Stowe Boyd

Better Social Plumbing For The Social Web

My recent talk from the Web 2.0 Expo in Berlin, which I retitled “Better Social Plumbing For The Social Web” instead of ‘Better Media Plumbing…”. I include the notes that I prepared, with minor tweaks.

My topic is not new — in the sense that I started writing about it a year ago. But I think it is of growing importance.

The basic premises that underlie social media — the fundamental relationships that link authors to the community of readers — are changing in the face of new and different social tools. In a sense, I am chasing the elusive question about where community resides, once again.

In this presentation, I plan to explore the root causes of today’s social media plumbling — the stuff that makes it social — and to outline the stresses that new social metaphors are creating. Lastly, I wave my hand at where it might all be headed.

I am best known these days for my writing (and the thinking behind it) at /Message and other blogs at www.stoweboyd.com. I have been involved with the development of various interesting ideas, like social tools, flow applications, workstreaming, web culture, edglings, microblogging, and new localism. I have spoken at dozens of conferences in the past ten years, like Lift, Reboot, Web 2.0, Enterprise 2.0, Defrag, Supernova, Under The Radar, Mesh, Next, and many, many others.

I work with a lot of start-ups building social tools, and larger companies trying to make sense of them.

Apologies. It was blogging that did this to me. No neat conclusions. A barrage of conjecture, wisecracks, and one-liners, disguised as a presentation.

My work is social tools, but my goals lie beyond.

We have inherited the Web 1.0 vision of the Web as a giant network of documents, linked to each other, where you can wander forever.

Where are the people? Oh. It’s the authors. They are people. They create links, but there are not in the immediate foreground. It’s documents, all the way down.

So in the Web of Pages — Web 1.0 — pages are more important than people.

And why are we here? “I must be here to read about things and follow links” since that’s what is the most natural thing to do, not to interact with people.

The Web of Pages wants us to be hunter/gatherers. We search, find a link, click it, see if it’s want we want, if not, we keep following leads.

It exploits the spatial sense in our minds.

People create the links that connect pages, asserting relationships between the info on the pages, and by extension, relationships between the people reading and writing the various pages. But it all seems pretty far away from human conversation. You write something, I find it, and write my own thing, and point at yours…. Its more like sending letters than conversing.

Yes, Google builds its page rank based on people’s actions — creating links and the identity of who created them — but all that seems way down in the subbasement, far far away.

Links are obliquely social, but direct conversation — through chat, social networking messages, and (most central to our subject today) blog comments — is where web media really become social media.

Asynchronous in nature, but can be near real time.

A blog is a long-lived repository of discourse, a place to ask questions, contend, agree, make suggestions, enlist support, and offer counterpoint. Nearly every possible sort of conversational interaction can be shoehorned into the lowly comment thread.

This is one of the reasons that the blogosphere has gone from a fringe phenomenon in 1999 to mainstream in 2009, where the leading print media outlets of the pre-web world have embraced the blogging paradigm at a fundamental level.

Social media is distinct from pre-social media in many ways — web based, individual voice dominates, etc. — but it is the social dimension that defines the difference. Social media is contrived (at least the best instances are) around the premises of open discourse between individuals.



While the publishers of blogs retain (in general) the ability to moderate comments, largely the model seems open. And the authorship — based on the identities of the bloggers and the commenters — makes the relationships that seemed so oblique in web 1.0 much more obvious and direct.

The heart of social tools is the individual, which is why I say social = me first. In the world of blogging that translates — not too well, actually — into ‘bloggers first, commenters second.’

The current baseline of blog technology puts very strong controls into the hands of the blogger, and hardly any in the hands of the commenters.

What sort of sociality is going on in a comment exchange? It’s a strangely unequal forum, where the blogger has nearly absolute controls, and the commenters — even in some sort of collective fashion — have no power. The blogger can delete any comment, or mark it as spam, while the commenters can’t change a single comma in the blog posts they comment on. (Of course, some commenters become bloggers themselves, and then the comment exchange becomes slower and distant, cross-linked by trackbacks or URLs, but pulled out of the comment stream.)

This asymmetry in control and ownership — who gets the revenue from the blog ads? — has ramifications throughout the web. For example, the page rank for a blog is attributed to the writing skills of the blogger, but it is the links that make the rank, not what is on the page. What is on the blog — the posts and the comments — are what lead to strong reputation for a blog, and that is due to the commenting — the sociality at the blog — but the author claims all the benefits, including all the revenue.

It’s interesting that social media is often heralded as proof that companies have lost control of their ‘message’ but at the same time that blogging keeps so much control in the hands of the blog publisher.

