“Habits as sensorimotor life-forms: modelling self-maintaining behaviour with an iterant deformable sensorimotor medium”, Tuesday 18th June, 11am, Carlos Santamaría Building, B14.
Date and Time: 11am, Tuesday 18th June, Carlos Santamaría Building, B14
Title: Habits as sensorimotor life-forms: modelling self-maintaining behaviour with an iterant deformable sensorimotor medium
Abstract: Artificial Life has not yet explored in depth the analogy between life and mind that is hidden in the nature of habits: their self-sustaining dissipative structure as ecological sensorimotor entities. We present a new dynamical model for habits implementing what we call a node-based “iterant deformable sensorimotor medium” (IDSM). The IDSM has properties designed such that trajectories taken through state space increase the likelihood that in the future, similar trajectories will be taken. We couple the IDSM to sensors and motors of a simulated body in a simulated environment and show that under certain conditions, the IDSM resonates with the other parts of the simulation, forming self-maintaining patterns of activity operating over the IDSM, the body, and the environment. These patterns of activity are similar in many respects to habits, patterns of activity that are self-reinforced. We present various environments and the resulting ‘habits’ that form in them, studying the sensorimotor coordination patterns that stabilize in the process. We discuss how this model and extensions of it can help us understand and model self-sustaining patterns of behaviour as building blocks for a theory of cognition that does no rely on representations
Date and Time: 11am, Tuesday 18th June, Carlos Santamaría Building, B14
Title: Habits as sensorimotor life-forms: modelling self-maintaining behaviour with an iterant deformable sensorimotor medium
Abstract: Artificial Life has not yet explored in depth the analogy between life and mind that is hidden in the nature of habits: their self-sustaining dissipative structure as ecological sensorimotor entities. We present a new dynamical model for habits implementing what we call a node-based “iterant deformable sensorimotor medium” (IDSM). The IDSM has properties designed such that trajectories taken through state space increase the likelihood that in the future, similar trajectories will be taken. We couple the IDSM to sensors and motors of a simulated body in a simulated environment and show that under certain conditions, the IDSM resonates with the other parts of the simulation, forming self-maintaining patterns of activity operating over the IDSM, the body, and the environment. These patterns of activity are similar in many respects to habits, patterns of activity that are self-reinforced. We present various environments and the resulting ‘habits’ that form in them, studying the sensorimotor coordination patterns that stabilize in the process. We discuss how this model and extensions of it can help us understand and model self-sustaining patterns of behaviour as building blocks for a theory of cognition that does no rely on representations
Abstract
Artificial Life has not yet explored in depth the analogy between life and mind that is hidden in the nature of habits: their self-sustaining dissipative structure as ecological sensorimotor entities. We present a new dynamical model for habits implementing what we call a node-based “iterant deformable sensorimotor medium” (IDSM). The IDSM has properties designed such that trajectories taken through state space increase the likelihood that in the future, similar trajectories will be taken. We couple the IDSM to sensors and motors of a simulated body in a simulated environment and show that under certain conditions, the IDSM resonates with the other parts of the simulation, forming self-maintaining patterns of activity operating over the IDSM, the body, and the environment. These patterns of activity are similar in many respects to habits, patterns of activity that are self-reinforced. We present various environments and the resulting ‘habits’ that form in them, studying the sensorimotor coordination patterns that stabilize in the process. We discuss how this model and extensions of it can help us understand and model self-sustaining patterns of behaviour as building blocks for a theory of cognition that does no rely on representations