In behavioural ecology the `interference function' relates the intake rate of a focal forager to the density of conspecifics, and has been (and remains) the subject of extensive theoretical and empirical activity. This paper describes a novel agent-based model of interference characterised by the use of genetic algorithms to evolve foraging behaviours in a spatially explicit environment. This model provides a flexible platform for modelling interference phenomena, and addresses several conceptual and practical problems with orthodox approaches. Its flexibility is demonstrated by an exploration of how, in the context of interference, some instances of individual irrational behaviour can be understood as a consequence of adaptation to a group context.