metaca

with apologies^* to douglas hofstadter, melanie mitchell, and james marshall

This code creates "meta-" cellular automata spacetime diagrams. That is, the CA's evolve their own rules locally, rather than evolving an abstract state according to one global rule. A MetaCA can be coupled (in one or both directions) with a standard CA.

Previous work with cellular automata feature evolving global rules, including the recent work on PICARD by Theodore Pavlic et al. The current work is the first example that I am aware of that evolves the evolutionary rule locally.

In addition to these 1D examples, our paper (and slides) presented at AISB'15 consider 2D examples. The code for those experiments is in ewenmaclean/2dCA.

Citation

Corneli, J. and Maclean, E. 2015. The search for computational intelligence. In: Erden, Yasemin J., Giovagnoli, Raffaela, and Dodig-Crnkovic, Gordana, eds., Social Aspects of Cognition and Computing Symposium, Proc. Annual Convention of the Society for the Study of Artificial Intelligence and Simulation of Behaviour (SSAISB), University of Kent, Canterbury, UK, 20-22nd April 2015.

*: See also

• Douglas Hofstadter, "Prologemena to any future Metacat", in Fluid concepts and creative analogies, pp. 307 - 318, 1995

• Melanie Mitchell, Peter T. Hraber, and James P. Crutchfield. Revisiting the edge of chaos: Evolving cellular automata to perform computations. Complex Systems, 7:89--130, 1993.

• James B. Marshall, Metacat: A Self-Watching Cognitive Architecture for Analogy-Making and High-Level Perception, (PhD thesis) Indiana University, Bloomington 1999.

• Pavlic, Theodore P. and Adams, Alyssa M. and Davies, Paul C.W. and Walker, Sara Imari, Self-referencing cellular automata: A model of the evolution of information control in biological systems In: Artificial Life 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems, The MIT Press, 522--529, 2014. (arXiv:1405.4070)