Jump to content

Social Machines

From Wikimania 2014 • London, United Kingdom


Once upon a time 'machines' were programmed by programmers and used by users. The success of the Web has changed this relationship: we now see configurations of people interacting with content and with each other, blurring the line between computations performed by machine logic and algorithms, and those that result from input by humans, arising from their own psychological processes and life experience. Rather than drawing a line through such Web-based systems to separate the human and digital parts (as computer science has traditionally done), we can now draw a line around them and treat each such compound as a `social machine', a machine in which the two aspects are seamlessly interwoven.[1]

In the field of Social Machines, and across a whole cluster of related fields including Human Computer Interaction, User Experience, Media Theory, Communications Theory, Social Networks, Virtual Communities, Complex Systems, Economics and Psychology, people are studying how these constructs function. Wikipedia is one of the largest and most visible examples, what can we learn from it? How can such systems be designed and engineered, and to what ends? Can a review of the field help us improve the way that Wikipedia itself functions?

Sir Nigel Shadbolt

Barbican Hall: Friday 10:00 – 10:30

“Society will change because of three developments:

Open Innovation – allows any one to participate and contribute to the generation of new insights, content and knowledge. It challenges those who have a vested interest in closed data and methods, restricted markets and knowledge;

Linked Data – the next generation will have at their disposal a web of linked data. They will be able to find things much more easily, leading to the rapid emergence of new ideas, products and services, disrupting existing business models;

Collective Intelligence – the web connects people, computers and data to produce systems much more powerful than the component parts. These systems will be essential to solve the challenges humanity faces, from climate change to public health, social inequality to crisis management. We need to understand how to conceive, design and maintain these systems.”

Sir Nigel Shadbolt is a Professor of Artificial Intelligence at the University of Southampton and the principal investigator in the Theory and Practice of Social Machines project. He is also Chairman and Co-Founder of the Open Data Institute, which launched in December 2012 to develop the publishing of and demand for open data.

Follow Sir Nigel Shadbolt on Twitter and explore the Theory and Practice of Social Machines project

Marc A. Pelletier

Barbican Hall: Saturday 10:00 – 10:30

“Internet has changed humanity; information propagates at the speed of light, unfiltered, from every corner of the world and our perspective is forever altered as a result. It's a revolution of the same magnitude as agriculture and the printing press, and the consequence will be just as profound − our societies hardly recognizable a century hence. I don't know what the future brings, but it's exciting and I'm loving every minute of it."

Marc works with the Engineering and Product Development team of the Wikimedia Foundation as an operations engineer.

He has been an active volunteer editor since 2006 on the English Wikipedia where he have served as administrator and arbitrator. He was also involved with the technical and procedural aspects of automated editing (so-called bots), having written and operated a copyright-violation detection bot himself for several years.

In his professional life, he has been a Unix system administrator and occasional computer science instructor for 20 years, in fields ranging from telecommunication and robotics to game development.

Marc's User Page

Raph Koster

Barbican Hall: Saturday 10:30 – 11:00

"Noise is any pattern we don't understand. If we perceive something as noise, it's most likely a failure of ourselves, not a failure of the universe."

Raph Koster is a pioneering designer of virtual environments. He was lead designer of Ultima Online, Chief Creative Officer at Sony Entertainment and is founder and president of Metaplace, a platform for virtual online worlds.

Follow Raph Koster on Twitter

Yaneer Bar-Yam

Barbican Hall: Saturday 11:30 – 12:30

"We, each of us, are parts of a greater whole. This relationship is shaping and will continue to shape much of our existence. It has implications for our lives as individuals and those of our children. What is generally not recognised is that the relationship between collective global behaviour and the internal structure of human civilisation can be characterised through mathematical concepts that apply to all complex systems.

An analysis based upon these mathematical concepts suggests that human civilisation itself is an organism capable of behaviours that are of greater complexity than those of an individual human being."

