2022
On November 22 (local time), Meta announced the “AI that achieved human-level performance” in the strategy game “Diplomacy”, “CICERO” (supposedly named after the Roman politician Cicero). The company claims to have achieved more than twice the average score of human players in an online version of the game against humans, ranking it in the top 10%. https://www.itmedia.co.jp/news/articles/2211/23/news050.html
The inner mechanism is like a partially observed Markov decision process, updating policy assumptions based on observed data.
This page is auto-translated from [/nishio/CICERO by Meta](https://scrapbox.io/nishio/CICERO by Meta) using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.