Carr-Benkler wager revisited

Nick Carr wants to settle the Carr-Benkler wagerHe thinks he won.  I would have thought that by 2012 it was clear that the Web is in fact focused on social media, participation in which is driven primarily by people’s desire to express themselves with and to each other, rather than on sites that have figured out how to pay their top contributors in order to draw audiences.  Peer production, the term I had introduced in 2001 in Coase’s Penguin, covered the fact that a critical source of value on the Net was the diverse experience and insight of many different people, with diverse backgrounds, experiences, and motivations.  As I wrote in 2001, “By connecting a very large number of people to these potential opportunities to produce, the e-text projects, just like Clickworkers, Slashdot, or Amazon, can capitalize on an enormous pool of underutilized intelligent human creativity and willingness to engage in intellectual effort.”  The reference to Amazon here was to the then-innovative way in which Amazon had integrated unpaid user reviews as a central part of the value Amazon the site provided when people were selecting books, by comparison to professional book critics. A feature that has become so normal in our view, that no one even stops to think what a profound shift we have undergone when the main selling point of Angie’s List is that it doesn’t use reviews other than those of real people.  It’s become the new normal.  I then embedded this economic observation in a broader sense of the implications of these trends for democracy and autonomy, and outlined the political and regulatory battles that the incumbents of the 20th century are engaging in so as to squelch the rise of social production and engagement in The Wealth of Networks in 2006.

It was in 2006, in response to the book, that Carr used Jason Calacanis’s effort to recruit the top contributors to Digg by paying them as a hook to criticize my view and argue that the future of the Web was to become a platform for delivery of professional content.  The question he raised was clear:

“I think that what Calacanis is getting at is that the reason “social media” has existed outside the price system up until now is simply that a market hadn’t yet emerged for this new kind of labor. We weren’t yet able to assign a value – in monetary terms – to what these workers were doing; we weren’t even able to draw distinctions between what they were contributing. We couldn’t see the talent for the crowd. Now, though, the amateurs are being sorted according to their individual skills, calculations as to the monetary value of those skills are starting to be made, and a market appears to be taking shape. As buyers and sellers come into this market, we’ll see whether large-scale social media can in fact survive outside the price system, or whether it’s fated to be subsumed into professional media. Which is mightier – Benkler’s dream or Calacanis’s wallet?”

“Benkler’s dream,” in The Wealth of Networks, is that what is critically new about the networked environment is that it allows people to come together, to act with and for each other in productive relationship.  In commenting on Carr’s point, I suggested that we look and see whether in the future the most influential sites will be those that allow people to come together and create with and for each other for social motivational reasons, rather than those that would be based on being able to clear markets, draw the best talent by paying it, and draw the most audience by providing that paid for, professional media content.  So, I wrote on his blog:

“The reason is that the power of the major sites comes from combining large-scale contributions from heterogeneous participants, with heterogeneous motivations. Pointing to the 80/20 rule on contributions misses the dynamic that comes from being part of a large community and a recognized leader or major contributors in it, for those at the top, and misses the importance of framing this as a non-priced social process. Adding money alters the overall relationship. It makes some people “professionals,” and renders other participants, “suckers.” It is not impossible to mix paid and unpaid participants, as we see in free and open source software and even to a very limited extent in Wikipedia. It is just hard, and requires a cultural form that is definitely not “now at long last we can tell who’s worth something and pay them, while everyone else is just worthelss.” What Calacanis is doing now with his posts about the top contributors to Digg is trying to alter the cultural interpretation of what they are doing: from leaders in an engaged community, to suckers who are being taken for a ride by Ross.Maybe he will succeed in raining on Digg’s parade, though I doubt it, but that does not mean that he will succeed in building an alternative sustained social process of peer production, or in replacing peer production with a purely paid service.”

To test this, I suggested:

“I’m happy to accept this wager as a measure of the quality of my predictions about the long term sustainability of commons-based peer production. The shape of the wager, however, should be clear. We could decide to appoint between one and three people who, on some date certain – let’s say two years from now, on August 1st 2008 – survey the web or blogosphere, and seek out the most influential sites in some major category: for example, relevance and filtration (like Digg); or visual images (like Flickr). And they will then decide whether they are peer production processes or whether they are price-incentivized systems. While it is possible that there will be a price-based player there, I predict that the major systems will be primarily peer-based.”

Now Carr thinks he has won.

