Swift-Hohenberg Model of Liquid Artificial Humans Mimicking Human Graded Reactions

Abstract

Humans and liquid systems have in common that they are both complex, self-organizing systems that produce under certain circumstances spatio-temporal patterns. The current study exploits this cross-disciplinary point of view and studies human graded reactions in terms of continuous confidence ratings from a pattern formation perspective. The wellknown V-shaped ratings are modelled by means of amplitude equations that are generic for a plenitude of pattern formation systems. Experimental data from three independent research groups are shown to fit simple, two-parametric amplitude equation models with goodness of fit measures of 90 percent. Importantly, the Swift-Hohenberg model describing the emergence of roll patterns in liquids is solved to show that under appropriate conditions the amplitudes of emerging roll patterns produce the same type of V-shaped functions as observed in humans. A new research avenue is opened for liquid artificial humans in terms of liquid systems that mimic human reactions.