The rapid advancement of artificial intelligence technology is profoundly transforming the gaming industry, evolving from early rule-based opponent modeling to the integrated application of deep reinforcement learning, natural language processing, and generative agents. Existing research indicates that deep reinforcement learning demonstrates formidable decision-making capabilities in complex strategy and real-time strategy games, while natural language-driven human-machine interaction and content generation significantly enhance immersion and personalized experiences. In recent years, the emergence of large-scale model technology has propelled research toward Decision AI Models, driving a shift from AI-generated content (AIGC) to AI-driven decision-making (AIGA). This paper aims to review the application and innovative practices of artificial intelligence algorithms in gaming, focusing on breakthroughs in deep reinforcement learning for agent training, the potential of natural language processing in interaction and narrative generation, and the leading role of Decision AI Models in shaping the future gaming ecosystem. Research findings reveal that AI technology not only enhances the intelligence and entertainment value of games but also provides significant opportunities for interdisciplinary research and industrial upgrading.
Liu Yihan (Wed,) studied this question.