Beyond Boole and Shannon

Boole and Shannon never studied the physics of computation. Obviously Boole simply could not have, as none of the required physics was even known in his day, and Shannon was nearing the end of his career when such considerations were just beginning. And yet, both Boole's algebra and Shannon's information concepts to make many of our calculations. This chapter touches on how fundamental physics—the uncertainty principle from quantum mechanics, and thermodynamics, for example—constrain what is possible, in principle, for the computers of the far future. It argues that while there are indeed finite limitations, present-day technology falls so far short of those limits that there will be good employment for computer technologists for a very long time to come.

EMBODIED COMPUTATION: APPLYING THE PHYSICS OF COMPUTATION TO ARTIFICIAL MORPHOGENESIS

We discuss the problem of assembling complex physical systems that are structured from the nanoscale up through the macroscale, and argue that embryological morphogenesis provides a good model of how this can be accomplished. Morphogenesis (whether natural or artificial) is an example of embodied computation, which exploits physical processes for computational ends, or performs computations for their physical effects. Examples of embodied computation in natural morphogenesis can be found at many levels, from allosteric proteins, which perform simple embodied computations, up through cells, which act to create tissues with specific patterns, compositions, and forms. We outline a notation for describing morphogenetic programs and illustrate its use with two examples: simple diffusion and the assembly of a simple spine with attachment points for legs. While much research remains to be done — at the simulation level before we attempt physical implementations — our results to date show how we may implement the fundamental processes of morphogenesis as a practical application of embodied computation at the nano- and microscale.