So what’s happened over the last five years? Let’s look at the blogosphere. Many of the most popular sites in 2006 were strictly amateur productions, often written by a lone scribbler. Today, the most popular blog sites are almost all corporate productions, usually written by teams of wage-earners employed by corporations, often large media corporations. That doesn’t mean the amateurs have gone away; it just means they’ve been marginalized. Video? In 2006, YouTube was a playground of amateur videographers, uploading their work for kicks. The amateurs are still there, but the most popular videos today are corporate productions – from TV networks, film studios, recording companies, publishing companies, game studios – and even a lot of the amateur productions are wrapped in advertisements. Beyond YouTube, online video is dominated by sites syndicating professional productions: Hulu, Vevo, Netflix, and the various TV networks, et al. Online music? Definitely dominated by sites offering professional productions from record companies (iTunes, Pandora, Spotify, Amazon, etc.). Again, there’s still plenty of amateur music online, but the dominant outlets are fundamentally commercial. Even open-source software – one of Benkler’s core examples of peer production – has shifted in the last five years toward commercial dominance, with many contributors to the largest open-source projects being salaried employees of software firms.
There are certainly places where unpaid contributors continue to play a dominant role online – photography, Wikipedia – but they are exceptions to the net’s general evolution away from a populist medium and toward a commercial one.

 

Let’s take a look, then.

Start with Carr’s original motivating example.  It turns out that Digg did not lose to Weblogs, despite Calacanis’s wallet.  It was Reddit, which flanked Digg from the peer production side.

Lets take a small set along the lines I proposed.  I don’t really feel like doing much more on a Sunday morning.  In parenthesis I have (Alexa rank) next to each site, not because it’s the most perfect ranking but because its easy to get and a first cut of what a real comaprison would look like, and there’s no systematic bias relevant to the question at hand. I have in brackets [Y] or [N] for my sense of whether it supports [Y]ochai or [N]ick.

Discovery, or relevance and accreditation (how do we know where to look for new and interesting sites): Twitter [Y](5); Facebook[Y](2); Reddit[Y](124); Tumblr[Y](37); StumbleUpon[Y](147)

Images: Flickr [Y](48); Google images[Y]; Photobucket[Y](153); Corbis [N](7,424)

Video: YouTube [Y](3)(if Nick is reduced to winning this social bet by having to argue that YouTube is an example of Calacanis’s wallet, all I can do is smile); Google video[Y]; Vimeo [Y] (116) as compared to Hulu [N] (214); Vevo [N](2,413); Netflix[N] (101)

Travel: Tripadvisor [Y](253); as opposed to Lonleyplanet [N](1259), Frommer’s[N](5,440) pr Fodor’s [N] (6,277).

Restaurants reviews: Yelp[Y](181); Zagat[N?Y?] (14,665) [the question mark is because, while older and more established, Zagat's original book form innovation was that collective judgments of many diners would be more useful than the views of highly-paid professional food critics].

Music distribution: Music sites harder to pin down; Nick has it right about the legal sites, but clearly wrong about the peer to peer uploading, the cyberlockers, and the increased use of YouTube as a distribution site.  So if you are looking at discovery and distribution, social networks and peer to peer recommendation are much more important now than payola or iTunes; Facebook is critical precisely because of the peer element of recommendation, not because it is a fantastic broadcast platform.

Music funding: Kickstarter is an interesting question.  Clearly its about raising money to support artists, so it supports the interest in professionalism, a la Nick.  But the model of funding is more “Benkler’s dream” than “Calacanis’s wallet”: going to fans and raising risk capital on social motivational model, in exchange for no share in the upside, but based on intrinsic motivations to support the work.  Overall, in 2006, would you predict that artists would get funded by people ponying up real money without expecting to make big bucks in return because of social motivations; or would you predict that the industry would find ways of monetizing it all and re-asserting control over independent artists.  If you were the latter; you’re with Nick, and you wouldn’t have predicted Kickstarter.  The former is where I was standing.

Another way of getting at this is just looking at the top sites.  Again, taking the top 20 sites in terms of Alexa ranking for US sites, in order.  I treat search engines as neutral because they capture both kinds of production.

Google[-]
Facebook [Y]
YouTube[Y]
Yahoo![-]
Amazon[N/Y] (the sales is critical; but its not about paid providers; and user reviews are an important component of the value; on balance I’d give this one to N)
Wikipedia[Y]
eBay[N]
Twitter[Y]
Cragislist[Y]
LinkedIn[Y]
Blogspot[Y]
Windows Live[-]
MSN[-]
Go[-]
Bing[-]
Pinterest[Y]
PayPal[-] (it’s a payment utility, not a paid-content site)
Tumblr[Y]
t.co[Y]
ESPN[N]

So who won is easy to infer, unless you are willing to say that people flock to social networks because of all the paid content, rather than because that’s where its awesome to hang out with others; or unless you are willing to completely distort my argument and say that whenever anyone is making money out of distributed, socially-motivated productive engagement it stops being peer production, as opposed to what I have said for over a decade: for investors, no less than for social innovators and activists, the place to be is in sites that build platforms for social interaction and engagement, not places that figure out how to pay for the best content (I suggest you ask Fred Wilson and Brad Burnham at USV whether when they invested in Twitter or Tumblr, Kickstarter or StackExchange because they thought they had found the next big Calacanis’s Wallet, or because they were investing in peer production).

