. In a decentralized structure, multiple Feeds in aggregate constitute an equivalent function to a List. Authors are Readers and as an Author has a choice to offer a Feed, as a Reader what to consume. We subscribe to people we trust not to waste our time. New Authors are revealed through social filtering (subscribed Authors referencing them). As Readers, they have the lowest transaction costs available for administering their consumption.

Decentralizing authorship, readership, administration and moderation pushes costs to the edge. This is made possible from a base of standards. Web nativity, an optional structure and social filtering process keeps spam out of trusted personal networks.


2:14 pm |

Paul Resnick on "Impersonal Social Capital"

By Sébastien Paquet

Paul Resnick (whose RSS feed is here, in case you've been looking) has been trying to pin down the mix of ingredients that enables coordination among strangers.

If neither friendship nor instutitional membership is a pre-requisite for coordinated activity among strangers, what are the pre-requisites? I've coined the phrase "impersonal social capital" to refer to whatever those enablers are. It's some combination of networks of acquaintances, generalized trust, assurance through reputation or other accountability mechanisms, and a big dose of technology to reduce communication and coordination costs. I took a first stab at trying to sort this stuff out in a paper titled, "Beyond Bowling Together: SocioTechnical Capital" a few years ago. It may be time to take a closer look again, now that there are more examples to draw from.


2:09 pm |

Ideant: Slipping between the one and the many

By Clay Shirky

There's a post over at Ideant on email in its many one2one2many2many guises, picking up on earlier posts from me and Ross, and making the point that the mode changes are a key part of the puzzle:
As a feature, it is very convenient because it allows email to metamorphose from one kind of technology (one-to-one) to another (many-to-many) [...] My point is that dialogue is a dynamic and unfinished business, and to the extent that technologies can support these one/many-to-many/one/self shifts, the more useful they are.

9:30 am |

Tuesday, August 19, 2003

Ask /.: Group communications tools go DIY

By Clay Shirky

I know the plural of anecdote isn't data, but 3 recent "Ask Slashdot" posts caught my eye: Implementing Intercom-like videoconferencing? -- "I'd like to improve day-to-day interactions between everyone here and I'm thinking of setting up (for lack of a better term) a video-intercom."   Network Chat as a Tool for Corporate Communications? -- "Does your company utilize [Instant Messenger or IRC] as a communication tool (to communicate with customers, between employees and Pointy Haired Bosses?"   and Workgroup Messaging -- "...allow a user to create pop-up messages for display on the remote screens."

The replies to these posts are fairly dull unless you are looking for similar software (although I was interested to see how many of the respondants were running Jabber.) The interesting thing to me is that so many Ask Slashdot questions are about bottom-up installation of workgroup software in such a short time.


10:01 pm |

Six Degrees as YASNS: It all comes full circle

By Clay Shirky

SixDegrees.com has entered the ontologically suspect category of "former ex-website". It went 404 in 2001, but now seems to be back as "WorldShine," albeit still under the sixdegrees URL. Having stolen a page from Friendster's book (a chapter, really), it positioning itself as a soft-sell dating site, with a frisson of remote hook-ups: "Imagine a dating site merged with a travel guide, all within six degrees of seperation..."

As seems to be required in the YASNS world, the About Us page also embraces the fiction that a six degrees network is both personal and exclusive -- "ALL Worldshine members have been invited by existing Worldshine members which keeps the community within six degrees of separation at all times and greatly increases the quality and frequency of the information exchanged" even though the very meaning of the phrase six degrees of separation implies a literally global inclusiveness.

I assume the logic of a re-launch is that there is some value in the old sixdegrees site, but given the re-branding, and the fact that Friendster has far more users than the old sixdegrees did, I predict a wipeout here.


8:42 pm |

SMSing around Hollywood's marketing machine

By Clay Shirky

Interesting article in the LA Times (via BoingBoing) on SMS use by audience members affecting a movie's box office revnues almost instantly:
In the U.S. these days, the pace of chat is fast enough, in some cases, to affect a movie's box office results from its Friday opening to Saturday night. "Charlie's Angels: Full Throttle" signaled it was in trouble when it dropped 11% overnight. Conversely, a hit like "Pirates of the Caribbean: The Curse of the Black Pearl" can show its mettle instantly; the Disney film, which opened on a Wednesday, actually went up 17.3% from Friday to Saturday, according to Nielsen EDI.
I wrote about the economics of this effect in 1999, in The Internet and Hit-Driven Industries:
This threat -- that after many months and many millions of dollars the fate of a movie can be controlled by actual fan's actual reactions -- is Hollywood's worst nightmare. There are two scenarios that can unfold in the wake of this increased community power: The first scenario (call it "Status Quo Plus") is that the studios can do more of everything they're already doing: more secrecy about the product, more pre-release hype, more marketing tie-ins, more theaters showing the movie on opening weekend. This has the effect of maximising revenues before people talk to their friends and neighbors about the film. This is Hollywood's current strategy, having hit its current high-water mark with the marketing juggernaut of The Phantom Menace. The advantage of this strategy is that it plays to the strengths of the existing Hollywood marketing machine. The disadvantage of this strategy is that it won't work...

6:57 am |

New Definition of Social Software

By Clay Shirky

New definition of social software:
Any arbitrary collection of algorithms, protocols and metadata that allows friendless agoraphobics to pretend otherwise.

“I’m having trouble deciding which node in my social software network I’m going to ask to the e-prom."

From The New Devil's Dictionary. See also, FlashMobs:
An impromptu gathering, organized by means of electronic communication, of the unemployed.

6:38 am |

Monday, August 18, 2003

Fakester.net: Someone doesn't get it

By Clay Shirky

Some fakesters are putting up a refuge site for entities from the Great Fakster Purge. That in itself is not so interesting, as it's a predictable reaction. What's interesting is that they are being just about as clueless as they could possibly be.

The message making the rounds reads:

We're currently putting together the sites, which, based on the widespread use and recognizability of the word "Fakester" and the other content on the sites, we confidently feel will be the one-stop site for deleted Fakester profiles.

It won't be just that, though, no no. There will be quite a lot of content- humor/entertainment, and socially relevant commentary and news, both of which will appeal to a lot people, the Friendster hip/ young/ college-age/ indie-rock/ counterculture demographic in particular. [...]

There won't be the whole full-fledged profile and inter-connected thing that Friendster has, simply because we just came up with this - the back-end software for Friendster took months and hundreds of thousands of dollars to build. There will be a chance to post pictures though, and we're looking into what level of profile-ish- ness we can offer.

Well there's a relief. If you're like me, when you heard that Friendster was deleting the fakesters, your first thought was "Oh no! A valuable source of humor/entertainment content, of the sort that appeals to the hip/ young/ college-age/ indie-rock/ counterculture demographic gone! Oh, the humanity!" Now, thanks to Fakester.org, your worries are over.

This is the mysterious thing to me about these "Site changes deal with users/social antibodies boil over" events: many of the reactions to the change exhibit not even a modest understanding of what made the service worth worrying about in the first place.

