As we started having our first meetings with video game companies, we quickly realized that everyone was very familiar with the term AI. But it was also clear that we were not talking about the same thing!
Decisive AI offers superior Artificial Intelligence to the video game industry, but since we started out, we’ve had to swiftly reposition our offering so that it was clear that we were not talking about the same old AI everyone is used to. Existing “AI” is typically a series of hard-coded patterns created by a human. At Decisive, when we say “AI”, it is describing using Reinforcement, Supervised, and Machine Learning to develop Intelligent Artificial Players (IAPs), which play games.
AI is what controls the Non-Player Characters (NPCs) in a game, whether these are part of the environment (for example, people walking down the street) or an individual character (someone you play with or against). Consequently, we realized that NPC is another term in which we needed to differentiate ourselves. What Decisive AI does is control the actions of what is (or should be) an intelligent player. Think boss battles, party members, or opponents in strategy games. Not the people walking down the street.
Since current NPCs are effectively controlled by a hard coded set of patterns written by very skilled developers, our Intelligent Artificial Players behave very differently when instead controlled by an heuristic function generated by AI through a true Machine Learning process. They also update and change, as opposed to the rigid behaviours programmed into games right now.
So why is it better to use AI to write the behaviour of an Intelligent Artificial Player?
The video game industry is well known for being early adopters of new technologies, whether software or hardware, and their need to be on the cutting edge is paramount.
We aim for our IAPs to be fair (no cheating), strong (a good player) and most importantly, adaptive. Not unbeatable and not useless, but a fun challenger to train/play with or against.
For the video game industry, there is a well known and direct correlation between the quality of a game’s AI, the player retention it produces, and the lifetime revenue it yields.
There is a lot of awareness of what is coming, as breakthroughs in the advancement of AI are announced more often than not intertwined with games (whether board or video) as the perfect field for its research and development.
Using AI to develop adaptive IAPs is the only viable path to personalizing the gaming experience, and that’s what it’s all about.