Artificial intelligence (AI) is not a new topic in video games. From Cortana in Halo series to Detroit: Become Human, there are tons of content that tackle the use and issues of AI. But is there only that to artificial intelligence in gaming? AI technology is not just a faint dream of the future, but something that is already being used in modern games. It helps to create a more responsive, adaptive, and challenging gaming experience while also staying hidden from the player.
For a long time, game developers had to manually create any activities and events that NPCs could take place in. Whether that be a street fight between two characters, admiring the landscape, or eating in a local café – each sequence was created by hand. While this is not too challenging, it definitely is time-consuming, especially in open-world games. Now, by implementing AI, game creators can make entire functional ecosystems of NPCs with their own interactions and changing activities. NPCs can finally actually ‘do their own thing’.
Hardly anyone was always satisfied with a recommended route to the objective when playing open-world games. Such algorithms usually greatly struggle to understand how fast players can move on different roads – slim gravel one will make one much slower than wide track one and so on. Meanwhile, with a help of AI and machine learning, developers can create pathfinding algorithms that improve over time, taking into consideration the real player’s speed in certain areas. It might be a small thing, but certainly a quality of gaming experience improvement.
This one is… not too viable today, but we already can see the first steps in the right direction. Nvidia Canvas allows the creation of realistic images of landscapes with ease while AI art generators such as DALL-E attempt to transform words into images. It is still far from creating fully blown-out quests and storylines, but this use of AI seems like a promising way to help create textures and illustrations for games.
It is common in single-player games for NPCs to cheat – spawn in unreachable points, move faster, take sharper turns when racing, etc. This allows an increase of the difficulty of the game without putting too much effort as creating algorithms that will actually control NPCs better is far more challenging. With the help of AI, game developers can create more seamless and interactive ‘cheating’ that suits individual players’ skills better.
While supersampling does not change the game itself, it provides higher frame rates, which can lead to a smoother and thus more enjoyable gaming experience. SS works by rendering the game at a lower resolution, then upscaling it and using AI and machine learning to fill in the missing details. By doing so, the game looks like it was running at a native higher resolution while also having higher frame rates.
For example, AMD FSR (FidelityFX Super Resolution) has had version 2.0 already released which has greatly improved the quality of the upscaled game. It sees the view not only in colors as we do, but also has a sense of depth (provided by the game itself) and motion vectors, which allow it to actually understand what is happening on the screen. It is a new technology that can allow playing games on weaker computers while not sacrificing performance.
Stay Tuned to All The Latest News on Gaming
Interested in learning more? Follow Breach for the latest industry news & don’t forget to Sign up! And be one of the first ones to join Breach MVP, which is rolling out soon!