How could anyone see the fun of inserting Jesus and San Francisco into a living social context, and then conclude that a site with only fake bios, in an inert and non-social setting ("There won't be the whole full-fledged profile and inter-connected thing that Friendster has...") will preserve any of the value of the fakester movement?

(And, answering my own question, I suppose the anser is "Denial is a powerful tool." Believing that something as stupid as Fakester.org is a good idea seems preferable to simply accepting that Friendster management has all the cards here, and can destroy any aspect of the social fabric they like with impunity.)


11:48 am |

Ito on social software and getting a job

By Clay Shirky

Joi Ito has a blog entry up on the value of social software in getting a job:
I recently hired two people who were IRC regulars. I felt very comfortable after "getting to know them" over the last few months on IRC. Of course face to face meetings and interviews were essential, but the time spent with them on IRC really added to my ability to judge their character. I realize now that I am actively recruiting from my network of weak ties [...] The Net has always been a big part of my arsenal of networking tools, but I think it's reaching a whole new level.

11:15 am |

Sunday, August 17, 2003

Should be One-to-One, Not Is

By Ross Mayfield

I must cede the point to Clay that email has always been a many-to-many medium. In general, it was and unfortunately still is.

I could make a point that email originally used FTP for transport, which is one-to-one, but this was soon replaced by SMTP and then Dave Crocker and others came along and made it what it is.

Unfortunately, the unintendend consequence of CC is occupational spam, unavoidable by opt-in and opt-out. If only we could disable this field and shift mass distribution to publishing mediums where subscription is optional.

So email is a many-to-many medium, but that doesn't mean it should be. Should be as in Policy, not Value.


1:32 pm |

Negotiation and Social Software

By Ross Mayfield

Negotiations are arguments. Agrumentation, not a derogatory term, is a practice of achieving a common sense through parties taking contrary positions. Debate is not only helpful in discovering compacts, but the essence of constructive social interaction.

There are three kinds of arguments: Fact, Value or Policy. You can argue over what is, what should be or how it should be.

In general, determining the winner in an argument of Fact or Policy can be relatively easy with pre-defined criteria. Cases of Value often embroil in emotion and winners are difficult to determine.

Social software can support negotiation, at the least, by revealing what kind of argument is in play. Every argument is different, but bringing parties to the same table, making positions clear, revealing differences and overlaps in preferences provides a basis for debate. Tools that allow mediators the flexibility to structure dialogue while deemphasizing personalities can accelerate constructive conversation. Tools that deemphasize personality and make positions incrementally explicit reveal sidetracking Value-based arguements, allow Fact to be resolved with fact and support collaborative development of Policy.

Michael Helfrich relates a case of using a shared space in support of negotiation:

The Virtual Negotiation Table in Southern Asia/New York/Helsinki: Groove was used less than eight weeks ago to broker peace in a nation in southern Asia. During the mid-80's, tension between the majority and the separatists on this island nation erupted into full blown ethnic war, with 10's of thousands of people losing their lives. Leveraged by some very bright folks from the Nobel Peace Laureate, and with the wisdom and guidance of James A. at Groove, a set of "Peace Tools" was developed and deployed to assist in a new round of peace negotiations.

Peace negotiation is intricate business. Bringing warring factions to a physical table is often tense, and can result in people getting shot. It was envisaged that Groove could bring this nation's leadership to the same "virtual" table as the separatists. It worked. Groove was embraced by both constituencies because of the virtual nature of the shared space. While one shared space served as the meeting place for the factions, each had separate spaces to discuss their positions and provide context for the negotiators. The shared space became the trusted and neutral enclave for all parties involved to lay out their positions, and to jointly work through the options. And no one gets shot.

This is a fantatstic case of social software invoking constructive agreement.

At Socialtext we have found shared wiki spaces for negotiation work well to foster trust because they reveal positions without personality and do not constrain . In a typical negotiation, say with a client and a vendor in a solution sale, no party is in "control," so giving up a little can drive resolution. Peter Morville comments that with wikis: "mutual openness and shared vulnerability led to a strong sense of shared trust."

Negotiation is a fact of life not only for organizations, but people within organizations. Meetings are primarily status contests, office politics perhaps the greatest drain on productivity and tension is often the norm.

Weblogs and Wikis inside organizations are public spaces where participants can undertake activities and communication with full knowledge they are revealing their preferences. The resulting text is an informal common sense. When the activity is developing a social compact itself, by contrast to the chest-thumping patterns of a typical meeting, the focus is constructing something together. Incremental success in construction, taking on increasingly politicized issues, creates shared achievements that turns parties into a group. And groups can achieve compacts.


1:17 pm |

Great Danah Boyd posts on the Fakester genocide

By Clay Shirky

Danah Boyd, of the connected selves blog, has a fantastic set of posts, both her own musings and documentary from elsewhere, about Friendster's decision to kill off the fakesters, the non-human or unreal characters that were all over the board (e.g. Jesus, San Francisco, War).

The pieces start with Goodbye Fakester:

See, the thing is that Friendster does not get away from the coarse descriptors that make interactions peculiar and sexualized (see Sexing the Internet). When it comes to meeting people outside of the Familiar Stranger/Friend of Friend degree, the profiles collapse a person's identity into a set of forms without much meaning. Most frequently, they give little to allow the start of a conversation. [I mean.. what am i going to say when someone writes: "Koyaanisqatsi: good movie!"]
and continue back through posts documenting the user's reactions: Start a Revolution ("...establishment of user rights for “fakesters” – parody or homage profiles of celebrities, religious figures, philosophers, abstract concepts, clothes, bands, fictional or mythic characters, martyrs, pop culture icons, household items, and the like..."), the Friendster Manifesto ("I. Identity is Provisional / Who we are is whom we choose to be at any given moment, depending on personality, whim, temperament, or subjective need..."), and Open Letter from a Friendster User ("The corporate masters at Friendster should be thrilled that they have such a vibrant online community as they now have on their hands. What they forget is that a living community, by definition, has a life of its own..."), and a pointer to the Yahoo Group FriendsterRevolution, which has 250+ members and is getting something like 30 messages a day.

There's lots of bits of commentary out there on the Friendster/Fakester activity, but Boyd seems to have made the story her own, and connected selves will obviously be the place to go as it continues...


7:44 am |

Saturday, August 16, 2003

Non-human nodes on Friendster

By Clay Shirky

The SF Weekly has a marvelous article on non-human, long-deceased, or falsely identified Friendsters:
Batty and numerous other Friendsters routinely violate the site's user agreement by creating fictional characters as profiles instead of, or in addition to, their "real" profiles. These "fakesters" portray themselves as everything from inanimate objects like the World Trade Center to celebrities like Paris Hilton to historical forces like War (which lists its profession as "resolving disputes").
This is related to the earlier piece about Ben Discoe's Friendster visualization efforts, which showed that many of the non-human nodes are the most highly connected. It also reminds me of the LiveJournal pattern, where the most highly linked-to 'friends' are often performance pieces, as with Billiam, the ur-wigger, Miss_Cleo, the phone psychic (since deleted), or Kim Jong Il's illmatic LJ page.

