Capcom Experimenting With Generative AI to Create 'Hundreds of Thousands of Unique Ideas' Needed to Build In-Game Environments

Capcom is exploring the use of generative AI to streamline the creation of in-game assets. The company faces the challenge of generating "hundreds of thousands" of unique design ideas for environmental elements, a process traditionally very time-consuming and labor-intensive.

As game development costs escalate, publishers are increasingly turning to AI tools to enhance efficiency and reduce expenses. This trend has sparked controversy, with instances like Activision's alleged use of AI for Call of Duty cosmetics and loading screens drawing criticism. EA has even declared AI as "the very core" of its operations.

In a recent interview with Google Cloud Japan, Capcom's technical director Kazuki Abe (known for his work on titles like Monster Hunter: World and Exoprimal) detailed the company's AI experimentation. Abe highlighted the significant workload involved in generating diverse designs, even for seemingly simple objects like televisions, each requiring unique logos and shapes. He stated that hundreds of thousands of ideas, including unused ones, are needed.

Abe developed a system leveraging generative AI to process game design documents and automatically generate design proposals. This system, utilizing models like Google Gemini Pro, Gemini Flash, and Imagen, significantly accelerates the process, provides self-feedback for refinement, and improves overall efficiency. Internal testing has yielded positive results. The AI's implementation promises substantial cost reductions and potential quality enhancements compared to manual creation.

Currently, Capcom's AI integration is focused on this specific application, with other crucial aspects of game development – including core gameplay mechanics, programming, character design, and overarching game design – remaining under human control.