And there is the ‘comment dispersal’ problem, where very active commenters wind up with their participation spread all over the web, and no collation of their contributions.

This is one of the reasons that solutions like Co-comment and Disqus have been making ground, because an individual’s stream of contributions can be pulled together somewhere, and in a sense ‘owned’ by the participants and not just the blog publisher.

The seed for the change in the blogosphere was a seemingly small advance.

RSS feeds are a way to receive the posts from blogs without visiting them. Instead, using an RSS reader, a webhead can instead have posts from any number of blogs deposited into an RSS tool, like Google reader. This is a sort of out of body experience, since the user doesn’t hunt and gather anymore, wandering around looking for new information by search and following links. Instead, a steady diet of bite sized morsels simply appear. And generally without the comments.

So the shift to RSS by the most technically hip websters meant that

1. they were seeing posts out of social context, without the conversational interaction that made the blogosphere a third place, and not just forty million people standing on soapboxes, and
2. over in the RSS reader the former participant — now a reader — is at least one and maybe more clicks away from adding comments. And of course there is a subtle devaluation of comments because they are read less often and harder to get to.

So, people find it easier to take other actions using tools outside the blogs, rather than comment. For example, creating shared bookmarks a la Delicious may seem of higher value for the individual and that individual’s network of friends than writing blog comments.

And right on the heels of RSS feeds, we began to see the rise of other social tools where conversation was the centerpiece, not the sidebar.

And these tools rely on our sense of timing, not a spatial sense. We are not wandering around looking for things to read: instead, things to read just show up in some regular timeframe.

Tools like Digg and Techmeme share a few key characteristics, and define two ends of a continuum. There is a stream of new information that finds its way to the pages of the website: in the case of Digg this arises from individuals recommending pages for others’ attention, and in the case of Techmeme an algorithm looks are the clustering of a short list of A list bloggers to see what is getting the attention of many. In both cases, the user is presented a stream of information, ordered by the observed actions of others.

Here we find the key attribute of all important social tools going forward: the collective actions of some group of people shape — order, filter, embellish — a stream of information. One sort of embellishment is a comment, like the comments traditionally left on blog posts or in bookmarks. But these comments never find their way back to the blogs, if they were sparked there.

Also, the way the sites work feature emergent properties from the community, collectively, that can’t happen on a single blog all alone. On Digg things rise (or fall) relative to others; and on Techmeme stories rise and fall by clustering of authors. In this regard, the evaluation of the value of a single blog post is one of the outputs of Digg and Techmeme.

As a result, participants’ behavior can change when exposed to these tools. It becomes more fun to Digg a blog post than to comment on it. As a reader of Techmeme, you find it unprofitable to comment on blog posts — even of posts currently active on Techmeme — since a/ comments don’t show there, and b/ the comments have no influence on the Techmeme algorithm.

So, there are now dozens of streaming applications — Digg, Facebook, Jaiku, Pownce, Twitter, Social|Media, Threads, Friendfeed, and more — where the social dimension is people interaction in (potentially large scale) open discourse via the ‘follower|following’ model, and without recourse necessarily to blogs.

Once a person begins to experience the dissociation of blogs and commentary — once commentary moves to these streams away from blog comments — it seems odd to go back. Like using a computer that is not connected to the web.

My hypothesis is that people will find it most natural to have the most active conversation where the flow feels fastest: meaning, where there are many people so that any given topic or link creates a great deal of commentary in relatively short order. However, this is an added incentive to comment directly in the streaming app like Digg, friendfeed or Twitter.

There is a cost to leaving behind the community of commenters on a blog, but if a core group defect en masse to some flow app, that community can remain largely entact, with even the blogger coming along.

For example, someone I follow on Twitter posts a Tweet with a link “someone Wrote a new post on XYZ topic. See www.tinyurl.com/y78YD889Ww.” My natural reaction is to click it, and then write a comment in Twitter to @someone, like ‘@someone have you considered the writing of Borges?”

And from the point of view of Twitter use, I am keeping the covenant: I received the message there, so I respond there. But from the blog-centric view, I am breaking faith, since I read the post that was published there, but I am merely treating the blog as a repository for posts, not a centerpoint for community.

So we are seeing a second wave of defection that defines a new era in social media. We defected from traditional mainstream media, where they broadcasted to us as passive members of an audience. And now we are defecting from the Web 1.0 model of social media, where the blog publisher hold all the power, and the world is a feudal patchwork of blog-based communities. We are moving into a era of flow, where blog posts will just another bit of conversation streaming in the flow.

Since there are dozens — and perhaps soon hundreds — of these streaming apps, each with different although overlapping communities, what can we expect in the near, medium and long term?