As my former student Alicia Cervini pointed out when studying Friendster last year, Friendster has never been able to enforce its policies of "real people, real relationships" -- people have been faking bios and friending one another to drive up network density with no regard for real life. The faskesters phenomenon is part of that larger trends, and part of what makes Friendster entertaining.


1:25 pm |

Context of a Post

By Ross Mayfield

Its interesting that every time there is a sudden news event, like today's blackout, that bloggers who take it upon themseves to cover it do so by continually revising their first post. Not by making new post after post. Very wiki-like, something weblogs are not supposed to do according to some definitions.

Reason being, covering an emerging event before significant context is available shouldn't be done with permanent chunks of text. Those chunks could be referred to out of context.

During the initial SARS crisis, a complex event with uncertain context, the Wikipedia SARS page became the best source of information for many. Living pages perform well when sources are evenly and disparately distributed -- and information is turbulent.


1:19 pm |

Friday, August 15, 2003

Bizarrely loving portraits of archetypal flamers

By Clay Shirky

Ah, Fridays in August: Mike Reed, an illustrator, has posted a guide to different kinds of "Flame Warriors", such as ALLCAPS, Grammarian, and Rebel Without a Clue:
ALLCAPS attempts to compensate for his limited rhetorical weaponry through the extravagant use of capitalized words - something netizens refer to as SHOUTING. Sure, a sprinkling of capitalized words can add spice to an attack, but overuse is like too much tarragon in the stew. Even worse from a tactical point of view, too much shouting alerts other Warriors to the opponent's verbal WEAKNESS and emotional EXCITABILITY.

This would fall beneath the threshold of interest, except for the fact that Mr. Reed has produced descriptions and illustrations for 81 different types of flamers. If he spent even 5 minutes on each text entry and another 5 on each drawing, that's several days of work spent on the project. Each illustration has a CafePress link below it, but I can't imagine the revenues will make up for even a couple of hours of the work, putting this in squarely in the labor of love camp, and Mr. Reed on the threshold bonafide net.kook.


5:22 pm |

Context as a Post

By Ross Mayfield

Its interesting that every time there is a sudden news event, like today's blackout, that bloggers who take it upon themseves to cover it do so by continually revising their first post. Not by making new post after post. Very wiki-like, something weblogs are not supposed to do according to some definitions.

Reason being, covering an emerging event before significant context is available shouldn't be done with permanent chunks of text. Those chunks could be referred to out of context.

During the initial SARS crisis, a complex event with uncertain context, the Wikipedia SARS page became the best source of information for many. Living pages perform well when sources are evenly and disparately distributed -- and information is turbulent.


3:38 pm |

Thursday, August 14, 2003

Subjectivity in Triad City: One for Jessica

By Clay Shirky

This is really Jessica Hammer's turf -- the mixture of game and narrative -- but there's an interesting piece about Triad City, a massive literary MUD, detailing the differences between TriadCity and other MMOs. A lot of it is marketing hype -- no fair claiming "really really big" as a point of differentiation in the age of Star Wars Galaxies -- but the listed difference that caught my eye was Subjectivity:
So far as we know, Triad is unique in its ability to present highly individualized views of the world to specific players or classes of players. One example of what this means is that, if you and I walk together into the same room, I might see a set of room and item descriptions which differ subtly or radically from the ones you see. Subjectivity of this sort is largely under the control of the world author, enabling fictional techniques such as voice and point of view.
I've mentioned the possibility of subjective drift before, in the "social forking" design of Guild Wars, but it's interesting to see it offered not as a thing to be fought against but as a design tool.

I am skeptical that TriadCity will get the 30,000 players they think they need -- MUDs tend to have very high churn -- but it will be worth seeing how the early experiments play out. (And I hope Jessica is keeping an eye on it, as she'll have more to say about it than I do.)


3:34 pm |

Email has always been a many-to-many medium

By Clay Shirky

This blog seems to have made it my destiny to quibble with Ross in public; this time it's about email as a many-to-many medium.

In his post about Rafe Needleman's smart thoughts on broadcast IM, Ross says

Email is dying because its a one-to-one medium being stretched for one-to-many and many-to-many uses
This seems wrong to me. Once email was ported from the TENEX system to the ARPAnet, organized mailing lists appeared almost immediately. Furthermore, some of these lists, like SF-LOVERS, remained viable for decades. SF-LOVERS preceeded such new fangled inventions as the domain name system, usenet, and even the formal changeover to IP addresses, and managed to outlast all the actual hardware of the version of ARPANet it launched on. Like the Great Red Spot of Jupiter, SF-LOVERS persisted as a collection of forces, rather than as a thing.

Far from being a one-to-one mediuum that is being stretched, email has been many-to-many from its inception -- by allowing multiple To addresses, email instantly shed the atomic constraints of "one object, one address." And, like the horseless carriage and compact disc before it, its name, designed to make it easy to understand, actually hid the degree of its radicalness. "Ad-hoc asynchronous group communications protocol" would have been a much more descriptive name, though one, of course, that would have doomed it. Much easier to think "It's like mail, but electronic", without having to think too much about the ways in which it wasn't like physical mail at all.

I hate spam as much as the next person, but I was around in Ye Olden Tymes, pre-spam, and mailing lists were a critical part of the social infrastructure -- they were in fact the first piece of social software on ARPAnet. Spam is orthagonal to the distribution pattern of mail; spam is a negative externality of cheap messaging and wide distribution of addresses. Spam may kill email, but that doesn't mean it was not or is not a many-to-many medium; it simply means that the original designers didn't provide enough barriers to this particular form of abuse.


3:15 pm |

Meaning of a Tag

By Clay Shirky

In a piece related to Ross quoting Matt Webb quoting Simon Roberts quoting the State of the Language(whew), Tim Bray has a related piece on the social construct of markup semantics:
Where does the meaning come from? I’m only aware of two ways for markup to take on semantic weight:

1. The designer of the markup asserts that some tag or attribute is used to identify content with a particular semantic.

2. The broad community of authors and programmers exhibits consensus as to the semantic of an item of markup.

This is not just a theoretical formulation: the venerable UL tag has been around for a long time in a succession of dialects including HTML. The name is an abbreviation for “unordered list,” and at one point the idea was that you could treat the contents like a relational table and sort them any way that was convenient. Due to a cascade of implementations (I first saw it in Mosaic), UL eventually grew the semantic of “ordered list with bullets.”

Worth reading as a companion piece to the Atom Wiki.

