Skip navigation
DOI: http://dx.doi.org/10.7551/978-0-262-31709-2-ch037
Pages 248-249
First published 2 September 2013

Synthetic signalling protocell networks as models of neural computation

Matthew Egbert, Gerd Gruenert, Gabi Escuela, Peter Dittrich

Abstract

Modern conventional computers are programmable, predictable and relatively easy to understand and engineered—at least compared to most complex non-linear systems. These properties are the result of various dynamical constraints that are universal to conventional computers, such as the clock mechanism that synchronises the update of logic gates and other components; the ubiquitous discretization steps (where continuous values are discretized into binary 1s and 0s); and the almost complete isolation of internal processes of computers from the environment of the computer. We are investigating an alternative computational medium composed of signalling synthetic protocells to explore the implications of relaxing some of these dynamical constraints that are typical of conventional computers. Is it possible to build useful and/or programmable computers out of unconventional media such as protocells that do not have a synchronizing clock? Or that do not employ a conventional representation of 0s and 1s? Or that are less decoupled from their environment?