Towards Declarative Content Generation for Creativity Support Tools

Procedural content generation (PCG) represents a class of computational techniques in the context of games, art, graphics, and simulation, for algorithmically producing novel digital artifacts more typically handcrafted by artists and designers, such as planetary terrain, texture assets, level map layouts, and game quests and stories. There is growing interest in PCG for its potential to enable creativity support tools (CST): software that can assist and proactively collaborate with human practitioners to iterate on generative algorithms towards a creative design goal. While declarative programming tools, especially logic programming and solvers, offer many appealing benefits for PCG, there remain barriers to their adoption. In this paper, we focus on the specific challenge of inspecting execution: the important role in PCG of being able to visualize a generative algorithm's intermediate states and see them as meaningful.