US Edition
ChatGPT Defeated at Chess by 1970s-Era Atari 2600
- Citrix engineer Robert Caruso conducted an experiment over the weekend where OpenAI's ChatGPT played chess against the Atari 2600's chess engine emulated by Stella.
- After discussing the history of chess and artificial intelligence with ChatGPT, Caruso initiated an experiment to test a 1979 chess game running on an eight-bit console from 1977 against the AI.
- During the 90-minute match, ChatGPT repeatedly confused pieces, lost track of the board, blamed abstract Atari icons, and requested multiple restarts before conceding defeat.
- Caruso reported that for a full hour and a half, he frequently had to intervene to prevent ChatGPT from making poor moves and repeatedly adjust its understanding of the chessboard throughout each turn, highlighting the AI's difficulties despite advanced hardware backing.
- The outcome highlights that ChatGPT, a general language model, lacks reasoning for games like chess, while a simple 1.19 MHz, 128-byte RAM Atari engine outperformed it in this task.
Insights by Ground AI
Does this summary seem wrong?
26 Articles
26 Articles
All
Left
1
Center
6
Right
1
An engineer lets ChatGPT compete against the chess engine of an Atari 2600 from 1977. The language model delivers desolate performance and loses a lot. The attempt shows the limits of today's omnipresent AI chatbots.
Good old Atari, he can still do it: In an experiment, the old computer system defeated a new AI. ChatGPT was probably looking for excuses.
·Germany
Read Full Article
One user has faced the mythical video game console that emerged in the late 1970s with the OpenAI AI. No one could have guessed the result.
·Madrid, Spain
Read Full Article
Coverage Details
Total News Sources26
Leaning Left1Leaning Right1Center6Last UpdatedBias Distribution75% Center
Bias Distribution
- 75% of the sources are Center
75% Center
13%
C 75%
13%
Factuality
To view factuality data please Upgrade to Premium