Professor Yaneer Bar-Yam is Founding President of the New England Complex Systems Institute. His research focuses on developing complex systems concepts and applying them to diverse areas of scientific inquiry and to major social problems.

Follow Yaneer Bar-Yam on Twitter

Social Machines

Wikipedia is more than a repository of information - it is 'the largest collaborative project of humankind'[2], a constant global conversation mediated by text, images, data and code. It is a unique mode of activity managed and facilitated by the site’s design, a huge interacting network of users with profiles, messaging, walls and newsfeeds, constantly being adapted and improved by users engaged in the conversation, allowing unprecedented levels of knowledge to be processed and shared. Its growth represents a pivotal moment when web design changed the way that we think about knowledge itself, opening both its consumption and its creation to anybody with an internet connection.

This is the core concept behind the idea of a ‘social machine’. Wikipedia and its sister Wikis have long been a pioneer of supporting purposeful human interaction on the internet, empowering its users to collaborate and mutually resolve issues and problems without resorting to remote ‘experts’ in their own and other organisations. Such is the fascination with and exponential growth of the movement, Sociam.org, a specific website run in conjunction with prominent institutes such as the University of Oxford, Edinburgh and Southampton has evolved with the express aim of analysing and understanding the ‘the theory and practices of social machines’. Sociam.org understandably points to Wikipedia as a forerunner of the social machine movement, reflecting on the incentives for participation as varying from ‘reciprocity to social responsibility to altruism’. Such values have long been echoed in the community of Wikipedians. In particular, the concept of educational altruism ties in splendidly with Wikipedia’s stated aim as ensuring that the sum of all human knowledge is readily accessible to all humankind. Social machines have long been paramount to constructive and reflective debate. Sir Tim Berners-Lee, inventor of the World Wide Web, has been a prominent proponent of social machines as a means of debating various systems, allowing us to look at evidence behind claims made as factual statements in politics, democracy and science. Despite this progress, Berners-Lee and other commentators have claimed that today’s interactive applications are very much the early forerunners of social machines owing to their lack of connection. Human communities still remain co-existent on the web and there is a lack of mechanism which works across all domains. This strive for cross-community connection is at the forefront of the thinking of Wikipedia and The Wikimedia Foundation. Ideas such as flow (see below), free culture, open data and Wikidata, all with relevance to social machines will be discussed in detail at Wikimania 2014.

Even though many Wikipedians work hard to welcome new members, there's concern that the community is shrinking. Though there is an ongoing effort to improve its accessibility and enable smoother user journeys, there are many points where users may be unsure of what to do, hindered by an arcane and confusing interface. An open community of designers is working collaboratively to make things better. New uses for Wikimedia’s platforms are also being explored. The work of finding and keeping contributors is multidisciplinary, running the gamut from mediation to metrics analysis. For example, the Growth Team runs experiments to iterate on existing on-boarding processes to continually improve the user journey of an editor.

Below are some examples of the kind of work being undertaken in the Wikimedia movement as part of its mission to enable new levels of digital collaboration. These projects themselves are inherently collaborative, and interested readers are encouraged to get involved. Help to build and test prototypes, give feedback, assist in decision-making and otherwise use your technical and social skills. Anyone can collaborate in the constant construction of Wikipedia’s community.

Community Dynamics

Wikipedia’s challenges and successes are instructive to anybody interested in autonomously organised labour, open democracy and online voting. An extraordinary case study in pseudonymity and semi-autonomous group organisation, the Wikipedia community is both fascinating and notorious. It is both consensus-oriented and collaborative, and in the same time occasionally conflict-driven and fueled by dissent[2].

Users self-organise into Wikiprojects, that enforce standards and highlight areas in need of further work. Administrators are elected by the community to assist its day-to-day functioning, and an elected committee of arbitrators take an active role in settling disputes and discussing policy matters. The organizational structure is, on one hand, extremely egalitarian, participative, and democratic, and on the other sometimes perceived as permeated by power dissonance and inequality[2]. As a result, community management is a core concern of leadership and has a tangible effect on user experience. And yet, the exact mechanisms driving this community (or rather, communities, as different language Wikipedias are run by their own rules) remain poorly understood.

