How to win board games like Diplomacy with AI

The day came when artificial intellect You can “play war”. At least diplomacy is about war and literally a game. Created by artificial intelligence purpose call Pika managed to defeat several players of the popular online game Diplomacywhich is a long historical board game.

Artificial intelligence created by the company Mark Zuckerberg It aims to mimic human decision making.

Chess or checkers are well known. However, this popular online game has a new variable based on it European political relations. Its action begins with a conversation between players.

The goal was to test natural language analysis using artificial intelligence and then make a binding decision. This is one of the biggest obstacles to this technology. Although it could reproduce human text, it could not fully understand how natural language works, much less reproduce it.

In case of game Diplomacy Adds the factor of interaction, influence on third parties and subsequent decision making. Opponents must talk to form bonds and alliances.

It remains to be seen whether they will be compromised or hacked, putting international diplomacy at risk, at least in-game. The goal of this board game is to survive the power of cities that are considered supply centers. Later, one must decide whether to attack or support the opponent each turn.

You may be interested in:

These were the results of Cicero

Cicero, the meta’s AI, has gained the ability to study whether his opponents are abiding by or breaking diplomatic agreements. All starting with the basics of the game: conversational negotiations.

Here, Cicero can negotiate tactical plans, reassure an ally, discuss the larger strategic dynamics of the game, or even talk about things a human player can discuss.

Cicero’s account entered webDiplomacy.net It was really devastating. AI was able to close 40 games against 82 men, Each lasts for two hours. Games where he always remained anonymous so as not to reveal his AI status. Finally, he got it Twice the average of other players. Much higher than average.

The AI ​​has learned from its opponents’ mistakes

Like all other AIs, Cicero began by learning from its closest opponents. Which means he started by repeating human mistakes: he lied a lot. And he was wrong too. Subsequently, he refined the technique until he achieved a balance between what to follow and what not to follow. as well as their ability to influence the decisions of third parties.

Before it went online, the research team responsible for its AI fed the system data from thousands of games Diplomacy. A total of 125,261 anonymous games, 40,000 conversations and 12 million messages.

For the meta, the purpose of this AI is the obvious focus Metaverse future and otherwise uncertain. Cicero creates a deeper social experience and interacts more naturally with people. In fact, they claim theirs website that Cicero is open sourced to allow third parties to rely on its AI technology for responsible use both within and outside of its ecosystem.

Continue reading: