AI & The Future of Gaming
Artificial intelligence and the video gaming sector have an intertwined history that goes as far back as 1951, in the earliest forms of computer based chess and later on seeing use in games like Tetris and Space Invaders.
Overtime, the video gaming industry has gone through many advancements since then. They now deliver a far more immersive and realistic gameplay experience than their forefathers did. A good part of this can be attributed to advancements in the hardware that is used to run them, and then the other tools used in game development. As for what goes on under the hood of how Non-playable Characters (NPC or bots), not much has changed.
One could think of the inner workings of a video game’s pseudo mind that operates in an environment governed by a strict set of rules, which the AI follows as a guide on how to respond to external stimulus (the players actions).
If you have had the chance to play one of those earliest forms of video games, you might remember how hard or downright impossible it is to win against those AI’s at the higher levels of the game. The godlike abilities of these AI players is the result of the closed nature of the environment in which they operate, coupled with the fact that they have perfect sets of information to work off of.
Advancements in the gaming industry have somewhat changed this over time. However this still isn’t the advanced “machine learning” kind of AI that conjures up fond memories of the most enthralling sci-fi flicks to have come to our screens.
What looks like realism to the end consumer is actually just very good mimicking of realism. With the help of newer development tools, game developers have become better and better at adapting the very same techniques used in old games to create an illusion of intelligence. This is according to Mike Cook, a research fellow at Queen Mary University of London’s Royal Academy of Engineering Research.
“Game developers got really good at being efficient with technology. They realised that they couldn’t create perfectly intelligent creatures. They’ve realised that they couldn’t solve all of these problems. So they figured out how to juice what they could do, they’d get the maximum out of it.” said Cook.
The kind of examples of artificial intelligence that have been used to dominate human champions at their own games, like the one that beat chess legend Garry Kasparov, or the Google/DeepMind’s AlphaGo bot that beat all the world’s best at a 2000 year old Chinese board game called Go, have not seen much in any other kind of modern games like Call of Duty.
There is a very good reason that developers prefer the smoke and mirrors approach over the use of machine learning-trained bots. This is because the latter could entirely ruin the player’s gaming experience by doing something so unpredictable that telling the game’s story becomes impossible. This would defeat the purpose.
However this does not mean that the advanced forms of AI are completely useless for game developers. In time to come, AI could be very helpful to developers who build open world games. Cook’s visions of the future of AI in gaming as a scenario where the AI assists the developers through the process of creating a game. Something that works like predictive text.
“I think you’re going to see tools that allow you to sit down and just make a game almost without thinking, as you work, the system is recommending stuff to you. This doesn’t matter whether you’re an expert game designer or a novice. It will be suggesting rules that you can change, or levels that you can design.” Cook said
What is even more interesting is the development of AI in gaming could bring about an era of games that are designed to give each individual a unique experience that is unpredictable and continually changing. An ever changing environment that never gets old, automated game design would add that “first time playing a game experience” over and over again.