27 November 2021
It is nearly impossible for anyone even a chess grandmaster to beat Alphazero in chess. What’s interesting is that Stockfish, the strongest chess engine, was defeated by this system.
In short, what humans can only think of as certain “moves” ahead, Alphazero looks a bit further by using different planning and optimizing methods. Beating Alphazero by any human is not possible as of now. Hence, a human can’t beat Alphazero or even a chess grandmaster.
Table of Contents
Alphazero: A brief intro
It was in late 2017 that Alphazero, a single system that can master chess within a few hours, came into existence. Alphazero has managed to bag the award for beating some of the world’s best human minds in the game of chess.
The system figures out what the next set of moves will be and behaves accordingly. Alphazero’s games are ground-breaking and highly dynamic. Moreover, the system utilizes an unconventional style of playing that is different from other chess-playing engines that evolved before.
What is the difference between Alphazero and other competitors?
The main difference between Alphazero and other AI-powered software is that the machine-learning abilities of the former are at par compared to the general concept of playing chess. Apart from the general rules, it works by playing itself repeatedly and acquiring reinforced knowledge.
As a result, Alphazero can overtake the human-like approach needed to look out for moves. What’s more, the system can process about 80,000 positions within a second compared to the famous Stockfish 8’s 70m. Alphazero won several games of chess when played against Stockfish 8.
Alphazero won 4 games, one starting with white and three starting with black. Furthermore, 72 games resulted in a draw. In addition, Alphazero learned shogi in about 2 hours and defeated another program Elmo in 100 matches.
Alphazero won 90 games, lost 8, and managed to draw 2. So, those were the staggering records that Alphazero holds. Moreover, the new Alphazero has won against its former version Alphago in the Chinese game of Go. It received 8 hours of self-training before achieving the feat of winning sixty games and losing forty games.
Many experts believe that the results were impressive and Alphazero has the potential to beat a variety of applications. That way, it can complement human knowledge to a great extent. However, some computer scientists and researchers do believe that these tasks are discreet and may take some time to claim near-human abilities in terms of thinking.
Famous chess players who tried to defeat Alphazero:
Sadly, no chess player has ever played and defeated Alphazero. However, the only human to defeat Alphago, the predecessor of Alphazero was Lee Se-dol. While it was a 5-match showdown, he lost four times but managed to beat the computer once.
So, the question remains whether or not any humans can beat Alphazero in the future. To be precise, no human or even a chess grandmaster can beat Alphazero. Moreover, the world champion in chess, Magnus Carlsen failed to defeat Alphazero.
Starting playing the game randomly with general knowledge about the rules, Alphazero achieved a superhuman level of intelligence while playing chess, shogi, and Go as well. Also, it defeated convincingly the world champions in each case.
Even the chess grandmaster Gary Kasparov who was defeated by a computer program several years ago realized that artificial intelligence would only grow stronger in the future. Viswanathan Anand, the former world champion from India, found Alphazero to be an amazing system that can learn and become a master in just four hours.
Also, he was impressed that the system learned everything from scratch. For thousands of years, humans crafted the game of chess only to be taken over by computers. With the increasing power of these software programs, their intelligence took a drastic turn.
In May 1997, IBM’s Deep Blue defeated Gary Kasparov. And, from then onwards, it has become nearly impossible for a human to defeat computers in a game of chess. While all the chess players worldwide have reacted in their way, most of them have praised its ability to become masters from scratch and without help from humans.
Can any human beat Alphazero in the future?
While it can only be anticipated that the number of wins against artificial intelligence programs goes with the humans, the future is uncertain. There is no denying that humans have certain limitations and these are bound to remain almost the same in a few decades or so.
For example, humans are likely to evolve much in the next 10 years. But Alphazero will get better and better in the next ten years. After all, it took about 4 hours to learn the different techniques used in a game of chess. Apart from that, the system developed a technique to teach itself.
As a result, it can play chess at an international level without trouble. Moreover, the most intelligent man on Earth doesn’t even try to beat Alphazero. Hence, there is no question that humans can defeat these super-intelligent computers anywhere in the future even at their strongest intelligence level.
Is it good to compare results from the humans and Alphazero?
Comparing humans with artificial intelligence-driven computers is not the same. It is like asking Usain Bolt to beat a car in a race. Computers have lots of advantages over humans. They never get exhausted, lose concentration, get scared, or question themselves.
Now, these are crucial when it comes to playing chess. Apart from that, computers can have access to libraries while playing a game and humans cannot. Rather, humans use memories to get an upper hand over their competitors. While computers seldom make inaccurate predictions, humans make blunders often.
You may have seen many chess players losing the game even with a knight and bishop in place. Alphazero, on the other hand, won’t make these mistakes. The tactical ability of computers doesn’t work like humans and it is quite astounding.
Machine learning abilities combined with artificial intelligence will only get stronger year after year. Before Alphazero it was Stockfish. And, there will be a lot in the future as well.