Players navigating through the 18,700+ levels of Candy Crush Saga might not realize that artificial intelligence (AI) plays a significant role in designing these puzzles. The Swedish game developer, King, uses AI to help create and refresh levels, keeping players engaged by constantly offering new challenges. The AI helps streamline the process, especially when updating old levels, making sure players don’t get bored or frustrated. Todd Green, general manager of Candy Crush, mentioned that without AI, it would be nearly impossible for designers to update so many levels efficiently.
Within the gaming industry, AI is seen as both a tool to assist with mundane tasks and a way to enhance game development. While some game creators embrace AI for improving interactive non-player characters or creating more immersive worlds, others fear it might threaten jobs, particularly in areas like acting or game design. Concerns about AI in gaming were part of the reason why game performers went on strike in July.
Green emphasized that AI isn’t about replacing workers or creating AI-driven gameplay but rather about improving workflows behind the scenes. It helps designers work more quickly and accurately, allowing them to focus on higher-level tasks. He noted that AI assists in making updates to Candy Crush levels faster, reducing the manual workload for designers.
In 2024, consumer spending on video games in the U.S. rose to $51.3 billion, with mobile games accounting for half of that. Candy Crush, which first launched on Facebook in 2012, continues to update regularly, with its 300th client version recently released. Activision Blizzard acquired King in 2016 for $5.9 billion.
Joost Van Dreunen, author of “One Up: Creativity, Competition, and the Global Business of Video Games,” pointed out that Candy Crush’s massive player base and decades-long popularity make it ideal for using AI to create new content on a large scale. AI is particularly helpful for rapidly generating new puzzle boards, which is key to keeping players engaged with fresh challenges.
King uses AI in two main ways: generating new levels and revisiting older ones to keep them relevant. For experienced players, especially those who might have taken breaks, the goal is to make sure the puzzles are fun right away, a challenge when there’s little player feedback before the level is released. Green described AI as a behind-the-scenes assistant, rather than a tool for direct player interaction.
AI also enables King to streamline the process of improving multiple levels at once. Instead of working on a few hundred levels each week, AI allows them to refine thousands of levels, making the development cycle more efficient. Player feedback, both quantitative and qualitative, is key to determining how well the levels are received, such as the pass rate and whether a board needs to be reshuffled.
Ultimately, AI in Candy Crush is used to refine game mechanics and ensure players have a satisfying experience, while also speeding up the level design process.