Our study has shown that genetic programming is an effective way to evolve hive intelligence. Our tests lend themselves to many avenues for research in the field of computer science. Though fairly comprehensive, our testing could be extended beyond the scope of our project. For instance, the graph of program length versus the number of generations had begun to exhibit a downward trend perhaps suggesting that over time the programs might get shorter while maintaining or increasing the fitness. Also, the ghosts’ use of pheromones was just one of many ways for them to communicate. What if we gave the ghosts the ability to instantaneously communicate or limited them to touch-only communication? If we left the maze the same for each generation, would the ghosts adapt to the specific terrain and take advantage of its peculiarities? If we changed the maze, how would this affect the fitness of the ghosts and how long would they take to adapt to their new surroundings? In general our programs mimic many principles of survival of the fittest and natural selection. Perhaps increasingly complicated simulations could be used to simulate biological processes in the body or the effects of pollution on a population of organisms.

     There are many real world applications as well. In fact, the only limit is our imaginations. For instance, many possibilities exist in the field of robotics. In terms of exploration, there are many places that are far too dangerous or hazardous for a human to venture. Instead, robots could be deployed to explore the unfamiliar terrain. Deep-sea probes running computer-evolved routines could roam the bottom of the sea at much deeper levels than the best human carrying subs. In outer space, many planets have extreme temperatures or gravities that would make it impossible for humans to explore. Yet, robots running evolved routines could explore these planets for us. None of these probes would need to be directly controlled by a human. Instead they would follow a series of basic commands preset by the user. They would prioritize and accomplish these tasks all on their own, while transmitting their findings as they explored. Some military applications could include espionage probes working in teams to map out buildings, to locate individuals, and to monitor transmissions. In the near future, genetic programming might lead to great advances in the generation of artificial intelligence and the systems that use such routines.