Do self-evolving systems operate holorithmically rather than algorithmically?

Whereas an algorithm is a procedure specifying the next step, given the present or prior steps, a holorithm is the outcome of the entirety of the relationships exerting an influence within a system.

Algorithmic and holorithmic processing are nearly equivalent if the number of relationships is small and the algorithmic processor is fast.

Holorithmic processing takes place at the cost of space rather than time. (Can computation take place spatially rather than linearly?) The information in holorithms is implicit within the relationships among the points, whereas information in an algorithm is implicit within the phase space explored under the algorithm over time.

Protein conformation "computed" by amino acid sequence is an example of holorithmic information processing. The entirety of the steric interactions among the constituent atoms in the sequence, the solvent, and so on, all result in a particular set of emergent characteristics.

The sequence that results in a given set characteristics would take an exceedingly long time to predict algorithmically, yet the results are inherent almost instantly within the universe of interactions taking place as the protein forms.

Holorithmic computation evolves through interaction rather than through linear construction, in the same way that the grooves in a phonograph are shaped by the context of vibrations of a needle rather than by being calculated in advance.

Michael Webb, 2002

home ]