Emergent Signs in GCN

Emergent Signs in GCN

Signaling is used by living systems, both social and biological, to coordinate lower level entities into emergent upper level systems. The GCN uses this signaling as a self-organizing and open-ended mechanism to create an improving multiresolutional mechanistic simulation of the human body. By "self organization," we mean that lower level entities react to each other without centralized guidance until their arrangement forms something that lasts longer than the individual entities. By "open endedness," we mean that they can continue to improve the realism of their mimicry of living systems, and by "emergence," we mean that something new emerges from the entities' relationships.

The fact that coordination happens because of signals helps open-ended self-organization in four ways. First, by making signs that are slipable, "fuzzy," or open to different meanings. This makes it possible for new ideas to be found and accepted before they are spelled out more clearly. Agent AI "tabula rasa"'s that haven't made up their minds yet and are still flexible, when put into a society with less flexible agents, are better able to interpret signs in ways that are useful to society in new ways. Second, agents act on other agents based on how they see their roles, not as individuals. This means that relationships that work well have a network-coordinating effect, not just a private one. The first (slipable) way is flexible, so many different kinds of agents can fit into any one role. Third, signs create a functional space, or culture, that shows new agents how to do the things that have worked for agents in the past.This learnable space accumulates past innovations as a scaffold for agents while at the same time being open to new innovation.

Fourth, interpretable, implicit signs can become explicit signs with precise instructions once they have become so ingrained and certain in a system that they no longer need to be guessed, freeing up implicit understanding for innovation. These are four endogenous open endedness mechanisms that make emergent systems receptive to exogenous open endedness mechanisms such as the availability of many different challenges to shape the system. Through the data absorption technique, we will investigate how multiple challenges can grow a data-driven complex adaptive system simulation given these mechanisms.

GCN's signaling dynamics can be viewed as a software emulation of the process of social emergence of symbols from subsymbols via symbolic interaction. Smolensky developed the idea of the emergence of symbols from subsymbols in the context of neural networks,⁴ using an example of a neural network that could detect if a room is a kitchen based on whether it had items in it like a stove or a refrigerator, etc. His point was that a concept like "kitchen" is a statistical entity derived from many instances of kitchens, which cannot be expressed in strict rules because any single thing in a kitchen could be absent, but the room would still be considered a kitchen.

⁴ Smolensky, P. (1988) On the Proper Treatment of Connectionism. Behavioral and Brain Sciences 11:1-23. Available at: https://home.csulb.edu/~cwallis/382/readings/482/smolensky.proper.treat.pdf