Innovation

From Blinky to beyond: Inside the evolution of AI in gaming

April 11, 2024 | By Chris Mullen

about In Tech

In Tech is our regular feature highlighting what people are talking about in the world of technology — everything from crypto and NFTs to smart cities and cybersecurity. 

In the evolving landscape of technology and entertainment, the convergence of artificial intelligence and gaming stands as a testament to human ingenuity and innovation. From the early days of simple rule-based systems with the likes of Inky, Blinky, Pinky and Clyde to today's complex neural networks, the journey of AI in gaming has been nothing short of extraordinary.

Pixel pioneers

In the early years, AI functions were rudimentary yet revolutionary, laying the groundwork for future innovations.

When Pac-Man dropped in 1980, it changed the landscape forever. The genius behind this simple maze game? Simple rule-based AI for those colorful ghosts. Each villain had its own playbook: Some chased Pac-Man, others retreated and patrolled their pre-defined spots. For the game developers, it was all about finding the sweet spot for players between fun and frustration. And sure, it may seem basic now, but back when Reagan was in office, it was cutting-edge stuff that had the quarters flowing in arcades, roller rinks and movie theaters worldwide.

But let’s rewind even further. Way before the era of consoles and joysticks, pioneers like Donald Michie were hard at work pushing the boundaries of AI. He started in the 1960s with simple tic-tac-toe machine learning games, and his work eventually led into the complex world of chess. His achievements in machine learning and game theory were like blueprints for understanding how computers could get inside our minds. While limited in scope compared to modern AI chess engines like Deep Blue or AlphaZero, Michie’s work represented a groundbreaking achievement in the application of AI to gaming, paving the way for future advancements in strategic gameplay.

Mortal Kombat’s AI odyssey

By the ‘80s and ‘90s, gaming was even more immersive and in-your-face. In 1992, the fighting game Mortal Kombat became a cultural phenomenon and soon a franchise. Enthusiasts everywhere were glued to their screens, mastering those brutal secret moves.

While the game was incredibly popular in the two-player format, developers had their work cut out for them. How would single-players experience the same euphoria? That’s where AI opponents came in.

In the early days of Mortal Kombat, the developers relied primarily on predetermined moves “taught” to the AI, offering a predictable yet challenging experience. Each AI character had a set sequence of attacks and defensive maneuvers, with varying degrees of difficulty across different settings. While groundbreaking for its time, the AI in early Mortal Kombat titles lacked the complexity and adaptability seen in later iterations and the result was repetitive gameplay and limited strategic depth.

But every fighter (and gaming platform) must evolve, right? Now, decades later, the series has seen a drastic evolution in how the developers employ its AI opponents against human players and boasts more dynamic and adaptive AI systems that respond to player actions in real time. The franchise has even introduced AI fighters that the player can train and allow to battle opponents on their own.

AI’s bone-chilling role in modern horror games

In recent years, AI technology has given smaller game studios the ability to create deeper immersive experiences, and that’s come to the fore especially in the horror game genre. Phasmophobia, released in 2020, has emerged as a groundbreaking example of AI-driven horror gaming. Developed by Kinetic Games, it places players in the shoes of paranormal investigators tasked with exploring haunted locations and identifying supernatural entities. What sets Phasmophobia apart from other games, however, is its innovative use of AI to simulate the behavior of ghosts, resulting in a dynamic and terrifying experience for players.

Each ghost in the game possesses unique traits and behaviors, such as roaming patterns, preferred hunting grounds, and methods of attack, all governed by AI algorithms. As players conduct investigations using an array of ghost hunting equipment, the AI-driven ghouls react to the human players’ inputs and behavior in unpredictable ways. It’s that kind of creepy programming that keeps the players constantly on edge.

Phasmophobia also uses voice recognition technology to enhance player immersion by allowing for real-time communication with the ghosts — no, they don’t talk back, but they can react to what the players say. This integration of AI-driven gameplay mechanics not only amplifies the horror atmosphere, but also fosters cooperative gameplay as players work together to uncover the mystery at hand.

Envisioning gaming’s AI frontier

Looking ahead, the future of AI in gaming holds boundless potential for innovation and evolution, with exciting developments on the horizon. As the technology continues to advance, expect game studios to develop bigger and more immersive experiences to delight their player bases. One promising avenue is the further enhancement of AI-driven procedural content generation, in which developers create variables that the AI uses to build out vast, ever-evolving worlds depending on player or multi-player interactions.

AI even may hold the key to unlocking new levels of personalization and player engagement in gaming. With advancements in machine learning and player modeling techniques, developers can create AI-driven experiences that tailor themselves to players and create game experiences  that are unique to the humans playing them.

Chris Mullen, contributor