GAs are less likely than other learning techniques to get stuck at a local maxima because mutations and crossover can cause to them to search entirely new areas

GAs are less likely than other learning techniques to get stuck at a local maxima because mutations and crossover can cause to them to search entirely new areas in the space (unlike ANNs for example).