Self-integrating and self-improving systems must be socially sensitive

Abstract

In this talk I will discuss two areas of recent research relating to socially-inspired technical systems. The first of these is socially-sensitive systems design: The proposal for a new way thinking about the design of complex systems that explicitly considers systems' runtime awareness of, and behaviour based on social factors with the potential to affect their operation. Drawing on research in self-organisation and sociology, I will present a conceptual framework for socially-sensitive systems design, comprising aspects related to social values, social relations and social organisation. A key benefit for such systems will be derived from the notion of better positioning through increasing social potential, enabling entities within the system to achieve goals quickly, through the establishment of shared social values, norms and networks. Secondly, I will highlight the role and importance of social learning in complex technical systems where fast integration is needed. Using examples from animal behaviour, I will explore the role of social and asocial learning in self-integration and self-improvement, highlighting that the two can sometimes be in tension with each other. I will conclude by identifying challenges that arise in integrating these two areas, and in doing so propose directions for future research in self-integrating and self-improving systems.