A team of researchers at the University of Brightloch have announced that 9×9 go has been solved. Inspired by recent improvements in computer play by Montecarlo algorithms, they hypothesized that “If a computer can play at random andplay good games, what prevents a more powerful device (like a brain) to do something similar?”
So the researchers picked a team of 100,000 chickens and taught them to play go. “It was hard in the beginning, but once a few knew how to play they started teaching others,” they report. “After a couple months all were playing as 30k players and we set them loose in a field filled with 9×9 go boards and bowls with stones. Also some grain and water.”
A month later the chickens were gone, and the same position was repeated on every 9×9 board: perfect play, with white winning by 0.5 points with 7.5 komi. The chickens had also left a note. It said “So long, and thanks for all the grain.”
- based on a report on Nordic Go Academy by RBERENGUEL; thanks to Go Game Guru, which posted it on their Facebook page.