2:47 pm |

Wednesday, August 13, 2003

Meaning of a Page

By Ross Mayfield

Matt Webb forwards this gem on language:
Via Simon Roberts in email:

"The meaning of a word is not a matter of fact (which is why an argument about it can't be settled by recourse to the dictionary), and it is not a matter of opinion (which is why an argument about it mustn't be unsettled by a refusal to have recourse to the dictionary). The meaning of a word is a human agreement, created within society but incapable of having meaning except to and through individuals. We may find evidence for such agreements, but we can't find proof of them. A language is a body of agreements. Some lapse; others change; new ways form". (From The State of Language (Ricks).)

Wikis foster the development of a shared vocabulary through such agreements. Each page has a title, initially created by one contributor. The contributor shares an initial definition, but puts it into a social space for others to redefine.

Once a term is introduced as a title, it rarely goes away. Someone may clone the page and create a new title, but links to the original page exist, so a forwarding link at least remains in reference. The context of links help define a term (and with the English language, there are three degrees of seperation between words).

The body of a wiki page is where agreement is reached on its meaning. Since its editable space, without social agreement there is a natural tension. Create a page called Abortion in a space shared with strangers (e.g. Wikipedia) and disagreement is quickly revealed. Text is deleted, struck through, wiped over, pushed down or imposed. The cycle can repeat itself in these worst case scenarios. But eventually an agreement is struck. Tenuous peace by linking to opposing new pages, Pro-choice & Pro-life for example. Perhaps what remains is a simple chunk of text, the lowest common denominator of what can be agreed. As Clay points out, other tools would amplify personality to the point where discord would dominate. Wikis allow opposites to exist in a community without heavy-handed moderation.

Sometimes a term is introduced that is outside the bounds of acceptability to the community. Attempts to wipe it clean and banish it from existence usually result in re-occurence. But replacing its definition with a clear rationale for why it was censored fosters understanding and learning within the community.

Wikipedia and the relatively new Wiktionary are the clearest example of people reaching agreement over language in this form.

The Atom Wiki is an interesting case where the ability to create new terms fostered an explosion of language, but agreement has yet to be struck. The naming of a product or standard is a difficult process even when all the participants are branding or IP experts. Perhaps Phase II will lead to agreements because consensus failure can be a driver for conciliation. The practice (not process!) continues.

Reaching agreement on a smaller scale within organizations as they define their shared vocabulary. There the objective is not defintion, but it is a positive externality of the communication and activities they undertake. A body of implicit and explicit agreements emerges that enhance trust and social capital.


7:15 pm |

Friendster Graphing software

By Clay Shirky

Ben Discoe at Washed Ashore wrote a piece on his Friendster graphing software

Says Discoe:
It turns out that many of the most well-connected Friendster hubs are not people.  There are Friendster IDs created for a whole host of non-human entities, including cities (San Francisco has only 84 friends?), drugs (Ecstasy has a lot of friends.... Ketamine fewer) and body parts (Gay Penis is wildly popular).
Most interesting to me is the insane bushiness of the networks drawn by walking the graph. Like many social networking services, Friendster engages in the fiction that someone 4 or 5 degrees from you is part of your network ("Hey, I know Barry Bond's accountant's brother's co-worker's daughter. I bet we could get tickets to the game!")

This strategy seems designed to make sparsely populated networks look dense in the early days of a social networking service, but it has little to do with reality.


2:45 pm |

Dash trashes Flash Mobs (+Rheingold roundup)

By Clay Shirky

Anil Dash has a funny little piece up called Flashmobs are 99% Bad ( a gloss on Jakob Nielsen's "Flash is 99% bad" title):
Those who are worried about the weblog fad flaming out, fear not. Mobs are truly the flagpole-sitting of the new millennium. As Joshua astutely observed, Flash Mobs are striking in that they are an affinity event for people who have no affinity group. A Meetup for people who like Meetups. How much more meta can it get? None. None more meta.
I've been telling press people who call that flash mobs are a cross between streaking and being in a marching band with very short routines, but I may have to switch to likening them to flagpole sitting.

And, just in case you hunger for more news about flashmobs, Howard Rheingold has been posting frequent news roundups. (In an interesting bit of meme-jacking, flashmobs started out life as flash crowds, but Howard's smartmobs meme was so much in the air that the name has shifted in just a few weeks.)


11:49 am |

Interesting MMO experiment in pop density

By Clay Shirky

Over at Skotos, Dave Rickey has a piece taking on Richard Bartle's skepticism of game design shaping game community, in which he outlines an interesting real-world experiment in technological shaping of population.

The experiment was about population density on the "Zek" servers on EverQuest, which allow essentially anarchic palyer-vs-player killing (PvP). Like many, he noticed that these servers were sparsely populated, but he didn't believe the classic explanation: the servers currently reach all players who want PvP worlds. Instead, he advanced a counter-hypothesis, namely that the number of players was a function of density. Since the number of PvP encounters will roughly rise with the square of users (Aside to the reader: You're obsessed with scale too, right? It's not just me is it?), Rickey concludes that "...the primary instrument of population pressure was the 'gankage' factor, how often the players were likely to lose battles."

And lo! and behold! when new Zek servers were deployed, the old equilibrium was upset, and the population did grow, until the gankage factor did causeth the populace to settle into a new, higher equilibrium. Score one for the predictive value of social science.

This in turn sugggests that deploying additional space, rather than additional expansion packs, may be the way to stimulate growth in otherwise stable MMOs.


11:46 am |

Economics of Online Games Roundup

By Clay Shirky

Bunches of stuff coming out on the economics of online games:
  • Wired has a piece on gaming companies using the crack dealer strategy (the first hit is free), to get people into their games, a strategy that looks eminently sensible in light of Sir Bruce's "your stable population is governed by your early growth" subscription curves.
  • Programmer's Heaven has an interesting article on using the Castronova thesis (in-game value is inseperable from real value) as a game design input:
    By creating and encouraging channels to allow money to flow from the real world back and forth into the virtual world, the game host can open up a whole new source of revenue. Virtual objects in the game world cost nothing for the host to create, but may be of sufficient economic value to players to warrant spending money to obtain them. [...]

    Virtual objects in the game world can also be tied on a one-to-one basis to valuable objects in the real world in order to raise the stakes and therefore the interest level of game players. For example, there might be a dragon in the game world guarding a treasure horde. One of the treasure items might be a virtual jewel that is actually a digital title to a real precious gem in a vault in the real world. The player who defeats the dragon and obtains the virtual gem may “cash it out" for the real gem.

  • Finally, RPG Vault has an interview with Mark Kern of World of Warcraft, where he talks about "instancing", a form of social forking where some areas of the game are open to all and some are private to each player. (Guild Wards does this too, and the problem was first noted in the early 90's in Habitat.) Instancing reduces scarcity of things like puzzle-based gameplay, as well as helping keep the economy in check by limiting player's ability to generate in-game value to sell to other players on eBay:
    When you play a game with thousands of other players, there are times when you want to be social and feel like part of a populated world, and there are times where you want to feel heroic and special. Instancing gives players their own private dungeon, or part of the world, so that they can have epic encounters and special events that aren't going to be spoiled by crowds or long lines of people "waiting to kill the dragon." Instancing is going to be a very key part of allowing us to provide those types of special encounters that wouldn't be possible otherwise. It will also cut down on camping and help control item farming.

