In a fascinating experiment, infrastructure engineer Robert Caruso pitted a vintage gaming console against cutting-edge AI technology in a chess match. The unlikely contenders were the Atari 2600, released in 1977, and OpenAI's ChatGPT. Surprisingly, the Atari 2600 emerged victorious, outperforming ChatGPT in a game of chess.
The showdown began when Caruso engaged in a conversation with ChatGPT about the history of chess. The AI boldly challenged the Atari to a match, citing "International Chess," a classic game developed for the Atari console. Caruso discovered that "Video Chess," the game in question, posed a challenge for beginners and was somewhat threatening to intermediate players, prompting him to stage this battle across eras.
The results were astonishing. Caruso reported that ChatGPT was "completely crushed" in the beginner-level match. Even with the chessboard layout provided, the AI confused rooks with bishops and frequently lost track of the pieces. Despite Caruso's attempts to guide ChatGPT using standard chess notation, the AI made numerous mistakes reminiscent of a novice player.
Throughout the 90-minute match, Caruso intervened repeatedly to help ChatGPT make more rational moves. Yet, the AI failed to improve its performance, even claiming it would do better if given a second chance. In the end, ChatGPT had to admit defeat, leading Caruso to quip, "Have you played Atari today?" to which the AI responded, "I wish I hadn't."
This human-AI chess match sparked discussions on retro technology and modern artificial intelligence. Caruso has also issued a challenge to readers who love retro devices, encouraging them to try their old gaming consoles against AI and share their results. The highlight of this event was the Atari 2600's victory over ChatGPT, with the AI making frequent mistakes and losing awareness of the chessboard.