Skip to content

Social Architecting and the Narrows

March 3, 2011

As of a few weeks ago, the Web Ecology Project concluded an experiment called Socialbots, essentially an event in which three teams competed to program bots to enter and influence large groups of users on Twitter over a two week, all-out battle of automated social shaping. We’re still sifting through all the data generated, though the code from the competition has been made available open source, and we’ve been talking about the project to a variety of different folks — trying to think through the implications of what’s come out of it all. But, wanted to drop out this post talking about how it all turned out, and addressing the next project that we’re planning to take on given those results (and looking for people who might be interested in collaborating with us).

The results: tremendously exciting! With only two weeks of coding time, the three competing teams were able to develop bots that, even following rudimentary patterns of behavior, were able to elicit an enormous amount of activity in the social cluster. The winning team alone built 107 mutual connections between its bot and the targets, and elicited close to 200 responses (@ replies, RTs, etc). In all, the bots collectively generated close to 250 responses, and received mutual connections from close to half of the entire target set. And the bots also had a strong effect on the topology of the social graph as well — in the two weeks, the bots were able to heavily shape and distort the structure of the network. This included bringing people together not originally connected, and bringing together a community of activity around the bots themselves (the picture above was the final end-state network graph in the game).

The ultimate “so-what” of this? Beyond just competitions, it opens the possibility of building a class of technologies that could be used to do targeted social shaping on a very large scale. Essentially, Socialbots demonstrates that proof of concept. To that end, swarms of bots with statistically-predictable social outcomes could be built and used to actively sculpt and rewire the connections of social groups online consisting of thousands (or perhaps hundreds of thousands of users).

So we’re setting our sights a bit higher now. What we’re working on now is the “The Narrows,” the first ever robo-constructed social superstructure leveraging and extending the technology from Socialbots to really engage in building mega-scale community architecture. That’s pretty abstract, but the idea behind the project is simple and concrete: we’re going to survey and identify two sites of 5,000-person unconnected Twitter communities, and over a six-to-twelve month period use waves of bots to thread and rivet those clusters together into a directly connected social bridge between those two formerly independent groups. The bot-driven social “scaffolding” will then be dropped away, completing the bridge, with swarms of bots being launched to maintain the superstructure as needed.

In any case, we’re getting a team together to embark on this project and form a social architecture team, the first construction project of a group that we’re calling “Pacific Social.” If you’re interested in playing a role, drop me a line at!

13 Comments leave one →
  1. March 27, 2011 11:39 pm

    Dear Tim — fascinating research! I’m following the evolution of the internet and information sharing/manipulation myself (who isn’t these days) and think this is a powerful idea but that it could easily lead to denigration of “truth” if used for, say, political motives (for example, If you don’t mind my asking, for what purpose will your company use this technology (good, evil, other)? Would you say this technique is somewhat analogous to “Google Bombing” for social networks instead of search engines?
    Thanks, -Arun

  2. Bob Timmins permalink
    March 28, 2011 8:01 pm

    Very clever but will ultimately be used for evil and will result in the demise of social systems if there is the possibility of non-human responses or traffic generation. Thus leading to ….. SPAM

    This is an automated response… #IdeaFail

  3. March 29, 2011 4:32 pm

    Perhaps if they are used for evil, then the integrity of the web as we know it will be changed – people will have to go out of their way to not ‘silo’ themselves, get their information from a variety of sources, and make their own informed decisions rather than ‘going with the herd’ which has long been manipulated by traditional media.

  4. March 30, 2011 10:46 am

    It’s hard to imagine any worthwhile use for this, seems primarily good for spam. Clever as it surely is. Also, Twitter is not society.

  5. October 28, 2011 3:31 pm

    This is exciting. I was just conversing with a friend of mine, josh walker, about away to prevent the “end of history”, by embedding the social layer, incorporating networks of trust, into the fabric of the web itself. By the “end of history”, we were positing a time in which sophisticated AI could easily replicate the form, content, and links of web pages, and other sources of information, to create a near endless variation of sources of “Truth”, thereby making it near impossible to ascertain “true history” from “false history.”

    Embedding a social layer into the web, would give priority over “histories” that were already associated with trust networks, however, if bot networks can gain our trust, then this would render this method noneffective.

  6. October 28, 2011 4:25 pm

    thanks for this article if you are interessted in programming your own bot in processing I just released a real time twitter gateway (fo open sound control) on github:
    ~ works now with oauth
    ~ has a settings interface
    ~ authenticate to YOUR twitter account
    ~ track words, or user id and get a real time stream of tweets
    ~ made with processing so simple to extend (see

  7. October 28, 2011 5:23 pm

    can you put me in the next target set? i want to experience it 🙂


  1. Done SocialBots… What’s next?… The Narrows « @AeroFade's Blog
  2. Turing tests en Twitter « PLAN H
  3. MediaBerkman » Blog Archive » Radio Berkman 177: Retweeting Robots
  4. Is anyone using bots for social engineering yet? - Quora
  5. Competição de socialbots: quem influencia mais? | Caos Ordenado
  6. Weekly List Bookmarks (weekly) | Eccentric Eclectica @

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

  • Twitter Updates

  • Elsewhere!

  • Pages

  • March 2011
    M T W T F S S
  • Archives

  • Meta

  • %d bloggers like this: