Overcoming scale domain gaps
Reflections on an agent-based model and its plugin for human agents
Due to the epistemological gap between the behaviours produced by an agent-based model and its known initial settings, these models are often referred to as black boxes. Building on an experimental design process and theories of non-scalability and embodied knowledge, the paper presented in the colloquium maps out aspects of this epistemological disconnect and speculates about possibilities to overcome them.
The design underlying the discussion was created in the framework of the IXDM project Thinking Toys for Commoning, which looks at ways how toys, games and play make complex commoning processes tangible and negotiable. One of the outcomes of the project is an agent-based model (ABM) addressing labour that is invested by members of the community to maintain the commons. An experimental physical computing setup attached to the model – the ‘Human agent ABM plugin’ – explores how interacting with computer agents as a human agent affords for a better grasp of the ABM’s inner life.
Designing and playing with the ‘Human agent ABM plugin’ most prominently allowed for reflections about the temporal scale domain gap (between the fast computer agents and the slow human agent) which will constitute the central aspect of the analysis.