Combining peer production, social networking and P2P approaches productively to filter for high quality in- formation requires fundamental theoretical and technological advances. Theoretically, we need to develop new and realistic models of user behaviour and the dynamic social structures that they create (i.e. new social the- ory). It is clear that quality can, in some way, emerge from these behaviours and structures. Yet individual behaviour and social structures co-evolve, and constrain and enable each other. It is not sufficient simply to model individual behaviour; one needs multi-level models and methods. Technologically speaking, we must design and deploy large-scale systems that self-organise (without central control) in generally low trust, open environments. Such systems need to protect users from unscrupulous, malicious, erroneous or self-interested behaviour that aims to dilute quality.
It is a fundamental premise of this project that by combining the above approaches, we can make major advances because:
A key aim of our project therefore is to support the creation and dissemination of quality artefacts by facilitating the emergence and self-recognition of communities that create them and consume them. Such communities share, by definition, some agreement on desirable content and quality within the domain and hence have in- centives to share resources to gain access to quality artefacts. We term such communities Quality Collectives (QLectives). One general function of QLectives is to operate as a form of “social filter” which, through the wisdom of the collective, can rate the quality of artefacts (e.g. scientific or media contents).
More generally, QLectives can be defined as cohesive and cooperative resource sharing communities di- rected towards the peer production of commonly defined high quality artefacts, services and experiences.
Our vision of future ICT is a global ecology of self-organising QLectives supporting diverse services and communities - moving away from the “one size fits all” approach and into the “long tail”. Current Web2.0 and P2P systems can be seen as constrained manifestations of a movement in this direction. Only by combining and extending these promising approaches can our vision be realised.
In order to achieve this, we propose the application, generalisation and extension of methods that have been successful within the natural sciences, to advance the social scientific understanding and engineering of QLectives. These methods involve: formulating models, theories and hypotheses; designing empirical experiments to test the hypotheses; carrying out experiments with suitable control and measurement infrastructures, and finally completing the loop by modifying, refining or validating models and theories based on the results of the experiments. For our experiments we will use “living labs”.