  • 11:24 am |

    Distributed Trust Metrics

    By Clay Shirky

    There's an interesting "Ask Slashdot" entry on someone who needs to provide filtering on a political website. However, the site operates at too small a scale to use slashdot-style user ratings. (Reading the web for me is like a permanent episode of Pee-Wee's Playhouse, where "scale" is always the magic word.)

    The response wasn't enormous (the question wasn't on the homepage) and mostly had the usual ideas (Bayesian filtering, Epinions Web of Trust), but included this little gem of social engineering:

    2) enable the ability to 'bozo-bin' someone: their account can be made so that they can still post, and they can see their posts, but noone else can. Most bozos won't even know they've been binned, and thus will not try to create a new account to get around it. Think of it as a honey pot for trolls.

    Perhaps most important, though, is that our love of the uncensored view has finally subsided enough that someone who wants to run a political website can ask about how to filter content on slashdot ("Home of the Techno_Libertarian Fringe!®") and not get flamed. This presages a huge amount of experimentation with conversational filters over the next couple of years.


    11:23 am |

    Too Much Mediation Damages Relationships

    By Clay Shirky

    The Seattle Times has a story on Tim Sanders, Yahoo's Chief Solutions Officer (ah, the 90s, that heady time...), and his desire to test what he calls New Economy Depression Syndrome, which he defines as "a state of work-related stress brought on by information overload, constant interruption by technology (think e-mail, instant messaging and cell phones) and the increasing personal isolation that technology affords us."

    He reports that after speaking on this issue in public, someone mailed him about a work environment where severl groups of engineers had gone to completely electronic communication:

    "The turnover rate was astronomical," Sanders said of those groups. "They were the worst people to work with. They took nerd and created monster or troll. These were very lonely, depressed, negative, anti-social, brilliant people."
    Interestingly (and related to the earlier note about the diagnosis for web addiction), Sanders is not content with asserting that this syndrome exists; he's actually arranging to have it tested:
    Because most of his examples are anecdotal, Sanders has teamed with Heart Math, a research, training and consulting company, to scientifically test his theory.

    Heart Math, which advises Fortune 500 companies on reducing stress and enhancing employee performance, administered a digital lifestyle survey, the results of which are expected to be released Labor Day weekend.

    It seems like a "Duh" conclusion, but many of us (me very much included) didn't believe it for a long time: in many situations, mediated communication is worse than face to face communication. The principal value of social software lies not in the fact that it is better than real world contact, but that it is alot better than nothing, when physical presence are impossible.

    11:04 am |

    Monday, August 11, 2003

    Keep it One-to-One

    By Ross Mayfield

    Rafe Needleman writes about how MessageCast is trying to turn IM into a broadcast medium:
    ...But instant messaging is largely a person-to-person service. Corporations will support it because they recognize its value for increasing worker productivity and for sending out real-time messages, but these same businesses are unlikely to get truly enthusiastic about IM until they can use it for the type of communication that they see as strategic: broadcast.

    The issue with IM as a broadcast medium, of course, is that nobody wants IM spam. Fortunately, instant messaging is less likely to become a spam channel since, unlike e-mail, the various popular IM systems are closed; you can't blast a million instant messages as easily or cheaply as you can a million e-mails.

    So there's a conflict: Businesses want channels that can be used for sending time-critical or mass messages, but the IM channel is set up for person-to-person communication, not business-to-the-masses blasts...

    Rafe is right to point out the danger of this direction. Email is dying because its a one-to-one medium being stretched for one-to-many and many-to-many uses. Unlike Email, IM doesn't have channels to exploit for broadcast. Creating this channel has commercial purpose, but beyond commercial spam I would be seriously concerned about the productivity impact of a new form of occupational spam compounded by IM's interruption tax.

    12:28 pm |

    Richard Bartle on Real Voice in Virtual Worlds

    By Clay Shirky

    GameGirlAdvance has an interview with Richard Bartle, one of the designers of the first MUD, on the addition of a voice channel to X-Box games, which he regards as a sign of The End Times®:
    If you introduce reality into a virtual world, it's no longer a virtual world: it's just an adjunct to the real world. It ceases to be a place, and reverts to being a medium. Immersion is enhanced by closeness to reality, but thwarted by isomorphism with it: the act of will required to suspend disbelief is what sustains a player's drive to be, but it disappears when there is no disbelief required.

    Adding reality to a virtual world robs it of what makes it compelling - it takes away that which is different between virtual worlds and the real world: the fact that they are not the real world.

    Voice is reality.

    Though he misunderstands the degree to which creating a virtual world is a goal for all game designers (Counterstrike, for example, is not a virtual world in his sense, and might well be enhanced by voice), he is right that a voice channel on the Everquest/Star Wars Galaxies class of games will be problematic (not least because men playing women and children playing adults will both be obvious to listeners, as they are not now.)

    He also predicts that voice-enabled games will attract newbies more quickly, and lead even moderately experienced players to leave more quickly. If he's right, this will show up several months from now on Sir Bruce's game population graphs as a round-trip curve instead of a roughly logarithmic one.


    10:31 am |

    Economists vs Criminals: Bet on the Criminals

    By Clay Shirky

    I've often posted about Prof. Edward Castronova's work on the real value of virtual goods, but a lot of the work has had a necessarily predictive or "What if" characteristic to it -- by extrapolating from eBay auctions, Castronova is making an argument about future trends, an argument that often gets short shrift from mainstream economists.

    Now, there's proof from an unexpected quarter that his predicted osmosis between virtual and real economies is real and will continue accelerating -- Korean cyber-criminals. The Korea Times is reporting that the number of online crimes is growing, and that more than half of the crimes are committed against property on gaming sites:

    'This is because these online games are not treated just as online games, and their money not treated as imaginary money,' Chang said. 'A lot of cyber crimes involve cyber money, but people are buying and selling this cyber money with real cash.'
    Remember Napster. Neither the RIAA nor the audiophiles believed that it could work, but the growth rates didn't lie, and file trading indicated a massive arbitrage of the cold dead hand of the music cartel. Now, even though mainstream economists are dismissive of the idea of the real value of virtual goods, the criminals are voting with an investment of time and effort. Not only is virtual property worth spending time and energy to acquire, its worth spending time and energy and the threat of jail time.

    GrepLaw also has a recent interview with Castronova:

    OK, we’ve recognized that there’s an economy. That means there are political classes, with resources and both communal and conflicting interests. What governments emerge? At the moment, the governments are basically like families. Think of Sicily [...] Similar thing is happening in synthetic worlds: the companies reserve the monopoly of force to themselves, but they don’t deploy it in any significant way. So, close-knit relationships are the only form of social order. Libertarians take note: Absence of formal government is not paradise, it sucks. Big time.

