Collective Cognition — Quantifying Distributed Inference: Difference between revisions

organized by Jessica Flack, Bryan Daniels, Chris Ellison, and Philip Poon (Wisconsin Institute for Discovery)

A living organism is a reflection of its past environment: When it is able to sense environmental regularities over evolutionary or developmental time, an organism and its behavior can be viewed as a hypothesis about environmental states that it will encounter in the future. In viewing the organism and its behavior as a prediction, we can recast adaptation and cognition as two parts of a broader process of inductive inference. Furthermore, when aggregates of adaptive organisms generate regularities at new temporal and spatial scales, this may allow for collective cognition: The new scales provide structure for further interactions and make them more predictable, reducing social uncertainty and facilitating adaptation at the aggregate level. In this workshop, we will consider two broad questions: (1) How can adaptation and cognition be quantitatively understood in terms of statistical inference and information theory? (2) Can we use such a quantitative framework to understand coordinated aggregates, explaining how components with only partially overlapping interests may join to form a more predictive whole? Is there any insight to be gained from thinking about collective behavior as a computation?