Within artificial intelligence, evolutionary algorithms refer to a set of programming methods that draw inspiration from concepts stemming from evolutionary biology. More specifically, they are algorithms capable of selecting the most appropriate solution (individual) from a large set (population) through the evaluation of its fitness (how well it adapts to the problem, the environment). Evolution thus takes place through the repetition of such selection.
The use of evolutionary algorithms has roots in the 1950s and has spread through many fields, from engineering, the financial market, chemistry, mathematics, and data mining. Nowadays there’s a great variety in evolutionary algorithms, ranging from simple genetic algorithms (seeking the solution through recombination or mutation) to neuroevolutionary algorithms (where the “genomes” are represented artificial neural networks). They differ mainly in the amount of populations in use and the operators responsible for introducing change.
It seems possible that with enough computer power we would be able to produce a superintelligence using a sufficiently complex evolutionary algorithm. This method does not require the understanding of intelligence needed to create an AGI nor the scanning equipment needed to create a WBE.
Further Reading & References
- Genetic Algorithms and Evolutionary Computation by Adam Marczyk
- Evolutionary Algorithms 1 Introduction by Hartmut Pohlheim
- Evolutionary Computation Repository
- Bäck, T., Fogel, D., Michalewicz, Z. (1997), Handbook of Evolutionary Computation, Oxford Univ. Press.
- Yang X.-S., (2010), "Nature-Inspired Metaheuristic Algorithms", 2nd Edition, Luniver Press