A large amount of the layouts and designs created by The Department of Notation are realized with the help of specialized component-based machines that create random variations in size, position and orientation. Because of the random nature of these tools, finding a desirable design can take a long amount of time. In order to expedite this part of our process we have been developing a toolset based on two sub-fields of artificial intelligence – neural nets and genetic algorithms. This presentation will cover the development and use of our first two tools – NurtureNet and GeneticAesthetic. NurtureNet is an implementation of a back-propagation neural net that is trained by the user to rate the results of our design machines. Once it is properly trained, it can be “shown” the results of one of our “machines” and then assign it a degree of perceived beauty. GeneticAesthetic allows the user to define a the input of one of our machines as a genome. The user can then treat instances of the machines like organisms and subject a population to the rules of evolution. It is possible to create random populations, breed the most fit organisms, and even apply random mutations. A trained NurtureNet can even be used to help quantify the fitness of organisms.