Wikipedia’s community is complex in structure and in practice. Usability initiative studies have shown that the number of contributors is diminishing, and this is partly attributed to a combative mode of discourse common in the community[3]. While users’ strict adherence to policy guidelines is informed by a desire for harmony and consensus, the effect of the legalistic debates about how to apply those guidelines can be intimidating to outsiders not yet familiar with the site’s conventions and jargon[4]. Since policy has been co-created as a result of years of experience optimising the process behind editing articles and resolving disputes, it is important to find ways to give users easy access to the valuable knowledge and methodologies it represents.


Millions of people across the globe are engaged in Wikimedia’s global conversation. This grand-scale knowledge sharing is often carried out on a micro-level, as people work on particular pages that interest them. These small conversations link out to the larger conversation at hand, by way of references to policy, links to user pages and edit histories, and the meta-information provided by wikiprojects that assess and assist with the ongoing refinement of the the sites’ content. At present, that link between the small conversation and the larger one is carried out through the practice of users as they edit talk pages. This requires that new users be socialised into the skills required to maintain those links.

Designers in the Wikimedia movement are working to automate those virtual social practices, so that users can focus on the work that they are passionate about. The core features team at the Wikimedia foundation has undertaken a project called Flow, to modernise communication and collaboration across Wikipedia and other Wikimedia projects. It will implement features that are taken for granted in other parts of the internet, such as automatic threading and automatic signatures, but that have until now been impossible to implement in wikitext. This is hoped to significantly lower the barrier for entry for new contributors, whose first interaction with Wikipedia’s talk pages is often met with a polite correction: ‘please sign your posts’.

Usability studies conducted in 2009 found that the talk pages system presented an intimidating learning curve for users, and lacked basic features that users were expecting from an online communication system, such as a ‘reply’ button and a comment field. Flow is currently in early prototyping stage, and aims to provide these features alongside those required for the smooth functioning of the Wikipedia community, such as links to lists of users’ contributions. It even introduces a button that allows users to ‘thank’ one another.


The Wikipedia Teahouse is a space designed for new editors to introduce themselves, ask questions, and be exposed to friendly veteran Wikipedians. The project has been up since 2012. The Teahouse Pilot Report shows promising results in helping retain new members. Compared to control-group counterparts, new editors who participated in the Teahouse pilot edited 10 times more articles, made 7 times more global edits, and had 2 times more content survive on Wikipedia. Two weeks after joining the Teahouse, 33% of the participants were still editing, compared to 11% in the control group. Overall, 29% of Teahouse participants were women, compared to 9% in the general editor pool.

The Teahouse is a strategic approach to help socialize new editors and to teach appropriate editing practice by providing access to friendly mentors. It is also successful in helping new editors find appropriate peers. Wikipedia is a large place, with editors participating in many different projects with varying degrees of enthusiasm and experience. According to common bond theory, by increasing the possibility that new editors see others that are similar to themselves, the Teahouse may be increasing the likelihood that new editors feel committed to the Wikipedia community.[5] At the same time, the Teahouse expresses a calm design and creates its own set of boundaries with a common symbol system. There is a potential for a sense of community within the Teahouse itself, a community that fosters a space of emotional safety and may encourage a kind of reciprocity that motivates members to edit.[6] After all, if someone took the time to answer my question, shouldn’t I at least put in some time to working on an article?

Projects like the Wikipedia Teahouse can be a good starting point to better learn the social conditions that can help Wikipedia retain and grow its editor base. With a better understanding of how people bond on Wikipedia, both to each other and to the group as a whole, better design decisions could be made to encourage participation and commitment.

