IBM has been working on bringing AI to esports in the form of “shoutcasting”. They are trying to use AI to feed highlights for people to commentate on live. This is just one example of the many uses AI can be in esports. As someone with a coaching background, the one that interests me most is the use of AI to improve and teach players.
AI has been used in gaming for a while already, almost every computer was pre-loaded with a chess game with the computer able to replicate a top level chess player. We’ve come a long way since then and have seen AI being used to defeat players in DOTA. Just like someone playing two separate chess games, one against a person, and one against a computer, they can learn from the moves the computer makes and repeat them to the real person. This is the kind of thing I want to be able to do in any game. (except for practice, not for cheating).
Riot has show cased technology to be able to “rewind” a game to the state it was in a few minutes ago. They’ve used it when a bug affected the integrity of the game. Technology like that can be combined with AI. Two teams playing a practice game can each stop, rewind the game, and then let AI take over and play the same exact 2 minutes that you just rewound. Of course, that technology isn’t quite here yet, but we have the pieces.
Not only can AI improve the viewing experience of hard to track games like CSGO and battle royale games, AI can be used to implement, in real time, advanced game theory algorithms. The hard part that I see for AI in games like league of legends and shooters like CSGO is being able to turn human players into data points. In chess, humans and AI play the same, moving one piece at a time. In modern games, there’s so many different variables that previously, only a human brain can keep track of.
AI is going to change the world, how it affects esports is just one piece of this exciting pie.