Abstract
Artificial broadcasts from extraterrestrial intelligences are a hypothetical class of celestial phenomena. Unlike known astrophysical objects, the societies that generate them may be able to replicate on galactic scales through interstellar travel. Different galaxies could thus have drastically different populations, with variations in abundance of many orders of magnitude. I present a probabilistic formalism to treat this shared history, in which societies and their broadcasts are described by distributions over basic properties like lifespan and energy released. The framework contains a hierarchy of objects related by a tree structure. Discrete societies, the sources of broadcasts, are organized into potentially interstellar metasocieties. The population of each type of object is represented by a random point process in an abstract parameter hyperspace, a haystack. When a selection like an observation draws a sample, the point process is thinned. Given assumptions of interchangeability and independence, observables are modeled with compound Poisson random variables. I present an example of how selection bias can favor sampling longer-lived objects. I rederive the Drake equation for societies in the limit of no expansion. When interstellar replication is present, however, the mean number of detected broadcasts can depend quadratically on stellar mass, suggesting a search strategy favoring large galaxies.
To read the full paper, click here.