The ever-increasing capabilities of high- performance optical chips, combined with a continuous trend towards miniaturization has made designing them a more and more complex task. A further difficulty is added by the fact that technology in the optics sector is not as well developed as that in standard microelectronics, which means designers of optical chips are bound by more technological constraints and restrictions. Furthermore, the design tools at their disposal are rudimentary at best and far away from the high standards of CAD tools used for microelectronics design. To alleviate this problem to some extent, a combination of genetic algorithms and neural networks has been used successfully to develop a tool, which is able to design chip parts and simple complete optical chips completely unaided by human interaction from the very first step. The completed designs are considerably more effective than those created by humans, as the artificial intelligence tool is far better able to cope with the complex relationships and boundary conditions that must be taken into account when designing such optical chips.