Stigmergy — auto-organization art
Collective Intelligence Inspirational Notes
Stigmergy is one of the meta-heuristic techniques or, in other words, algorithms that define a pattern of behavior used for performing a function or tackling a challenge. It can be mainly found in collective intelligence of the animal kingdom (ant colonies, schools of fish), but also in recent times in questions of algorithms for computer and robotic systems, in machine learning as well as web communities and human societies. Stigmergy embraces models and phenomena of emergent (biological) self-organization of (eco) systems.
The resulting structures and functionalities amaze us by the perceptual, physical and cognitive complexity of the capacities far exceeding the ones of participating organisms.
Thanks to the marking of the environment through signs (unintentional in the case of stigmergy, versus intentional signs that lead to more complex dynamics, which is the case, for instance, of language evolution), it leads to social communication driven by positive or negative feedback. Stigmergy makes it easier to prioritize solutions, in case of demanding situations, but also in the events of opportunities to be seized. Having clear priorities makes it easier to focus efforts on actions with the most options for success.
In order to function, stigmergy techniques are based on simple rules and local information, without the need for central coordination or supervision. These phenomena also have a high degree of robustness and resilience, as some individuals resist malfunction or behavioral deviations.
Examples of stigmergic self-organization are:
• aggregation (schools of fish),
• grouping and classification (ant colony: grains of sand, eggs, dirty dishes, collecting isolated objects, leaving clustered ones),
• Architectural construction of wasp and termite nests (where discrete stigmergic communication is used, because the animal changes its specific construction behaviors genetically encoded and through stimulus-response (discrete building rules) based on discrete perception patterns),
• collective selection of routes to reach the food
• division of work with gently specialized individuals (who can perform different roles) or strongly specialized (with the possibility of carrying out only specific activities). Examples of division of labor could be between queen bee, etc. To explain the change (or not) of the task, the response threshold (or not) model is applied. These models are widely used in video games, strategy, especially.
As we can appreciate, stigmergy is an inspiring coordination mechanism, decentralised and highly synergetic, which can be used for collective intelligence fostering.