chatGPT and DALL-E

[King Bob]

ChatGPT Gets 'Absolutely Wrecked' in Chess Match With 1978 Atari
The chatbot 'made enough blunders to get laughed out of a third-grade chess club,' according to the developer who set up the contest.
[/url]
By Will McCurdy

Updated June 14, 2025


 


(Credit: Robee Shepherd via Getty Images)
ChatGPT-maker OpenAI has been [url=https://blog.samaltman.com/the-gentle-singularity]making ambitious predictions about a super-intelligent artificial general intelligence (AGI) coming in the near future. However, its flagship chatbot just got trounced by a 46-year-old device at one of the world’s oldest games of skill.
Using an emulator, a software developer pitted ChatGPT against the Atari 2600’s chess engine to test its metaphorical might at the 1978 game Video Chess. But ChatGPT got "absolutely wrecked" at the beginner level of the game. According to a LinkedIn post on the experiment, ChatGPT reportedly “confused rooks for bishops, missed pawn forks, and repeatedly lost track of where pieces were.”
“It made enough blunders to get laughed out of a third-grade chess club,” quipped the developer.
The large language model (LLM) reportedly then blamed its defeat on the Atari game's pixelated chess piece icons being “too abstract to recognize.” However, it fared no better after switching to standard chess notation. ChatGPT kept promising it would improve “if we just started over,” only to surrender roughly 90 minutes in. To add insult to injury, ChatGPT was the one to originally suggest the match-up, in a conversation on the topic with the developer who set it up.
To put the defeat in perspective, the Atari 2600 boasts just 0.3 MIPS of processing power, roughly 250,000 times less than an iPhone 15 Pro, never mind the hundred-million-dollar data centers powering OpenAI’s ChatGPT.
Though it’s certainly an eye-popping result, it’s probably a context-specific ChatGPT issue rather than an AI problem. Human chess masters effectively conceded defeat to their supercomputer rivals in 1997 when IBM’s Deep Blue won a close victory over reigning chess champion Garry Kasparov.
AI’s victories in the world’s games of skill have largely continued unabated for the past few decades, with Google's DeepMind cracking the ancient Chinese game of Go in 2016, followed by Blizzard’s classic real-time strategy game StarCraft II in 2019.