Ideas Challenge Winners

 

Organization

Congratulations to our Ideas Challenge winners in the Organization track! These applicants explore novel theoretical approaches or acquisition of data around self-organization at one or more levels of scale. Learn more about them and their projects below.

	Jonathan Askonas Catholic University of America	From the microbiome to the global ecosystem, diverse entities interact to create symbiotic relationships and virtuous cycles beyond the comprehension or narrow self-interest of any of them. This project asks what the scientific study of ordering might look like. It explores how we might assess the environmental knowledge which the structure of an order encodes, whether it’s a metabolism, a liturgy, or a culture.
	Anna Guerrero Marine Biological Laboratory	A “cyborg cell,” or an organic cell with a synthetic, magnetically manipulable cytoskeleton, will help researchers identify how value-state recognition occurs in living systems by enabling 1) empirical separation of organelle structure from chemosensory signals without loss of maintenance response; 2) identification of which cellular components participate in recognizing a “beneficial” environment in order to maintain life at multiple scales; 3) the manifestation of cellular agency in juxtaposition to human agency. Researchers will observe chemotaxis in a normal cell, cell-mediated chemotaxis in a cyborg cell, and human-mediated chemotaxis in a cyborg cell. Beyond enabling empirical identification of cellular components associated with value-state determination, a cyborg cell represents a novel system for studying organizational and scalar hierarchies within complex living systems.
	Elliot Nelson IBM	The free energy principle (FEP) is a theoretical framework which draws on non-equilibrium thermodynamics, statistical physics, and machine learning, and aims to explain the high-level goal-oriented behavior of self-maintaining, self-organizing adaptive systems, not least biological brains. In recent years, advances in machine learning have led to powerful neural network based reinforcement learning algorithms, capable of modeling increasingly complex goal-oriented systems such as those governed by the FEP. We propose to apply these tools towards the study of FEP-based adaptation in simulated environments, with the goal of understanding the range of adaptive behaviors, cognitive capabilities, and evolutionary dynamics that can emerge from free energy minimization in either digital or biological settings.