AI’s Struggles with Chess: A Surprising Revelation
It appears that the fears surrounding AI-driven robots engaging in warfare may be premature, as recent findings reveal that certain AI models, particularly ChatGPT, struggle significantly with chess. This information surfaced from a post by software engineer Robert Caruso on LinkedIn, where he disclosed that the well-known AI chatbot experienced severe defeats while competing against an Atari 2600 running Atari Chess.
A Look Back at Atari Chess
Released in 1979 for the Atari 2600 console, “Atari Chess” was notable for its time as a groundbreaking 4k game, especially when many games were limited to just 2k. It featured eight difficulty levels, allowing the computer to take its time with each move. However, in comparison to modern chess engines like Stockfish, which can analyze millions of potential moves in mere seconds, Atari Chess appears exceedingly basic. The game lacked a strategic opening move approach and was notorious for making odd moves, including nonsensical “king walks” and unnecessary sacrifices of pieces.
ChatGPT’s Disappointing Performance
Caruso noted that ChatGPT was unable to perform even at a novice level, stating, “ChatGPT got absolutely wrecked on the beginner level.” The AI’s struggle came after a discussion about AI’s historical involvement in chess, during which it eagerly volunteered to challenge Atari Chess, curious to see how quickly it could win against a game that only considers 1-2 moves ahead. Over the course of 90 minutes and multiple attempts, the AI had difficulty keeping track of the pieces and their positions on the chessboard, even misinterpreting the moves of its antiquated opponent. Frustrated, it repeatedly requested to restart the game, believing a fresh start might yield better results. Nevertheless, Atari’s 8-bit system simply continued executing its programmed moves without complication.
The Contrast Between Knowledge and Skill
Although ChatGPT possesses an understanding of chess, it fails to generate effective strategies for gameplay. In contrast, the Atari 2600, dedicated solely to chess, proves to be a more effective player. This situation underscores the notion that having knowledge does not necessarily equate to possessing the ability to execute that knowledge effectively. In response to the AI’s defeat, DeepSeek, a competing AI from China, remarked that defeating the Atari 2600 in chess would be a trivial challenge, emphasizing that its capabilities are minuscule compared to advancements made in computer chess over the last four decades. DeepSeek likened the challenge to a race between a modern Formula 1 car and a classic Model T Ford, highlighting the vast difference in performance.
Celebrating Military Service and Community
We Are The Mighty serves as a platform honoring military service while aiming to entertain, inform, and inspire both active service members and their supporters. Our community includes current military personnel, veterans, families, and civilians seeking connection. Stay updated on the best in military culture and entertainment by subscribing to our newsletter.
