From Animal Crossing to Call of Duty, AI plays a major role in almost all modern video games. Shaun Grech, CTO and Co-Founder with Games Champions, discusses how the latest advancements in AI have changed gaming in recent years, and explores how the technology might influence the games we play in the future.
When it comes to designing an immersive, engaging, fully interactive world that gamers can truly lose themselves in, artificial intelligence (AI) is one of the best tools that developers have in their arsenal. Whether it’s designing a virtual world that feels incredibly realistic, or seamlessly adjusting the difficulty of the gameplay to better suit the player’s skill level, AI plays a vital role in almost all major video games produced today.
AI has become such an essential part of the gaming experience that we often don’t even notice it’s there until something goes wrong. We saw this happen back in 2020, after the release of the much-hyped Cyberpunk 2077: the graphics were stunning, the concept was ambitious, and expectations for the game were sky high. But, it didn’t take long for the cracks to start to appear. In the days after the game’s release, the internet was abuzz with players talking about the glitchy non-playable characters, buggy enemies, and vehicles seeming to appear and disappear at random. Although critics praised the ambitious concept, they were quick to note the many technical glitches, and the game received very mixed reviews (Metacritic).
Ultimately, a bug-riddled AI meant that the strong storyline and immersive world failed to mesh together — and it left players cold. The lesson we learned from the fiasco was clear: if the AI isn’t up to scratch, it can make it harder for the player to suspend their disbelief and truly get lost in the world of the game.
Clearly, strong AI is needed if a game is to succeed. But, how are developers currently using and advancing the technology, and how could it change the games we’re playing in future? Here, I’ll take a closer look at how advances in AI are pushing the boundaries in gaming.
Immersive, interactive virtual worlds
One of the most notable advancements that AI has facilitated in recent years has been in terms of how players interact with the virtual in-game environment. In particular, AI has greatly advanced the non-playable characters (NPCs) that populate many games — think of the townspeople in an RPG, or the pedestrians in an open-world city environment.
As AI has become more and more advanced, developers have been able to make the behaviour of these NPCs increasingly naturalistic and complex. AI can be used to dictate how these characters move and behave, and how they react when the player engages with them. Tools such as dynamic pathfinding mean that NPCs can enter and exit areas as they follow their own pre-set storylines, making interactions feel more unpredictable, fresh, and exciting. As a result, the virtual world can be more immersive, realistic, and interactive than ever before.
One notable example of this has been Rockstar’s modern classic, Red Dead Redemption 2. The game’s open-world wild west environment is populated by a cast of thousands of NPCs, all of whom seem to have a storyline and motivations of their own. More importantly, they can also react to changes in the game world as they happen. This makes the setting feels incredibly real — for example, if you start a gunfight in the street, NPCs will shoot back at you or try to run for cover, depending on their programming. It all feels very fresh and alive.
This complex use of AI also greatly increases the number of potential storylines and side quests available to the player, with a virtually inexhaustible amount of unique playable scenarios and interactions to discover. When you look at this way, it’s clear to see that games that can use complex AI can offer much greater value to the player.
While NPCs are the most visible form of AI in most modern games, it has other uses that are a bit less obvious, but equally important. One application of AI that we’re seeing more and more nowadays is in player experience modelling. This is a new type of programming where the game is able to collect data about the gamer’s ability and skill level as they play. It can then automatically adjust the gameplay to suit their ability, resulting in a more satisfying gaming experience.
This process is more commonly known as dynamic difficulty balancing, and it has been used in countless games. Naughty Dog’s Crash Bandicoot 2, released in 1997, is a famous early example: the game would take note of any sections in levels where the player was struggling, and then adapt the gameplay to help make it easier for them. For instance, the game might introduce more frequent checkpoints, provide extra power-ups, or slow down obstacles to help the player progress. It just goes to show how AI can help developers to create games that will please a broader audience.
Adding value with procedural generation
Developers can now create algorithms that will generate unique experiences for the player during gameplay, like a fresh level, map, or items, allowing players to experience something new every time they play. This AI is known as procedural generation and, so far, many games have successfully used it to keep players coming back for more, including Minecraft, No Man’s Sky, and Borderlands. The AI can also be used to generate new outcomes for quests in MMOs, to help ensure that gameplay doesn’t become repetitive.
It’s possible that this form of AI will become increasingly common in gaming. These days, lots of people argue that many blockbuster next-gen video games are overpriced, because don’t offer enough play time to justify rising prices. But, with procedural generation, the possibilities are endless, so the player will never run out of things to do. If developers want to ensure that gamers feel they’re getting the most bang for their buck, then this type of AI could be the way forward.
Better matchmaking for exciting multiplayer gaming
Another major application of AI is in multiplayer video game tournaments, where it can be used to ensure a level playing field. Matchmaking — the process where players are selected to compete against each other — uses AI to work out which players are most evenly matched in terms of skills, taking into account their previous performance and amount of experience. This way, newer players are less likely to compete against someone playing at a much higher level, ensuring a more satisfying and fun experience for everyone involved.
Here at Game Champions, we use a type of AI that studies a player’s behavioural patterns to help make sure each player is evenly balanced. For example, for our FIFA matchmaking platform, we use a machine learning algorithm to assess the player’s form, rating, and game history, which is then used to train our AI models. In our Champion’s Draft, each player is then assigned a rating to help balance the match, ensuring an advanced player is always assigned a weaker team when playing someone with less experience or skill. This helps keeps gameplay fun, fair, and fresh for all players.
AI has countless applications in gaming, and new possibilities are opening up every day. To keep up with the latest developments in the world of video games, check out Tech on the Go’s gaming hub, where you’ll find news about the latest releases and technologies.
One of the Editors for Tech on the Go, I love covering mobile tech and social media. I also manage the reviews we cover so all the writers stay on my good side!