    10:17 am |

    Saturday, August 9, 2003

    "Small World" Experiment Results

    By Clay Shirky

    Scientific American has a short piece on the results of Duncan Watt's small worlds email experiment and attempt to re-create Stanley Milgram's famous "Six Degrees of Separation" experiment online, by getting participants to forward mail to a specifc target, who they don't know.

    The authors of the study (still running at smallworld.columbia.edu) found that the average chain was indeed around 6 people, and that the usual search tool was "forwarding by characteristic" -- says Watts "If you’re trying to send something to Siberia, you don’t think, ‘Who do I know with a lot of friends?’" Watts remarks. "You think, ‘Who do I know who is Russian?’

    The most important finding, though, was that while chains can be connected in a few hops, few are. Of over 60,000 volunteers, only 384 chains, around 3%, were actually connected. This poses a challenge for the Friendser/LinkedIn/YASNS services -- there is a huge difference between latent and active social fabric.


    9:51 pm |

    Anti-social software: Defining web addiction

    By Clay Shirky

    CNN has a short piece on a proposed diagnostic for net addiction, going by the acronym MOUSE:
  • More than intended time spent online
  • Other responsibilities neglected
  • Unsuccessful attempts to cut down
  • Significant relationship discord because of use
  • Excessive thoughts or anxiety when not online.
  • This is interesting to me not just as a former net.addict (two years of mainlining usenet, to a degree that would be hard to overstate), but also because it indicates a more nuanced engagement with the hot-button issue of addiction.

    For years, the Party Line has been to downplay internet addiction, especially when it has been of a social character, as with the famous Amy Bruckman quote: "It is tempting but dangerous to impose value judgments on MUD players who are happy with how they are spending their time. Certainly, Foo is courting danger because he is neglecting his responsibilities at work. However, DePlane, despite MUDding 80 hours a week, still gets above average grades and holds down a part-time job to make his spending money."

    I first read that 10 years ago, and it still gives me the creeps today -- the psychological phrase for that is "high functioning", as in "high functioning alcoholic."

    As Pavel Curtis of PARC said, years ago:

    I am concerned about the degree to which people find virtual communities enchanting. We have people who use LambdaMOO who are not in control of their usage who are, I believe, seriously and clinically addicted. . . . These people aren't addicted to playing video games. It wouldn't do the same thing for them. They're communication addicted. They're addicted to being able to go out and find people twenty-four hours a day and have interesting conversations with them.
    One of the curiosities of social software is the way it denatures much of the anxiety of face-to-face communication (as anyone who has been involved in an email mediated courstship or breakup can attest), but the very fact of that denaturing can make it appealing enough to be addictive to some.

    9:04 pm |

    Thursday, August 7, 2003

    History, Personalities, Wikis Redux

    By Clay Shirky

    I have wanted to post in the "what effect has the wiki had on the development of !Echo?" thread, but I haven't, because, somewhat to my surprise, I've been too emotional about the subject. I was put off my feed reading Liz's first post on the subject, because while my analytical self recognizes that her critique of wikis is correct -- they put some users off in ways weblogs or mailing lists do not -- my gut *wants* the !Echo wiki to work, because I love wikis and I think what the !Echo group is trying to do matters. So, since I see what Liz is saying, but I don't feel it, I am suffering from cognitive dissonance.

    Now, having copped to an inability to think dispassionately about the subject, I want to quote Sam Ruby:

    In defense of the wiki - had this merely been a weblog post or a mailing list, I am confident that we wouldn't be having a naming discussion right now.  Or any discussion.  Quite simply, it was the wiki that made this project possible.
    Sam posts for me -- the initial problem the !Echo project faced was not a standards war but a personality war, and the wiki was a brilliant counter-move. Afte rthe launch of the wiki, there was more movement, faster, than I have ever seen in any standards effort with that many players; the wiki was a big reset button, and it worked. The current discussion about its effectiveness is the result of a success crisis -- the critique of the wiki only matters because the process has created enough material of value to make it worth trying to identify and alter anything that might be holding the process back.

    And Sam is right about the inferiority of the alternatives: Mailing lists and BBSes amplify personality, and one shudders to think what a standards effort conducted by blog would look like. Classic discussion tools have a number of forced moves built in, like allowing the users with the fastest posting tempo to rule the conversation. They are also subject to any number of delaying and distracting strategies like ad hominem attacks, vertiginous changes in scope (e.g. "Plus, we could add an agent-based micropayment market for reputational capital!") or meta-conversations about the value of the actual conversation (e.g. "As long as there's a digital divide, does any of this really matter?") that, despite their irrelevance, almost universally excite social antibodies and lead to flame wars or enervating ping-ping discussions.

    Wikis, on the other hand, are all about unforced moves -- don't like something you see point? Edit it, or add to it. Or both. Don't think a subject is being adequately addressed? Create a NewPage and lay out your case there. Or combine pages, rescue deleted material, re-name things. Linear discussion makes roadblocks easy to erect, but a wiki has no such linearity. In an environment with none of the authorial perogatives of a mailing list or BBS, a wiki makes it obvious who is actually advancing the work and who is merely running their mouth.

    I've read a lot of the background material about the problems with the !Echo wiki over the last few days, and most of the complaints seem relatively inconsequential, while one complaint -- unfocussedness -- seems quite real and serious. The complaints that seem inconsequential to me are:

    - What's the discussion blocked on? Naming. This is the least important part of the discussion, and the one where there are the fewest external metrics of success. Mark Pilgrim can post two examples of XML side by side for view to illustrate an obvious point, but two names have no such objective presentation. Arguing about what to name the project is like arguing what to call your garage band, an utterly absorbing conversation that matters a lot less than, say, rehearsing.

    - "Can't find earlier conversations." I think Phil's suggestion that he held back from posting on the wiki because he couldn't check the archives (unlike a mailing list) is at best a red herring and at worst a cultural misunderstanding. If he has a point of view not reflected on the wiki, he should post it, period. Whether something was considered earlier doesn't matter if it should be discussed now. The ahistoricism of the Wiki Now lets ideas see re-consideration in new contexts, rather than prematurely consigning them to the dustbin.

    - "There's too much there to follow." This has nothing to do with the wiki -- it is a standard complaint of all such successful efforts. Whenever busy people come together to do anything outside the framework of their current jobs, there will always be too much content for some people, and alas, the busiest people are often the ones you most want to have involved, and the ones who have the hardest time making the time. It's not that volume is not a problem, its just not a problem with the wiki -- its a problem with any successful conversation that is wide-ranging, long-lived, and matters to a large number of participants, regardless of format.

    - "I could follow it on a mailing list." This might be true in terms of volume -- mailing lists are idea for rapid scanning -- but the content of the wiki couldn't have been produced on a mailing list. (See above graf on forced moves in mailing lists.) It might have been a good idea to have a tracking list alongside the blogs, with links pointing back to the relevant pages.

