A genetic algorithm is a problem-solving method that borrows its strategy from natural selection. Instead of reasoning out a single answer, it starts with a population of many candidate solutions, scores each by how well it solves the problem, and then breeds the next generation from the best performers. It mixes pieces of good solutions together, a step called crossover, and occasionally makes small random changes, called mutation. Repeat this over many generations and the population tends to drift toward better and better answers, even when no one knows in advance what a good answer looks like.
The field was founded by John Holland, whose 1975 book “Adaptation in Natural and Artificial Systems,” published by the University of Michigan Press and reissued by MIT Press in 1992, laid out the theory of how such evolutionary search works and why it can be effective. Holland’s insight was that evolution is, in effect, a powerful general-purpose optimizer, and that its mechanics could be turned into an algorithm for any problem where solutions can be encoded, combined, and scored.
Why a business reader might care: genetic algorithms shine on messy optimization problems that have no clean formula, scheduling, routing, layout, antenna or part design, tuning the settings of a complex system. Because they only need a way to score candidates, not a smooth mathematical structure, they can tackle problems where conventional methods stall. They belong to a broader family called evolutionary computation, which remains a practical tool in engineering and operations research.
The honest limits are speed and certainty. Evolving a population over many generations can take a lot of computation, and a genetic algorithm offers no guarantee that the answer it returns is the best possible, only that it is usually good. For problems that do have efficient exact or gradient-based solutions, those methods are typically faster and more reliable. Genetic algorithms earn their keep precisely on the hard, irregular problems where nothing else fits well.