Techniques of evolutionary computing have proven useful for a diverse array of fields in science and engineering. Because of the expected low signal to noise ratio of LIGO data and incomplete knowledge of gravitational waveforms, evolutionary computing is an excellent candidate for LIGO data analysis studies. Using the evolutionary computing methods of genetic algorithms and genetic programming, we have developed, as a proof of principle, search algorithms that are effective at finding sine-gaussian signals hidden in noise while maintaining a small false alarm rate. Because we used realistic LIGO noise as a training ground, the algorithms we have evolved should be well suited to detecting signals in actual LIGO data, as well as in simulated noise. These algorithms have continuously improved during the five days of their evolution and are expected to improve further the more they are evolved. The top performing algorithms from generation 100 and 199 were benchmarked using gaussian white noise to illustrate their performance and the improvement over 100 generations.