Accessibility Skip to Global Navigation Skip to Local Navigation Skip to Content Skip to Search Skip to Site Map Menu

CSIS Seminar by Professor Jeremy Pitt

Friday, 31 March 2017

Speaker: Professor Jeremy Pitt, Intelligent Systems & Networks Group, Department of Electrical & Electronic Engineering, Imperial College London

Title: Interactional Justice and Knowledge Management in Open Systems

Location: Owheo G34 Owheo Building

Date & Time: Friday 31 March, 1:00 pm

Abstract:
This talk will present work (in progress) on the computational characterisation of interactional justice. Interactional justice is a user-centric aspect of justice required for realising requirements of fairness and inclusivity in self-organising rule-based open systems. We start from ideas due to Ashby about quasi-stability and Rawls about a well-ordered society, to give a definition of "rule quasi-stable". We then describe the implementation of a multi-agent system simulation required to bring about resource allocation in an economy of scarcity, and show that using opinion formation with respect to a social network and shared values (i.e. interactional justice) can deliver a system which is rule-quasi stable, whereas the same system which depended only on individual metrics would collapse. This will be used as a basis for a broader discussion of knowledge management in self-governing socio-technical systems, in particular addressing the fundamental design dilemma: on the one hand, the system must be sufficiently unrestricted (fluid, flexible), in order to enable a collective with a shared set of congruent values to achieve their joint purpose(s) in collective actions situations; on the other hand, the system must be sufficiently restricted (stable, robust) to prevent a subset of the group from exploiting this openness ‘against itself’, as it were, by usurping control of the system and running it for the benefit of their own sectional (un-shared) interests.This talk will present work (in progress) on the computational characterisation of interactional justice. Interactional justice is a user-centric aspect of justice required for realising requirements of fairness and inclusivity in self-organising rule-based open systems. We start from ideas due to Ashby about quasi-stability and Rawls about a well-ordered society, to give a definition of "rule quasi-stable". We then describe the implementation of a multi-agent system simulation required to bring about resource allocation in an economy of scarcity, and show that using opinion formation with respect to a social network and shared values (i.e. interactional justice) can deliver a system which is rule-quasi stable, whereas the same system which depended only on individual metrics would collapse. This will be used as a basis for a broader discussion of knowledge management in self-governing socio-technical systems, in particular addressing the fundamental design dilemma: on the one hand, the system must be sufficiently unrestricted (fluid, flexible), in order to enable a collective with a shared set of congruent values to achieve their joint purpose(s) in collective actions situations; on the other hand, the system must be sufficiently restricted (stable, robust) to prevent a subset of the group from exploiting this openness ‘against itself’, as it were, by usurping control of the system and running it for the benefit of their own sectional (un-shared) interests.