AI will teach humanity to play ancient games

Artificial intelligence can not only search for answers to the questions of hundreds of users, but also serve as an assistant in the revival of ancient traditions. In particular, what board games people played thousands of years ago.

New Scientist notes that artificial intelligence is helping to restore long-lost rules of games. They are not definitive, but probable, based on similar games. Named games include — Shahr -i Sokhta, also known as Ancient Burnt City Game. This board game was discovered in the 1970s in Tomb No. 731 at Shahr-e Sokhta, located in Iran. It is approximately 4600-4700 years old, and is considered the oldest complete board game ever found.

The game consists of 27 pieces and is played on a 20-square board. It is the earliest known example of a complete 20-square game. It stands out among over 100 types of ancient intellectual games found in the Middle East, played by different cultures rather than a single civilization. For the first time, the board has been reconstructed and ancient rules, which were probably played in that era, have been added very recently. The scientific paper was written by Sam Jelveh from the School of Computer Science and Electronic Engineering at the University of Essex and Dr. Hossein Moradi.

The game has some similarities to the Royal Game of Ur from ancient Mesopotamia, whose long-lost rules were deciphered in 2007 from cuneiform in the British Museum. To put the rules as simply as possible, players take turns rolling a set number of dice. One piece is moved according to the number. If it lands on an opponent's piece, that piece is captured and returned to the start. Some areas may be “safe”, where the piece cannot be captured. Other ancient “boards” are currently awaiting their rediscovery — the Roman game Ludus Latrunculorum and the Egyptian game Senet.