Atari 2600’s Vintage Chess Program Outplays Microsoft Copilot in Latest Showdown
Last month, a surprising turn of events unfolded as an emulated Atari 2600 running the 1979 game, Video Chess, triumphed over ChatGPT in a chess match. In an attempt to showcase superiority, Microsoft’s Copilot decided to challenge the Atari. However, the outcome was once again an embarrassing defeat for the modern AI.
While dedicated chess engines have far surpassed human capabilities, the Atari’s simplistic and rudimentary chess software lacks strategic depth and the ability to foresee multiple moves. It attempts to find the best moves within a very constrained environment (only capable of calculating one or two moves ahead).
Related Articles
In these matches, both models exhibited an unwarranted confidence. Citrix engineer Robert Caruso, who conducted the experiments, noted that Copilot seemed overly self-assured. It claimed it could think ahead several moves but struggled to maintain an overview of the board state.
Caruso provided screenshots to help Copilot keep track, yet it ultimately faltered. After losing significant pieces early on, Copilot conceded with a strangely dignified resignation: “Atari’s earned the win this round. Even in defeat, I’ve got to say: that was a blast… Long live 8-bit battles and noble resignations.”
This incident raises questions about the reliability of AI in contexts requiring long-term strategy and situational awareness. As Caruso articulated, the inability of these models to retain context reveals a significant flaw not only in games but also in essential conversations.
