Explainability via Responsibility
This addresses the need for explainable AI in co-creative game design systems, though it appears incremental as it builds on existing PCGML methods.
The paper tackles the problem of black-box machine learning models in procedural content generation for games, which are hard for human designers to understand and debug, especially in co-creative systems. It presents an approach using training instances as explanations for AI agent actions, evaluating it by approximating how well it helps users understand and cooperate with the AI.
Procedural Content Generation via Machine Learning (PCGML) refers to a group of methods for creating game content (e.g. platformer levels, game maps, etc.) using machine learning models. PCGML approaches rely on black box models, which can be difficult to understand and debug by human designers who do not have expert knowledge about machine learning. This can be even more tricky in co-creative systems where human designers must interact with AI agents to generate game content. In this paper we present an approach to explainable artificial intelligence in which certain training instances are offered to human users as an explanation for the AI agent's actions during a co-creation process. We evaluate this approach by approximating its ability to provide human users with the explanations of AI agent's actions and helping them to more efficiently cooperate with the AI agent.