Artificial Intelligence: CICERO, An AI That Plays The Diplomacy Game “At A Human Level”

November 23, 2022 | Artificial Intelligence, News

CICERO is claimed to be an AI that can engage and compete with people in gameplay using strategic reasoning and natural language.

Meta (formerly Facebook) (NASDAQ: META) announced on Wednesday the launch of CICERO, the first artificial intelligence (AI) agent to achieve human-level performance in the popular strategy game Diplomacy. So far, Diplomacy was a huge challenge to AI given that the game needs human like capabilities such as persuasive conversation, complex planning, strategies and their adjustment and an understanding of other players’ perspectives and motivations. (Meta)

Diplomacy: A game about people rather than pieces

CICERO demonstrated its abilities by playing on, an online version of the game, where it achieved more than double the average score of the human players and ranked in the top 10 percent of participants who played more than one game.

This shows CICERO can negotiate, persuade, and cooperate with people.

CICERO is made up of a controllable dialogue model for Diplomacy coupled with a strategic reasoning engine. At each point in the game, CICERO looks at the game board and its conversation history, and models how the other players are likely to act. It then uses this plan to control a language model that can generate free-form dialogue, informing other players of its plans and proposing reasonable actions for the other players that coordinate well with them.

According to Andrew Goff, Diplomacy world champion, CICERO “plays the situation as it sees it.”

“So it’s ruthless in executing to its strategy, but it’s not ruthless in a way that annoys or frustates other players.”

Open code

Meta has open-sourced the code and published a paper to help the wider AI community use CICERO to spur further progress in human-AI cooperation.

“Open science and open source are crucial for the health of the AI research ecosystem,” said Percy Liang, Director, Center for Research on Foundation Models at Stanford University’s Institute For Human-Centered AI. “It is great to see Meta AI continuing its culture of openness with this release, which will allow researchers to build on top of this work.”

Related Story: Robot Playing Chess In A Russian Tournament Fractures Child’s Finger

Free Industry News

Subscribe to our free newsletter for updates and news about alternatives investments.

  • This field is for validation purposes and should be left unchanged.


Latest Alternative Investment News
Digital Assets: Mike Novogratz Hits Out At Sam Bankman-Fried, Calls For His Prosecution
December 2, 2022     Digital Assets, News

Mike Novogratz, the billionaire CEO of Galaxy Digital (TSE: GLXY), was unequivocally critical about Sam Bankman-Fried’s virtual New York Times (NYT) Dealbook Summit interview. “I’m hoping that the authorities get…
FinTech: Monzo To Turn Profitable In 2023, Says CEO
December 2, 2022     FinTech, News

“We will be profitable in the next financial year,” TS Anil said at the summit on Thursday, though Monzo reported pre-tax losses of £119mn for the 12 months to the…
Artificial Intelligence: OpenAI Launches ChatGPT, A New Conversational AI System
December 2, 2022     Artificial Intelligence, News

OpenAI has unveiled ChatGPT, a new AI system that interacts in a conversational way. As part of a dialogue format, ChatGPT answers followup questions, admits its mistakes, challenges incorrect premises,…
Artificial Intelligence: Monarch’s Electric, Driver-Optional, Smart Tractor Now In Commercial Production

Monarch Tractor has rolled out the first production model of the MK-V tractor that combines electrification, automation, and data analysis to help farmers reduce their carbon footprint, improve field safety,…