Capcom talks about using generative AI in game development

Capcom's technical director reveals how the company plans to use generative AI to create thousands of ideas to speed up game development.

author-image
Mayank
Updated On
New Update
Resident Evil 4

Capcom game development with AI (Photograph:Capcom)

Capcom is using generative AI to improve productivity. Using generative AI the company is exploring different ideas for game development. AI has been quite a controversial topic in the gaming community due to ethical reasons, however, Capcom may be using generative AI in a bit different way than other companies.

Advertisment

In a new interview with Google Cloud Japan, Kazuki Abe, a technical director at Capcom talked about the process. Abe said that coming up with ideas is one of the most time-consuming tasks as each game needs to have different item models to make them unique. This is where AI can come in to help the team quickly come up with new ideas for things they have in mind.

“One of the most time-consuming and labour-intensive aspects of game development is coming up with ideas for creating the game's worldview. For example, if there is a television in the game space, we can't just use a real-life product,” he said.

Further explaining that each idea requires illustrations, texts, and other things which can take a lot of time to create. This is where Abe saw the use of generative AI. Although, due to its controversial nature, it’s hard to tell what fans will think of this. The main goal here seems to be using the reference created using generative AI to then make their assets. Rather than using the AI to create assets or other gameplay elements.

Also Read:Capcom brings huge discount on Okami HD with the announcement of a sequel

Rising costs and revenue goals leading toward AI

The increasing production costs of video games are skyrocketing and companies need to reduce the cost and time somewhere. As per Abe, cutting development costs and time from the start is a great way. For their current plans, he developed a prototype system using Gemini Pro, Gemini Flash, and Imagen.

“We are still in trial operation, but we are pleased with the progress we have made in being able to output a large number of diverse ideas quickly, and the evaluation of the development team that has actually tried it out has been excellent,” Abe said.