DistantNews
Support us
🇦🇹 Austria /Technology

Why Is AI Better Than Humans at Chess?

From Die Presse · () German

Translated from German and summarized by DistantNews. Read the original for the full story.

At a glance

Explainer Named sources Context piece
  • Chess engine Stockfish has a rating of 3650, far above world number one Magnus Carlsen’s 2823.
  • Researcher Vito D. P. Servedio says Carlsen might draw some games against Stockfish but would probably not win a 100-game match.
  • The study suggests that AI manages tension and complexity differently from top human players, though it finds correlation rather than proving causation.

Stockfish wins at chess, and according to the article, it wins every time. Even Magnus Carlsen, currently the world’s highest-ranked player, would probably have no chance of winning a match against the engine.

Vito D. P. Servedio of Vienna’s Complexity Science Hub estimates that if Carlsen and Stockfish played 100 games, Carlsen might secure a few draws but would not win a match, at least according to statistical probability. About 1,000 chess players worldwide have ratings above 2500. Carlsen leads the human rankings with 2823 points. Stockfish stands at 3650.

If the two played each other 100 times, Carlsen might perhaps achieve a few draws, but, at least according to statistical probability, he would not win a match.

· Vito D. P. ServedioThe Complexity Science Hub researcher estimates the likely result of repeated games between Magnus Carlsen and Stockfish.

Servedio studied why artificial intelligence is so dominant on the 64-square board. His findings are due to appear the following week in the journal APS Open Science. One possible explanation, he says, is that AI handles complexity differently.

One of the reasons could be that AI handles complexity differently.

· Vito D. P. ServedioServedio offers a possible explanation for why Stockfish outperforms human chess players.

Human players build tension during a game. Pieces face one another and create threats, while players decide whether capturing an opposing piece will improve their position. Top players often build considerable tension early, then begin exchanging pieces and reduce it.

AI appears to follow another pattern. It does not build as much tension at the start, but keeps the existing level for longer or reduces it only slightly. Servedio compares the approaches with two armies. One accepts heavy losses while weakening the opponent, eventually reaching a smaller force on both sides before exploiting an error. The AI instead reduces both armies earlier and maintains the balance over a longer period.

While playing, both opponents build tension: pieces stand opposite one another and threaten each other.

· Vito D. P. ServedioHe describes how human players create and manage tension during a chess game.

That second approach seems more successful on the chessboard. Servedio cautions, however, that the research establishes a correlation, not a proven causal relationship. Human grandmasters also generally have no more than two strategies in mind, while AI can choose its next move rapidly.

Even world-class players usually have a maximum of two strategies in mind.

· Vito D. P. ServedioThe researcher contrasts the limited strategic framing of human players with AI’s rapid move selection.
About this summary

Originally published by Die Presse in German. Translated, summarized, and contextualized automatically by DistantNews, with a note on how the source frames the story. Not individually reviewed before publishing. How this works.

More Stories

From Our Blog