The Evolution of AI In Gaming

​You might not realise it, but Artificial Intelligence has been used in popular gaming for almost half a century.


You might not realize it, but Artificial Intelligence has been used in popular gaming for almost half a century. Just like the evolutionary chain of man, you can see it develop in ability, intellect, and social skills over time.

Artificial Intelligence has been integrated with gameplay since its inception, developing over the years to incorporate more complex and challenging scenarios. It’s mainly used to imitate humanlike characteristics, enhance gameplay and improve user experience. Game AI is mainly associated with the development of behaviors in non-player characters (NPC’s) although not exclusively personified. It allows NPC pathfinding from A to B, traversing various terrain and navigation around the virtual world to appropriately avoid collision with enemies or find allies.

Arcade games are one of the earliest instances of commercial gaming using stored patterns to direct enemy movement. The later inclusion of microprocessors allowed for more random movement patterns. One of the first examples of this in popular gaming is in iconic fixed-shooter game Space Invaders (1978). Arcade games used early AI in the form of stored patterns, simulating random movements of the aliens that were actually pre-programmed into the game. This AI successfully kept players hooked and slotting coins into the machines!

Pac-Man (1980) also hit the arcades bringing AI to a complex maze based game. But, forget about the hungry yellow blob – it’s all about the ghosts! This innovative AI programming allowed these colorful creatures to exhibit different personality traits, deceiving players into thinking they’re playing against 4 individual entities.

Fast forward a couple of decades to when real-life simulation game The Sims (2000) let you assume the role of an all-powerful being from the comfort of your armchair. By using need-based formulas, different objects affected the character’s behaviors and relationships. Characteristics defined by the player at the beginning of the game impacted The Sims character’s choices. Pre-programmed variables in gameplay are also applied in sports games. Outcomes are based on variables like coaching style or abilities of managers and players.

Combat games have garnered much attention for AI development creating realistic opponents for users to play against. Finite State Machines (FSMs) are used in development to indicate the flow of action of AI in-game in games like the Tekken and Mortal Kombat Series. The FSM can be used to prompt enemy action in each single state – to heal or attack.

Building on the generic attack/defend mode, emphasis is put on how the NPC can hunt players through audible and physical disturbances to the environment, like footprints or a twig snapping underfoot. These complexities seen in stealth games like Metal Gear Solid 5 (2015) have allowed for a more immersive experience of open world gaming. More than hunters, NPC’s are survivors. The AI includes pre-set markers for them to improve health levels, reload ammo or seek cover during battle.

These techniques are important for improving gameplay as online multiplayer gaming dominates the market. Gamers want opponents with more impulsive and erratic behaviour, which they get playing online. Single-player games need to keep enhancing open world environments and non-player characters to rival human gameplay. AI is the way to do this.

Insanely successful GTA5 (2013) pushed open world to the next level with the most lifelike characters seen in gaming, set in a simulation of the real world - Los Santos. It’s come a long way from the bird's-eye view shots of the original release almost two decades earlier.

Characters go about their business throughout gameplay and interactions with them are incredibly realistic – a huge factor behind the controversy surrounding the game. Watch out Uber - these AI taxi drivers can get you from A to B with no input to pathfinding from the player. It’s not exactly David Attenborough but you can even watch a live feed of the modded game of an AI deer living life in the world of San Andreas. It’s not just AI characters though; the in-game website updates throughout the game and you’ll also notice characters having realistic conversations with backstories with other AI on phone calls.

Data Mining has helped developers automate difficulty levels to create more interactive and challenging worlds. Professor Arend Hintze recently created an AI that adapts the video game to the player’s behaviour to match their ability. The AI increases as the gamer engages in more play, feeding the system with more data.

Big Data in gaming is capturing information on how and when we game like never before. Tracking data points is allowing developers to tailor gaming experiences to individuals whilst increasing revenue from trends in spending habits. By incorporating such methods into gaming, expect to see more enhanced player experiences unlocked.

When thinking about the future of gaming, it’s hard not to mention the latest craze - augmented reality hit Pokémon Go. With over 30million active users in less than a month of being released, it’s popularity could signify what’s to come. If you can catch a Jigglypuff on your way to work now, just imagine defeating Dr.Neo Cortex in your office, building on your home Sim’s-style or living out a digital zombie apocalypse through your device. Virtual Reality is also set to play a part in the development of open world gaming, allowing full simulation of alternate realities to play in.

The future of gaming is a much-debated subject with a variety of people backing different corners. Whichever comes out on top, the gaming industry will continue to evolve rapidly like it always has, with AI and big data leading it into new unchartered territories.

Matt Reaney is the Founder of Big Cloud. Specialists in Big Data recruitment - connecting innovative organisations with the best talent in Data Analytics and Data Science.

Follow us on LinkedIn

Vision small

Read next:

Big Data Forecasting In Pharma