Researchers have built a new artificial intelligence chess engine that has been trained to play like a human

The first time a computer had beaten a human chess master occurred in 1997 when IBM Deep Blue defeated the champ, Garry Kasparov. Artificial intelligence systems have so advanced in their ability to play chess that no human has defeated a computer in a tournament in the last 15 years. A team of researchers from Tisch University has developed an artificial intelligence chess engine that does not seek to outrun humans.

Rather, this AI Chess game The engine tries to play like a human to create a more enjoyable chess playing experience for people. The researchers on the project note that it also highlights how computers make decisions differently from humans and how this can help humans learn to do better. One of the researchers on the project, John Kleinberg, says that chess is something that humans study throughout their lives until they become masters at it.

However, computers are better in every sense of the word than humans at playing chess at this point. Chess is a place where we can understand human skills through the lens of super-intelligent AI, says Kleinberg. The chess engine created during the research is known as Maia.

Maia was released on the free online chess server and participated in more than 40,000 matches in its first week. The algorithm used in the new chess engine mathematically determines the errors that are typical of a player’s level. It can pinpoint the mistakes people have to work on and the mistakes they shouldn’t work on because they are still difficult to understand.

One of the challenges facing AI systems today is that because they approach problems completely differently than human learning from them, this is difficult and potentially dangerous. This project aims to create an artificial intelligence that reduces the disparities between human and algorithmic behavior by training the computer using the effects of individual human steps rather than teaching the AI ​​itself to complete an entire task. Artificial intelligence has been trained in individual human chess moves rather than the general goal of winning a game, and it is primarily taught to imitate human behavior.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button