Do we need a theory in the Era of Massive Data Flow?

Massive Data Flow (MDF) is everywhere these days; from data about neural cells, social insects and genetic networks, to Lifelog (digital storage of a person’s visual and audio life log) and SNS (Social network service) data streams. Current web and device technology has made it possible for us to record detailed and massive data flows of artificial and real living systems. But how can we analyze and understand MDF? Can a simple toy model based on a plausible narrative and simulation still tell us something? Concepts like “the edge of chaos” and “self-organized criticality” once helped us to understand living systems, but we do not know whether the same concepts can be useful to MDF. I think studies of artificial life in MDF need larger models, because we need the strength of models that overcomes MDF. Possible larger models do not have to mimic existing living creatures but can be larger, in the sense of novel invention and utilization of space and time. In other words, to understand the complexity of MDF is to recast and reconfigure it into a larger artificial model. Indeed, I myself made a large model called “MTM” (Mind Time Machine) in 2010 that ran for three months in an open space, receiving massive visual data from the environment with 15 cameras, processed by internal neural dynamics with a learning capability, and showing sustainable complex adaptive dynamics. We need a theory to make large artificial life models and to take them out into the real world.