AI is such a broad term these days, but in Bizio it directly refers to the decision-making logic implemented for the non-human players in a game. You first need to understand the mechanics of a turn before you can understand how the AI leverages these mechanics to make decisions like a real human.
As a player the system captures your actions as a “turn action”. These turn actions contain the relevant data required to replay what you wanted to do on the server. If you decide to make an offer of $1,500 a week to John Smith to work for you then the system will generate a turn action with that information. There are several different kinds of turn actions and these are the backbone of Bizio’s game play.
There are limits to how many of each kind of turn action you can perform in a given turn, and the system bases these limits on how much non-project-related skill effort your staff puts into them. For example, hiring someone requires your company to accrue some amount of accounting skill points. The more employees you have with the accounting skill that are not using for a project, the more capacity you will have to hire new employees.
As a human player you make decisions based on the complex landscape of your company, the market, and the projects available. As a non-human player these decisions can get pretty complicated pretty quickly. The turn action system simplifies and organizes this decision-making and the result is an AI that feels like another human.
Continuing with the hiring example, when you want to hire a new employee you consider many factors: what skills do I need, how much can I afford to pay them, etc. You are able to quickly synthesize the pieces of information necessary to begin your pursuit of your next employee. The AI also follows the path. Whenever the AI has capacity to perform a given turn action, like hiring an employee, it first considers whether it needs to perform that action. In this case it looks at how many employees it has, how many projects it’s working on, and other factors.
If it thinks hiring someone new makes sense then it begins scoping out the perfect* candidate. It will consider factors just like you would: cost, skills, age, etc. Once it has decided on who to hire the result is the same as a human player: the system creates a turn action with the necessary pieces of data.
We’ve designed the AI to act as closely to human as possible and turn actions are a big part of naturally bounding them to the same rules as a human player. We want games to include any number of AI players and have the them feel organic, competitive, and quirky.
* Perfect, to the AI, is incredibly subjective. Stay tuned for our next update which will help explain what “perfect” means to the AI by explaining the really rich personality system.