Interactive digital storytelling stands at a transformative frontier as advanced AI technologies become viable for integration into game design. This paper introduces ‘Neuro-Game Design’, a paradigm that leverages neural network-based artificial intelligence to co-create dynamic, adaptive gameplay experiences, reimagining the relationship between human designers, players, and intelligent agents. Drawing from the author’s past research, we present the ‘Ludic Framework’ as a complementary foundation emphasizing interaction-signs (ludics) and affective design principles. Traditional game design has relied on deterministic mechanics and the “illusion of intelligence” in game AI, which constrained the medium’s potential for genuine emergent storytelling. By integrating modern AI systems with the Ludic Framework’s semiotic approach, Neuro-Game Design enables game worlds that interpret player actions and learn and adapt in real time, facilitating co-authored narrative experiences between human and non-human agents. This integration offers new opportunities for interactive digital narratives that transcend predefined branching story trees, embracing the emergent complexity arising from intelligent systems working in concert with human creativity. Importantly, we distinguish Generative AI (neural models that create new content, e.g. GPT text generation) from Predictive/Reactive AI (models that learn decision-making or anticipate player behavior, e.g. reinforcement learners like AlphaGo). This clearer terminology underpins our discussion. We also address how Neuro-Game Design differs from existing emergent narrative approaches, engage with knowledge representation frameworks (e.g. knowledge graphs and hybrid authoring systems), and consider practical constraints such as scalability and real-time performance. We distinguish generative vs. predictive neural methods, situate Neuro-Game Design among emergent-narrative approaches, integrate recent knowledge-representation work, and address practical constraints.