The past year has seen several Sikkim Game that showcase the potential of this technology. Both Ubisoft’s NEO and GoodAI’s AI People are examples of playable experiences that could become more commonplace in the years to come, should business models and technologies reach a point of maturity.

The main thrust of these games is to utilise contemporary AI technologies, notably Large Language Models, and fit them into traditional game AI architectures to make non-player characters (NPCs) more believable and interactive. The resulting NPCs are hunters that can stalk players, observing their behaviour and searching for signs of their presence such as audible and physical disturbances in the environment such as footsteps or a twig snapping underfoot.

From Pong to GPT: A History of AI in Games

NPCs are also able to learn from their actions and exhibit emotions, which helps to increase their realism and engagement levels. They can also react to player actions – for example, by adjusting the difficulty level to provide a balanced experience tailored to individual skill levels and making NPCs more aggressive or docile depending on the player’s performance.

Other uses of AI in games include reducing development time and streamlining production, freeing up designers to focus on the things that matter most. As Shigeru Miyamoto, creator of Mario, once said: “A delayed game is eventually good, but a rushed one is forever bad.” AI can help deliver on this promise by automating repetitive tasks and enabling developers to create a more diverse and dynamic world with ever-changing gameplay experiences.