A genetic algorithm is a program that uses Darwinian principles of random mutation to improve itself. The algorithms are lines of computer code that act like living organisms. Different sections of code haphazardly come together, producing programs. Like Darwin's rules of evolution, many chunks of code compete with each other to see which can best perform the desired solution the aim of the program. Some chunks will even become extinct. Those that survive will combine with other survivors and will produce offspring programs

Expert systems can capture and preserve the knowledge of expert specialists, but they may be slow to adapt to change. Neural networks can sift through mountains of data and discover obscure relationships, but if there is too much or too little data they may be ineffective—garbage in, garbage out. Genetic algorithms, by contrast, use endless trial and error to learn from experience—to discard unworkable approaches and grind away at promising approaches with the kind of tireless energy of which humans are incapable.
The awesome power of genetic algorithms has already found applications. Organizers of the Paralympic Games used it to schedule events. LBS Capital Management Fund of Clearwater, Florida, uses it to help pick stocks for a pension fund it manages. In something called the FacePrints project, witnesses use a genetic algorithm to describe and identify criminal suspects. Texas Instruments is drawing on the skills that salmon use to find spawning grounds to produce a genetic algorithm that shipping companies can use to let packages "seek" their own best routes to their destinations. A hybrid expert system-genetic algorithm called Engeneous was used to boost performance in the Boeing jet engine, a feat that involved billions of mind-boggling calculations.

Computer scientists still don't know what kinds of problems genetic algorithms work best on. Still, as one article pointed out, "genetic algorithms have going for them something that no other computer technique does: they have been field-tested, by nature, for 3.5 billion years.