Genetic algorithms work via survival of the fittest. A set of algorithms are created to solve a problem. 

The most successful ones are carried over to the next generation and crossed over to create better versions of the algorithm.