    Then there's the one complaint that seems serious and core to the problems cited:

    "Things seem unfocussed." This is, I think, the real and core issue -- the !Echo wiki isn't wiki-patterned enough. The early addition of the "RefactorOK" tag set three unwikish precedents -- editing would mostly be by appending or inserting paragraphs, as opposed to re-writing or restructing existing work; paragraphs would be signed, pushing each page towards annotated document or even in-line BBS format, as opposed to "sense of the meeting" emergence; and the wide use of RefactorOK undercut the presumptive right of each to edit all. As a result, many of the pages looks like a crappy BBS instead of a good wiki, and pages multiplied without being folded together.

    While the "from wiki to blog" publishing pattern was interesting, it was a broadcast version of a living document -- useful for those of us looking in from the outside, but not a replacement for the continual pruning, condensing and combining of pages, refactoring of names, and deletion of material that no longer represents a live point of view that makes a wiki work.

    I don't know the actual history of the RefactorOK decision or the subsequent lack of wiki gardening, but the background elements include David Winer's cancellation of his mailing list when the conversation became not to his liking, Sam Ruby's strikethrough treatement of Shelley Power's postings on his site, and Sam's not wanting to seem overbearing despite being one of the logical candidates for gardener. It may be at the outset that the re-emergence of You Own Your Own Words pattern helped bring people in, but the drag on the project from that pattern is now large.

    The solution here is to fix the problems, by refactoring and by offering alternate views of the content, on e.g. an accompanying mailing list. It may be that this will not work, or that !Echo will fail for any of the myriad reasons that standards efforts fail, but the early burst of energy came from inviting everyone to one place, in an environment that mitigated against conversational derailment.


    2:19 pm |

    Game Subscription curves

    By Clay Shirky

    If there's one thing I've learned in my long life, it's that chicks dig it when you use the words logarithm and homeostasis in the same sentence, so here goes:

    Sir Bruce has posted an updated graph of MMOG subscription numbers, and the results show, even more clearly than last time, a tendency towards long-term homeostasis, via roughly logarithmic growth.

    A few things jump out about this graph:

  • Rapid early growth seems (visually, I don't have the underlying numbers) to correlate with long-term population.
  • No game of the 19 listed here ever fell to zero users after launch, and only Motor City Online even suffered a moderately significant loss.
  • Exapnsion packs seem to correlate with population growth only weakly, if at all.

    This in turn suggests that the Star Wars Galaxies "get big fast" strategy is the right one, that expansion packs should be viewed as one-time revenue hits, and that the key to managing games like this long-term is knowing when and how to switch from fast growth mode, where you spend on marketing, to cash cow mode, where you simply control costs while collecting monthly rent from your long-term community.


  • 10:12 am |

    Wednesday, August 6, 2003

    The Network is the People

    By Ross Mayfield

    So I won the little bet, but there is little reason to gloat.

    You will recall that the reason I took the bet was the first point. Power laws exist when ties are weak. Say, with Clay's dinner money. We are all fascinated by the search prospects of weak ties, realizing how loosely connected we all are and that the horizon is not that far away. But what is of value is ties that are strong, real relationships.

    Private Referral Networks, like LinkedIn, work because they represent our transactional relationships based on social credit that drives relationships -- with discovery beyond our natural limits. This friction limits what is a tie, what is a "friend," because we put ourselves at risk when we seek reward. Friendster works because communal oversight out perform algorythms like Match.com's. LinkedIn makes social credit part of its process, which begets social capital. Relationships are full of friction, which protects us from overload and disrepute.

    Graph distribution is shaped by the friction of information flow. As Duncan Watts observed, "when the requirements for connections increase, connections diminish." By nature we all seek preferential attatchment. What keeps us from directly affiliating with the most connected node is the barriers kingpins errect to protect themselves and their natural limits.

    There are natural limits. With blogs as publishing, there is no limit for the amount of readers the writer will accept. Write once, runs everywhere. With blogs as communication, the limit (150) is the amount of conversations you can passively participate in. With blogs as collaboration, well, wikis mostly, the limit (12) is the amount of relationships you can actively manage.

    We aren't dolphins. If 1/3 of our network was lost, society would crumble. Power laws are indeed fractal, scale-free reaches small scale -- in absence of friction.

    The Network is the People. When we network, we have limits. Networkers within LinkedIn reached that magical upper boundary of 150. Sure, Joi and Reid (the two above 150 ties) may be cetacean delphi among us, but more likely they have allowed declarative ties for reasons beyond conversation. If we gave the bet more time, I am confident the rule of 150 would constrain the upper limit to flatten the curve.

    So we shall dine at the venue of your choice. Perhaps the splendor of Fiesta del Mar Too!, with mole poblano, habanero chiles, bottom shelf margaritas and a smattering of social software. Or New Bamboo for shaken beef and Singha.

    Im not going to gloat, as this was a close one. And there was another bet that it seems I will lose.


    3:12 am |

    Tuesday, August 5, 2003

    history, personality, and wikis

    By Elizabeth Lane Lawley

    On Tim Bray's blog there's a follow-up to yesterday's News.com story about the Pie/Echo/Atom project ("Battle of the Blog"). Tim talks about the reporter's attempt to get him to focus on the personality politics in the story--and Sam Ruby reports a similar conversation in his blog.

    Tim goes on, however, to say the following:

    You can't understand the real story--ever--without understanding the personalities and who said what to whom and when and why. Marxism had an alternate theory of history: that it was all predetermined by socio-economic forces and that the individual was not a factor in the story. There's a word for that theory: wrong.

    Contrast that with Clay's posting here last month on the topic of the project wiki:

    But there is a second reason, under the surface but possibly more important -- wikis denature personality. Echo exists not because there are things wrong with the RSS markup -- there are, but they could be easily fixed. Echo exists because there are things wrong with the RSS process. RSS is having not a technological crisis but a constitutional one, where who decides what concerning RSS is not clear, and will never be clear, because the people doing the deciding don't even see themselves as being part of a decision making body.

    Are there times when "denaturing personality" is useful? Sure. But Tim's points bring out for me where my greatest discomfort with wikis come from. I believe it matters who said what, and when. That context provides enormous "metadata" for me personally. And the wiki explicitly strips that. I understand why, and I do recognize its benefits. But I'm still uncomfortable with it.


    3:09 pm |

    Musing on Losing

    By Clay Shirky

    I lose.

    Three months ago, I spotted an incipient power law distribution appearing on LinkedIn, and asserted that it would a permanent feature of that service. Ross Mayfield disagreed, so we made a bet, based on 4 parameters derived from power laws. And though the official bet is not over til this Friday, its clear that I've lost, so Ross, I owe you dinner. (I hear Una Mas Taqueria in Santa Clara is quite nice.)

    Here is the image of the 956 people listed as being in my network under the category of internet, ranked in order of number of connections from 332 (most) to 1 (least):

    This is plainly a power law curve, with a fit of 0.95 (i.e. the real-world data fits a power law distribution very closely.) However, because the bet was on 4 characteristics, and I lost one of them, I lose.

