The Game Mechanics: What is Artificial Intelligence?

This is the first part in a series on Artificial Intelligence and its impact on board game hobby. In this first article we will explore what AI is and build a common set of terms and language that will be used throughout the series as we dive deeper into it. We will also take a look at the long history of AI and board games. Board games and AI have always been closely linked weather it be in real life practical ways like Deep Blue took on grand-master of chess Kasparov in pop culture as in War Games where Mathew Broderick's character takes on the WOPR computer in a game of thermal nuclear war.

AI is a super broad term that covers a lot of different technologies. Its not just about T-1000 and Data from Star Trek. AI is a wide area of technology science including machine learning and deep learning. We will take a look at these aspects and highlight them with real world examples. AI is becoming part of board gaming and has been for a long time. We will examine how board games and AI are intertwined and how they affect each other in this series.

First up lets take a look at what machine learning is. Machine learning is a technology that most of are familiar with if we know it or not. Ever had to click on the images with a boat in them to prove you were human, ever use Alexa or Siri? If you have then you have actually been helping train a machine learning algorithm. So what does that mean? Well if you have built an algorithm to identify pictures of boats you need to train the algorithm on what a boat is. One way to do this is to tag a bunch of photos of boats as boats and let it figure it out from there. This is similar to the way that Siri learns your voice patterns and weather it got your request right or not. If we go back to the example of Kasparov vs Deep Blue (IBMs supercomputer) what they did to teach the algorithm was to input more than 4,000 positions and 700,000 grandmaster games. Just like a human who would have studied chess this much the computer became really good at chess and beat Kasparov.

Deep learning is the next term in AI that we will go over. Machine learning needs structured data deep learning can use unstructured data. The information fed into the algorithms in deep learning can be unstructured meaning the data doesn't have to be prepared in a format that is easily understood by the algorithm. Applications like ChatGPT, language translation services and programs that add color to black and white photos. The ancient game Go has always been a hard game to get computers to tackle, however with deep learning technology in place a team at Google was able to get AI to learn the game of Go without any human intervention as reported by the Smithsonian Magazine.

Getting AI to play games with us is just the tip of the iceberg. There are companies using AI to generate art for games while other companies like Wizards of the Coast and Paizo are not allowing AI generated art into their games and books. The technology is very interesting and has some real interesting components that will forever change the way we make and play games.