The Wikipedia Adventure

Supported by the Individual Grants Program, the Wikipedia Adventure is an interactive game designed to teach new Wikipedians how to edit. As the game designers described it, lessons in community are at the heart of the game. The game lets players practice productive communication, collaboration and cooperation in a simulated and safe environment. It uses principles of game design to encourage a sense of mastery as participants learn the new, unfamiliar task of editing. Compared to a control group of similar new editors not playing the game, the Wikipedia Adventure pilot has shown some success in retaining editors. The pilot reported that new editors who played the game made 1.2 times more edits than the control group. Players were also 1.2 to 1.7 times more likely to make 20+ edits. Among those who played the game, players who completed the last level made 3.2 times more edits than those who only started the first level, and were 2.9 times more likely to make 20+ edits.

WikiProject Council

When thinking about communities as going through a lifecycle, it is often useful to think of phases of community development as the ‘inception’, ‘establishment’, ‘maturity’, and ‘mitosis’ phases. Communities reach maturity when everyone who is interested has joined and activity and membership naturally plateaus. For very large communities similar to Wikipedia, reaching maturity phase may also mean a decrease in activity and membership because the large size makes it increasingly more difficult to maintain a sense of community but very easy to overload members with too much information. That is, unless the community moves to ‘mitosis’ and splits the overall community into distinct sub-communities. Each sub-community can then limit its focus, attract its own members, and continue to foster a sense of community within its purview. [7]

Wikipedia is well into its mitosis phase. The list of examples of sub-communities on Wikipedia is abundant, each with its own interest area, set of processes and sub-culture. For example, there are over 2,000 active projects in the WikiProjects Directory alone. WikiProject Council strives to document best practices for creating and maintaining active WikiProjects. The WikiProject Guide contains extremely helpful information for people looking to work with a WikiProject. However, although it refers to WikiProjects as fundamentally “social” constructs where each WikiProject should be more about the participating editors than the articles created, the guide is primarily focused on giving advice about article scope, talk page design, and policy structure. To be truly effective in its mission, work needs to be done in understanding the social dynamics of WikiProject participants, in mapping the natural life-cycles of WikiProjects, and in teaching would-be coordinators on how to best curate the right social dynamics to properly grow their projects.

Getting involved

Wikimedia can be hard to get involved with - that's the entire problem! But the output of the process is something that benefits us all. Wikimania will be the largest gathering of Wikimedians ever, so there's no better time to come and find out what's being made and how you could get involved.

Register for tickets, or consider volunteering! If you can't wait till August, check out the fringe programme, which has an event specifically on social machines.


  1. Towards a classification framework for social machines Nigel Shadbolt, Daniel Alexander Smith, Elena Simperl, Max Van Kleek, Yang Yang, Wendy Hall, Web and Internet Science Group, University of Southampton, UK
  2. 2.0 2.1 2.2 Jemielniak, Dariusz (2014) Common Knowledge? An Ethnography of Wikipedia, the Largest Collaborative Project of Humankind, Stanford: Stanford University Press
  3. Halfaker, A., Geiger, R. S., Morgan, J., & Riedl, J. (2013). The Rise and Decline of an Open Collaboration Community. How Wikipedia's reaction to sudden popularity is causing its decline, American Behavioral Scientist 57(5), pp. 664-688
  4. Butler, B., Joyce, E. i Pike, J. (2008) Don't look now, but we've created a bureaucracy: the nature and roles of policies and rules in Wikipedia, 26th annual SIGCHI conference on Human factors in computing systems Florence (Italy), ACM. 1101-1110
  5. Ren, Y., Kraut, R., Kiesler, S. Identity and Bond Theories to Understand Design Decisions for Online Communities, Human-Computer Interaction Institute, Carnegie Mellon University
  6. McMillan & Chavis (1986). Psychological Sense of Community: Theory of McMillan & Chavis
  7. Millington, R. "Buzzing Communities: How to Build Bigger, Better, and More Active Online Communities" (2012)