Free service, meet free service. FIGHT! Google's issue is that, to remain free, it's got to be cheap. Very cheap. They can't be paying higher fees! The math just doesn't work out. But this situation illustrates another issue with our expectations of free: if free services have to be cheap, does that mean we can expect lower quality service? Does that make free services a race to mediocrity? Google Voice certainly seems to be heading that way. What's frustrating is that many people won't care… until a service they use gets cut in cost savings.

What's even more troubling is that Google is arguing that it's not a phone company, so it can play by its own rules. Federal laws prohibit phone companies from blocking numbers, precisely because universal access is a communicative public good (using Fulk et. al's terminology). But Google just provides a piece of software that piggy-backs on the web. Well, excuse me, but that is bullsh*t, and regulators will soon show Google Voice that. When they do, will it be the end of Google Voice?

(Click for a larger image…)

Anyway, I think Dahlen's got a great point. In fact, this is how many examples of user-generated content have worked from the beginning, especially if we step back from the idea of 'machine-mediated' to the more general 'system-mediated'.

Wikipedia

In Aaron Swartz' now famous blog entry "Who Writes Wikipedia?", he shows that Wikipedia has always (or at least since 2006 when Swartz wrote this) followed the model of UGMMC. Swartz does a quick analysis showing that most of Wikipedia's content gets put into the system by a diverse array of users, many of them casual contributors. Those raw materials then get massaged and organized by a committed core of editors. This is system-mediated, and I would guess this model continues to be the thing that makes Wikipedia as good as it is.

Team Fortress 2

I'm thinking of two examples that speak to this issue here. The first is Valve's model for interacting with its community. Via its forums, Valve's developers are always asking for ideas about how to improve gameplay and add content to classes. It's hard to tell how much of the final product is actually user-generated, but I would imagine a good bit. We can think of this model in lots of places where companies solicit ideas from the community, and then polish them into something solid.

But Valve has sometimes gone even further than that with its UGMMC. Take for example the classic user-generated video, FYI I am a Spy (don't worry, it's only 7 seconds long):

When it updated the Spy class with new content, Valve added an achievement based on this video:

FYI - I am a Spy: Backstab a Medic who has healed you in the last 5 seconds.

One other feather in Team Fortress 2's cap: Valve has built a reputation on polished gameplay and clever content. But it has also left the door open to unrestrained user-generated content in intelligent places. Case in point: sprays. While you're in-game, TF2 allows you to "spray" a small picture (animation possible) on the wall inside any level. Restrictions – it can only live in one place, so spraying it again removes the last spray, and when levels restart the sprays go away. So, Valve left the door open for players to put anything they want in a level, but they've restricted users from really changing the look of the game by covering the walls in graffitti. Result: interesting and hilarious uses of sprays. Note, this video is much longer and potentially NSFW. The whole thing is pretty funny, but skip to 3:58 for creative use of sprays:

Antin, J., and C. Cheshire. 2009. “Readers are not Free-Riders: Reading as a Form of Participation on Wikipedia.” in Proceedings of the ACM 2009 conference on Computer supported cooperative work. Savannah, GA. (PDF)

You can check out the full advance program too. Looks like lots of fantastic papers in there!

Coye and my chapter in a new book also co-edited by Coye is out: eTrust – Forming Relationships in the Online World. This is the latest volume in the Russell Sage Foundation's series on trust, and it's chock full of great stuff.

Amazon doesn't list it as released yet, but you can get it from Russell Sage directly.

The article brought up some thoughts I started knocking around last summer while working at Yahoo! with Elizabeth Churchill. Namely this contradiction, the form of which I borrow from Larry Downes' Law of Disruption:

Attitudes about privacy change incrementally, but revenue models for free online services change exponentially.

Result: what companies need to do to support free services clashes with privacy attitudes. Note that I'm not making any statement about the goodness or badness of behavioral targeting. My only point is that the expectation of free online services sometimes doesn't jive that well with the fact that the companies that provide those services have to find a way to make money.

As concerns about privacy, control of personal data, and behavioral targeting escalate, I predict we're going to find at least some companies crying foul that consumers want to have their cake and eat it too. This could lead to a move towards subscription services that relieve companies from the burden of having to make money from targeted ads, market research, and the like.

Update: Here's the original study.

Hoffman, David A., and Salil Mehra. n.d. “Wikitruth Through Wikiorder.” SSRN eLibrary. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1354424.

The authors present a nice history of Wikipedia's dispute arbitration process, and a solid analysis of published arbitration decisions. Very cool. By far the best part, however, is this sentence on the challenges of coordination in online collective action:

Altruists, like cats, are hard to herd.
 
 

(Thanks to Megan for the heads up on this one!)

(Click for a Larger Image)

If you've been up Half Dome, you know all about this. If you haven't, let me try to explain how ridiculous it is. You begin in Yosemite Valley at about 4000 feet, and begin the climb straight up past two waterfalls, Vernal and Nevada. We chose the Mist Trail (not so misty this time of year), which consists of a series of steep granite steps and switchbacks. About 2000 vertical feet and almost 3 miles later, you're at the top of the falls. Now hike another 2 miles (and 1000 vertical feet) to the base of sub-dome, where the hard work begins. A vertical-seeming face with switchback steps cut into it. Get to the top, and your reward is pictured to your left: a final, slick face of 50+ degree rock face with two thick metal cables running up it. Metal poles and slats are placed every so often. Grab a pair of gloves and haul yourself up. Stupid, right? Not as stupid as trying to get down, shimmying from slat to slat because your boots have almost no traction at all.

