devlog

Saul suggested that we run a “Open Social Graph” session at next year’s Open Knowledge Foundation conference. Great! I thought. Hang on, what’s an open social graph?

Bread Fitzpatrick sets out some broad assumptions and goals about this open social graph effort. It’s a reaction to user annoyance at having to recreate a “social network” on different trendy online services that rely on a map of contact lists to spread messages about.

I recalled and exhumed a short essay written a couple of years ago and never really aired, about the forces behind the bubble-and-collapse of social network data sharing during the last major effort to do this, using FOAF at the peak of the Great Social Networking Fad of 2003/4.

Why is it happening again now? The userbase of such services must be many thousandfold what it was at that time; there is more momentum to it, more “added value” now. Perhaps the momentum is caused by a lot of people going into these online services from a “mainstream”; they won’t understand the need to have to “link up” in facebook, and linkedin, and dopplr, to get more access to the presence of overlapping sets of people.

But no-one directly *wants* to get this data out into a more “commons” like environment - it’s just the easiest way of bridging around the gaps. A relatively small and vocal group of people want to make this happen, self-interest is flourishing among them as it does in everyone.

Does anyone seriously want “a thousand new social applications to bloom”? I don’t want to be twittered and doppled and jickled to an early death. I already am involved in more email conversations than i really have the cognitive capacity for. I would much rather have less code, than more. And the prospect that an “open” effort raises, of a blooming of “Mom and Pop” social graph based service hosts is, in more than the short term, a pipe dream or a bad joke.

The underlying structural design looks like it tends to agglomeration, and eventual monopoly or oligarchy. There’s only one Wikipedia, one OpenStreetmap, one Flickr, holding an inexorable momentum over any minor adaptation, spinoff or clone. In real-world business, the kind that accompanies these waves of excitement on the internet, we see a cycle of buyout, burnout, merger, consolidation, buyout…

There’s no reason to believe this will stop. Meanwhile, there is space in this debate about the “identity” of shared descriptions and the terms on which they are shared, to rest for a bit, and not go rushing forward.

I’m not saying “sleepwalking into a surveillance society“, as the UK’s Information Commissioner put it a few years ago; but stop and think. What needs is such an effort really trying to address? What place does much of this information have on the Internet, let alone the Web?

Exposing cleanly another layer of information, rendering it that much easier to hook into other information resources… the pressure in the structural design doesn’t go away. We can “democratise” this layer of “social graph” services all we like, but they will go the way of independent ISPs and free web hosting and email providers, unless we find a way of building them that hasn’t involved papering over the cracks exposed in all the underlying layers. In the meantime I, for one, welcome our Google Social Graph.