Learning

Chess Bots: The Evolution of AI in Chess Engines

Chess Bots: The Evolution of AI in Chess Engines

Preamble:

I have a fun mode right now. Let’s start today’s intro with a bot-alike approach from the following:

Chess is a two-player strategic board game played on a 64-square grid. Chess is considered to have evolved from chaturanga before the 7th century.Chess’s popularity throughout the global pandemic is heading to a revolutionary trend. There’s no better moment to study how AI can improve chess.

See? How does it look? “So-so”? That’s where Artificial Intelligence comes into play. AI’s involvement has brought the chess bots a revolution. Besides, AI has many real-world uses.

You got the vibe of our today’s article, right?

Let’s begin then!

History:

Alan Turing published a chess program in 1951. In subsequent years, improvements proceeded. This period saw new chess games and engines. AI chess systems had yet to attain higher-level success.

In the 1980s, Garry Kasparov claimed AI chess computers could never defeat grandmasters. In 1996, he successfully defended his throne against IBM’s Deep Blue in six games, 4:2. In 1989, he defeated IBM’s Deep Thought, Deep Blue’s forerunner.

Deep Blue beat the world title a year after losing. Kasparov lost to Deep Blue 2.5:3.5. Although there’s some debate about the engine’s win, it’s generally accepted. In a recent interview, Kasparov agreed that he was beaten fairly.

Since the 1950s, chess engines have improved dramatically. Let’s examine the recent evolution of chess engines to understand AI’s chess achievements better. Multiple Kasparov, being Kasparov, said IBM cheated. It’s been decades.

Today, Magnus Carlsen wouldn’t beat the world’s most outstanding chess software in 100 games. Carlsen’s current Elo rating is 2845, whereas Stockfish 9’s is 3438. (the engine ratings are not FIDE ratings, but the player pool for engines is much stronger than for humans, so theoretically, a FIDE rating for Stockfish 9 would be even higher).

In chess, comparing two players separated by 600 Elo points is like comparing Portland Pickles and Red Sox, Justin Bieber and Bob Dylan, or Windows Phone with any other phone. Who’s the boss? He’s many things. Not modest people. In the 1980s, he said no chess program could beat him. In 1989, he beat IBM’s Deep Thought twice.

In 1996, Kasparov overcame its successor Deep Blue 4:2 but was the first chess world champion to lose a tournament game to a computer. The following year came.

Kasparov lost to Deep Blue 2.5:3.5. Deep Blue’s spontaneous, inventive game stunned the world. Kasparov, being Kasparov, said IBM cheated. It’s been decades.

Modern Evolution:

Many different concepts are used in the Artificial Intelligence systems built for human play. While it’s difficult to say what each engine employs for its functionality and operation, we know that neural networks, deep learning, and neural net-type automation are used by a few popular engines like Alpha zero. Leela Chess Zero is based on an open-source version of AlphaZero, a chess learning program that employs self-play games and deep reinforcement learning.

image

Modern chess engines have advanced to the point where they will not lose a single game to human players. In 100 games, even the current defending world champion failed to beat the modern best chess engine once. Across all formats, the world champion has a FIDE rating of over 2800. The competition is typically held in traditional times. Stockfish 9 was the opponent.

It has a 3438 rating (the engine ratings are not FIDE ratings, but the engine’s player pool is much stronger than for humans, so theoretically, a FIDE rating for Stockfish 9 would be even higher). More information about this topic can be found in the reference section.

According to Wikipedia, the outcomes of two recent chess engine battles are as follows. In a 100-game competition, AlphaZero, a neural net-based digital automaton, defeats Stockfish 28–0 with 72 draws. In a 100-game match, Leela Chess Zero (LCZero v0.21.1-nT40.T8.610) defeats Stockfish 19050918 53.5 to 46.5 to win the TCEC season 15 title.

These modern findings, which have been achieved with the help of chess engines, neural networks, and deep learning-based chess networks, are a significant sign of the possibility for more and enormous opportunities. Let us delve a little deeper into the potential impact of artificial intelligence in the chess universe.

AI & It’s Influences:

Once supposed to benefit chess players, AI determines whether individuals perform better than expected. AI has come up with some cheat-detection techniques.

Metal detectors prevent board game cheating. Online cheating detection is difficult. Four thousand players from 55 federations are in the 2020 European Online Chess Championship. 5 of the 6 B-group (1400-1700) players were disqualified. Most disqualified players were young. EOC cheating.

DeepMind, the maker of MuZoo, is developing 100% accurate cheat detection algorithms. On Chess.com, a cheat detection system analyzes millions of games to identify the slightest chance a human will match or surpass an engine’s best decisions. Analyze cheating reports. “Month in Review” summarizes monthly results. Most games were rapid chess. Therefore technology may help longer games. FIDE Online Arena cheat-detection technique. Many internet sweepstakes were accurate during the pandemic.

image

Online gaming fraud detection is controversial. Cheaters cheat. Players lose. After multiple online games, these systems distinguish engine-aided plays.

A human expert would rate a perfect chess machine’s player. AI-based cheat detection program finds chess engine users faster and more accurately than a human referee. Professionals review all fraud cases.

Algorithms identify unexpected moves. The software analyzes cheater chess positions (ideally hundreds, but a handful would do). This unlikely move raises cheating flags. Games improve accuracy. Top-player moves can detect cheating. Top players never risk a tournament. Best players rarely cheat because chess enhances fairness. Non-chess players are entering online competitions due to online chess’s popularity. Some use chess engines.

Conclusions:

Artificial Intelligence is a revolutionary phenomenon that has undoubtedly altered the landscape of chess, unlike many other platforms that have been fortunate enough to be adopted by AI. Artificial Intelligence’s impact on chess can be debated from both sides, with both sides having legitimate points.

However, the influence, enthusiasm, and excitement generated by AI are undeniable. Artificial Intelligence has a lot of potential in chess in the future. Modern chess engine discoveries and advancements are continually developing. In the contemporary realm of chess, we can achieve far more incredible feats of achievement.

I’m looking forward to seeing where these developments take us. I also hope you will all try to make your chessboard or chess engine a fun project!