    To recap, the bet was:
    1. The person in the #10 position will have less than 15% of the number of connections of the person in the #1 position. (Actual: 20.1% (Private to Ross: Ever been to Chubby's Broiler, over there in Sunnyvale?))
    2. The top 20% of the result set will account for ~80% (+/-5%) of the links. (Actual: 75.8%)
    3. The average number of connections will be at least double the median number of connections. (Actual: 6.2::2, more than triple))
    4. Neither the median nor mode number of connections will be in the double digits. (Actual: Median of 2, mode of 1)

    This also means that 79.5% of the LinkedIn users listed here are of below average connectivity.

    A couple of things are striking to me about the data. First, I lost because I bet on the head as well as the overall distribution. (Ross, the Falafel Drive-In on Steven's Creek is the *best*!). Though the trendline fit with an overall power law was quite strong, a perfect fit would have required a #1 position of over 1700, 5 times the actual size of Joi Ito's 332 links. So the curve seems to have flattened a bit on some sort of upper limit to human connectivity. (Even for Joi.)

    Next, there are only 5 people with connections in the triple digits, and these 5, accounting for less than half a percent of the total number of users, make up 17% of all connections. In random distributions that tend towards bell curves, removing such a small data sample would hardly have any effect, but here, those 5 users make up a huge part of the overall connectivity of the LinkedIn network.

    Furthermore, they are not spaced evenly -- #1 and #2 are closer than you would expect, as are #3-5, and those two sets of points are further from one another than you would expect. (Let's go to the close-up of the top 25)

    This may simply be noise, but I wonder if competitive behavior causes close clustering in the heads of other such distributions.

    Finally, these numbers are very close to the numbers I calculated in early June, meaning the predicted homeostasis did appear quite early.

    So I called it mostly, but not completely, right. (Paging Dr. Mayfield: Please meet your party at the Chi-Dog drive-thru window on 3284 El Camino Real.)


    1:01 pm |

    Sunday, August 3, 2003

    wiki backlash?

    By Elizabeth Lane Lawley

    There's an interesting thread on Phil Ringnalda's blog today regarding the naming process for the !(echo/atom/pie) syndication format project.

    Sam Ruby, who maintains the wiki for the project, asks in the comments why Phil feels unable or unwilling to re-open "Pie" as a naming option. He wonders whether it's the tone of the wiki name vote page that's keeping Phil from doing so.

    Phil responds thusly:

    Mostly, yes: it's the "here are the only seven possible names, there was ample time to suggest others, and we've done a ton of work to be sure that these are the only possible names already, now pick one" tone of the vote page.

    But it's also the lack of archives: if (Ghu forbid) it had been done on a mailing list rather than a wiki, I would know that there had to have been at least one discussion where Pie got thrown out, and could look back to see the complete history, and if it had been casually tossed out before the difficulties of finding a workable name were really clear, I could say "why not reconsider Pie, now that we know we can't just pick anything we want?" As it is, I have to assume there's some good reason (surely they didn't just throw your name out for no reason at all, did they?), but even if there is a WhyYouCantHavePie page, I can't assume that it actually tells the whole story, just the story that the last few people to edit it wanted told.

    I like compromise, where someone clearly gives up something they want in order to get forward progress, with the feeling that maybe next time their ox won't be the one to get gored. I don't think I like consensus, where everyone silently joins the faceless crowd just because it is a crowd.

    Phil's not the only one with concerns about wikis for these types of collaborative processes. It's interesting to read the comments on a recent post by Sam regarding the wiki process. Several extremely capable people--including Shelley Powers and Dare Obasanjo have expressed discomfort with the chaotic nature of the project wiki.

    I see a backlash building on wikis, mostly in a good way. The choice of a tool for online group-forming or process facilitation is an important one, and the pros and cons of specific tools need to be carefully considered by anyone wanting to implement them.

    I'm not yet at the point where I see wikis as adding sufficient value to any process I'm involved with to justify the installation, configuration, and learning curve for users necessary to add another tool to my social software arsenal. Like Phil, I continue to be troubled by the inherent ahistoricism built into the wiki environment; like Shelley I find the lack of social cues to tell me if I'm treading on someone's toes by changing content to be inhibiting; like Dare, I find that large-scale active wikis are often too chaotic and disorganized, making it difficult for me to find what I'm looking for. But I'm still willing to be convinced.


    9:50 pm |

    Virality

    By Ross Mayfield

    According to Reuters Webhookups are Blamed for Rise in AIDS
    A growing number of gay and bisexual men in the United States are engaging in risky sex with partners they meet on the Internet, raising fears that the AIDS virus could be poised for a major comeback in the group hardest hit by the epidemic.

    Online chatrooms and Web sites are replacing gay bathhouses and sex clubs as the most popular meeting point to arrange high-risk sex, according to two new studies presented on Tuesday at the 2003 National HIV Prevention Conference.

    People are using these tools to meet for all kinds of reasons. A good thing. Before anyone gets in some social conservative reactionary jerk, let's talk about how the tools could help.

    A social networking model could be developed for groups to privately and voluntarily share health status. Participants anonymously contribute their symptoms and/or diagnosis. In return, participants are given a probability of contagion from their real social network based on known epidemiological science. Knowing this probability would lead to better prevention, reduce exposure and inform practitioner queries -- without invading privacy.

    In high health risk environments, like the petri dish that is preschool, people talk about their family's health all the time. This would be a convenient way of talking about health to improve it. Something tells me the marketing would be viral.


    9:17 pm |









    Copyright 2002-2003 Corante. All rights reserved. Terms of use


    Archives, RSS & Email


    Access our archives, syndicate by XML/RSS or subscribe by email:


    powered by Bloglet

    Your Hosts
    Jessica Hammer
    Elizabeth Lane Lawley
    Ross Mayfield
    Sébastien Paquet
    Clay Shirky

    Blog Neighborhood
    Danah Boyd
    Stowe Boyd
    Timothy Burke
    Stewart Butterfield
    Marc Canter
    Tom Coates
    Flemming Funch
    Anne Galloway
    Dan Gillmor
    Meg Hourihan
    Joi Ito
    Steven Johnson
    Matt Jones
    Peter Kaminsky
    John Lebkowsky
    Adina Levin
    Jim McGee
    Cameron Marlow
    Jerry Michalski
    Matt Mower
    Ray Ozzie
    Mark Pilgrim
    Shelley Powers
    Mitch Ratcliffe
    Howard Rheingold
    Doc Searls
    Jeneane Sessum
    Dave Sifry
    Jon Udell
    Matt Webb
    Dave Weinberger
    Kevin Werbach
    Dave Winer

    Blogging & Social Software Sites
    AllConsuming
    Blogdex
    Blogroots
    Blogshares
    Daypop
    Meetup
    Metafilter
    Ryze
    Socialtext
    Technorati
    Weblogs.com

    Listed on BlogShares

    SSA