Near term — tower of Babel as more people find value in one or more streaming conversational tools, the conversation — the third space — will be subdivided ten times over. And less and less conversation will happen on blogs.

Medium term — Comment tools — like Disqus and IntenseDebate — provide a way to pull the commentary from streaming apps back onto the blog posts. I write a blog post, and a handful of people comment on it at Social|Medium (for example), and my future commenting solution would display those comments as if they had been made on my blog, along with comments from other streaming tools and native one from my blog. This fills a gap in the plumbing, but doesn’t really change the experience of people in each of the streams, since they will only see a subset of the total comments.

How does this feel from the perspective of the individual at the micro level?

Once you adopt the flow attention model, things change.

Here’s my desktop, or one part of it. I usually work with my laptop plugged into a 30” monitor, and I do my ‘work’ — blog posts, email, writing, reading — on the 30” monitor. I keep the laptop screen for flow apps. Here, from left to right I have Snackr (an RSS newsticker app), Twhirl (a twitter client), Friendfeed’s RealTime Beta, and Flickr’s new Activity Stream. Another stream I use is Backpack’s Journal.

I am not saying that everyone is going to become me, but the flow model — where pertinent information is filtered by my contacts and finds its way to me instead of me finding it — is simply better, simpler, and less time consuming, so long as you can make the shift from manual to automatic transmission.

In the long term the static inequalities in blogging make it hard to fit into the coming web of flow. We need a world in which comments, posts, bookmarks, and recommendations are really different aspects of the same thing. Why have we devised a web where posts and comments are so different? Or so different from a bookmark?

All of these ideas share core principles: a person authors a post, comment or bookmark. It is created in some context, probably represented by other open windows on the screen or selections on those screens. (For example, a ‘bookmark’ is the storing of a URL for a page, plus a title, a note, and some tags. The same information could be created as a blog post, right? Or a comment is based on some open post or comment, and has a link to the post or comment. And so on.)

In today’s world, the URL associated with anything created on the web is in effect it’s unique ID, but it is a physical location not just a logical handle. Imagine a web in which the physical location of things was simply an archive, a place to access the definitive data and metadata, but otherwise was used principally as a unique identifier. Imagine all these sorts of conversational particles bouncing around the world through a gazillion streaming apps that use various sorts of social and algorithmic models to order, filter, and aggregate the particles in various ways. These bits could flit from one streaming app to another, or users could create a post in one, a comment in another, and see them come together in a post+comment form in a third. In this world, the bits float around and are experienced in the flow apps, and people might never go back to the original URL associated with the bits.

Also in this world, the bits might be owned by the author, no matter where they are streamed. So if I want to, I can put an ad in every post, bookmark, and comment I make. Likewise, when a service creates new value from aggregation, algorithm, or emergent property of some social sling, they should be able to have their own ads in that context, but they should also honor the ads embedded in the bits. Basically, it’s a fractal world, where ownership is associated with the smallest bits, and larger aggregates. But these can travel — even the aggregates — from one streaming context to another. We will need standards for this, of course, but they will arise to meet demands.

So imagine the following scenario: I post an observation about, say, the future of Web 2.0. A few others see this floating past in various streams and create comments linking to my post, or Digg it — which means it shows up in the Digg stream and several others, like a future friendfeed or social|medium. All of these people add their own ads, and when Digg streams voting results (with associated comments) that has Digg ads embedded. A future version of Techmeme notices that my post is collecting a lot of heat, so it creates a story ‘cluster’ around my post and other posts and comments linking to my post (including in the future, the digg object pointing at my post), and Techmeme drops that into its stream, with an embedded ad, too.

While it might be possible to go down, down, down to the URLs associated with these objects, who ever would? Why leave the flow, where the live things are, to look at the river bottom, where all the dead things fall?

“The Internet doesn’t know what it is doing” - Clay Shirky. It isn’t build to push just one sort of thing around, but all sorts of things.

I think this generation of flow apps will move past a static and deterministic model of what sorts of conversational particles exist, how they should be related to each other, and what apps can show which ones — to a much more fluid model, where all sorts of new associations can be made.

We need to agree on the metaphor of the web of flow — of the bloodstream — and then people can create all sorts of particles that can be streamed through it, and we can get to a more egalitarian social medium that we have now, one that is firmly in the Web of Flow, past the Web of Pages.

There have been a few posts by various folks who heard the talk (here, here, and here), and I think a recording will be posted at http://dogearnation.com/ on Monday or Tuesday. The slideshare.net version can be accessed, and the comments at the Web 2.0 Expo site, too.