Well, we survived. Certainly there was spectacular scenery, but the most memorable part of the trip for me was being on the cables. It's like nothing I've ever done. On the side of this slick rock, hanging precariously off of a metal cable along with fifty or so other people. But somehow the only drama is a fallen water bottle or two.

In my research I frequently use the notion of social value orientation (SVO). The idea is that people have certain general dispositions towards the distribution of rewards from themselves and others. You've got the self-interested folks who worry about themselves. You've got the competitive folks who try to maximize the difference between themselves and others. And you've got the pro-social folks who worry about others.

Well, being the typical geek that I am, I was hanging off of the side of Half Dome and thinking about how my experience speaks to SVO. Getting up and down those cables safely requires that everyone works together. Some people are just worried about their own adventure, bitching about slower climbers and failing to make room for others when they need to. Some people just want to beat everyone – like the intrepid souls who decided to climb outside the wire and blow right past everyone. But most people are considerate and encouraging (pro-social), they help others, they're patient, and concerned for each others' safety. It's the only reason more people don't die out there.

I think it says a lot about human beings – or at least the people who go to Half Dome – that we can end up hanging on a wire, in a life and death situation (albeit a minor one), and that everyone looks out for each other. My experience mirrors what a lot of the participants in my interview research have been saying about Wikipedia. Why does Wikipedia exist? Why do people put in the time and effort? They believe it's because people are generally good and giving. Because they want to share, and they want to help. I think there are many other motivations that drive people, but I love that people believe in the good as a primary reason. Having been up and down Half Dome recently, it seems like a perfectly reasonable assumption.

First, let's make sure we're on the same page about public goods. Public goods have two properties – when one person takes advantage of the good, it doesn't reduce the amount available to anyone else (that's called non-rivalry) and you provide the good to everyone or no one, you can't selectively exclude some people from taking advantage of the good (that's called non-excludability). Traditional examples of this are things like clean air and national defense. My breathing easier, my being safer doesn't take away from your breathing easier or being safer. And we all breathe easy, we all benefit from safety.

So, is Twitter a public good? Well, yes and no. First let's start with the 'no'. Depending on how it's used, Twitter is a point-to-point or broadcast communication tool. In that capacity, we could argue that it does not constitute a public good. No more than email or letters do, anyway. If I post a tweet about my breakfast, where is the public value there? Where is the 'good' in the sense of something which can benefit many?

But then again, there is a 'good' there. It's a derivative good, but it's important. When people make their tweets public, they are doing at least two things: first, they are communicating with friends and family (or strangers). This is probably what they were trying to do. But second, they are contributing a bit of information to a collective body of real-time information about what people are doing and thinking about.

This is the power of Twitter trends and Twitter analytics. (See Twitter Search, Twitalyzer, Tinker, Brizzly, or Trendistic, just to name a few…) By aggregating all those tiny bites, we get a public body of information that can tell us a lot about what's going on. If I want to know what people are thinking about, paying attention to today (or at least what the tiny fraction of Americans who Tweet are thinking about), I can use Twitter to find out. And just like any public good, there's a social dilemma there. I read my Twitter stream all the time, but I almost never Tweet myself. I search the stream and look at trends, but I don't add my bites to the stream. I'm a taker, but not a giver. So technically I'm a free-rider.

But there's another interesting dimension there. If the Twitter public good is a derivative by-product, that means that many or most people are contributing to it without intending to. Like i said, they're just sharing some info. about their day. Many of them may not be aware that their info. is helping to make Twitter Trends more powerful. So how can we call them free-riders? It's like saying the people who don't edit Wikipedia because they don't know they can are free-riders. From one point of view they are, but the definition of a free-rider requires an informed choice to take advantage of others. So, we'd better look more closely.

Here's another example: Netflix' movie rating system. If I never rate movies, I still benefit from the algorithms that suggest movies I might like, but I'm not putting in my little bites. This may be because I don't even know that this is how the movie recommendation algorithm works. Am I a free-rider, even though I never knew about the derivative product of my ratings? On the other hand, I might rate movies all the time, but only because I like to have a reminder of how I liked movies I've seen over time. I don't know or care that those ratings get used for anything else.

Anyway, I think there are some interesting issues here. But getting back to the original question – Is Twitter a public good? – I'm going to come down on the side of a definitive 'Yes!'. But there's a lot of interesting thinking and research to do on the question of exactly how it's a public good.

I'm a gamer. I play video games. But unlike many gamers, I play only one game. So, I guess I should say 'I play video game.' I've dabbled around in others, but mostly it's just the one. I play Team Fortress 2 (TF2, for short), created by Valve. I bought this game a few years ago, and I've been hooked ever since. But, you know, I'm not addicted. I could quit any time.

I continue to be amazed at how it keeps my attention over time. In fairness, I was enabled in this time-sucking pursuit by an unlikely conspirator: my academic adviser, Coye Cheshire (also a closet gamer). As it happens, being directed towards this particular game was fortunate given my areas of research. For someone who studies online collaboration, social psychological incentives, and computer-mediated communication, TF2 is like a giant sandbox with one of those amazing lever-operated wooden backhoes. Engaging deeply with TF2 has convinced me of these things, which I hope also to convince you of:

1. Valve is brilliant. Team Fortress 2 is brilliant.
2. With respect to incentives and community management, Team Fortress 2 is the most socially advanced game of its type.
3. TF2 should be a model for designing and implementing effective incentives for online participation.

Over the next few weeks I'm going to post on a variety of topics and, I hope, 100% convince you of these statements.