The software used many concepts originally observed in biology. For example, Genetic Algorithm (GA).
Artificial life (AL) reveals / uses several principles of biology, such as resistance to imperfect code fragments , addressing by content , imperfect reproduction (in some implementations, as well as sexual, i.e. multi-orginanisms-driven, reproduction) and target function utility . An interesting result of AL is the spontaneous production of macroelements observed in areas such as ecology or epidemiology (domains highly dependent on biology), such as the appearance of parasites and even organisms that use parasites or subtle predators, prey relationships.
Perhaps we can say that the software has become a "full circle" with some experiments in the calculations, which include real (based on carbon) DNA (or RNA) molecules! An original experiment in this field (PDF link) by Professor Alderman (RSA fame), who encoded various elements of the graphics problem (gram of the Hamiltonian) with various DNA molecules, and let the massive parallel computing power of biochemistry do the rest and solve the problem!
In the digital world, but with strong inspiration from biology and, indeed, from the anatomy of the cerebral cortex and many theoretical and clinical observations in the field of neuroscience, Neural Networks (NN). In the field of NN, perhaps deserves special attention, the Numenta strong> hierarchical temporal model model , which, although it reproduces [the understanding that we have] the neocortex very weakly, introduces the idea that the same algorithm is used in all areas and at all levels of the cognitive process fed by the brain, the idea is largely supported by biological, anatomical and other forms of evidence.
mjv
source share