And that’s before we think of how protester videos were the most important source of international video coverage of the Arab Spring; before we think of where we got images of pepper spraying police in Occupy protests; before we learned to understand Andy Carvin’s approach to coverage of the Arab Spring, or how the Guardian is innovating in integrating social reporting into its model.

It doesn’t seem to me to be a close question as to who was right.

3 thoughts on “Carr-Benkler wager revisited

  1. Yochai,

    Thanks for your considered reply. I’m sure you’ll be shocked to discover that I disagree. I think you’re doing more than a little cherry-picking here, both in choosing categories and in defining categories. Most important, you’re shortchanging the actual content that circulates online, which I would consider the most important factor in diagnosing the nature of production. For example, you highlight “music distribution” and “music funding,” but you ignore the music itself (the important thing), which continues to be dominated by price-incentivized production – even on p-to-p networks, the vast majority of what’s traded is music produced by paid professionals. The same goes, for example, for news reporting, where peer production is a minuscule slice of the pie. Where is the great expansion of the peer production model that you promised six years ago? The fact is, it just hasn’t happened. Even open-source software has moved away from peer production and toward price-incentivized production. I don’t say that this is necessarily a good thing, I just say that it’s the reality of what’s happening.

    Here’s my (quick and dirty) alternative rundown of whether the most influential sites in major categories of online activity are generally characterized by price-incentivized production (pi) or peer production (pp):

    News reporting/writing: pi (relatively little pp)

    News opinion: pi (Huff Post has mix, but shifting toward more pi, both by Huff Post paid staff and for others writing as paid employees of other organizations; other popular news opinion sites are dominantly pi)

    News story discovery/aggregation: unclear (are Twitter, Reddit, Stumbleupon et al. really the most influential, or are they secondary to the editorial decision-making by sites like NY Times, WSJ, BBC, Daily Mail, and other traditional papers and magazines? I don’t think this is an easy one to figure out. My own sense is that traditional news organizations play dominant role in shaping what news is seen and read online)

    General (non-news) web content discovery/aggregation: pp

    Blogging: pi (big change from 2006)

    Other writing: pi

    Music (creation): pi

    Music (discovery/aggregation): difficult to say, as a lot of sites have mixed models, but I would give the edge to pi based on dominance of big retailing sites (iTunes, Amazon), big streaming sites (Spotify, Rhapsody), and artist and record company sites. Certainly since 2006, there’s been a shift toward commercial sites.

    Photographs: pp

    Video (creation): pi (with all due respect to YouTube amateurs, professionally produced TV shows, TV show clips, movies, movie clips, music videos, advertisements, now dominate online viewing through YouTube, Hulu, Netfllix, tv sites, newspaper sites (e.g. WSJ video), professional blog sites, etc.

    Recipes: pi

    Charitable giving: pi (i.e., operated by paid staff)

    Reviews: pp (a big, diffuse category that could be broken up in many ways, but I think overall pp dominates in use, if not influence)

    How-to advice: pi

    Encyclopedias: pp

    E-books: pi (that goes for even self-published e-books)

    Apps: pi

    Games: pi

    Advertising: pi (except classified advertising, which goes pp thanks to Craigslist)

    Education and training: pi

    Weather: pi

    Scholarship: pi

    Pornography: pi (again, with all due respect to the intrepid amateurs)

    The fact is that peer production hasn’t revolutionized models of content production online, even if we grant (as I do) that peer production continues to play an important (and occasionally dominant) role in some areas.

    Social networking? Yes, social networking is, by definition, social networking, but even in what we define as social networking, do you really dispute that price-incentivized content is not playing an ever larger (rather than smaller) role? Here, to illustrate, are the Top 20 most popular Facebook pages as of the end of 2010 (most recent stats I could find):

    1. Texas Hold’em Poker (Zynga page)
    2. Facebook
    3. Michael Jackson
    4. Lady Gaga
    5. Family Guy
    6. YouTube
    7. Eminem
    8. Vin Diesel
    9. Twilight Saga
    10. Starbucks
    11. Megan Fox
    12. South Park
    13. Coca-Cola
    14. House (TV show)
    15. Linkin Park
    16. Barack Obama
    17. Lil Wayne
    18. Justin Bieber
    19. Mafia Wars (Zynga)
    20. Cristiano Ronaldo

    Of the ten most viewed YouTube videos of all time (as of this month), nine are professional productions and only one (yes, it’s “Charlie Bit My Finger”) is a non-price-incentivized production. Even Twitter is far from a pure p-to-p site; many contributors contribute as part of their jobs. I think what we’re learning – and it’s hardly a surprise – is that when a for-profit corporation operates a social network, it becomes steadily more commercialized.

    I don’t expect that I’ve convinced you that I’ve won the bet, but it sure would be nice if you would acknowledge the web’s shift, since 2006, to a medium ever more dominated by professionally produced content and ever more controlled by commercial interests, for better or for worse.

  2. Pingback: The Carr-Benkler wager and the peer-powered economy — Tech News and Analysis

Comments are closed.