The Challenge of Producing Human-Competitive Results by Means of Genetic and Evolutionary Computation

Human-competitive results include those equivalent to new scientific results published in peer-reviewed scientific journals, solutions to long-standing or indisputably difficult problems, patented inventions, and results that tie or beat human contestants in regulated competitions. We argue that the pursuit of human-competitive results is not only a worthy goal in itself, but a useful compass for guiding the future growth of the field. We say this for reasons of utility, objectivity, complexity, and interminability. We believe that the continuing generation of evermore important human-competitive results relies on progress in three areas of research: multiobjective optimization, parallel computing, and the development and perfection of competent genetic and evolutionary search methods. Addressing the characteristics of human-competitive problems is one way to expand the theoretical underpinnings of the field of genetic and evolutionary computation.