A thirty-year-old computer outperforms AI on the chessboard

Garry Kasparov lowers his gaze, places his hands on his head, and suffers. It's 1997, New York, and for the first time in his career he's lost a match. The best grandmasters in the world weren't enough; it took a computer: Deep Blue . In truth, perhaps few people know this, Kasparov only lost the rematch; he had won the first match the year before in Philadelphia. But it doesn't matter; in chess, the result counts, and the result was historic.

Computers have been playing chess for decades, but they were simply a pastime for beginners, experiments for IT technicians, and few of the greats (one of them is former world champion Mikhail Botvinnik) took an interest. Then, within a few years of the new millennium, they became unrivaled, almost flawless. Faced with this victorious epic (those who are curious can delve deeper into the rise of one of the most recent and innovative software, Alpha Zero) , we could today be witnessing the writing of a new chapter in this history, one that would bear the name "Kaggle Game Arena," the tournament in which, for the first time, not dedicated chess engines, but the leading artificial intelligences participated: Open AI, Grok, Gemini, DeepSeek...

These twenty-first-century technological marvels, which have already replaced humans in many areas, have not, however, replaced them when it comes to chess. Every self-respecting chess player watching their games cannot help but feel profound embarrassment, laughter, or even pity for these "poor creatures" who not only fail to reach the level of traditional computers, but even struggle to play an entire game without committing illegal moves. Even Hikaru Nakamura, the current world number two who commentates live on the games, often fails to make sense of their moves (which just as often make no sense).

The technical reasons for this scarcity aren't overly complex, and can be found in the probabilistic nature of AI models . However, some argue that we may be faced with immature technologies that, just like traditional computers, will quickly abandon their grotesque game. Who knows? What is certain is that a human primacy in chess still remains: not skill, but natural stupidity, which is much more interesting than artificial intelligence.

More on these topics: