This engineer made ChatGPT play chess against a 1977 video game console and the results were embarrassing

ChatGPT

In an unusual face-off between old-school gaming and modern AI, ChatGPT recently lost a chess game to a video game console from 1970s — the Atari 2600.

The match was the brainchild of Robert Caruso, an engineer at Citrix, who wanted to see how ChatGPT would perform against Video Chess, a game released in 1979 for the Atari 2600. Using a software emulator to run the classic title, Caruso pitted OpenAI’s AI chatbot against the retro game — and the results were anything but flattering for the AI.

“ChatGPT confused rooks for bishops, missed basic pawn forks, and kept forgetting where pieces were,” Caruso said in a LinkedIn post. At first, the chatbot blamed the Atari game’s abstract icons, but when the pieces were switched to standard chess notations, it still struggled. Badly.

“It made enough blunders to get laughed out of a 3rd-grade chess club,” Caruso joked. “ChatGPT got absolutely wrecked at the beginner level.”

The 90-minute game didn’t improve with time. In fact, Caruso says the chatbot repeatedly asked to restart the match, clearly flustered by its poor performance.

While it might sound shocking that a powerful AI lost to a 46-year-old console, there’s a simple reason. ChatGPT is a language model — designed to understand and generate human-like text, not to process complex game logic or calculate strategic chess moves like a dedicated chess engine.

This isn’t the first time machines and chess have intersected. Back in 1997, IBM’s Deep Blue made history by defeating world champion Garry Kasparov. But unlike Deep Blue, ChatGPT wasn’t built for board dominance.

So, if you want to chat about chess strategies, ChatGPT might be your pal. But if you're planning to play a match, even a decades-old video game might just come out on top.

OpenAI hasn’t responded to the match just yet — maybe they’re still recovering from the loss.