Description: This post explains the interesting ways that game developers are using AI in their games and how it will affect the game industry.
AI, AI Everywhere
So if you’ve been on the internet for long enough, then you’ll that there’s an active discourse surrounding a new piece on coding technology called AI. AI (for those who don’t know) stands for artificial intelligence, and it thinks for you. Think of it as a more powerful search engine, with the ability to get any answer to any question. That means that it can do more complex tasks that people don’t want to do. It can even simulate and imitate human emotions (which is a whole other can of worms that I will not be touching today).
People have been using AI for stuff like coding, making art, and even solving complex math problems. However, this has created some controversy surrounding this topic, as it can act as a replacement for people. Companies are already researching ways for more complex jobs to be taken over by AI machines and people are outraged. I also briefly mentioned that AI can create art, and while that is technically true, it traces other people’s art and the art it makes is… well let’s just say it’s not the best. However, it is getting better, which is making people worry.
However, there are ways that people are using AI responsibly (specifically in video games because it’s in the title). Game devs are using AI in ways that were previously not even thought of before, and many of these ideas are incredible and terrifying. So in this blog post, we’ll be talking about a few of these games and how future game devs can use them in their games (while also going over other responsible ways for non-coders to use AI).
Rain World
Now while AI is blowing up with all the advancements that people are making, its has been around for a while, and it’s been used in video games to simulate behavior patterns in NPCs and mobs before. One perfect example of this is in the video game Rain World. It’s one of my favorite games and I’ve been wanting to talk about it forever on this website.
Rain World is a survival game where you play as a slugcat, the guy grabbing the broken pole you see above the text. Your goal is to make it back to your family by getting out of an abandoned urban area, left by whatever sentient life that was originally here. However, you are a tiny creature in a huge, and I mean HUGE world. Predators will attack on sight and will even fight each other if they are hungry enough. And then there’s the rain, which isn’t like the rain we have in our world. If you get caught in this rain you just die, meaning that finding shelter before the day ends is important.
Rain World’s brutality is amplified by its unpredictable AI. Creatures in the game move with surprising realism because their AI controls their bodies, including their joints. Unlike typical games, where code automates actions like walking or fighting, Rain World’s AI must manage survival, navigation, and hunting for food. Below is a visualizer showing how the AI operates.
Imagine the yellow creature is a lizard and the white blob is the player. Both can see the room layout, but the lizard only spots the player within its field of vision (shown by the red line). The lizard plans a route, crawls through tunnels, and attempts to catch the player. If successful, it takes the player to its den to eat and stay alive.
That’s the other thing, while no two creatures in Rain World are the same, they all have one goal, to survive. Every creature (including the player) has a hunger meter that they have to refill so that they don’t die or suffer the consequences of starvation. This makes the entire world feel like its own little ecosystem. Not only that, but each creature has its own ways of movement, its attack patterns, and is given a random personality to create a level of unpredictability in the game.
If you are interested in the AI of Rain World and want to learn more, I’ll leave a link to this video, which goes into more about how it works and how the devs trained the creatures to survive. Maybe I’ll gush about Rain World in its blog post, but I’ve talked about it enough in this one, let’s move on to another game.
Minecraft
When you think of Minecraft, you may not think about AI, but Minecraft has actually been using an interesting type of AI in their games for years now. It’s how you get a unique world every time you play, making the game fresh and fun to explore. It’s also a common type used to create random biomes and places. It’s called Procedural World Generation.
Procedural World Generation, or PWG for short, is an AI structure tool that is used to create a constant or random level every time you start a new game or enter a new room. Minecraft uses PWG to create an entire world meaning that no 2 worlds are ever exactly alike. This also leads to players find once in a lifetime locations, creating a desire to explore every new world. “But how does it work?” I hear you saying (yes I can hear you, don’t question it).
Well, Minecraft specifically uses something called a seed system. “Whenever the game has to generate a new world, it calls upon an algorithm known as Perlin noise. This algorithm outputs a pseudo-random value that is then used to determine the characteristics and features of the world. However, the algorithm always outputs the same value each time for a constant starting point (seed). Thus, the same seed generates the same terrain every time.”
In other words, while every world in Minecraft is randomly generated, the world generation has to abide by certain rules to make sure that key locations show up in the world. If these locations didn’t show up, the game would be unprogressable. So the AI memorizes those structures and places them around the world for the player of explore and find them.
If you want to learn more about PWG, and how the seed system in Minecraft works, I recommend going to the link above to chack out the Minecraft wiki. It has a lot of information from both experienced players and past devs, so I highly recommend it.
Suck Up
For this last one, we’re going to go into the more touchy subjects surrounding AI and why people in the games industry are worried about more and more companies using AI. Suck Up is a game where you play as a vampire who has to convince his neighbors to let them into their homes. It’s a relatively funny game about wearing different disguises and talking to different people, and it makes for some fun interactions.
But the game has a dialog system that we don’t really see used much right now in games, and it may be used more and more as time goes on. Suck Up uses a ChatGPT model to allow the player to directly communicate to the NPCs. Simply type what you want to say and hit enter, and the AI will think of a way to respond to the situation. The AI will also make a note of what your character is wearing and point it out from time to time, forcing you to have to pretend to be someone.
This made the game blow up online as one of the funniest games to come out. However, the implications of the tools used in this game speak for themselves. As I’ve mentioned before, companies are trying to incorporate stuff like this into more of what they do, and that includes video games. Ubisoft has already unveiled that they’re using a similar AI to Suck Up to create their story for them, they also expressed their interest in using AI for other things. So this AI could potentially take jobs away from programmers and writers in the game industry.
Conclusion
So, what do we do about AI? Honestly, there’s not much we can do about AI right now, but we can make a note about how to use AI. All of the games that I mentioned only used AI for certain aspects of their game and didn’t completely rely on it to make the entire game. While AI is a useful tool, we have to remember that it is a tool, not a replacement. Welp, that’s all from me today, see ya next time!
