Meta has shown the highlights of AI in the gaming industry

by Meghmala November 25, 2022


CICERO, the AI ​​that plays games for Meta, has human abilities to make alliances

The first artificial intelligence (AI) agent, CICERO, performed on a human level in the well-known strategy game Diplomacy. While CICERO can only be used to play diplomacy, the technology that underpins it applies to a host of other uses. Because it involves players understanding people’s motivations and views, formulating intricate plans and change strategies, and using language to persuade individuals to form alliances, diplomacy is seen as an almost insurmountable problem in AI.

In an online version of the well-known strategy game Diplomacy, in which seven players compete for supremacy over Europe by moving pieces on a map, Meta has developed an AI that can surpass humans. Players in Diplomacy must communicate with each other, form alliances, use negotiation strategies and recognize when others are bluffing, unlike other board games that AI has mastered, such as chess and Go. Compared to 82 human players playing 40 online games, the AI ​​Cicero finished in the top 10%. (Who didn’t know they were up against a bot). Cicero won an eight-match competition against 21 other contestants. Meta described her work in a publication that appeared in Science.

Meta combines two separate types of AI to create Cicero; one, a reinforcement learning model that decides what moves to make, and two, a huge language model that engages in conversation with other players. Based on the current state of the board and his previous conversations with the other players, Cicero predicts how the other players will behave at each stage of the game. It then determines how participants can work together for everyone’s benefit and creates messages to further those goals. It is important to learn how to play diplomacy for several reasons. Each step is preceded by a lengthy negotiation in which players converse in pairs to form alliances or take on competitors. Not only does this game involve multiple players performing actions at the same time, but it also involves multiple players. Players then choose which pieces to move and whether or not to fulfill a deal after this negotiation round.

Cicero was not flawless. Still, it occasionally released communications that conflicted with its plans or strategic errors. It’s also important because, unlike games like chess or go, which have winners and losers at the end, real problems often don’t have such clear-cut solutions. Often finding compromises and workarounds is more valuable than succeeding. According to Meta, people often prefer to work with it over other players. Cicero, according to Meta, is a step toward AI that can help with a variety of complex problems that require trade-offs, such as routing around heavy traffic and contract negotiations.

CICERO ushers in a new era for AI that can collaborate with humans while playing games using strategic reasoning and natural language processing, and the lessons learned from such technology could one day result in collaborative intelligent assistants.

Share this article

Do the sharing thing

About the author

More info about author

Leave a Comment