Whenever you hear about a video game, you typically think of entertainment or a mere way to pass the time. However, video games go much deeper than that, as they are facing the frontier of artificial intelligence head-on. If you have ever played a fairly new video game, then chances are it has used artificial intelligence before! We see AI being utilized in almost every aspect of video games, from controlling non-player characters (NPCs) to generating infinitely diverse terrain.

A widely used algorithm for many video games to make their NPCs seem intelligent is the Finite State Machine, or FSM. Basically, this algorithm generates each possible outcome an NPC could encounter, and then determines a way to react to each outcome. Therefore, the NPC always has a way to react to a given situation and seems to be able to “think” before it acts accordingly (2). The drawback to this method, however, is that it becomes fairly easy to notice the patterns. The player then can learn to counteract its predetermined reactions. And that is the key—learning. Researchers constantly strive to make their video game AIs capable of learning from past decisions and outcomes. With recent breakthroughs in machine learning, such as the victory of Google’s AlphaGo program over the best Go player in the world, it is fairly apparent that computers dominate the realm of video games (4). As a result, artificial intelligence is not only being utilized for the gameplay itself, but for other facets of video games.

Computer scientists at Italy’s Politecnico di Milano created an algorithm called the Generative Adversarial Network to generate levels in the first-person shooter video game, DOOM. They fed this network thousands of sample levels from hundreds of developers and hobbyists. Basically, the neural net utilizes two artificial intelligence systems to to mimic man-made levels. While DOOM is a fairly old and simple game, the researchers are hoping to “scale” this technology up to be able to generate more involved levels for complex games. If the project is a success, it could drastically reduce the workload needed for video games, and consequently their cost (1). Another researcher from the University of Falmouth created an AI named Angelina, capable of creating hundreds of experimental video games from scratch. His program initiates the next step of creating an algorithm capable of thinking ideas humans have never thought of before. Not only that, but the popular game of No Man’s Sky utilized hundreds of algorithms to create a boundless world with infinite amounts of creatures: most of which the creators never saw before! The use of AI in video games to create the unimaginable is only growing by the minute (5).

Nowadays you may be scratching your head wondering how new video games are becoming so realistic. Thanks to artificial intelligence, video games can have improved graphics and more life-like physics. Different algorithms can analyze thousands of hours of people walking, running, and jumping, and incorporate that into video games to allow NPCs to mimic the fluid motion of human behavior. Several algorithms can also predict the layout of a person’s face and can project an actual person’s face onto a digital player (3). With the advent of virtual reality and augmented reality (popularized by the success of Pokémon Go), this AI becomes powerful when being able to incorporate a person’s everyday life into these video games.

If you asked anybody a few years ago, they would probably never expect video games to become such a melting pot of science, art, and technology. With an ever-growing number of gamers and increasing competition in the game market, developers are pushing the boundaries of video game capabilities. They are in an arms race against the consumer, and as a result, video games have become a field where different AIs and algorithms can be tested and analyzed. With that, video games have the potential to become the very artificial intelligence we have been seeking for so long.


(1) AI Designed DOOM levels so good they seem man-made. (2018). Futurism. Retrieved 1

July 2018, from https://futurism.com/artificial-intelligence-video-games/

(2) AI in Video Games: Toward a More Intelligent Game – Science in the News. (2017).

Science in the News. Retrieved 1 July 2018, from

http://sitn.hms.harvard.edu/flash/2017/ai-video-games-toward-intelligent-game/

(3) Boyle, E. (2018). Game on! How AI is transforming video games forever. TechRadar.

Retrieved 1 July 2018, from https://www.techradar.com/news/game-on-how-ai-is-

transforming-video-games-forever

(4) Eastwood, G. (2018). How video game AI is changing the world. Network World.

Retrieved 1 July 2018, from https://www.networkworld.com/article/3160106/artificial-

intelligence/how-video-game-ai-is-changing-the-world.html

(5) Parkin, S. (2017, November 30). AI is dreaming up new kinds of video games.

Retrieved from https://www.technologyreview.com/s/609482/ai-is-dreaming-up-new-

kinds-of-video-games/

0 comments on “The Blurred Line Between Video Games and AI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: